The Tasmanian Devil Biology, Facial Tumour Disease and Conservation

Allen Greer Visit: sugargum.wix.com/tasmaniandevil Edition: 3 May 2016

“At the present time the future of the Tasmanian Devil looks good .... The only danger would seem to be a return of the disease which has devastated the population several times in the past” (Guiler, 1992)

“Wholly protected, common and secure, at present” (Mooney, 1992)

“Barring further epidemics … its status appears to be secure” (Jones, 1995)

“The devil is not threatened…” (Jones and Rose, 1996)

Introduction

For many years, the public attitude towards most Australian animals was one of indifference at best and hostility at worst. In the past, fish were mainly “good to eat;” frogs “slimy;” reptiles weird, sinister or dangerous; small and some medium-sized birds attractive, “useful” or “good to eat,” but many medium-sized to large birds pests or vermin, as were most mammals. And while many of these attitudes are still prevalent, there is an increasing appreciation of animals either for what they “do for us” or, even more encouragingly, just as interesting creatures in their own right; indeed, beings with which we share both an evolutionary past and an uncertain future, and without which we would be truly alone in the universe.

Just how much attitudes have changed can be judged by contrasting attitudes in the 1950s and 1960s with those of today with regard to animals such as sharks, kangaroos, whales and dingoes. No species has benefited from this makeover as much as the Tasmanian Devil, a small dog-sized carnivorous mammal now living only on the island state of Tasmania. The Devil was almost universally loathed up until the 1960s but is now one of the few species for which the overused word “iconic’ is truly appropriate, not only in its home state of Tasmania but also in Australia generally (Darby 2003a). In May 2015, it was named Tasmania’s official animal emblem (ABC News, 30 May 2015; Tasmanian Government media release, 31 May 2015). But with the timing of a romantic tragedy, humankind’s recent reconciliation with the Devil has been tempered by a severe population decline with the threat of extinction, due to a rare transmissible tumour, Devil Facial Tumour Disease, sweeping through the population.

As a result of the renewed interest in the Devil both in its own right and because of its disfiguring and lethal tumour, research on the animal’s biology has proliferated and a multi-faceted conservation effort put in place. But to date, no one has tried to bring all this information together and provide a concise overview of all aspects of the species’ biology, its strange disease and current conservation. Such an overview might be especially interesting to those who take an interest in or have responsibility for the

6/05/2016 8:41 AM 2

Devil but lack the background or the time to access and assimilate the relevant scientific and associated literature. The primary purpose of this book is to provide such an overview.

The high public interest in the Devil’s plight has also meant that people working on the Devils biology, disease and conservation have provided a vast amount of information to the public. This on-going flow of information provides an unusual opportunity to see how science operates on a near daily basis. A second purpose of this book, therefore, is to look at the sometimes “all too human” aspect of science on display in the effort to save the Devil. In this regard, the last two sections of this book (Devil Facial Tumour Disease and Conservation) are a small contribution to the history of science in Australia and can be read profitably along with essays on the Wollemi Pine (Greer, 2007) and cloning the Thylacine (Greer, 2009 ).

How the Devil Got its Names

The only written record of how the Devil got its common English name comes from Gould (1863) who said that its “black colouring and unsightly appearance” led to the name Devil or Native Devil. Indeed, early settlers with a Christian background may have found many of the Devil’s features suggestive of an alliance with the Devil. Devils are largely black and mostly nocturnal and when annoyed, they scream and their ears turn red. Furthermore, what more appropriate place to find the Devil’s home than the “other side of the world?”

At the time of first European settlement in Tasmania, the Devil’s common name was “Native Devil” (Harris, 1808; Waterhouse, 1846; Gunn, 1852). At that time, it was unknown if the species occurred on the mainland as well as Tasmania. But when it became clear the Devil did not occur on the mainland, the common name “Tasmanian Devil” came into usage. In the mid-1850s, the Devil was also occasionally referred to as the “Tasmanian Land Devil (Howitt, 1855). As a logical corollary to the species’ common name, the Devils’ young became known as “imps.” Nowadays, they are also called “joeys” and “pups.”

Although the common name “Devil” stuck, a mammalogist working in the British Museum was scathing about the propensity of British colonists to give inappropriate names to local animals. “The Ursine Dasyure [his preferred common name] is called by the colonists “The Devil,” or “Native Devil;” I see no necessity for adopting such barbarous names, nor do I think it desirable to adopt other names given by our colonists, when they convey an erroneous impression of the nature of the animals which have received them (Waterhouse, 1846).”

A number of scientific names have been applied to the Devil. The species was first formally described and named for science in 1808 by George Harris, the Deputy Surveyor General for the new colony of Van Diemen’s Land. He named his new species Didelphis ursina, as “many of their actions, as well as their gait, strikingly resembled those of the bear (Harris, 1808).” From his description, Harris was obviously impressed with the species’ large mouth, diverse array of teeth, and array of muzzle bristles. He noted correctly that males are larger than females and have a pendulous scrotum carried in a pouch, and that females have a pouch and give birth to small, naked and blind young which adhere firmly to the teats with their mouths.

6/05/2016 8:41 AM 3

Fig. 1. The first European image of the Tasmanian Devil (in Harris, 1808)

He also noted that the species burrows; preys on small vertebrates; uses the front feet to convey food to the mouth, and is querulous, with females dominant.

There are only two errors in Harris’ observations. He miscounted the number of premolars and molars (“grinders”) as four and five in each the upper and lower jaw, respectively, instead of six and six (below). He also observed that the tail was “slightly prehensile” (repeated in Geoffroy Saint-Hilaire, 1810 and Boitard, 1841: “prenant”) for which there have been no further observations, although there is a photo of a young animal with its tail curled around a fern stem (Sanderson et al., 1979: fig 1).

And in an observation that would prove to be important in determining post-European changes in population size, he noted that the first settlers of Hobart Town found the species to be “very common.”

Interestingly, the specimen(s) that Harris used to describe his new species are now lost. He notes that the specimen he illustrated was to be sent to Sir Joseph Banks in England, but it “died before the vessel left New South Wales (Harris, 1808).”

After reading Harris’ paper, the French zoologist Éntienne Geoffroy Saint-Hilaire realised that the Devil was not particularly closely related to American marsupials, as was implied by the generic name Didelphis, but instead was most closely allied to other Australian marsupials. Consequently he placed Harris’ species in the existing Australian genus Dasyurus. Thus, it was realised relatively quickly after the Devil became known scientifically that its relationships lay with Dasyurus. This generic name was also used by the Dutch zoologist Coenraad Jacob Temmnick in 1827 and the Tasmanian botanist Ronald Campbell Gunn in 1838 and again in 1852, and somewhat oddly, even by the Victorian comparative anatomist William Colin Mackenzie as late as 1921.

In 1837, however, the French zoologist Frédéric Cuvier thought the Devil differed enough from Dasyurus to warrant its own generic name and proposed Sarcophilus. This name is derived from two Greek words, one for “lover” (philus) and the other for

6/05/2016 8:41 AM 4

“meat” (sarco). Cuvier had no first hand knowledge of the species’ habits but interpreted them from Harris’ original description and its dentition. Hence, the Devil’s scientific name became Sarcophilus ursinus.

A few years later, in 1841, John Edward Gray (1841) of the British Museum, in what was perhaps a bit of cross channel rivalry, proposed, without explanation, a new generic name for the species, Diabolicus. Under current rules of zoological nomenclature, however, whimsical name changes such as this have no validity.

In this same year, 1841, the French naturalist Pierre Boitard proposed, again without explanation, another completely new name, Ursinus harrisii. The new generic name was perhaps Boitard’s own attempt to impose his nomenclatural will, but it was doomed to fail for the same reasons Gray’s name did. But the new species name was perhaps proposed in recognition that this part of the Devil’s name was “unavailable,” due to its application eight years earlier to the Common Wombat, Didelphis ursina (now Vombatus ursinus). Parenthetically, there is some justice in this, as the wombat is surely more “bear-like” (ursine) than the Devil. Boitard’s new species name harrisii graciously recognised the original describer of the species.

Much later, the British zoologist Oldfield Thomas thought he was the first to realise that the second part of the name Sarcophilus ursinus was unavailable for the Devil and proposed his own substitute name, Sacrophilus satanicus (Thomas, 1903). Nine years later, however, he realised his mistake and acknowledged Boitard’s much earlier substitute name (Thomas, 1912).

Finally, in 1987 the Swedish palaeontologist Lars Werdelin, while working briefly in Australia, proposed a subspecies name for the recently extinct Devils in western Victoria, dixonae, after Joan Dixon, the then curator of mammals at the then National Museum of Victoria (now the Museum of Victoria). He gave some dental measurements to justify the taxonomic recognition of this population, but he did not compare them statistically with the Tasmanian population. This name has never been used in any context.

There is one other twist in the history of the Devil’s scientific name, but it is discussed under the section on “Distribution.”

Phylogenetic Relationships

All living mammals are members of a single lineage that evolved sometime in the early Mesozoic. This lineage split into the monotremes or protherians (egg laying mammals) and the therians (live-bearing mammals), and the therians then split into the metatherians (marsupials) and eutherians (placentals). The dates for the origin of mammals and the three branching points of its three living constituent groups are uncertain. The earliest for each of the three groups all date from about 125 Ma. But inferences from molecular genetics suggest that mammals evolved about 325 Ma (Shedlock and Edwards, 2009): the protherian – therian spilt occurred about 220 Ma and the metatherian – eutherian split about 175 Ma (Madsen, 2009; see also Bininda- Edmonds et al., 2007). Australia is the only continent with all three major groups of living mammals.

6/05/2016 8:41 AM 5

The marsupials appear to have first evolved in the great northern landmass, Eurasia and then extended into the great southern landmass, Gondwana. They went extinct in the North American part of Eurasia about 15 to 20 Ma but persisted in Gondwana.

Sometime between about 70 and 55 Ma a distinct lineage of marsupials, the australidelphians, arose in Gondwana. Shortly thereafter, this lineage divided into two sub-lineages, one of which is represented today by only one species, Dromiciops gliroides, in South America and the other of which is represented by all Australian marsupials, the euasutralidelphians (Nilsson et al., 2010; see also Meredith et al., 2008 and 2011; Springer et al., 2009). How these two lineages came to be associated exclusively today with South American and Australia is unclear.

Four major groups are recognisable within living Australian marsupials: the kangaroos and their relatives (diprotodontians); the marsupial moles (notoryctemorphians); the possums and their relatives (peramelemorphians), and the quolls and their relatives (dasyuromorphians). The Tasmanian Devil is in the last group.

The relationships between these four groups are uncertain. The highly specialised notoryctemorphians (two burrowing, blind mole-like species) being especially difficult to place (Nilsson et al., 2010; Gallus et al., 2015b). Indeed, it is not certain that the relationships of these four groups will ever be resolved as a neat dichotomously branching tree (Gallus et al., 2015b).

The dasyuromorphians evolved some 25 to 45 Ma according to various accounts (Krawjewski, Wroe and Westerman, 2000 and Nilsson et al., 2004 and Nilsson-Janke et al., 2013; Gallus et al., 2015a-b; see also Westerman et al., 2015). The group contains a variety of living and extinct species, most of which are relatively small in size; primarily ground dwelling, and predaceous, mostly on insects. Of the living and recently extinct species, the Thylacine, with a weight of about 30 kgs, was the largest and the Long-tailed Planigale at about 6 gms the smallest.

Although the dasyuromorphians are clearly a group sharing a single common ancestor, the relationships among them are only partly understood (see Westerman et al., 2015 for a review). One of the most strongly supported relationships within dasyuromorphians, however, is that between the Devil, Sarcophilus, and the native cats, Dasyurus. This relationship has been recognised since shortly after the Devil’s discovery by science (Temminck, 1827) and has been supported by evidence from morphology (Temminck, 1827; Cuvier, 1837; Waterhouse, 1986; Archer, 1982; Wroe and Mackness, 2000, but see Kirsch and Archer, 1982 where there was strong support for a Thylacinus + Sarcophilus relationships), molecules (Baverstock et al., 1982; Thomas et al., 1989; Kirsch et al., 1991) and genetics (Thomas et al., 1989; Krajewski et al., 1992; Krajewski et al., 1993; Krajewski et al., 1994: fig. 2; Krajewski et al., 1997; Krajewski, Woolley and Westerman, 2000; Krajewski, Wroe and Westerman, 2000; Krajewski et al., 2004; Menzies et al., 2012; Nilsson-Janke et al., 2012; Zemann et al., 2013; Gallus et al., 2015a-b; Westerman et al., 2015).

An estimate of divergence times based on mitochondrial sequences differences and calibrated against the fossil record suggest that the Sarcophilus + Dasyurus lineage arose in the mid-Miocene about 12 Ma and the Sarcophilus - Dasyurus split occurred

6/05/2016 8:41 AM 6 sometime before 11 Ma (Krajewski, Wroe and Westerman, 2000). An estimate also based on the mitochondrial sequence data estimates the two events at 14 and 13 Ma, respectively (Nilsson-Janke et al., 2012; see also, Gallus et al., 2015a-b). A combined nuclear and mitochondrial DNA analysis returned a mean divergence date for the latter split of 13.5(range: 11.3-15.3) Ma (Westerman et al., 2015).

In most studies, the detailed relationship between Sarcophilus and the species of Dasyurus has been unresolved (Baverstock et al., 1990) or has been returned as a sister-group paring: Sarcophilus + Dasyurus. In a few studies, however, the relationship has been returned as Sarcophilus nested within Dasyurus (Kirsch and Murray, 1969; Archer, 1982; Krajewski et al., 1992). Interestingly, the latter result is supported by a shared physiological similarity, high blood acid phosphatase levels, in Sarcophilus and Dasyurus maculatus (Parsons and Guiler, 1972b; Sallis and Guiler, 1977). The only other Dasyurus examined for acid phosphatase levels is D. viverrinus (Sallis and Guiler, 1977), and its values are low. Although the current weight of evidence is strongly in favour of a sister group relationship, the possibility that Sarcophilus is actually derived within Dasyurus is worth remembering. Even in this context, however, the assertion in the draft species recovery plan for the Devil that its closest living relative is Dasyurus viverrinus (DPIPWE, 2010d) has no published research to support it.

The relationship between the Sarcophilus + Dasyurus group and other dasyuromorphians has been more problematic (Wroe and Mackness, 2000). Genetic evidence has suggested its relationship with: a group consisting of two genera from New Guinea, Phascolosorex and Neophascogale (Krajewski et al., 2004; Westerman et al., 2015 – latest and most comprehensive analysis); a monotypic genus from southwest Australia, Paranatechinus (Gallus et al., 2015a, but this study did not include the two New Guinea genera), or the Myrmebecobiidae from Australia, with one living species, the Numbat (Zemann et al., 2013). The extinct Thylacinidae is almost certainly close to this group, but whether it lies inside or outside it remains uncertain.

Most phylogenetic studies make it clear that the Sarcophilus + Dasyurus group is highly derived within dasyurids. In this context, the group’s relatively large size and carnivorous feeding habits can be seen as being evolved features within dasyurids. In addition, taking the inference one step further, the Devil’s scavenging habits can be seen as having evolved from the group’s carnivorous habits.

Interestingly, a recent study of the rates of evolution within living mammals based on body size suggests that the increase in body size of the Sarcophilus + Dasyurus lineage was the fastest recorded (Venditti et al., 2011). What drove this large and sudden increase in size is unclear.

Physical Characteristics

Size and Shape

Devils are often described as being the size of a small dog (cocker spaniel, corgi and terrier have all been mentioned) with a large head relative to the rest of the body (Sanderson et al., 1979). In fact, both their head and fore-quarters are robust

6/05/2016 8:41 AM 7

compared to the rest of the body, which slopes posteriorly. This distribution of bone and muscle are probably related to their scavenging habits. The large head may be an adaptation to rendering bone, cartilage and skin, and the large forequarters may be an adaptation to pulling off, by tugging, pieces of flesh from carcasses.

Devils can weigh up to 12 kg, but they are strongly sexually dimorphic (below) and only males attain this size.

Devils appear to vary geographically in size, but the accounts are conflicting. An early field worker noted that Devils from a northwest population were “smaller and weigh less” than those from a northeast population, but attributed this to nutritional status (Guiler, 1970a, 1978). The difference in mean mass between the larger and smaller population was 14 percent for males and 20 percent for females (data in Guiler, 1970a). And a Tasmanian government source says “devils on the East Coast of Tasmania are much larger than west coast individuals (DPIPWE, 2010b; see also Gliddon, 2005).” In contrast, a media story stated that the Devils from the far northwest are “among the biggest … in Tasmania (Denholm, 2010e).” This view could also be supported by the statement of a field worker, speaking of the Woolnorth Devils in far northwestern Tasmania, that “They grow them big up here (S. Fox, in Bain, 2010).”

Limbs

Devils have five digits on the front foot but only four on the rear foot (Bensley, 1903: plate 7, fig. 12; Pocock, 1926: figs. 34 and 37a; Moeller, 1972a: fig. 3; Guiler, 1985: fig. 3.4 d-e, 1992: fig. 15, 1998: p. 53, figs d-e; Tasmanian Parks and Wildlife Service, 2006, p.1; DPIPWE, 2010d, p. 2; Vogelnest and Allan, 2015: fig. 7.5). The first digit (hallux) is the one that has been lost evolutionarily on the rear foot (Pocock, 1926).

Devils are unusual in that the front and rear limbs are of nearly equal length (contra Eisenberg and Golani, 1977), and they are long relative to the vertebral column (Finch and Freedman, 1988). In the absence of a robust phylogeny among the dasyuromorphians, it is difficult to tell the direction of evolutionary change. But functionally, front and rear limbs of nearly equal length usually indicate an ability to run swiftly, hardly a noted capacity of the Devil.

Occasionally it is said of Devils that the front legs are longer than the rear legs (DEH, 2006a-b; DEWHA, 2009a-b). But in fact, the ratio of mean front limb to mean rear limb length is 0.96 (Finch and Freedman, 1988).

Colour and Pattern

Devils appear mostly black due to the colour of the outer fur (Green, 1967). Some older animals, however, appear slightly brownish (Guiler, 1964). This is probably due to the thinning of the outer-fur in older animals, which lets the dark brown, or henna, under-fur show through (Green, 1967; Mooney, 1992; Jones, 1994). The tip of the tail, especially, in older individuals may not only be a fluffier but also brownish.

6/05/2016 8:41 AM 8

Most individuals also have white markings. These markings can be conceptualised as being a transverse blaze across the chest and shoulders and a transverse across the rump. The chest and shoulder mark is often split into a central chest mark and two shoulder marks, each of which can be quite small. The rump mark is always complete but can vary in size (Green, 1967). There is also often a small white spot above each eye. Some Devils, however, are completely dark (Guiler, 1992: p. 2 and fig. 35).

Very rarely Devils may be mostly (Guiler, 1976: fig. p.156, 1992: fig. 8) or completely (Owen and Pemberton, 2005) albino.

The markings do not change with age (Pemberton, 1990), and the sexes do not differ in the distribution of the markings (Pemberton and Renouf, 1993).

The proportion of all black animals is about 13 to 15 percent in populations from northeast Tasmania (Pemberton, 1990: fig. 5.1 and Pemberton and Renouf, 1993, and Green, 1967, respectively) and about 5.6 percent (N = 263) in the northwest (Guiler, 1976). One field observer noted that “most of the Devils in … [the] south-west region were almost completely black, with the merest traces of the usual large white patches … seen in Devils farther north and east” (Fleay, 1952; see also Guiler, 1964).

There are two explanations for the Devil’s colour pattern, neither mutually exclusive. The black and white pattern may serve to break up the animal’s overall size and shape in dark conditions (Mooney, 1992). Or the markings, which are variable, may allow the animals to recognise each other as individuals (Buchmann and Guiler, 1977). The fact that some Devils have similar patterns or lack white markings all together need not diminish the discriminatory function of the patterns entirely (Pemberton and Renouf, 1992), as what pattern is recognisable could still assist in the overall identification of many individuals.

Observers and handlers generally find the pattern of white markings useful in identifying individual animals in the field (Gould, 1863; Guiler, 1978; Pemberton, 1990; Pemberton and Renouf, 1993; Jones, 1998; I. Albion, in Denholm, 2005; Hamede et al., 2008; Hughes et al., 2011; STDP website: “Population monitoring,” 25 September 2012), enclosures (Smith, 1993; Koncza and Pryor, 2001; Newsletter: June 2010; STDP website: “Populations persist at ‘ground zero,’ ” 1 June 2011; STDP Annual Report, 2012/13) and lab (Waterhouse, 1846; Jones et al., 2005). Devils may find it useful as well.

Viewers of the documentary “Terrors of Tasmania” (Parer and Parer-Cook, 2004), may like to use the unique white markings to count the number of Devils playing the role of the documentary’s star, “Manganinnie” (hint: it is more than two).

Parenthetically, the most famous, or perhaps infamous, Devil to date, Cedric (below), was all black (Woods, 2008).

Vibrissae

Vibrissae are long, slightly stiff hairs that usually occur in tuffs, most noticeably on the face, where they are often called, informally, whiskers. The vibrissae probably provide sensory information (“touch”) in dark narrow spaces.

6/05/2016 8:41 AM 9

Devils have five sets of facial vibrissae: supraciliaries, mystacials, genals, submentals and interramals. The genals, the vibrissae on the side of the head, are especially long, reaching back beyond the ear when adpressed to the side of the head (Pocock, 1914: fig. 1b, 1926: fig. 25; Lyne, 1959: fig. 5a).

Devils also have vibrissae, the ulnar carpals, on the inside of the middle front leg (Pocock, 1926: fig. 34; Lyne, 1959: fig. 5a; Guiler, 1970b; Jones, 1994).

In general, Devils have more vibrissae in all areas than other dasyurids, including the larger Thylacine (Lyne, 1959). The implication is that they are more sensitive to tactile stimulation in these areas than are other dasyurids. Perhaps this is due to the amount of time they spend in tight situations such as crevices, burrows and dens.

Sexual Dimorphism

Devils are strongly sexually dimorphic in mass (Jones, 1997: table 1). As adults, males are larger and heavier than females (Fleay, 1935, 1952; Table 1). At a northeast locality, the mean body mass of males and females was 9.0 kg and 6.5 kg (N = 73 and 83), respectively (Guiler, 1970a). And at northwest locality it was 7.7 kg and 5.4 kg (N = 18 and 20), respectively (Guiler, 1970a). Interestingly, it is “common” for the weight of an individual Devil to vary as much as ± one kg in one day (Pemberton, 1990). This represents a minimum eight percent change in the weight of the largest animals and even more in smaller animals (Table 1).

Table 1. Reported weights (kg) of Devils.

Males Females Maturity Reference Range Mean n Range Mean n 5.5-11.8 8.3 19 4.1-8.2 5.5 14 - Green, 1967 (original in lbs) - 9.0 73 - 6.5 83 - NE Tasmania, Guiler, 1970a - 7.7 18 - 5.4 20 - NW Tasmania, Guiler, 1970a - 6.9 100 - 5.4 100 - NW Tasmania, Guiler, 1978 - 8.2 50 -9 6.1 50 M NW Tasmania, Guiler, 1978 - 4.8 50 - 4.4 50 IM NW Tasmania, Guiler, 1978 - 10.2 - - 7.1 - - Pemberton and Renouf, 1993 9-12 10 - 5-8 6 - - Jones, 1994 5.0 – 10.4 8.4 56 3.6-8.5 5.4 138 M Jones, 1995: table 3.4 - 8.4 - - 5.4 - - Jones, 1998 - - - 5-8 - - M Jones and Barmuta, 1998 - 8.5 - - 5.4 - - Jones, 1997, 1998; Jones and Barmuta, 2000

6/05/2016 8:41 AM 10

- 10 - - 7 - - Lachish et al., 2009 - 8.3 34 - 6.6 50 M Peck et al., 2015: table 11

6/05/2016 8:41 AM 11

1. Combines data from two localities, one in the west and one in the east, and female data include some females with pouch young.

Adult males also have a “massive head and a broader neck compared with adult females” such that this difference is used to distinguish the two sexes in the field (Hamede et al., 2008).

Interestingly, head width compared to head length is not significantly wider in males and than in females (Fig. 2). This suggests that there has been neither trophic nor sexual selection in this parameter in either sex. There is no behavioural evidence of males fighting over females, and this is borne out by the lack of a wider head in males. The parameter may simply be driven by selection for strong bite force arising from the Devil’s scavenging habits.

2.2

2.1

2.0

Log Head Width Log Head 1.9

1.8 2.0 2.1 2.2 2.3 Log Head Length

Fig. 2. Plot of log mean head length vs log mean head width in male (x) and female (○) Devils. Data in Pemberton, 1990: appendix 2.

Canine strength in Devils is said to be greater in males than in females and was tentatively ascribed to sexual selection (Jones, 1995). But the comparison was made on the assumption that the primary (tooth size) and secondary (tooth strength) relationships within each sex are linear. A more robust test would take into account the possibly of a curvilinear relationship in one or both sexes. The issue, therefore, remains open, and its cause is problematic, especially in light of the lack of sexual dimorphism in head width (above).

Mature male Devils have an external scrotum, which is usually carried retracted into a pouch-like cavity in the abdominal skin (Guiler and Heddle, 1970: fig. 1; Hughes, 1982: fig. 2; Reppas et al., 2010: fig. 1; Keeley et al., 2011: fig. 1). The scrotum’s mean length, width and height are 3.25 cm ± 0.3, 2.50 cm ± 0.4 and 1.92 cm ± 0.3, respectively (Keeley et al., 2012b). The stratum is attached to the body through a peduncle stated to be approximately 50 mm in length in one account (Guiler and

6/05/2016 8:41 AM 12

Heddle, 1970) and measured at a mean length of 3.8 (± 0.6) cm (N = 44) in another (Keeley et al., 2012b) . This scrotal pouch would no doubt offer protection to the testes during normal activity. It is unclear, however, under what circumstances the scrotum descends from the pouch.

The narrow scrotal peduncle contains a rete mirabile consisting of 3 to 15 small arteries and 13 to 26 small veins (Guiler and Heddle, 1970: fig. 2). Such structures often function as counter-current heat exchangers, the influent veins absorbing heat from the effluent arteries and returning it to the body before it reaches the extremities. However, temperature measurements do not support such a function in Devils (Guiler and Heddle, 1970).

A layer of fat underlies the scrotal skin (Keeley et al., 2012b), but it is unclear what function it may serve.

Devils have the scrotum in front of the penis, which astute readers will realise is the opposite of humans.

Dentition

Devils have only one set of teeth during their life. The young do not have deciduous teeth as humans do.

The dental formula for Devils is: incisors 4/3; canines 1/1; premolars 2/2, and molars 4/4 (Laurent et al. 1839: plate 4, figs 2-3; Owen, 1840-1845: plate 98, fig. 2; Owen, 1841: plate 70, fig 5 (uppers only); Gerrard, 1862; McCoy, 1882; Thomas, 1887; Lord and Scott, 1924; Gill, 1953: plates 6; Guiler, 1964; Lundelius and Turnbull, 1978: figs 20-21; Archer, 1982 (“all dasyuroids”); Dawson, 1982; Guiler, 1985; Miles and Grigson, 1990; Holz, 2008). One report, in passing, gives the number of lower incisors as four (Guiler and Heddle, 1974a: fig. 1), but this is presumably a lapse. There are six sets of illustrations showing three (Owen, 1840-1845: plate 98, fig. 2; Owen, 1877: plate 5, nos 5-6; Archer, 1976: fig. 8 f; Mooney, 1992: fig., p. 58; Holz, 2008, p. 363; Jones, 1994: fig. p. 41), and in 40 individuals in the Australian Museum, I found three lower incisors in all 76 countable sides.

A fifth, only partially grown, upper molar occurs in some Devils (Green, 1967: plate 3), which may represent a molar otherwise lost in evolution. There is also the curious observation of a specimen with both P3 teeth rotated 180º (Miles and Grigson, 1990).

In general, the teeth erupt in an anterior to posterior sequence, but there is individual variation in the precise sequence and degree of eruption (Guiler and Heddle, 1974).

The first incisors on the upper jaw are relatively slow to erupt and do not do so until about the time of weaning. The resulting gap, however, provides a groove for the nipple in the suckling young (Guiler and Heddle, 1974).

Devil teeth have recently been discovered to “continue to erupt throughout life” or “over-erupt” (Jones, 1995, 1997, 2003a). Although the details of this “in prep.” (Jones, 1995) and “unpublished manuscript” work (Jones, 1997) have not yet been published, it has been concisely described as the teeth “continue to grow throughout

6/05/2016 8:41 AM 13

life (Jones, 1995).” And the canines have been specifically described as increasing in length and diameter “with age” such that in some old individuals the enamel-dentine junction has grown out to almost the tip” of the tooth (Jones, 1995, 1997).”

As in most mammals, the Devil’s teeth along a row have different functions. The delicate incisors are presumably used primarily in prehension. The “grossly enlarged (Archer, 1976),” long and pointed canines are used in fighting and killing. The two “rounded stumpy (Archer, 1976” premolars have no obvious function as they do not even occlude; their shape is “directly correlated with the “extreme shortening of the premolar row” (Archer, 1976; see also Thomas, 1887). The four molars are used for grinding and slicing, the relative area for the former decreasing and the relative area for the latter increasing from front to back (Jones, 2003a). The second molar appears to most of the bone crushing, to judge by degree to which it is worn down compared to the molars (Jones, 1994, 2003a).

Considering the way Devils use their teeth in cracking bones and fighting (Fleay, 1935) it is not surprising that the teeth often show loss, damage, wear and discolouration. In museum skulls, for example, nearly 70 percent of adult Devils have at least one broken tooth, and of the teeth broken, 15 percent are the canines (Jones, 1995, cited in Jones, 1997; see also Pemberton, 1990).

The four molars wear differentially from front to back, with the first two molars showing eventually the most wear (Pemberton, 1990: fig. 2.8; Robson and Young, 1990; Jones, 1994, 2003a). Older Devils in general have teeth that are yellowed, worn, broken and missing (Green, 1967; Mooney, 1992; Jones, 1994; Darby, 2011).

In the only reference to mention it specifically, Devil teeth showed no signs of decay (Green, 1967).

The enamel structure of Devils has been described in detail and compared with other dasyuromorphians but no phylogenetic or functional conclusions could be drawn (Stefan, 1999).

Convergence with Hyaenidae

Several observers have noted the similarity between Devils and hyenas, in body shape, skull morphology and strength and foraging strategy (Vrolik, 1852; McCoy, 1882; Fleay, 1935, 1952; Werdelin, 1986; Tate, 1947; Flannery, 1994; Jones, 1994, 2001, 2003a). The similarities include large head and massive fore-quarters, scavenging habits and the ability to crush and eat bones. This latter is achieved by the evolutionary movement of the bone crushing dentition closer to the hinge of the jaw, where the bite forces are greater (Werdelin, 1986). There are major differences, however. For example, Devils pass the bones they eat largely undigested in their faeces, whereas hyenas digest them completely (Lundelius, 1966).

Temporomandibular Joint

Although the Devil shares virtually all of its anatomy with other dasyuromorphians, it is seeming unique in one regard, the absence of the articular disc in the temporomandibular joint, the joint between the skull and the lower jaw. In Devils, the

6/05/2016 8:41 AM 14

armadillo and monotremes, the cartilaginous disc that separates the bone on the two sides of the joint is in direct contact (Parsons, 1900; Hayashi et al., 2013, 2014). Why Devils lack this basic “force-buffering” structure is unclear.

The best approach to this question would be to survey increasingly phylogenetically distant relatives of the Devils to determine when in marsupial evolution the articular disc disappeared and the life history characteristics associated with its loss. Is the Devil truly unique in this regard? Or is it a feature shared with only Dasyurus or with some or all dasyuromophians?

Finally, several figures of radiographs of the Devil’s skeleton and dentition are available in Vogelnest and Allan, 2015.

Distribution

In European times, Devils have occurred naturally only in Tasmania: on the main island and only two of the 334 smaller islands surrounding the main island. The Devil probably occurs throughout the main island (Rounsevell et al., 1991: fig. 8; DPIWE, 2005a: fig. 2; DPIPWE, 2010d), although its density varies greatly with habitat (below).

The Devil also occurs today on Robbins Island (Hawkins et al., 2008; DPIPWE, 2010d; STDP 2011b), which is separated from the main island on the northwest coast only by tidal flats and is accessible to Devils during spring low tides (PHVA, 2008).

There are compelling historical records of Devils on Bruny Island in the mid-1800s, (Medlock and Pemberton, 2010; see also N. Mooney, on ABC Rural, Bush Telegraph, 20 September 2011 and 25 May 2012). There are even relatively recent reports of roadkills on Bruny (R. Denne, pers. comm. to Hird, 2000), but in the absence of a specimen, these recent observations are hard to credit. An analysis of a Tasmanian state government database, mostly for the period between 1967 and 1989, was incorrectly interpreted as showing Devils on Bruny (Hird, 2000). In fact, the analysis in question stated categorically, although incorrectly, that the Devil is “absent from all islands (Rounsevell et al., 1991).” In 2015, some Devil researchers tabulated Bruny Island as being “devil-free (Fancourt, 2015a; Fancourt et al., 2015” but did not give their reasons. Considering the historical, if not the relatively recent, evidence for Devils on Bruny, the “devil-free” determination and the statement that Devils “… have never been present [on Bruny] (Hollings et al., 2014)” may be a bit too categorical.

If Devils did occur on Bruny in European times, it raises two questions. First, why were they on this island but no similarly isolated island at the time of European settlement. And second, what caused them to go extinct? To answer the first, Bruny may have been the last large island to be isolated by rising sea levels. The depth of the intervening D’Entrecausteaux Channel is only about 10 m in many places and the island’s large size may have mitigated the impact of isolation on the population. And the answer to the second may have to do with the effect of European clearing and killing may have finally had on the population within the century following European arrival in Tasmania (Medlock and Pemberton, 2010).

6/05/2016 8:41 AM 15

Interestingly, aside from Bruny, Devils occur on no other off-shore island around the main island or in Bass Strait and yet these islands would have all been attached to the main island and hence had Devil populations on them during the height of the last glaciation (see bathymetric chart in Cosgrove et al., 2014: fig. 14.1). This suggests that, all else being equal, Devils could not survive on islands up to the size of even the largest island, Flinders Island (1333 km2). Assuming a “high” density of 2 Devils/km2, this would suggest that isolated Devil populations smaller than about 2700 individuals may not have been self-sustaining.

In 1996, unnamed persons introduced “two or three pairs” of Devils to Badger Island in the Furneaux group off the northeast corner of Tasmania (Wade, 2005b; Hawkins et al., 2008; PHVA, 2008; DPIPWE, 2010d; STDP, 2011b; Mooney, 2012). The animals came from the Northern Midlands and Bridport area, were ferried to nearby Flinders Island and flown to Badger (STDP, 2011b). This illegal translocation was possibly done out of spite due to the island being handed back to its Aboriginal owners in that year. Subsequently, the owners requested that the Devils be removed, and removal was underway by early 2005 (Wade, 2005b; PHVA, 2008). Where the Devils were originally sourced from is unknown.

The introduced Devils bred, and 120-130 individuals may have occurred on the 14 km2 island at the time of removal. However, by August 2007 all the Devils were “thought to have been removed” (N. Mooney, in Hawkins et al., 2008, but see Mooney on ABC Bush Telegraph, 22 May 2012 and Insurance Population: Off-shore islands, below). One hundred and twenty-six trapped animals were released at various sites on the mainland, including Bridport, Epping Forest and Georgetown (STDP, 2011b), and two female Devils with their young were sent to a captive breeding facility at Bicheno (Radio National, Innovations: Devil Island Project, 26 May 2008). Interestingly, the authorities were not able to catch all the animals (below), and they may still inhabit the island.

In the not so distant past, Devils, or at least animals in the genus Sarcophilus, occurred widely on mainland Australia, at least in its more mesic periphery, as well as on Tasmania. But some time prior to about 3000 years before present, that is, during the mid-Holocene, Devils disappeared from the mainland (Brown, 2006). The remains of mainland Devils, which consist entirely of bones and teeth, are in some cases so recent as to not even be mineralised (McCoy, 1882; Kershaw, 1912; Mahony, 1912; Gill 1953).

Some early estimates placed the Devil’s demise on the mainland are as recent as about 430 ±160 years (Archer and Baynes, 1972; Archer, 1979). Two subsequent reviews, however, cast doubt on all dates earlier than about 3000 years, including the 430 year date (Brown, 2006; Letnic et al., 2014). Despite this, many commentators continued to cite, without qualification, the recent dates, including the most recent one of c. 430 years (Siddle et al., 2007a; Jones, 2008; University of Adelaide media release, 3 November 2008; Kreiss et al., 2009a; J. Weigel, on ABC World Today, 14 July 2010; Belov, 2011; Bender, 2010; DPIPWE, 2010d; STDP website: “Background,” 23 March 2010; STDP, 2010b; Woods and Sharman, 2010; ABC News, Behind the News, 14 June 2011; Newsletter, March 2011; Boness, 2012; Tovar, 2012; Milman, 2013b; Warren, 2013; Doman, 2014; Gallus et al., 2015a; Marris, 2015).

6/05/2016 8:41 AM 16

Why Devils Disappeared from the Mainland

There are three hypotheses for the Devil’s disappearance from the mainland, all of which could have acted simultaneously (Johnson and Wroe, 2003). First, Dingoes, which were introduced into Australia about 5000 years ago by seafarers from southeast Asia (Gollan, 1984; Corbett, 1995), may have led to the extinction of Devils on the mainland (Smith, 1909; Le Souef and Burrell, 1926; Troughton, 1938; Fleay, 1952; Bauer, 1961; Guiler, 1983a; Flannery, 1994; Jones, 1995; Megirian et al., 2002; Morris et al., 2013) through predation (Wintle, 1886; Sharland, 1940; Johnson, 2014), competition (Fleay, 1952; Guiler, 1982, 1983b; Taylor, 1986; Mooney, 1992; Flannery, 1994; Jones, 1997; Johnson, 2014) or the introduction of disease, or some combination of these (Johnson, 2006; Morris et al., 2013). This hypothesis is attractive because the dates for the dingo’s arrival and the Devil’s demise correspond closely, and in Tasmanian today, free-ranging domestic dogs will kill Devils if given the chance (Hawkins et al., 2008). The trouble with this idea, however, is that although Dingos never made it to Tasmania, wild dogs have been there since European settlement, and Devils have thrived despite their presence. It could be argued, however, that wild dog numbers have been kept low. Another difficulty with the Dingo hypothesis of Devil mainland extinction is that Devils also disappeared from Kangaroo Island (Calaby and White, 1967; Hope et al., 1977), and Dingoes never occurred there.

Second, humans, who arrived in Australian possibly as long ago as 60 000 years ago, may have driven Devils to extinction (Guiler, 1982) either through hunting (Flannery, 1994; Johnson and Wroe, 2003; Johnson, 2006; Johnson, 2014), competition (Guiler, 1982; Johnson, 2014) or habitat change due, for example, to altered fire regimes (Johnson and Wroe, 2003; Johnson, 2006; Prowse et al., 2013). One observer even opined, somewhat confusedly, that “… competitive killing [?] by Dingoes and Aboriginal peoples probably expedited their demise ….” (Jones, 2008). The trouble with this idea, however, is that Devils co-existed with Aboriginals on the mainland up until relatively late in Aboriginal occupation and in Tasmania throughout Aboriginal occupation. The counter argument, however, is that the Devil’s demise on the mainland correlates with an apparent increase in Aboriginal technology, density and distribution about 3000 years ago but these changes did not affect the by then isolated Tasmania (Johnson and Wroe, 2003; Prowse et al., 2013).

Third, climate change in the last ten thousand years may have led to the Devils demise on the mainland. One version of this view focuses on the climate’s increasing aridity in this period (Guiler, 1982; Jones, 1995, 1997, 2008; Johnson and Wroe, 2003) while another version focuses on its increasing variability (Brown, 2006). Presumably, the critical feature in both cases is a lack of food, mainly carrion, in the leanest times. It seems odd, however, that Devils did not find refuge in the mountainous parts of the southeastern mainland, which are not so different from Tasmania. Perhaps the Tasmanian enclave was just benign enough during the worst of times to allow Devils to hang on.

Currently, the climate change hypothesis seems to fit all the data best. Tasmania would have experienced the general aridity of the mid-Holocene, and there is evidence the Devil population declined at this time (Brüniche-Olsen et al., 2014). Further, Tasmania did not have the other two factors suggested as causes of the

6/05/2016 8:41 AM 17

decline and extinction on the mainland, Dingoes and a technologically advanced Aboriginal population. Hence, the mid-Holocene aridity caused the mainland population to decline and then go extinct but only caused the more temperate Tasmanian population to decline.

Since 1912, members of the public have found six Devils on the mainland, all in Victoria. This number included two live animals and four dead ones, in the latter case all road kills (Kershaw, 1912; Troughton, 1967; Guiler, 1992; Bevilacqua, 2001; White, 2009; Mainland Devils Website). This low but steady flow of specimens has led to speculation that Devils may still be living naturally on the mainland or at least did so up until very recently. However, the number of mainland specimens is so small compared to those encountered in Tasmania over the same period that it is almost certain that these mainland animals have either escaped from captivity or been purposely released. In support of this interpretation is the fact that up until the latter part of the 1900s, it was legal to catch and transport native mammals anywhere is Australia. Furthermore, Devils are proven escape artists. They have absconded from a government research facility in Tasmanian (ABC News, 27 and 28 July 2006), a government quarantine facility in Tasmania (ABC News, 28 July 2006) and zoos both in Tasmania (Roberts, 1915; Paddle, 2012; Andrewartha and Campbell, 2006) and on the mainland (L. Souëf, in Kershaw, 1912; The Argus, 13 February 1939; Fleay, 1935, 1952; Robertson, 2012). Indeed, as early as 1882, two Devils “escaped into the bush from a menagerie in the Clarence River district (Launceston Examiner, 15 July 1882). The most recent specimen to turn up on the mainland, a road kill, had escaped from a local zoo in Grantville, Victoria about two weeks previously and had travelled about 30 km before being killed (White, 2009).

There is a claim that the last population of wild Devils on the mainland was in Victoria (Owen and Pemberton, 2005). But the claim is unsupported and there is no evidence for it.

In 1992, an honours student attempted to sequence a highly variable part of the mitochondrial genome to see if four Devils found in Victoria in the 20th Century were distinct from the Devils in Tasmania and hence to tell if there might be a remnant mainland population. However, technical difficulties made the attempt uninformative (Blake, 1992). It is surprising this approach has not been revisited, especially in light of the advances in genomic technology.

A related issue is why no Devils occur today on virtually none of Tasmania’s offshore islands, almost all of which would have been connected to the main island during the c. 130-150 m lower sea levels of the Pleistocene’s last glacial period 18 to 20k ybp. Tasmania has 334 off-shore islands, ranging in size from little more than a barren rock to 1334 km2 Flinders Island, and Devils occur on only one naturally, 99 km2 Robbins Island, which is connected to the mainland across Robins Passage at very low tides.

As post-Pleistocene sea levels rose, Devils might not have lasted very long on small islands due to the islands’ small population sizes, which would be subject both to inbreeding and random events such as bushfires. Devils, however, reached densities of about 8.6/km2 on 12.4 km2 Badger Island in Bass Strait in the 11 years between their

6/05/2016 8:41 AM 18

introduction in 1996 and their removal in 2007 (Hawkins et al., 2008), suggesting that they can thrive, at least in the short term, on relatively small islands.

The introduction of Devils onto 115 km2 Maria Island off the east coast of Tasmania (below) may shed more light on this mystery. If the Devils survive and a stable population develops, the cause might be inferred to have been something other than island size. But if the Devils decline, island size could be part of the answer.

Fossil Devils

Fossil and sub-fossil remains of the genus Sarcophilus occur in early Pliocene (5.3 – 3.6 Ma, age dates) to Holocene (10 ka - present) deposits in Australia. On Tasmania, the only place where the genus occurs today, the fossil record goes back to at least the late Pleistocene (Garvey, 2007).

To date, fossil devils are known from all parts of Australia, except the northwest. References to the remains are given below chronologically by region.

Southwest: Glauert, 1912, 1914, 1948; Cook, 1960, 1963a-b; Lundelius, 1960, 1963, 1966; Douglas et al., 1966; Butler, 1969; Archer and Baynes, 1972; Baynes et al., 1976.

South-central: Lowry and Lowry, 1966; Baynes, 1987; Walshe, 1994.

Southeast: McCoy, 1882; Mahony, 1912; Hale and Tindale, 1930; Colliver, 1938; Keble, 1946; Gill, 1953, 1971; Hale, 1956; Mulvaney, 1960; Wakefield, 1964a-b; Flood, 1973, 2007; Hope, 1972, 1978; Macintosh et al., 1970; Macintosh, 1971; Marshall, 1973; Balme et al., 1978; Gillespie et al., 1978; Williams, 1980; Shawcross, pers. comm., in Walshe, 1998; Stefan, 1999; Reed and Bourne, 2000; Gerdtz and Archbold, 2003b; Garvey, 2007.

East-central: Longman, 1921, 1925, 1945; Sobbe, 1990; Hocknull, 2005; Cramb et al., 2009.

Northeast: Horton, 1977.

North-central: Calaby and White, 1967.

Central: Megirian et al., 2002.

Islands: Kangaroo Island - Calaby and White, 1967; Hope et al., 1977; Flinders Island - Hope, 1972; Guiler, 1992).

Fossils from various early Pliocene to late Pleistocene localities in southeastern mainland Australia are indicative of an animal larger than the living species. Estimates of the (presumably linear) size difference between the two forms vary, with one-third larger (Owen, 1838), 15 to 50 percent larger (Gill, 1953), an average about 15 percent larger (Marshall, 1974; Marshall and Corruccini, 1978) and an average of approximately 14 percent larger (Dawson, 1982) all having been mentioned. The latter two estimates appear to be based on the largest sample sizes and hence are

6/05/2016 8:41 AM 19

probably the most nearly accurate. A “fragment of a megafaunal devil jaw in the Queen Victoria Museum and Art Gallery in Launceston” that is “about 50 percent larger than that of the extant species (Owen and Pemberton, 2005)” had apparently heretofore gone undetected.

On the basis of fossils from Wellington Caves in central western New South Wales, the British palaeontologist, Richard Owen, named the species represented by the larger fossil specimens Dasyurus laniarius on the basis of its proportionately larger canines (laniaries) (Owen, 1838). Later, he transferred it to the genus Sarcophilus (Owen, 1877) and in addition to its larger size, distinguished it from the living species by features of the skull and mandible.

In addition to these large (Sarcophilus laniarius) and medium-sized species (S. harrisii), there are a few fossils from some late Pleistocene to early Holocene localities indicative of an animal smaller than the living species (Dawson, 1982; see also Horton, 1977). This smaller animal seems to have been referred to once in a diagram under the name S. laniarius dawsoni (Werdelin, 1987: fig.9). However, the name is a nomen nudum and has no nomenclatural validity.

There is also a small Sarcophilus from what are tentatively thought to be early Pleistocene (Crabb, 1982) or Pliocene (Wroe and Mackness, 2000) deposits in southwest New South Wales. The animal was named S. moornaensis and interpreted as an ancestor of S. harrisii (Crabb, 1982). But it was not compared with, or even discussed in the context of, S. laniarius and the smaller unnamed species, both of which were discussed in another chapter in the same volume (Dawson, 1982). If the two small species are the same, the name S. moornaensis would apply to it. It is worth noting, however, that taxonomic work never published apparently considered S. moornaensis the same species as Galucodon ballaratensis (below; Aplin and Hope, J. Hope pers. comm. in Werdelin, 1987).

Finally, subfossil remains from the Holocene of one Victorian locality have been named Sarcophilus laniarius dixonae in a study where the species name laniarius was used for all Sarcophilus forms, fossil and living (Werdelin, 1987). The diagnosis was based on tooth measurements, which would have to be tested in a wider comparison (below) to be convincing. The name has never been used subsequently.

The chronology of the recognised fossil forms of Sarcophilus is shown in Table 2.

Table 2. Chronology of the four forms of Sarcophilus currently recognised.

S. S. S. S. Epoch laniarius harrisii small moornaensis Holocene Holocene-Late Pleistocene Pleistocene, late Pleistocene, mid Pleistocene, early Pleistocene-Pliocene Pliocene, late Pliocene, mid Pliocene, early

6/05/2016 8:41 AM 20

It is important to remember the main distinguishing feature among all these forms is size or size related. Some researchers have noted additional osteological differences (Owen, 1877; Calaby and White, 1967) and dental proportional differences (Crabb, 1982; Dawson, 1982), but no one has investigated these differences in the context of the ontogenetic and geographic variation in the living species (Wells, 1978). For example, Devils from (warmer) northeast Tasmania may be, on average, 14 percent (males) and 20 percent (females) heavier than Devils from (cooler) northwest Tasmania (see “Size and Shape,” above).

Various authors have speculated about phylogenetic relationships for the forms of Sarcophilus known at the time, but as the main guides have been chronology and size (or size related) the inferences are highly problematic. Several authors have suggested that S. laniarius evolved into S. harrisii at the end of the Pleistocene when conditions were warming (Lydekker, 1887; Marshall, 1973; Marshall, 1974; Marshall and Corruccini, 1978). This hypothesis seemed reasonable in terms of the known chronology at the time and the well known temporal and spatial tendency for some mammal species to be larger in cool climates and smaller in warmer climates (Bergman’s Rule). However, the chronologies of the two species are now thought to overlap broadly, and contrary to the “rule,” Devils from the cooler and wetter northwest corner of Tasmania are lighter in weight than those from the warmer and drier northeast corner.

The original describer of Sarcophilus moornaensis thought it was “… probably in the direct evolutionary line leading to S. harrisii …,” but he based his interpretation on the basis of a trend in the size of the teeth among some closely related fossil and living forms (Crabb, 1982).” He did not mention S. laniarius.

An early reviewer of Sarcophilus fossils suggested two possible schemes of relationship for the three forms she dealt with. In the first, the small form gave rise to S. harrisii and S. laniarius was on a separate line, the two lines presumably joining earlier in time (but not so indicated) (Dawson, 1982: fig. 3.1). In the second scheme, S. laniarius gave rise to S. harrisii while the small form was on it own earlier line (Dawson, 1982: fig. 3.2). The reviewer did not mention S. moornaensis by name or material, although it was described in the same volume.

The last opinion expressed about the relationships of the forms of Sarcophilus was (moornaensis, small form, (laniarius, (dixsoni, harrisii))) (Werdelin, 1987).

The hypothesis that Sarcophilus laniarius evolved directly into S. harrisii (above) has implications for the name of the living species, because Owen’s 1838 name for the extinct animal, Sarcophilus laniarius, has priority over Boitard’s 1841 name for the living animal, Sarcophilus harrisii. Hence, it would be the “correct” name for the living species under the International Code of Zoological Nomenclature (Stephenson, 1963; Werdelin, 1987; also Marshall and Corruccini, 1978). Indeed, for a while, some researchers did use the name S. laniarius for the living species (e.g., Jones, 1997, 1998; Jones and Barmuta, 1998, 2000; Jones et al., 2003a-b; Harington et al., 2006; McGlashan et al., 2006; Obendorf and McGlashan, 2008). Later, however, they reverted to the more familiar name, for reasons never explained.

6/05/2016 8:41 AM 21

Given that the four fossil forms of Sarcophilus are distinguished mainly by differences in size and form of the teeth, the reasonable null hypothesis of their relationships is they just represent temporal and spatial variation within a single species (Werdelin, 1987). In this regard, it is worth noting that where the stratigraphy is well understood, there is no indication of sympatry among the forms.

With this in mind, it would be useful to revisit the differences among all the Sarcophilus fossils and living form with a single multivariate study of tooth measurements in a study of all available specimens that was blind to previous names, localities and age.

Finally, the fossil genus Glaucodon needs mention. The single species in the genus, G. ballaratensis, is known from two fossils from the Late Pliocene of Victoria. A relationship between Glaucodon and Sarcophilus and Dasyurus has long been recognised, but its precise phylogenetic relationships remain uncertain (Stirton, 1957; Ride, 1964; Archer, 1976; Gerdtz, 2001; Gerdtz and Archbold, 2003a and references therein). The most recent interpretation of its relationships is, ((Sarcophilus + Glaucodon) + (Dasyurus)) (Gerdtz and Archbold, 2003a).

Habitat

Devils occur in a variety of terrestrial habitats. These include open forest, woodland, shrubland, heathland and coastal scrub. They also occur in ecotones between these natural habitats and human-made clearings (Guiler, 1970a, 1983a, 1992; Hawkins et al., 2006; Jones, 2008). Devils occur in their highest densities in coastal heathland and woodland (Guiler, 1970a; Jones, 1995; Hawkins et al., 2006). They occur only in low densities, if at all, in alpine areas, moist closed forest, low heathlands and open grasslands as well as extensively cleared agricultural areas (Guiler, 1970a, 1992; Jones et al., 2004a-b; Jones, 2000). Devils may occur in wet eucalyptus forests with a clear understory and few obstructions, that is, where the horizontal visibility is high (Jones and Barmuta, 2000).

These habitat preferences mean that within their total range, Devils are most common in the east, north and northwest parts of Tasmania (Hawkins et al., 2006; Jones, 2008).

Some people fear that logging will imperil Devils locally (ABC Hobart, 5 July 2010; The Mercury, 8 September 2010). No doubt, some forestry operations might drive some Devils away or kill them outright, and dens could be destroyed or made useless through exposure. But whether there would be a net loss of Devils though the habitat destruction depends on the type of forest. Densities of Devils in thick forest is often low, ≤ 2 animals/km2 (Jones et al., 2007a), and in these areas the density of Devils in the ecotones created between regenerating logged areas and unlogged areas might be much higher.

Some field workers have suggested that the clearing for agriculture and forestry might be favoured by such practices, as they are likely to lead to an increase in the number of carcasses available (Driessen and Hocking, 1992).

6/05/2016 8:41 AM 22

An equal or even greater worry about forestry operations in any area as yet unaffected by facial tumour disease is the increased access by diseased Devils through the creation of roads and jigging trails.

Activity

Other than on the roads at night and at tourist facilities that attract Devils to large carcasses, Devils can be very difficult to see. For example, in 12 years one experienced field worker saw only half a dozen Devils at the field site where she had been working for 12 years (Grzelewski, 2002). Most of what is known about Devils in the wild has been gathered either by trapping or telemetry.

Daily Activity

In the wild, adult Devils are usually described as being primarily nocturnal (Gunn and Gray, 1838; Guiler, 1970c, 1992; Pemberton, 1990; Jones et al. 1997). They may, however, also be out in the twilight after sunset (Pemberton, 1990), presumably just after emergence, or in the first light of pre-dawn (Bauer, 1961), just prior to retiring. Wild juveniles are also occasionally active during the day (Jones, 2008), as are adults in deep snow (Jones, 2008).

At one study site in the northeast, in both summer and winter, Devils emerged between 34 and 128 (mean = 74 ± 31.2 SD, N = 34) minutes after sunset. In summer, they retired between 127 minutes before and 56 minutes (mean = 5 ± 50.1 SD) after sunset (data were insufficient for winter retiring times; Pemberton, 1990).

A sentinel camera study at 11 sites in three areas across the eastern part of the state and carried out to avoid the confounding effect of season and location found most Devil activity was between 1800 h and 800 h in all three areas. Intriguingly, peak activity seemed to occur at approximately 400 h in the northeast, 000 h in the centraleast and 2200 h in the southeast (Fancourt et al., 2015: fig. 6a). These possible differences were not analysed statistically, however.

Another sentinel camera study at 48 sites across the northern part of Tasmania showed that “Devil activity peaked within the hours of darkness with most activity just after sunset, but with another peak a few hours before dawn (Hollings, 2013).”

There is no significant difference between males and females in the time and distance travelled during a complete foraging period, that is, from emergence to retirement (Pemberton, 1990).

Although almost all experienced observers regard Devils as being generally nocturnal, one observer found that Devils on the remote west coast were “quite diurnal” and were out and about in the “middle of the day.” He attributed this diurnal behaviour to both increased competition for food (a population at carrying capacity) and a lack of human persecution (STDP Website; “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011; see also Mooney, 1992). In February 2011, a group of tourists also saw a recently weaned Devil foraging along the beach on the west coast (STDP website: “A seaside odyssey,” first published 22 March 2011). Oddly, a virtually identical account was reported as having happened in

6/05/2016 8:41 AM 23

February 2012 (Kempton, 2012a). It is not clear, therefore, if this is one or two observation of diurnal foraging in a young Devil.

In captivity, the Devil’s daily rhythm is often different from what it is in the wild, possibly because the animals are encouraged to be active during the day by the schedules of their keepers. They may be fed, for example, during the day when their keepers are at work. And in the case of zoos, they have to be out during the day to be seen by the paying customers.

Devils show a circadian rhythm in their daily activity times, at least to judge from two captive animals, a male and female, in an experimental situation (Packer, 1966). When activity was measured by how much time the captive Devils spent running in a wheel in a 12 hour light/12 hour dark cycle, the animals became active after dark. But the male became active about three hours after “dusk” and remained active over a core period. The female, however, became active about an hour after dusk and remained active for a shorter period than the male but then had a second period of activity an hour or so before dawn. Unfortunately, the small sample size precluded any possibility of determining whether the differences in activity times were due to the sexual or individual differences.

And when one animal, the female, was then put into a continuous dark cycle, the onset of activity slowly drifted forward (earlier). In contrast, when the other animal, the male, was put into a continuous light cycle, the onset of activity slowly drifted backward (later). But whether this was indicative of a real sexual difference as opposed to an individual difference is unknown. And in both cases, the discrete bouts of activity seen under the day/night cycle broke down and activity became somewhat more continuous throughout the 24 hour period.

When asleep, Devils show bouts of both Slow Wave Sleep (SWS) and Paradoxical or Rapid Eye Movement Sleep (REM), as do most mammals. Behaviourally, SWS is characterised by general body quiescence. And REM sleep is indicated by rapid eye movements, twitching vibrissae and ears, mastication movement of the lower jaw and uncoordinated twitches and movements of the limbs (Nicol and Maskrey, 1980). Humans slow the same general patterns. In humans, of course, dreaming occurs during REM sleep.

Seasonal Activity

In the northeast part of the state in summer, Devils emerge about 30 minutes after sunset and have retired no later than about an hour and a half after sunrise. In winter, however, they emerge about two hours after sunset but when they retire is unrecorded. The earlier emergence in summer might be due to the stronger odours emanating from carcasses at this time (Pemberton and Renouf, 1993: fig. 2).

In the central highlands however, Devils begin and cease their nocturnal activity at about the same time throughout the year. On average, they became active at 2105 h and inactive at 0205 h (N = 11), spending on average about 7.9 h active each night (Jones et al., 1997) And based on one animal, the time a female spends away from her den with and without young in the den did not differ significantly (Jones et al., 1997: fig. 1).

6/05/2016 8:41 AM 24

At higher altitudes, Devils are just as active during severe winter weather as during milder conditions, but they move less when it is sleeting or snowing and when there is deep snow on the ground. Because of their weight and narrow foot, Devils sink into the snow, making it more difficult for them to move, and hence impeding the extent of their movements (Jones et al., 1997).

Devils are active year around. They may be less active during winter at higher elevations during winter, but they are still out and about (Jones et al., 1997).

There is no evidence that Devils undergo seasonal torpor at any time in any part of their range, including winter in the central highlands (Jones et al., 1997; Jones and Barmuta, 1998). If a lack of torpor is any indication of the Devil’s resilience in cold conditions, it is of further interest to note that Devils lived in southwestern Tasmania during the last glacial maximum (Gravey, 2007).

Devils are susceptible to unusually harsh seasonal conditions. During the drought of 2001, adult Devils were in noticeably poor condition coming into the breeding season (M. Jones, in Hamede et al., 2008).

In outdoor enclosures, where they were encouraged to be active during the day by being fed at this time, Devils were more active in winter than in summer, therefore requiring more food. They were minimally active on hot days in summer (Kelly, 1993).

Movements

Like the young of most mammals, young Devils usually disperse after weaning. And they may move large distances in their search for a home area of their own. A young (18 months) female, for example, moved 110 km in two and a half months, an average of 1.2 km per day (Newsletter, 2007). And another sub-adult female moved 25 km in 10 days and then another 8 km over the next two days (M. Jones, in Jones et al., 2004a). In general, Devils 1-2 years old move more than older animals (Brandenburg, 2010).

A male Devil of unspecified age, but “out of home range,” moved an astonishing 50 km in one day (M. Jones in Jones et al., 2004a; Parer and Parer-Cook, 2004; McCallum and Jones, 2006; C. Pukk, in Hawkins et al., 2006). And another adult male was said to have moved 50 km in 48 hours (Jones, 2008). Another animal was reported to have made “a regular 40-kilometre return trip to visit a rubbish tip (Jones, 2008; see also M. Jones, in Parer and Parer-Cook, 2004).

Individual Devils can easily move three km in a night (Guiler, 1970a), and movements of seven km in a night have been recorded (M. Jones, in Hamede et al., 2008; see also Pemberton, 1990: fig. 6.6A). Indeed, one lactating female moved 17.7 km (11 miles) in two days (Guiler, 1970a), which gives an average distance of 8.8 km per day. There is also mention of a movement of 16 km in one night (Guiler, 1978).

Over a ten day period, animals in a 10-20 km2 grid moved between 0 and 6.5 km/day with an average of 1.0 km (Brandenburg, 2010).

6/05/2016 8:41 AM 25

A Devil may travel 8 km without stopping (Pemberton, 1990a).

The sexes differ in the distance of their movements. An early study found that the mean distance between recaptures for males was 1.8 km and for females 2.3, but no statistical comparison or dispersion numbers accompanied these figures. It was suggested that females move more than males due to their greater foraging activity when they were rearing young (Guiler, 1978).

There may also be a seasonal effect on movements in the sexes. In summer, males move more than females; whereas in winter, females move more (Brandenburg, 2010).

Weather seems to make little difference in Devil movements; at least it seems to have minimal effect on trapping success (Pemberton, 1990).

Devils often follow runs and trails made by other species, both exotic and native (Fleay, 1952; Green, 1967; Guiler, 1970a, 1978; Mooney, 1992). They also follow vehicular tracks (Guiler, 1970a, 1978; Taylor et al., 1985; Mooney, 1992; Lachish et al., 2007), beaches (Mooney, 1992), creeks (Mooney, 1992; Lachish et al., 2007; Hawkins et al., 2008) and rivers (Hocking and Guiler, 1983), although the attraction of the latter two may actually be their associated runs and trails. Perhaps they would follow any trail, track or road if it suited their intentions. In fact, researchers find these pathways and their junctions with each other and with creeks preferred trapping areas (Guiler, 1978; Hawkins et al., 2006; Lachish et al., 2007; C. Hawkins, in Hughes et al., 2011).

When unperturbed, Devils move steadily at about 3-4 km/h. One Devil, perhaps attracted by the smell of a carcass, moved 4 km in 45 minutes, an average speed of 5.3 km (Jones, 1005). The individual that moved 50 km in a day (above) would have averaged about 2.1 km/h. Humans walk at about 4-5 km/h.

When chased, Devils can run at about 12 km/h for short distances (Guiler, 1983a, 1983b, 1992). Over a short distance, at least, a human can run them down (Meredith, 1853; Guiler, 1992). One experienced observed reported Devils as “… capable of speeds up to 25 km/h when running along roads” (Jones and Stoddard, 1998) and later as “… running along roads at over 30 km/h” (Jones, 2003a). The top speed, clocked from a car along 300 m, was 35 km/h (Parer and Parer-Cook, 2004; M. Jones, in Owen and Pemberton, 2005). Humans run at about 10-16 km/h.

As an argument that Devils are unlikely to be able to catch healthy lambs, and by implication adult sheep, one experienced observed stated that “a lamb [after its first few days of life] can run much faster than a devil (Guiler, 1970c).”

The stamina of Devils can be partly inferred from experimental situations in which the animals run on treadmills moving at a fixed speed. In such a situation, maximum values were 30 minutes at 7 km/h, 22 minutes at 9 km/h and 14 minutes at 10 km/h (Nicol and Maskrey, 1986: figs 2-3). An experienced field worker also noted that Devils were capable of maintaining 10-15 km/h for several kilometres (N. Mooney, in DPIPWE, 2010d).

6/05/2016 8:41 AM 26

When breaking into a burst of speed, Devils use anaerobic (oxygen-free) metabolism, which delivers a greater amount of energy over a short period than does aerobic metabolism but at a cost in terms of debilitating by-products that accumulate temporarily in the blood. If the activity is intense and prolonged, the animal may fatigue and stop what it is doing, and the metabolites may take more than an hour to be metabolised aerobically (Nicol and Maskrey, 1986).

Digging

Although it is known that Devils will dig to get at buried carcasses, to clear an obstruction to their shelter and to escape an enclosure, it is not clear to what extent Devils will dig to get live prey or to construct their own shelters.

Those who keep Devils in outdoor enclosures have to be sure their walls and fences are secure, either by embedding them in concrete (Guiler, 1971) or burying them at least 800m deep (Newsletter, June 2010; see also Slater, 1993).

There is no description of how Devils dig for any purpose. They have been observed to dig with the front feet (Koncza and Pryor, 2001), and the fact that the front claws tend to be more worn that the back claws (Green, 1967) suggests that when they dig, they rely more heavily, perhaps exclusively, on their front feet.

Climbing

Devils are primarily ground dwelling, but they will climb, the young being especially adept (Roberts, 1915; Guiler, 1970a, 1971, 1992: fig. 14; Jones, 1994, 1996; Jones and Barmuta, 2000; E. Kirchner, in Owen and Pemberton, 2005; N. Mooney, pers. comm. and S. and R. Gale, unpublished data, in Pemberton et al., 2008). They will climb shrubs, small trees with trunk diameters ≥ 10 cm and large trees with trunk diameters > 40 cm and leaning from the vertical 30 - 40º or more. The highest they have been recorded climbing is 7 m (Jones and Barmuta, 2000).

In captivity, Devils will juveniles and sub-adults will “shimmy up a vertical star picket (Edwards, 2006).” Video footage of a young Devil climbing in captivity shows that it, initially at least, use both rear feet simultaneous to saltate up a wooden post (STDP website: “Young devil displays gnarly climbing technique,” first published 3 May 2011).

An assertion that Devils are “active rock climber[s]” (Finch and Freedman, 1988) is without foundation.

Devils are good at crossing road grids and “… use their tails and chins like a fifth and sixth leg” (D. Pemberton, on Professionals Australia website: Here’s a devil of an engineering challenge, 27 July 2012).

Swimming

Devils can swim. If given the choice, however, they appear to prefer to cross creeks and streams via dams (Guiler, 1978) or logs (Jones, 1994) instead of swimming. They

6/05/2016 8:41 AM 27

swim well, but they seem only to enter water when frightened. One fled across an “icy” river when released from a trap (Fleay, 1946, 1952). And another swam “powerfully across [a] fast-flowing, 50-metre wide … [river]” (N. Mooney, in Owen and Pemberton, 2005).

Locomotion

Devils have two different gaits, one when walking and another when running. The Devil’s slow walking gait has been described as being “lop-sided, shambling” (Bauer, 1961) and a “shuffle” (Guiler, 1964).

The Devil’s running gait has been described as “cantering” (Fleay, 1952); a rocking horse motion (Guiler, 1992, also 1983a) and a “rocking amble (Mooney, 1992).” It leaves a distinctive track (Guiler, 1983b: fig. on p. 119, 1992: fig. 45; Owen and Pemberton, 2005, p. 32; Nationalgeographicstock.com, picture ids 1081295 and 1101192; sunsetholidayvillas.com, little angels 049; Bob Brown’s Tasmania Catalogue online, image 42), consisting of two approximately side by side prints and two single prints one well separated from the other. The paired prints and the one front print closest to them represent the two rear feet and one front foot while the single more remote print represents the other front foot.

The sequence of footfalls in this gait can also be determined from still (DPIW, 2007a: p. 11; STDP website: “Status,” 3 August 2011 (these two images are the same but printed different side down); Sydney Morning Herald, Good Weekend, 26 May 2012, p. 3; davewattsphoto.blogspot.com, Tasmanian Devil, 065.2; Proceedings of the Royal Society, B, vol. 282, issue 1814: cover) and moving images (Parer and Parer- Cook, 2004). The isolated front foot can be called the leading foot because at the beginning of the foot fall cycle it is put forward, lands and then is passed over by the other, trailing, front foot and the two rear feet brought up more or less simultaneously.

It is difficult to determine, however, if there is a bias in the lead front foot as it cannot be guaranteed that the images were all printed in the aspect in which they were taken; there is one case (above) in which this was clearly not the case. It can be determined from the images, however, that the lead front foot has no correlation with the most forward (slightly) placed rear foot.

There may also be an asymmetry in the force exerted by the rear feet to judge from the depth to which each foot penetrated on sand (Guiler, 1992: fig. 45), which in this one case was on the opposite side of the leading foot.

This gate has been asserted to be “energy efficient (Brown, 2006),” but there was no supporting data or documentation.

There may also be a third “gate.” There are images of Devils with hind legs side by side on the ground, front feet side by side and either pointing upward and forward, in line with the body or tucked under the chest (Brüniche-Olsen and Austin. 2014: fig.). These Devils look as if they are pouncing in the former case or bounding in the latter.

Juveniles are said to be more agile than adults and jump whereas adults cannot (DPIPWE, 2010d).

6/05/2016 8:41 AM 28

Devil’s can stand up-right on their rear legs (Owen and Pemberton, 2005: plate), presumably, the better to see or smell (Edwards, 2006).

Home Range

Devils do not have territories, but they do have home ranges (Guiler, 1970a and c, 1971, 1978, 1992; Buchmann and Guiler, 1977; M. Jones, in Jones et al., 2004a). In northeastern Tasmanian, the home ranges vary in size from 4.0 to 26.7 km2, with a mean of about 13 km2 (N = 9; Pemberton, 1990). A popular article’s rounding off this estimate to “in the order of 10 to 20 square kilometres” (Jones, 1994) could be potentially misleading.

The surmise that a Devil will return to use a latrine after a week or more “implies a form of territoriality (Owen and Pemberton, 2005” is unsupported by other observations but is in accord with over-lapping home ranges, which is well documented.

Likewise, the narrator’s statement in an STDP-supported documentary on the Devil that “the biggest and most experienced males dominate the best places (Devil Island, 2013e)” has no published support.

Male home ranges are larger than female home ranges on average (M. Jones, in Jones et al., 2004a).

Home ranges both within and between sexes can overlap broadly (Guiler, 1970a, 1992; Pemberton, 1990; Hamede et al., 2009). As an indication of the degree of overlap, one experienced field observer noted that “as many as ten different animals could be caught in the same trap over a short period of trapping (Guiler, 1992).”

Once a Devil has established a home range at maturity, it is said to rarely shift its boundaries appreciably or move to another (Jones et al., 2012).

Typically, Devils emerge from their dens after dark and use familiar tracks and forest edges to move around their home ranges. These movements are punctuated by periods of stillness, lasting up to half an hour. Towards the end of the night, Devils move steadily for their home den, generally arriving before dawn (Pemberton, 1990).

Shelter Dens

Devils usually use a den for daytime shelter, mating and rearing young. It is unclear, however, if dens used for these purposes differ from one another and if so, how. In general, all that is known of dens is the location of their entrances. There are some data on the size of the openings of dens in general (Jones, 1998), but these have not been analysed.

Shelter dens may be situated under grass sags (Owen and Pemberton, 2005), Lomandra tussocks and coral fern (Pemberton, 1990); in or under hollow logs (Guiler, 1960; Bauer, 1961; Bevilacqua, 2005f), and in cavities in rocks, small caves or burrows (Jones et al., 1997), including those of another species, such as wombats

6/05/2016 8:41 AM 29

(Green, 1967; Guiler, 1992; STDP, 2011b). Devils may also shelter in the foundations of buildings (Green, 1967; Jones, 1995; STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011). In one case, three Devils were found asleep inside the body cavity of a dead cow they were eating (Guiler, 1992).

Within a coastal area that included coastal scrub bounded by tea tree forest, eucalypt woodland and buttongrass moorland, the tea tree forest was said to be ideal denning habitat for Devils (N. Mooney, in Pemberton et al., 2008).

Shelter dens may have a strong and unpleasant odour (Bauer, 1961), but it is not clear what causes the odour, old food scraps, faeces or the residue of natural secretions.

A Devil may use between one and ten dens (mean = 3.8, mode = 3) within its home range (Pemberton, 1990: table 6.2; Guiler, 1992; Mooney, 1992; Owen and Pemberton, 2005; DPIPWE, 2010d), which they move between every one to three days (Pemberton, 1990). A den may be used for as many as 25 successive days (Pemberton, 1990). The dens probably help define the Devil’s home range and are used by the same individual as long as it occupies its home range.

There is no significant difference between the number of dens used by male and female Devils (Mann-Whitney U = 12.5, P = 0.20; pers. obs. on data in Pemberton 1990: table 6.2).

Good shelter dens may be in short supply, raising the question of whether they are ever contested among Devils themselves or with other burrow-sheltering species. There are reports of Devils sharing their den with at least one other species, the Common Wombat (Pemberton, 1990; Guiler, 1992), but supporting observations are lacking.

Given the high value of a good shelter den, it is easy to imagine they pass from one Devil to another (Owen and Pemberton, 2005) for virtually as long as the dens last.

Devils’ ranges often overlap broadly, but it is not known if different Devils use the same shelter dens at the same time or even at different times. One experienced field observer said that “more than one devil will use the same den.” But whether this means at the same or different times is unclear. And in any event, there were no supporting observations (Guiler, 1992; see also Jones, 1995). A proximity data-logger study of individual contacts among 12 male and 15 female Devils between February and June (inclusive) in one year showed that almost no contact longer than about 40 minutes, as would be expected if animals were sheltering together routinely during the daylight hours (Hamede et al., 2009: fig. S1).

In captivity, adult Devils will sometimes consistently sleep together, often in physical contact, quite peacefully (Bauer, 1961; E. Kirchner, in Owen and Pemberton, 2005). But whether this is a reflection of wild behaviour or an adaptation to captivity is unclear.

It is not clear to what extent Devils dig their own shelters. In the wild, they will dig to re-gain their den under the foundations of a house, if humans attempt to block the

6/05/2016 8:41 AM 30 entrance with stones and rocks (D. Sadler, in Owen and Pemberton, 2005). In captivity, they dig “shallow burrows” (Fleay, 1935; see also D. Schaap, in Newsletter, March 2011), a behaviour that can begin in Devils as young as about 30 weeks (Kelly, 1993). And one captive female dug a den under concrete slabs in which to rear her young (Oglesby, 1978). The front claws are usually more worn that the rear claws, indicating greater use (Greer, 1967), possibly due partly to digging with the front limbs.

It is also not clear to what extent Devils furnish their dens. At one site in northeastern Tasmania, dens on the surface under thick vegetation, contained no bedding material (Pemberton, 1990). Dens in caves, however, have been found with bedding material (Pemberton, 1990). Devils may furnish their retreats under houses with articles of clothing and bedding taken from their upstairs neighbours (N. Mooney, in Owen and Pemberton, 2005). And young Devils raised in a house will carry loose items of clothing into their shelter (STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).

In captivity, Devils will furnish their shelter sites with softer material as available. One observer whose experience was primarily with captive animals said the both males and females carry straw into their retreats (Fleay, 1935) and that both males and females make “comfortable beds” out of bark, button grass and leaves (Fleay, 1952). Two females, which used the same shelter site, would carry the den lining materials with them when they changed sites (E. Kirchner, in Owen and Pemberton, 2005). One female lined her newly dug den with grass and straw (Oglesby, 1978), and another female carried straw into a shelter site used by her and her nearly weaned young (Roberts, 1915). And, most convincingly, an adult of undetermined sex in a large, “free-range” enclosure was videoed biting off a large clump of tussock grass near its base and carrying it into a nearby den (STDP Website: “Survivor: Desert Island,” 20 March 2010).

On the other hand, an observer, whose experience was with both wild and captive animals, said “neither grass nor vegetation is carried in to make a bed” (Guiler, 1992). But to judge from the observations mentioned above, this unsupported observation seems either too categorical or even just plain wrong.

In captivity, Devils have been observed to not defecate or urinate in their “nest box (Packer, 1966).” Presumably the same would apply in the wild.

Considering adult Devils have relatively few predators, it is likely that the use of a shelter den during the day has more to do with their physiology than protection (Pemberton, 1990). The equitable temperature and humidity parameters of a shelter burrow, at least, might provide energy savings associated with temperature and water regulation.

Mating Dens

There are relatively few observations on mating dens in the wild, and there are no observations about how the mating dens may be chosen, if they are chosen at all. It is unclear, for example, if a mating den is either the male or female’s primary den, or if

6/05/2016 8:41 AM 31

it just any den in which a male and female happen to shelter den on any day in the mating period?

There are only two observations of mating dens in the wild (Pemberton, 1990), and one of these was described simply as being “an underground mating den (Owen and Pemberton, 2005).”

There are also only two observations from the wild about the time a pair spends in a mating den and both involved the same male with two different females in two different dens. In the first instance, the pair stayed in the den for eight days and in the second they stayed together for five days (Pemberton, 1990; Owen and Pemberton, 2005).

In captivity, males will keep females in dens during the mating period and defend them against intruders (Roberts, 1915; Fleay, 1935). This behaviour can begin suddenly and is usually associated with a reversal of the usual dominance roles between males and females.

In an account of one pair in which the female was usually dominant, with the onset of the mating season, the male became dominant and kept the female confined in a den for “ten or twelve days.” If the female attempted to leave the den, the male would drag her back in by the ear or “any other part.” During this time, the male would bring the female food. After the brief mating period, the female became dominant once again (Roberts, 1915). In another account, a male kept a female in his den for ten days (D. Schaap, in Macey, 2008).

One Devil researcher said that when in a mating pair, neither individual feeds (Jones, 1994). But the source of this information and whether it derives from observation of wild or captive animals is unclear.

Rearing Dens

In the wild, females use their “primary,” or most-used, shelter den as the rearing den for their young. However, they may also continue to use other shelter dens; that is, they may not return each night to the rearing den, especially as the end of weaning approaches (Pemberton, 1990). The only explicit location of a rearing den in a natural situation was in “fragile hollow log complex (Mooney, 2014).”

In captivity, female Devils may use an existing shelter den as a rearing den, or they dig a new one shortly after mating (Oglesby, 1978).

In the wild, females will defend their rearing dens from large predators such as humans. One female, for example, rushed a human who was growling speculatively at its entrance (Mooney, 2005; STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011). Rearing dens (note plural) may also have “narrow cub-sized tunnels” that might protect young Devils from marauding adult Devils (Jones 1995).

Little seems to be known of the location of rearing dens in the wild. One experienced field observer reports that female Devils with unweaned young “covet” dens “in rock

6/05/2016 8:41 AM 32

substrate” (Jones, 2008). Another experienced observer is reported as saying that maternal dens may be clustered together if there is limited soil suitable for burrows (D. Pemberton, in DPIPWE, 2010d and in STDP, 2011b). This suggests that rearing dens, and therefore a female’s primary shelter den (above), may be determined by substrate type, which in turn may be a limiting factor for population size.

Another experienced observer noted two rearing dens under the same house, separated by part of the foundation and with the entrances only 25 m or so apart (STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).

In April 2012, a student began research on the characteristics of dens in general and rearing (maternal) dens in particular (UTAS media release, 19 April 2012; also STDP website, “Scholarship recipient to demystify devil dens,” 19 April 2012), but the results of this research have not been published.

In the introduced population on Badger Island, “different breeding females may be caught outside the same den site … (N. Mooney, unpubl.)” (Hawkins et al., 2008). It is not clear, however, exactly what “breeding females” means in this context: mature females in the reproductive season; in oestrus; with pouch young, or young in the den?

In captivity, and no doubt the wild, females furnish their rearing dens with a variety of plant materials such as bark, grass, leaf litter, sticks, straw, twigs and even small logs (!; out of desperation?) (Roberts, 1915; Oglesby, 1978; Kelly, 1993; Smith, 1993; Parer and Parer-Cook, 2004). The females carry the materials into the den with the mouth and manipulate it with the paws (Roberts, 1915; Kelly, 1993). Furnishing may occur anytime between shortly after mating (Oglesby, 1978; Kelly, 1993; Parer and Parer-Cook, 2004) and when the pouch young are three and a half to four months old (Smith, 1993; see also Roberts, 1915).

The observations of “breeding females” being caught outside the same den and of rearing dens being “clustered together” (above) suggest that females with denned young are tolerant of one another. Indeed, some brief observations of captive females sharing rearing duties suggest that cooperation may even occur. Two females with pouch young shared the same den and occasionally appeared to suckle each other’s young. Later, they certainly carried each other’s young on their backs. They would, however, vocalise and fight in the presence of their young (Smith, 1993).

There is little information on how far a female with young in a den may go to forage or how long she may be away. But a lactating female was trapped twice in two days over a distance of 17.7 km (straight-line) (Guiler, 1970a). This observation suggests that some females may spend many hours away from their young and travel long distances to forage.

Young Devils appear to rarely, if ever, venture outside the rearing den until shortly before they are ready to abandon it entirely. One experienced observer has seen Devils as small as 400 g “more than 100 m away from known dens” and “they usually run quickly back to the den (STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).” But another experienced

6/05/2016 8:41 AM 33

observer stated that the “… young are never observed outside the dens until near weaning (Pemberton, 1990).”

Video camera footage of an apparently wild Devil was interpreted by the original observer to show “… a female covering the opening to her den as she set out for a night’s foraging. She likely had joeys in there that she was trying to protect (Abernathy, 2016).” If this interpretation is correct, it is the first observation of such behaviour, and calls into question protocols from identifying rearing dens.

A poster available on the STDP Website (“Life cycle of the Tasmanian Devil,” undated) states that in the den the mother teaches her young what to eat and “about dangers.” It also says that when the young “hear a noise, they will hide in crevices in the den, keeping very quiet, until the mother calls to let them know it is safe.” Although these assertions are plausible, there is nothing in the literature to support them.

Hibernation

There is no evidence that Devils hibernate, not even those living at the highest elevations (Jones et al., 1997; Jones and Barmuta, 1998). Indeed, as scavengers, Devils are able to take advantage of harsh conditions that might lead to the deaths of less robust species.

Densities

The density of Devils varies according to habitat. In general, densities are said to be lowest in wet rain forests and sedgelands and in cool alpine moorlands, intermediate in wet sclerophyll habitats and higher in dry sclerophyll habitats (Pemberton, 1990).

Devil densities are considered to be “low” between 0 and 0.4 per km2; “intermediate” between 0.5 and 0.9; “high” between 1 and 1.9, and very high” ≥ 2 “very high” (Hamede et al., 2008). A density of about 0.5 per km2 seems to be the single figure used in most general comparisons (Jones et al., 2004a).

In “unmodified habitat,” typical densities are thought to have been relatively low, about 0.3-0.7/km2 (M. Jones, unpublished, cited in DPIPWE, 2010d). In the absence of DFTD, in moist forest and highland herblands densities are ≤ 2 animals/km2 (Jones et al., 2007a); in the northwest they vary from 0.8-2.9/ km2 (Hamede et al., 2008) to 3.3-4/ km2 (DPIPWE, Annual Report for 2010), and in lowland open woodland adjacent to grazing lands they could be as high as 12/km2 (Pemberton, 1990; see also Guiler, 1970a; Lachish et al., 2007; Mooney, 1992). An experienced field researcher’s early report of “over 100” Devils having “been taken from about 100 acres” (Guiler, 1964) on the “west coast” equates to about 247/km2, which seems extraordinarily high.

On some pastoral properties with low intensity stock management, Devil densities may exceed 4/km2 (N. Mooney, in DPIPWE, 2010d). And on some sheep properties, which generate a supply of dead stock, densities can be as high as 8.8/km2 (Buchmann and Guiler, 1977; see also data in Guiler, 1983b). Before they were removed, animals in the introduced population on Badger Island reached a density of 8.6/km2, but this

6/05/2016 8:41 AM 34

was due to food supplementation in the form of wallaby culling and sheep over- stocking (Hawkins et al., 2008; see also DPIPWE, 2010d; STDP, 2011b).

Devil “activity,” possibly a surrogate for density, can vary widely within habitats within essentially the same period. In the DFTD-free northwest, Devil activity as measured by the number of “hair traps” that captured hairs varied more than sixfold in “coastal scrub” in northwestern Tasmania (Hollings et al., 2015b: fig. 2). Interestingly, however, the daily activity scores summed within each of the four different habits – coastal scrub, dry eucalypt forest, mature wet eucalypt forest and wet eucalypt forest – were similar (Hollings, 2013: fig. 3.5; the lack of accompanying dispersion statistics precluded reader statistical analysis of the results).

Devils are usually surveyed by catching them in traps baited with meat. Experience shows that in any one area, the number of animals captured peaks in about three or four nights and then falls away. Hence, estimates of densities based on trapping results need to take this into account. Trapping also shows that even in “Devil country,” individual traps may never catch anything (Guiler, 1978).

Local populations can fluctuate over time in ways that are often difficult to understand (Guiler, 1978, 1982, 1983b; Hollings et al., 2014: fig. 2, Devil “absent”). At one locality, a population decline seemed to be associated with weight loss in the animals (Guiler, 1983b), but what caused this loss is unclear.

Population Fluctuations

Most populations fluctuate over time due to a variety of reasons and Devils are no exception.

Prehistoric fluctuations

Studies using ancient DNA and advanced inferential statistics suggest that Devils have gone through two declines prior to European settlement. The first decline was approximately 29-48k ybp, a period roughly coincident with the last glacial maximum, and the second was approximately 2.4-3.9k ybp. The aridity of both periods may have caused the Devil population(s) to contract (Brüniche-Olsen et al., 2014).

Historic fluctuations

It is widely believed that the Devil population in Tasmania has undergone strong fluctuations between the time of European settlement and the present (Jones et al., 2004a; Bradshaw and Brook, 2005; Siddle et al., 2007a-b; Cheng et al., 2010; Morris et al., 2013). One early Devil researcher asserted “Tasmanian Devil populations suffer collapses from time to time and may not recover for many years. The cause of these crashes is not known (Guiler, 1992).”

The shape of the Devil’s total population curve since European settlement, however, is unclear. Notions of “collapses,” “crashes” and “recovery” imply strongly there were two or more population cycles, most usually interpreted as, a decline, a recovery, another decline, another recovery and then the current decline due to the

6/05/2016 8:41 AM 35

Devil’s facial tumour (see Bradshaw and Brook, 2005: fig. 1). The evidence, however, suggests that the curve may not be so neat. Most critically, there is little evidence of the first recovery. If this is the case, then the population curve would consist of an initial decline, a low plateau followed or not by a second further decline, recovery to an unprecedented high and then a decline to present day numbers.

Early historic decline

Ecologists estimate that at the time of European settlement, there were about 45 000 Devils in Tasmania (M. Jones unpublished data, in Jones et al., 2004a). The earliest account of the Devil notes, “these animals were very common on our first settling in Hobart Town [but] as settlement increased, and the ground became cleared, they were driven from their haunts near the town to the deeper recesses of forests yet unexplored (Harris, 1808).” This brief assessment could very well describe the fundamental dynamic shaping the Devil population up until the mid-20th century. In the first quarter of the 19th century, Devils still appeared to be common in newly settled areas (L.A. Meredith, 1853 – “shepherds caught in a pitfall, during one winter, no less than a hundred and forty-three devils” (emphasis added); see also Meredith, 1880; L.A. Meredith, in Walch, 1890 and Guiler, 1992, who places the event in 1825).

Devil numbers appear to have declined noticeably, however, by the middle of the 19th century (Meredith, 1853 – “they … are gradually becoming rare in all the occupied parts of the island;” The Argus, 20 October 1862 – “an exceedingly rare animal” and “only … outdone in rarity by the [Thylacine];” Gould, 1863 - “it has now become so scarce in all the cultivated districts that it is rarely, if ever, seen in a state of nature;” Caledonian Mercury, 5 February 1866 – “now almost extinct,” but from a possibly self-serving commercial advertisement; Hobart Mercury, 20 April 1866 – “now almost extinct;” South Australian Register, 30 April 1867 – “these animals are becoming very rare.” There were, however, a few dissenting opinions (Krefft, 1868 - “still plentiful in some of the wild districts).”

First recovery or continuing low numbers?

Whether Devil numbers increased during the latter half of the 19th century or stayed relatively low is the key question in determining if the Devil population cycled in historic times. The notion of an increase rests almost entirely on the influential, but unsupported, assertion of an early Devil expert that “Devils were scarce around 1850 but had built up again by the 1890s (Guiler, 1992).”

The few published accounts from this time that bear on the issue of Devil numbers, however, clearly support scarcity, at least in settled areas (Launceston Examiner, 30 September 1879 – “devils are not very common even in the back settlements;” Meredith, 1880 – “much scarcer than formerly;” Hobart Mercury, 26 April 1884 – “becoming scarce on the depasturing section of Woolnorth; ” Hobart Mercury, 10 September 1888 – “the curator of the Tasmanian Museum wrote [to the Hamilton Municipal Council] expressing his desire to obtain … Tasmanian devils, and offering to pay expenses by coach or rail;” Hobart Mercury, 3 April 1896 – “wanted … four Tasmanian devils, alive. State price. Apply Curator of Tasmanian Museum;” Launceston Examiner, 6 April 1897 – “correspondence was read from the curator of the Hobart Museum and Art Gallery, requesting the cooperation of the [Deloraine

6/05/2016 8:41 AM 36

Municipal] Council in collecting specimens of the … Tasmanian devil;” New York Times, 22 January 1899 – “ the devil is now very scarce and will soon be extinct”). There is nothing to suggest other than a scarcity of Devils, let alone an increase in Devil numbers, in the latter half of the 1800s. The advertisements from the two museums may be especially indicative, as they would have presumably had established contacts they could have called upon before resorting to public advertisements.

Continuing low – or going even lower?

Beginning in the early 1900s, a steady stream of comment bears on the number of Devils. In most cases, the comments indicate scarcity (H. Scott, in letter of April 1904, referred to in Paddle, 2012; W. McGowan, in letter of November 1906 quoted in Paddle, 2012 – ‘ “very difficult to obtain” ’); Lucas and Le Souef, 1909 – “has been driven back to the very roughest country;” Flynn, 1911 – “increased scarcity;” Lord, 1919a – “now only met with in the rugged unsettled districts;” Lord, 1919b – “now met with only in the most unfrequented localities; Taylor, 1920 – “decreasing rapidly;” Hobart Mercury, 3 November 1920 – ‘ “even scarcer than the “tiger” ’; Mackenzie, 1921 – “almost extinct;” Flynn, 1922 – “rare;” Hobart Mercury, 27 July 1922, 14 August 1925 – “now practically extinct,” 9 September 1926 – “numbers have diminished considerably within the last six or seven years,” 31 December 1926 – “exceedingly scarce;” Bernie Advocate, 20 December 1928; Carter, 1933; Hobart Mercury, 29 November 1935 – “now nearly extinct,” 3 June 1936 – “extremely rare,” 24 February 1937 – has “not been seen in the district [Perth, NE Tasmania] for more than 30 years;” Flynn, 1939 – “very rare and indeed verging upon extinction” and “now to be found only in the unsettled mountainous parts of Tasmania from which it is rapidly disappearing;” N.J.B. Plombley, Director, Launceston Museum, as reported in the Examiner, 17 May 1946 – “it was rarely that the devil frequented inhabited areas, being found mostly in rugged mountainous country and Tasmania’s west;” see also, Guiler, 1961,1964, 1970c, 1972, 1978, 1982; Atkinson, 2001).

Three accounts from the first half of the 20th century, however, indicate Devil numbers were high, at least in some parts of the state. Lord and Scott (1924) wrote, “Although it has been driven from the more settled districts, yet it is not by any means in such danger of extermination as is the Thylacine…. It is content to dwell in the rocky and almost inaccessible portions of the bush, and, considering the extent of this class of country in Tasmania, we are of the opinion that it is not in immediate danger of extermination. …. We are aware that certain Zoologists are of opinion that the species is in danger of extinction, and we will readily admit that it is difficult to secure perfect specimens when required, but that is readily explained by the class of country which the species now inhabits.” The comment “now inhabits” suggests that the Devil had disappeared from more settled areas. Two years later, Le Souef and Burrell (1926) wrote that the Devil is “… rather numerous and,” and in 1928, Lord wrote, “in the rougher sections of the country this species exists in fair numbers;” see also Owen and Pemberton, 2005).

What is uncertain about the low numbers in the early 1900s is whether they were simply a continuation of the lows of latter two-thirds of the 1800s or whether they were a further step down. The case for continuing low numbers is that the remarks about relative scarcity in the early 1900s are rarely in the context of a noticeable

6/05/2016 8:41 AM 37

recent decline. There is only one report of “increased scarcity (Flynn, 1911),” but the time period involved is unstated.

The claim of an early 1900s decline was based, again, on an influential but poorly supported comment by an early Devil researcher: “It was suggested that about 1908 all the Dasyures … were affected by a disease which decimated their numbers…. The Devils certainly became scarce due to some cause or causes (Guiler, 1964; also 1992).” More recently, the disease hypothesis was resurrected to account, at least partly, for the pre-extinction decline of the Thylacine as well as the Devil (Paddle, 2012).

If disease did reduce the Devil population in the early 1900s, it probably just further reduced a population that had been kept suppressed by anti-Devil practices and attitudes (“vermin”) prevalent throughout the 1800s and indeed up until the 1950s. There is no evidence that an early 1900s disease, if it was that, reduced a “recovered” or “recovering” population.

It is worth noting in passing that one Devil researcher’s implied identification of a population “high” between the early 1900s and the 1940s (Jones, 1995 - “populations in Tasmania have fluctuated widely in the recent past, being rare in the first decade of the twentieth century and gain in the 1940s …”) is without foundation.

“Modern” recovery

Devil numbers clearly began increasing in the 1950s (Hobart Mercury, 26 May 1951 – “once almost as rare [as the Tasmanian Tiger], is increasing in numbers;” Burnie Advocate, 26 May 1951 – “not uncommon in North-Eastern districts and the Lake country,” also the Launceston Examiner of the same date); Guiler, 1960 – “very common in many parts; Guiler, 1964 – “certainly on the upgrade of a cycle … and now is known from areas where it has not been seen in living memory…. now have recovered very fully;” Anonymous, 1966 – “we meet the Devil quite frequently in Tasmania” and “common among wooded ranges, in parts of lowland scrub, and about the fringe of farms;” Guiler, 1970a – “not until 1960 that field notes showed evidence of their widespread distribution” and “nowadays, there are few areas of scrub where the species does not occur;” Guiler, 1970c – “began to reappear in numbers about 1955 and by 1960, in some districts, they were plentiful;” Guiler, 1978 – “the population throughout Tasmania has increased steadily since 1958 so that they now are numerous in many places where they were unknown 20 years ago;” Guiler, 1992 – “in 1950 a Tasmanian devil was brought into the Zoology Department at the University of Tasmania and everyone rushed to see this unusual animal;” Guiler, 2001 – “by 1950 … numbers had built up …grant permits for their capture;” see also Green, 1967; Guiler, 1972, 1992; G.J. Hocking pers. comm.., in Jones et al., 2004a; Bradshaw and Brook, 2005: fig. 1) and remained high until the advent of Devil Facial Tumour Disease in the mid-1990s.

Some researchers have suggested that the number of Devils may have been at its highest ever just prior to the onset of Devil facial tumour disease (Green, 1967; Guiler, 1970c, 1982, 1992; Mooney, 1992), with estimates as high as 150 000 (see Population Decline, below). However, the most recent estimate suggests a population

6/05/2016 8:41 AM 38

of about 60 000 (Beeton, 2012, as cited in Hollings, 2013). Recall that at the time of European settlement, Devil numbers were estimated to be about 45 000.

This modern recovery was due to several factors. In 1941, Devils were protected by law from being taken or killed. In the early 1950s, the use of strychnine to control rabbits was ended (Stratham, 2006; McCallum and Jones, 2006); Devils would no longer die from scavenging strychnine-killed rabbits. People were also beginning to take pride in Australian flora and fauna. And the ecological importance of predators and scavengers was being appreciated. Furthermore, once protected, the Devil population could increase in size rapidly due to the absence of a once important competitor and predator, the Thylacine (Mooney, 1992; N. Mooney, in Bevilacqua, 2003a; Hawkins et al., 2006; N. Mooney, in Circular Head Chronicle, 4 October 2013); the large number of carcasses generated by grazing (Guiler, 1970a, 1978, 1992; Mooney, 1992) farming and road traffic, and the mountains of garbage created by an increasingly affluent society.

Modern decline

The recent rapid decline of Devil numbers with the advent of the species’ facial tumour disease is discussed below. Here it only needs to be noted that the decline has reduced the Devil’s population from perhaps as many as 150 000 to perhaps about 30 000, that is, even less than at the time of European settlement.

Cause of the early 1900s “decline”

The decline in Devil numbers in the early 1900s, if it was that, has been attributed to disease (Bauer, 1961; Guiler, 1964, 1983a, 1992; Green, 1973; Mooney, 1992; Jones, 1995; Paddle, 2012). But what the disease may have been is unclear. A first guess would be that it was something that occurred benignly in an introduced species, domestic or wild, and then jumped to naïve native species with devastating effect. The disease seems to have been specific to dasyuromorphians (Bauer, 1961; Guiler, 1964, 1970c; Paddle, 2012) so another facet of the mystery is why only this carnivorous (in Tasmania) group of marsupials were susceptible.

Toxoplasmosis, a disease caused by a protozoan parasite, has been mentioned in relation to some historic declines in Australian wildlife. Toxoplasmosis does occur in Devils (Hollings et al., 2013; see also Munday, 1966), but it is not clear if has a debilitating effect.

A distemper-like disease has been mentioned specifically with regard to the Devil’s decline (Guiler, 1970c; Mooney, 1992; Jones, 1995; K. Morris, on Ninemsn, 5 December 2012; Morris et al., 2013). This idea seems to go back in a comment by the same Devil expert who suggested Devil populations “suffer crashes from time to time” and who supported the idea of a crash in the early 1900s. He wrote, "Several men have stated to me that the decline was very rapid and occurred almost simultaneously throughout Tasmania and one grazier stated that all the Dasyures disappeared about 1910, claiming that a disease like distemper killed them (Guiler, 1961; see also Guiler, 1970c)."

6/05/2016 8:41 AM 39

The best chance of identifying the disease would be to apply modern forensic techniques to a preserved specimen with the disease, if one could be found (Morris et al., 2013). In the early 1900s, museums were reluctant to preserve “atypical” or “unhealthy” specimens, so the chances of this are slim. Parenthetically, one measure of how much museum practices may or may not have changed in the last 100 years would be the number of DFTD-affected Devils now in Tasmanian museums.

Observing Devils

When assessing the reality or extent of broad population cycles in Devils, it worth bearing in mind that Devils are both hard to detect by casual observation and the numbers in local populations can change rapidly.

Devils can be hard to see casually at the best of times. For example, in four years, an enthusiastic naturalist encountered only 11 Devils in the bush (Bauer, 1961). Two other field researchers had similar experiences on their study sites just before the arrival of the facial tumour disease. One worker was reported as having seen, while on foot, only a dozen Devils in three years on a site that had some of the highest densities of Devils at the time (Owen and Pemberton, 2005). The other was reported as having seen only “half a dozen” Devils on her 200 km2 study site in 12 years (Grzelewski, 2002). The only reliable way to census Devils, therefore, is through a standard monitoring protocol, such as bounty payments (in an earlier day), regular trapping, spot-lighting and counting roadkills.

Furthermore, standard monitoring protocols show that Devils can disappear quickly from an area (Guiler, 1992; see also “Densities”).

Summary

A conservative interpretation of population “fluctuations” in the Devil population since European settlement based on published accounts leads to a quite simple pattern. As European settlement spread through Tasmania beginning in the early 1800s, the Devil population, through cumulative local declines, began falling from an estimated 45 000 animals. Settler animosity drove the local declines and ensured there was no recovery. Local declines were not limited to the actual cleared areas but extended into the local bush through the activity of trappers.

As settlement spread, the general Devil population declined. Local populations remained suppressed until attitudes toward the Devil (and other wildlife) began to change in the mid-1900s, at which point the population may have been at its all time low. The population size at this time has never been estimated, but it possibly could be, by applying appropriate density estimates to areas not then settled. As the, arguably, most astute pre-modern Devil observer probably correctly noted, “in the rougher sections of the country this species exists in fair numbers (Lord, 1928).”

From the mid-1990s, the population increased rapidly to an historic high of perhaps 150 000 until the advent of facial tumour disease in the mid-1990s. The population then began its second serious decline, to possibly 30 000 today, and presumably continuing.

6/05/2016 8:41 AM 40

Prehistoric fluctuations

In 2014, geneticists published an analysis of 10 microsatellite loci from nine recent sites using sophisticated statistical methods which indicated that the Devil population had undergone two declines in the distant past, one 22-48 k ybp and the other 2-4 k ybp (Brüniche-Olsen et al., 2014; Brüniche-Olsen and Austin, 2014). The former period was during the last glacial maximum when Tasmania was still connected to the mainland and Devils were widespread. The latter period was several thousand years after Tasmania had been separated from the mainland by the flooding of Bass Strait and Devils may or may not have still been on the mainland (see “Distribution”).

What was the size of the Devil population just prior to and after each decline? The authors speak of an overall 83 percent decline relative to the current population, but it is not clear to which decline this refers. The percentage decline suggests the earlier population(s) were nearly six times larger than the population at the time the microsatellite samples were taken, mostly 2004-2005 (N = 306) but some from 1964 (N = 24). DFTD had reduced the Devil population by no more than 50 percent in 2004-2005, so at this time it may have consisted of 75 000 animals. Hence, at one or both of the ancient pre-decline highs the population may have consisted of about 450 000 animals. This is a plausible figure for the first pre-decline population, because it would have presumably included the mainland population. But is it a plausible figure for the second pre-decline population if the only Devil population was the one on isolated Tasmania?

Diet

The Devil’s diet in the wild is known primarily from analysis of its stomach contents (Green, 1967) and scats (Jones and Barmuta, 1998). These analyses along with other observations suggest the Devil is mostly carnivorous. Indeed, it is often touted as the world’s largest living marsupial carnivore (Pemberton and Renouf, 1993).

Stomach contents are more useful than scats in determining the Devil’s diet, because the food items are less processed than they are by the time they are formed into scats. Only three of the several studies bearing on the diet in the Devil’s in the wild, however, examined stomach contents, about 137 stomachs in total (Green, 1967 - 30; Guiler, 1970a – 98, and Pemberton et al., 2008 - 9). This may sound like a lot, but the analysis of more stomachs would allow a better assessment of the diet of just weaned Devils, the frequency of seemingly “under-represented” groups of animals (below) and perhaps even more importantly the frequency of cats, an exotic predator which some think may be suppressed by Devils (see “Re-introduction to the Mainland,” below).

Road-killed Devils provide an excellent opportunity to get more information about the species’ stomach contents. Thousands of Devils are killed on Tasmania’s roads each year and the state government has trained staff and volunteers to take biological samples from some of the reported kills. But the training apparently not include the taking of stomachs. Between February 2001 and June 2010, the Tasmanian authorities received reports of 898 road-killed Devils (Lawrence and Donnelly, 2011). Of these, 128 had been reported by individuals with enough interest and training to have

6/05/2016 8:41 AM 41

collected the stomachs. Had all these stomachs been preserved and analysed our knowledge of the Devil’s diet from this important source would have nearly doubled.

The remains in scats that are most useful for inferring the Devil’s diet are the generally difficult-to-digest keratin parts of its food such as claws, feathers, fur, scales and teeth (Fleay, 1952; Guiler, 1964). Teeth are usually readily identifiable as to species. The scale and fur can usually be identified to species or species group using microscopic features. Feathers, however, are more problematic. Remains unidentifiable visually can now also be assessed through DNA analysis.

Most of the material in a scat, by weight, is bone. In a collection of 50 dried scats, nearly half the weight (47 percent) consisted of 3868 bone fragments. These had an average length of 11.7 mm and a maximum length of 34 mm (Marshall and Cosgrove, 1990; see also Lundelius, 1966 and Hall and Jones, 1990). One researcher even found the forearm of a wallaby in a scat (Jones, 1994). The high bone (calcium) content of the scats often gives them a whitish cast (Guiler, 1992: fig. 26; Jones, 1994).

In the wild, Devils eat a variety of animals: invertebrates, fish, lizards, snakes, birds, antechinuses, cattle, horses, pademelons, possums, quolls, rats, wallabies and wombats (Table 3).

Despite the variety of items eaten, some interesting disparities remain between the items eaten and the abundance of potential items in the wild. Frogs are mentioned in only one early popular account (Fleay, 1952), but frogs would be common along water bodies night when Devils are foraging. Reptiles also figure only rarely in lists of dietary items, and although mostly diurnal, many species such as skinks are common and could be readily found in their nocturnal shelters. Among mammals, no house mice have been listed as a Devil dietary item, despite the small rodent being widespread and common.

One ambiguity in the dietary list for Devils is worth comment. The Lowland Copperhead, Austrelaps superbus, has been listed specifically as a food item in a technical account (Guiler, 1970a, as Denisonia superba), but the popular account cited as the reference mentions only “snakes (Guiler, 1964);” perhaps this is a subsequent more specific identification by the same author.

Table 3. List of animal food items found in the stomachs or scats of Tasmanian Devils taken from the wild or animals observed being eaten in the wild. An asterisk indicates species the healthy independent individuals of which are probably too large to be predated by Devils and were most likely to have been scavenged. Fly maggots are assumed to have been ingested with carrion and are omitted (e.g., Taylor, 1986). Within marsupials and placentals, the species are arranged in increasing order of maximum mass.

Group Scientific Name Common Name Reference Invertebrates Insect Fleay, 1952; Guiler, 1960, 1964; Pemberton et al., Oncopera sp. Moth (larva) Guiler, 1970a,c; Owen and Pemberton, 2005 Coleoptera Beetle Taylor, 1986 Scarabaeidae Cockchafer Mooney, 2012

6/05/2016 8:41 AM 42

B. pachydermatina Marine ascidian Guiler, 1970a,c, 1992 Marine crab Green, 1967 Kelp maggot N. Mooney, in Owen and Pemberton, 2005 Fish Fish Green, 1967; Guiler, 1992; Mooney, 1992; Pembe 2008 Frogs Frog Fleay, 1952; Guiler, 1970c Tadpole Owen and Pemberton, 2005 Reptiles Guiler, 1970c C. casuarinae She-oak Skink Fleay, 1952 Lizard Guiler, 1960, 1964 Skink Jones, 1995 N. scutatus Tiger Snake Fleay, 1952; Mooney, 1992 A. superbus Lowland Copperhead Guiler, 1964 Snake Fleay, 1952; Guiler, 1964; Jones, 1995 Birds Guiler, 1960, 1970c, 1992; Green, 1967; Taylor, 1 1992; Jones and Barmuta, 1998, 2000; Pemberton STDP website: “Maria Island Devil Translocation Update, December 2014,” 23 December 2014; STD Report, 2013/14; STDP, 2015a. Eggs STDP website: “Maria Island Devil Transloca Project Update, December 2014,” 23 Decembe STDP Annual Report, 2013/14; STDP, 2015a Parrots Le Souef and Burrell, 1926 Coturnix sp. Quail D. Mann and B. Bauer, pers. comms in Pemberton 2008; also Le Souef and Burrell, 1926 N. novaeseelandiae Boobook Owl Mooney, 1992 E. minor Little Penguin Shine, 2014b; Jones and McCallum, 2011a (assert 2015a G. domesticus Chicken Fleay, 1952; Guiler, 1964, 1970a P. tenuirostris Short-tailed Shearwater Jones and McCallum, 2011a (asserted); STDP, 20 C. coronoides Australian Raven Guiler, 1970a Monotremes O. anatinus Platypus Pemberton et al., 2008 P. aculeatus Echidna Guiler, 1970a; Mooney, 1992; Jones, 1994; Pembe 2008; STDP website: “Maria Island Devil Translo Project Update, December 2014,” 23 December 20 Mooney and Huxtable and Bayes, pers. comms in McCallum, 2011a; STDP Annual Report, 2013/14 2015a Marsupials C. nanus Eastern Pigmy Possum Guiler, 1970a Antechinus sp. Antechinus Taylor, 1986 A. swainsonii Dusky Antechinus Jones, 1994 P. breviceps Sugar Glider Jones, 1994; Jones and Barmuta, 2000 P. peregrinus Ring-tail Possum Guiler, 1960, 1964, 1970a; Taylor, 1986; Marshall Cosgrove, 1990; Jones, 1994; Jones and Barmuta, Pemberton et al., 2008; STDP website: “Maria Isla Translocation Project Update, December 2014,” 23 2014; STDP Annual Report, 2013/14; STDP, 2015 Perameles “Bandicoot” Guiler, 1970a; STDP Annual Report, 2013/14; ST P. gunnii Eastern Barred STDP website: “Maria Island Devil Translocation Bandicoot Update, December 2014,” 23 December 2014 P. tridactylus Long-nosed Potoroo Guiler, 1970a

6/05/2016 8:41 AM 43

B. gaimardi Eastern Bettong STDP website: “Maria Island Devil Translocation Update, December 2014,” 23 December 2014; STD Report, 2013/14; STDP, 2015a. D. viverrinus Eastern Quoll Guiler, 1970a (as D. quoll); Jones, 1995; Jones and 1998 D. maculatus Spotted-tail Quoll Jones, 1995; Jones and Barmuta, 1998 Dasyurus spp. Quolls Jones et al., 2014 T. vulpecula Brush-tail Possum Guiler, 1960, 1964, 1970a; Green, 1967; Jones, 19 Jones and Barmuta, 1998, 2000; Pemberton et al., website: “Maria Island Devil Translocation Projec December 2014,” 23 December 2014; STDP Annu 2013/14; STDP, 2015a. S. harrisii Tasmanian Devil Green, 1967; Guiler, 1970a; Jones, 1995; Jones an 1998; Mooney, 2012; Jones et al., 2014. T. billardieri Tasmanian Pademelon Fleay, 1952; Guiler, 1960, 1964; Taylor, 1986; Ma Cosgrove, 1990; Jones, 1994; Pemberton et al., 20 website: “Maria Island Devil Translocation Projec December 2014,” 23 December 2014; STDP Annu 2013/14; STDP, 2015a. M. rufogriseus Bennett’s Wallaby? Guiler, 1960, 1970a, 1992; Green, 1967; Jones, 19 Pemberton et al., 2008; STDP website: “Maria Isla Translocation Project Update, December 2014,” 23 2014; STDP Annual Report, 2013/14; STDP, 2015 V. ursinus Common Wombat* Fleay, 1952; Guiler, 1964, 1992; Marshall and Cos 1990; Jones, 1994, 1995, 1996; Pemberton et al., 2 website: “Maria Island Devil Translocation Projec December 2014,” 23 December 2014; STDP Annu 2013/14; STDP, 2015a. M. giganteus Forester Kangaroo* STDP website: “Maria Island Devil Translocation Update, December 2014,” 23 December 2014; STD Report, 2013/14; STDP, 2015a. Placentals P. higginsi Long-tailed Mouse Taylor, 1986; Jones, 1992 “Rats” Le Souef and Burrell, 1926 “Native rat” Guiler, 1964 R. lutreolus Swamp Rat Guiler, 1970a; Taylor, 1986 R. rattus Black Rat Guiler, 1964; Pemberton et al., 2008 R. norvegicus Norway Rat Pemberton et al., 2008 O. caniculus Rabbit Guiler, 1960, 1964, 1970a; Pemberton et al., 2008 F. catus Cat Guiler, 1964, 1970a C. familiaris Dog* Guiler, 1970a O. aries Sheep* Green, 1967, 1992 Arctocephalus sp. Fur Seal* Pemberton et al., 2008 Equus caballus Horse* Guiler, 1970a, 1992; Owen and Pemberton, 2005; et al., 2008 Bos taurus Cow* Guiler, 1964, 1970a, 1992; Owen and Pemberton, Pemberton et al., 2008 Cetacea Whale* Pemberton et al., 2008; Mooney, 2012

In captivity, Devils will eat natural foods such as marine invertebrates, insects, fish, frogs, birds, bird eggs, macropods, mice, monotremes (echidnas), possums, guinea pigs, rabbits and rats as well as meat from cattle, macropods and sheep (Fleay, 1952; Hartz, 1966; Guiler, 1971; Weiss and Richards, 1971; Moeller, 1972b-c; Collins,

6/05/2016 8:41 AM 44

1973; Sallis et al., 1973; Sallis and Guiler, 1977; Oglesby, 1978; Green and Eberhard, 1979; Williams, 1981; Finnie and Thomas, 1982; Sobbe, 1990; Kelly, 1993; Slater, 1993; Smith, 1993; Jones et al., 2005; Albion, 2006; Edwards, 2006; Hesterman et al., 2008a; Kreiss et al., 2008; Hesterman and Jones, 2009; Keeley et al., 2012d). They will also eat manufactured food such as dog biscuits and pellets (Guiler, 1971; Moeller, 1972b; Sobbe, 1990) and mink food (Hartz, 1966).

Although a trivial point, the assertion that adult Devils are “obligate flesh-eaters” as opposed to invertebrate eaters (Jones et al., 2014) has never been tested by experiment. The variety of invertebrates in the diet of wild Devils and the manufactured food in the diet of captive Devils raises the possibility Devils could get by on a diet of non-fleshy foods. One field worker noted that all the Devil scats found in the Tarkine for a couple of weeks were full of cockchafers (Mooney, 2012).

Although Devils eat almost all of a carcass, observations on captive Devils show that they usually avoid fecal matter in the intestine, by pushing it out with the feet (Edwards, 2006).

Devils devour most of their food, leaving little behind. They eat bone as well as keratin in all its forms: scales, feathers, fur, claws and hooves. About the only parts they do not eat are heavy bones such as wombat skulls (Guiler, 1964; Jones, 1994) and cow long bones, pelvis and skull (Guiler, 1964). Given the Devil’s proven ability to crush and devour all but the largest bones of the largest carcasses within its domain, it is no surprise that, based on an indirect measure involving skull dimensions and jaw muscle cross-sectional area, Devils have the strongest bite force relative to size of any living carnivore (Wroe et al., 2005; Attard et al., 2011; see also Macalister, 1872; Lord, 1928 and Gill, 1953).

Devils can “badly damage” traps made of 10 gauge wire, which is 2.38 mm in diameter, whereas cages made of 6.35 mm thick steel mesh restrains “most” animals (Guiler, 1970a). Working in the reverse direction, Devils will chew through traps made of chain link fencing material to get at small animals inside (D. Randall, in Owen and Pemberton, 2005).

The importance of vegetable matter in the Devil’s diet is unclear. In the wild, one early field researcher noted that “a surprising amount of vegetable matter may be found in some samples. This is not an indication of starvation of privation, as such animals are in good to excellent condition with much fat on the intestines.” He went on to suggest, “the vegetable matter may be eaten as an aid to digestion (Guiler, 1964; see also Jones, 1994).” Subsequently, however, he seemed to change his mind, writing, “Devils do not eat vegetable matter, any grass found in their stomachs probably has been ingested accidentally … (Guiler, 1992).” In captivity, Devils eat apples, carrots and celery (Kelly, 1993). These conflicting observations make it difficult to determine the actual importance of vegetable matter in the diet, but they seem to open the possibility that although generally thought of as a carnivore, Devils might be more omnivorous than given credit for.

Devils rarely reject any form of flesh. Some captive animals, however, have apparently refused certain kinds of meat. One female refused to eat meat from either cattle or horse (Fleay, 1935); two young animals, a male and female, left two

6/05/2016 8:41 AM 45

bandicoots largely untouched (Douglas et al., 1966); two young animals of unstated sex would not eat rats (Collins, 1973), and some Devils, specifically identified as being from Southport, would not eat chicken (Albion, 2006).

Scavengers or Predators?

The analysis of stomach contents and scats has revealed most of what we know about what Devils eat, but it does not tell us how the animals obtained those food items, whether by scavenging or by hunting (predation). Even the most experienced field workers seem uncertain as to the balance between the two methods of foraging. Indeed, some of the most knowledgeable field workers seem uncertain about the balance.

One experienced field worker called Devils “primarily scavengers (Jones, 1994) but a year later wrote it was “almost certain” the two Tasmanian quolls and the Devil were “primarily predators (Jones, 1995).”

Another worker said in 1990, “There is no evidence from this study to show that devils can predate on live prey,” and “if predation does form a significant part of a devil’s diet I suggest … this would only occur in the absence of the available carcasses of large animals, such as macropods (Pemberton, 1990).” Eighteen years later, however, he and his co-authors would write, “When our results are compared and combined with those of other studies that on the prey of Tasmanian Devils, it emerges that the habits of the Tasmanian Devil include both scavenging and direct predation foraging strategies that target both mammals and birds. The prey species that are available for direct predation occur both on the ground and in trees (Pemberton et al., 2008).

Another field worker stating “Devil facts” in a popular article wrote, “All manner of animal matter eaten, usually fresh and most scavenged (Mooney, 1992). In another popular article he wrote, in an inset summarising the Devil’s biology, “Nocturnal scavenger on animal carcasses…. Also an opportunistic hunter of vertebrates and invertebrates (Mooney, 2004).”

The earliest experienced observer clearly believed Devils are primarily scavengers (Guiler, 1970a, 1983a, 1992). He once wrote, “Thus, they appear to be versatile scavengers rather than hunters (Buchmann and Guiler, 1977).” Later, he wrote, “… these … little animals are too small, light and slow to catch and kill large animals. As a result they act as scavengers …: and food larger than themselves must be dead or disabled before it can provide sustenance (Guiler, 1983b).”

The STDP website states that “The Tasmanian Devil is mainly a scavenger and feeds on whatever is available” (STDP website, “About Tasmanian Devils,” published 3 August 2011).

Three other lines of evidence also suggest the Devil is a scavenger first and an opportunistic predator second. First, most solitary predators show stalking behaviour in that they have to get close enough to their prey to be able to catch it in a quick final attack. There is no evidence from either the wild or captivity that Devils stalk potential prey. They is also no evidence that Devils lie in ambush.

6/05/2016 8:41 AM 46

Second, Devil stomach contents usually consist of pieces of large animals, not a large number of small animals (Taylor, 1986), as is the case with animals that are predators first and scavengers second, such as cats (Taylor, 1986).

Third, Devils seem to eat all manner of inanimate objects emitting potential food odours (below). Eating such objects is more suggestive of the habits of a scavenger than a predator, again, such as a cat.

Devils as Scavengers

As scavengers, wild Devils eat items that vary from still quivering to partially decayed, as evident from the maggots found in stomachs (Guiler, 1970a, 1992). In captivity, Devils will eat carcasses that have begun to stink (Ewer, 1969).

Devils have almost certainly benefited from European agriculture through the introduction of exotic stock animals (Pemberton, 1990; Guiler, 1992), which often die under the harsh Tasmanian conditions. They have also benefited from the native animals shot and poisoned as “pests.” And they have done well out of the rubbish tips established around towns and homesteads (Guiler, 1992). The carcasses arising from vehicular traffic, however, would have been a mixed blessing for Devils, as these new feeding fields also became their new killing fields.

Devils will also ingest pieces of manufactured materials, both inorganic and organic in origin (Table 4). such as cartridge wadding, gloves, rubber and leather boots, pieces of worn clothing, pieces of plastic, rubber thongs, running shoes, scourers, silver foil, worn socks and tea towels (Jones, 1994; Mooney, 2004). Clothing made from animal skins, such as boots and gloves may contain enough residual smell of the original owner to be attractive, and articles made of non-animal materials may pick up carrion- like smells (smelly sandshoes). In any event, to the extent that these items are purposely eaten, they show that Devils are generalist scavengers.

Table 4. List of manufactured objects injested by Devils.

Item Reference Possible Attractant Odour Cartridge wadding Guiler, 1970a ? Cigarette butt Mooney, 1992 Residual food Silver foil Mooney, 1992; Jones, 1994 Residual food Styrofoam Mooney, 1992 Residual food Tea towel Jones, 1994 Residual food Plastic pieces Guiler, 1970a, 1992; Residual food Mooney, 1992 Stock ear tags Mooney, 1992 Residual wearer’s odour Docking ring, lamb’s Mooney, 1992 Residual wearer’s odour Scourer, metal Mooney, 1992 Residual food Collar, dog or cat Mooney, 1992 Residual wearer’s scent Boots, rubber Guiler, 1970a Sweat Boots, leather Jones, 1994; Mooney, 1992; Source’s odour; wearer’s sweat ABC, World Today, 26 Nov. 2013 Running shoes Jones, 1994 Sweat

6/05/2016 8:41 AM 47

Thongs, rubber Jones, 1994 Sweat Socks, worn Mooney, 1992 Sweat Gloves Guiler, 1970a Sweat Clothing, worn Guiler, 1970a; Mooney, 1992 Sweat; residual oils or fats

Devils will not attack a sleeping human, but they will eat a corpse (Guiler, 1992). If a person were to die while bushwalking in Tasmania, the family might not have much left to bury. Devils will also tunnel down to dead animals that have been buried (G. Sutton, in Owen and Pemberton, 2005), suggesting colonial bush burials may have been problematic.

Scavenging Behaviour

Devils seem to rely heavily on a sense of smell, because when an animal travels, it “noses the ground almost continuously in its slow meandering gait” (Green, 1967). Indeed, Devils are said to be able to smell food up to a kilometre away (Owen and Pemberton, 2005). A farmer who provides a “devils dining” wildlife experience drags roadkill several kilometres to his “restaurant” made the interesting observation in passing that “the devils will know which way it [the dragged roadkill] is going (G. King, on ABC Landline, 29 June 2003; see also Edwards, 2006).” This suggests that they may be able to detect a subtle odour gradient.

Devils may also rely on their sense of hearing, as they seem to be attracted not only to the vocalisations of other Devils around a carcass but also to the distress calls of animals (Mooney, 1992).

The assertion that “if a Devil finds a large meal or one too difficult to butcher alone they announce the fact” by vocalising (Devil Island, 2013b) is implausible on theoretical grounds, despite it apparently being signed off on by an expert adviser.

The importance of sight is controversial. One field observer thought sight was important in “socialising, navigation and when hunting small, active animals (Mooney, 1992). But a keeper said that Devils’ “eyesight goes to about four or five feet (1.2-1.5 m)” (G. Irons, in Cart, 2011). An experienced field worker thought that the “Devil’s sight is defective in light (Guiler, 1960).” Perhaps vision, at night at least, is important for close-in work but not over great distances.

One experienced field observer notes that the Devil’s food is “mostly scavenged” and its foraging method seems to be “almost ceaseless patrolling … [and] favourite feeding areas are carefully examined then the animal moves more rapidly to the next likely area (Mooney, 1992).”

Devils will also dig up buried food, which they almost certainly detect by smell. For example, wild Devils will dig up meat baits buried 15 cm below the surface (Hughes et al., 2011); have dug through sand to feed on a buried horse (Owen and Pemberton, 2005), and will dig under the baited end of traps (Jones, 1995). Captive Devils will dig up food that is buried in their enclosures to diversify their activities (Owen and Pemberton, 2005). This ability to dig up buried flesh has raised a concern that Devils may dig up poisoned meat baits set (buried) for fox control (The Mercury, 9 July 2010). This concern, however, may be misplaced (below).

6/05/2016 8:41 AM 48

Devils are capable of eating almost all the body parts of the largest terrestrial animals they find dead. The jaws and teeth are powerful enough to crush most of the bone of the largest animals. Only the skulls of wombats (Guiler, 1964; Jones, 1994) and the skull, long bones and pelvises of cattle (Guiler, 1964) are left unbroken (see also Mooney, 1992).

Crushing the bones of other animals is not without its risks, however, as Devils sometimes cut themselves on the jagged bones (Buchmann and Guiler, 1977).

Devils usually devour a carcass in its entirety quickly. For example, a group of Devils can devour a large carcass, such as that of a wallaby, overnight (Green, 1967). On the other hand, one researcher wrote, “they will return to a cow skeleton for months after it was killed” (Guiler, 1992).

Devils are sometimes called “specialised scavengers” (Hawkins et al., 2008; Jones, 2008) in that many of their morphological adaptations allow them to eat all but the most resistant remains of large animals, such as thick bones.

Early bushmen reported that “one or more” Devils would follow a Thylacine on its hunting trips and eat the remains of its only partially eaten kills (Lord, 1928; Fleay, 1935). No doubt, Devils finished such kills, but whether the Devils actually followed Thylacines with this “in mind” seems problematic.

Devils as Predators

In the wild, Devils will attack warm-blooded prey such as chickens (Roberts, 1915), ducks (Roberts, 1915; Fleay, 1935), young rabbits (Guiler, 1992), cats (A.D. Fergusson, in Fleay, 1952), new-born lambs (Guiler, 1992), platypuses (N. Mooney, in Munday et al., 1998), sick sheep (Guiler, 1992) and young dogs on a chain (Green, 1967). None of these species, except for possibly the cat and dog depending on their size, would be able to offer much resistance.

Devils will also catch and kill “cold-blooded” vertebrate prey such as lizards and snakes (Fleay, 1952; Guiler, 1992).

They also forage for invertebrates such as colonial ascidians (Guiler 1970a), kelp maggots (N. Mooney, in Owen and Pemberton, 2005) and moth larvae such as “corbie grubs” (Oncopera spp.) (Guiler, 1983a). In eating insects, the Devil is not only Australia’s largest carnivore, but also its largest insectivore (C. Dickman, in Darcé, 2010).

The assertion that Devils are “effective predators of medium-large prey such as wallabies” (Jones, 2008) are unsupported by observations in the literature (below).

The assertion that Devils “… sometimes hunt for small mammals and birds” (Woods et al., 2015) is unsupported in the cited reference (Hawkins et al., 2006).

The assertion that Devils have been recorded “killing” spiders, possums and wombats (Cheng et al., 2015) is unsupported in the published literature.

6/05/2016 8:41 AM 49

Prey Capture

There are some general statements in the literature of how Devils capture and kill prey, but some of these are contradictory and few are supported by first hand observations.

Some of the generalisations about how Devils kill prey are quite categorical. For example, “Devils kill by attaching themselves to the head or chest region of the prey, adjusting their bite and hanging on until the animal succumbs” (Jones, 1996).

And some of the generalisations are contradictory. For example, one observer has said that Devils “… kill prey using a crushing bite to the skull, nape, or chest, depending on the relative sizes of the predator and prey, in which the canines penetrate the skull (smaller prey) or the nape or chest (larger prey …)…(Jones, 2003a, p. 292). But the same observer in the same article also said “… their strong, crushing, but generally non-penetrating, killing bite to the chest, head or nose of the prey … (Jones, 2003a, p. 292) (italics added).

The details of prey capture and killing have been published for captive individuals. In one series of observations, when offered a live rat, a young male usually attacked the head or neck, killing the prey with a single bite (Ewer, 1969). But in series of observations involving several individuals, no animal showed a propensity to attack any particular part of the prey, although the final, “definitive” bite was invariably delivered to the head (Buchmann and Guiler, 1975).

When fed live prey, such as small mammals, in captivity, Devils are often hesitant to attack the prey and may even avoid it. When they do attack, they often seem inept and uncertain at first, with only some individuals becoming more proficient with practice (Ewer, 1969; Moeller, 1974, as cited in Buchmann and Guiler, 1977; Buchmann and Guiler, 1976).

Prior to biting, Devils may stab or grasp the prey with a front foot (Buchmann and Guiler, 1975).

In captivity, a litter of four young with no previous experience in killing, attacked, killed and ate a bandicoot that accidentally entered their enclosure. The prey was initially immobilised by a bite into the backbone and then killed with repeated bites by all the young Devils (Buchmann and Guiler, 1977).

A mature male in captivity tried, unsuccessfully, two methods of catching c. 20 cm long fish placed live in its 63 x 52 x 8 (deep) cm water basin. One method involved standing on the edge of the basin and scooping with one front foot and the other involved standing on the edge of the basin or in the water and grasping with both front feet (Moeller, 1972a).

Predatory Behaviour

Although Devils are considered partly predatory, there are relatively few observations of Devils actually hunting, catching and killing prey in the wild.

6/05/2016 8:41 AM 50

Despite this, some researchers frequently use the word “prey” for Devils’ food when, in fact, the origin of that food is unknown, because it has been found in scats or stomachs (e.g., Jones, 1997; Jones and Barmuta, 1998, 2000; Jones et al., 2014). This lax usage may contribute to the belief that Devils are more predaceous than they really are.

In captivity, there are three reports of Devils eating chicken eggs but two do not make it clear if the eggs were offered whole or broken (Fleay, 1935; Kelly, 1993). The third report notes that Devils dexterously manipulate the eggs with the front feet until thy gain a secure hold and then break them with the rear teeth; then they lap up contents leaving the shells behind (Moeller, 1972c). The smaller eggs of Blackbirds are eaten whole (Moeller, 1972c). Devils immediately recognise whole emu eggs as food and can get their mouths around them, but it may take them a day to learn to break the shell (Williams, 1981). If Devils can eat bird eggs as large as those of emus in captivity, they should be able to eat virtually any bird egg they encounter in the wild.

Some of their smaller prey species, such as moths and tadpoles (Own and Pemberton, 2005), are unlikely to be found either as carrion or in or around carrion and therefore are likely to have been predated instead of scavenged. There are also specific observations of Devils foraging for kelp maggots along the strand line (N. Mooney in Owen and Pemberton, 2005). One hand-reared young female devil would catch and kill mice. This same animal also found a den with young rabbits, dug into it and ate the young (Pemberton, 1990). Presumably, therefore, completely wild Devils would do the same.

Devils can kill animals of intermediate size, that is, animals about their own size. In the wild, Devils will attack, apparently, cats in houses (A.D. Fergusson, in Fleay, 1952) and young dogs on chains (Green, 1967). In captivity, Devils will also attack animals that happen to enter or are introduced into their enclosures (Table 5). This does not mean, however, they could catch all these species if they were free-ranging. The largest animal killed by a Devil in an enclosure appears to be a Muscovy Duck, which can reach a weight of 8 kg (but weight of individual actually killed unknown). Interestingly, a mature male Devil was able to catch but not kill a c. 6 kg rabbit released in its enclosure (Moeller, 1974).

Table 5. List of live animals Devils recorded as having caught and killed in captivity (free ranging in a confined space). Species listed in approximate increasing order of size.

Species Reference Insects Moeller, 1974, as cited in Buchmann and Guiler, 1977 Locusts Moeller, 1972a-b (wings clipped prior to feeding) Mice Ewer, 1968; Moeller, 1972b Chicks Eisenberg and Leyhausen, 1972 Frogs Moeller, 1972a-b Rats Ewer, 1968, 1969; Moeller, 1972a-b; Buchmann and Guiler, 1977 Birds Moeller, 1974, as cited in Buchmann and Guiler, 1977 Blackbird Moeller, 1972b (wing feathers clipped) Pigeons Moeller,1972a-b (wing feathers clipped) Small mammals Moeller, 1974, as cited in Buchmann and Guiler, 1977

6/05/2016 8:41 AM 51

Guinea Pig Moeller, 1972a Rabbits Moeller, 1972a-b Tamarin Monkey (juvenile) Hobart Mercury, 3 January 2001 Southern Short-nosed Bandicoot Buchmann and Guiler, 1977 Chickens Many Muscovy Duck Fleay, 1935, 1952

Devils may also be able to catch the young of larger animals, either in the open or in nests and dens. Again, however, there are no first hand observations. One experienced field observer once wrote, “It is recorded that a young devil dug out a rabbit burrow and ate the young (Guiler, 1992),” but this sounds like a second-hand account. It has been said that Devils “… will kill juveniles [of wombats] …” (Jones, 2003a), but there are no published details.

The Devil’s climbing ability (above) has led to the supposition that they may climb to catch birds (Guiler, 1970a, 1983) and perhaps mostly arboreal mammals such a Ring- tail Possums (Green, 1967). Again, however, there are no direct observations of bird or possum predation. Contrariwise, there are observations of Devils visiting a feeding station at a tourist facility, searching for bits of meat among the offerings but showing no predatory interest in a group of possums that had also been attracted to the station (Jones, 1994). In fact, one researcher believes adult Devils may climb simply to gain a vantage point (Jones, 1994).

One experienced field worker has made numerous statements about Devils hunting and killing large prey, but all the assertions are either unsupported, unpublished or second-hand. Here is a sampling. “… there have also been numerous observations of Devils perhaps opportunistically chasing, grabbing onto and in a few cases killing animals ranging in size from possums to wombats (Jones, 1994, italics added).” “A number of accounts, however, suggest that at least chasing macropods and wombats is not uncommon (Pemberton, 1990; pers. comm.., Joe’s uncle Marrawah 1990, Vicki Shilvock 1991, Leon Barmuta 1992, Dave Watts 1993, Klaus Toft 1993 (Jones, 1995).” The Devil is a “… predator of large terrestrial prey …” (Jones and Barmuta, 2000). Its prey is “primarily wombats, wallabies, pademelons and possums … (Jones, 2003a, italics added; see also Jones, 2008). “Devils are known to attack adult wombats (30 kg) and to kill juveniles (10 kg) but their success rate is unknown (Jones, unpubl. data) (Jones, 2003a).”

Among the second-hand reports, six have been cited in support of the statement that “… chasing macropods and wombats is not uncommon (Jones, 1995). Five of these are personal communications and the sixth is a reference to a publication that only contains three personal communications with respect to the pursuit of pademelons.

More specifically among the second-hand reports, four deal with Devils and pademelons (P. Naughton, M. Mooney and N. Brothers, pers. comms in Pemberton, 1990; A. Osborne, pers. comm. in Jones, 1995), a species that may be as large as 12 kg or about 15 percent larger that the largest Devil at 10.4 kg. Three accounts involved a chase and one a discovery. Three involved a bite. Only one led to a kill (A. Osborne, pers. comm. in Jones, 1995).

6/05/2016 8:41 AM 52

There is one second-hand report of a Devil chasing an adult wombat. This report involved the Devil biting the snout leading to a “tug-of-war”, but the outcome was not observed (D. Probert, pers. comm. in Jones, 1995).

There are, to be sure, observations of Devils eating possibly killing still-living macropods caught in snares, but, to reiterate, there are very few published first-hand observations of Devils pursuing, let alone actually catching and killing animals larger than themselves when free ranging.

In summary, therefore, there is only one specific observation of a Devil attacking and killing a wild animal potentially larger than itself (a pademelon of unstated size and health). Perhaps the most conservative interpretation of these observations is that Devils, with some unknown frequency, chase animals slightly larger than themselves in order to reveal any disability that might make an attack profitable. Indeed, the Devil’s apparent stamina over long distances may be especially well suited to the steady pursuit of the infirm. But catching and killing a sick or captured animal is closer to scavenging than predation.

Considering their importance for an understanding of Devil predation on animals as large or larger than themselves, it is unfortunate none of the “pers. comms” on which they are largely based have been described in the literature. The local natural history magazine, the Tasmanian Field Naturalist, would have been the perfect outlet.

There are also numerous statements in the literature about how Devils pursue and kill prey larger than themselves, but none are first-hand accounts.

One statement says, “Devils have been seen to hunt pademelons by persistently following them at a steady lope …” (Jones, 1995).”

When Devils were released on Maria Island, the Tasmanian Minister for Parks and Wildlife said that some of the Devils had been seen “hunting together” to bring down a possum or wallaby (B. Wightman, on ABC Hobart, 21 October 2013). There are two remarkable aspects to this observation. First, there are totally unsubstantiated reports of Devils hunting in a group (Owen and Pemberton, 2005; Pemberton, 1990), but this is the first such observation attributed to a source. Moreover, the observation relates to two very different species and hence potentially to two independent observations. Second, the outcome is ambiguous but crucial. Did the Devils actually capture the prey, or did they only chase it? The difference could be between a scavenger testing the capacity of a sick or injured animal and a predator actually pursing a healthy animal. The observer, therefore, should publish the details. Until then, there is nothing to credit in this observation.

Several authors have stated or suggested that predation in Devils is a combination of ambush and pursuit: “They appear to capture prey either by ambush or persistence, wearing the prey down” (Jones, 1994); “…hunting … is probably a mixture of ambush and persistent pursuit (Jones et al., 2004a);” “hunting strategy involves …ambushing, or persistently pursuing, prey” (Jones, 2008), and “… Devils … actively hunt their prey … using a combination of ambush and short, moderate-speed pursuits … (Hawkins et al., 2008)” (see also Wintle, 1886; Jones and Barmuta, 2000; Jones, 2003a; Jones, 2015a).

6/05/2016 8:41 AM 53

Some authors have also described the predatory habits of Devils as “pounce and pursuit” (Jones, 1995; 2003a). Pounce-pursuit predators “… use a moving search culminating in a pounce or chase and rarely grapple with prey” (Jones and Stoddart, 1998).

It is not certain what actual observations, if any, prompted these assertions of “ambush,” “persistent pursuit” and “pounce and pursuit.” The first mention of such behaviour goes back to two accounts published in 1909. A British traveller wrote, “being slow and clumsy in its movements it lies in wait for a sheep and then springs at its neck, fastening its hold with bull-dog tenacity (Smith, 1909).” Two Australian resident naturalists wrote, “Being slow in movement …, it captures its prey by a sudden spring …. (Lucas and Le Souef, 1909).” The two accounts are suspiciously similar and both seem conjectural.

A popular account from 1962, said that the Devil is “… tireless in tracking its prey and wearing down a victim in a long pursuit, if not striking it down from ambush beside a game trail leading to grass of water (Sharland, 1962).” This seems like embellished conjecture.

Later, a pioneering field worker wrote, “Le Souef and Burrell’s suggestion (1926) that they [Devils] capture animals by lying in wait and pouncing requires confirmation (Buchmann and Guiler, 1977).” The cited work, however, contains no such suggestion (see above).

The only attributed account of prey killing is the following, “… observations suggest that the method for killing mammals larger than themselves (e.g., wallabies and wombats) involves gaining a grip with the jaws somewhere on the anterior half of the body, adjusting the bite until a vulnerable area such as the chest, neck, or head is gained, and hanging on with a crushing bite, even being dragged, until the prey succumbs (A. Osborne and D. Probert, personal communications)” (Jones, 1997; emphasis in original).

The fact that the Devil has a “relatively simple” retinogeniculate system (the part of the brain onto which the stimuli to the retina project; Sanderson et al., 1979) also suggests that the species may not pursue its prey, because such pursuit would seem to require sophisticated image processing.

Unfortunately, these poorly documented assertions about the Devil as predator acquire a cache of authority and certainty as they are repeated and taken up uncritically by other researchers. One researcher wrote, “… [Devils] are also active predators capable of killing mammals well above their own body weight (Jones and Barmuta 1998; M. Jones pers. comm.)” (Johnson, 2014; emphasis added). But the first reference contains only the unsupported assertion quoted above, and there is no reason to believe the “pers. comm.” has any more support than what this communicator has offered as “evidence” in the past. In another paper, three researchers wrote, “… ecological studies indicate devils are adept hunters … (Jones and Barmuta 1998; Jones 2008)” (Letnic et al., 2014, emphasis added), but the actual “observations” behind this claim do not necessarily support it. Another group of very experienced researchers wrote that medium-sized macropods (Bennett's

6/05/2016 8:41 AM 54

wallaby, Macropus rufogriseus; Tasmanian pademelon, Thylogale billiardierii), and the common brushtail possum (Trichosurus vulpecula) were “major prey species (Hollings et al., 2015b).” One of these researchers had written earlier that Brushtail Possums are “regularly preyed upon,” and are “a common prey item” of Devils (Hollings, 2013). As seen above, however, the evidence for actual predation on any of these species is limited to one pademelon of unknown health status

Devils are sometimes used to infer the biology of similar fossils. The analogy is often based on the belief that Devils are predators. For example, in the interpretation of a mid-Cenozoic thylacinid, Devils were said to be “… opportunistic hunters and are known to prey on mammals that may exceed their body mass Attard et al., 2014),” citing two references for the statement. However, one reference says of Devils, “… there may be killing of weak animals [sheep] (Guiler, 1970a, emphasis added).” The other says that Devils “may … use an element of surprise to kill their prey [Tasmanian Pademelon, Thylogale billardieri] (Taylor, 1986, emphasis added).” Hence, neither reference provides a first hand observation of Devils killing prey heavier than themselves, only conjectures.

Devil as an Apex Predator

The first reference to the Devil as a “top” predator was in 2003 (Jones, 2003; Wood, 2003a: Wroe and Jones, 2003) and as an “apex” predator in 2011 (Hollings et al., 2011a). It is now commonplace to refer to the Tasmanian Devil as a “top” (Hawkins et al., 2006; Jones et al., 2007a; Kreiss et al., 2009a; Jones, 2010, 2013; Jones and McCalllum, 2011b; Brown, 2013; D. Bowman, in O’Neill, 2014; Brüniche-Olsen and Austin. 2014; Hunter et al., 2015; Hunter and Letnic, 2015; Huxtable et al., 2015; Pye et al., 2015c) or “apex” (Ritchie et al., 2012; Hollings, 2013; ARC grant LP140100508, 2014; Hollings et al., 2014; Jones et al., 2014; Rewilding Australia website, accessed 10 January 2015; Hollings et al., 2015a-b; Jones, 2015a) predator in popular, applied and theoretical discussions. As discussed above, however, the importance of the Devil as a predator/scavenger in general is unquantified and as a predator on animals as large as or larger than itself largely without evidence.

The question has to be asked: Is an animal that has been recorded in the wild as catching and killing only one immobile individual of unknown health status of one species only slightly larger than itself, a pademelon, and is recorded in captivity as catching and killing no animal larger than a Muscovy Duck (max. 8 kg) really a top or apex predator?

In this light, it is interesting that a group of five researchers, including two Devil researchers, applied for and were awarded a taxpayer funded grant of $1.36 million over five years to investigate the “keystone effects of Australia’s top predators: dingoes, devils and biodiversity (grant: DP110103069).” The purpose of the grant was to: “… study the interactions of Australia's two largest predators, the dingo and Tasmanian devil, with other species. The project … [would] help develop an understanding of the value of these predators in maintaining ecosystem processes and diversity, and guide their management in the future” (STDP website: “Tasmanian Devil to benefit from ARC research grants,” 25 October 2010; italics added).

6/05/2016 8:41 AM 55

It is interesting to recall that there are at least three other species that would qualify for “Australia’s top predator,” ahead of the Devil: the Carpet Python (15 kg), Amesthystine Python (15 kg) and Salt-water Crocodile (several hundred kgs), all of which are unambiguously predators of prey larger than themselves. Even in Tasmania, the smaller (relative to Devils) Spotted-Tailed Quolls (Dasyurus maculatus) and Tiger Snakes (Notechis scutatus) probably predate more animals than do Devils.

It would be closer to the truth to refer to the Devil as Tasmania’s “top” or “apex” terrestrial scavenger. No doubt, while foraging, Devils test the health status of suspect animals – looking for the living dead and in doing so cause healthy to remain vigilant and cautious (Hollings et al., 2015b). However, whether they kill enough healthy animals to affect population numbers seems unlikely on the available evidence; it is certainly untested.

Devils as Cannibals

Devils can be cannibalistic in the wild (Green, 1967; Guiler, 1970a, 1971, 1983b, 1992; Jones and Barmuta, 1998; Mooney, 2012). Devil scats and stomach contents occasionally contain the remains of other Devils. But as with other food items found in scats or stomachs, there is no way to tell whether the food has been scavenged or predated.

Devils can also be cannibalistic in captivity, although how often cannibalism occurs is unclear as zoos are not wont to let bad news stories out. There is, however, a record of one older Devil eating a younger one in Adelaide Botanical Gardens in 1872 (Launceston Examiner, 5 October 1872). And there is another record of a mother and possibly her other young eating a dead young, even though other food was plentiful, in a zoo in the early 1930s (Fleay, 1935, 1952).

In the wild, newly weaned and independent Devils probably run the greatest risk of being killed and eaten by another Devil (Guiler, 1983b). In communal feeding at carcasses, subadults are more vigilant than adults, even occasionally rearing up on their rear legs to look around, as adults rarely ever do (Jones, 1998).

Amounts Eaten

Devils can eat large amounts in one meal. One researcher recorded a Devil as being 2 kg heavier after having fed overnight on a sheep. Unfortunately, the Devil’s “before” body weight was not published (Guiler, 1970a), although a figure of 25 percent was published later as a general comment by the same author (Guiler, 1992). More recently, researchers found that Devils can eat up to 40 percent of their body weight in one meal (Pemberton and Renouf, 1993). This means an adult female Devil could eat all of an adult Ring-tail Possum or juvenile Brush-tail Possum and an adult male Devil, all of an adult Brush-tail (Jones and Barmuta, 1998). With a “good” carcass, a Devil can consume its maximum capacity in about half an hour (Owen and Pemberton, 2005), and it can eat its fill every two or three days (Owen and Pemberton, 2005).

6/05/2016 8:41 AM 56

An adult Devil might spend about an hour and a half feeding at a carcass and may leave the carcass for a while and then return to feed again (Jones, 1994). After feeding heavily, a Devil may not return to feed for 3-8 days (Pemberton and Renouf, 1992).

When gorged, a Devil’s belly becomes noticeably distended (Jones, 1994), and they may not move far from the carcass to digest a prodigious meal (Owen and Pemberton, 2005). Indeed, on one occasion, Devils were seen emerging from the interior of a carcass as if they were using it for a shelter site.

Devils can make short work of a large carcass. In one instance, they devoured a horse about 160 cm at the shoulder (16 hand) in three days, leaving only some bones and tail hairs (Guiler, 1983b, 1992). They may also return over long periods to gnaw on a carcass. In one case, Devils kept visiting the remains of a cow months after it had died (Guiler, 1992).

In one area of the central highlands, there was large overlap year around in the diets of males and females. In fact, the most noticeable general trend was for larger Devils to include a larger proportion of larger animals in the diet (Jones and Barmuta, 1998).

There was also a seasonal shift in diet. In summer, the greatest proportion of the diet of both sexes was large mammals in summer but medium-sized mammals in winter (Jones and Barmuta, 1998).

In captivity, keepers may feed an adult Devil a rabbit a day (Bauer, 1961) or a chicken or small rabbit every other day (Oglesby, 1978). In terms of grams, one captive diet was 68 gm of horsemeat daily supplemented with two eggs and 23 gm of fish weekly (Oglesby, 1978); 500 gm of chicken, rabbit of rats spread evenly over a week, about 70 gm/day (Oglesby, 1978); 200-300 g of birds, fish, mice and rats (Slater, 1993) or, variably over six days in the week, 250-800 g portions of kangaroo meat per day (Hughes et al., 2011). Devils about nine months old are fed 320 gm per day for males and 250 gm per day for females (Daily Liberal, 21 January 2015). In one case, four adult males maintained their weight and four adult females slowly gained weight over a six month period when fed six days a weeks with food approximately 10 percent of their body weight (Jones et al., 2005: fig. 1).

Feeding and Food Handling

Observations on feeding and food handling in the wild come largely from observations made at large carcasses. Wild Devils usually enter the fresh carcasses of large animals through the belly (Jones, 1996), where the skin is softest and there is direct access to the soft intestines.

One study of feeding times at carcasses found that the mean time spent feeding continuously was 29.6 ± 4.7 min and discontinuously, 45.0 ± 13.1 min, the difference not being significant, with a combined mean, therefore, of 34 ± 4.5 min (Pemberton and Renouf, 1993). Adult males fed longer than all other Devils, perhaps simply because they needed more time to feed to satiation (Pemberton and Renouf, 1993).

6/05/2016 8:41 AM 57

Although not quantified, one experienced field worker thought that the feeding rate at carcasses increased “substantially” as the group size rose from one to three (Jones, 1998).

Wild Devils apparently eat chunks of meat nearly whole with little chewing, because trappers often find the intact meat baits in Devils’ stomachs as well as the intact feet of smaller animals such as bandicoots and rabbits (Guiler, 1970a).

Devils will drag carrion away from where it is found and eat it elsewhere. Six planted possum carcasses were dragged between 15 and at least 170 m in a zigzag path, with some indication of having been eaten along the route (Jones, 1995). Furthermore, Devils will often run off with pieces they can detach from carcasses (Jones, 1995).

How frequently wild Devils carry carrion/prey or bits of carrion/prey back to their diurnal shelters is unclear. An archaeologist noted that “they have never been shown to bring prey into their lairs (Solomon, 1985) (Hall and Jones, 1990).” But in fact, very little seems to be known of what their active lairs contain.

Much more is known about feeding and prey handling in captivity, but how applicable this may be to life in the wild is unclear. In an earlier day, live prey was sometimes offered as food, and in one case, the observer pointedly recorded that his Devil did not stalk its prey but simply bit out at it when it was close (Ewer, 1969; see also Eisenberg and Leyhausen, 1972). Whether this observation is relevant to wild Devils and the issue of their hunting versus scavenging behaviour is unclear.

A hand-raised and presumably captive female would catch mice and kill them with a neck bite and shaking (Pemberton, 1990).

Captive Devils consume small prey rapidly, devouring a small bird or mammal in its entirety in five to ten minutes (Buchmann and Guiler, 1977).

Devils often use the front feet to hold and manipulate pieces of food (Ewer, 1969; Eisenberg and Leyhausen, 1972; Sanderson et al., 1979: fig. 1; Guiler, 1992: fig. 11; Hartigan, 2013; Devil Island, 2013d). They will use both feet to hold a large piece of meat on the ground while they pull parts of it off with the teeth. They will also hold a smaller piece of meat in one foot to pull off a piece. And they use their front feet to “cram lengths of intestine into the mouth like spaghetti” (Buchmann and Guiler, 1977).

Some observations on captive Devils indicate they start eating small mammalian prey (rats) head first, apparently using the lay of the pelage to indicate where the head end is (Ewer, 1969). But other observations indicate Devils start eating at the part of the prey closest to them, although this is often the head as a Devil’s “final, definitive bite was invariably directed at the skull” (Buchmann and Guiler, 1975).

One captive Devil chewed the head off dead fish before eating the body (E. Kirchner, in Owen and Pemberton, 2005). Another Devil would push off the skin of a freshly killed rat from neck to tail, turning it inside out. It would then eat the flesh but leave the discarded skin (Ewer, 1969). This is the only observation of a Devil not devouring its prey’s skin.

6/05/2016 8:41 AM 58

Fresh blood appears to be highly stimulating. For example, Devils can be induced to start eating at any spot on a carcass where there is fresh blood (Buchmann and Guiler, 1975). Spilled blood is “licked up meticulously (Buchmann and Guiler, 1975; see also Edwards, 2006)”.

In captivity, Devils fight over small food items. Often all the Devils in an enclosure will be pull vigorously on a single chunk of meat, a behaviour that exhibitors take advantage of to offer a spectacle to their visitors. If one Devil gets possession, it will often run to its shelter (Buchmann and Guiler, 1975), closely pursued by the others. Needless to say, one Devil will take food from the mouth of another if it can (Eisenberg and Leyhausen, 1972). Sometimes a Devil will lie on its food to prevent it being taken by another individual (Buchmann and Guiler, 1975).

In captivity, Devils may eat just-killed prey on the spot or carry it to a “usual” feeding spot (Ewer, 1969) or a more secluded part of their enclosure, often one they use for resting (Buchmann and Guiler, 1975). When a food is moved, it is either picked up and carried or dragged, depending on its size (Buchmann and Guiler, 1975).

Caching

There is no evidence that Devils in the wild cache food, even during winter in the central highlands (Jones, 1995; Jones and Barmuta, 1998).

During the first month or so when wild Devils are put into captivity, they eat and drink very little, but they do occasionally cache food in their nest boxes. However, they rarely appear to eat this food (Jones et al., 2005).

Drinking

Devils drink and make a noisy “clop-clopping” sound when they do, at least in captivity (Fleay, 1935, 1952). The young will lap milk from a bowl (Hartigan, 2013).

Gut Passage Times

Fur- and bone-free food takes food about 24 hours to pass through a Devil’s gut (Hesterman et al., 2008a). Food containing fur and bone, however, takes 2 to 3 days to clear the gut (Douglas et al., 1966).

Scats, Defecation and Urination

Most researchers find Devil scats easy to identify (Mooney, in Warner, 2001). But others are not so sure Devil scats can be readily distinguished from quoll scats (Mumbray, 1992). Devil scats are relatively large, averaging about 15 cm in length but occasionally reaching 25 cm (and 3 cm in width), and slightly twisted (Guiler, 1964, 1966; Lundelius, 1966: plate 28 a-c; Mooney, 1992; Triggs, 1996; Owen and Pemberton, 2005; Pemberton et al., 2008) but apparently less so than sympatric native cats, Dasyurus. When fresh scats dry, they “spring open” and appear even larger (Mooney, 1992).

6/05/2016 8:41 AM 59

Opinions differ as to where Devils defecate. Field workers comment that Devils often deposit their faeces along the trails they use (Guiler, 1992; Jones, 1994). Devils also often defecate in the same area, creating so-called “latrines” (E. Guiler, pers. comm. in Marshall and Cosgrove, 1990; Pemberton, 1990; Guiler, 1992: fig. 27; Mooney, 1992; Pemberton et al., 2008; STDP website: “Insurance Devils settle into Freycinet FRE [Free-range Enclosure],” 24 February 2011; Newsletter, 2011). One field study found that latrines occurred mainly along paths or tracks crossing creeks but were also be associated with dams, tracks, firebreaks or even granite slabs (Pemberton, 1990; table 6.8). A latrine is usually less than 15 m2 in area and may vary in shape from roughly linear to square (Pemberton 1990). One latrine may contain as many as 46 scats (Pemberton, 1990: table 6.8). The latrines are used year around (Pemberton, 1990).

It is not clear how many Devils may use a latrine or how often they might visit it. The assertions that “dozens of devils will defecate in one area” and the “same spot will be used by a devil after a week or more (Owen and Pemberton, 1990),” are undocumented in the literature.

There is one observation of a Devil’s lair in the wild containing “droppings,” and “badly stinking (Bauer, 1961).” It seems counter-intuitive that an animal, especially a carnivorous one, would defecate in the shelter site that it returns to repeatedly, as the one just noted apparently did. It is remarkable how little is known about “live” Devil dens in the wild, that is, how they are constructed and furnished and what they eventually come to contain. In this regard, even observations from captivity could be informative.

Devils have also been observed to defecate at sites baited with food and food odours. In this case, defecation may serve as a signal to other Devils or other scavengers (Lazenby and Dickman, 2013).

Devil scats, like the droppings of most mammals, contain odorous compounds indicative of an animal’s sex, age and physiological state, and hence latrines may play a role in social interactions (Pemberton, 1990). Indeed, the apparent attractiveness of latrines is such that researchers use “artificial latrines” to draw animals to particular sites (STDP website: “Populations persist at ‘ground zero,’ ” 1 June 2011).

Some archaeologists, who take an interest in the remains of what Devil have been eating because they can be confused with the remains of what Aboriginals may have been eating, have claimed that Devils do not use latrines but “… tend to defecate at the kill or scavenge site” (Marshall and Cosgrove, 1990). However, they gave no references and proceeded to analyse the remains of scats found in several sandstone rockshelters in central Tasmania. In any event, archaeologists have, in fact, collected scats from a Devil’s lair in a limestone cave (Hall and Jones, 1990).

Apparently, gardeners have spread Devil scats in their gardens in the belief it deters possums. In 2000, the Tasmanian Parks and Wildlife Service was going to trial the efficacy of Devil scats in repelling herbivores, in the hope they could be spread along roadsides to reduce the roadkill (Read, 2000). It is not clear what the outcome of the proposed trials was. A few years later, extracts from fresh Devil scats were found to

6/05/2016 8:41 AM 60

inhibit feeding in captive goats and Eastern Grey Kangaroos, but not as effectively as extracts from Tiger (Cox et al., 2010).

In the wild, urination has been observed in animals approaching a carcass with other Devils already in attendance (Pemberton and Renouf, 1993). In this instance, the behaviour could convey information about the urinating Devil to the other Devils, or it could be simply “load lightening” in the event of being involved in conflict. Urination has also been observed at sites baited with food and food odours. In this instance, the behaviour could convey information to other Devils or other scavengers (Lazenby and Dickman, 2013).

In captivity, urination is associated with activity, specifically running on a wheel, although it was not clear whether the urination occurred before or after a bout of running (Packer, 1966). The biological relevance of this association is not clear.

Competition for Food

Devils are primarily scavengers. Hence, their greatest competitors for food are likely to be other scavengers. And competition is likely to most intense around food large enough to attract a number of individuals before it is completely consumed.

Intraspecific

There is no doubt that Devils’ most serious competitors at carcasses are other Devils. As many videos show, Devils are always noisy and sometimes physically aggressive when feeding at a carcass. Devils can “lose out” at a carcass if they are intimidated or chased away, or if they are distracted from feeding by being vigilant of other individuals. Not surprisingly, the young and smaller individuals are at the greatest competitive disadvantage (Jones, 1998).

Interspecific

Devils’ most serious current native competitors for food, at least at carcasses, are almost certainly the two species of quolls (but see below), which like Devils, are nocturnally active. Adult Devils, however, easily displace quolls at carcasses (Mooney, 1992; Jones, 1996; Jones, 1998; Jones and Barmuta, 2000).

Wedge-tail eagles and currawongs are also potential competitors even though they don’t overlap with Devils in their activity time.

Among exotic species, house cats, foxes and wild dogs would also be attracted to the same carcasses as Devils and at the same time of day. It is unclear who would have the upper hand with the first two species, but odds would be on the Devil. Indeed, there is video footage of a cat beating a hasty retreat from a carcass as a Devil enters the scene (STDP website: “Do devils suppress cats?,”as updated on 5 October 2010). Dogs might be another matter, as dogs can and will kill Devils (see “Exotic predators”).

In the past, Devils may have competed with Thylacines for carrion on both the mainland and Tasmania. And when they lived on the mainland, Devils may have

6/05/2016 8:41 AM 61

competed goannas over carcasses, although they were out and about at different times of the day.

Devils have been described as part of a “carnivore guild” along with quolls and the Thylacine, with implied competition and its evolutionary consequences among its members (Jones, 1997). But the Devil is primarily a scavenger while the two species of quolls, at least, are primarily predators, so the idea of a guild is a bit of a stretch. If a guild concept were to be pursued, it might make more sense to put the Devil into an Australia-wide “scavenger guild” along with several large goanna species and put the quolls into a “predator guild.”

Reproduction

Knowledge about the Devil’s reproduction is perhaps second only to knowledge about its diet. Anatomists described the Devil’s reproductive tracts as part of the early interest in how marsupials differed from placentals and among themselves. Early field biologists worked out the Devil’s basic reproductive attributes and cycles as part of their efforts to understand the species’ general ecology. And physiologists worked on the basics of Devil reproduction as part of a general interest in variation among marsupials as well as a potential aid in captive breeding.

Age and Size at Maturity

Little is known about the age and size at sexual maturity in males due to the fact that it is difficult to judge if they are mature by visual inspection. In contrast, females show their state of maturity by their pouch condition and the presence of young.

Several observers have stated that in general Devils become sexually mature at two years of age (Guiler, 1978, 1992; Hughes, 1982).

With regard to males, specifically, one observer noted that males may be physiologically capable of breeding before the age of two, but he gave no evidence (Guiler, 1992). Another noted that spermatogenesis can occur in males as small as 3.5 kg (Hughes, 1982). One Tasmanian zoo apparently had a one-year old male (Studbook No. 1074) that bred successfully (Hibbard et al., 2012).

Most females reproduce first at the age of two (Fleay, 1935; Guiler, 1970b, 1978, 1992; Hughes, 1982; Pemberton, 1990), but between 0 and 12.5 percent of one-year old females also breed (Hughes, 1982; Pemberton, 1990; Guiler, 1992; Jones et al., 2008). Presumed one year old females can breed before their tibial epiphyses, assessed at the knee joint, are fully fused (Hughes, 1982).

The modal number of breedings for a female is three (McCallum and Jones, 2006; STDP website: “The disease,” 13 September 2011).

At a western coastal locality, the smallest reproductive (lactating) female weighed 4.5 kg (N = 39 based on Guiler, 1970a and b: table 1), but at an eastern lowlands locality, the smallest reproductive female (either pregnant or with a litter, which not stated) weighed 4 kg (N = 44; Hughes, 1982).

6/05/2016 8:41 AM 62

In an east coast population, one year-old breeding females had a mean weight of 5.5 kg (Lachish et al., 2009).

Studbook data for captive animals shows a median age of “first offspring” as two years for both males (N = 83) and females (N = 103) (PHVA, 2008; see also Lees and Andrew, 2012). These data suggest that in both sexes, but notably males as well for which data from the wild are scarce, some individuals are physiologically capable of breeding successfully before they are two.

Urogenital System

There are descriptions of the urogenital systems of male (Flynn, 1910, 1911, 1922; MacKenzie and Owen, 1919; Mackenzie, 1921: fig. 98; Hughes, 1982: fig. 26; Keeley et al., 2012b: fig. 1) and female Devils (MacKenzie and Owen, 1919; Pearson, 1945, 1950; Pearson and De Bavay, 1953; Hughes, 1982: fig. 3).

In females, at the beginning of the breeding season the reproductive tact increases in size and vascularisation. Prior to ovulation, the two uteri increase as much as tenfold in weight, and during pregnancy, they increase as much as 100 fold (Hughes, 1976a, 1982).

The urogenital opening increases in both length and width as the oestrous cycles unfolds. It also changes in colour from pale to pink in oestrus (Hesterman et al., 2008: fig. 4).

Testes

There is a difference in reported testis size in Devils. One researcher reported mid- winter (July) testis size in 19 males as ranging from 11 x 15 mm to 17 x 20 mm (Green, 1967). Another researcher reported mean testis size in 30 individuals collected over an unspecified yearly as 26 x 32 mm (Guiler, 1970b).

One study (Pemberton, 1990) was ambiguous about whether it measured testis (text and fig. 7.6 top y-axis label) or scrotum (fig. 7.6 caption). In any event, 113 individuals from throughout the year had monthly means ranging from approximately 27 - 28.5 mm in length and 35 - 37.5 mm in width, with no evident seasonal trend (Pemberton, 1990: fig. 7.6).

Testis mass averaged 1.71 g ± 0.26 in seven healthy Devils (Keeley et al., 2012b).

Among Devils, heavier individuals usually have larger testes (Green, 1967), and among carnivorous marsupials, Devils have relatively heavy testes compared to body mass (Taggart et al., 2003).

The absence of testes in one male has been attributed to a congenital condition (Kolkert, 2005, as cited by Keeley et al., 2012b).

The epididymis mass averaged 0.76 g 0.13 in seven healthy Devils (Keeley et al., 2012b).

6/05/2016 8:41 AM 63

Accessory glands

Like other marsupials, male Devils have only two kinds of accessory glands, a prostate and two pairs of Cowper’s glands. In seven healthy Devils, the prostate mass averaged 2.84 g ± 0.90 and the mass of the first and second pairs of Cowper’s glands averaged 0.42 g ± 0.20 and 0.34 g ± 0.19, respectively (Keeley et al., 2012b).

Sperm

The information on the possible seasonality of sperm production in mature Devils in the wild is incomplete and partially conflicting. In one study, spermatogonial mitoses, the early stage of sperm production, were plentiful in June, but the later stages of spermatogenesis, including mature sperm, were absent (Sharman, 1959). Another study involving 24 males weighing ≥ 3.5 kg captured in April and May (N = 18) and in August (N = 6) found active spermatogenesis, but only males ≥ 5 kg contained mature sperm (Hughes, 1982). A recent study of 17 males aged ≥ 1.5 y collected between January and June found they carried “viable sperm (Keeley et al., 2011).”

Devil sperm has been described in some detail (Hughes, 1965: fig. 1l; Kelley et al., 2011: figs. 2-3). The sperms’ “overall morphology … is typical of that of most other marsupials …” (Hughes, 1982; also Hughes, 1976a), although they tend to be larger than in most other marsupials (Hughes, 1982; Keeley et al., 2011).

The gross morphological changes in the final stages of sperm maturation have been described briefly and are similar to other marsupials (Matson et al., 2005; see also Keeley et al., 2011).

Ova

The ovulated egg consists of the vitellus (yolk) surrounded, concentrically, by the zona pellucida, mucoid coat and shell membrane. The vitellus has a mean diameter (N = 20) of about 210 μm (Hughes, 1976a, 1982), the zona pellucida a mean width of about 4 μm, the mucoid coat 28 μm and the shell membrane 5 μm, respectively (Hughes, 1982). The whole “egg” has a mean diameter of about 300 μm when the zona pellucida is still adherent to the vitelline membrane and about 435 μm when it has lifted off to form the perivitelline space (Hughes, 1982).

The granulosa-oocyte complex, when stripped enzymatically from the ovary, has a diameter of 265 μm (Czarny and Rodger, 2010). The state of maturity may account for the smaller measured diameter of what is a more inclusive structure than the ovulated egg.

Oestrus

In the wild, most females usually have only one oestrus per year (Hesterman et al., 2008b), and this occurs between early February and mid-April (Guiler, 1970b: Hamede et al., 2009). However, if a female fails to conceive in her first oestrus, or the young from a successful mating perish, she may undergo a second oestrus after a refractory period of about seven to 13 weeks. In one population, seven percent of the population of breeding females (one in 14) had either a second or a very late oestrus

6/05/2016 8:41 AM 64

(Hamede et al., 2009). Female Devils, therefore, are facultatively polyoestrus (Hesterman et al., 2008b; Jones, 2008; Hesterman and Jones, 2009), that is, they do not ovulate upon copulation but, rather, when they are physiologically ready.

Females that breed when only one year old, instead of the usual two, apparently come into oestrus rather late in the season, in late April and May (H. Hesterman, S. Jones and M. Jones, as cited in Lachish et al., 2009).

In captivity, mature females that fail to conceive in their first oestrus may undergo a second or even third oestrus (Hesterman et al., 2008a; Kelly et al., 2012; T. Faulkner, in Edwards, 2012b; Hibbard et al., 2012; Devil Ark newsletter, Autumn 2012). A female may not conceive until her third cycle (K. Masters, on ABC News, 26 September 2012; T. Faulkner, in Edwards, 2012b). However, females experiencing one unsuccessful cycle may have no further cycles in that reproductive season (Keeley et al., 2012d). The interval between the onset of the first and second oestrus is reported as 47-68 days in one account (Hesterman et al., 2008a) and 48-82 in another (Keeley et al., 2012d: table 3).

Interestingly, in one case, half (8/16) of hormonally monitored one-year old female Devils went through oestrus (Keeley and Eastley, 2012). As these young Devils had no mating opportunities, however, it is unknown whether their oestrus would have let to pregnancy and birth. The apparent high rate of oestus in these yearling females may have been due to their highly subsidised lives.

In captivity, success rates, that is, proportion of females producing young was 44 percent for the first cycle; 10 percent in the second and, as noted above, zero for the third (Keeley et al., 2012d).

An observation from captivity, however, seems to contradict the notion that successfully reproducing females have only one oestrus per season. Three of five females in a large open-air enclosure gave birth early in the season and at least two raised their young to weaning. But all five females are said to have undergone a second oestrus (Sinn et al., 2010). Whether this second oestrus in the two female that weaned young is an artefact of captivity or an indication of a more frequent second oestrus in the wild is unclear.

A complete oestrous period, measured hormonally (below), lasts about 32 days (Hesterman et al., 2008a).

In captivity, oestrus has been recorded between the end of January and late June (Fleay, 1935; Andrewartha and Campbell, 2006; Hesterman et al., 2008; Sinn et al., 2010) and is said to last between two and five days (Smith, 1993).

Females close to oestrus are said to solicit males by making a growling whine, sometimes quite high pitched, and short soft barks. The male is generally silent or responds with snorts (M. Jones, in Jackson, 2003).

Keepers use “behavioural and physical cues” to detect the approach of oestrus (Hesterman et al., 2008; Keeley et al., 2010, 2012d). The former include increased activity (D. Schaap, in Hibbard et al., 2012), diminished appetite (Hibbard et al.,

6/05/2016 8:41 AM 65

2012) and the on-set of nesting behaviour. The latter include the development of the neck roll and changes in pouch condition (below) (Keeley et al., 2012d and references therein). A common behavioural cue is diminished appetite. In one case, a previously dominant female started avoiding contact with her cagemates and came last instead of first to feed before losing interest in food altogether; she mated shortly thereafter (Kelly, 1993). This sudden loss of dominance may have been the prelude to the temporary loss of dominance in females during the mating period.

In captivity, some oestrous females will apparently fight if confined together (Turner, 1970), but others, at least long-time cagemates, are compatible, at least until the young are born when they may become querulous, although remaining compatible (Smith, 1993).

Reproductive Hormones

Males

In wild males, testosterone levels, as measured in blood samples, show no significant differences between juveniles, and adults and do not vary seasonally (Pemberton, 1990: fig. 7.5-6). This seems surprising, considering the role of testosterone in reproduction and the strong seasonality of reproduction in Devils.

Anesthetised males show no significant difference in testosterone levels between day and night (Keeley et al., 2012a).

In captive males, androgen levels, as measured mainly in the faeces and to a lesser extent in the plasma, are high in mid- to late summer, just prior to the period when females enter oestrus, fall away during autumn and early winter and then rise abruptly again toward mid-summer (Hesterman and Jones, 2009). It is unclear whether the mid-summer peak is part of an endogenous rhythm or was engendered by pheromones from the females, which were kept separate but in the same captive facility.

Females

Oestrus, the brief period in which mature females become physiologically able to ovulate and behaviourally ready to mate, consists of two phases, a follicular phase, when the follicles in the ovaries become ready to release their individual eggs and a luteal phase, when the eggs have been released and the follicle develops into a corpus luteum. Mating occurs during the follicular phase, possibly most reliably during the latter part (Hesterman et al., 2008a; Keeley et al., 2012d).

As elucidated in captive females, the follicular and luteal phases of oestrus are defined by two rises in progesterone levels, as determined both indirectly from metabolites in the faeces and to a lesser extent directly in the blood plasma (Hesterman et al., 2008a: fig. 1a; Keeley et al., 2010, 2012d: fig. 1).

In the follicular phase, the source of the progesterone is probably the maturing follicle. The hormonal peak is low and is followed by a return to baseline. This phase, as measured by fecal analysis, lasts 8-23 days (mean ≈14-15) in one account

6/05/2016 8:41 AM 66

(Hesterman et al., 2008a) and 12-26 (mean ≈ 19) in another (Keeley et al., 2012d; table 3).

In the luteal phase, the source of the progesterone is probably the corpus luteum. The hormonal peak is much higher than in the follicular phase. The phase, as measured by fecal analysis lasts 12-25 days (mean ≈ 17-18) in one account (Hesterman et al., 2008a) and 11-15 days (mean ≈ 12.5) in another (Keeley et al., 2012d: table 3). Birth occurs “within days” of the luteal phase’s sharp decline in progesterone (Hesterman et al., 2008a).

Pouch Condition

The pouch opens backward in Devils (Pocock, 1926: fig. 41c; Fleay, 1952; Guiler, 1970b; Hughes, 1982), not forward as in the more familiar macropods. When the pouch sphincter contracts, however, the opening shifts more toward the centre of the pouch (Hughes, 1982). The size of the pouch varies with the number of young (Guiler, 1970b).

The condition of the pouch changes appreciably during the oestrous cycle and provides a practical means of assessing a female’s reproductive state. Researchers have identified four stages (Hesterman et al., 2008b).

Stage 1.) Shortly before the breeding season, the pouch of mature females is pale and shallow – about 50 mm wide and 30 mm deep. The teats are everted but small (≤5 mm) even in females that have reproduced the previous year.

Plasma progesterone, faecal preganediol (a progesterone metabolite) and faecal total oestrogens are low.

Stage 2.) At the beginning of the breeding season, the pouch skin turns pink and the interior begins to secrete a pink to red oily substance (see also Fleay, 1935, 1952; Parer and Parer-Cook, 2004). This stage marks the beginning of the follicular phase of the oestrous cycle, the period just prior to the eggs being shed into the oviducts.

Plasma progesterone, faecal preganediol and faecal total oestrogens mean levels are about 1.8, 1.5 and 1.3 times higher than resting, respectively.

Stage 3.) The pouch becomes slightly deeper and the oily exudate more copious, sometimes even forming a crimson ring around the pouch edge. At this stage, copulation occurs.

Plasma progesterone, faecal preganediol and faecal total oestrogens mean levels are about 3.2, 2.9 and 1.8 times resting levels but begin declining.

Stage 4.) At its most developed, the pouch is c. 50 mm deep and has a thickened edge. The interior is highly vascular and damp and a thin clear fluid replaces the red oily exudate. After 2-3 weeks, and irrespective of whether the female has mated or not, small, white “studs” cover the floor of the pouch and persist for 2-3 weeks. The studs, which are probably enlarged sebaceous glands, appear at least a week before the young are born. This stage marks the beginning of the luteal phase of the oestrous

6/05/2016 8:41 AM 67

cycle, the period when the corpus luteum secretes the hormones that support pregnancy.

Plasma progesterone and faecal preganediol rebound markedly to rise to mean levels about 10.7 and 4.4 times resting but faecal oestrogens return to resting levels.

The condition of the pouch, especially the secretions, may provide important cues to males as to the female’s reproductive condition, because males have been observed “investigating” a female’s pouch before mating (Hesterman et al. 2008b).

Ovulation

The two ovaries together produce many more eggs than will ever go on to be young. One study found that female Devils ovulate 11-56 eggs (mean = 39, N = 6; Hughes, 1982; see also Flynn, 1922, 1939), but another put the number as high as 114 (Guiler, 1992).

The eggs in a female’s oviduct may be in a variety of early cleavage stages, ranging from uncleaved to 61 cells (Hughes, 1980, also 1982: figs 10-15). It is not clear why there is such a disparity in cleavage stages or how it relates to the young actually born.

During oestrus, hormonal indicators suggest that ovulation is likely to occur within a window of about seven days on average (Hesterman et al., 2008a).

Devils are asserted to be spontaneous ovulators (Flynn, 1939), that is, ovulating without the stimulus of mating, but there is no supporting evidence.

Details of the ovulated but unfertilised egg have been provided by Flynn (1939).

Mating

The condition of the pouch in mature females in January (Hamede et al., 2009) suggests they have not yet begun oestrus and hence have not begun mating.

There are many general comments in the literature about when most mating occurs in the wild: February to March (Hamede et al., 2009; M. Jones, in DPIPWE, 2010d); mostly in February and March (Mooney, 1992; Jones, 2008); a six week period in February and March (Jones et al., 2007a); from the end of February into March (Guiler, 1992); from approximately February 20 to end of March (Pemberton, 1990); March (Guiler, 1964); a three week period in March/April (D. Pemberton, in ABC News, 21 July 2008); March to April (Guiler, 1983a, b; Jones et al., 1997; Jones and Barmuta, 1998), and March through May (Hamede et al., 2012a).

The actual recorded dates of mating are between mid-February (The Advocate, 15 February 2012) and 22 March (Guiler, 1970b).

Some mating occurs up until June (Jones et al., in press, cited by DPIPWE, 2010d). Presumably, these late matings are due to the second oestrus following the failure of the first.

6/05/2016 8:41 AM 68

In captivity in Tasmania, mating has been observed or can be strongly inferred (males keeping females in their dens) between 17 February and 12 April (Roberts, 1915; Turner, 1970; Kelly, 1993; Hesterman et al., 2008b; Hesterman and Jones, 2009; see also Roberts, 1915). On the mainland, mating was observed in early April (Oglesby, 1978) and inferred (from the age of a joey) to have occurred possibly as late as the end of June (Smith, 1993).

The beginning of the oestrus cycle, as judged hormonally in captive females, occurs between late January and late July (Keeley et al., 2012d).

Females from a northeastern lowland locality, Avoca, carried small pouch young in the period 26-31 March, suggesting that they had mated shortly before this time (Hughes, 1982).

Females from unspecified localities contained unfertilised eggs and/or blastocysts (fertilised eggs in the early cleavage stages) in the period 21 April – 10 May (Hughes, 1982), which suggests those with unfertilised eggs may have been ready to mate and those with blastocysts had just mated.

Curiously, there are also second hand anecdotal observations that suggest non- oestrous females may defend oestrous females in order to prevent them from mating with interested males (M. Jones and A. Kelly, in Sinn et al., 2010).

Females are said to have, what was described in a St Valentine’s Day news item, as a “lovely, whining call” to attract potential mates (M. Jones, in Larkins, 2012).

In any event, when it comes to the actual mating itself, females are said to be “… quite choosy, they really like to mate with older males and with larger males and they’re also quite choosy as to personality of the male that they mate with. They really do like dominating males. They would not mate with a male that will not take charge and grab them by the neck and drag them around and push them around and dominant them” (M. Jones, Radio National – Innovations; 2 June 2008; see also M. Jones in Parer and Parer-Cook, 2004, Darby, 2006 and Larkins, 2012; Jones, 2008).

This also seems to be the experience of zookeepers. One said “You’ve got six kilo and ten kilo males and if you put a six kilo male in with the girl, she’s not going to have a bar of him, because they want the dominant one (T. Faulkner, in Cubby 2008).”

Males stay close to females in oestrus (Guiler, 1970b) and fend off other males (Guiler, 1971). When in oestrous, females are said to be receptive for up to five days in one report (Jones, 2008), and “… from anything from two days up to a couple of weeks” in another M. Jones, on Radio National – Innovations; 2 June 2010).

Just prior to mating, the female goes into “an almost trance-like state (M. Jones, in Larkins, 2012).

Each mating event is said to last “…about one and a half hours” and there may be two to four matings, after which the male guards the female against other males in order to

6/05/2016 8:41 AM 69

“protect his paternity” (M. Jones, on Radio National – Innovations; 2 June 2010; see also M. Jones, in Larkins, 2012).

In mating, the male mounts the female from behind and either grasps her with his front limbs (Guiler, 1970b, 1971; Jones, 1994: fig. p. 39) and by the scruff of her neck or anterior back with his mouth (Eisenberg et al., 1975; Oglesby, 1978; Kelly, 1993; Slater, 1993; Koncza and Pryor, 2001). When resting between copulations, the male may release his grip on the female’s nape and lick her head and neck, but quickly re- grasp her neck if she moves (Kelly, 1993). In the run up to the mating season, the scruff of the neck in females becomes swollen (Jones, 2001, 2008; Parer and Parer- Cook, 2004), which might mitigate the effects of the male’s mating bite. It is not clear what causes this seasonal swelling.

In captivity, copulations occur within a period ranging from one to at least 12 days (Kelly, 1993; Smith, 1993; Hesterman et al., 2008a; Hesterman and Jones, 2009).

Copulation itself takes from a few minutes to several hours (Guiler, 1971), although the condition under which these observations were made are unclear. One captive pair copulated between four and eight hours (Eisenberg et al., 1975). Another pair, which were in a mating den for four days, “copulated and rested in alternating 10 minute bouts (Kelly, 1993).” And a documentary on Devils said that during their time together, a pair may mate “many times” and “for hours on end (Parer and Parer-Cook, 2004).”

Mating may occur during both the day and night (Guiler, 1970b).

During copulation Devils may be silent (#) or quite loud, growling and hissing (Oglesby, 1978). . If a female does not want to be mounted, she rolls over on her back, throwing the male off (Eisenberg et al., 1975).

In captivity, a mating pair left the den once to walk to drinking water, the male griping the female’s nape with his mouth the entire distance. After drinking, the female started to trot away, but the male caught her, grasped the skin of her back (nape?) and dragged her back to the den where they copulated again (Kelly, 1993).

When a captive female defended the entrance to her nest box from her male partner outside, both animals would engage in a duet of whine-growling (Eisenberg et al., 1975).

In captivity, mating males eat little and females may not eat at all, although both sexes continue to drink (Kelly, 1993; Smith, 1993). Not surprisingly, both males and females lose condition. Males, for example, are said to experience 25 percent weight loss, anaemia and reduced immunocompetence (unpublished work cited by Jones et al., 2008; see also Jones and Barmuta, 1998). In its first breeding season, one captive- bred and possibly overweigh two-year old male released on Maria Island (see “Off- shore islands,” below) lost one-third of its weight during the breeding season (Devil Island, 2013f). After mating, the female’s appetite returns within 24 hrs of separation from the male (Kelly, 1993).

6/05/2016 8:41 AM 70

Observations on mating in captive Devils suggest that the amount of physical harm a male inflicts on a female with his nape grip during mating can vary from no obvious harm to deep gashes (Kelly, 1993; Slater, 1993).

There are unpublished observations that litters may consist of young from matings with more than one male (Jones, 2008; M. Jones, in DPIPWE, 2010d and Sinn et al., 2010; M. Jones unpubl. data in Jones and McCallum, 2011b). In fact, one account said that DNA typing showed that usually between two and four males were involved in siring a single litter (Parer and Parer-Cook, 2004; see also M. Jones, in Larkins, 2012). This is surprising considering the relatively short oestrus, male guarding of females and mated-female defensiveness.

Reference has been made to anecdotal reports that “… reproductive suppression, whereby females that are in heat are guarded by other females that are not (thus disrupting mating by males), can occur (M Jones, A Kelly, pers. comm.) (Sinn et al., 2010).” This is another interesting idea about the Devil for which there is as yet no published evidence.

Considering the extra activities associated with mating, it is not surprising that at least for males, body weight is heaviest just before the mating season in summer and lightest just after the mating season in autumn (Peck et al., 2015). Females are heaviest in winter, when they are carrying pouch young (Peck et al., 2015), but there appear to be no data for seasonal changes in female body weight corrected for pouch young weight.

Sperm Storage

Females are receptive to mating a few days prior to when the eggs are released from the ovary, as inferred by the luteal phase of the hormone cycle associated with oestrus (above). This implies, therefore, a period of sperm storage (Hesterman et al., 2008a; Keeley et al., 2012d). Estimates for this period are approximately seven (Hesterman et al., 2008a) or eight (Keeley et al., 2012d) days.

Gestation

The number of fertilised eggs and early-stage embryos in the uteri of any one female can be as many as 15 (Guiler, 1970b), and possibly higher on the basis of one composite count of eggs and blastocysts as high as 56 (Hughes, 1982).

There are several statements in the literature as to the general period when females are carrying embryos, that is, when they are pregnant. For example, one early worker said the majority of pregnancies occur between March and May (Hughes, 1976a). Later he reported finding pregnant females in April and May (Hughes, 1982).

It is difficult to determine the precise gestation time, that is, the time between fertilisation and birth, because it is difficult to know precisely when either event occurs, especially the former. Early estimates, however, were “about three weeks” (G. Wood, in Hughes, 1982), 31 days (Adsell, 1964; Guiler, 1964, 1971, 1983a), 35 days

6/05/2016 8:41 AM 71

(Guiler, 1983b). Later estimates, however, were generally shorter: 18 to 21 days (Guiler, 1992) and 19 days (Hughes and Rose in Rose, 1989; Slater, 1993).

Based on examination of the development stages of specimens from the wild, an embryologist suggested that gestation might take 13 to 14 days (R.L. Hughes, in Keeley et al., 2012d).

In captivity, the first or last copulation can be objectively determined and is sometimes taken as a proxy for fertilisation, although there is probably some delay. Based on the first copulation, gestation lasts 20-31 days (Hesterman et al., 2008b) and based on the last copulation, 19-27 days (mean ≈ 24) (Hesterman et al., 2008a-b).

The onset of the luteal period in the oestrous cycle, as indicated by a hormonal surge (below), is another objective proxy for fertilisation. On this basis, gestation ranges 14- 22 days (mean ≈ 18) (Hesterman et al., 2008a).

Recently, a worker reported, as a personal communication from an earlier worker, a “pregnancy” period of 13-14 days during which “implantation” occurs during the final 2-2.5 days” (R.L. Hughes, as reported in Keeley et al., 2012d). If “pregnancy” means gestation, then this is the shortest recorded gestation period for the Devil. It would also be the first record of any sort of “implantation.”

Birth

In captivity, pregnant females are said to “hide away,” not return for several days and “won’t eat (T. Faulkner, in Schliebs, 2005).”

In the wild, births occur over a broad period, from February through June (Pemberton, 1990, as cited in Lachish et al., 2009) with most in March and April (Guiler, 1992; Mooney, 1992; Jones and Barmuta, 1998). One group of researchers, for example, said that about half of all births occur in March (McCallum et al., 2009), and another researcher estimated that all 21 litters examined were born in April (Hughes, 1982). A three year field study in north-eastern Tasmania found that births occurred over a three week period with a modal birth date of approximately 10 April each year (Pemberton, 1990). And recently, one group of field workers set a birth date of 20 March for fixing subsequent life history events (Hamede et al., 2012a).

In captivity, most births have occurred between early autumn and early winter (19 March to 22 June) (Fleay, 1935, 1952; ABC World Today, 5 April 2007; Currumbin media release, 2008; Devils@Cradle website: “Successful breeding season,” 22 December 2008; Stapleton, 2008a-b; Taronga Conservation Society website: “Zoo successful with first Tasmanian Devil breeding season,” 17 July 2008; Hesterman and Jones, 2009; Sinn et al., 2010; Joyce, 2011; Australia Zoo website: “Have a go at our five new Tasmanian Devil joeys,” 19 August 2011; Ward, 2011; Albert & Logan News, 19 October 2012; Sunday Mail, 15 January 2012; Stokes, 2015). There are, however, a few accounts of births at zoos occurring outside this period. The unusual dates are: September (ABC Wildlife at the zoo, episode 7, 30 May 2012; October, ABC News, 11 March 2014; Gloucester Advocate, 12 March 2014; Griffiths, 2014); early December (STDP website: “Founder devils home from mainland Insurance Population,” 9 December 2010), and late January (Hobart Mercury, 30 January 1931).

6/05/2016 8:41 AM 72

Some captive-breeding facilities believe out of season births are an indication of the success of their programs (ABC News, 11 March 2014).

Birth, as inferred as the end of the luteal phase of the oestrous hormonal cycle (above) occurs between 19 February and 10 May (mean = 22 March) for conception in the first cycle; between 14 April and 4 June (mean = 9 May) for conception in a second cycle, and between 25 June and 18 August (mean = 19 July) had conception occurred in a third cycle (Keeley et al., 2012d).

The female gives birth while standing (Guiler, 1992) on all four legs with the hindquarters raised and the head down and in an almost trance-like state (frame from NHNZ, 1999 in Rose and Shaw, 2006: fig. 1a; Parer and Parer-Cook, 2004). This position provides a downhill path for the young from the cloacal opening to the backward opening pouch (Rose and Shaw, 2006). Video footage shows that the young are emitted within a stream of cloacal fluid, which is initially pale pink in colour and thin in consistency and then becoming slightly more viscous and paler (Parer and Parer-Cook, 2004). The female gives no assistance to the young during birth, other than to lick the fur (Guiler, 1992), presumably along the track to the pouch.

Females give birth to between 20 and 40 rice-size, naked, sightless young (Guiler, 1971; Pemberton, 1990; Parer and Parer-Cook, 2004; M. Jones on Radio Australia – Innovations, 2 June 2008), which in order to live must crawl from the cloacal opening to the pouch, enter the pouch and then find one of the four teats to which to attach their mouth. Once attached, the teat enlarges inside the newborn’s mouth, making it difficult for it to become detached (Guiler, 1992). If a new-born fails to find and attach to a teat, it perishes. Hence only a few of the embryos produced survive, and selection in the short crawl from cloaca to pouch must be intense (Grueber et al., 2015a).

For the record, Devils first bred in captivity in the Beaumaris Zoo, Hobart, in 1913 (Roberts, 1935).

It is difficult to tell from the available data how quickly Devils sent from Tasmania to zoos in the Northern Hemisphere can shift their reproductive cycle to the new seasonal cycle. Two pairs that arrived in Copenhagen in the early northern autumn (early October) mated in winter (late February) and the females gave birth in late spring (April) (ABC News, 14 June 2013, where the export date was given as November; Hobart Mercury, 2 April 2013; Kempton, 2013a, c; Glaetzer, 2014a). A pair that arrived in Toronto in early (northern) winter (29 December) had a litter the following late spring (3 May; data in Koncza and Pryor, 2001). These animals seemed to have maintained their original Tasmanian breeding cycle. A female that arrived from Tasmania in Fort Wayne (Indiana) in late autumn (14 December) gave birth in the following late autumn (30 November; data in Koncza and Pryor, 2001). A pair of Devils of unstated origin that arrived in Chicago in early autumn (September) mated first in the following early spring (19 March; (Hartz, 1966). Young in the Basle zoo were born in early spring (24 March; Hartz, 1966). Arguably, in all these cases the new arrivals appear to have kept their home seasonality, albeit with a bit of a lag in the case of the Fort Wayne animal.

6/05/2016 8:41 AM 73

A pair that had been in the Northern Hemisphere for more than a year had a litter in early winter (1 January; data in Koncza and Pryor, 2001). A pair of “half-grown” Devils of unstated origin arrived at the Washington D.C. zoo in spring and tried to mate for the first time in the following winter. After two years, these animals were trying to mate in spring (Eisenberg et al., 1975). In these two cases, the animals appear to have been unable to adjust after more than a year in the Northern Hemisphere.

Litter Size

Female Devils ovulate up to 21 eggs (Flynn, 1922) and as many as 15 embryos have been found in the two uteri together (Guiler, 1970b). However, Devils have only four teats (Guiler, 1970b). The maximum possible litter size, therefore, is four.

The litter size, however, is often less than four. In 53 litters from three localities, the number of young ranged between one and four, with a mean of 2.9 (Guiler, 1970b). In 97 litters observed over ten years at one locality at a northwestern locality the mean yearly number of young ranged 2.0 - 3.3 with a grand mean of 2.7 (Guiler, 1978). In 82 litters observed in three years at one northeastern locality, the mean number of young was 2.9 (pers. obs. on data in Pemberton, 1990; fig. 4.4). In 44 litters observed over nine years at an east coast locality, the mean number of young was 3.4 (Lachish et al., 2009). And on the basis of “limited field data at high and low densities,” means were 3.3 and 3.7, respectively (Lees and Andrew, 2012).

Observations on whether mean litter size varies among with female age varies. One experienced field worker noted that “two year old females consistently have the largest litter size, averaging as high as 3.59 young per animal; this figure decreases for older animals (Guiler, 1992),” but he provided no supporting data or analysis.

In contrast, data from a northeastern population indicate that there is no significant difference in litter size between females two years old and older than two years (P = 0.38; Kruskal-Wallis Test adjusted H = 0.75, P = 0.38, pers. obs. on data in Pemberton, 1990: fig. 4.5).

Data from this population for litter sizes for 17 females captured over two or three seasons indicated a just not significant, but suggestive, indication that older females have smaller litter sizes (Kendall’s Test of Concordance, P = 0.06; Pemberton, 1990: table 4.10 and appendix 5). Analysis of the observed frequencies of the three possible changes between two successive litters of individual females, that is, increase (N = 4), stay the same (N = 10) or decrease (N = 6) suggested they are not significantly different (Kolmogorov-Smirnov One Sample Test; pers. obs. on data in Pemberton, 1900: table 4.10).

Data from three populations taken together indicate that litter size is inversely proportional to female weight (Spearman rank correlation = -0.89, P < 0.03 for litter size on weight class of female; pers. obs. data in Guiler, 1970b). In other words, lighter females have larger litter sizes (Guiler, 1970b, 1992). This could be due to the larger energy demands placed on a female by a large litter compared to a smaller litter. It is probably not evidence of a decrease in litter size with age, as Devils “…

6/05/2016 8:41 AM 74

cease mass gains early in … life (Pemberton, 1990).” In any case, weight can vary widely within and among adults depending on a variety of factors.

The proportion of adult females in the population has no association with the mean litter size (rs = 0.05, P = 0.88; pers. obs. on data in Guiler, 1978: table 9). Therefore, whatever factor may influence female participation rate seems not to affect litter size. Specifically, total rainfall in the year preceding birth – the only factor for which there is environmentally relevant data available – seems to have no significant influence on either participation rate (rs = -0.07) or litter size (rs = -0.20; pers. obs. on data in Guiler, 1978: tables 1 and 9).

Parenthetically, the Save the Tasmanian Devil Program almost certain has a huge amount of data that could provide insight into the environmental parameters affecting a variety of the Devil’s basic biology, such as participation rate and litter size, but it has yet to be analysed.

In captivity, mean litter sizes are not obviously different from mean litter sizes in the wild. Mean litter sizes may be smaller in facilities that are more intensively managed. In 155 litters born in what might be thought of as small zoos and wildlife parks, the number of young ranged from one to four, with a mean of 2.7 (PHVA, 2008; Lees and Andrew, 2012). In an unspecified number of litters born in what might be called a large zoo or wildlife park, the number of young ranged one to four with a mean of 3.0 (pers. obs. on data in Lees and Andrew, 2012). And in an unspecified number of litters born in large free range enclosures (below), the number of young ranged one to four with a mean of 3.0 (pers. obs. on data in Lees and Andrew, 2012). The last two groups, however, had identical frequency distributions, raising the question of whether there might have been a lapse in the table data entry (Lees and Andrew, 2012: table, p. 54).

In captive females, there is no apparent difference in litter size with age, at least among two- to four-year olds, for which there are adequate data (Fig. 3). These data for captive females are in line with the data for wild females (above), which indicate no significant directional change among a female’s successive litters.

6/05/2016 8:41 AM 75

Mean Litter Size vs Female Age in Years

3.5

2.5

1.5

Mean Litter Size 1

0.5

0 1 2 3 4 5 Female Age in Years

Fig. 3. Mean litter size by age in females in captive breeding facilities in the 2012 breeding season (data in Hibbard et al., 2012). Sample sizes: 1, 14, 15, 4, 1, respectively.

Interestingly, wild-caught captive females have significantly larger litters than captive-bred captive females, means = 2.8 vs 2.1, F = 7.8, P = 0.006; Hibbard et al., 2012).

Considering a litter size of four, one litter per year and a four year reproductive period (two to six years old), a female is unlikely to have more than 16 young in her lifetime.

Diapause

In contrast to many marsupials, Devils have no diapause; that is, they do not hold young in an arrested state in the uterus until the older pouch young have vacated (Guiler, 1970b).

Development and Growth

Information on the Devil’s development and growth is one of the areas in which animals kept in captivity have contributed substantially due to the close daily observation these conditions provide. Appendix 1 summarises developmental events in relation to age and size.

Embryonic Development

When the eggs are shed from the ovary, they are swept into the two uteri where they will be fertilised by the waiting sperm. A developmental biologist has collected and described embryos from the uterus ranging from just two cells resulting from the first cleavage up to a “well-formed unilaminar blastocyst, that is, a “single-layered, hollow ball of cells without an inner cell mass (Hughes, 1976a).” These blastocysts still retain the mucoid coat and complete shell membrane evident in ovulated eggs (Hughes, 1982). There is no morula stage (Hughes, 1976a).

6/05/2016 8:41 AM 76

The interval between the appearance of the neural streak and birth has been stated as 2.67 days, much shorter than that for a bandicoot (5.5 days) and a kangaroo (11-12 days) (Hughes and Hall, 1988).

Neonates

Newborn young, that is, those young entering the pouch are pink (furless) and about the size of a gain of rice. Two early studies gave the weight of the “new born” as ranging 0.18 to 0.29 g (Guiler, 1970b) and averaging about 0.244 g (N = 4; Hughes, 1982). Two latter studies, however, gave the weight of < 24 h old young as 30 mg (= 0.030 g) (Hughes and Hall, 1988; Hughes and Rose in Rose, 1989; table 1). The latter figure is almost certainly correct based on a comparison of the weights of other similarly sized marsupial species. The earlier figures may have been based on young a few days old (R. Rose, pers. comm.).

In comparison to other marsupials, such as bandicoots and kangaroos, Devils are born in a relatively undeveloped state. Their small size may make them susceptible to hypothermia and desiccation as they crawl from the urogenital opening to the pouch.

At birth, the young have well-developed front limbs with “conspicuous, functional claws,” which no doubt help the vulnerable embryo to gain quickly the safety of the pouch. The rear limbs, in contrast, are “rudimentary immobile paddles” that clasp the recurved tail (Hughes, 1982: fig. 25; Hughes and Hall, 1988: fig. 2.1).

Pouch Life

Estimates of how long the young stay in the pouch vary from about 105 (15 weeks) to 140 days (Guiler, 1983b), or specifically: 105 to 112 days (15 to 16 weeks; Guiler, 1970b, 1983a; McCallum et al., 2009), 130 days (Pemberton, 1990) and 140 days (Guiler, 1983b). The shortest period, that is, 105 days has been attributed to animals raised in captivity where food was plentiful and hence growth rates likely to have been accelerated (Pemberton, 1990); however, this period was determined from wild animals and preserved specimens (Guiler, 1970b).

In the wild, young can be found in the pouch as early as April (Roberts, 1915; Lazell, 1984; Hamede et al., 2009; Sinn et al., 2014) and possibly as late as late December (Newsletter, August 2007). Recorded months are: April (Roberts, 1915), June (Hamede et al., 2008), July (Launceston Examiner, 17 July 1872; Fleay, 1935; Green, 1967; Hamede et al., 2008), June-July (Hamede et al., 2008), August (Hobart Mercury, 6 September 1870; Guiler, 1970b; Pemberton, 1990; Jones and Barmuta, 1998), September (Roberts, 1915; Green, 1967; Guiler, 1970b; Newsletter, December 2007), October (Green 1967; Guiler, 1967) and December (“shortly after … Christmas,” Newsletter, August 2007, but possibly a captive animal).

In captivity, young can be found in the pouch as early as the end of April (Sinn et al., 2010) and as late as August (Andrewartha and Campbell, 2006). Recorded months are: April (Sinn et al., 2010), August (Andrewartha and Campbell, 2006)

6/05/2016 8:41 AM 77

The pouch young of any one litter are not always at the same developmental stage. In a litter of three, one individual was “slightly less developed than her two pouch mates” (Green, 1967). In a litter of four, one young was said to be about 10 days behind its three siblings in both its state of developmental and weight (Guiler and Heddle, 1974). And in a litter of three, completion of furring was spread over a week (Guiler, 1970b, 1992). These differences in development might reflect normal variation with one male parent or might be due to multiple male parents.

The young are relatively secure when in the pouch. One field researcher reported no loss of young in 97 females, despite repeated handling (Guiler, 1970b, 1992).

The constituents of Devil milk are similar to that of other marsupials (Green, 1984; see also Hewavisenti et al., 2014). It does not contain more iron as stated in DPIPWE, 2010d.

In captivity, young born at the beginning of June first released their continuous hold on the teat, the first step toward an independent life, on 1 October, at an age of about 4 months (Fleay, 1952).

Pouch Exodus

There are few observations from the wild as to when the young leave the pouch. In a three year study in northeast Tasmania, the earliest date recorded for young being left in the den was 6 August and all young had been “denned” by the middle of August (Pemberton, 1990). In the central highlands, young are said to leave the pouch in August (Jones and Barmuta, 1998).

There is one observation of a female with one pouch young and two lactating teats, suggesting that the young may not all leave the pouch simultaneously, assuming the missing young were denned and had not died (Pemberton, 1990).

In captivity, young have been recorded as being outside the pouch for the first time in early August (Oglesby, 1978; Joyce, 2011).

Estimates of the age at which the young can first be found completely outside the pouch range from “just over 12 weeks” (Kelly, 1993), through 15 weeks (Fleay, 1935) to about four and a half months, that is, about 18 weeks (Roberts, 1915).

In captivity, and probably in the wild, the young give up pouch life only slowly. After their first trip “outside,” they will return at times of stress or danger (Roberts, 1915; Fleay, 1935).

Young Devils’ exodus from the pouch also marks the beginning of their starting to cling to the female’s back by hanging onto her fur with their teeth and claws (Fleay, 1935: plate 9).

Females may exhibit a sudden change of behaviour when the young first leave the pouch. One captive female was docile and easy to handle while her young were still in the pouch, but afterwards, she suddenly became “quite fierce” and difficult to handle (Fleay, 1935). Another captive female became very aggressive toward the male parent

6/05/2016 8:41 AM 78

when her young were approaching the time of first leaving the pouch, so much so that he had to be removed from the same cage. Circumstantial evidence suggests that the male subsequently killed one of the emerged young, when the two were inadvertently left in the same enclosure overnight (Smith, 1993).

Once a teat has been stimulated by a suckling young, it continues to be active for the remainder of the breeding season, even if no further suckling occurs (Lachish et al., 2009). This makes it impossible to know if an active but vacant teat represents a young left in a den or the death of a young (Pemberton, 1990).

Den Life

In the wild, lactating females found on their own are circumstantial evidence that they have left, temporarily, unweaned dependent young, in a rearing den (Guiler, 1970b; 1983a). There are, however, few such recorded instances. In a three year study in northeast Tasmania, the earliest date for nine lactating female without pouch young was 6 August and the latest 30 August (Pemberton, 1990; see also Guiler, 1970b). In southeast Tasmania, two lactating females without pouch young were found on 7 September (Newsletter, December 2007). In northwest part of the state, “more than half of the 63 females captured in a November survey were lactating (Newsletter, March 2012). And in the central highlands, lactating females without pouch young have been found between October and January, although the young are said to have left the pouch in August (Jones and Barmuta, 1998). These observations suggest that for the species as a whole, unweaned young may be in rearing dens between early August and January, but perhaps with a tendency for this part of the breeding cycle, at least, to occur somewhat latter in the year in cooler regions than in warmer ones.

There is little information on how long each day female Devils may be away from their denned young. But one tracked female with seven month old young, spent as much time away from the den as she did when she did not have denned young (Jones et al., 1997).

A fortuitous observation of four unweaned young Devils emerging alive but emaciated after being lost for in a rat warren for eight days (Fleay, 1952) suggests that young at this stage might also survive unattended this long in the wild if necessary.

There is one media report suggesting that females stimulate their unweaned young to defecate and urinate, probably by licking the anal area with their tongues (simulated by carers wiping the bums of unweaned orphans with a moist towel) (Newsletter, December 2010).

There are second hand reports based on captive animals that males “assist the females in caring for the young” (Collins, 1973). But there is no evidence to support this assertion.

Back-riding Behaviour

In captivity, it is common for large, but still not completely weaned, young to ride on their mother’s back or belly, clinging on with their teeth and front claws (Fleay, 1935: plate IX; Fleay, 1952: fig. p.275); Turner, 1970; Oglesby, 1978; Kelly, 1993; M.

6/05/2016 8:41 AM 79

Hingley, in Marszalek, 2007; Edwards, 2012a). This behaviour has been observed first in various litters in August (Oglesby, 1978). The age of first back-riding has been given as about 22 weeks (Kelly, 1993).

There is also one report of the young clinging to the father (Turner, 1970), but this needs confirmation.

This clinging behaviour has only rarely been observed in the wild (Pemberton, 1990). One field worker saw it only once in his long experience and this was when he caught a female with her young in a trap. The young were clinging tightly to their mother with their front feet, but when released from the trap, they abandoned her for him and were difficult to remove (Guiler, 1970b, 1992).

The significance of the back-riding behaviour is unclear. Perhaps the mother uses it only when it becomes necessary for some reason to move young joeys away from their den, perhaps to a new one. Or perhaps it is her way of getting older joeys to leave the den. If they drop off and are left behind, they are on their own.

There are two recorded observations of a female handling one of her young with her mouth, both on video. In one case, a female under a house is carrying a young animal, presumed to have fallen out of her pouch, across its hind-quarters; she moves a few paces in one area and then climbs through a hole into another area and disappears (ABC News online, 13 November 2015). In another case, a captive female picks one of her free-ranging young out of a creek, which it had fallen into while drinking with the female and litter mates (Parer and Parer-Cook, 2004). The female grasps the pup by the front leg with her mouth and pulls him out.

Weaning

Weaning, the period when a young Devil makes the transition from feeding entirely on its mother’s milk to feeding entirely on other food, is difficult to determine precisely, because it happens gradually and its beginning and ending is difficult to observe.

One clear morphological marker for the onset of weaning, however, appears to be the eruption of the first upper incisor on each side of the mouth. The late eruption of this pair of teeth closes a gap in the surrounding teeth through which the teat passed into the mouth of the suckling Devil (Guiler and Heddle, 1974).

One study in the wild estimated that weaning occurred about 278 days after birth (Pemberton, 1990).

Observations from the wild suggest weaning can occur anytime between October and February. The earliest date, October, is based on one young of the year caught in a baited trap (Jones, 1995; Jones and Barmuta, 1998; also N. Mooney, in Wood, 2003b). But estimates of the earliest and latest dates vary. As to the earliest dates, one experienced field observer with much trapping experience said on one occasion that females lactate until November-December (Guiler, 1970b), although he later modified this to no females were lactating by the end of November (Guiler and Heddle, 1974) and then to weaning is finished by late October to early November (Guiler, 1983b).

6/05/2016 8:41 AM 80

As to the latest day, weaning is finishing in January on the east coast (Freycinet Peninsula; Hamede et al., 2008). Weaning also finishes in January in the central highlands (Jones, 1995; Jones and Barmuta, 1998; see also Jones, 2008). The latest date is from an unpublished thesis where the weaning period is given as between December and February (Jones, 1995; Hesterman, 2008, cited in McCallum et al., 2009). In any event, all these observations indicate that the young Devils start fending for themselves during late spring to mid-summer, when food is most abundant (Guiler, 1983b; Hamede et al., 2015).

Observations of young Devils in captivity, where their development can be closely monitored, give widely varying estimates of the actual ages at weaning. A litter of three had “practically ceased to rely on the mother’s mammary glands for nourishment” at about 154 days (22 weeks or five months) (Fleay, 1935; see also Fleay, 1952). Two litters totalling five young were completely on solid food between about 152 - 183 days (five to six months; Smith, 1993). In one litter of four, weaning occurred at an age of about 200 (± 5) days (Guiler and Heddle, 1974). But in another litter of four, it occurred at approximately 232 days (Phillips and Jackson, 2003). Finally, a young of about 213 days (seven months) as still at heel with its mother (Hughes et al., 2011); presumably it would not still be following her unless it was still suckling, at least in part.

A general statement about weaning times identifies the period from mid-December to early February (DPIPWE, 2010d).

A general observation based on several captive-born litters was that young started eating solid food at about 126-154 days (18-22 weeks) and are completely weaned at about 280 days (40 weeks; Kelly, 1993). There is also an unsubstantiated estimate of weaning at about 244 days (eight months; Guiler, 1971).

There is as yet no explanation for the wide variation in estimated weaning times. But one observer wrote that “the exact time [is] somewhat dependent on availability of food (Mooney, 1992).” The observational basis for this comment is unclear.

In what may be part of the weaning process, females have been observed bringing food to the den for the young both in the wild (Guiler, 1992; Jones, 1994) and in captivity (Roberts, 1915; Parer and Parer-Cook, 2004).

One captive female was observed running up and down outside a natal den with food in her mouth. This was interpreted as an enticement for the young to leave the den (A. Anderson, in Owen and Pemberton, 2005).

In captivity, females with weaning young would not eat if separated from their young and when together would let their young eat their food (Smith, 1993).

In the wild, lactating females will shelter in a den that does not contain their young. This suggests that the young are not fed everyday. Indeed, there is some evidence females are “away” more often toward the end of the weaning period, suggesting that increasing hunger may be part of the weaning process (Pemberton, 1990).

6/05/2016 8:41 AM 81

In the wild young Devils have been seen outside and away from any obvious den in late September (STDP website: “Survey for orphaned Devils at Taroona,” 30 October 2014).

Little is known about what just-weaned Devils eat in the wild. Young Devils killed in the wild by dogs and cars often contain invertebrates in their stomachs (STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011). Devils as young as “approximately 8 months” old are occasionally caught in traps baited with meat (Peck et al., 2015). Generally, however, Devils up to the age of three months post-weaning are rarely (3 in 276 individuals) caught in traps baited with meat (Guiler and Heddle, 1974; Buchmann and Guiler, 1975; Guiler, 1992) suggesting their diet differs from that of adults.

In captivity, some observers report that young Devils eat the “same types of food” as adults (Buchmann and Guiler, 1975) while others note that some, at least, will chase and eat spiders (STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).

In captivity, the first solid food of young consisted of mealworms and earthworms (Oglesby, 1978). Young perhaps just days from being fully weaned will eat meat brought to them by their mother, and newly weaned young will eat meat on their own (Roberts, 1915).

Juveniles are said to become fully independent at about nine months of age (Lachish et al., 2007, 2009).

Oddly, young Devils appear to have greater manual dexterity than adults do. This is evident in both their greater agility in climbing vertical structures such as tresses and fences (Roberts, 1915; Buchmann and Guiler, 1977; Jones, 1996) and in grasping prey (Buchmann and Guiler, 1975). A good grip would undoubtedly be useful for young in the late stages of their association with their mother, when they climb on her back and are carried about.

In captivity, young Devils up to about one and a half years of age seem more generally agile than older Devils in all sorts of activities, including jumping, climbing and burrowing (Kelly, 1993). Perhaps young Devils’ greater manual dexterity (climbing) and agility (flight) would help them avoid older Devils during dispersal.

There seem to be no observations of females foraging with young tagging along. This suggests that if the young learn anything from the mother it would be what they see in the den, which would be mainly food handling. Indeed, it is said, “the young do not accompany the mother at night and she does not teach them to hunt (referenced as “Jones et al. in press,” in DPIPWE, 2010d).”

Dispersal of Young

Between the time they leave the natal den and the time they are one year old, young Devils are usually called “juveniles,” and between one and two years of age they are usually called “subadults (Lachish et al., 2007, 2010; see also Peck et al., 2015).” It is quite possible, of course, that a young Devil may reach its first birthday, that is

6/05/2016 8:41 AM 82

become a subadult, and still not be settled in an area. Similarly, the small percentage (up to 12.5 percent) of Devils that breed in their first year (above), may be subadult for a much shorter period than the majority that breed at two years of age.

The literature indicates that the overall dispersal period for juvenile Devils is between early November and April (Pemberton, 1990: table 4.4; Guiler, 1970, 1992; Lachish et al., 2007; S. Blackhall, on ABC Hobart, 24 October 2011), the extremes, of course, representing the tails of the distribution.

Just-weaned juvenile Devils may stay close to their den for a few days prior to striking out on their own, as evidenced by being caught in baited traps set just outside their den in November (Pemberton, 1990: table 4.4).

The young may move large distances before settling into a home area of their own. For example, “dispersal movements of 10-30 kilometres have been recorded directly … [and] average dispersal distance is probably much greater (Jones, 2008).”

In nine captive-raised “orphan” Devils released into the wild, four c. 10-14 month-old animals dispersed an average of c. 750 m in a three to eight month period whereas five c. 20 month-old animals dispersed an average of c. 150 m in a c. eight month period (Sinn et al., 2014: fig.4 and table 1). This result provides an insight into the Devil’s age-related drive to disperse from its place of birth.

The fact that Devils do not have territories, that is, areas which they defend from other Devils, is likely to facilitate dispersal. Presumably, dispersing young would not be hindered in either moving or settling.

A three-year field study in the northeast found that most young males tended to disperse but most young females tended to stay close to the areas where they were born (Pemberton, 1990; see also DPIWE, 2005a). A later field study also found that “primarily males … disperse after weaning” (Jones et al., 2007a) and that “dispersal of young animals is male biased; some females disperse” (Jones, 2008; also Jones, 2003b). The supporting data for these early observations have yet to be published.

Genetic evidence for a sex bias in dispersal is contradictory. One study found evidence for greater dispersal in males than in females (Lachish et al., 2011a), but another study found no significant difference (Brüniche-Olsen et al., 2013).

In terms of actual measured dispersal, an on-going broad scale survey of marked, subadult animals found no significant difference between the sexes in actual dispersal. Eight males dispersed a mean distance of 20.3 (± 7.4, 95 percent confidence interval) km and seven females dispersed a mean distance of 42.0 (± 28.1) km, with an overall range of 12.5 – 109.3 km (straight-line distance) and mean of 30.3 (Lachish et al., 2011a, but these data include both pre- and post-disease populations).

Young Devils are more crepuscular than adults are (Owen and Pemberton, 2005), but the reason for this is elusive.

The post-weaning dispersal stage is probably the most dangerous time of a Devil’s life. The animals are small, inexperienced and often in unfamiliar surroundings.

6/05/2016 8:41 AM 83

Occasionally there are reports of “plagues” of Devils. These observations are probably based young dispersing Devils.

Growth

There are observations, from captivity, on weight increase in the pouch young of four litters. In one young of unspecified age, weight increased from 112 g to 160 g, an average rate of 4.4 g/day (#). In a litter of three, also of unspecified age, mean weight increased from 130 to 230 g in 11 days, an average rate of 9.1 g/day (Nicol, 1978). In a litter of four, between 142 days old and weaning at 232 days, mean weight increased from about 300 g to about 1500 g, an average of 13.3 g/day (data in Phillips and Jackson, 2003). In a litter of four 144 days old, means weight increased nearly linearly from about 490 g to about 1960 g in about 96 days, an average rate of 15 g/day (Guiler and Heddle, 1974: fig 1, from which data were estimated; weaning occurred halfway through the observation period).

In four young from two litters, growth between the ages of about 110 and 180 days (N = 1) and about 140 and 225 days (N = 2) and 260 days (N = 1) was slightly exponential and between five and six months of age averaged 12 g/day and at an age of seven months averaged about 38 g/day. And in three young from one litter, growth between about 135 days and 240 days (N = 2) and about 285 days (N = 1) was nearly linear to slightly exponential and “initially” averaged 14 g/day and at approximately seven months of age averaged 22 g/day (Smith, 1993: fig. 3).

In the wild, “juveniles” showed a growth rate of about 0.5 kg per month (e.g., about 16g/day; Guiler, 1970b).

In captivity, five immature animals (from three litters) aged approximately 7 months grew at rates of between 30 and 50 g/d over a 19 to 39 day period (data in Sinn et al., 2010: table 3).

In captivity, juveniles, that is, one year old Devils, weight between 5 and 7 kg (STDP Annual Report, 2011/12).

In general, female Devils attain their “final adult weight” at about 18 months and males at about two years (Jones, 1995; Jones and Barmuta, 1998), although two-year old females have also been called “small” (Jones et al., 2008).

Two wild males showed exceptional weight increases. Beginning at 2.28 and 3.5 kg, they increased their weight over a six month period at a rate of 27 and 32 g/d, respectively (Guiler, 1978).

There are static growth curves for head length, head width, rear limb length and rear foot length for the Mt William population (Pemberton, 1990: figs. 2.4 - 2.7). The raw data (Pemberton, 1990: appendix 2) have been replotted here (Figs. 4-7).

6/05/2016 8:41 AM 84

Head Length vs Age in Devils

175

165

155

145

135

Length (mm) Length 125

115

105 5 10 15 20 25 30 35 40 45 50 Age (Months)

Fig. 4. Head length (mm) versus age (months) for male and female Devils from Mt William. Males blue triangle and females pink dots. Data from Pemberton, 1990: appendix 2.

Head Width vs Age in Devils

125

115

105

Length (mm) Length 85

65 5 15 25 35 45 55 Age (Months)

Fig. 5. Head width (mm) versus age (months) for male and female Devils from Mt William. Males blue triangle and females pink dots.

6/05/2016 8:41 AM 85

Rear Leg Length vs Age in Devils

170

160

150

140

130

Length (mm) Length 120

110

100 0 5 10 15 20 25 30 35 40 45 Age (Months)

Fig. 6. Rear limb length (mm) versus age (months) for male and female Devils from Mt William. Males blue triangle and females pink dots. Data from Pemberton, 1990: appendix 2.

Rear Foot Length vs Age in Devils

70 Length (mm) Length

60 5 10 15 20 25 30 35 40 45 50 Age (Months)

Fig. 7. Rear foot length (mm) versus age (months) for male and female Devils from Mt William. Males blue triangle and females pink dots. Data from Pemberton, 1990: appendix 2.

Two observations can be made about these plots. First, they show the marked sexual dimorphism that is evident in overall size and weight in Devils. Further, the dimorphism becomes evident at a very early age, usually before the age of 12 months.

Second, the inflection point in all measurements for both sexes occurs just before or just about the age of 24 months, that is, roughly the age at sexual maturity for most Devils. In other words, Devils appear to stop growing, in linear dimensions at least, just before they are ready to mate.

6/05/2016 8:41 AM 86

Breeding Participation

Maturity is one thing, but actual breeding success is another. For example, although most males are probably mature at two years of age, one researcher thought that males “may not mate until they are three or even four” (Jones, 2008), although no supporting evidence was provided. It has also been said that “… a few male devils dominate paternity” (Jones, unpublished, cited in Jones et al. 2004), but again the supporting evidence is yet to come. It is also thought “more dominant (aggressive) males are likely to achieve higher rates of paternity (M. Jones, unpublished data (Hamede et al., 2013).”

The results of an unpublished study (R. Hamede) were provided in graph form provide some insight into male participation in mating. The graph shows the mean number of partners (minimum of eight hours together spent within 30 cm of one another) adult males and females of different age classes had during the mating season (Jones et al., 2007a: fig. 1 and re-drafted here as Fig. 8). There were no sample sizes, dispersion statistics or accompanying statistical analysis so it is difficult to say how indicative the data are. Nonetheless, taking the results at face value suggests that second year males participate in more matings than males in other age classes, perhaps as a result of the combination of experience and vigour. First year males gain experience, while males older than two may lose vigour.

2.5

1.5

1 Mean Number of Mates

0.5

0 2 3 4 5 Age Class (Years)

Fig. 8. Mean number of mating partners of Devils of different sex (males, black; females, grey) and age (year) classes at Narawntapu National Park, Tasmania recorded using proximity loggers (paraphrased from legend to Jones et al., 2007a: fig. 1).

6/05/2016 8:41 AM 87

Not all mature females breed (carry pouch young) in any one year. In a ten year (1966-1975) study in an area in northwest Tasmania, the proportion of breeding adult females ranged from 31 to 80 percent, with a mean of 58.1 (Guiler, 1978). In a three year (1966-1968) study in an area in northeast Tasmania, the proportion of breeding adult females ranged from 47 to 100 percent, although only one population exceeded 80 percent (Guiler, 1970b). And in another three year study in the northeast, the proportion of breeding adult females varied from 69.7 to 76.7 percent (Pemberton, 1990). The basis of a general assertion that “about 75 % of females of breeding age carry young (Guiler, 1992)” is unclear. The lowest value in the combined observations (31 percent) means that as many as 69 percent of mature females in a local population may not breed in a particular year.

Occasionally, a female may never breed. In one study, a female trapped over five successive reproductive seasons never carried pouch young (Guiler, 1978).

At the northwest locality there was a strong positive correlation between mean female weight (all ages exclusive of pouch young) and the proportion of pregnant females (rs = 0.82, P < 0.01; data in Guiler 1978: tables 6 and 9). This suggests that the general condition of the females in any one year may partially determine how many breed.

The proportion of breeding adult females can also vary greatly among local populations within any one breeding season. In one year (1966) in the northeast area mentioned above, the range among five local populations was 47 to 80 percent (Guiler, 1970b).

If the graphical representation of the unpublished data on the number of matings by age class and sex is correct (above; Fig. 8), then females appear to participate in matings according to their vigour.

In captivity, breeding success of mature females is slightly lower than in the wild. In 2009 and 2010 in the mainland population, the overall breeding success, defined as the production of viable young, among mature females given an opportunity to mate was 53 and 40.5 percent, respectively (Keeley et al., 2012d). In the period between 2006 and 2011, the breeding success, defined as the production of weaned joeys, was 41.8 percent. In this period, there was no apparent significant difference in breeding success between wild-caught and captive-bred females (Hibbard et al., 2012; see also Table 18). In the period 2007-2011, the breeding success, defined as the production of joeys, was stated as 26.3-53.0 percent (Keeley and Eastley, 2012). In 2012, breeding success, defined as pouch young, was 41.8 percent (Hibbard et al., 2012). For other comments about the generally lower breeding success of captive females, see Jones et al., 2007a; Hesterman, 2007; Kent, 2009; ABC, Behind the News, 14 June 2011; Lees and Andrew, 2012).

As might be expected, breeding success differs between facilities. In the 2011/12 breeding season, breeding success, as measured by the presence of pouch young, was higher in free-range enclosures (below) than in zoos, 44.2 percent and 36.4 percent, respectively, a difference that may have been due to age differences in the two populations (Hibbard et al., 2012).

6/05/2016 8:41 AM 88

Breeding success varies among facilities of a similar kind. In 2012, for example, one small free-range enclosure (below) had a breeding success of only 20 percent while another small free-range enclosure had a 55 percent (6 of 11) success rate (Newsletter, September 2012; Hibbard et al., 2012). Another free-range enclosure achieved a 61 percent (14 of 23) success rate (data in Egan, 2012a; Hibbard et al., 2012).

It is important to make the distinction between the two different definitions of breeding success, that is, number of pouch young vs the number of weaned young, in the captive breeding program, because the latter may be much lower. For example, of the number of young born in free-range enclosures in Tasmania in 2010/11 only 65 percent survived to weaning (Hibbard et al., 2012). It would be interesting to know the source of this mortality.

It is also important to remember that breeding success varies with age, at least in captivity. Eighteen years of studbook data showed that amongst females, 34 percent of two and three year olds and 10 percent of four year olds reproduced (PHVA, 2008).

The reasons for the generally lower breeding success rate in captivity are unclear, but one keeper noted that obesity “negatively affect[s] reproduction (Slater, 1993).” How obesity affects reproduction is unclear, and its implications, if any, for wild animals remain to be investigated.

Criteria for Determining Age

Pouch young can be aged fairly accurately for about 140 days from the time they are fully furred (but still attached to the teat) using, in combination, the stage of tooth eruption, shank length and weight (Guiler and Heddle, 1974).

Free-ranging Devils can be aged fairly precisely up to three years of age by using a combination of dental criteria, such as canine over-eruption, molar eruption, tooth wear and the distance from the dentine-enamel junction to the gum (M. Jones, as reported in Loh, 2006; Jones, Sinn and Beeton et al., unpublished manuscript, as cited in McCallum et al., 2009; Lachish et al., 2009, 2010a; M. Jones, in Hamede et al., 2012a-b; Keeley et al., 2012b; Lane et al., 2012; Hamede et al., 2015), head width (Hamede et al., 2015), body weight and “general appearance” (Jones et al., 2008).

For example, only mature Devils have the fourth molar fully erupted (Jones et al., 2005) and Devils about three years old usually have the second molar worn down to a “blunt stump” (Jones, 1994).

The condition of the canines and molars is especially useful in age determinations. For example, the canines are reported to be “short/very sharp” in the first year; “long and sharp” in the second year; “long, robust and slightly blunt in the third year; “more blunt” in the fourth year and “thick and rounded” in the fifth year and older. All the molars are “sharp” in the first year; “tips worn off” M1 and M2 in the second year; M2 “one-third” denuded in the third year; M1 and M2 “increasingly blunt/rounded” in the fourth year and older (M. Jones, as reported in Loh, 2006).

In general, “the older the Devil, the more battle-scared it appears (Obendorf, 1993).” Older Devils may be missing distal limbs and digits and even an eye (Obendorf,

6/05/2016 8:41 AM 89

1993). And old male Devils may lose the hair on the snout and rump (Guiler, 1971, 1978).

In a reversal of the usual trend in mammals, the young and immature have more pointed snouts than adults do (Fleay, 1952; Hamede et al., 2008).

Learning Ability

There are no formal tests of Devils’ learning ability, but their capacity to learn has been demonstrated serendipitously both in the wild and captivity. In the wild, for example, trapping success tends to decline over time. In one northwest area, trapping success, measured as the number of animals per trap per night, declined from about 0.35 to about 0.06 over a six month period (rs = -0.85, P < 0.01; data in Guiler, 1970b: table 1). And at any one locality, trapping success often peaks after just a few days and then declines (Guiler, 1970a, 1978). This suggests that many animals find the trapping experience unpleasant, despite the food reward, and avoid the traps the more familiar they become with them.

In contrast, a few Devils enter baited traps night after night in order to get an easy meal, and after having fed will curl up to sleep while waiting to be released (Guiler, 1978; Jones, 1994). One male, for example, was caught in the same trap over six successive nights (Guiler, 1978) and another was caught 28 times, although over what period was not specified (Guiler, 1992). These “trap happy” animals seem to have had a different experience of the trap routine and are content to exploit its benefits.

Other trappers report that Devils can apparently learn how to remove the wired meat baits at the end of treadle traps without getting caught (Jones, 1995).

Interestingly, sentinel cameras at baited sites also show a significant decline over time in both the number of “encounters” and the number of images per encounter (Thalmann et al., 2014). The fact these cameras involve no physical stress to Devils suggests that novelty contributes to the initial interest in a camera site and this falls away over time.

Another example of Devils learning capacity is the ability of some to learn to come to a keeper’s call (Roberts, 1915) or whistle (Slater, 1993), a sound they probably associate with feeding.

Another possible indication of Devils’ intelligence comes from some experiments in exercise physiology when researchers tried to train animals to run on treadmills (Bell et al., 1983; Nicol and Maskrey, 1986). The experience of one lab was that “the temperament of Tasmanian Devils is such that they do not respond well to training, and some animals, after quite good initial runs, would progressively become more refractory (Nicol and Maskrey, 1986).” Perhaps the Devils’ post-experience recalcitrance is a sign of their intelligence.

Brain size relative to body weight is often taken as an indicator of comparative intelligence among similar species, say, all marsupials or all mammals. There are five studies concerning brain size in Devils, two dealing with the Devil only and three comparing the Devil with near relatives. Of the two Devil only studies, one reported

6/05/2016 8:41 AM 90

that three Devils weighing 4.4-5.9 kg had brain weights of 13.0-14.5 g (Sanderson et al., 1979) or about 0.25-0.3 percent of body weight (assuming a correlation between brain and body weight) and the other reported that three Devils weighing 8.2-10.1 kg had brain weights of 11.2-12.2 g or 0.12-0.14 percent of body weight (Patzke et al., 2014). The data support the expected general relationship of decreasing relative brain weight with increasing body weight.

The three comparative studies returned conflicting results. An early study comparing the relationship between braincase volume and braincase length among two species of quoll, the Devil and the Thylacine demonstrated graphically (no statistical analysis was provided) that the Devil had a slightly larger volume relative to the two quolls but somewhat smaller volume relative to the Thylacine (Moeller, 1970). A subsequent study using 34 specimens reported the relationship between log10 endocranial volume 0.249 (ECV) and log10 body weight (BW) was: ECV = 1.755(body weight) (Ashwell, 2008: table 2). And in contrast to the first comparative study, in a plot of these two parameters among Australian Region carnivorous marsupials (Dasyuridae + Myrmecobiidae + Notoryctidae + Thylacenidae), the Devil appeared to have a relatively small brain (Ashwell, 2008: fig. 1b). However, this may be due to the Devil’s more massive (and hence heavier) skull (and possibly other bones as well) rather than a truly smaller brain. Most recently, a PhD thesis asserted that among three species of quoll, the Devil and the Thylacine, the Devil had the smallest brain case relative to condylobasal length, but there was no supporting analysis (Jones, 1995, 2003a).

Finally, when presented with a mirror, Devils apparently think the reflection is another Devil (J. Clarke, on ABC AM, 20 October 2012; Devil Island, 2013a). It is interesting that Devils apparently recognise the image as another Devil, suggesting that an image alone is enough to recognise a conspecific. It is also interesting that Devil’s do not recognise the image as “self;” however, only a few other animals, such as some primates, cetaceans and proboscidians, all regarded as intelligent on other criteria, do recognise their own image.

“Personality”

One researcher describes Devils as having “loads of personality (Jones, 2008; see also M. Jones, in Bevilacqua, 2003b; Sunday Tasmanian, 13 September 2003; Fraser, 2005 and ABC Stateline, 23 September 2005).” And another says, “each has their own personality (G. Woods, on Sixty Minutes, 24 July 2008). A field researcher says, “Some are crazy bastards…. Some are slugs (N. Mooney, in Darby 2004; see also Ninemsn Sunday, 16 May 2004).”

Devil handlers and keepers also often express the opinion that the Devils in their charge show “personality” differences (E. Kirchner and T. Virgis, in Owen and Pemberton, 2005; D. Schaap, in Macey, 2008; M. Pepperday, in ABC World Today, 1 January 2009; S. Mulhall, in Tweed Daily News, 14 July 2009; T. Britt-Lewis, in Harris, 2010; J. Pelham, in Smith, 2010; G. Irons, in Cart, 2011; A. McLeish, in Dobbin, 2011; D. Cobbold, in Zaw, 2012; D. Saliba, in ABC News, 27 September 2013, Broadstock, 2015; N. Wilson, in Finlay, 2015; D. Reid, in White, 2015; see also Sinn et al., 2010; S. Fox, in Shannon and Lehman, 2015). And one experienced field worker has noted that personality traits are “passed on from one generation to the next

6/05/2016 8:41 AM 91

(M. Jones, in Parer and Parer-Cook, 2004). But whether these differences are a result of nature or nurture is unclear. One keeper noted that some young Devils were bold and others shy with regard to new items placed in their enclosures (C. Pukk, on ABC News, 6 December 2007).

Keepers also report that an individual Devil may respond to individual humans in different ways (Tweed Daily News, 14 July 2009).

Population Structure and Dynamics

Sex Ratio

Opinions about whether the sex of newly born Devils can be determined at birth are conflicting. One early field worker thought this was possible, drawing attention especially to the presence of the scrotum in males (Guiler, 1970b). One keeper also thought the male’s external genitalia could be readily distinguished (Oglesby, 1978). Two keepers monitoring the development of a litter of captive born Devils also thought sex could be determined from day one, but they gave no criteria (Phillips and Jackson, 2003).

In contrast, a developmental biologist disagreed and said that only at a head length of about 11.0 mm were the rudiments of the male scrotum and female pouch evident. Interpolating from the field biologist’s growth data these young were thought to be about ten days old (Hughes, 1982; see also Pemberton, 1990). Furthermore, field workers have said “sexing very small pouch young is not possible as they still possess an undifferentiated phallus (Lachish et al., 2009).”

Sex ratios in the wild

Conflicting views have been expressed about whether the sex ratio in pouch young in the wild is at parity or in favour of females. An early worker commenting on data from a population in the northeast wrote, “Females are in larger numbers than males in the pouch young … (Guiler, 1970a.)” He also said about this same population, “the sex ratio of … the pouch young … [favours] the female …. (Guiler, 1970b).” But a few years later when commenting on a population if the northwest, he noted there was no significant deviation from parity in the sex ratio of the pouch young (Guiler, 1978). Two latter field studies also indicated parity in the cumulative sex ratios of 16 litters from a population in the northeast (Pemberton, 1990) and of 44 litters from a broad area of Tasmania (Lachish et al., 2009). However, in a study of the Devil’s general reproduction, the sex ratio in pouch young was 27/48 (X2 = 5.88, P = 0.015; data in Hughes, 1982). Although these latter data indicate a strong sex bias in favour of females, the researcher considered “the small sample size was … inadequate for this to be firmly established (Hughes, 1982).” Analysis of the data in the literature for wild pouch young (Table 6), shows that the number of males (318) and females (359) is not significantly different from parity (X2 = 2.36, P = 0.12, pers. obs.).

Table 6. Sex ratios for wild Devils as reported in the literature. Number of asterisks (*, **) indicates 0.05 and 0.01 levels of significance.

Males/Females Ratio Location Date Maturity Reference

6/05/2016 8:41 AM 92

58/68 0.85 Cradle Mountain N/A ? Jones and Barmuta, 19 19/14 1.36 Icena 1964 ? Green, 19 26/37 0.70 Takone area 2011 ? STDP website: Mapping the disea ? 0.88 Tasmania ? ? Guiler, 197 143/126 1.13 Tasmania N/A ? Pearse et al., 20 28/32 0.88 Freycinet 1999 Ad/Immat Brüniche-Olsen et al., 20 25/22 1.14 Freycinet 2006 Ad/Immat Brüniche-Olsen et al., 20 14/20 0.70 Little Swanport 1999 Ad/Immat Brüniche-Olsen et al., 20 17/12 1.42 Little Swanport 2006 Ad/Immat Brüniche-Olsen et al., 20 17/21 0.81 Marrawah 1999 Ad/Immat Brüniche-Olsen et al., 20 17/19 0.89 Narawntapu 1999 Ad/Immat Brüniche-Olsen et al., 20 29/25 1.16 Narawntapu 2006 Ad/Immat Brüniche-Olsen et al., 20 18/18 1.00 Pawleena 1999 Ad/Immat Brüniche-Olsen et al., 20 15/7 2.14 Pawleena 2006 Ad/Immat Brüniche-Olsen et al., 20 15/10 1.50 Granville 1966 Adult Guiler, 19 64/74 0.86 Granville 1967 Adult Guiler, 197 14/15 0.88 Granville 1967 Adult Guiler, 19 11/9 1.22 Granville 1968 Adult Guiler, 19 20/25 0.90 Granville 1969 Adult Guiler, 19 10/8 1.25 Granville 1970 Adult Guiler, 19 11/8 1.38 Granville 1971 Adult Guiler, 19 18/10 1.80 Granville 1972 Adult Guiler, 19 31/34 0.91 Granville 1973 Adult Guiler, 19 49/39 1.36 Granville 1974 Adult Guiler, 19 35/26 1.35 Granville 1975 Adult Guiler, 19 48/31 1.55 Mt William 1984 Adult Pemberton, 19 42/28 1.50 Mt William 1985 Adult Pemberton, 19 30/31 0.97 Mt William 1986 Adult Pemberton, 19 23/23 1.00 Narawntapu 2006 Adult Hamede et al., 20 32/38 0.84 Freycinet 1999 Adult Lachish et al., 201 19/12 1.58 Little Swanport 1999 Adult Lachish et al., 201 17/18 0.94 Pawleena 1999 Adult Lachish et al., 201 3/12 0.25* Granville 1966 Immature Guiler, 19 2/8 0.25 Granville 1966 Immature Guiler, 197 4/11 0.36 Granville 1967 Immature Guiler, 19 6/8 0.75 Granville 1967 Immature Guiler, 197 0/4 N/A Granville 1968 Immature Guiler, 19 5/7 0.71 Granville 1969 Immature Guiler, 19 1/2 0.50 Granville 1970 Immature Guiler, 19 3/7 0.43 Granville 1971 Immature Guiler, 19 2/5 0.40 Granville 1972 Immature Guiler, 19 17/12 1.42 Granville 1973 Immature Guiler, 19

6/05/2016 8:41 AM 93

6/15 0.40* Granville 1974 Immature Guiler, 19 1/5 0.20 Granville 1975 Immature Guiler, 19 55/58 0.95 Mt William 1984 Immature Pemberton, 19 45/28 1.61* Mt William 1985 Immature Pemberton, 19 32/49 0.65 Mt William 1986 Immature Pemberton, 19 70/79 0.89 Cape Portland 1966 Pouch Young Guiler, 197 13/18 0.72 Granville 1966 Pouch Young Guiler, 197 12/8 1.50 Granville 1966 Pouch Young Guiler, 19 14/25 0.56 Granville 1967 Pouch Young Guiler, 19 9/12 0.75 Granville 1967 Pouch Young Guiler, 197 12/8 0.25 Granville 1968 Pouch Young Guiler, 19 11/27 0.41** Granville 1969 Pouch Young Guiler, 19 12/8 0.25 Granville 1970 Pouch Young Guiler, 19 10/6 1.67 Granville 1971 Pouch Young Guiler, 19 16/12 1.33 Granville 1972 Pouch Young Guiler, 19 29/14 2.07* Granville 1973 Pouch Young Guiler, 19 17/15 1.13 Granville 1974 Pouch Young Guiler, 19 8/10 0.80 Granville 1975 Pouch Young Guiler, 19 1/2 0.50 Icena 1964 Pouch Young Green, 19 26/33 0.79 Mt William 1984 Pouch Young Pemberton, 19 24/33 0.73 Mt William 1985 Pouch Young Pemberton, 19 7/1 7.00 Mt William 1986 Pouch Young Pemberton, 19 27/48 0.56* Tasmania ? Pouch Young Hughes, 19

The study in the northwest part of the state found a significant difference in the sex ratio in immature Devils, with females (80) much more numerous than males (42) (X2 = 11.84, P = 0.001; Guiler, 1978). Analysis of the data in the literature for all wild immature Devils also suggests females (231) are more numerous than males (182; X2 = 5.58, P = 0.02, pers. obs.).

An early worker commenting on a population in the northeast said “the sex ratio of … adults … [favours] the female …. (Guiler, 1970b).” Analysis of the data in the literature for all wild adults, however, indicates that the number of males (421) and females (371) is not significantly different from parity (X2 = 2.69, P = 0.10, pers. obs.).

The analyses above support the early observation of “… a change in the sex ratio from almost equal sexes in the pouch young to a female dominated juvenile population and then a return to a population which, although not significantly different from 1:1 ratio does show a tendency to male dominance in the adults (Guiler, 1978).” One explanation of these results is that with a bias toward male dispersal, dispersing juvenile and subadult males (see “Juvenile Dispersal” above) may experience greater mortality than juvenile and subadult females but reproducing adult females may suffer greater mortality than adult males.

6/05/2016 8:41 AM 94

It should be mentioned that the STDP has extensive data on the sex of Devils gathered over a number of years and from throughout Tasmania. These data would provide useful information on the temporal and spatial patterns in the Devil’s sex ratios.

Sex ratio in captivity

There is little information on sex ratios of Devils born in captivity. Adelaide Zoo reported a cumulative sex ratio of eight males and six females, which is clearly not significantly different from parity (data in Smith, 1993). Some zoos have reputedly reported either a male or female bias in offspring (Sinn et al., 2010), but no supporting details are available. Between 2006 and 2011, the sex ratio was in favour of males in two years, presumably significantly so, but approximately at parity in the other years (Hibbard et al., 2012).

Age Structure

In a northeast population (Mt William) over three successive years the population was relatively stable over the first two years but with relatively fewer juveniles and more subadults in the third year (Fig. 9).

Percent of Juveniles, Subadults and Adults at Mt William

100% 90% 80% 70% 60% Adults 50% Subadults 40%

Percent Juveniles 30% 20% 10% 0% 1983/84 1984/85 1985/86 Year

Fig. 9. Percent of juveniles, subadults and adults at Mt William, northeast Tasmania in the years 1983/84 to 1985/96. Plotted from Pemberton, 1990: table 4.1.

In a northwest population (Granville Harbour) over ten successive years, the trappable population, that is, animals one year and older, consisted of between 9.0 and 38.5 percent animals one year old and between 61.5 and 91.0 percent (mean = 76.2) of animals two years and older (Guiler, 1978: table 10; Fig. 10).

6/05/2016 8:41 AM 95

Percent of Subadults and Adults at Granville Harbour

100 90 80 70 60 50

Percent 40 30 20 10 0 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 Year

Fig. 10. Percent of subadult (white) and adult (black) Devils at Granville Harbour, northwest Tasmania in the years 1966 to 1975. Plotted from data in Guiler, 1978: table 10.

In another northwest population (West Pencil Pine), which has been mildly affected by facial tumour disease, the proportion of one year olds ranged approximately 18 to 40 percent and the proportion of two years and older ranged approximately 60 to 82 percent (data in Hamede et al., 2012: fig. 2: Fig. 11). Interestingly, in this population the proportion of one, two and three year olds each ranged more than two fold between years (data in Hamede et al., 2012a: fig. 2), which suggests that the population may be subject to some strong, fluctuating environmental influences.

Age Structure at West Pencil Pine

100% 90% 80% 70% 4+ years old 60% 3 years old 50% 40% 2 years old Classes 30% 1 year old 20%

Proportion of Age of Proportion 10% 0% 2006 2007 2008 2009 2010 Year

Fig. 11. The age structure in the West Pencil Pine population during winter (July- August). Re-plotted from Hamede et al. 2012a: fig. 2.

In the general northwest area of Tasmania in October 2009, shortly after DFTD arrived in the area, the population consisted of almost equal numbers of one, two and

6/05/2016 8:41 AM 96

three year old animals but many fewer four and five year olds (data in Fox et al., 2010b: fig. 18; Fig. 12).

Total Number of Devils Caught in NW Tasmania in October 2009

80 70 60 50 40 30 20

Total Number Captured Number Total 10 0 1 2 3 4 5 Age in Years

Fig. 12. The age structure in the general northwestern population in early October 2009, shortly after DFTD arrived in the area. Re-plotted from Fox et al., 2010b: fig. 18.

A interesting feature in these data on population structure is that with the subadult + adult group, in the one northeast population sampled, Mt William, the percentage of subadults was above 50 percent in all years samples whereas in the three northwest populations, Granville Harbour, West Pencil Pine and “NW Tasmania,” the percentage of subadults never exceeded 41 percent and in most years was much lower.

In likely reference to eastern areas, approximately half the presumably trappable individuals in local populations were said to be mature (Hawkins et al., 2008), meaning, presumably, that about 50 percent of a population consists of animals one year of age and 50 percent consists of animals two years and older.

And over two years just prior to the arrival of facial tumour disease, the trappable animals in one east coast population (Freycinet) contained between 51 and 53 percent of one year olds and 47 and 49 percent of two years and older (Lachish et al., 2009: fig. 2).

The difference between the western and eastern populations in the proportion of one year old vs two year and older animals (greater proportion of one year olds in the west) is intriguing but needs more testing.

Mortality and Survivorship

Pouch Young

6/05/2016 8:41 AM 97

Mortality of the pouch young was thought by an early field worker to be very low, possibly close to zero and hence survivor ship very high, possibly close to 100 percent (Guiler, 1970a-b, 1992). However, in the Mt William population in northeast Tasmania over three years, seven of 23 lactating females were lactating on fewer teats on subsequent recapture (Pemberton, 1990). Assuming, generously, that females started with four young each and the seven females lost only one young each, the resulting mortality of pouch young would be at least 7.6 percent (7/92). On the assumption of three pouch young each, mortality would rise to just over 10 percent (7/69; pers. obs.).

Mortality in the first year of life is unknown, due to the difficulty of assessing mortality during the time weaning young are in the den. However, mortality in the few months between the time young animals have left the den and the time they turn one year old has been estimated in the Mt William population to vary between 50 and 65 percent (Pemberton, 1990: table 4.7).

Juveniles

Mortality in “young devils,” presumably those one year old, has been estimated in a northwest population as varying from 45.0 to 94.8 percent (Guiler, 1978).

In the Mt William population, mortality in the one-year old cohort varies from 25 to 42 percent (Pemberton, 1990: fig. 4.7).

Survivorship in “juvenile devils,” that is the difference between the number of pouch young and the number of devils in the first year of life has been estimated at about 3 percent in a northeast population and 30 percent in a northwest population (Guiler, 1970a).

Subadults and adults

In the Freycinet Peninsula population in eastern Tasmania, annual survivorship among subadults (two-year old cohort) appeared to be about 46 percent and in adults (three years and older) about 51 percent (Lachish et al., 2007, as estimated from fig. 4; see also McCallum et al., 2007). But numbers applying to a graph showing the absolute size of age classes in the three years up to and including the first appearance of facial tumour disease in the population (Fig. 13) indicate that attrition between the one year and two year age classes varies from 55 to 62 percent and from the second to third year varies from 48 to 68 percent (Lachish et al., 2009: fig. 2 but actual data from STDP, 2011b: table 6).

6/05/2016 8:41 AM 98

140

120

100 4 years and older 80 3 years old

60 2 years old 1 year old 40 Number of Individuals of Number 20

0 1999 2000 2001 Year

Fig. 13. Number of individual Devils of each age class captured during winter (June/July) field trips each year on the Freycinet Peninsula. (Redrawn after Lachish et al., 2009: fig. 2 with data from STDP, 2011b: table 6).

In a modelling exercise, the annual survival probability for one, two, three and four year olds was taken as 0.45, 0.51, 0.54 and 0.26 (Beeton and McCallum, 2011: table s1). In general, these mortality rates, at least up until age four, are indicative of a Type II survivorship curve, that is, one in which attrition is more or less constant per unit time. A similar survivorship curve seems to pertain to Devils in captivity as well (Hibbard et al., 2012: fig. 2a).

Mortality rates in Devils 4 years and older is almost certainly higher than for younger adults (Jones et al., 2013).

There are conflicting views as to the sexual bias in survivorship. One early field study concluded that mortality rates were higher in females than in males (Guiler, 1978: fig. 11). But a latter study indicated that mortality rates were higher in males than in females (Pemberton, 1990). The latter determination, however, has been criticised as not being robust (Lachish et al., 2007).

Longevity

In the wild, Devils are thought to live rarely longer than six years (DPIWE, 2005b; Buchmann and Guiler, 1977; Guiler, 1978; Jones et al., 2008; Lachish et al., 2011a). In an early survey in the northwest, one male was thought to have been about eight years old at final capture (Guiler, 1978), and in a more recent survey also in the northwest, some animals were found to be “nearly seven years old” (S. Fox, in Mounster, 2010; Beeby, 2010 – as seven years old; Devil Island, 2013f – as just turned seven).

In captivity, one Devil was alive at seven years of age (Owen and Pemberton, 2005); another died at seven and a half (Owen and Pemberton, 2005; Fort Wayne Daily News, 15 February 2010); two were alive at nearly eight (Todd, 2008), and another was alive at eight (Nowak and Paradiso, 1983; Hesterman and Jones, 2009). Studbook

6/05/2016 8:41 AM 99

data show a maximum age of eight for females and 11 for males (PHVA, 2008; Lees and Andrew, 2012). And a 13-year-old male “from a wildlife park in Sydney” was mentioned as an experimental animal (Lang et al., 2005). As of October 2012, the oldest Devil in captivity is eight (Hibbard et al., 2012: fig. 2a-b).

Both the wild and captive data, therefore, indicate that few if any Devils live longer than eight years. An early encyclopaedic assertion that the life span is approximately seven to eight years (Walker, 1964) would be a maximum, not an average, estimate.

Senescence

It is not known at what age Devils living in the wild lose the ability to reproduce and hence whether any individuals live to this age.

In captivity, Devils six years and older are generally considered to be senescent (PHVA, 2008). Male Devils “generally cease breeding before age 6 (Lees and Andrew, 2012).” A captive male sired his last young at five years of age, although whether this reflected opportunity or ability was not clear (Owen and Pemberton, 2005). A 6.5 year old male had maturing sperm in the testes and epididymides (Matson et al., 2005), although there was no information about its overall quantity or quality. Female Devils”…usually cease breeding at age 4 (Lees and Andrew, 2012).”

Population Turnover

The only estimate of population turnover is for an area around Cradle Mountain where the annual turnover was 88.6 percent, a very high figure (Jones, 1998). In other words, year on year nearly 89 percent of the animals observed were new.

Social Interactions

Devils interact with each both at a distance and in proximity. They interact over space through vision, sound and smell. They can also interact over time through smell. The spectral sensitivities of Devils are unstudied, so it is unclear what wavelengths they respond to and how they see the world. It would be surprising, however, if Devils’ white and black markings were not just as distinctive and recognisable to them as to a human observer (see “Colour and Pattern). So too would be their body postures and gaits. Their vocalisations are as varied as any marsupial (see “Vocalisations and Other Sounds”), and although their loudest vocalisations would carry for long distances, all seem to be used only in close range interactions. Devils body odours would also convey information “up close” while their anal markings (see “Anal Dragging”, scats and latrines (see “Scats, Defecation and Urination”) would be informative across time

In terms of social interactions, in the wild, Devils are largely solitary animals (Jones et al., 1997). They usually occur together for any length of time in one of only three situations: feeding on a carcass; courtship and mating, and a female with her dependent young (Pemberton, 1990). Feeding at a carcass can involve up to 22 individuals (Pemberton and Renouf, 1993; see also Guiler, 1970c, 1983a, Jones, 1998); courtship and mating usually involves just a single male and female, and a family group includes a female no more than four young.

6/05/2016 8:41 AM 100

In captivity, Devils can show levels of sociability not evident in nature. For example, up to six subadults or even adults may sleep together in the same nest box day after day (Albion, 2006).

Interactions at Carcasses

Interactions at carcasses, even when staged for tourists, probably reveal completely natural behaviours, because they are undoubtedly part of life in the wild.

Although Devils often feed on carcasses by themselves, the proportion of time a carcass is visited by more than one Devil increases with the population density, as does the amount of biting (Hamede et al., 2008). When one Devil first finds a carcass, its crunching of bone and cartilage is enough to attract other Devils to the feeding opportunity (M. Jones, in Parer and Parer-Cook, 2004), and as other Devils arrive and begin to feed, their vocalisations serve as a further alert (Mooney, 1992).

There is a difference of expert opinion about whether or not Devils around a carcass form a dominance hierarchy. One field worker noted that a newcomer to a carcass was only challenged by the previous newcomer and once accepted became the challenger to the next newcomer (Pemberton and Renouf, 1993). If correct, this shows an interesting distribution of motivation.

In contrast, another field worker thought the Devils around a carcass might form a diffuse (not linear) dominance hierarchy. She noted, in effect, that when the ratio of Devils to carcass size is low, the animals tend to space themselves somewhat evenly around it with minimal bickering. But as the ratio increases (due to more Devils or less carcass), the bickering increases and dominance appears (Jones, 1994). And a Devil’s place in the hierarchy seemed to be determined by degree of satiation (Jones, 1994: Mooney, 1992), size (Jones, 1994) and perhaps reproductive condition, at least among females, with breeding females being dominant (Mooney, 1992). Young may actually reach the carcass but then be too fearful to feed (Jones, 1994).

An animal’s relative position in the hierarchy can be important, because it can “influence the proportion of bone, fur and flesh” it consumes (Jones, 1995, cited by Jones and Barmuta, 1998), that is, the quality of the food.

At carcasses, Devils are usually dominant over other species, notably those smaller than they are, such as the two sympatric species of quoll (Jones, 1998; Jones and Barmuta, 2000) and feral cats (STDP website: “Do devils suppress cats?,”as updated on 5 October 2010).

Among wild males, most animals one year old, that is, between 13 and 23 months, show relatively little scarring (2 percent), whereas animals two years and older show more frequent scarring (76 percent) (Pemberton, 1990). This pattern accords with observations that juveniles hang back at carcasses and the fact that they would not be involved in mating efforts.

Male-Male Combat

6/05/2016 8:41 AM 101

It is often assumed that during the breeding season males fight over females (Hamede et al., 2008). There is, however, no direct evidence for such behaviour in the wild (Hamede et al., 2008). Indeed, the principle in one field study of remotely monitored social interactions of Devils in the breeding season (Hamede et al., 2009) said, “I was expecting to see much more male-to-male fighting, but “… there were hardly any male-male encounters (BBC News, 19 August 2009).”

In captivity, males will fight (Roberts, 1915), but the circumstances are so uncontrolled that it is difficult to know the basis of the conflict, that is, access to females, defence of a familiar space or social dominance under confined conditions.

In one group of closely observed captive Devils, one of the five mature females showed indications of approaching oestrus, that is, a tailing off of her previous dominant position, and the three mature males started to become aggressive toward each other. Ultimately, the largest and oldest (4 y) “easily subdued” the other two (2 y) but “without any serious fighting (Kelly, 1993).”

Other observations indicate males will tolerate each other well. For example, two males kept in a divided box overnight, tore down the separator and were found curled up together asleep the following morning (Mooney, 2005; STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).

Courtship and Mating

Courtship and mating involve a few days of intense interaction between a male and female. But this phase of a Devil’s life has been observed only fleetingly in the wild. The statement that “a female loves and leaves several males in succession over a period of a week” (Jones, 2001) has no published details to support it.

Courtship and mating are much better known from observations in zoos, although in such circumstances, the interaction may be influenced by an inability of females to break away from an overly aggressive partner.

In captivity, females appear to dominate males during most of the year, sometimes biting them severely (Roberts, 1915). During the mating season, however, males become dominant (Roberts, 1915; Fleay, 1935; Turner, 1970; Guiler, 1971; T. Virgis, in Owen and Pemberton, 2005; E. Kirchner, in Owen and Pemberton, 2005; Slater, 1993; Smith, 1993). In some cases, the male will drag the female around the enclosure by the scruff of her neck (Slater, 1993; E. Kirchner, in Owen and Pemberton, 2005). Critically, females regain their customary dominance by the time they have young, which need defending.

In captivity, a sexually-interested male Devil ingested the fresh faeces of a mature female cage mate (Eisenberg et al., 1975). And the same male sniffed the female’s urine spots prior to interactions with her (Eisenberg et al., 1975; Eisenberg and Golani, 1977).

In captivity, males will fight when put together after the mating season (Roberts, 1915). And one female was reported as killing both males and females (Fleay, 1935).

6/05/2016 8:41 AM 102

During the mating season, an active and successful male may lose up to one-quarter of its body weight (Parer and Parer-Cook, 2004; Jones et al., 2012; see also STDP Annual Report, 2012/13).

Female and Young

Almost everything that is known about the interaction between the female and her young and the interactions among the young comes from observations on closely confined animals.

In disputes over food in captivity, a female may push her threatening young away or occasionally even butt them violently with her head (Buchmann and Guiler, 1975).

In captivity, litter mates may form hierarchies shortly after weaning. A litter of four, for example, formed a linear hierarchy within three weeks of “full weaning,” and showed most of the behaviours of adults in a similar hierarchy (Buchmann and Guiler, 1975).

With one exception, males appear to take only a predatory interest in their own young (Roberts, 1915; Harz, 1966). One early keeper, however, reported that “males wash and clean the young in a solicitous way (Turner, 1970).”

This same keeper also reported that he had successfully fostered three young Devils abandoned by their mother with a domestic cat (Turner, 1970). Both observations are so unusual, they need confirmation to be accepted.

Social Significance of Facial Scarring

Some observers note that among wild adults, males have more facial scarring than females (Fisher Exact Probability Test P = 0.023, pers. obs., data on adults, scar scores combined within each sex, in Pemberton, 1990: table 2.5; Guiler, 1992; Pemberton and Renouf, 1993; Kreiss, 2009: fig. 1.1f). The wounds leading to such scars almost certainly arise out of social interactions among Devils, but the nature of those interactions is unclear. Perhaps the most likely explanation is they reflect greater aggression among males at carcasses, if for example, males, being larger, need to spend more time feeding (Pemberton and Renouf, 1993).

Importantly, among wild juveniles, males also have more facial scarring than females (Fisher Exact Probability Test P = 0.021, pers. obs. data on juveniles, scar scores combined within each sex, in Pemberton, 1990: table 2.5). This suggests the wounds leading to the scars are not inflicted during male combat, a presumably adult behaviour, if it exists in Devils.

Just weaned Devils show no scarring at all (Pemberton, 1990: table 2.5).

Odour

Although young Devils do not have much of an odour, at least to a human observer,

6/05/2016 8:41 AM 103

adults have a distinctive odour, variously described as: “musty, waxy” (Owen and Pemberton, 2005; also Buchmann and Guiler, 1977); “… fur with a lanolin smell and a slightly pungent, musky, doggy aroma on its feet” (M. Jones, in Connellan, 2005; see also Fleay, 1935; Mooney, 1992; Jones, 1994; M. Jones, in Grzelewski, 2003). Males are said to emit a stronger odour than females (Guiler, 1992).

It makes sense that young Devils have no odour, because they need to avoid being detected by predators and are too young to be involved in adult social interactions for which odour may be important. In contrast, adults engage in both intra- and inter- sexual activities in which odour may play an important role.

Devils are also said to emit a distinct scent when frightened, although the nature of the scent and its source were unspecified (M. Jones, in Grzelewski, 2003).

Anal Dragging

Devils drag their hind-quarters along the ground in a variety of social contexts: courtship, hierarchy formation and feeding in the presence of other Devils (Fleay, 1935; Eisenberg et al., 1975: fig. 2b; Buchmann and Guiler, 1977; Lazenby and Dickman, 2013: fig. 4). There are two pairs of two glands near the anal opening (Chatin, 1876; M. Jones, in Connellan, 2005), and presumably during dragging, the gland’s secretion is spread on the substrate. Presumably, this behaviour, also known as ano-genital and cloacal dragging, leaves an odour on the substrate, which can be interpreted by other Devils (Eisenberg et al., 1975; Buchmann and Guiler, 1977). A secretion could contain information about an individual’s sex, state of maturity and reproductive status.

The behaviour begins in unweaned young (c. 140 to 210-214 days, according to various authors) and continues throughout life in both sexes (Buchmann and Guiler, 1977; Eisenberg and Golani, 1977; Kelly, 1993; Pemberton and Renouf, 1993).

In captivity, anal dragging occurs almost inevitably in a social context, that is, when two or more Devils are put together (Buchmann and Guiler, 1977; Eisenberg and Golani, 1977; Kühme, 1983), usually in a way they never would be in the wild. In these circumstances anal dragging may precede or elicit an attack; precede, interrupt or follow a display, and be performed by either a “winning” or “losing” individual. However, as a dominance hierarchy forms in captivity, the incidence of anal dragging diminishes in the subordinate members (Buchmann and Guiler, 1977).

In captivity, anal dragging also often seems to be associated with aggression over food. An individual may drag around or over food, and an individual may drag even with the food in its mouth (Buchannan and Guiler, 1977).

In wild, anal dragging has been observed when two or more Devils are feeding on a carcass (Pemberton and Renouf, 1993). In this context, the behaviour has two parallels with what occurs in captivity: individuals in an aggressive interaction due to an unusual degree of proximity and in association with contested food. It has also been observed at sites baited with food and food odours, but the method of observation (automatic camera) was not able to determine if conspecifics were nearby

6/05/2016 8:41 AM 104

(Lazenby and Dickman, 2013: fig. 4), as the other observation of anal dragging in the wild might suggest.

If anal dragging does involve a secretion useful in social communication, it is interesting that there are no observations of direct anal sniffing among Devils.

Anal dragging leaves no scent detectable to a human observer (Kühme, 1983).

Vocalisations and Other Sounds

Vocalisations are difficult to describe due to lack of agreement on how sounds are rendered into words. For this reason, vocalisations are usually recorded and then analysed from sonographs, which display the frequency and amplitude of the sound. Vocalisations are also difficult to interpret in terms of behaviour, because they can be used in different contexts.

Devils are said to have the “largest vocal repertoire identified for any marsupial (Eisenberg et al., 1975).” Early work on two pairs of captive Devils identified ten vocalisations – huff, hiss, click (emitted with snorting or huffing), snort, bark, growl, moan, screech, snort and whine and three mechanical sounds – click (emitted by jaw clapping), lip smacking and foot stamping (Eisenberg et al., 1975: fig. 6; Eisenberg and Golani, 1977).

Later work, on adult animals feeding on a carcass in the wild, identified 11 vocalisations: bark, click (single or in a rapid series), crescendo, growl 1, growl 2, growl 3, humph-growl, shriek, snort, whine and yip (Pemberton and Renouf, 1993: figs 3-4 and table 3).

Some of the vocalisations almost always occur together in a specific sequence. The best known complete vocalisation, for example, is that given by two animals in a high intensity confrontation, often involving food, such as a carcass. This vocalisation has been described, variously, in terms of the call types mentioned above as: hiss – growl – whine – growl – terminal shriek (Eisenberg et al., 1975; Eisenberg and Golani, 1977) or growl 1 – growl 2 – growl 3, crescendo and shriek (Pemberton and Renouf, 1993). Note the different renditions for this well known, indeed partly defining, Devil vocalisation.

Devils will often respond to a human-made devil-like sound with a similar sound (Eisenberg et al., 1975). Devils will also respond in kind to a sound made by a physically separate Devil. The function of such imitation is unclear, but may serve to indicate the level of competency on the part of the responder to a query or challenge thrown out by the initiator.

Some vocalisations can carry over long distances. Calls broadcast into the bush can bring animals out from cover and elicit response calls (Eisenberg et al., 1975; Buchmann and Guiler, 1977).

And for what it’s worth, in one of those “my-tax-money-paid-for-that!?” kind of experiments, researchers in the UK found that out of 34 sounds ranked as the most

6/05/2016 8:41 AM 105

unpleasant, the Devil’s “sound,” which was unspecified but likely to have been its scream (shriek), came in at number 11 (ABC News Online, 26 January 2007).

There seem to be few observations on vocalisation in very young Devils. In captivity, young just over 15 weeks of age and just off the teat utter “anxious whimpering, yapping cries, similar to those of a very young puppy” when lifted away from their mother (Fleay, 1935). And young just old enough to leave the pouch may “cry” if left behind in the nest when the female leaves with the other young in her pouch (Roberts, 1915). Young a few days older and more independent may make “a faint attempt at a bark (Roberts, 1915).” Young at 20 weeks of age uttered “high-pitched moaning snarls” when handled (Fleay, 1935) and at 22 weeks and almost weaned made “whining howls and sharp sniffs (Fleay, 1935).”

Agonistic Behaviour

The Devil’s agonistic behaviour has been observed in both captivity and the wild. In captivity, such behaviour has been studied mostly in animals (cagemates) that have been confined more closely and over a much longer period than would ever occur in the wild. In one study, the animals were a male - female pair, which were occasionally put in the same cage, thereby engendering some “get-to-know-you-again” agonistic interactions (Eisenberg et al., 1975; Eisenberg and Golani, 1977). And in another, they were “captive devils observed over several years” in a university facility (Buchmann and Guiler, 1977). In the wild, the behaviour was studied at a large staked out carcass, thus replicating a natural situation (Pemberton and Renouf, 1993).

Low level interactions

When Devils engage each other, they have a number of ways of showing intent and perhaps capacity. For example, when aroused, blood perfuses the Devil’s external ears, the pinnae, which are only sparsely haired, turning them a deep reddish hue (Buchmann and Guiler, 1975; Eisenberg et al., 1975; Jones, 1994). Similarly, the hair, especially on the tail, may become erect (Buchmann and Guiler, 1975; Eisenberg et al., 1975: fig. 2c). Note that ear profusion is also used in thermoregulation.

In the wild at least, a Devil arriving at a carcass where other Devils are feeding may stand off and simply sit and stare or sit on its rear legs and tail with front legs relaxed (Pemberton and Renouf, 1993). The latter behaviour may simply enable the newcomer to get a better look at the overall scene.

In the early stages of an interaction, a Devil may present its head or entire body laterally to the antagonist (Buchmann and Guiler, 1975; Eisenberg et al., 1975: fig. 2c; Pemberton and Renouf, 1993). And one Devil may initiate contact by simply pushing the other with its shoulder (shouldering) or rump (rumping) (Eisenberg et al., 1975; Pemberton and Renouf, 1993; Jones, 1994). Another form of apparent low level contact involves stabbing the chest of an opponent with the front feet (Pemberton and Renouf, 1993).

Early in an interaction, a Devil may gape (Eisenberg et al., 1975; Eisenberg and Golani, 1977; Guiler, 1983a; Pemberton and Renouf, 1993) and champ the jaws (Guiler, 1983a). It may also hiss; huff; snap the jaws creating a clapping sound, and

6/05/2016 8:41 AM 106 lick the lips making a smacking sound (Eisenberg et al., 1975; Eisenberg and Golani, 1977).

Occasionally at a carcass in the wild, a newcomer will run straight up to a feeding Devil, causing it to flee (Pemberton and Renouf, 1993). Perhaps the sheer confidence of the arriving animal is enough to send a message to the displaced animal, or alternatively, perhaps the displaced animal recognises the arriving one and acts on a memory of a past interaction with it.

In general, the intensity of vocalisation, especially growling, increases with the intensity of the interaction, both on the rise and the descent (Eisenberg et al., 1975; Eisenberg and Golani, 1977).

Hesitancy

In the initial stages of a conflict, Devils may show several behaviours that have been interpreted as being indicative of hesitancy or fear. Such behaviours observed in captivity include extension and shortening the body and treading movements of the front feet (Buchmann and Guiler, 1975).

Yawning (Fleay, 1952: fig. p. 276) is another open mouth display, but in contrast to a gape, which seems to indicate aggressive intent, yawning is thought to indicate stress (Wood, 2003d) or fear (Buchmann and Guiler, 1975; Grzelewski, 2002) and is often a precursor to flight (Buchmann and Guiler, 1975). In captivity, yawning first occurs at about 22 weeks of age (Kelly, 1993). Yawning also occurs in captive adult Devils when they are confronted with prey (laboratory rats) they find intimidating, one rat even crossing “safely between the yawning jaws of an adult Devil (Buchmann and Guiler, 1977).”

In the wild at a carcass, Devils have been observed to raise one foot off the ground, while holding the head up and the tail down (Pemberton and Renouf, 1993). This posture seems tentative and hence might indicate hesitancy.

Appeasement

In the early stages of an interaction, one Devil may appease the other by offering its cheek (Eisenberg et al., 1975: fig. 3d), which may elicit only a feigned bite or light nip (Pemberton and Renouf, 1993).

Fighting

The only actual trigger for a fight that has been identified is, oddly enough, sneezing - “a sharp, exhalant noise with a somewhat querulous note.” The call is said to be most often emitted first by the eventual winner of an encounter and can be returned by the other animal (Buchmann and Guiler, 1975).”

In an intense interaction, two individuals may begin to grapple with each other in a crouching or standing posture and then rise up on their rear legs, grasping each other with their front legs, bringing their open mouths into proximity while emitting the characteristic growl – scream call (Eisenberg et al., 1975: fig. 4c; Pemberton and

6/05/2016 8:41 AM 107

Renouf, 1993). The combatants may lock their jaws briefly (jaw wrestling) or back down without actually making jaw contact (Buchmann and Guiler, 1975; Pemberton and Renouf, 1993). A male-female pair familiar with each other may simply end the bout with mutually licking (Eisenberg et al., 1975).

In the wild, biting, which almost certainly occurs in agonistic encounters, leads to flesh wounds on the head and body (manifest as scars), broken teeth and even damaged skull bones (Buchmann and Guiler, 1977). In some cases, fresh wounds may suppurate freely and old wounds may permanently distort facial features (Green, 1967). As one very experienced field worker observed, “facial wounds and abscesses are common in older devils (N. Mooney, in DPIW, 2007a).”

When several Devils are around a carcass, there are many open-mouthed threats, much piercing screaming and occasional nipping at exposed parts. Some Devils back into a carcass, thus presenting their backsides to receive any greeting nips and bites (Mooney, 1992; Jones, 1994). It is perhaps as a consequence of being bitten on the rump when being chased from a carcass as a younger animal (below) and backing into it as an older one that older Devils can have well-scarred rumps (Mooney, 1992: fig. p. 59).

Disengagement

A fight usually ends when the “loser” backs away. The dominant animal usually just watches the retreating animal. But it may also chase the loser a short distance (Buchmann and Guiler, 1975; Pemberton and Renouf, 1993; Jones, 1994) or even pursue it for tens of meters, biting its rump and both animals vocalising (Pemberton and Renouf, 1993; see also Mooney, 1992). A fleeing animal may have “its tail in the air and fur fully erect” (Pemberton and Renouf, 1993; see also Jones, 1994; DPIPWE, 2010d). Hence, erect hair is a sign of both aggressive intent (above) and fear.

Agonistic play

Not surprisingly, the play of young Devils involves behaviour that forms the basis of agonistic behaviour in adults. For example, young Devils at an age of about 20 weeks and almost weaned will rush at each other with open mouth (Fleay, 1935).

The observation that juvenile Devils drag one another by the ear (Owen and Pemberton, 2005) is interesting as this is usual a behaviour used by temporarily dominate males when controlling temporarily subordinate females during mating. It would interesting to know if this behaviour is more common in juvenile males than in females or equally frequent between the sexes.

Social hierarchies

In captivity, two adult Devils kept together for prolonged periods usually form a lasting dominant – subordinate relationship and agonistic interactions between them diminish (Eisenberg et al., 1975; Buchmann and Guiler, 1977). Such relationships can even be achieved without physical contact, as becomes evident over time when two animals are kept in separate cages close together (Buchmann and Guiler, 1977). Size

6/05/2016 8:41 AM 108

is an important determining factor in an animal’s position (Buchmann and Guiler, 1977).

In captivity, if a subordinate animal comes under attack, it may “stretch out with [the] underside hard against the ground (Buchmann and Guiler, 1975).” In the wild, similar behaviour may be shown by a newcomer trying to get a place at a carcass: “the newcomer … [lies] down, belly on the ground, with fore and hind feed extended anteriorly and posteriorly (Pemberton and Renouf, 1993).”

In the wild, Devils are generally solitary and hence the issue of social hierarchies does not arise. There are two situations, however, in which they may occur. In captivity, just weaned young appear to form a hierarchy in (Buchmann and Guiler, 1975), suggesting that weaning and just-weaned young in and around their natal den may do so as well.

In the wild, the Devils feeding around a carcass may also form a hierarchy, although as mentioned, this is a matter of expert difference of opinion.

Grooming

Devils groom themselves, but they only groom one another during one aspect of courtship.

When Devils wash themselves, they put the front paws together making a cup-like depression, thoroughly lick them and rub them over the face (Roberts, 1915; Rose et al., 2006).

It is not clear if Devils groom the rest of their body. Devils feeding inside a rotting carcass can emerge “gooey (Jones, 1994),” but how they deal with the goo on their body fur is unknown.

In captivity, Devils will enter their water dishes (Roberts, 1915; Fleay, 1935; Guiler, 1971 as well as accept water being poured over them (Roberts, 1915). Young Devils are said to play in their water trough (Turner, 1970). This behaviour seems to be a way of keeping cool, but whether it also has a cleaning aspect is unclear.

During courtship, a pair of Devils may lie on their sides facing each other and mutually lick their partner’s muzzle, check and ear. This mutual grooming may last “from thirty or forty seconds or longer” (Eisenberg and Golani, 1977; fig 1h). Its function is unknown.

A captive-bred female treated with parturition-promoting hormones “… briefly cleaned her urogenital opening” (Rose et al., 2006). However, other females both in the control and in the experimental groups showed no such behaviour, nor did males (Rose et al., 2006).

Predators of Devils

6/05/2016 8:41 AM 109

Predation is difficult to preserve, especially in largely nocturnal prey. And one animal eating another is not evidence of predation, unless it is known that the animal doing the eating will not touch carrion.

Native predators

Devils have no known native predators (Guiler, 1978; Jones 1998). But it seems likely that at least Tiger Snakes (Notechis scutatus), Masked Owl (Tyto novaehollandiae), the Spotted-tailed Quoll (Dasyurus maculatus) and wild dogs might take young Devils (DPIPWE, 2010d; STDP, 2011b). Larger Devils might also take small Devils (Guiler, 1978). The recently extinct Thylacine might have also been a predator of Devils (Mooney, 1992).

Exotic predators

Domestic dogs can easily kill an adult Devil (Widowson, 1829; Launceston Examiner, 23 March 1917; Green, 1967; Guiler, 1970a, 1978, 1983a, 1992; Jones, 1995; Jones et al., 2014). There is one account, possibly apocryphal, of an Australian kelpie sheep dog in a kennel with her litter of pups apparently killing six Devils bent on eating her young (Guiler, 1992). In captivity, dingoes can easily kill Devils (The Argus, 14 May 1946). Indeed, the STDP lists “predation by poorly controlled dogs” as one of several human impacts on Devils (STDP Annual Report, 2012/13).

The assertion that Devil “adults have been known to kill dingoes (Devil Island, 2013b)” is unsubstantiated in the literature.

It is difficult to know how many Devils are killed by dogs annually. There is an unsupported report of about 20 dogs killing about 50 Devils, although the length of time and the geographical area of this mortality was not included (Hawkins et al., 2008). But in that these reports were made voluntarily with some risk and little to be gained, the actual number of death by dog must be much higher.

Apparently, one of the worst times of year for dogs killing Devils is during the Easter long weekend when people flock to their week-enders in the bush with their pet dogs. This is also the time when young and naïve Devils are dispersing (D. Pemberton, in video entitled: “Racing to save the Tasmanian devil,” episode 9 in Earth Touch News Network’s “Earth Touch Insider.”

Foxes, if they were ever to become established in large numbers, would also be potential predators on young Devils.

Human predators

Although there is good evidence from what is now mainland Australian that Aboriginals killed Devils for both material items and food, it is not clear how widespread this practice was. There is no pre-historical or recent evidence from Tasmania that Aboriginals killed Devils for any use.

The early settlers killed Devils by trapping, shooting and poisoning, and although trapping and shooting are now largely prohibited, poisoning continues. In the early

6/05/2016 8:41 AM 110

days, people used strychnine, but when this was banned in the early 1950s, they switched to 1080, which is much less lethal to Devils.

Between 1941 when Devils received their first-ever legal protection to 2002 when they were given full protection, Devils could be killed under permit and the first such permit was issued in August 1960 (Guiler, 1982). Indeed, in 1974, in one district alone (Auburn), 250 Devils were killed “as a control measure (Guiler, 1978),” although a subsequent publication seem to put the year as 1973 and the total killed as 192 (Guiler, 1982: table 1). In the mid-1990s, it is thought that about 10 000 Devils per year were killed, most illegally. Illegal killing continues but the numbers are now in the 100s (N. Mooney unpublished, in DPIPWE, 2010d; STDP, 2011b). The last conviction for illegal killing was in 1998 (DPIPWE, 2010d).

Today, the main concern is not with poisoning targeting Devils specifically but as “collateral damage” in targeting other species, mainly with 1080. Devils usually respond initially to 1080 poisoning by vomiting, which onsets within about 20 to 80 minutes of ingestion. After vomiting, some individuals may recover, especially those that vomit sooner rather than later. Those that do die, however, do so in distress (McIlroy, 1981). Devils are much more resistant to 1080 than are other carnivores, both exotic and indigenous, in Tasmania. For example, the LD50 for Devils is 4.24 mg/kg but 3.72 for Eastern Quolls, 0.10 mg/kg for foxes and 0.06 mg/kg for wild dogs (DPIPWE, 2010b; see also Mooney, 1992).

Humans not only poison Devils directly but also indirectly when Devils scavenge the bodies of the animals for which the poison was intended, usually rabbits or wild dogs.

Roadkill

The advent of the motor vehicle was a new source of human-caused mortality, although in this case, largely inadvertent. And it didn’t help that the density of the road network matched closely the density of Devils. It is not surprising, therefore, that vehicles account for a large proportion of the recorded mortality for Devils. For example, vehicles accounted for up to 50 percent of Devil mortality at Cradle Mountain over 17 months and 20 percent at Freycinet National Park over 12 months (Jones, 2000; Jones, in Hawkins et al., 2008). More broadly, the statewide roadkill in 1998 was estimated to be about 5000 Devils (N. Mooney unpublished, in DPIPWE, 210d). And a survey between June 2001 and November 2004 estimated the annual roadkill over the entire state was about 3390 Devils or 3.8 – 5.7 percent of the population (Hobday and Minstrell, 2008).

The Devil’s dark pelage makes in one of the most difficult animals to see on the road at night from a moving vehicle (Hobday, 2010).

As roads improve, a constant priority in the bush, they become more lethal to wildlife, including the Devil. As a winding, bumpy, narrow dirt road becomes a straight, smooth, wide, tarred road the number of vehicles increases, the average vehicle speed goes up and the number of animals killed rises (H. McCallum, on the ABC, World Today, 11 June 2010). When Woolnorth Road and Arthur River Road, both in northwest Tasmania, were sealed, roadkill increase by 400 and 50 percent,

6/05/2016 8:41 AM 111

respectively (unpublished, S. Plowright and G. King, respectively in DPIPWE, 2010d).

Roads are attractive to Devils because they provide easy travel and are a source of food – other roadkills. These incentives along with the steady immigration of “new” Devils moving in to take up the “vacancies” creates a mortality sink for Devils.

Roadkill can perhaps be reduced somewhat by taking pre-emptive measures, such as erecting cautionary signage (STTDP website: “Sobering roadkill trials,” 15 May 2008; Tasmanian government media release, 16 October 2010), reducing the speed limit (Jones, 2000), making road surfaces light in colour (Darby, 2013; Dawtrey, 2013), using noisy road surfaces (Darby, 2013; Dawtrey, 2013; Richards, 2014), installing drainage that discourages vegetation growth along road edges (Dawtrey, 2013) and erecting warning posts and lights in hot spots (“virtual fences”) (Richards, 2014; McIntyre, 2015; Mather, 2016a; Vinales, 2016).

The latest effort to reduce the road kill in Devils, at least those to be released from captivity into the wild, is negative conditioning. This involves instilling in pre-release Devils a fear of roads and/or motor vehicles (Vinales, 2016).

The concept of “hot spots” for Devil road kills raises the issue of just why these spots are so “hot.” There seems to have been no research into makes these spot “hot” for Devils and how they might be made cooler.

Predator Defence

When confronted by a human in the wild, Devils may “freeze” (Denholm, 2010b; S. Fox, in Beeby, 2010), try to slip away or put on an impressive display. When taking the quiet way out, Devils may slowly creep away until the human observer moves when they “freeze,” only to resume their slow retreat when the human observer stops (Jones, 1994). But if startled, they may flee with tail raised and fur standing on end (Mooney, 1992; Jones, 1994).

When confronted by a human in captivity, Devils may threaten mildly by gapping (Guiler, 1971; Eisenberg et al., 1975), snapping the jaws to make a clapping sound (Eisenberg et al., 1975) and licking the lips making a “soft smacking sound (Eisenberg et al., 1975). A more severe threat may involve a rush and a “bark, followed foot stamping or snorting interspersed with barks (Eisenberg et al., 1975).”

In captivity, young Devils at about 20 weeks of age and off the teat but still not fully weaned will make sharp sniffing noises and high-pitched moaning snarls and gape when handled (Fleay, 1935).

No doubt, the Devil’s full-blown defensive display helped convince people with little close association with it that it was a dangerous animal. But while it is true that an aroused Devil could inflict a very nasty bite, it is also true that even wild-caught Devils can become “quite docile and easily handled” (Roberts, 1915; Green, 1967; also Jones, 1994). Devils’ ability to modify their behaviour to suit the circumstances may be an indication of their level of intelligence.

6/05/2016 8:41 AM 112

A media story mentioned that a wild Devil survived despite being bitten “several” times (Hoggett, 2012). This raises the intriguing possibility Devils may have some immunity to tiger snake venom. This would not be surprising if Devils frequently preyed on tiger snakes.

Parasites and Diseases

Like every other vertebrate species, Devils have both external and internal parasites. Some of these are quite general in their taxonomic distribution; some are restricted to the Devil and its close relatives, and some are specific to Devils.

Interestingly, “counting parasites” was and perhaps still is a part of the standard assessment procedure of devils captured in monitoring the spread of the disease (Johnston, 2007). But as yet, none of this information has been published.

Ectoparasites

As might be expected, Devils in the wild carry a variety of ectoparasites, that is, parasites on the outside of the body. To date, these include one flea, Uropsylla tasmanica; one mite, Satanicoptes armatus, and three ticks, Ixodes fecialis, I. holocyclus and I. tasmani (Green, 1967; Beveridge and Spratt, 2003; Vilcins et al., 2009).

Fleas

The endemic Tasmanian flea, Uropsylla tasmanica, is unusual in its biology, because it spends almost its entire life cycle on the host, the larval and pupal stages occurring in the subcutaneous tissue of the host (Dunnett, 1970; Pearse, 1981; Williams, 1986, 1991; Beveridge and Spratt, 2003). This is in contrast to most fleas, on domestic animals for example, where the immature stages fall off the host used by the adult.

This species occurs only on Thylacinus cyanocephalus, the Thylacine; Sarcophilus harrisii, the Devil; Dasyurus maculatus, the Tiger Cat, and D. viverrinus, the Eastern Quoll (R. Warneke, pers. comm., in Obendorf, 1993; Beveridge and Spratt, 2003). The status of an earlier report of the flea occurring on D. geoffroii (Williams, 1991) is unclear.

Occasionally Devils can be highly infected with fleas, presumably this species. One early observer said a dead Devil so swarmed with fleas that “they formed a continuous brown under-coat all over its body … and the place … where it was laid down for a few moments, became covered with them … in myriads … (Meredith, 1853.)”

Mites

Two mites have been described from Devils and, to date, are known only from Devils.

Diabolicoptes sarcophilus was recovered from the faeces of a wild Devil held in captivity for ten weeks and was presumed to be a skin parasite from either itself or

6/05/2016 8:41 AM 113 one of its food items while in captivity (Fain and Domrow, 1974; Fain and Laurence, 1975)

Satanicoptes armatus was first recovered from patches of sarcoptic mange on an animal in the London Zoo (Fain and Laurence, 1975). The areas of mange caused by the mite can be especially large in older Devils (L. Stevenson, in Owen and Pemberton, 2005; Jones et al., 1997).

Ticks

Of the three species of ticks that occur on Devils, Ixodes fecialis occurs widely in eastern and southern Australia and has been taken from a variety of marsupial and placental species. I. holocyclus, the paralysis tick, extends along the eastern seaboard of Australia from Cape York Peninsula to northeast Tasmania; it also occurs on a number of marsupial species. I. tasmani, the marsupial tick, is the most common tick in Tasmania and very widespread in Australia; it likewise occurs on a number of marsupials.

All three tick species are possible vectors for are variety of potentially pathogenic microbes that could inhabit the blood of vertebrates. To date, however, only two of the three species have actually been found to carry potentially disease-causing microbes. A survey of 44 Ixodes tasmani recovered from an unspecified number of wild Devils, found that 55 percent carried the rickettsia, Rickettsia tasmanensis (Izzard et al., 2009; Izzard, 2010; see also Steel 2010). And another study of 44 samples recovered from 35 wild Devils and potentially containing both Ixodes holocyclus and I. tasmani found that 45 percent of the samples carried another rickettsia of the Rickettsia massiliae group; 34 percent carried a hepatozoon and 14 percent carried both microbes (Vilcins et al., 2009: see also Steel, 2010). It is not known, however, if the infected ticks picked up their microbial pathogens from the Devils from which they were sampled or from another species (Vilcins et al., 2009).

Rickettsias cause a variety of diseases in humans and other animals, such as cats and dogs. Hepatozoons occur in most tetrapod vertebrates. They are usually benign in most host species, although they can cause disease in dogs.

Endoparasites.

Only three groups of metazoan endoparasites are known to occur in wild Devils: cestodes (tapeworms), nematodes (round worms) and trematodes (flukes). Devils also carry a number of microbial blood parasites.

An article on captive Devils but written by a researcher with extensive field experience of wild Devils cites three species of cestodes, two of nematodes and two of trematodes, as having been recorded from Devils. But it is unclear if these were from wild or captive individuals (Guiler, 1971). The difference is important, because captive Devils can pick up parasites they would not ordinarily have in the wild.

Cestodes

6/05/2016 8:41 AM 114

Five species of cestodes occur in Devils. Anoplotaenia dasyuri was first described from an animal that died in captivity in the Gardens of the Zoological Society in London (Beddard, 1911; see also Beddard, 1915; Cameron, 1931 and Obendorf and Smith, 1989). However, it occurs in almost all wild Devils (293 out of 294 examined) and usually in large numbers, often a 1000 or more (range: 2 - 5600). There is a strong positive correlation between Devil body weight and the number of worms, suggesting that Devils continue to acquire worms throughout life. There is no significant difference in the number of worms in males and females.

The fact that the worms occur in large numbers without the Devils showing any noticeable ill-effects suggests that the parasite, even in large numbers, is relatively harmless (Sandars, 1957b; Gregory et al., 1975; Gregory, 1976; see also Kelly, 1993; Ladds, 2009) and that Devils have little trouble feeding the worms as well as themselves. Interestingly, the metacestode, immature, stage of the parasite does not occur in Devils

Among native species, mature Anoplotaenia dasyuri occur in the Spotted-tailed Quoll, Dasyurus maculatus, and the Eastern Quoll, D. viverrinus, as well as in Devils (Obendorf and Smith, 1989; Sandars, 1957b). Among exotic species, mature worms also occur in cats and dogs (Gregory et al., 1975).

Among native species, metacestodes (immature stage) of Anoplotaenia dasyuri occur naturally in only some species, such as macropods, that are likely to be scavenged by Devils, but not others, such as bandicoots, possums, rodents and wombats. They also occur in the two species of Tasmanian Dasyurus and in Thylacinus, but not the Devil itself (Obendorf and Smith, 1989). The immature stages also do not occur in exotic animals such as rabbits and deer, which might also be scavenged by Devils (Beveridge et al., 1975; Gregory et al., 1975).

Dasyurotaenia robusta was also first described from a Devil that died in the Gardens of the Zoological Society in London (Beddard, 1912). It has apparently been found in only one place in the wild in Tasmania, Collings Gap, northwest of Hobart (D.L. Obendorf, pers. comm. in Beveridge and Spratt, 2014). The worm inhabits the Devil’s small intestine (Beddard, 1912, 1915; Beveridge, 1984; Beveridge and Spratt, 2003). As the species occurs only in Devils, its fate is tied to that of its host, a situation recognised by it too being listed as an endangered species (Hawkins et al., 2008; Beveridge and Spratt, 2014).

The relationships of both Anoplotaenia and Dasyurotaenia among cestodes are obscure (Beveridge and Jones, 2002).

Spirometra erinacei. Rudolphi, 1819; Mueller, 1937; Spratt et al., 1991;

Taenia pisiformis. This widespread species of tapeworm appears in a list of worm parasites in the Devil based on one register number in the national worm collection in Adelaide (Spratt et al., 1991). No further details are available.

Traces of cestodes are common in scats found in the bush (Fleay, 1952).

6/05/2016 8:41 AM 115

Devils do not seem to be a natural host to any species of tapeworm affecting sheep (B.L. Munday pers. comm. in Howkins, 1966; Gregory, 1972, 1976), despite the fact that they eat the entrails of dead animals. Furthermore, attempts to infect experimentally Devils with the immature stages of three species of tapeworms involved with debility in sheep, Echinococcus granulosus, Taenia hydatigena and T. ovis, were successful only in the case of T. hydatigena (Gregory, 1976). It is possible, in fact, that Devils and other dasyurids are refractory to Echinococcus granulosus (Jenkins, 2006).

Nematodes

Eight species of nematodes occur in Devils. The strongylidan Angiostrongylus cantonensis is known from captive Devils (Samuel, in Munday, 1988).

The ascarididan Baylisascaris tasmaniensis has as its definitive hosts only the Devil and the two species of quolls in Tasmania (Sprent, 1970; Sprent et al., 1973; Obendorf, 1993). The nematode’s intermediate hosts include Brush-tail Possums (Obendorf, 1993) and wombats (Munday and Gregory, 1974). A North American species of the genus can infect humans, and it has been suggested the Tasmanian species may also have this potential (Bradbury, 2015).

The filarioid Cercopithifilaria johnstoni is recorded for the Devil, but with no additional details (Ladds, 2009). The worm is carried by ticks and hence occurs in a large number of mammals (Ladds, 2009).

The spiruridan Physaloptera sarcophili was described from the Devil without any reference as to its anatomical location (Johnston and Mawson, 1940). Apparently, it has not been recorded in any other species subsequently.

The spiruridan Cercopithifilaria johnstoni. Baker and Chabaud, 1982; Mackerras, 1954; Spratt et al., 1991

The spiruridan Cyanthospirura dasyuridis. (Mawson, 1968; Spratt et al., 1991)

The strongylidan Woolleya sarcophili was described from a Devil that died in the Scottish Zoological Park (Cameron, 1931, in the genus Nicollina). The parasite has not been reported subsequently in any other Devil or species, raising the question whether it occurs naturally in the Devil or was picked up from another species in the zoo.

The tricocephalidan Trichinella pseudospiralis occurs in both its larval and adult stages in the muscles and intestine, respectively, of Devils (Obendorf et al., 1990; Obendorf, 1993; Booth, 1994; Alford et al., 1998; Cuttell et al., 2012). The parasite also occurs in the two species of quolls in Tasmanian and, in very low frequency, the Brush-tail Possum. It does not seem to occur in the native macropods or the exotic cats, pigs or rodents. This taxonomic distribution suggests that the life cycle is confined almost entirely to the three extant native carnivores, through scavenging, predation, cannibalism or coprophagy (Obendorf et al., 1990).

Trematodes

6/05/2016 8:41 AM 116

Three species of trematodes have been recorded from Devils. Mehlisia acuminata (Johnston, 1913; Spratt et al., 1991).

Neodiplostomum diaboli was described from preserved specimens found in a collection (Dubois and Angel, 1972; Spratt et al., 1991; Cribb and Pearson, 1993). The species has also been recorded from Dasyurus viverrinus (Cribb and Pearson, 1993).

Neodiplostoma sarcophilus was described (originally in the genus Fibricola; see Cribb and Pearson, 1993) from specimens from the “stomach and intestine” of one or more devils “from Tasmania” (Sandars, 1957a), suggesting they were of wild provenance. This parasite may have been the same as the “Alaria sp.” mentioned earlier in the literature from a Devil that died in the Scottish Zoo (Cameron, 1931; Sandar, 1957; Spratt et al., 1991; Cribb and Pearson, 1993). Other than these two instances, the parasite has not been reported again. If the species is truly restricted to Devils, it should be listed as endangered as is the Devil-specific Dasyurotaenia robusta (above).

Microbes

A small percentage of Devils may carry the cysts (muscles) or sporocysts (intestine) of a single-celled apicomplexan in the genus Sarcocystis or a near relative (Munday and Mawson, 1978; Munday et al., 1978). These observations suggest that at least a few Devils could be both an intermediate and definitive host for the protozoan. There is no evidence that the parasite has a debilitating effect on Devils.

Some Devils also test seropositive for another apicomplexan, Toxoplasmosis gondii (Hollings et al., 2013; see also Munday, 1966). This intracellular parasite was probably introduced into Australia through the introduction of the cat. Devils are probably infected by eating other animals containing the parasite rather than picking it up through oocysts in the ground (Hollings et al., 2013). The parasite has no obvious debilitating effects on Devils.

The bacterial flora in the oral cavity of Devils is likely to be rich, but to date only a few groups have been identified. Antibodies to the bacterium Leptospira have been detected in wild Devils (Wynwood et al., 2015), but despite the finding being called “leptospirosis” (University of the Sunshine Coast media release, 18 August 2015 and subsequent media stories), the disease caused by the bacterium, apparently no disease effect has been observed.

Both wild (Obendorf, 1993) and captive (Kelly, 1993) Devils carry Salmonella bacteria, although the Devils appear to be asymptomatic.

A few species of Pasteurellaceae have been recognised in captive Devils (Brix et al., 2015; also Georghiou et al., 1992). The bacteria showed a range of susceptibilities to the 22 anti-microbial agents tested (Gutman et al., 2016).

A broad survey of the bacterial flora of the oral cavity of Devils, foreshadowed in July 2015, should extend greatly the knowledge of these microbes (Taylor et al., 2015).

6/05/2016 8:41 AM 117

The lungs of 25 wild Devils were surveyed for the spores of the fungus Emmonsia, a pathogen common in burrowing animals in the Northern Hemisphere, and all were negative (Munday, 1966; Green, 1967).

Four Devils have been examined for the presence of trypanosomes, and all were negative (Hamilton et al., 2005a-b; Thompson et al., 2014).

Devils carry a unique herpesviruses, Da-VH-2. There is no association between DFTD and the presence of the virus, but captive Devils are significantly more likely to be infected than wild Devils. There is no evidence the virus causes disease in Devils (Stalder, 2013; Stalder et al., 2015).

A study of the blood smears for 187 wild Devils found no hemoparasites (Peck et al., 2015).

Commensal microbes

The availability of high-throughput sequencing machines, made it possible in 2015, for researchers to survey in wild and captive Devils, using 16S ribosomal RNA sequences, the myriad bacteria commonly associated with large organisms, most of which are essential for the health of those animals. The survey included the oral cavity, gut (sampled via faecal material), skin (chest-abdomen) and in-active pouch. It found 37 major groups of bacteria, the most common five of which were Actinobacteria, Bacteroidetes, Firmicutes, Fusobacteria and Proteobacteria. In wild Devils and in the oral cavity, the proportions of major groups Bacteroidetes, Firmicutes, Fusobacteria and Proteobacteria. In the gut, the proportions of the major groups were similar to other carnivores. And the skin and pouch skin were similar in their bacterial compositions. Interestingly, captive Devils differed significantly in the bacterial compositions in all body areas, oral cavity, gut and skin (both areas) (Cheng et al., 2015).

Diseases

In the wild, Devils may develop skin sores (Guiler, 1971); cysts, both external (nipple; Guiler, 1978; Obendorf, 1993: fig. p. 44) and internal (uterus; L. Hughes in Guiler, 1978), and haemorrhoids (Guiler, 1971). There is also something that field workers call “scabby groin (Beniuk, 2013).” And one wild-caught Devil had pneumonia (Jones et al., 2012).

Older Devils may show evidence of osteoarthritis, “rheumatism” and neuro-muscular disabilities (Guiler, 1978, 1992; Kelly, 1993; Obendorf, 1993; Holz and Little, 1995), which can affect their movements.

In captivity, Devils have are subject to a variety of diseases, including infections, perforated ulcers and cancers. In terms of infections, septicaemia associated with an intestinal tract infection was fatal (Guiler, 1971). A perforated duodenal ulcer was also fatal (Guiler, 1971; Kelly, 1993). Skin tuberculosis has been recorded, with an unreported outcome (Kragic et al., 2015). Two small skin infections presumably caused by a common soil bacterium and possibly incurred during fighting, were

6/05/2016 8:41 AM 118 treated successfully with antibiotics (Reppas et al., 2010; also P. Holz, in Michael and Sangster, 2010; Kragic et al., 2015).

An intestinal bacterium of the genus Helicobacter has been recorded in Devils, but there was no reported associated pathology (Coldham, 2004; Ladds, 2009).

Cancers observed in captive Devils include those of the adrenal gland (Guiler, 1971; Munday, 1988; Kelly, 1993; Holz and Little, 1995; Koncza and Pryor, 2001), intestine (Guiler, 1971), brain (Kragic et al., 2015), central nervous system (Holz and Little, 1995), liver (Griner, 1979; Tovar et al., 2011), histiocyte – an immune system cell (Tovar et al., 2011), lung (B. Rideout, in Loh, 2006), lymph nodes (Griner, 1979; Munday, 1988; B. Rideout, in Loh, 2006), mammary gland (Effron et al., 1977; Griner, 1979; Koncza and Pryor, 2001; B. Rideout, in Loh, 2006; Australian Registry of Wildlife Health, 12 November 2007; Tovar et al., 2011), ovary (Tovar et al., 2011); perianal gland (B. Rideout, in Loh, 2006), “pouch area (Koncza and Pryor, 2001),” sebaceous gland (Effron et al., 1977), skin (Patske et al., 2014; Scheelings et al., 2014; Kragic et al., 2015), thyroid (Lombard and Witte, 1959; Effron et al., 1977; Griner, 1979; B. Rideout, in Loh, 2006), skin (Ratcliffe, 1933; Lombard and Witte, 1959; Effron et al., 1977; Griner, 1979; Montali, 1980; Canfield and Cunningham, 1993; Holz and Little, 1995; Ladds et al., 2003; Matson et al., 2005; K. Rose, in Loh, 2006; Tovar et al., 2011), smooth muscle (Griner, 1979) and stomach (Ladds et al., 2003).

An early news item based on an interview with a researcher at the University of Tasmania noted “the Tasmanian devils which have died in captivity have all been found to have died from cancer (Anonymous, 1969)”, although it is not clear what animals (University of Tasmania research labs?) or what numbers of animals it was referring to. One early researcher even suggested, presciently perhaps, that Devils might be genetically predisposed to the development of cancer (Griner, 1979).

Cancers observed in wild Devils include those of the adrenal gland, bone, connective tissue, lymph nodes, mammary gland, sebaceous gland, skin and stomach (Loh, 2006).

Cancers are thought to be “common” in dasyurids, including Devils, in both the wild (Munday, 1988) and captivity (Griner, 1979, 1983; Canfield et al., 1990; Booth, 1994, and P. Holz, in Edwards, 2007; see also Anonymous, 1969; A. Kelley, in Millikin, 2003; S. Pyecroft, on Radio National Earthbeat, 6 December 2003; M. Jones, ABC Science Show, 3 September 2011, and M. Jones, video on The Australia newspaper website, 7 September 2011; Jones, 2012). Indeed, it has been said that Devils are particularly susceptible or prone to cancers (A. Papenfuss, in Borrell, 2009; DPIW, 2007b; Brown et al., 2011; van der Kraan, 2011; Brown, 2013; B. Ujvari, in Azvolinsky, 2015).

It is difficult, however, to know what to make of these assertions (Twin and Pearse, 1986), because the basis of comparison is rarely stated and never tested statistically. And in the case of captive animals, the claims may not take into account the conditions under which the animals were kept. For example, the fact that Devils in zoos are usually forced to be outside during the day in order to support the commercial interests of the institution means the animals are exposed to levels of

6/05/2016 8:41 AM 119

solar radiation with which they have not evolved. The practice is likely to increase their chances of developing more radiation-induced cancers than would be the case with more diurnal marsupials such as kangaroos.

Curiously, the STDP website seems to support explicitly diurnal display, because it asserts, along with video evidence, that “Tasmanian Devils love nothing better than to play in the sun and then cool down with a dip in the ‘pool’ (STTDP website: “Devil summer – a swim and a ball game,” 24 February 2011).”

Captive devils can also have arterial degeneration (R. Parsons, in Anonymous, 1969; Munday, 1988), cardiomyopathy (Ratzke et al., 2014; Kragic et al., 2015), blocked kidneys (Munday, 1988), meningitis (Blanchard, 1991, as cited in Obendorf, 1993), multiple sclerosis (Cerabona, 1997) and prolapsed intervertebral discs (Ratzke et al., 2014). Indeed, very old Devils that have spent most of their lives in captivity may show a variety of maladies (Michael and Sangster, 2010).

Interestingly, if even some of the high-incident Devil diseases, such as cancer, are the result of institutions keeping the animals under conditions to which they are not adapted, such as prolonged exposure to solar radiation (Twin and Pearse, 1986), there might be a case to answer under animal welfare laws.

Be that as it may, all the diseases listed above pale into insignificance in comparison with a disease first noticed in 1996 and now sweeping through the wild population: an infectious cancer cell line causing what has come to be known as Devil Facial Tumour Disease. This disease and what is being done about it is the subject of the last two chapters of this document.

In recent times due to the species’ facial tumour disease many more Devils have been brought into protective captivity, and as a consequence, “… new medical conditions have been observed which have not been previously reported in this species (Peck et al., 2015).” It is not clear if these conditions are due to captivity or have only become recognisable in captivity.

Physiology

Physiologists take special interest in the Devil due to its position due to its large size and calm temperament, which make it easy to handle, the animals sometimes even falling asleep in the experimental chamber. As the largest living marsupial carnivore, it provides a window onto the even larger marsupial carnivores of the past such as the Australian thylacoleonids. And as the largest living dasyurid weighing 6-8 kg, the Devil provides the “must have” heavy anchor for all weight based comparisons within this diverse family. The first study of Devil physiology was the quantification of several blood parameters published in 1966.

Metabolic Rate

In the field, there was no significant difference in the metabolic rate, as measured by 22sodium turnover, between a 7.1 kg winter animal (407 kJ/kg/day) and a 7.9 kg summer animal (328 kJ/kg/day) (Green, 1997: tables 9.2-3).

6/05/2016 8:41 AM 120

In captivity, the lowest mean metabolic rate, as measured by oxygen consumption in one male Devil during the middle of its inactive period, was 0.28 ml/g/h and occurred at an ambient temperature of 31.0º C and a body temperature of 36.8 º C (MacMillen and Nelson, 1969). The mean metabolic rate for five Devils during the inactive period and at 20º C was 0.29 ± 0.02 ml/g/h. After acclimatisation for two weeks at 3-5º C and tested again at 20º C, the rate went up significantly to 0.33 ± 0.03 ml/g/h. This suggests that Devils have a capacity for non-shivering thermogenesis, although the physiological basis for the increased heat production is unclear (Kabat et al., 2003).

When awake, a Devil’s metabolic rate is lowest at an ambient temperature of about 28º C and rises on either side, although the rate of increase is greater and more variable with cooling than with warming, probably because the animals shiver periodically at lower temperatures, which raises the rate. When asleep, the same pattern holds, but the rise during cooling is less than during waking (Nicol and Maskrey, 1980; fig. 2).

When awake, a Devil’s metabolic rate is fairly uniform at any particular ambient temperature. When asleep, however, the rate cycles, and the amplitude of the cycle increases with decreasing temperature. Within in each cycle, increases are associated with shivering (Nicol and Maskrey, 1980; fig. 1).

When the Devil’s ambient oxygen levels are experimentally halved from about 20 percent to about 10 percent, its metabolic rate rises from 2.5 W/kg to about 4.3 W/kg. This is probably due to the increased breathing rate and cardiac output mounted to deal with the diminished oxygen and raised blood carbon dioxide content in the blood (Nicol, 1982: fig. 1).

The Devils is said to “appear” to show “shallow daily torpor (Nicol, 1982),” but the evidence for this in the original work (Nicol and Maskrey, 1980) is obscure.

Body Temperature and Thermoregulation

The Devil’s body temperature is related to ambient temperature at the high end of the ambient range and probably at the low end as well. Hence, it is useful to distinguish between body temperatures recorded without accompanying ambient temperatures and those with them.

Body temperature without regard to ambient temperature

In the wild, the only body temperatures for Devils are from four animals measured with telemetry in the central highlands. These animals’ body temperatures ranged between 31.3 and 37.5º C, with a mean of 35.7º C (Jones et al., 1997). Importantly, this “in-the-wild” temperature range encompasses most of the other observed body temperatures, all of which were made in experimental situations (maximum = 38.1 º C).

In the three experimental situations unlinked to ambient temperature measurements, the body temperatures of conscious, resting Devils averaged 37.6 º C in one adult (Robinson and Morrison, 1957); ranged 36.1 - 37.7 º C in one male (Morrison, 1965), and ranged 36.5 - 38.1 º C in four males and 34.6 - 38.0 º C in nine females (Parsons

6/05/2016 8:41 AM 121

et al., 1970), while the body temperatures of anaesthetised Devils subject to an invasive procedure ranged 33.0 - 34.9 º C in two adult males and one lactating female (Parsons et al., 1970).

When ambient oxygen levels are experimentally halved, that is from about 20 percent to about 10 percent, the Devil’s body temperature falls from 36.6 ºC to about 35.8 º C. This is probably due to the increased evaporative water loss due to an associated increase in the breathing rate (Nicol, 1982: fig. 1).

Body temperature with regard to ambient temperature

The most comprehensive analysis of body temperature in relation to ambient temperature involved three Devils, two males and a lactating female in both wakeful and sleeping states observed over ambient temperatures ranging from 2 to 44 º C. The resulting “response curve” was an upward opening, left hand truncated parabola with its vertex at a body temperature of about 34 º C and an ambient of about 20 º C. From this vertex, body temperature rose to about 36.5 º C at a minimum ambient of about 2 º C and to about 38 º C at a maximum ambient of about 44 º C. And at any one ambient temperature, the results showed a range of about 2 º C, although this was less when animals were held for long periods, 7 h, at any one temperature (Nicol and Maskrey, 1980: fig. 3).

This “response curve” suggests that Devils will not let their body temperatures fall below about 34 º C. It also shows that they will let their body temperatures rise, probably passively, with rising ambient temperatures, at least up to a point, which may not have been tested by the available range of ambient temperatures. And most interestingly, it shows that as ambient temperatures drop below about 20 º C the animals actually increase their body temperature, probably by shivering, up to at least about 36.5 º C. Why they don’t just maintain their temperature at the lowest tolerated level of about 34 º C is unclear.

In contrast to the parabolic response to ambient temperature described above, in another experiment over ambient temperatures between 11 and 39 º C a male Devil was graphed as showing a linear rise in body temperature from about 36.5 to 37.4 º C (Morrison, 1965: fig. 8). In other words, there was no detectable decline in body temperature over intermediate ambient temperatures as was the case in the parabolic response. However, there were only three data points, and they were on either side of the critical low area in the parabolic response curve.

In an experimental setting and at an ambient temperature of 31 º C, the body temperature in a single adult male Devil averaged about 36.8 º C (MacMillen and Nelson, 1969). This body would be at the high end of the range for its ambient temperature in the parabolic response curve.

In another experiment, at an ambient temperature of 19-21 º C an adult had a body temperature of 34.5 º C and at 39-41 º C a temperature of 39.0 º C (Hulbert and Rose, 1972). The former body temperature is well within the range for its ambient temperature in the parabolic response curve, but the latter temperature would be at the high end of the range for its ambient temperature.

6/05/2016 8:41 AM 122

In two adult males under ether anaesthesia and subject to an invasive procedure, one at an ambient temperature of 18.5 º C had a mean body temperature of 33.3 º C and the other at an ambient temperature of 20.1 º C had a mean body temperature of 35.5 º C (Guiler and Heddle, 1970: tables 1-2). Both body temperatures are well within the range of body temperatures observed at those ambient temperatures in the parabolic response curve.

In terms of energy expenditure in regulating their body temperatures, Devils probably save energy by letting their body temperature rise with ambient temperature and only begin expending energy by panting (below) when their body temperature approaches dangerously high levels. But Devils appear to expend more energy than seems necessary to actually raise their body temperature by shivering with falling ambient temperature. In terms of metabolic rate, Devils appear to achieve the greatest economies in energy expenditure over an ambient temperature of about 20 to 30 º C, with a peak perhaps at about 30 º C. This is probably the ambient temperature range in which they probably spend much of their lives, that is, in their diurnal shelter sites year round and in their nocturnal activity times from late spring to early autumn.

Daily variation

The Devil’s body temperature varies cyclically during the day, with the highs occurring at night when the animals are normally active and the lows in the day when they are normally inactive. In the wild, body temperature in two males and two females, in aggregate, was slightly higher in the evening when the animals were active and slightly lower during the day when they were inactive, but with a mean amplitude of only 0.6 º C. The daily patterns are “surprisingly stable” throughout the year (Jones et al., 1997).

In captivity, the Devil’s body temperature varies during the day in the same way, whether the animals are kept under ambient, that is, naturally fluctuating, temperatures (Guiler and Heddle, 1974: figs 2 and 3) or constant temperatures (Guiler and Heddle, 1974: fig. 4 and table 2). There is, however, a great deal of variation within individuals. For example, the mean hourly body temperatures of a large male recorded for 12-17 days at ambient room temperature had standard deviations ranging 0.8-2.4 (mean = 1.7; data from Guiler and Heddle, 1974: fig. 3). And the significantly different mean dark (35.6 º C) and light (35.0 º C) body temperatures (Mann-Whitney U = 138.5, P < 0.001) for this same animal suggest a daily amplitude similar to that observed in the wild.

Interestingly, results from the observations just described suggest that much of the within hourly variation in the single Devils body temperature may be accounted for by the within hourly range of ambient temperatures (r2 = 0.71, P < 0.001; data from Guiler and Heddle, 1974: fig. 3). Assuming that the within hourly ambient temperature range reflects the within hour variability of those temperatures, it would appear that much of the variability in the Devil’s body temperature might be due to the variability of its ambient temperature. This, along with would suggest further that the Devil thermoregulates only loosely instead of precisely.

In contrast to the daily pattern observed in both wild and captive animals, an early observation on the daily variation in the body temperature of a captive male found a

6/05/2016 8:41 AM 123

high of about 37.7 º C shortly before mid-day and a low of about 36.1 º C around midnight (Morrison, 1965: fig. 4), with an amplitude of 1.6 º C. The apparent phase shift in this animal was attributed to its having been borrowed from a zoo and therefore having an activity cycle adapted to the day-active viewing public (Guiler and Heddle, 1974).

The daily temperature rhythms in Devils are probably entrained by their normal activity times, but just how quickly they can be re-entrained by a shift in activity time, as in the case of the zoo animal just mentioned, is unclear.

Body temperature and activity

In experimental situations, Devils’ body temperatures increase with exercise, rising to as much as 4-5 º C over resting levels, with a recorded maximum of 41.7 º C (Nicol and Maskrey, 1986). After exercise, body temperatures return to normal in about 40 – 60 minutes (Nicol and Maskrey, 1986), apparently with no ill-effects.

Heating and cooling mechanisms

Devils warm themselves by shivering (Nicol and Maskrey, 1980; also Nicol and Maskrey, 1986) and cool themselves by panting (Hulbert and Rose, 1972; Nicol and Maskrey, 1980; Nicol, 1982; but see Guiler and Heddle, 1970 for failure to record panting). Even when sleeping, Devils will shiver at low ambient temperatures, but they do so only in slow wave sleep, not rapid eye movement sleep (Nicol and Maskrey, 1980). Interestingly, there are no observations of Devils licking themselves to keep cool.

In terms of their own body temperatures, Devils start panting at about 37 º C, and body temperatures “rarely” exceed 38 º C, even at ambient temperatures in the high 30s and low 40s (Nicol and Maskrey, 1980; Nicol, 1982). Such, apparently, is the efficiency of panting.

There are only a few other body temperatures with associated ambient temperatures and these fall within the general parabolic pattern described above. Two adults at an ambient temperature of 19-21 º C, for example, had a body temperature of 34.5 º C, and at an ambient temperature of 39-41 º C, it rose to 39.0 º C (Hulbert and Rose, 1972).

Shivering as a way to raise body temperature is only a short-term response to brief periods of cold. When kept under cold conditions for longer terms, Devils can raise their basal metabolic rate and presumably also their temperature without shivering. Non-shivering thermogenesis is common in eutherians, but it has difficult to demonstrate in metatherians. It appears, however, that Devils have this capacity, because when kept at 3-5 º C for two weeks, their basal metabolic rate increased by 11 percent (above). In eutherians, the heat produced in non-shivering thermogenesis is usually generated by the metabolism of brown fat. But Devils appear to lack such fat, leaving the muscles as the most likely place for the increased heat production (Kabat et al., 2003).

6/05/2016 8:41 AM 124

In captivity, Devils may become extremely lethargic on hot days (Ewer, 1969), which suggests that they are either inhibited in or actively reduce their activity. In either case, the decreased activity would minimise heat gain.

In an early experiment, an adult Devil was kept at 40º C for six hours to see how it would cope with this relatively high ambient temperature. The animal’s rectal temperature rose over the first half hour to about 39.4º C, but over the next two and a half hours it fell back and over the last three hours was close to what it was before the animal was moved to a warmer environment (c. 37.4 º C). The animal’s respiratory rate followed a similar pattern. How the animal kept its body temperature below ambient was not clear, as it was not observed to pant (at least after its temperature had returned to near pre-experimental levels) or lick its fur (Robinson and Morrison, 1957). The suggestion at the time was that the animal must have been sweating (also Morrison, 1965), but subsequent work showed that Devils do not sweat (Hulbert and Rose, 1972; Bell et al., 1983). If this animal truly did not pant, it is difficult to see how it could have maintained its body temperature below ambient.

In another experiment, a Devil with a body temperature of 39 º C suffered heat collapse and had to be removed to a cooler temperature. This was possibly the experiment in which the body temperature of an animal kept for two hours at 50 º C, as this was the only reported constant experimental temperature above 35 º C, the only reported fluctuating ambient temperature above 37.8 º C and the only reported body temperature above 37.9 º C. Given that some of the body temperatures in this experiment approached or even reached at least 37.9 º C and in one case attained the stressful 39 º C, it is interesting that there was no observation of “panting, tongue lolling or licking (Guiler and Heddle, 1974).”

It is worth noting that there is no mention in the literature of Devils panting with the mouth open, as in dogs. Panting, that is, an increase in the breathing rate in association with high temperature, appears to occur only with the mouth closed. It is possible, therefore, that observations of a lack of panting in Devils might hinge on this difference in what constitutes panting.

Under hot conditions, experimental animals will wallow in their water dishes (Ewer, 1969) and drink large amounts (Robinson and Morrison, 1957). And zoo animals are said to “enjoy bathing” and will do so communally if the water container is large enough (Guiler, 1971; see also Roberts, 1915; Oglesby, 1978; Edwards, 2006). Devils probably wallow and bathe in order to stay cool (see T. Faulkner, on ABC Newcastle, 10 February 2011) as they become used to being out during the day. Indeed, it is difficult to imagine wild Devils, active at night, ever needing to cool down by bathing. Importantly, wild Devils newly brought into captivity remain nocturnally active and rarely enter their water dishes to the degree that established captive Devils do (Edwards, 2006).

In the wild, but near a facility frequented by humans, young Devils have been seen sunbaking (STDP, 2014a). The function of such behaviour is unknown

In captivity, Devils are often active during the day, probably in response to the schedule imposed on them by their human keepers. Curiously, day-active Devils will bask, with the “head and hind-quarters quite flat on the ground and turned out at right

6/05/2016 8:41 AM 125

angles, somewhat as a frog” (Roberts, 1915; also Fleay, 1935, Kelly, 1993 and Slater, 1993). Basking Devils may have bright red ears, as if they were dissipating heat from them (Roberts, 1915; Fleay, 1952). If this is the case, their basking may be more related to some function other than thermoregulation, such as digestion or ectoparasite control.

Devils have a system of blood vessels at the base of the brain that appears to act as a counter-current heat exchanger to keep the brain cool. The bilateral system consists of convoluted anterior and posterior branches of the internal carotid arteries, which join just below the brain in close apposition to a sinus of venous blood. The system carries arterial blood on its way to the brain and the sinus carries venous blood returning from the nose and mouth. The architecture is such that it appears as if heat in the arterial blood from the warm body would be lost to the venous blood returning from the cool nose and mouth. Such a cooling system might be advantageous in keeping the brain cool during high ambient temperatures or strenuous activity when a Devil’s body temperature rises and it starts panting (Shah et al., 1986, 1989).

In terms of thermoregulation, it is easy to imagine Devils in the wild having to cope with extremely cold conditions, as they are out and about at night in winter at high altitude (e.g., ≥ 760 m). On the other hand, it is hard to imagine they would have to deal with maximum temperatures much above what might be expected on a very warm summer’s night, say 35 º C. In this regard, the Devil’s ability to keep its body temperature low in the face of high ambient temperatures than it would probably never experience in the wild is remarkable.

In captivity, localised heating can be used to attract Devils to a particular part of their enclosure (Slater, 1993). Presumably, this is in cooler conditions.

Body Water

As with most other vertebrates, most of the Devil’s body weight is water, specifically, between 54 and 83 percent (Nicol, 1978). Although the sample size was small (N = 4), the two lactating females measured had the lowest percentages, 54-60 vs 77-83. The loss of water through lactation was dismissed as being relatively small, but no alternative was suggested (Nicol, 1978).

Water Flux

Devils, like other most other vertebrates get their water from drinking and their food. Also like other vertebrates, they loose water through evaporation from the skin, eyes, nose and mouth and in the faeces and urine. And as among mammals, males lose water in the ejaculate and females when lactating. The amount of water that passes through an animal per unit time per body weight is known as the water turnover rate.

For Devils the water turnover rate per day ranged 115.6 to 141.4 (mean = 129; range calculated from data in reference, pers. obs.) ml/g0.82 in five animals at 13 ±1 º C (Green and Eberhard, 1979) and 91.6 to 106.0 (mean = 100.4 ± 7) ml/g0.82 in four animals at 20 ±1 º C (Nicol, 1978, 1982), the higher value at the lower temperature being attributed to the higher metabolic rate at the cooler temperature (Nicol, 1982).

6/05/2016 8:41 AM 126

Under standard conditions, the water turn-over relative to weight is very similar for all mammals examined to date. About the only variation in the relationship is a tendency for animals from more arid areas to have lower rates of turnover; that is, they seem to conserve water slightly better than animals from more mesic areas (Nicol, 1978). The data for water turnover relative to body weight in Devils fits well within the relationship demonstrated for mammals in general (Nicol, 1978).

In the field, water influx as measured by labelled water, was 105 ml/kg/day in a 7.1 kg winter animal and 92 ml/kg/day in a 7.9 kg summer animal (Green, 1997: tables 9.2-3).

Although Devils do not sweat, they do lose water passively through the skin, and the higher the body temperature the more water they lose. This is probably due to the higher skin temperature, which increases the water vapour pressure at the skin surface. At an ambient temperature of 19-21º C the cutaneous water loss in two adult 2 Devils averaged 24.3 ± 5.6 g H2O/m /h, but at the near lethal ambient of 39-41º C, it 2 average 36.8 ± 6.4 g H2O/m /h, which was only an increase of just over 50 percent increase (Hulbert and Rose, 1972).

Breathing Rate

At an ambient temperature of 19-21º C, two resting adult Devils had an average breathing rate of 20 ± 2 breaths/minute, but at an ambient of 39-41º C for an ambiguously stated length of time, the rate went up to 144 ± 27 breaths/minute as the animal panted (Hulbert and Rose, 1972). Similarly, at an ambient temperature of 40 º C for six hours, one adult Devil’s respiration rate rose from about 116 breaths/min to a peak of about 250 in the second hour before returning to, and remaining at, about 127 (Robinson and Morrison, 1957). The increase in breathing rates in both experiments was presumably an effort to lower body temperature through evaporative cooling along the moist respiratory passages.

When the level of ambient O2 is experimentally halved, that is, reduced from just over 20 percent to 10 percent, the Devil’s breathing rate rises to about 55 from 20; the amount air inspired and expired at each breath, the tidal volume, falls only slightly from about 9 ml/kg to about 6 ml/kg, but the total ventilation nonetheless rises from about 100 ml/kg/min to about 330 ml/kg/min (Nicol, 1982: fig. 1). The only time the ambient oxygen is likely to be reduced is, say, at the end of its daily resting cycle and at the end of a deep shelter burrow, or when two mating Devils had spent several days at the end of a deep mating burrow.

When blood CO2 levels rise, Devils respond by increasing both their breathing rate and, most markedly, their tidal volume (Nicol, 1982: fig. 1). Such a situation would arise in the wild during strenuous activity or perhaps at the end of a burrow.

Heart Rate and Output

Recorded heart rates also range widely. Minimum heart rates for two alert animals with body weights of 4.3 and 5.9 kg were 54 and 66 beats per minute, respectively (Kinnear and Brown, 1967). These figures fall well within the body weight dependent

6/05/2016 8:41 AM 127

range for marsupials in general, which are well below the range for eutherians (Kinnear and Brown, 1967).

Observations (N = 20) on an unspecified number of animals with unspecified histories recorded heart rates that “often fell to about 50 beats per minute (Nicol, 1982).

Experimental procedures may induce an elevated heart rate. For example, heart rate in three ether-anesthetised adults had a mean of 180 beats per minute (Parsons et al., 1970). And heart rate in five animals with elevated body temperatures and possibly slightly anaemic following a surgical procedure had a mean rate of 142 (± 30) beats per minute (N = 29; Nicol, 1982).

The Devil’s normal cardiac output is about 100 to 200 ml/kg/min. But when the ambient oxygen falls, cardiac output rises, to about 580 ml/kg/min in an atmosphere of 10 percent oxygen (Nicol, 1982: fig. 1).

Blood Pressure

In three ether-anaesthetised mature animals, aortic blood pressure averaged 94/42 mm Hg, and pulse pressure averaged 52 mm Hg (Parsons et al., 1970). In three sodium pentobarbitone – anaesthetised mature animals, femoral artery pressure ranged 120/85 – 125/95 mm Hg (Weiss and Richards, 1971). And in five intact and resting animals, systolic pressures averaged 131 ± 26 mm Hg and diastolic pressure 98 ± 23 mm Hg (28 observations; Nicol, 1982). The latter measurements, due to the alert and resting state of the animals, are probably the most indicative.

Blood Constituents

Prior to the discovery of Devil Facial Tumour Disease in 1996, researchers interested primarily in the differences between marsupials and placentals and the variation among marsupials, elucidated many of the standard parameters of Devil blood composition and function. This early work was often basic. Sample sizes were small; many of the animals had been in captivity for long periods, and handling in some cases was stressful. The first Devil examined in the pre-DFTD period was a hapless captive, probably in the Basel Zoo (Bartels et al., 1966). Subsequent studies in the pre-DFTD period were largely opportunistic and uncoordinated (Parsons et al., 1970; Parsons et al., 1971a-b; Parsons and Guiler, 1972a-b; Nicol, 1982; Isascks et al. 1984). With the advent of DFTD, however, researchers began to study systematically the Devil’s blood, especially as it functions in the immune response, mainly to determine blood constituents or function was contributory to the spread of DFTD. A handful of researchers determined quickly the Devil was not the problem, as it had robust blood function. As this result unfolded, attention turned to DFTD’s ability to evade the Devil’s immune response. Work has now shifted to this aspect of the problem. In the short time the Devil was the focus, however, researchers catalogued much basic information about the animal’s blood constituents and function. Below is a resume of what has been discovered.

Table 7. Counts for certain cell types in Devil blood.

Provenance N Parameter Units Range Mean SD,se Referenc

6/05/2016 8:41 AM 128

Wild 182 Red Blood Cells 1012/l 5.43-8.99 6.89 0.83,- Peck et al., 2015 Captive 1 Red Blood Cells 1012/l 6.70 - -,- Bartels et al., 19 Captive 4 Red Blood Cells 1012/l 5.84-6.59 6.33 -,- Nicol, 1982 Captive 4 Red Blood Cells 1012/l - 6.6 -,0.2 Simmonds et al. Captive 4 Red Blood Cells 1012/l - 6.46 0.4,- Isaacks et al., 19

Wild 182 Nucleated RBCs 109/l 0.0-23 3.5 6.3,- Peck et al., 2015

Wild 182 White Blood Cells 109/l 8.6-21.2 14.0 3.1,- Peck et al., 2015 “Wild”1 1 White Blood Cells 109/l 15 - - Parsons et al. 19 Wild2 6 White Blood Cells 109/l - 16.6 5.9,- Kreiss et al., 200 Wild3 14 White Blood Cells 109/l - 10.1 2.9,- Kreiss et al, 200 Wild 5 White Blood Cells 109/l 6.3-13.9 10.3 3.1,- Kreiss, 2009 Wild (2), 6 White Blood Cells 109/l - 8.16 2.65,- Woods et al., 20 captive (4) Captive 4 White Blood Cells 109/l 8.6-14.6 12.2 Nicol, 1982 Captive 4 White Blood Cells 109/l - 9.8 1.2,- Isaacks et al., 19 Captive 4 White Blood Cells 109/l - 7.0 -,0.7 Simmonds et al. Captive 9 White Blood Cells 109/l - 7.1 1.9,- Kreiss et al., 200 Captive 6 White Blood Cells 109/l 6.7-10.7 9.1 1.7,- Kreiss, 2009 Unstated ? White Blood Cells 109/l 7.5-14.8 - -,- Munday, 1988

Wild2 6 Basophils 109/l - 0.3 0.2,- Kreiss et al., 200 Wild3 14 Basophils 109/l - 0.1 0.1,- Kreiss et al., 200 Wild 5 Basophils 109/l 0 0 -,- Kreiss, 2009 Wild (2), 6 Basophils 109/l - 0.05 0.06,- Woods et al., 20 captive (4) Captive 9 Basophils 109/l - 0.2 0.2,- Kreiss et al., 200 Captive 6 Basophils 109/l 0 0 -,- Kreiss, 2009

Wild 183 Eosinophils 109/l 0.00-2.12 0.81 0.55,- Peck et al., 2015 Wild2 6 Eosinophils 109/l - 0.05 0.05,- Kreiss et al., 200 Wild3 14 Eosinophils 109/l - 0.02 0.01,- Kreiss et al., 200 Wild 5 Eosinophils 109/l 0-0.7 0.04 0.02,- Kreiss, 2009 Wild (2), 6 Eosinophils 109/l - 0.10 0.09,- Woods et al., 20 captive (4) Captive 9 Eosinophils 109/l - 0.05 0.0,- Kreiss et al., 200 Captive 6 Eosinophils 109/l 0-0.1 0.04 0.04,- Kreiss, 2009

Wild 183 Neutrophils 109/l 2.84-12.18 6.17 2.39,- Peck et al., 2015 Wild2 6 Neutrophils 109/l - 12.6 5.5,- Kreiss et al., 200 Wild3 14 Neutrophils 109/l - 5.5 1.9,- Kreiss et al., 200 Wild 5 Neutrophils 109/l 2.7-5.4 4 1,- Kreiss, 2009 Wild (2), 6 Neutrophils 109/l - 4.04 0.75,- Woods et al., 20 captive (4) Captive 9 Neutrophils 109/l - 3.4 1.8,- Kreiss et al., 200 Captive 6 Neutrophils 109/l 1.8-4.2 2.8 0.9,- Kreiss, 2009

6/05/2016 8:41 AM 129

Wild 183 Lymphocytes 109/l 2.58-13.06 7.07 2.67,- Peck et al., 2015 Wild2 6 Lymphocytes 109/l - 2.6 1.5,- Kreiss et al., 200 Wild3 14 Lymphocytes 109/l - 3.0 1.7,- Kreiss et al., 200 Wild 5 Lymphocytes 109/l 1.4-6.0 3.3 1.7,- Kreiss, 2009 Wild (2), 6 Lymphocytes 109/l - 2.50 2.12,- Woods et al., 20 captive (4) Captive 9 Lymphocytes 109/l - 2.0 1.2,- Kreiss et al., 200 Captive 6 Lymphocytes 109/l 2.2-6.1 4.4 1.3,- Kreiss, 2009

Wild 183 Monocytes 109/l 0.00-0.55 0.11 0.17,- Peck et al., 2015 Wild2 6 Monocytes 109/l - 0.7 0.2,- Kreiss et al., 200 Wild3 14 Monocytes 109/l - 1.1 0.4,- Kreiss et al., 200 Wild 5 Monocytes 109/l 0.5-2.2 1.4 0.4,- Kreiss, 2009 Wild (2), 6 Monocytes 109/l - 0.87 0.53,- Woods et al., 20 captive (4) Captive 9 Monocytes 109/l - 1.1 0.4 Kreiss et al., 200 Captive 6 Monocytes 109/l 1.1-2.0 1.4 0.6,- Kreiss, 2009

Wild 182 Platelets 109/l 115-600 412 110.49,- Peck et al., 2015 Captive Platelets 109/l - 218.0 -46.0 Simmonds et al. 1. From the wild and tested within five days. 2. From DFTD-free areas. 3. From DFTD areas, but healthy.

The counts for three blood cell types in Table 7 are worth comment. For eosinophils, there is an order of magnitude diference between the higher counts of Peck et al., 2015 and others. Perhaps this is due to the much larger sample size in the former and simple sampling error, but it may be worth a further look. For neutrophils, the numbers for captive animals seem lower, probably significantly so, than the numbers for wild animals. Perhaps wild animals are more likely to carry wounds and injuries than more pampered captive animals and hence carry fewer infections (Kreiss et al., 2008). For lymphocytes, the counts of Peck et al., 2015, are higher than those of other researchers. This difference was attributed to the use of isoflurane as an anesthetic, a compound associated with lower white blood cell counts in other species (Peck et al., 2015).

Red Blood Cells

Devil red blood cells are biconcave discoids (Simmonds et al., 2011: fig. 1b). In one long-term captive, mean “thickness” and “length” were 1.70 μm and 7.26 μm, respectively (Bartels et al., 1966). In 24 cells from an unstated number of captives mean diameter was 6.8 (± 0.29 sd) μm (Benga et al., 1992).

In one long-term captive, the mean RBC surface area was calculated as 101.3 μm2 (Bartels et al., 1966).

The mean RBC volume was 70.1 μm3 in one long-term captive (Bartels et al., 1966); averaged 64.8 (± 1.0) μm3 in four captives (Isaacks et al., 1984); averaged 65.0 (±

6/05/2016 8:41 AM 130

1.1) μm3 in four other captives (Simmonds et al., 2011), and ranged 64 to 68 (mean = 65.5) μm3 in four individuals of unstated provenance (Nicol, 1982).

Devil RBCs are at the low end of the size range in marsupials (Benga et al., 1992), but the functional significance of red RBC size is not understood so no explanation to their small size is forthcoming.

Reports of the Devil haematocrit range widely. In three wild individuals, it ranged 28- 38 percent (mean = 33.4) (Parsons et al., 1970). In 181 wild Devils it ranged 0.38- 0.54) (mean = 0.44, SD = 0.04) (Peck et al., 2015). In one long-term captive, it was 47 percent (Bartels et al., 1966); in four captives it averaged 41.6 (± 2.2) percent (Isaacks et al., 1984), and in four long-term captives it averaged 42.5 (± 1.8) percent (Simmonds et al., 2011). In four individuals of unstated provenance, it ranged 39.1- 41.7 percent (mean = 40.8) (Nicol, 1982) and in an unspecified number of individuals of unstated provenance it ranged 39-49 (Munday, 1988: table 1). The overall range of 28 to 54 percent encompasses almost the entire range for Australian marsupials (Munday, 1988: table 1) and seems large for a group of healthy animals.

Reported haemoglobin levels in Devils generally fall within a relatively narrow range. In three freshly-caught, ether-anaesthetised adults it ranged l1-13 g/dl (mean = 12.2) (Parsons et al., 1970). In 182 wild animals, it ranged 12.7-21.2 g/dl (mean = 15.5) (Peck et al., 2015). In one long-term captive Devil, it was 20.1 g/dl (Bartels et al. 1966); in four captives, it averaged 15.3 g/dl (SD = ± 0.7) g/dl (Isaacks et al., 1984), and in four other captives, it averaged 15.5 g/dl (± 0.5) (Simmonds et al., 2011). In four individuals of unstated provenance, it ranged 13.8 to 15.0 g/dl (mean = 14.0) (Nicol, 1982) and in an unspecified number of animals of unstated provenance it ranged 13 to 18 g/dl (Munday, 1988). In addition, haemoglobin values in four to six animals observed pre- and post exercise had average values of 12.0 (± 0.6) g/dl at pre- exercise rest, 13.3 (± 0.4) g/dl just after exercise, 13.5 (± 0.4) after 10 minutes, 12.1 (± 0.3) g/dl after 30 min and 11.9 (± 0.3) min after 70 min (Nicol and Maskrey, 1986). In general, these values are well within the range of other marsupials (Nicol, 1982; Munday, 1988: table 1).

The mean corpuscular haemoglobin (MHC) ranged 21-24 pg (mean = 23, SD = 0.87) (Peck et al., 2015) in 181 wild Devils. It was 30 pg in one long-term captive (Bartels et al., 1966); averaged 23.8 (± 0.6) pg in four captives (Isaacks et al., 1984); averaged 23.8 (± 0.3) pg in four other captives (Simmonds et al., 2011), and ranged 22.3 to 24.1 (mean = 23.2) pg in four animals of unstated origin (Nicol, 1982).

The mean corpuscular haemoglobin concentration (MCHC) ranged 31.6-44.9 g/dl (mean = 34.9) 180 wild Devils. It was 42.7 g/dl in one long-captive animal (Bartels et al., 1966); averaged 36.9 (± 0.6) g/dl in four captives (Isaacks et al., 1984); averaged 36.4 (± 0.47) g/dl in four other captives (Simmonds et al., 2011), and ranged 34.6 - 35.5 g/dl (mean = 35.2) in four individuals of unstated provenance (Nicol, 1982).

The mean corpuscular volume (MCV) ranged 51.8-73.9 (mean = 64.8, SD = 5.25) in 181 wild Devils (Peck et al., 2015).

Red blood cell ATP and its metabolites

6/05/2016 8:41 AM 131

Red blood cells use ATP as an energy source and as it is used and re-formed several intermediate metabolites are created. Some of these have been measured in the Devil (Isaacks et al., 1984), and all are within usual mammalian values.

Red blood cell electrolytes

Levels of calcium (0 m-equiv., N = 3), magnesium (4.7 m-equiv., N = 1) and phosphate (7.8-9.1 mg %, N =3) in RBCs are within the general mammalian range (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972b).

Red blood cell membrane permeability

The permeability of the red blood cell membrane to water in Devils seems typical of most Australian marsupials, the variation that does exist being poorly understood (Benga et al., 1993).

Red blood cell sedimentation rate

The red blood cell sedimentation rate in a single lactating female was 18 mm/h (Parsons et al., 1970). In humans, this value varies with age but in adults, it ranges about 12 to 23 mm/h.

White Blood Cells

As far as surveyed, Tasmanian Devils have the full complement of white blood cells, in normal proportions and in the expected lymphoid organs based on other marsupials and eutherians (Kreiss et al., 2009; Howson et al., 2014).

The white blood cell count (WBC) found by various researchers are summarised in Table 7. These overall counts are at the high end of the range for marsupials (Nicol, 1982; Munday, 1988: table 1), which may not be surprising in a scavenger often with open wounds in the mouth and body and not adverse to eating rotting flesh.

There are two major groups of white cells or leucocytes: granulocytes (or polymorphonuclear leucocytes) and agranulocytes (or mononuclear leucocytes). In Devils, the number of granulocytes as a proportion of the total WBC count ranged from 73 to 76 (mean = 74) percent in three “wild” (above) Devils (Parsons et al., 1970) and from 44 to 52 (mean = 48) percent in four of unstated provenanceDevils (Nicol, 1982). The reason for this large difference between the two reports is unclear.

There are three types of granulocytes: basophils, eosinophils and neutrophils. Counts for each in Devils are shown in Table 7.

Neutrophils are phygocytes and form one of the first responses to invasion by microbes. In vitro, Devil neutrophils phygocytosed bacteria and appeared to digest them; that is, they performed competently (Woods et al., 2007; Kreiss et al., 2008).

There are three types of agranulocytes: lymphocytes, monocytes and monocytes’ in- tissue derivatives, macrophages. Counts for the first two types in Devils are given in

6/05/2016 8:41 AM 132

Table 7. There are no counts for macrophages, possibly due to the difficulty of counting them in tissues.

In Devils, the number of lymphocytes as a proportion of the total WBC count ranged from 20 to 27 (mean = 24) percent in three “wild” Devils (Parsons et al., 1970) and from 45 to 51 (mean = 48) percent in four others of unstated provenance (Nicol, 1982). The proportion of monocytes ranged 0 to 4 (mean = 2) percent and 2 to 6 (mean = 3.5) percent, respectively. The reason for the large difference in the proportion of lymphocytes as between the two reports is unclear, as it was for the proportion of granulocytes (above).

A study that distinguished lymphocytes found that among a cell population consisting of granulocyte neutrophils (as contaminants) and all agranulocytes, the percentage compositions (± SD) were: neutrophils 7 (± 5), T lymphocytes 55 (± 8), B lymphocytes 33 (± 8), natural killer-like cells 4 (± 1) and monocytes 5 (± 3) (Brown et al., 2011; see also Brown, 2013; table 4.1).

The counts of these white cells, which are an important component of the immune system are all well within the normal range for mammals.

Two common types of lymphocytes are T cells and B cells. Both can proliferate as part of the immune response. To test the ability of the lymphocytes in Devils to proliferate, researches challenged the lymphocytes in vitro with several standard mitogens, compounds known to induce cell proliferation. The lymphocytes responded competently, although the strength of the response was variable both within and among individuals (0-100 + fold increase) (Woods et al., 2007; Kreiss et al., 2008; Stewart et al., 2008). The significance of this variable response is unclear. There was no significant difference in the responses of males vs females or adults vs juveniles (Kreiss et al., 2008).

The less common natural killer cells also respond competently to foreign cells, both alive and irradiated (Brown et al., 2010, 2011; Brown, 2013).

To the extent they have been studied, Devils have all the types and subtypes of T and B lymphocytes expected in a mammal (Howson et al., 2014).

Platelets

Platelets, or thrombocytes, are small cells involved in the clotting process in mammals. The number of platelets in 182 wild Devils ranged 115-600 x 109/l (mean = 412, SD = 110.5) (Peck et al., 2015). In four captive Devils, they averaged 218.0 x 109/l (± 46.0) (Simmonds et al., 2011).

Fibrinogen

Fibrinogen is a plasma soluble glycoprotein involved in the clotting process. In 180 wild Devils, it ranged 1-6 g/l (mean = 2, SD = 1.28) (Peck et al., 2015). In four captive Devils, it had an average of 2.0 g/l (± 0.0) (Simmonds et al., 2011).

Blood Serum Electrolytes

6/05/2016 8:41 AM 133

Table 8. Blood serum electrolytes in Devils.

Provenance N Compound Units Range Mean SD,se Reference “Wild”1 3 Calcium meq 3.4-5.1 4.5 -,- Parsons et al., 1970 ? 4 Calcium meq 3.4-6.7 - -,- Parsons et al., 1971a Wild 185 Calcium mmol/l 2.01-2.83 2.46 0.22,- Peck et al., 2015 “Wild”1 3 Chloride meq 100-102 101 -,- Parsons et al., 1970 ? 3 Chloride meq 100-102 - -,- Parsons et al., 1971a Wild 185 Chloride mmol/l 105-115 110 3,- Peck et al., 2015 “Wild”1 3 Magnesium meq 2.0-2.25 2.1 -,- Parsons et al., 1970 ? 4 Magnesium meq 1.8-2.4 - -,- Parsons et al., 1971a Wild 184 Magnesium mmol/l 0.73-1.18 0.88 0.11,- Peck et al., 2015 Wild 184 Phosphate mmol/l 1.3-2.68 1.91 0.37,- Peck et al., 2015 “Wild”1 3 Potassium meq 4.6-5.0 4.8 -,- Parsons et al., 1970 ? 3 Potassium meq 4.6-5.0 - -,- Parsons et al., 1971a Wild 183 Potassium mmol/l 4.4-9.4 6.1 1.3,- Peck et al., 2015 “Wild”1 3 Sodium meq 135-136 138 -,- Parsons et al., 1970 ? 3 Sodium meq 135-142 - -,- Parsons et al., 1971a Wild 185 Sodium mmol/l 137-147 142 3,- Peck et al., 2015 1. From the wild and tested within five days.

The resting levels of calcium (3.4-6.7 m-equiv., N = 4), chloride (100-102 m-equiv., N = 3), magnesium (1.8-2.4, N = 4), potassium (4.6-5.0 m.-equiv., N = 3) and sodium (135-142 m-equiv., N = 3) are within the normal mammal range (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972b).

Calcium levels were apparently significantly higher in male than in female Devils (Peck et al., 2015: table 7; unreported non-parametric test).

Calcium levels in adult males showed significant group differences in spring (2.30), summer (2.22), autumn (2.51) and winter (2.47) groups (Peck et al., 2015).

Calcium levels in adult females showed significant group differences in early lactation (2.4), mid-lactation (2.2) and end of lactation (2.2) groups (Peck et al., 2015). The higher levels in early lactation were suggested to be “… due to increased bone metabolism to provide calcium to nursing young via the milk (Peck et al., 2015).” An earlier report based on a one lactating female and two adult males had suggested calcium levels might be lower in lactating females (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971). It was not clear, however, what lactation stage the female was in.

Calcium levels are significantly higher in juveniles (2.58) than in adults (2.32) (t = 9.42, P < 0.001; pers. obs. on data in Peck et al., 2015: table 6), a common result in growing young animals (Peck et al., 2015).

Chloride levels in adult males in spring, summer, autumn and winter groups apparently showed no significant differences among the groups (Peck et al., 2015).

6/05/2016 8:41 AM 134

Chloride levels in adult females also showed significant differences in early lactation (111), mid-lactation (112) and end of lactation (112) groups (Peck et al., 2015).

Magnesium levels in adult males showed significant group differences in spring (0.87), summer (0.94), autumn (0.84) and winter (0.82) groups (Peck et al., 2015:table 4).

Phosphate levels in adult males in spring (1.73), summer (1.69), autumn (1.63) and winter (1.48) groups showed a significant difference among the groups (Peck et al., 2015; table 4).

Phosphate levels are significantly higher in juveniles (2.09) than adults (1.64) (t = 10.08, P < 0.001; pers. obs. on data in Peck et al., 2015: table 6), a common result in young growing animals (Peck et al., 2015; table 6).

Potasium levels did not differ significantly between male (6.0) and female (6.0) Devils (t = 0.000; pers. obs. on data in Peck et al., 2015: table 7).

Blood Serum Enzymes

Many enzymes occur in the blood serum, and a few of those found in Devils have had their levels measured in healthy animals (Table 9).

Table 9. Serum enzymes in healthy Devils. Note this is only a few of the many enzymes in serum. Provenance N Enzyme Units Range Mean SD,se “Wild”1 2 Acid phosphatase KA 193-287 240 -,- Par ? 9 Acid phosphatase KA 310-462 - -,- Par Wild 10 Acid phosphatase KA 211-462 381 72,- Par Wild 183 Alanine transaminase IU 28-113 54 22, - Pec Wild 184 Alkaline phosphatase IU 31-287 140 70, - Pec “Wild”1 3 Alkaline phosphatase IU2 14-43 33 -,- Par ? 4 Alkaline phosphatase IU2 14-128 - -,- Par “Wild”1 3 Amylase IU3 98-135 110 -,- Par Wild 184 Aspartate aminotransferase IU 63-344 122 61, - Pec Wild 184 Creatine kinase IU 357-4878 1342 948, - Pec Wild 184 Glutamate dehydrogenase IU 51-726 134 174, - Pec “Wild”1 3 Glutamic oxaloacetic transaminase IU 59-70 66 -,- Par “Wild”1 3 Glutamic pyruvic transaminase IU 13-15 20 -,- Par “Wild”1 3 Lactate dehydrogenase IU4 255-313 286 -,- Par 1. From the wild and tested within five days 2. Original in King Armstrong Units. Conversion factor: 1 IU = 7.1 x l KA unit. 3. Original in Somogyi Units. Conversion factor: 1 IU = 1.85 x 1 Somogyi unit. 4. Original in Wroblewski Units. Conversion factor: 1 IU = 0.482 x 1 Wroblewski unit.

Acid phosphatase levels are very high in both the serum and tissues of adult Devils (Parsons et al., 1970; Parsons, Heddle and Guiler, 1971; Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972a-b; Sallis et al., 1973; Sallis and Guiler, 1977; Sallis

6/05/2016 8:41 AM 135

et al., 1984). Levels of blood total acid phosphatase, for example, can be two orders of magnitude higher (310-462 King Armstrong Units, N = 8) than in most other mammals.

The comparatively high levels in Devils raises the issue of the levels in Devils’ near relatives. Interestingly, Dasyurus viverrinus has (low) levels comparable to other mammals (Parsons, Guiler and Heddle, 1971), but D. maculatus has levels even higher than Devils (Parsons and Guiler, 1972a-b: Sallis and Guiler, 1977; Parsons et al., 1982). It is unclear whether the high levels in the blood are due to high levels of release from an organ, which remains unidentified, or from low rates of clearance from the blood (J. Sallis, in Nicol, 1982).

Plasma acid phosphatase levels seem to increase throughout the first two years of life, that is, the period when Devils are reaching maturity. This suggests that the enzyme levels may be responding to a change in diet (Sallis and Guiler, 1977), perhaps from milk to omnivory to meat.

Captivity appears to have little effect on acid phosphatase levels, as Devils kept in captivity for up to three months were said to be not significantly different from those shown by wild Devils (Sallis et al., 1973).

The origin of the Devil’s high acid phosphatase levels are unclear (Sallis et al., 1984), although an unpublished suggestion was that it was not due to “tissue spillage but from “certain chemical or structural properties that afford the enzyme a long circulating life” (J. Sallis pers. comm., in Nicol, 1982).

Presumably the Devil’s (and Spotted-tailed Quoll’s) comparatively high levels of acid phosphatase has some functional significance, but this remains to be discovered (Sallis et al., 1984).

Alanine transaminase in adult males showed significant group difference between spring (44), summer (47), autumn (50) and winter (55) (means in IU unites given as indicative; Peck et al., 2015: table 4).

Alanine transaminase in adult females showed significant group differences between early lactation (50), mid-lactation (58) and the end of lactation females (44) (means in IU units given as indicative; Peck et al., 2015: table 3).

Alkaline phosphatase in adult females showed showed significant group differences between early lactation (87), mid-lactation (62) and end of lactation (87) (Peck et al., 2015: table 3).

The mean level of alkaline phosphatase was significantly (t = 11.23, P < 0.0001) higher in immature (177) than in mature (89) Devils (pers. obs. on data in Peck et al., 2015: table 6 although no provision made for repeated samples as was done in original work). The higher level in immatures was ascribed to “… osteoblastic activity during growth ….” (Peck et al., 2015), that is, the dynamic re-shaping of bone as an animal grows.

6/05/2016 8:41 AM 136

Aspartate aminotransferase levels were apparently significantly higher in male Devils (130) than in female (115) Devils (Peck et al., 2015: table 7, unreported non- parametric test).

Aspartate aminotransferase in adult females showed significant group differences in early lactation (102), mid-lactation (120) and end of lactation (102) (Peck et al., 2015: table 3).

Creatine kinase levels were apparently higher in juvenile () than in adult () Devils (Peck et al., 2015: table 6; unreported parametric test with a repeated measures provision).

Glutamate dehydrogenase in adult females, however, showed significant group differences between early lactation (70), mid-lactation (86) and end of lactation (89) (Peck et al., 2015: table 3). The significance of this difference remains to be determined.

Blood Serum Lipids

Table 10. Blood lipid levels in Devils. Provenance N Compound Units Range Mean SD,se Reference “Wild”1 3 Cholesterol mg % 67-123 94 -,- Parsons et al., 1970 ? 5 Cholesterol mg % 54-135 - -,- Parsons et al., 1971a Wild 91 Cholesterol mmol/l 2.2-6.1 4 0.95,- Peck et al., 2015 ? 2 Phospholipids mg % 110-174 142 -,- Parsons et al., 1971a “Wild”1 3 Triglycerides mg % 50-71 58 -,- Parsons et al., 1970 ? 2 Total Lipids mg % 399-442 420 -,- Parsons et al., 1971a 1. From the wild and tested within five days.

The blood levels of cholesterol (54-135 mg%, N = 5), phospholipids 110-174 mg%, n = 2), triglycerides (50-71 mg %, N = 4) and total lipids (399-442 mg%, N = 2) (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972b) are all within the general range for mammals.

Blood Serum Organics

Table 11. Blood serum organic compounds measured in Devil serum.

Provenance N Compound Units Range Mean SD,se Reference “Wild”1 1 Bilirubin mg % 0.5 0.5 -,- Parsons et al., 1970 Wild 162 Bilirubin μmol/l 0.7-4.6 2.7 1.1,- Peck et al., 2015 Wild 185 Creatinine μmol/l 27-56 41 8,- Peck et al., 2015 “Wild”1 3 Glucose mg % 102-252 155 -,- Parsons et al., 1970 Wild 157 Glucose mmol/l 4.2-9.0 5.6 1.2,- Peck et al., 2015 “Wild”1 3 Urea mg % 60-76 66 -,- Parsons et al., 1970 ? 9 Urea mg % 53-104 - -,- Parsons et al., 1971a Wild 186 Urea mmol/l 8.3-17.8 12.3 2.6,- Peck et al., 2015 “Wild”1 3 Uric acid mg % 1.0-3.8 2.2 -,- Parsons et al., 1970 1. From the wild and tested within five days.

6/05/2016 8:41 AM 137

The levels of blood glucose (102-252 mg %, N = 3), urea (53-104 mg %, N = 9) and uric acid (1.0-3.8 mg %, N = 3) and the levels of serum bilirubin (0.5 mg %, N = 1) are all within the normal range of mammals (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972b).

Bilirubin levels in adult males showed significant differences among spring (12.6), summer (11.8), autumn (11.8) and winter (9.9) groups (Peck et al., 2015: table 4). Bilirubin also showed significant differences among early (2.1), mid- (2.8) and end of lactation (3.2) groups (Peck et al., 2015: table 3).

Creatine levels were significantly higher in male (44) than in female (39) Devils (t = 4.52, P < 0.001; pers. obs. on data in Peck et al., 2015: table 7).

Glucose levels in adult females showed significant differences among early lactation (5.9), mid-lactation (5.4) and end of lactation (5.5) groups (Peck et al., 2015: table 3).

Glucose levels were significantly higher in juveniles (6.2) than adults (5.6) (t = 3.26, P <0.001; pers. obs. on data in Peck et al., 2015: table 6). This was attributed to the possible greater stress that young Devils experience when first trapped compared to older Devils that may have been trapped more than once and knew the drill.

There may be a “handling effect” with glucose (and other blood constituents). An unpublished honours thesis found a decrease in glucose levels of between 1.4 and 2.9 mmol/L after handling Devils (Dewar, 2013, as reported by Peck et al., 2015).

Urea levels in adult males showd significant differences among spring (2.2), summer (3.6), autumn (3.1) and winter (2.8) groups (Peck et al., 2015: table 4). This compound also showed significant differences among early (10.4), mid- (13.0) and end of lactation (11.7) groups (Peck et al., 2015: table 3).

Blood Serum Proteins

Table 12. Serum proteins in Devils.

Provenance N Compound Unit Range Mean SD,se Reference s Wild 184 Albumin g/l 29.7-39.1 34.6 2.4,- Peck et al., 2015 ? 1 Albumin g% 3.5 3.5 -,- Parsons et al., 1971a Wild 184 Globulin g/l 22-43 33 5,- Peck et al., 2015 ? 1 Globulin, α1 g% 0 - - Parsons et al., 1971a ? 1 Globulin, α2 g% 0.77 - Parsons et al., 1971a ? 1 Globulin, β1 g% 0.81 - Parsons et al., 1971a ? 1 Globulin, γ g% 1.45 - Parsons et al., 1971a 184 Albumin/Globulin 0.76-1.63 1.07 0.21 Wild 185 Total Protein g/l 56.2-79 68.0 5.4,- Peck et al., 2015 ? 1 Total Protein g% 6.5 - -,- Parsons et al., 1971a

The levels of total protein (6.5 g%); pre-albumin (0) and albumin (3.5 g%), and the globulins alpha 1 (0 g%), alpha 2 (0.77 g%), beta (0.81 g%), gamma (1.45 g%, N = 1

6/05/2016 8:41 AM 138 in all cases) are within the range of normal mammalian values (Parsons, Guiler and Heddle, 1971; Parsons and Guiler, 1972b).

In 180 wild Devils, the total plasma protein (serum + clotting factors, mainly fibrinogen) ranged 62-85 g/l (mean = 71, SD = 5.6) (Peck et al., 2015).

Albumin levels in mature males differed significantly among spring (34.1), summer (36.0), autumn (34.4) and winter (34.7) groups (Peck et al., 2015). Albumin in lactating females also differed significantly among early lactation (35.6), mid- lactation (31.7) and end of lactation (34.0) groups (Peck et al., 2015: table 3).

Albumin levels were significantly higher in juvenile (35.4) than in adult (34.0) Devils (t = 4.25, P < 0.001; pers. obs. on data in Peck et al., 2015: table 6).

Globulin levels were significantly higher in adult (37) than in juvenile (31) Devils (t = 9.01, P < 0.0001; pers. obs. on data in Peck et al., 2015: table 6). This may reflect the increased number of immunological challenges Devils experience, not only as they age, but when their social interactions involve more biting and their carnivory increases the incidence of open wounds in the mouth. An interesting comparison usuing the apparent available data set (Peck et al., 2015) would be to test the correlation between globulin levels and age and incidence of wounds current and past.

The albumin:globulin ratio in mature females differed significantly among early lactation (1.03), mid-lactation (0.92) and end of lactation groups (0.91) (Peck et al., 2015: table 3).

This ratio was significantly lower in adult (0.93) than in juvenile (1.16) Devils (t = 8.94, P < 0.001; pers. obs. on data in Peck et al., 2015: table 6).

Total serum protein in mature females also differed significantly among early lactation (70.6), mid-lactation (66.7) and end of lactation (71.1) groups (Peck et al., 2015: table 3).

Blood Plasma Nucleotides, Nucleosides and Bases

A variety of plasma products of nucleic acid metabolism have been surveyed in the Devil and four other Australian marsupial species, but while the results are highly reproducible and differ among species, there is as yet no functional explanation for the differences (McKeag et al., 1981; Sallis et al., 1984).

Other Blood Constituents

The level of protein-bound iodine was 1.4-1.8 μg %) in two adult males and 16 μg %) in one lactating female (Parsons et al., 1970; Parsons, Guiler and Heddle, 1971). The male values are within the normal mammalian range, but the female is high and probably relates to the fact that she was lactating.

Blood Gases, pH and Bohr Effect

6/05/2016 8:41 AM 139

The oxygen tension at which half the haemoglobin is saturated with oxygen, the P50, has been determined twice under the same conditions (37 C and pH 7.4) with similar results: 41.2 mm Hg and 44 mm Hg (Bartels et al., 1966 and Nicol, 1982, respectively). These values are well within the range of dasyurid marsupials (Hallam et al., 1995), but dasyurid values are high relative to other marsupials (Hoversland et al, 1973; Bland and Holland, 1977: fig. 1; Henty et al., 2007) and mammals in general (#). The value for a healthy human would be about 26.6 mm Hg under similar conditions. Dasyurids’ high P50 values are indicative of their blood’s low affinity for oxygen (as half of the bound oxygen is “given up” at oxygen pressures that are higher than in most other mammals). This has been interpreted as an adaptation to a high level of activity associated with predaceous (actively hunting) habits (Hallam et al., 1995).

The partial pressure of oxygen, PO2, in awake, resting animals had a mean of 98.5 mm Hg (sd = 15; 44 observations on six animals) for arterial blood and 57.4 mm Hg (sd = 10; six observations on one animal) for venous blood (Nicol, 1982). In a single sedated but struggling animal, the PO2 had a mean of 60.7 mm Hg (sd = 73, six observations; data in Bartels et al., 1966). In a single animal under anaesthetic the value was a low 62 mm Hg for blood from a cardiac (arterial?) puncture (Parsons et al., 1971), suggesting an effect of the anaesthetic (Nicol, 1982).

The partial pressure of carbon dioxide, PCO2, in the same (above) awake, resting animals had a mean of 30.8 mm Hg (sd = 3.95) for arterial blood and 33.1 mm Hg (sd = 3.8) for venous blood (Nicol, 1982). In the single sedated but struggling animal as above the PCO2 had a mean of 42.5 mmHg (sd = 21.5, six observations; data in Bartels et al., 1966). The animal under anaesthetic had a high of 39 mm Hg (Parsons, Guiler and Heddle, 1971).

- The HCO3 content for the awake, resting animals had a mean of 19.3 mmol/l (sd = 2.2) for arterial blood and 21.1 (sd = 2.9) for venous blood (Nicol, 1982), whereas the animal under anaesthetic had a slightly lower value of 18 mmol/l (Parson, Guiler and Heddle, 1971).

The pH for the awake, the resting animals had a mean of 7.44 (sd = 0.04) for arterial blood and 7.44 (sd = 0.03) for venous blood (Nicol, 1982). The pH for the sedated but struggling animal had a mean of 7.43 for venous blood (sd = 0.23, data in Bartels et al., 1966). The animal under anaesthetic had a slightly lower value of 7.29 (Parsons, Guiler and Heddle, 1971), suggesting that the procedure had caused a degree of acidosis (Nicol, 1982).

When the oxygen content in the ambient air is experimentally reduced from about 20 percent to about 10 percent, the PO2 in the venous blood falls slightly and while the PO2 of the arterial blood falls markedly to just above that of the venous partial pressure. This is natural enough as the amount of available oxygen is simply less. The PCO2 and pH in the arterial and venous blood drops and rises, respectively, however, due to the increased breathing rate and greater clearance of CO2 from the blood (Nicol, 1982: fig. 1).

There are two determinations of the effect of the blood pH on the P50, the Bohr effect, and both are very similar: ∆ log P02/∆pH = -0.47 and -0.44 (Bartels et al., 1966

6/05/2016 8:41 AM 140

and Isaacks et al., 1984, respectively). These two values are higher than the only other marsupial for which there is data, the Brush-tail Possum, with a value of -0.60 (Henty et al., 2007). The lesser Bohr effect in the Devil may reflect its higher P50, which achieves a similar effect, unloading oxygen as needed, by a different means.

Blood Clotting

Observers experienced at taking blood samples from Devils and other marsupials have noted that Devil blood clots very rapidly (Parsons et al., 1970; Sallis et al., 1973). Perhaps rapid clotting facilitates wound closure in a species noted for intraspecific fighting, which can lead to severe wounding.

Digestion

Fed on rats, Devils digest between 77.9 and 79.9 (mean = 79.1) percent of their food’s dried weight and between 85.8 and 88.0 (mean = 87.2) percent of its energy content (Green and Eberhard, 1979). These digestion and energy efficiencies are in line with other carnivores.

Fat

Devils store fat most obviously in the tail (Mooney, 1992; Denholm, 2010b; Devil Island, 2013e), which gives it a carrot shape (Jones, 1994). In captivity, the mean width of the tail as measured 2-3 cm from the base in four captive females varied by as much as 30 percent over a six month period (October to March) (Jones et al., 2005).

When over-fed in captivity, fat accumulates on the neck, back and lower abdomen (Kelly, 1993). Very fat devils may also have a roll of fat just above the shoulder and another roll above the base of the tail (Sinn et al., 2010). All these areas would probably also be fat storage areas in wild Devils, should conditions ever be so good.

Dissection of Devils in good condition reveals large amounts fat stored just under the skin (Green, 1967; Parsons et al., 1970; Parsons, Heddle and Guiler, 1971; Guiler, 1992), where it can be 12 mm thick, and around the viscera (Guiler, 1964; Green, 1967), where it can completely cover the intestine and kidneys (Green, 1967: plate 5).

There is also a band of back fat deep to the subcutaneous back fat that runs from the neck to the rump with short extensions into the base of the limbs (Parsons et al., 1970).

There is also a description of an “apron of fat” (Parsons, Heddle and Guiler, 1971), but it is unclear just where in the, presumably, abdomen this fat lies.

The described colour of the fat appears to vary among accounts and perhaps within the body. The subcutaneous fat and the “body fat in general” has been described as yellow, while the band of back fat is described as “whiter” (and firmer) than the body fat (Parsons et al., 1970; see also Parsons, Heddle and Guiler, 1971). Somewhat confusingly, however, the neck fat has also been described as “white” (Parsons, Heddle and Guiler, 1971).

6/05/2016 8:41 AM 141

There is no evidence that Devils have brown fat (Hayward and Lisson, 1992; Kabat et al., 2003), the type of fat that supplies the energy for non-shivering thermogenesis in eutherians.

There is limited evidence that Devils’ body condition varies seasonally. Males are said to show a decline in condition during the mating season (Pemberton, 1990: fig. 7.6; M. Jones unpublished data in Jones and Barmuta, 1998). In terms of actual measurements of mass, the lowest period is just after the mating season, in May and June (Jones, 1995). On the same measure, mass relative to pes length, females show no seasonal variation (Jones, 1995).

In pastoral areas, Devils are heaviest during spring, (September-October), due to the availability of lambs, which are born (and perish) at this time of year (Guiler, 1970a; Pemberton, 1990: fig. 7.6 for males).

Interestingly, Tasmanian wildlife authorities apparently have survey data that would further address the question of seasonal fluctuation in body weight generally (Sinn et al., 2010; Newsletter, May 20133), but there is as yet no published analysis.

There is also unpublished evidence that body condition can vary in an exceptional season. Devils on Freycinet Peninsula, for example, were said to be in “poor body condition as they were coming into the mating season” in the drought year 2001 (M. Jones, unpublished data 2001 as cited in Hamede et al. 2008).

Adrenal Function

The total mass of the two adrenal glands in three adult Devils ranged 0.79-1.20 g. Adrenal mass relative to body mass ranged 163-182 mg/kg. The two adrenal glands differed in size, the left being about 13 to 23 percent heavier than the right (data in Weiss and Richards, 1971: table 1).

It is unclear what the true baseline levels of corticosteroids are in Devils, because all determination methods involve some level of stress, involving at the very least, capture, restraint and blood letting (Table 13).

Also, to judge from one survey based on anaesthetised animals, cortisol levels, at least, are significantly higher during the day than during the night (Table 13), as might be expected in an animal that is primarily nocturnal in its activity time Keeley et al. 2012a).

Table 13. Plasma/serum cortisol levels in Devils.

Cortisol mg mL-1 Sex Health Status Reference Range Mean n 28-37 32.5 2 ♂ Healthy, wild Weiss and Richards, 1971 - 23.7 ± 13.1 9 ♂ Healthy, captive Keeley et al., 2012a - c. 14.51 4 ♂ Healthy, wild Jones et al., 2005 75 75 1 ♀ Healthy, wild Weiss and Richards, 1971 - c. 391 4 ♀ Healthy, wild Jones et al., 2005

6/05/2016 8:41 AM 142

- 15.9 ± 23.7 32 ♂ Diseased, wild Keeley et al., 2012a - 24.8 ± 4.7 ? ♂ Diseased, wild, day Keeley et al., 2012a - 7.7 ± 5.3 ? ♂ Diseased, wild, night Keeley et al., 2012a 1. Mean of five monthly means estimated from Jones et al., 2005: fig. 4.

Only two different corticosteroids have been found in the Devil’s peripheral plasma, cortisol and corticosterone. In contrast, 14 different steroids have been found in the Devil’s adrenal venous blood, ten of which have been identified (Weiss and Richards, 1971). Cortisol appears to be the most common steroid (Weiss and Richards, 1971; A. Bradley, in Pemberton, 1990).

Interestingly, of seven Australian marsupials (four herbivores, one herbivore/insectivore and two carnivores) surveyed for ACTH-induced cortisol secretion, the Devil had the highest level of all with no overlap at all with the other species (Weiss and Richards, 1971).

Within Devils, males appear to have lower corticosteroid levels than females (Weiss and Richards, 1971; Pemberton, 1990: fig. 7.1; Jones et al., 2005: fig. 4).

Corticosteroids appear to be relatively stable throughout the year in males but are cyclic in females. In one field study, adult males showed no significant seasonal differences. Adult females, however, had rising values from mid-summer (January) to mid-autumn (April) when births occur, followed by the declining values to early winter (June) when the young are still in the pouch and then rising to mid-spring (October) when the young are just about to become independent, followed by a decline (Pemberton, 1990: figs 7.1 and 7.3).

A subsequent study of seasonal variation in corticoidsteroids found plasma cortisol increased in male Devils during autumn and winter (Kolkert, 2005 as reported in Peck et al., 2015).

When caught and brought into captivity, wild Devils show increased levels of the stress-associated corticosteroid, cortisol. Levels returned to what are perhaps normal resting values, however, after about 48 hours in males but not for about a month in females. Indeed, the levels remained significantly higher in females than in males for at least up to six months, the length of the experiment. It is uncertain, however, if the females in this experiment had prolonged higher levels due to their special circumstances (one female caged with one male) or to their apparently higher levels in general (above; Jones et al., 2005).

Lymphatic System

Anatomically, the Devil’s lymphatic system is similar to that of other mammals. There are lymph nodes and a spleen, and a thymus is clearly evident in juveniles and subadults as well as adults (Mackenzie, 1916, 1921; Woods et al., 2007; Kreiss et al., 2009a). Devils also have palatine tonsils (Hett and Butterfield, 1909). Histologically, therefore, there is nothing to indicate that the Devil’s lymphatic system is other than fully competent (Woods et al., 2007; Kreiss et al., 2009a).

6/05/2016 8:41 AM 143

As an historical aside, before the immunological importance of the thymus was discovered, it was briefly considered to bear some inverse relationship with muscle development and regeneration (Mackenzie, 1921).

Immune Response

General observations would suggest that Devils have robust immune systems. Adults with open wounds on the head can be found tucking into the intestines of rotting carcasses looking none the worse for wear. Here are descriptions of some of their injuries. “One [male] had pieces of the lips, jowls and face torn away (Guiler, 1992).” “One male … had no skin over the whole rump exposing the raw flesh. The wound was clean and the animal in good condition … (Guiler, 1992).” “… tooth damage was almost universal particularly in the older animals … (Guiler, 1992).” “Devils are constantly getting broken teeth and rotten teeth and bones stuck in their mouth (N. Mooney, in Cordingley, 2007b).” “Devils always have injuries around their mouth ….They get [echidna] spines stuck in the roof of their mouth. They get possum claws wedged between their teeth. And they get beautiful gum infections. They get shards of bone jammed up between their molar teeth. And they get festering infections (Jones, 2012).” As well, their age-related scarring shows that open wounds are a normal part of a Devil’s life.

Furthermore, when it comes to wounds, a veterinarian has said, Devils seem more resilient than cats or dogs (S. Peck, in Devil Island, 2013a).

It would be remarkable, therefore, if Devils did not have a robust immune system (Woods et al., 2007). Nonetheless, researchers on the Devil’s Facial Tumour Disease thought it expedient, as their first line of investigation, to examine healthy Devils’ immune capacities. In all the ensuing examinations and tests, their immune system proved to be robust.

For example, research demonstrated that the Devil’s neutrophils, the white blood cells responsible for attacking microbial infections, are both numerically (above) and functionally normal (Woods et al., 2007; Kreiss et al., 2008).

The Devil is also fully able to produce antibodies to foreign proteins. For example, when injected with horse red blood cells, two adult Devils produced “high antibody titres,” which were boosted again following a second injection (Kreiss et al., 2009b). Indeed, the response was two orders of magnitude greater than the response mounted by the New World Opossum in a similar experiment (Woods et al., 2007).

Devils also produce cytotoxic and antibody responses to injected tumour cells from other species even when those cells lack the MHC-I proteins on their surface that identify them as “other,” one of the methods, at least, that Devil Facial Tumour cells use to evade the immune response (Brown et al., 2011; Brown, 2013).

Devils also reject reciprocal skin grafts as robustly as other mammals (Kreiss et al., 2011b).

6/05/2016 8:41 AM 144

Finally, healthy Devil cells, quite normally, express MHC-I genes and transport the proteins to the cell surface for presentation as “self” to the immune surveillance system (Tovar, 2012).

Sedation Drugs

A variety of drugs has been used to relax or anaesthetise Devils in zoo, laboratory and fieldwork.

For sedation, zoo veterinarians have used tiletamine-zolazepam (Smeller et al., 1977; Booth, 1994; Holz, 1992, 2002), but the degree of relaxation is variable and recovery times can be in excess of six hours (Holz, 1992, 2002). Field workers have used xylazine hydrochloride (Pemberton and Gales, 1991). Lab researchers have used a combination of detomidine and butorphanol (Loh et al., 2006a-b; Pearse et al., 2012).

For anaesthesia, lab researchers have used, variously, ether followed by sodium pentobarbitone (Guiler and Heddle, 1970; Weiss and Richards, 1971); ketamine hydrochloride (Kabat et al., 2003, 2004); sodium pentobarbitone (Parsons and Guiler, 1972b; Keeley et al., 2012a-b), and phencyclidine (Bartels et al., 1966). Field workers have used ketamine hydrochloride (Tribe and Middleton, 1988; Pemberton and Gales, 1991).

Recently, zoo (Booth, 1994; Holz, 2002; Keeley et al., 2012d; Scheelings et al., 2014) and lab workers (Kabat et al., 2003, 2004; DPIWE, 2005a; Loh et al., 2006a-b; Siddle et al., 2007b; Woods et al., 2007; Kreiss et al., 2008; Stewart et al., 2008; Kreiss, 2009; Brown et al., 2011; Kreiss et al., 2011b; Simmonds et al., 2011; Keeley et al., 2012a; Pearse et al., 2012; Phelan et al., 2013; Kreiss et al., 2015; Patchett et al., 2015) have used isoflurane (gas) anaesthesia.

Interestingly, in the case of phencyclidine, despite the “large” dose administered (3mg/kg injected intramuscularly), the drug failed to relax the Devil it was being used on (Bartels et al., 1966). The significance of this lack of response is unclear.

Devils can be euthanased with an intravenous injection of sodium pentobarbitone (Loh, 2006; Loh et al. 2006a-b; Keeley et al., 2012d; Pearse et al., 2012).

Genetics

Work on the Devil’s genetics began with the description of its karyotype in 1923 and most recently saw the sequencing of its genome in 2010 and 2011.

Chromosomes

The normal Devil karyotype consists of 14 pairs of chromosomes, including an XY (males) or XX (females) sex pair (Greenwood, 1923: figs 22-25; Koller, 1936: fig. 1b; Martin and Hayman, 1967; Sharman et al., 1990; DPIWE, 2005a: fig. 1 on p. 25; Bostanci, 2005: fig. on p. 1035; Eldridge and Metcalf, 2006: fig. p. 30; Pearse and Swift, 2006: fig. 1; Connellan, 2006: fig. on p. 109; Pyecroft et al., 2007: fig 2; Quammen, 2008: fig. p. 1; Kreiss et al., 2011a: fig. 4a; Bender, 2010: fig. 23.2; Lachish et al., 2011b; Bender et al., 2012: fig. 1; Deakin et al., 2012: fig. 1; Tovar,

6/05/2016 8:41 AM 145

2012: fig.1.2a; Pye et al., 2016b:fig. 3). This is the karyotype in all dasyurids (Hayman et al., 1974; Young et al. 1982; Rofe and Hayman, 1985; Westerman and Woolley, 1989).

The dasyurid karyotype is not only conservative in terms of gross morphology but in the position of the genes as well. Of 105 genes mapped onto the Devil karyotype, for example, all but four mapped to the same chromosomes as in other dasyurids (Deakin et al., 2012).

From a practical point of view, a gene found only on the Y chromosome provides a molecular means of determining sex from tissue as well as an external body part such as hair and even scats. The gene is isolated, its DNA amplified and then the amplified DNA run on a gel for visualisation (Lachish et al., 2011b).

There is little information on variation within the Devil karyotype, although a pericentric inversion, not seen in other dasyurids and hence probably derived within Devils, occurs as an uncommon variant in pair 6 (originally identified as pair 5; see Table 15) in both sexes across much of the Devil’s range (Australian Geographic, Issue 84; AusVet, 2005; Pearse and Swift, 2006; Pearse et al., 2007; Pearse et al., 2012). In the wild, the variant occurs in about 2.6 to 4.5 percent of the population (Pearse and Swift, 2006; AusVet, 2005; Pearse et al., 2012). It usually occurs as a heterozygote, but a homozygous individual has been bred in captivity (Newsletter, December 2010, September 2011; STDP website: “Trowunna devils join Insurance Population,” 19 August 2011).

It was originally thought that this chromosome variant might play a role in the Devil’s facial tumour disease (Pearse et al., 2007; Pyecroft et al., 2007; possibly also Pearse in Denholm, 2007; see also Newsletter, September 2011; below), but nothing further has been heard about this idea.

For its relevance to the Devil’s facial tumour disease discussed below, it can be noted here that the Major Histocompatibility Complex of the Devil’s genome maps to the long arm of chromosome 4 (Cheng et al., 2012a: fig. 3; see also Cheng and Belov, 2014).

Devils show a dichotomy in the length of their telomeres, the structures at the end of the chromosomes that help maintain the physical integrity of the chromosomes. In each pair of chromosomes, one member has long telomeres and the other member has short telomeres. In the sex chromosome pair, the X, which is inherited from the mother, has short telomeres and the Y, which is inherited from the father, has long telomeres. This led researchers to infer that the chromosomes with short telomeres are inherited from the mother and the chromosomes with long telomeres are from the father (Bender et al., 2009; Bender et al., 2012). In fact, this dichotomy in telomere length seems to be a feature of dasyurids among marsupials, although the two other dasyuromorphian families were unavailable for analysis (Bender et al., 2012).

Based on a comparison with mouse telomere length, it was suggested that in dasyurids, telomere length has been lengthened in males and shortened in females (Bender et al., 2012). However, a comparison with non-dasyurid marsupial telomere

6/05/2016 8:41 AM 146

length might have been more indicative in determining whether only one or, indeed, both sexes have experienced a change in telomere length.

Telomere length, as estimated by telomere copy number, is said to not vary significantly (post-hoc Conover-Inman test, P = 0.52) between chromosomes from lung (N = 5) and spleen (N = 5) (Ujvari et al., 2012). However, re-analysis of the data (Ujvari et al., 2012: fig. 3) shows the difference lies just outside significance (Mann- Whitney U = 21.5, P = 0.056, pers. obs.), suggesting the reported P value is a typographical error. Given the sexual dimorphism in telomere length (above), it is also possible a sexual bias in the sample may have biased the comparison. In any event, the near level of significance suggests the issue of a difference between lung and spleen needs to be addressed with larger sample sizes and until then, the possibility of differences in telomere copy number among different tissues remains open. Whether various tissues differ significantly in the telomere copy number is important, because it bears on the interpretation of the telomere copy number in the Devil’s facial tumour (below).

In terms of chromosomal abnormalities, two Devils that were “internally female” but with a “scrotal rudiment” in place of a pouch had XO karyotypes based on leucocyte culture (Sharman et al., 1990).

Genome size

The Devil’s nuclear genome has about 3.63 pg of DNA (Martin and Hayman, 1967) or about 2.89-3.17 (Murchison et al., 2012) or 3.3 gigabase pairs (Miller et al., 2011). These values are similar to other mammals and very close to humans (3.2 Gb).

The Devil genome has a G + C content of 36.4%, similar to that of the opossum (37.8%) but lower than that of humans (45.2%) (Murchison et al., 2012).

The Devil’s mitochondrial genome has about 1700 base pairs (V. Hayes, on ABC Catalyst Online, 18 June 2009).

Genome sequence

The rapid spread of the Devil’s facial tumour disease pushed the species to the top of the waiting list for an entire nuclear genome sequence. Sequences from both healthy Devils and the tumour would allow comparison of the two and identification of ways in which the tumour differed. This would provide possible insight into how the tumour originated, how it is changing and what avenues means might be able to treat the tumour or prevent its spread to other Devils.

Curiously, two groups undertook the sequencing effort independently of each other. The first group was initially interested in whether a virus may have had a role in the origin of the Devil’s facial tumour (below) and began asking the Tasmanian authorities for some DNA samples either in late 2005 or early 2006 (Bevilacqua, 2006c, f, 2007; Duncan, 2006). The authorities knocked back multiple requests, both written and telephoned, because the CSIRO’s Australian Animal Health Laboratory was already doing the work (Bevilacqua, 2006c, g; Duncan, 2006; Stedman, 2007; Waterhouse, 2007). However, the CSIRO group was apparently looking for a virus

6/05/2016 8:41 AM 147 with other techniques (Bevilacqua, 2006a; Hyatt et al., 2007; Waterhouse, 2007; CSIRO media release, 4 June 2007; ABC Radio Australia, 13 August 2007) than sequencing (but see Wang, 2007).

The group persisted with its requests, possibly spurred on when a June 2006 training course introducing a new sequencing machine to its institution examined the DNA from a healthy Devil and found it “laden” with retroviral DNA (Bevilacqua, 2006g, 2007). Oddly, this finding never seems to have been published.

The availability of a new generation sequencing machine (Bevilacqua, 2006b) plus the recently published insight that the tumour probably had a genetic rather than viral cause (Pearse and Swift, 2006) seems to have shifted the group’s interest to the genetic cause of the disease. In October 2006, the group’s leader detailed the apparently new research focus, “By comparing the DNA of a healthy devil to that of the cancer, we can start to understand the disease. I hope to use sensitive genomic assays to determine which genes have gone awry in the cancer and how misregulation of these genes could lead to an infectious lesion (Murchison, 2006).” Despite this clear shift in the group’s research goal, up until late March 2007, the media continued to report the group’s goal as a search for a causal virus (Bevilacqua, 2007; Daily Telegraph, 25 March 2007; Sunday Mail, 25 March 2007).

The group’s trouble in getting the necessary tissues was picked up by the media in November 2006, perhaps due to the realisation the group’s leader had been raised in Tasmania (ABC News, 13 November 2006; Bevilacqua, 2006b, d-g; Hobart Mercury, 21 November 2006; Paine, 2006). As a result, in late November 2006, the relevant Tasmanian government minister promised the group DNA samples from both healthy and infected Devils (Duncan, 2006; Waterhouse, 2007).

The group received the tissues in March 2007 (Bevilacqua, 2007; Sunday Mail, 25 March 2007; Daily Telegraph, 26 March 2007). These included body and tumour tissues from 25 DFTD-affected Devils, body tissues from nine healthy and two DFTD-affected Devils, tumour tissues from two Devils and tissues from nine healthy Devils (Murchison et al., 2010: table S1). The group also had tissues from what was probably the last Devil to die (from cancer, but not DFTD) in an overseas zoo, an animal at the Fort Wayne Childen’s Zoo (Murchison et al., 2010: table S1; see “Overseas facilities,” below).

The group appears to have initially sequenced tumour and healthy tissue in order to confirm the clonal origin of the tumour and to gain further insight into the cellular origin of the tumour (Murchison et al., 2010). This work was complete no later than 14 August 2009, the date the paper was received by the journal.

When this group actually began sequencing the entire genome is unclear. In August 2007, the group was reported as “using new DNA sequencing machines … to sequence DFTD genes (Newsletter, August 2007; see also STTDP website: Cold Spring Harbor Laboratory, New York,” 17 August 2007).” And in early 2008, the group told some visiting Tasmanian politicians that it was sequencing the Devil’s cancer genes (Science Daily, 28 January 2008; Anderson, 2008; see also PhysOrg, 29 January 2008). According to another account, however, the group began sequencing in the latter part of 2009 (Darcé, 2010). In any event, the group finished sequencing one healthy genome and two DFT genomes in September 2010 (Darcé, 2010);

6/05/2016 8:41 AM 148

announced their results at a genetics conference in Hobart in September 2010 (Business Wire, 16 September 2010; Darby, 2010b; Pennisi, 2010), and published them in February 2012. This group had STDP funding but no Australian government funding.

Parenthetically, the entire genome sequencing effort appears to have required additional samples, from the healthy Devil and the two tumours (Murchison et al., 2012). When these were obtained is unclear. Furthermore, presumably, the group would have made an application for the tissues to the relevant Tasmanian and Australian Commonwealth authorities, and this application would have contained details of what the tissues were to be used for. Thus, the Tasmanian authorities would have known that the group was planning to sequence the entire genome.

This group’s intention to sequence the entire genome of a healthy Devil seems, however, to either have not been widely known or appreciated in the general Devil facial tumour research community, because in November 2007, when the first group was complaining about not having access to Devil tissues, the second group of sequencers was looking for someone to sequence the Devil’s genome (below; see McDonald, 2007). And in early 2010, the official list of STDP collaborators listed this group’s project as “DFTD cancer genome sequencing (STDP website: “Collaborators,” 2 February 2010).” Even as late as 2010, a review of marsupial genetics and genomics, did not list the first group as the one sequencing the Devil’s genome (Papenfuss et al., 2010: table 6.1; Marshall-Graves, 2010b).

The second group seems to have had a specific mandate for the sequencing work, as a Save the Tasmanian Devil Program Genetics Workshop held in Hobart in November 2007 agreed that this group would begin work sequencing “… the complete genome using new generation sequencing technology (Newsletter, March 2008; see also V. Hayes, ABC Catalyst Online, 18 June 2009).” In May 2008, the STTDP’s Senior Scientist wrote, “a large scale proposal has been put forward” by this group (McCallum, 2008b). In early 2010, the official list of STDP collaborators referred to this group’s project as “Next G sequencing of devil and tumour genome (STDP website: “Collaborators,” 2 February 2010).” The late 2010 review of marsupial genetics and genomics also listed this group as the only one involved in sequencing the Devil’s genome, while the first group was listed as sequencing the Devil’s tumour (Papenfuss et al., 2010: table 6.1; Graves, 2010b).

The second group began forming toward the end of 2007 (Hayden, 2009; Nicholls, 2009; ABC Catalyst, 18 June 2009). It received what appeared to be its major funding in August 2008 (Gordon and Betty Moore Foundation; see also Minister for Innovation, Industry, Science and Research media release of 5 November 2008) and began sequencing in “late 2008” (Darcé, 2010). It finished sequencing two devils and a tumour from one of them in about July 2010 (Darcé, 2010). It submitted the results for publication in February 2011 and saw them published in June 2011 (Miller et al., 2011). This group apparently had no funding from the STDP or any Australian government source.

The second group bridled at the first group’s “publication” prior to peer review, but researchers in the two competing human genome projects in the late 1990s had

6/05/2016 8:41 AM 149

already established the precedent, at least in genomic work, for announcing results prior to publication (Pennisi, 2010).

In any event, the ability to sequence the complete nuclear genome within a relatively short time, about two months currently, makes it possible to compare genomes from different animals on a nucleotide by nucleotide basis for possible relevance to population variability, sub-structuring and disease development and resistance (V. Haynes on ABC Catalyst, 18 June 2009; Pennisi, 2010). Indeed, by late 2015, the whole genomes of seven other Devils had been sequenced (Wright et al., 2015), bringing the total to ten.

Finally, it should be mentioned that the second group had apparently sequenced the entire mitochondrial genome by, at the latest, June 2009 (V. Hayes, on ABC Catalyst Online, 18 June 2009). More recently, the complete mitochondrial genomes of 500 Devils from seven populations have been sequenced (Brüniche-Olsen, 2011), although the results and analyses have not yet been published. Reference was also made to mitochondrial data from samples c. 15 000 ybp (Brüniche-Olsen, 2011).

Immunome

The immunome is that part of the genome involved in the immune response. It involves major groups of genes scattered throughout the genome. Geneticists became interested in the Devil’s immunome in the mid-2000s, when it became evident the Devil’s immune system was not detecting and eliminating the Devil’s facial tumour (below). Prior to this, nothing was known of this part of the Devil’s genetics.

One of the major components of the genome is the major histocompatibilty complex (MHC). It functions in assessing protein fragments from both “self” and “other,” and setting in train the immune response against “other.” The MHC occurs only in vertebrates.

The Devil’s MHC maps to the long arm of chromosome 4 (Cheng et al., 2012a; see also Cheng and Belov, 2014). It encodes both Class I and Class II molecules. Class I molecules, occur on the surface of an individual’s nucleated cells, that is, virtually all cells except red blood cells, and bind the individual’s own antigens (intracellular protein fragments) thereby presenting those cells as “self” to the individual’s immune system. But should foreign cells find their way into another individual, that individual’s immune system would recognise the antigens presented by these cells as “non-self.” Class I molecules also present antigens from intracellular viruses, bacteria and mutated (cancerous) cells, which the immune systems recognises and reacts to.

Class I molecules can be divided into Class Ia and Ib molecules. Class Ia molecules have wide tissue expression and are relatively polymorphic while Class Ib molecules generally have more limited tissue expression and are less polymorphic.

Class II molecules occur on the surface of only a few cell types, mainly dendritic cells, macrophages and B lymphocytes. Class II molecules bind externally-derived “non-self” antigens for presentation to the immune system and the initiation of an immune response.

6/05/2016 8:41 AM 150

The Devil has three Class Ia genes, designated Saha-UA, -UB and -UC (Siddle et al., 2007a; Cheng et al., 2012a; Lane et al., 2012; Cheng and Belov, 2014; note: “Saha” derives from the first two letters of the species scientific name), with six, ten and seven recognised alleles, respectively (Lane et al., 2012: table 1). The UA gene shows geographic variation. It is present in all individuals in the southeast but is lacking in 54.8 percent of individuals in the northwest (Cheng et al., 2012a; see also Lane et al., 2012). All these alleles code for different amino acid sequences, that is, they are non- synonymous.

As expected, the Devil’s three Class Ia genes were found to be expressed in eleven different types of tissue (Cheng and Belov, 2014).

The Devil also has five Class Ib genes, designated Saha-UD, -UK, -UM, -MRI and -CDI, with four, one, one, one and four recognised alleles, respectively (Cheng et al., 2012a; Cheng and Belov, 2014). Two of the UD alleles code for the same amino acid sequence, that is, they are synonymous. UD was first recognised as an unclassified Class I gene (Siddle et al., 2007a; Cheng et al., 2012a; Cheng and Belov, 2014). Of the four CDI alleles, in wild Devils (N = 40) one occurs at a frequency of nearly 91 percent and the remaining three occur at frequencies of 2 to 4 percent. In captive Devils (N = 30), however, the three rare genes were undetected (Cheng and Belov, 2014), suggesting a bias of unknown cause(s) in captive Devils.

Four of the Devil’s Classs Ib genes, UD, UK, UM and CDI were expressed in only some, not all, eleven Devil tissues tested, but the fifth gene, MRI, was expressed in all eleven tissues (Cheng and Belov, 2014).

The Devil has two Class II gene families, DA and DB, although the latter is represented by only one non-functional remnant, or pseudogene, designated DBB (Cheng et al., 2012b). Within the DA family there are two genes, DAA and DAB. The DAA gene has one locus and one allele. The DAB gene has three loci, DAB1, DAB2 and DAB3, with three, six and three alleles, respectively (Cheng et al., 2012a-b). The allele frequencies are heavily skewed at each locus, with the dominant allele occurring at a frequency of 85 (allele 01), 82.5 (allele 03) and 90 (allele 05) percent at the three loci, respectively (Cheng et al., 2012b: fig. 5).

The DB family is represented by “multiple functional” genes in other Australian marsupials (Tammar Wallaby and Brushtail Possum), but where it was reduced to a pseudogene in the ancestry of the Devil will require further analysis of the Devil’s part of the evolutionary tree.

Natural killer cells, which are among the earliest cells to react in the immune response. They have two “super-families” of receptors on their cell surfaces, the natural killer complex (NKR).and the leucocyte receptor complex (LRC). The genes of the former occur on chromosome 5 and the genes of the latter to chromosome 3. While many of the genes encoding these receptors are shared between the Devil and the South American Opossum (Monodelphis domestica), the only two marsupials studied to date, the Devil has a number of genes that differ from the opossum, leading the authors to conclude “… devils could possess a collection of devil-specific receptors not found in eutherians or opossum” (van der Kraan et al., 2013). Perhaps the wording is just infelicitous, but without a broader survey of Australian marsupials,

6/05/2016 8:41 AM 151

it is premature to claim the unique genes identified in the devil (compared to the opossum) occur only in the devil.

Toll-like receptor genes are another major group of genes in the immunome. They occur in invertebrates and vertebrates and hence are among the earliest evolved genes in the immunome. The protein products of these genes occur on certain cells of the immune system, where they detect and bind molecules that are unique and widespread in invasive microbes such as viruses, bacteria and fungi; they also bind to molecules produced by damaged host cells. Once bound, the toll proteins activate inflammatory genes that then stimulate further immune responses. Immunogeneticists have found ten toll-like receptor genes in Devils (Cui et al., 2014, 2015a), all of which appear to be functional (Patchett et al., 2015a). Seven of these genes have only one allele and three have two (Cui et al. 2014, 2015a). By way of comparison, the Koala has ten identified toll-like receptor genes, one of which has one allele, two have two alleles, four have four, one has five, one has ten and one has 12 alleles (Cui et al., 2015b).

Immunoglobins are antibodies made by B cells. Immunoglobin genes vary due to somatic mutation and random rearrangement, and this variability allows for an eventual “fit” to any antigen. Once a fit is found, the cell line producing the antibody proliferates, and after the crisis has passed, the cell line continues, ready for a rapid response should the pathogen reappear (immunological memory). The genetics of the structural genes underlying the immunological active region of the immunoglobins has been characterised in Devils and found to be very similar to the two other marsupials similarly characterised to date, a South American opossum and the Australian Brush-tailed Possum (Ujvari and Belov, 2015). The researchers said their results provided further evidence the Devil’s “… immune system is capable of recognizing a broad range of antigens….”

The antigen receptors on T cells recognise protein fragments bound to an MHC molecule of another cell. These T cell receptors (TCRs) are composed of constant and variable domains. Devils have all five constant domains recognised in marsupials, α, β, γ, δ and μ with one, three, one, one and six (text) or seven (table 8) genes, respectively, identified to date. By way of comparison, the American Opossum has eight μ genes (Morris et al., 2015a).

The antigen receptors produced by B cells occur either in the B cells’ membranes as immunoglobins (Ig) or in soluble form as antibodies. Immunoglobins consist of a heavy and a light chain, and each chain has a constant and a variable region. Immunoglobins are characterised by the constant regions in their heavy chain. Immunoglobins also have two forms of light chains, κ and λ. Marsupials have five of the six Igs know, A, G, E and I, each represented by a single gene; D is not know in the group. The Devil is typical in its Igs. Marsupials also have both forms of light chains, the κ form has one gene and the λ form a variable number. Devils have one κ gene and at least four λ genes. In comparison, the American Opossum has seven λ genes (Morris et al., 2015a).

Chemokines are small chemoattractant proteins that stimulate the migration and development of cells, especially phagocytes and lymphocytes. Thirty-nine (text) or 41 (table 3) chemokine genes have been identified in the Devil. In comparison, the American Opossum has 36 and humans 47. The Devil has its own apparently unique

6/05/2016 8:41 AM 152

gene expansions (CCLD5-8 and 10) in a region noted for expansions in marsupials, and one unique and divergent gene duplication (CXCL10A-B) (Morris et al., 2015a).

Interferons are proteins involved most notably in the induction of cells’ responses to viruses. There are three classes of interferons, I – III. In mammals, Group I has six subgroups: α, β, δ, ε, ω and κ and of these the American Opossum has three: α, β and κ, and of these, the Devil apparently has only two α and β. Mammals have many α genes. Humans have 13, the opossum seven but the Devil only four, although it may have two additional α pseudogenes (essentially defunct genes). A β gene common to both the opossum and Devil has also been found. Group II has one group, γ, with one gene. Most mammals have this gene, including the Devil. Group III has one group, λ. Eutherian mammals have three genes. The Devil has two genes, which may have represent a unique duplication in marsupials (Morris et al., 2015a).

Interleukins are cytokines produced by a variety of cells, including leucocytes. There are a large number of interlukins, and they have a variety of functions, including cell activation, chemotaxis, growth, differentiation, maturation. Forty (text) or 45 (table 1) interleukins have been identified in Devils (Morris et al., 2015a). In comparison, 36 interleukins have been identified in humans. Of most interest, the Devils has three variants of IL18, whereas most species have only one, while the opposum has two. The Devil may also have two variants of IL22 whereas most species have only one (Morris et al., 2015a).

Tumour necrosis factors are proteins with a wide variety of functions, the most important of which, arguably, is the stimulation of apoptosis (cell death). Eighteen TNF genes have been identified in the Devil (Morris et al., 2015a: table 5).

Cathelicidins are proteins released from certain cells in vertebrates that have both antimicrobial and immuno-regulatory functions. In marsupials, they help protect immunologically naïve pouch young from infections. Immunogeneticists have identified six cathelicidin genes in the Devil and synthesised proteins of two that show anti-fungal activity (Peel et al., 2014).

A variety of colony stimulating factor and transforming growth factor genes that are widespread in mammals have also been found in Devils. Notably, the OSM gene, which encodes a protein that inhibits several types of tumour cell growth, is very divergent from eutherian humans (19 percent similarity in amino acid sequence) and even the marsupial opossum (23 percent similarity) (Morris et al., 2015a).

Transposable elements

Transposable elements (TEs) are sections of DNA that can move within or between genomes. TEs probably move between one organism and another in body fluids, either directly or via a parasite. TEs can occur in 106 copies in a single genome and represent a high percentage of an organism’s DNA, as much as 52 percent in some vertebrates. TEs can enter a genome at any time, and some are hundreds of millions of years old. Over time, TEs can lose their ability to transpose themselves.

There are two broad types of TEs, those that require transfer RNA for insertion (retroposons or Class I) and those that insert directly (Class II). In animals, Class I

6/05/2016 8:41 AM 153

TEs are an order of magnitude more common than Class II. TEs of both Class I (Nilsson-Janke et al., 2012) and Class II (Gilbert et al., 2013) have been explicitly characterised in Devils.

TEs are highly conserved, suggesting they have a function, possibly regulatory. As highly conserved sections of DNA, they are useful for inferring deep phylogenetic relationships (see “Phylogenetic Relationships;” Nilsson-Janke et al., 2012; Gilbert et al., 2013 and Zemann et al., 2013).

Genetic differences among tissues

A fortuitous finding during the course of work on the genetics of the Devil’s facial tumour disease was that nerve tissue had high levels of genomic AFLP polymorphism compared to brain, heart, kidney and liver tissue combined, 83 and either six (table) or eight (text), respectively and in the same individuals (Ujvari et al., 2013: table 1). The significance of this difference is unclear, but in that DFTD appears to have arisen in a nerve tissue, this tissues inherent higher variability is intriguing (Ujvari et al., 2013).

Data on the number of methylated sites between nerve and brain, heart, kidney and liver tissue combined have been published (Ujvari et al., 2012: table 1), but no statistical analysis of the differences was presented.

Geographic genetic sub-structuring

A number of studies bear on the genetic sub-structuring in the Devil population. In thinking about the results, however, it is important to remember that for the eastern and central parts of the Devils range, many, perhaps most, of the studied tissue samples have come from areas where DFTD was well established. If DFTD has exerted any selective bias within the affected populations (see Lachish et al., 2011a), the revealed geographic patterning will reflect post-disease conditions, not pre- disease.

In 2004, the analysis of 11 microsatellite loci in six populations (above) provided the first evidence of genetic sub-structuring across the Devil’s range when it showed that the semi-isolated population in the northwest part of the range was different genetically from five populations in the eastern part (Jones et al., 2004a; see also Brüniche-Olsen et al., 2013). In October 2010, sequencing work found that this population also had a unique single nucleotide polymorphism in the mitochondrial locus control region (Murchison et al., 2010: fig. S1, table S2). And in 2012, sequencing of the MHC Class II DAB gene found slightly higher heterozygosity at the three loci as well as a unique allele (SahaDAB*07) in the northwestern population (Cheng et al., 2012b).

A further insight into the genetic sub-structuring of the Devil appeared first in the media in March 2009 with a report that there were differences in certain SNPs, between western and eastern Devils (Hayden, 2009). It was not clear at the time whether these SNPs were in the mitochondrial or nuclear DNA. A few months later more results were released, again in the media. In an interview, one of the researchers said a survey of SNPs in the mitochondrial DNA of “almost 200” animals from across Tasmania had revealed five regional “groups” (haplotypes), one of which was the

6/05/2016 8:41 AM 154

western population from which most of the facial tumour disease-free captive breeding population has been taken (below). This was the only general area, apparently, in which there was no variation in the loci examined (V. Hayes, on ABC Catalyst, 18 June 2009).

When the researchers published their results in July 2011, it became clear that analysis of 17 SNPs in 175 individuals returned the five extant “region-specific” groups or haplotypes and one “ancient” group of apparent unknown provenance. Four of the extant haplotypes (a-d) were widespread in the central and eastern parts of Tasmania and one (e) was confined to the northwest (Miller et al., 2011: fig. 3a). The northwest population showed no internal variation. The published work also included an analysis of 702 SNPs in the nuclear DNA, which identified eastern, central and northwestern groups (Miller et al, 2011: fig. 3b).

Analysis of 1482 informative SNPs from across the genome in six populations identified two clusters corresponding with geography, one in the northeast (Forestier, Freycinet, Mt William and Narawntapu) and another in the northwest (Arthur River and Woolnorth). Furthermore, the two northwest populations had 13 and 11 private alleles, whereas the four northeast populations had between eith (Freycinet) and no (Forestier) private alleles (Brüniche et al., 2016).

In February 2012, other researchers reported results from sequencing the mitochondrial DNA of 92 Devils from 25 localities across Tasmanian. They found evidence for six “normal” mitochondrial halplotypes, three of which were widespread in central and eastern Tasmania and three of which were confined to the northwest (Murchison et al., 2012). While the number of haplotypes identified in this work (6) was similar to the number identified in the earlier work (5), only four correspondences were readily determinable from the two published maps (pers. obs.; Table 14). The researchers also published an unrooted tree of their six haplotypes, and the northwest haplotype (one of the three haplotypes identifiable in both works) was the most divergent (Murchison et al., 2012: fig. 3c). Why this group (paper received by journal on 9 September 2011) did not try to reconcile their haplotype designations with those of the first group (paper published 26 July 2011) is unclear, as two authors were common to both papers.

Table 14. An attempt to reconcile the mitochondrial haplotypes identified in Miller et al., 2011 and Murchison et al., 2012. Colour codes follow original designations.

Miller et al., 2011 Murchison et al. 2012 Pink Blue Green Dark Green Blue Red Yellow Pale Green Dark Pink ? ? Orange ? Pink

Interestingly, in early 2010, geneticists had sequenced an 1180 base pair fragment of the mitochondrial locus control region and found that devils from throughout the range were similar except for the animals from near coastal areas of northwestern

6/05/2016 8:41 AM 155

Tasmania that shared a unique single nucleotide polymorphism at position 15 711 (Murchison et al., 2010: fig. S1a).

In 2010, immunogeneticists published evidence for two distinct nucleotide sequence lineages, groups 1 and 2, in the Class I segment of the MHC. Most Devils were reported to have both groups, but some Devils in the northwest had only group 1 or group 2 (Siddle et al., 2010a-b). This report received enthusiastic media coverage (University of Sydney media release, 10 March 2010; AAP, 10 March 2010; ABC News, 10 March 2010; Denholm, 2010c; Mather, 2010; Smith, 2010), because it had implications for some Devils’ potential ability to mount an immune response to the DFT (below). The then federal innovation minister said “This discovery is great news for Australian research and for the Australian public” and the then head of the federal granting agency called it an “extraordinary discovery” (Mather, 2010; ARC media release, 10 March 2010).'' Unfortunately, the result was due to a technical error (Lane et al., 2012), which render both its geographical findings and immune response implications incorrect (Cheng et al., 2012a; Lane et al., 2012). Curiously, the correction was presented as a refutation of an hypothesis (that some Devils were genetically resistant) instead of the correction of an error (University of Sydney media release, 7 June 2012; The Conversation, 8 June 2012; ABC World Today, 12 June 2012).

The original 2010 article also identified 25 MHC types in the Class I α 1 domain and found that there was limited MHC and contained sequence diversity in the southeastern two-thirds of the state compared to the northwestern one-third (Siddle et al., 2010a: fig.1). This result added to the apparent distinctiveness of the two populations. However, the number of MHC types and the sample size for the 15 populations were strongly correlated (Spearman rank correlation = 0.78, P < 0.004, pers. obs. on data in Siddle et al., 2010a: fig. 1 and table 5), suggesting that at least some of the 15 samples may not have been large enough (N ranged 4-59) to capture all the local diversity.

Subsequently, passing mention was made of two western devils having three MHC I and two MHC II alleles that were not present in an eastern population analysed (Kreiss et al., 2011b). The details of this important finding have apparently never been published.

In 2012, immunogeneticists showed that a MHC Class I gene, UA, appears to show clinal variation across the state, occurring in 100 percent of the population in the southeast but declining to about 45.2 percent of the population in the northwest (Cheng et al., 2012a: fig. 1). In other words, the gene becomes less frequent from southeast to northwest and adds to the distinctive genetics of the northwestern population.

In 2015, immunogeneticists examined the diversity in the toll-like receptor genes in 25 Devils. The geneticists mentioned they had 14 Devils from the eastern “subpopulation” and 11 from the northwest. They did not, however, mention this geographic difference in the results or the discussion (Cui et al., 2015a), so presumably no differences were found.

6/05/2016 8:41 AM 156

The broad geographic variation of all these genetic markers and the nature of the transitions between them are still only poorly known. The most marked geographic substructuring is between northwestern and all other populations, a difference that is especially interesting in light of the apparent resistance of northwestern populations to DFTD.

It is also evident that there is enough geographical sub-structuring among Devils to question the rationale for translocating individuals for whatever reason.

Finally, it is worth noting that the genes in almost all examined healthy populations have been in Hardy-Weinberg equilibrium. For microsatellite loci, one study of 11 loci (Sh2b, Sh2g, Sh2i, Sh2p, Sh2v, Sh3a, Sh3o, Sh5c, Sh6e, Sh2L, Sh6L) found no deviations from HWE (Jones et al., 2003, 2004a). Another study of ten of 11 original loci (Sh6b said to be excluded but there was no Sh6b in the original) found significant deviations at two loci (Sh2p, Sh2v) at two sites (Lachish et al., 2011a). Another study of ten of the original loci (Sh3a excluded) found that one locality (Narawntapu) in one year (2007) deviated significantly from HWE. This was attributed to a DFTD-free population resisting immigration from DFTD-affected areas (Brüniche-Olsen et al., 2013). Another study of seven of the original loci (Sh2b, Sh2g, Sh2p and Sh3a excluded) found that one locus (Sh3o) was significantly different from HWE (Hogg et al., 2015).

For MHC class II genes, none of the three DAB loci examined (DAB 1-3) showed significant deviation from HWE, a result interpreted as evidence of random mating (Cheng et al. 2012b).

A survey of 94 SNP markers in a variety of immunome genes found that 30 deviated from HWE (Morris et al., 2015).

Several factors can cause a deviation to HWE but only one seems to clearly apply to Devils, non-random mating. Female Devils are said to prefer to mate with older, larger and more aggressive males and “… a few male devils dominate paternity” (see “Mating” and “Breeding Participation”). But whether this behaviour is responsible for any of the deviations from HWE in Devils is unknown.

Temporal genetic change within populations

Analysis of ten polymorphic microsatellite loci at two localities, Marrawah and Narawntapu National Park in northwest and central-north Tasmania, respectively, showed no significant change in genetic diversity between 1999, and 2007 and 2006, respectively (Brüniche-Olsen et al., 2013).

There was also no significant change in effective population size at either locality over the same period (Brüniche-Olsen et al., 2013). Neither result is too surprising, as the period represents only three to four generations.

Given these results, it is surprising that the inbreeding coefficient increased significantly in both populations in the same period. In Marrawah’s case, this was attributed to the small population size and lack of gene flow (Brüniche-Olsen et al., 2013). However, as any trend in either factor could not be sustained long term, the

6/05/2016 8:41 AM 157

factors either were new in the population or are variable. In Narawntapu’s case, the increase in the inbreeding coefficient was attributed to immigration, inferred from gene flow estimates (Brüniche-Olsen et al., 2013).

An analysis of potential genetic change in response to DFTD using separate temporal and spatial analyses of a large number of SNPs from across the genome at six localities concluded that the allele frequency changes observed were more likely due to genetic drift, arising from small effective population size, than selection (Brüniche- Olsen et al., 2016).

Genetic diversity of Devils

In 2004, an analysis of 11 polymorphic microsatellite loci in six populations sampled in 1999 from throughout much of the Devil’s range indicated that genetic diversity, as measured at each locus both by the mean number of alleles and the degree of heterozygosity was low in comparison with a wide variety of species, including other dasyuromorphians (Jones et al., 2004a). All samples bar one were pre-DFTD; the exception (Little Swanport) coincided with the arrival of DFTD.

Parenthetically, a subset of these same 11 microsatellite loci have been used in other studies (Lachish et al. 2011a - ten; Brüniche-Olsen et al., 2013 – ten loci; Hogg et al., 2015 – seven loci). Further, a set of 11 microsatellite loci in the MHC complex have also been characterised (Cheng and Belov, 2012a), and although it was not clear where or when these samples were taken relative to the arrival of DFTD, interestingly, the mean number of alleles, mean observed heterozygosity (HO) and mean expected heterozygosity (HE) did not differ significantly between the two groups (Mann Whitney U-Test on data in Jones et al. 2003b: table 1 and Cheng and Belov, 2012a: table 2; pers. obs.).

In 2011, another group sequenced the genome of two Devils from the extremity of the range (NW and SE) and found 914 827 single nucleotide differences between the two. Using comparable methods, they determined these were only about 20 percent of the differences between a Bushman and a Japanese person and 28 percent of the differences between a Chinese and Japanese (Miller et al., 2011). Apparently, there are no similar comparisons for any other marsupial species.

This group also sequenced the mitochondrial DNA in 13 Devils, including two from the extremes of the range (NW and SE) and found an average difference of 10 base pairs between any two individuals. This value was second lowest among 12 species ranging from Bushman humans (85) to thylacines (5), the highest, gorilla (92), omitted here on the chance that it included two species (Miller et al., 2011: fig. 2c). Interestingly, the two lowest ranking species, the Devil and the Thylacine, were the only marsupials in the comparison (Miller et al., 2011: fig 2c).

The group also estimated that the number of amino acid differences between these same to Devils was between 3000 and 4000, and the number of protein variants in Devils was about 1141. However, they made no comparisons with other species (Miller et al., 2011).

6/05/2016 8:41 AM 158

Immunogeneticists working on the Devil’s Major Histocompatibility Complex have also expressed opinions about the diversity in this part of the genome. Class I genome was said to show low diversity. However, the comparisons were not rigorous. In one case, it was based on a comparison with only two other species, humans and Gir Lions, both of which have high variability (Siddle, 2007b). There was no comparison with other species, especially other marsupials. In the second case, the opinion was based on a comparison with three other species, with Devils ranked third least variable on the criterion of total number of “variants” across the total number of “loci.” And although an allusion was also made to the level of sequence divergence between variants, no comparative data were given (Siddle et al., 2010a). In none of the comparisons was a statistical test applied.

The evidence for low diversity in the Devil’s Class II genome has been more convincing. For example, Devils have only one functional family of genes, DA, whereas other marsupials can have up to three families, Devils having probably lost the other two, DB (aside from a remnant pseudogene) and DC. Devils have only four DA genes, DAA and DAB1-3, which is about as many DA genes as most other studied marsupials (N = 5). The Devil’s DAA gene has only one allele, but so do all other studied marsupials. And its DAB genes have 12 alleles whereas other studied marsupials have between five and 47 alleles (Cheng et al., 2012b: table 3).

There is also evidence for low diversity in the Devil’s toll-like receptor genes. Of the ten identified genes, seven have only one allele while three have three alleles (Cui et al., 2015a). In comparison to the only other marsupial for which there is comparable information, the Koala, has two toll-like receptor genes, one of which has one allele, two have two alleles, four have four, one has five, one has ten and one has 12 alleles (Cui et al., 2015b).

A mid-2015 study examined further the putative low genetic diversity of Devils by examining SNP diversity in 167 immune genes, including some cytokines (chemokines, interferons, interleukins and tumour necrosis factors), natural killer cell receptors (leucocye and killer cell immunological-like) and other immune genes (Morris et al., 2015b). The study found “… extreme paucity in genetic diversity within these genes …” and concluded, “… genetic diversity at functional loci in the devil appears to be critically low ….” However, specific comparisons were made with only a few genes and a few species, two fish, a rodent and five primates. There were no comparisons with other marsupials.

In work published in October 2015, “over six million” SNPs were identified among ten whole Devil genomes (Wright et al., 2015). In comparison, as of June 2015, the number of identified SNPs in the four animal species for which there were the most complete data, humans had over 37. 5 million, domestic chicken over 10.9 million, domestic cat over 3.2 million and horse over 95 thousand. Perhaps only the human count is indicative due to the attention it would have received.

Opinion about when Devils may have lost genetic diversity have changed over time, especially with the analysis of ancient DNA from mainland fossil Devils. Originally, two possible time periods were advanced. First, it was suggested that the reduction occurred due to founder effect when the Tasmanian population was isolated by the opening up of Bass Strait some 14 000 years ago (Jones et al., 2004a; see also Cheng

6/05/2016 8:41 AM 159

et al., 2012a-b). There is an estimate of the size of the Devil population at this time of 500 (H. McCallum, in Pickrell, 2007), but it is unclear how this number was arrived at.

Second, it was thought repeated periods of low population size in the Tasmanian population during European times, that is, over the last 200 years may have led to reduced diversity (Jones et al., 2004a; see also Cui and Belov, 2012).

The analysis of fossil Devil DNA, however, has offered two other possible times. Originally, these were identified as “during the Pleistocene and Holocene” (Morris et al., 2013; see also Morris et al., 2011; University of Sydney media release, 5 December 2012; Morris, 2013). This was later specified as 2000 to 4000 years ago, a period of El Niño aridity and 22 000 to 48 000 years ago, a time approximately coincident with the last glacial maximum and also a period of aridity (Brüniche-Olsen et al., 2014; see also Brüniche-Olsen and Austin. 2014; Milman, 2014; Phillips, 2014; Shine, 2014). A bibliographic quirk of interest to those with such a turn of mind is that the later work dealing with these analyses of ancient DNA, Brüniche-Olsen et al., 2014 and other references cited, did not cite any of the earlier work, Morris et al., 2013 and other references cited, despite both efforts sharing a common principle researcher.

In 2014, a conservation article said Devils “… have already lost about half of their expected genetic diversity a long time in the past (Jones et al., 2014).” However, the reference cited for this statement (Jones et al., 2007a) contains no such information.

In 2015, a researcher who has worked on Devil genetics in the past claimed that Devils were “inbred’ (B. Ujvari, in Azvolinsky, 2015), a claim not made by any other Devil geneticists.

Regardless of its timing, the loss of genetic diversity has been attributed to two non- exclusive causes. The most common suggestion is small population size, usually in association with climate change. As populations become smaller, inbreeding increases and the chances that rare alleles will be lost increase.

The other suggested cause is a period of intense selection leading to the survival of only a small part of the Devil’s original genetic diversity (Cui and Belov, 2012; Morris et al., 2013). Such a period is thought to have occurred during the late 19th and early 20th centuries when a disease may have swept through many dasyuromorphian species (see “Cause of the early 1900s ‘decline’”). The evidence from ancient DNA, however, is that the reduction in the Devil’s genetic diversity predates this period.

There may have also been some recent additional loss of genetic diversity. Mitochondrial DNA from seven recent individuals and six museum specimens collected between 1870 and 1910 (precise date within this period unknown) and 1994 confirmed that genetic diversity in Devils was relatively low (above). But it also showed that while five haplotypes were common to both recent and historical specimens, a sixth haplotype was present only in a historical specimen. Interestingly, the apparently “missing” haplotype was from what was either the oldest or the second oldest specimen (1870-1910 vs 1908) (Miller et al., 2011), a period that includes one of the Devil’s reputed historic population lows.

6/05/2016 8:41 AM 160

Although the claim that Devils have low genetic diversity is now widely held, more could be done to substantiate it. Microsatellite data appears to support the claim, because in comparison to other marsupials in general and other dasyuromorphians in particular, Devil genetic diversity is low. The claim that Devils have low genetic diversity in other parts of their genome may be premature, however, as it rests on only very broad phylogenetic comparisons. More information on the genetic diversity from other parts of the genome in more marsupials is needed to support the claim and to determine when in marsupial evolution the Devil’s inferred low genetic diversity, if it is that, actually evolved.

In light of the prevailing sentiment that Devils have low genetic divereity, it is interesting that a recent report said some Devils near Lake Augusta in central Tasmania had “alleles” (note plural) that had never been seen before (D. Pemberton, on ABC News online, 13 November 2015). The report did not say where the alleles were in the genome, but the discovery of new alleles raises some interesting implications for the “Devils have low genetic diversity” claim.

Interactions with Humans

Devils and humans have had an intense relationship, perhaps since the beginning of human occupation of Australia. Aborigines considered it worthwhile to paint them and in one case create from them one of the most remarkable body adornments, a canine tooth necklace (below), ever recorded for Aboriginal people. For their part, Europeans first loathed them but more recently have accorded them “iconic” status and seen them as “a jewel in the crown of Tasmania’s biodiversity (STTDP, 2007c; see also DPIWE Annual Report 2003/04).”

Aborigines

Tasmanian Aborigines apparently called the Devil, “purinina” (Booth, 2007), or “tardebar” (Mooney, 1992), “tarrabah” (Guiler, 1992) or “tardibar” (Robertson, 2005). The word “tarrabah,” at least, is said to be onomatopoeic (Guiler, 1992).

The Aborigines had explanations for at least two aspects of its biology. They thought that the Devil got its pattern when, through greed, “it tried to catch animals in a bush fire and got burnt, rolling over and over in the soot (Mooney, 1992).” And it got its red ears because it “… is so bad tempered (presumably from the fire) that the other bush animals are constantly talking about it (Mooney, 1992).”

Whether Aborigines ate Devils is controversial. Some writers say they did (Guiler, 1992); others infer they did not (Noetling, 1910; Hall and Jones, 1990). In Tasmania, at least, no Devil bones occur in archaeological sites (Cosgrove et al., 2014). On the mainland, however, Devil remains are sometimes found in middens, suggesting that at least some Aborigines ate them (Kershaw, 1912; Mahony, 1912; Gill, 1953). Taken together, these reports suggest there have been regional differences both in the abundance of Devils and in the “tastes” of local people.

The discovery of a single Devil tooth necklace found on an Aboriginal at a burial site in southwestern New South Wales raises questions about the aesthetic and symbolic

6/05/2016 8:41 AM 161

significance of Devils to Aboriginals. The necklace was found around the neck of a rather tall male skeleton in a shallow grave, dated at about 7140 years (Pardoe, 1995). The necklace, which would have reached the pelvis of the wearer, contained 178 teeth, all canines (uppers and lowers), representing a minimum of 47 animals. Considerable effort had gone into making the necklace, because each tooth had to be pierced through the root in order to be strung on the necklace (Macintosh et al., 1970; Macintosh, 1971: plate 2; Flood, 1983: plate 6). The necklace is unique both as an ornament and as an example of Aboriginal people piercing teeth for any purpose. That such an object exists so fully developed and without precedent makes it truly astonishing.

The original describer of the necklace devoted considerable analysis and thought to the necklace and developed some interesting ideas about it. Careful analysis led to the conclusion that the individual teeth were of different ages and showed different styles of workmanship among themselves. This suggested that the necklace might have been in existence for several human generations and was important to those generations, perhaps as a symbol of a totem being. But the fact that the necklace was also finally “laid to rest,” as it were, suggests that its original relevance may have declined to such an extent that it could be “retired.” And the reason for its loss of relevance, may have been the decline of the Devil in this part of the mainland, as part of its general decline, and hence its loss of meaning to the local inhabitants. If the Devil did have totemic importance, however, it may have been very local. This might account for the necklace being unique.

Aboriginal people in northern Australia also depicted the Devil in rock paintings (Lewis, 1988: figs 1-2; see also Calaby and Lewis, 1977: figs 1-2 and McCarthy, 1976: fig. 1). The clearest painting has been dated stylistically at between 6000 and 9000 years (Lewis, 1988).

Settlers

The early settlers seem to have had only one direct economic use for Devils. They used the animals’ body fat as a leather preservative (Guiler, 1992). The convicts ate Devils, which were reported to taste not unlike veal (Harris, 1808), but Devils never became a staple in the settler diet. A trapper active on the remote west coast in the mid-1930s reported that Devils tasted like wallaby (S.L. Larnach, in Macintosh, 1971).

Early trappers had no interest in Devil fur. They were more interested in wallabies, possums and rabbits (Guiler, 1957). In fact, trappers came to hate Devils, because they ate the animals in the wire snares and would even chew through the snares if they got caught themselves (Fleay, 1952; Green, 1967; Guiler, 1972). One Devil caught in a snare was wearing six other snare “necklaces” (L.D. Connell, in D. Macdonald, 1927). Trappers were so bothered by Devils that they laid out strychnine and set auxiliary traps to control them in areas where they were snaring other mammals (Guiler, 1982). Devils can chew through more that the flexible snare wire. They can also chew through cyclone mesh fencing (Guiler, 1992: fig. 39; see also Fleay, 1946).

Early farmers and graziers also hated Devils, because they believed the animals destroyed stock. In 1930, the Van Diemen’s Land Company even put a bounty on

6/05/2016 8:41 AM 162

Devils on their properties in the northwest part of the state. The price was, in today’s money, about 25 cents (2/6d) for males and 35 cents (3/6d) for females (Mooney, 1992; Parks and Wildlife Service Tasmania, 2009). Devils probably do occasionally take a newborn lamb or incapacitated sheep. But they are unlikely to be able to catch and kill a healthy sheep (Guiler, 1972). Moreover, in scavenging dead stock Devils keep the paddocks clean and parasites like sheep tapeworm and blowflies in check (Guiler, 1972, 1992). This “cleaning service” is increasingly appreciated by a new generation of people on the land (Jones, 1994). For example, the manager of a property in northwest Tasmania has said that Devils help maintain “farm hygiene” (STDP website: “North-west Devils continue to thrive,” 21 June 2010).

Devils have always occurred, sometimes commonly, around the edges of cities and towns (Guiler, 1970a, 1982, 1983a, 1992; Jones, 1994, 1995, 2008; STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011), although their nocturnal habits mean they are often not seen. In rural and semi-urban areas, Devils occasionally take up residence under houses where they can be noisy (N. Mooney, in Wood, 2003b; MX (Australia), 15 July 2003; Owen and Pemberton, 2005; ABC Northern Tasmania, 16 December 2013; STDP Annual Report, 2013/14; ABC News online, 13 November 2015).

Devils with little familiarity with people are usually fearful and hence docile and easy to handle. But with increasing familiarity, some become bolder and more aggressive and difficult to handle (Albion, 2006; Edwards, 2006; Sinn et al., 2010; see also A. Anderson, in Owen and Pemberton, 2005). This may explain why some captive Devils are aggressive with their keepers.

Sometimes even naïve wild Devils, both young and old, can be immediately aggressive toward humans. But usually a Devil acts this way because it has been trapped or chased. When two wildlife volunteers released a newly weaned Devil from a trap, it chased the pair so aggressively that they had to retreat to their car (Owen and Pemberton, 2005). And when two wildlife officers wondered whether a Devil was fast enough to chase down prey, they released one and chased it, only to have the animal suddenly stop, turn and hiss at the pursuer (Owen and Pemberton, 2005).

Devils occasionally nip their handlers, but there are no reports of serious injuries. Whether this is due to good luck or under-reporting or indicative of the degree of the Devils’ “intent” is unclear.

When a captive Devil bit a five-year old boy on the finger, the finger became infected with the bacterium Paseturella multocida, a common infectious agent in animal bites. Antibiotics quickly controlled the infection (Georghiou et al., 1992).

Interestingly, there is an unsupported assertion but in a paper co-authored by one author with close ties to the zoo community, that “… keepers have reported that devil bites cause severe infections that are difficult to treat” (Cheng et al., 2015).

Humans seem to have taken a new interest in Devils since their facial tumour disease started diminishing their numbers. Visitors now go to see Devils in zoos and in the wild, where they can be attracted to staked-out road kills (Mooney, 1992).

6/05/2016 8:41 AM 163

Devil Facial Tumour Disease

In 1994, a Devil researcher wrote that “despite being quite rare earlier this century, numbers of Devils have been steadily increasing for the last 30 years and their status is considered to be secure” (Jones, 1994). And even in the fateful year of 1996 she could write that of all Tasmania’s living carnivores “only the Devils seems secure at the moment” (Jones, 1996; see also Jones and Rose, 1996). But in October 2006, this same researcher had come to believe that “almost every devil in Tasmania is going to die within a year of reaching adulthood of this disease. And that’s – stunning (ABC Catalyst, 19 October 2006).”

Fig. 14. Two Tasmanian Devils with advanced Devil Facial Tumour Disease (DFTD).

6/05/2016 8:41 AM 164

First Evidence of DFTD

The first well-documented evidence of the existence of Devil Facial Tumour Disease (DFTD) was in 1996 when Christo Baars, a Dutch wildlife photographer visiting Mt William National Park in far northeast Tasmania, took pictures of several Devils with large lumpy, red growths on their mouths feeding on a set carcass (Hawkins et al., 2006). The first specimen of a tumour available for clinical examination was from an animal captured in 1997 near Bridport, about 75 km east of Mt William National Park (Loh et al., 2006a, as “Bredport,” but later corrected (?) to “Waterhouse” (Loh et al., 2007; T. Knox, pers. comm. in McCallum et al., 2009; STDP, 2012a) and later expanded to “Waterhouse Point” (STDP website: “Status,” 18 September 2015).

It is unlikely that the discovery of the disease was contemporaneous with its origin. Instead, it probably started sometime before 1996. But just how much earlier is difficult to say. A wildlife researcher thinks he may have seen an animal with a facial tumour in 1984, at Mt William, where the disease was first documented (D. Pemberton, in Owen and Pemberton, 2005).

Even more intriguingly, on a study tour of European museums in 2004, a zoology curator at the Tasmanian Museum and Art Gallery believes she saw deformities in three skulls that were similar to the deformities caused by DFTD (K. Medlock, on ABC, Stateline Tasmania, 9 October 2004; K. Medlock, in Owen and Pemberton, 2005). If the skulls were from wild specimens, it could have important implications for the origin of DFTD. An email request (28 June 2012) enquiring about the curator’s current thoughts on what she saw went unanswered. None of the Devil skulls in Australian museums show any evidence of disease. Furthermore, none of the six biologists who trapped more than 2000 Devils between 1964 and 1995 can recall seeing anything like the disease in their time (STDP website: “The Disease,” accessed 17 June 2010).

The existence of the disease was first made public in a news story in March 2003 (Mooney, 2003; Wood, 2003a), more than six years after it was first noticed. The first government response was an information brochure published on 23 July 2003 (Clark, 2003; ABC, World Today, 24 July 2003; Hawkins et al., 2006).

Tumour Characteristics

The tumours develop quickly. One devil with a single lesion had four more after just eight days (C. Hamsen, in Darby 2011). The tumours first appear as small firm pink nodules (Hawkins et al., 2006; Holz, 2012: fig. 50-2) and grow steadily.

Rarely do the tumours stop growing or regress (Hawkins et al., 2006; Lachish et al., 2007). There are only four reports, none formally published, of tumours apparently regressing. One involved a wild Devil in which the tumour was recorded as having become smaller in the three months between captures but then having resumed growing, eventually killing the animal (STTDP website: “A facial tumour that appeared to shrink in an “interesting” devil,” 17 August 2007; Newsletter, August 2007; DPIW, 2007a: pictures p. 11).

6/05/2016 8:41 AM 165

Another, involved an unspecified number of animals injected with live cells either from an active tumour or from a tissue culture (the report did not say which) subsequent to which there was “… some degree of resorption and redevelopment … (Pyecroft, 2007b).”

The third became known at a public lecture in mid-July 2012 (Jones, 2012). The lecturer, a Devil field researcher, noted that the tumours in five Devils at West Pencil Pine had regressed over a period of three to six months (see “West Pencil Pine,” below). At least one of these regressed tumours was known as early as early as 2009. These five Devils were almost certainly the ones mentioned subsequently in passing as “…five animals who are still seemingly healthy after tumour regression (Belov and Hogg, 2012).” They were also probably the basis for reports of some tumours in one population regressing (Beniuk, 2012; see also M. Jones in UTas media release, 12 March 2013 and K. Belov, ABC Radio, 4 April 2013).”

The fourth was from a field worker who said, commenting on her long-term study sites, “We’ve seen seven, possibly eight animals whose tumors have regressed M. Jones, in Marris, 2015).”

The fact that at least some tumours regress is important, because it raises the possibility that these Devils are able to mount an immune response, even after the tumour has begun to grow.

Two infected Devils, both from Fentonbury, were early reported as showing “… very little apparent development in their tumours” after eight months (DPIWE, 2005d). There was an unresolved question, however, as to whether the biopsies taken to prove DFTD may have been contributory to the slow growth. A third Devil, from Bronte, was said to be a “similar case (DPIWE, 2005d).”

As the tumours grow, they become flaky and friable such that, in time, pieces come off easily. They ulcerate, become infected with bacteria and suppurate. The tumours also “stink” (M. Jones, in Parer and Parer-Cook, 2004). In their advanced stages, the tumours can push teeth out; erode bone such that the jaw might break off, and grow over and into the eye socket (Fig. 14). The primary tumours can impair and ultimately curtail the ability to eat, drink, breath, see and smell. A single tumour can reach a diameter of 12-15 cm before an animal dies (Hawkins et al., 2006; M. Jones, ABC Catalyst, 19 October 2006; DPIW, 2007b; M. Jones, on ABC Science Show, 21 April 2007; see also Pyecroft et al., 2007).

An individual Devil can have between one and four primary tumours (Hawkins et al., 2006; M. Jones, on ABC Science Show, 21 April 2007). Indeed, multiple tumours are more common than single tumours. In one study, 68 percent of 85 infected animals had more than one primary tumour (Loh et al., 2006).

The primary tumours often metastasize, presumably through the blood and lymphatic systems, and have been observed in 65 percent (N = 91) of diseased Devils. Metastases have been recorded from the adrenal glands, heart, kidney, lungs, lymph nodes, mammary glands, ovary, pituitary, rib surface and spleen, where they can eventually impair the organ’s function (Loh et al., 2006a-b; Bender, 2010: fig. 23.1;

6/05/2016 8:41 AM 166

Keeley et al., 2012b). Even small primary tumours can metastasise, although just how small they were was not made clear (Kreiss, 2009).

A single Devil can carry many tumours. One popular account put the total number of tumours in a single animal as high as 50 (Wood, 2003d; see also M. Jones, in Wood, 2003a). Most of these would have been secondaries.

It is also clear from the unpublished inoculation experiments on one Devil, “Cedric” (below), that metastases can be released at a very early stage in the growth of the primary tumour. By the time this Devil’s “pea-sized tumours” (UTas media release, 26 February 2009) were recognised and surgically removed, some tumour cells had already metastasised. This experiment also showed that a metastasis could have no noticeable effect on an animal’s health for up to 8 months (period between Cedric’s December 2009 surgery to remove his two primary tumours and his euthanasia at the end of August 2010 due to metastasis in his lung; below).

When tumours become friable, pieces can fall off or be dislodged easily. Tumour cells are said to be unable to survive for more than a few minutes outside the body (DPIPWE unpublished data, as cited in Hamede et al., 2013), but they may survive longer in larger masses.

Devil facial tumours often have to be identified in the field. There is a difference of opinion, however, about how reliably this can be done, especially in the early to mid- stages of the disease. One observer noted that the tumours were “distinctive and readily identifiable in the field (Lachish et al., 2010).” But another observer said “… Tasmanian devils suffer from other types of tumours that look remarkably like DFTD (Fox et al., 2010b; see also Hawkins et al., 2015).” Some empirical evidence bears on the issue. In one instance, 2 of 145 animals (1.4 percent) thought to have DFTD, proved not to have the disease when autopsied (Lachish et al., 2010). But in another instance, five of 53 adult males (9.4 percent) assessed in the field as having facial tumours actually had only “scar tissue” when assessed in the lab (Keeley et al., 2012b). Parenthetically, it would be interesting to know the cause of the other types of tumours that “look remarkably like DFTD.”

Presumably, field researchers record the size of the tumours, especially the primaries, either with actual measurements or with a relative scale. In their variability, growth rates by themselves could be indicative of resistance, and when compared to strain type or genetics they could narrow the focus on the cause of that resistance. If such an analysis has been done – and plenty of time has now passed in which it could have been done, it is curious the results have not been published, as even a nil result would be important. If such basic research hasn’t been done, one would have to wonder why.

Tumour Cytogenetics

Researchers began working on the cytogenetics of the Devil’s facial tumour in January 2004 (Pearse et al., 2007; Swift, 2007; Pearse et al., 2012). They have used increasingly sophisticated techniques, G-banding (Pearse and Swift, 2006; also Kreiss et al., 2011a; Pearse et al., 2012), DAPI banding (Murchison et al. 2011; Deakin et al., 2012), chromosome painting (Murchison et al. 2011; Deakin et al., 2012) and

6/05/2016 8:41 AM 167

gene mapping (Deakin et al., 2012), leading to increasingly greater insight into the tumour’s cytogenetics (Deakin et al., 2012).

Karyotype

A signature feature of facial tumour cells is a distorted chromosome complement. Devils normally have 14 chromosomes including an XX in females and an XY in males (above). The original facial tumour cells examined, however, had only 13 chromosomes, only some of which could be related to normal chromosomes. One member of chromosome pair 1 appeared to lack its long arm; chromosome pair 2, one member of chromosome pair 6 and the two sex chromosomes appeared to be absent, and four “new” unpaired “marker” chromosomes were of uncertain provenance but presumably reconstructed from the “missing” parts of the other chromosomes (DPIWE, 2005a: fig. 1; Pearse and Swift, 2006: fig.1; Swift, 2007).

The fact that there were no known “intermediate” conditions between the Devil’s normal karyotype and the DFTD karyotype raised the possibility the rearrangement was achieved not incrementally but in one big change (Deakin et al., 2012; Deakin and Belov, 2012).

Only part of the karyotype has been extensively re-arranged in DFTD. But intriguingly, the same chromosomes rearranged in DFTD also appear to be the ones largely rearranged among the two other well-studied marsupials, the wallaby and opossum (an American species) (Deakin et al., 2012). The significance of this similarity between the tumour aetiology and evolutionary diversification is unclear.

Subsequently, other researchers modified slightly the original interpretation of the tumour’s chromosomes in comparison to the traditional designations in the highly conserved dasyurid karyotype (above). One group thought it was Pearse and Swift’s chromosome 5, not 6, that was missing its homolog and identified the missing member in a marker chromosome (Kreiss et al., 2011a). Later, another group swapped the identity of the Pearse and Swift’s chromosomes 1 and 2. These researchers also found that it was chromosome 5, not 6, that was missing a homolog which they found in marker chromosome 2 (M2), apparently unaware of the first group’s earlier similar insights into these same two chromosomes (Deakin et al., 2012; Table 15).

Another large karyotype change figured but not explicitly identified is the apparent presence of only a truncated homologs of chromosome 5 (Pye et al., 2016: fig. 3), not its complete absence (through relocation to other parts of the karyotype). As one of the authors of the paper first describing the DFT karyotype was an author of the work cited, it is unlikely to be an error.

Table 15. Authors’ designations of Devil autosomal chromosome numbers that differ (red) from the numbers based on the standard dasyurid karyotype (bold).

Reference 11 22 3 4 5 6 Pearse and Swift, 2006 2 1 3 4 5 6 Pearse et al., 2007 ? ? 3 4 5 6 Pyecroft et al., 2007 2 1 3 4 5 6 Bender, 2010 2 1 3 4 5 6

6/05/2016 8:41 AM 168

Kreiss et al., 2011a 2 1 3 4 5 6 Holz, 2012 2 1 3 4 5 6 Deakin et al., 2012 as corrected in 1 2 3 4 5 6 PLOS Genetics Staff, 2014 Murchison et al., 20123 2 1 3 4 5 6 Pearse et al., 2012 2 1 3 4 5 6 Pye et al., 2016b 2 1 3 4 5 6 Notes: 1. Large submetacentric chromosome. 2. Large metacentric chromosome; 3. Authors recognised designation of chromosomes 1 and 2 was reversed but maintained the original usage.

When researchers found the pericentric inversion in chromosome 5 as a variation of the apparently normal Devil karyotype (above), they thought it might bear some relationship to the disease (AusVet, 2005; possibly A.-M. Pearse, in Denholm, 2007; Pearse et al., 2007; Pyecroft et al., 2007; AMMRF News 2010; STDP website: “Trowunna devils join Insurance Population,” 19 August 2011). This idea is apparently still being considered, because “investigation of the relevance of the AC5 chromosome” was listed as a research project for the 2012/13 year (STDP Annual Report, 2011/12; see also STDP Annual Report, 2012/13).

Strains

The original description of the tumour’s karyotype in January 2006 (Pearse and Swift, 2006) made no mention of any variants (Pearse and Swift, 2006). But in the following year, media and research reports began referring regularly to different chromosomal “strains” or “sub-clones” (hereafter, “strains”) of the tumour.

The number of different strains was reported as four in early 2007 (McGlashan et al., 2007; Pyecroft et al., 2007), nine in mid-2008 (ABC News, 25 July 2008; Brand Tasmania Newsletter, August 2008; DPIW Annual Report, 2007/08; Martain, 2008; Mather, 2008a-b; Newsletter, September, December 2008, March, June 2009), “up to 12” in early July 2008 (PHVA, 2008), 13 in mid-2009 (Newsletters: September, December 2009; March, June 2010; DPIPWE, 2010d), six in September 2009 (Kreiss, 2009), nine in early 2010 (Woods and Sharman, 2010), “at least 8” in the first half of 2011 (Hamede et al., 2012a), “multiple” in mid-2011 (STDP Annual Report, 2010/11), “at least 4 major … strains” in November 2011 (Pearse et al., 2012),”and “five” in mid-2013 (STDP Annual Report, 2012/13).

It is possible some of the variation in the number of strains arises from whether tetraploids are recognised as strains in their own right (Jones, 2012) or as forms of a diploid strain (Pearse et al. 2012).

Parenthetically, considering strains were widely recognised as early as 2007, the comment in January 2013 that “Up until a year ago we thought the tumour was completely stable (K. Belov, in Zimmer, 2013)” is difficult to understand.

Despite the many early references to “strains,” the concept was only formalised in 2012 as, “a karyotype of consistent chromosomal constituents that has been identified in multiple devils from the same geographic regions and is therefore transmissible (Pearse et al., 2012).”

6/05/2016 8:41 AM 169

An understanding of the strains is complicated by the fact that the only two articles describing and analysing the strains to date (Pearse et al., 2012; Deakin et al., 2012) each contain ambiguities. These ambiguities make it difficult to understand completely each paper individually and even more so in relation to each other. The two articles will be referred to as the first and second, not in order of publication, which was March and February 2012, respectively, but with regard to the historical involvement with strains of the two senior authors.

The first article (Pearse et al., 2012) used G-banding on 269 Devil tumours to identify the strain chromosomes and their parts. It covers all (?) four strains, numbered 1-4. Strain 3 has two variants, 3a and 3b. The diagnostic features of these strains are shown in Table 16. The article also tabulates (table 1) a “Strain CE,” from one locality (Dunalley) but does not discuss it. In addition, there is a caption reference (Pearse et al., 2012: fig. 4) to a “Strain N,” which is interpreted as a tetraploid derivative of tetraploid Strain 1, but use of the word strain for this tetraploid raises the question of why the tetraploids of the four strains themselves are not give “strain” designation.

Table 16. Chromosome rearrangements in the four Strain types identified in Pearse et al., 2012 with G-banding. Re-arrangements likely to have arisen evolutionarily in a strain type are in red.

Strain 2N Sex 1 2 3 4 5 6 Mar 1 Mar 1 13 -x-x/-x-y del 1q -2-2 t(5;6) + + 2 14 -x-x/-x-y del 1q -2-2 t(5;6) + + 3a 14 -x-x/-x-y del 1q -2-2 del 3p t(5;6) + + 3b 13 -x-x/-x-y del 1q -2-2 del 3p, in 3q1 t(5;6) + + 4 14 -x-x/-x-y del 1q -2-2 t(5;6); del 5p der 6 + + 1. In the chromosome formula, the change in 3q is given as “del 3q,” but in two places in the text, it is described as an “inversion.”

In addition, the four strain types are said to differ in the length of the cell’s dendritic processes, the length of the processes increasing in the order: 1>2>4>3 (Pearse et al., 2012: fig.). It is not clear, however, just how discriminatory these differences are by themselves, especially between Strains 3 and 4 where the difference is “slight.”

The first article described no further variation within the strain types other than a variable number of Dmins in Strain 4 and tetraploidy in Strains 1 and 2 (Pearse et al., 2012).

The second article (Deakin et al., 2012 as extensively corrected in PLOS Genetics Staff, 2014 following discrepancies first noted in the 26 July 2012 edition of this document) used G-banding, chromosome painting and genetic mapping on eight Devil tumours to identify and characterise the strain chromosomes and their parts. It covers three Strains, numbered 1-3. This article’s concept of Strains 1-3 appears to be the same as in the first article. But it goes further and identifies sub-strains 2a-d and sub-strains 3a-b and it is not clear if two sub-strains recognised in common between the two articles, 3a-b, are the same. However, this article gives no exclusive diagnostic features for any of its sub-strains. Furthermore, it makes passing mention

6/05/2016 8:41 AM 170

of “2c sub-strains” and “substrains of 2c and 3b”as if there are sub-strains of sub- strains. Parenthetically, in what is almost certainly a typo, the correction speaks of a sub-strain 3c, whereas the introduction says substrain 3c has been renamed 3b (PLOS Genetics Staff, 2014).

The second article makes no mention of Strain 4, and it is not clear if it was not recognized or not studied. In either case, without chromosome painting and gene mapping data from it, we can neither infer the evolutionary relationships among the four strains nor, more importantly, infer the features common to all strains associated with successful transmission.

The second article described additional gross variation in strains 1-3, and those identified largely by chromosome painting are summarised in Table 17. For the extensive variation identified by gene mapping see the original article and as corrected.

Table 17. Tumour strains recognised by Deakin et al., 2012 (as corrected in PLOS Genetics Staff, 2014) and variation in chromosome morphology as revealed largely by chromosome painting.

Strain 3 4 5 M4 M5 1 2A 2B del 3p: ? %1 t(4p;M4q): 100% 2C2 del 3p: ? %3 t(4p;M4q): 0 % + 42 %;- 58 % + 54 %; - 36 % 2D 3A 3B4 del 3p: ?%3 t(4p;M4q): 12% rcp (4/5): 87% + 36 %; - 64 % + 100 % rcp(4/5): 87 %

Notes: 1: Occurs in only one chromosome of the pair. 2. Two sub-strains of 2C are mentioned once in PLOS Genetics Staff, 2014 but without description or further discussion. 3. Occurs in both chromosomes of the pair. 4. Named as chromosome 3C in text of PLOS Genetics Staff, 2014, but introduction says 3C has been renamed 3B.

In terms of their distribution, Strains 1 and 2 are widespread throughout the island; Strain 3 is restricted to the Forestier Peninsula and Strain 4 is known from only two localities in the southeast, Buckland and the Freycinet Peninsula (Pearse et al., 2012; Deakin et al., 2012 and PLOS Genetics Staff, 2014). Strain 2 is said to appear “… to be emerging as the dominant DFT strain over most of Tasmania (excluding Forestier Peninsula (Pearse et al., 2012),” although the case for this assertion has not been made explicit.

Several general observations can be made about the strains. First, they are characterised by deletions, duplications and rearrangements both between and within chromosomes (Pearse et al., 2012; Deakin et al., 2012 and PLOS Genetics Staff, 2014). These changes notwithstanding, the strains are said to “show remarkable homogeneity in the chromosome rearrangements (Pearse et al., 2012) and to be “characterized by minor cytogenetic rearrangements (Deakin et al., 2012).” Indeed, this is why strains are recognisable in the first place.

6/05/2016 8:41 AM 171

Having said this, however, it is noteworthy that in analysing the tumours of only eight Devils with chromosome painting and gene mapping techniques, geneticists identified seven strains and sub-strains and many small chromosome and gene rearrangements (Deakin et al., 2012 and PLOS Genetics Staff, 2014). One begins to wonder just how many more variants might have been detected if even more tumours had been analysed with this detail.

Second, all four strains described to date are said to form tetraploids (Pearse et al., 2012; also Ujvari et al., 2014b). Oddly, however, in the two major review papers touching on tetraploids, none were tabulated for Strains 3 and 4 in one (Pearse et al., 2012: table 1) nor found for analysis for Strain 4 in the other (Ujvari et al., 2014b: fig. 2). As of April 2013 and overall, the proportion of tetraploids in the strains was c. 34, 14, 16 and 0 percent, respectively, Strain 1 having significantly higher proportion than the others (Ujvari et al., 2014b: fig. 2; note that in the supplementary data file for this paper, the coding for tetraploid and non-tetraploid samples appears to be reversed).

At one locality, the Forestier Peninsula, which consists almost entirely of Stain 3 and where diseased animals were culled between 2006 and 2010, logistic regression analysis indicated that the proportion of tetraploid tumours increased over time (Ujvari et al., 2014b: fig. 3; see also Fig. 15). The relationship was interpreted as evidence of the tumour’s ability to respond rapidly to novel selection regimes (Ujvari et al., 2014a-b). This may be the case, but the culling protocol was both more directed and intense than might be expected in nature.

Parenthetically, it is worth noting that this increase in tetraploidy was interpreted as showing, “… that this malignant cell line has once conquered genomic decay by increasing its chromosome numbers” and that “DFTD can rapidly adapt to anthropogenic selection via altering chromosome numbers … (Ujvari et al., 2016).” Aside from the curious anthropomorphism (“has …conquered”), the paper referred to (Ujvari et al., 2014a-b) contains no evidence whatsoever that “genomic decay” actually preceded the increase in polyploidy. There was also a claim that “younger strains” have “reverted to diploid karyotypes (Ujvari et al., 2016),” but again there is no published evidence for such a reversion.

Oddly, no tetraploids were reported for the Forestier Peninsula population in an earlier 2012 review (Pearse et al., 2012), despite the likelihood that the two studies (Ujvari et al., 2014a-b and Pearse et al., 2012) surveyed most of the same samples (inferred from sample sizes, dates of collection and special circumstances of the collection, a culling program).

6/05/2016 8:41 AM 172

Percent of Devil Facial Tumours Tetraploid in Forestier Population in Years 2006-2010

40 50 35 30 25 18 20 8 Percent 15 10 40 33 5 0 2006 2007 2008 2009 2010 Ye ar

Fig. 15. Percent of Devil Facial Tumours that were tetraploid in the years 2006-2010 in the Forestier Peninsula population. Sample sizes shown above bars. All tumours were Strain 3 bar one (Strain 1).

With further regard to the tetraploids, it is interesting that in Strain 1, the diploid form apparently shows telomere “erosion,” but the tetraploid form apparently has the capacity for “telomere replenishment (Pearse unpublished data, in Pearse et al., 2012).”

Third, the four strains appear to differ in growth rate in tissue culture, increasing in the order 2, 1, 3 and 4 (Pearse et al., 2012). Furthermore, the tetraploid of Strain 2 (from West Pencil Pine) is “extremely slow growing” (Pearse et al. 2012) and slower than the diploid form (A. Kreiss pers. comm. in Pearse et al., 2012; see also Hamede et al., 2015, where the phenomenon is reported for an unnamed locality that almost certainly WPP).

Fourth, the frequencies of the strains at different localities are not static (Pearse et al. 2012; Ujvari et al., 2014a-b), suggesting a dynamic that is not yet understood.

Fifth, there is no significant difference among the four strains in telomere copy number, a proxy for telomere length (Ujvari et al., 2012a; although see point two, above).

Sixth, metastases can show additional changes not seen in the primary tumour. These changes “include further translocations and tetraploidy, with subsequent chromosomal rearrangements and chromosome loss (Pearse et al., 2012).” This has led to the inference that transmissibility is due to some specific cytogenetic features that have yet to be identified (Pearse et al. 2012; see above). Considering the chromosomal variation between and within primary tumours and metastases, it would be useful for

6/05/2016 8:41 AM 173

the main protagonists to get together and characterise the features unique to transmissible and non-transmissible variants.

Lastly, preliminary comparisons using the same six of the 11 polymorphic microsatellite loci used to first assess genetic variation in healthy Devils (Jones et al., 2003b; Jones et al., 2004a) found no significant variation among the four tumour strains Pearse et al., 2012).

The phylogenetic relationships of the strains are thought to be: 1 → 2 or 2 →1 and 2 → 3 and 4 (independently) (Pearse et al., 2012: fig. 2); however, the 1 → 2 step is more parsimonious than the 2 → 1 step and should take provisional preference. Strain 1 seems to have undergone the fewest cytogenetic changes, so if an “explanation” for the origin of DFT lies in its cytogenetics, the minimalist condition of Stain 1, is the place to look. Indeed, the extensive fragmentation of chromosome 1 in this strain (to five other chromosomes), as well as the other strains, narrows the focus even more closely (Deakin et al., 2012 and PLOS Genetics Staff, 2014).

It is not clear when the strains first evolved, because there are no tissue samples from the early years. Strains 1 and 3 were extant in 2004, the first year DFTD was studied cytogenetically. But if Strain 3 is derived from Strain 2, it too must have been extant in 2004. Strain 4 was first found in 2006 (Pearse et al., 2012: table 1). Strain 1 has been called the “oldest strain (Ujvari et al., 2014b),” which would be true under the most parsimonious interpretation of strain evolution (above).

With the relatively comprehensive overview of the strains now at hand (Pearse et al. 2012), it is possible to make inferences about the identities of some of the strains mentioned in earlier media stories and research papers. The earliest mention of a strain number was a media account in which “Strain 2” was said to be widespread and known in “the general vicinity of Trowunna [a zoo near Mole Creek] (Sun Herald, 24 October 2006).” Strain 2 is widespread and does occur at Mole Creek (Pearse et al., 2012: table 1), suggesting that the early strain designation stands.

In 2007, the first described karyotpye was identified as “Strain 1” (Pyecroft et al., 2007). This was subsequently corroborated by the describers of the original strain (Pearse et al., 2012).

A recent media report mentioned a tumour type that was tetraploid, slower growing and found in “the western area” of the Devil’s range (Mills, 2011, based on conversation with M. Jones). It now seems likely that this was the tetraploid form of Strain 2, which has been described as “extremely slow growing [in culture] compared with the diploid forms (Pearse et al., 2012).”

A recent study of the genomics of two different Devil tumour cell lines also included a characterisation of the tumours’ cytogenetics. One tumour cell line, designated 53T, from Narawntapu National Park in the northcentral part of the state, had 32 chromosomes (Murchison et al., 2012: fig. 2a). This was a metastasis tumour and may have been one of the “progressions” of “Strain N” of Pearse et al., 2012, itself a tetraploid derivative of Strain 1 (Pearse et al., 2012). The other tumour cell line, 87T, from the Forestier Peninsula in the southeast had 13 chromosomes, heterozygous translocations on chromosomes 1, 2 and 3 and trisomy of the short arm of

6/05/2016 8:41 AM 174

chromosome 5 (Murchison et al., 2012: fig. 2a, c-e). This was a primary tumour and likely to have been a derivative of Strain 3, the only strain found to date on the Forestier Peninsula. It is unclear, however, whether the karyotype of this tumour represents a unique variant or is representative of another strain.

Parenthetically, there appear to be three errors in the literature regarding strain type and locality. Strain 3 is listed from “Brighton” (Pearse et al., 2012: table 1) and “Coles Bay” (Bender et al., 2012: table s2), but Strain 3 is known to occur only on the Forestier Peninsula (Pearse et al., 2012). Strain 4 is listed from “West Pencil Pine” (Ujvari et al., 2013), but Strain 4 is restricted to two localities in the southeastern part of the state (Pearse et al., 2012: table 1).

A strain can apparently mutate into new cytogenetic forms once it becomes established in an animal, although these new forms are not necessarily passed on to other animals, that is, they do not always become strains themselves (above). This possibility was first hinted at as early as 2005 when it was observed that “in a small number of affected devils which have survived for a period with the disease … further chromosomal transformations have occurred (DPIWE, 2005a).” The first published observations of this phenomenon came from what was almost certainly a Strain 1 tumour that also contained near triploid and near tetraploid karyotypes with “additional structural and numerical abnormalities (Kreiss et al., 2011a).”

As mentioned, recent work has indicated that the strains appear to vary greatly in their growth rates (Pearse et al., 2012). Interestingly, variable growth rates had been mentioned as early as 2007, albeit in a rather colourful manner. The STTDP’s science manager said, “Every time the cell divides, it becomes really clunky, so the cancer is growing more slowly (C. Boland in Darby, 2007).”

As yet, there are no formal studies of the virulence among the strains. A media report mentioned that one particular animal from a new location for the disease “… had a strain of the disease that was less deadly than in other areas (Mounster, 2011).” This may have been a reference to the tetraploid form of Strain 2 (above). And in early March 2010, a researcher commented, slightly apocalyptically, that “some new strains of the tumour … look very nasty (J.A.M. Graves, ANUchannel video, 29 March 2010).”

Two apparent trends in the evolution of the strains bear on the possible course of DFTD. First, it appears as if the two slowest growing strains (1 and 2) are the most widespread, whereas the two fastest growing strains are geographically restricted (Pearse et al., 2012). This suggests that, relative times of origin aside, there is no advantage, and perhaps even a disadvantage, to growing too quickly and killing the host too soon. Second, in that tetraploidy is a derived condition from diploidy and tetraploids grow more slowly than diploids (above), the evolution and persistence of these even slower growing forms of each strain suggest, again, no disadvantage and possibly even an advantage in prolonging the life of the host, the advantage being a greater chance of being spread (Pearse et al., 2012). The advantages of tetraploidy would come even more to the fore, if tetraploidy were found to have arisen more than once in each strain. In any event, both trends augur well for the eventual evolution of a more benign tumour and a long-term sustainable relationship between tumour and host.

6/05/2016 8:41 AM 175

Most recently, an abstract published late in 2014 contained two assertions, one startling and the other interesting and potentially important (Deakin et al., 2014). In the first instance, it was stated, “until recently, structural mutations in DFTs were seemingly restricted to particular genomic regions, predominantly regions consisting of chromosome 4, 5 and X material … (Deakin et al., 2014). This is startling, because chromosome 4, until this comment, has never been mentioned as being involved in any of the rearrangements seen in DFT; perhaps the number “4” is a lapse for another chromosome number. In the second instance, it was stated, “a more recent karyotypic strain of the DFT has been detected in a population where DFT disease prevalence is reduced and there has been a limited effect on population structure compared to other areas affected by DFT disease. This DFT strain has an additional large marker chromosome [emphasis added].” This is interesting because it is the clearest association yet of a genetic change and a change in the progress of the disease. The published paper is to be much anticipated.

In vitro strain changes

In July 2004, the Tasmanian government established a laboratory for DFT tissue culture (DPIWE Annual Report, 2004/05, 2005a), and cultured cell lines are now used widely in DFTD research (Pearse and Swift, 2006; Brown et al., 2011; Bender et al., 2012; Pearse et al., 2012; Siddle et al., 2013). Some DFT cell cultures started in 2005 were still providing cells for research as late as 2012 (Siddle et al., 2013).

Researchers report that cell lines remain “karyotypically stable in cell culture” over relatively short periods (< 48 h; Deakin et al., 2012),” but “frequently” show “increased aneuploidy and aneusomy” (Pearse et al., 2012) over the long term. In one case, a cell line developed a trisomy 6 (Murchison et al., 2012). It was not clear, however, when this anomaly appeared. In another case, after three months, a culture contained not only a standard Strain 1 diploid cell line but also near triploid and near tetraploid cell lines (Kreiss et al., 2011a). It was not clear, however, if these latter two lines were present among the foundation cells or developed in culture. Interestingly, the near tetraploid line contained a marker 5 chromosome, which does not occur in Strain 1 but does occur in Strains 2-4 (Pearse et al., 2012).

Tumour Genetics

The origin and spread of the Devil’s tumour has coincided with an exponential increase in genetics technology and knowledge. Devil genetic research has ridden this wave of innovation to much interest if not, as yet, to much effect.

The most powerful expression of the capacity of modern genetics with regard to the Devil is the recent sequencing of the genomes of the tumours from three Devils by two research groups (see “Genome sequence”). One group sequenced the genome of a cancer from a female from the southeast part of the state (Freycinet Peninsula) (Miller et al., 2011), and a second group sequenced the genomes of two cancer cell lines from animals of unstated sex, one from northcentral part (Narawntapu National Park) and the other from the southeast part (Freycinet Peninsula) (Murchison et al., 2012). The two cancers from the southeast part of the state were from different individuals (Deakin and Belov, 2012).

6/05/2016 8:41 AM 176

These three genomes are apparently just the first of many, as there is a project underway to sequence “hundreds more genomes” (Zimmer, 2013). This many sequences would provide insight into the tumour’s genetic evolution of the tumour and allow robust inferences as to the possible founder mutations.

There are two main reasons for an interest in the tumour’s genetics. First, it may provide clues about what caused the cancer, how it has evolved and how it is maintained. And second, it may help understand how the tumour evades the immune response.

General features

Like most tumours, the Devil’s tumour is a mixture of normal cells and cancer cells, and in the one tumour examined in this regard, about 30 percent of the nuclear DNA and 15 percent of the mitochondrial DNA was estimated to be derived from the host and the remainder from the tumour (Miller et al., 2011).

The tumour’s nuclear genome has 118 575 unique single nucleotide polymorphisms (SNPs) in comparison two healthy genomes that were homozygous at those positions. The tumour’s mitochondrial genome also has five unique SNPs in comparison to 13 healthy animals (Miller et al., 2011). Considering a single nucleotide mutation could theoretically lead to tumour development, it is likely to be some time before an inference as to the genetic origin of the Devil’s tumour can be made.

Genetic evolution

It is evident that the tumour has evolved not only in its cytogenetics (the strains, above) but also in its genetics. The two geographically distant cancers sequenced, for example, had accumulated 15 160 and 17 790 unique mutations since the time of their common ancestor. It is not known when this ancestor lived, but given the large geographic separation between the two host Devils, it could have been close to the time of the original cancer.

In comparison to the large number of mutations that evolve in most cancers before their host dies, the number of mutations that have accumulated over the years in the Devil’s cancer is low. This suggests that either the mutation rate is low (Murchison et al., 2012) or there is stabilising selection in the transmissible line of the cancer.

The tumour’s genome has evolved into at least 14 mitochondrial haplotypes, including that of the original tumour, since the mid-1990s (Murchison et al., 2012: fig. 3c). A phylogenetic tree for the 14 tumour haplotypes indicates that ten arose from the original tumour haplotype and the remaining three arose independently from one of the derivatives (Murchison et al., 2012: figs 3c; see also fig. 5). In other words, most of the tumour’s haplotype diversity as known to date has arisen from the original haplotype and the remainder from one of the original haplotype’s derivatives. This suggests that while the original haplotype has the ability to mutate viably, its derivative haplotypes may not.

6/05/2016 8:41 AM 177

There is evidence from the Forestier Peninsula in the southeast part of the state that between 2007 and 2010 one lineage of tumour variants declined while another increased. Whether this was due to chance or selection could not be determined (Murchison et al., 2012: fig. 5). At the time, however, researchers were actively culling diseased Devils from the peninsula in an attempt to suppress the disease (below). This raises the intriguing question of whether the cull had a role in the change in lineage variants.

Unfortunately, it is not yet clear what relationship, if any, there is between the tumour haplotypes, which are based on genes and the tumour strains (above), which are based on chromosomes.

Despite demonstrable evolution in some aspects of the tumour’s genetics, some regions of the genome seem unchanged. In 2007, geneticists published the first analysis of variation in the tumour’s genetics. This work assessed four microsatellite loci and several MHC (both class I and II) loci in 15 tumours from eight localities scattered throughout the then known distribution of the disease in central and eastern Tasmania. They found no variation (Siddle et al., 2007) and by implication, no evolution.

In 2012, geneticist assessed six polymorphic microsatellite loci among the four tumour strain types and found no variation (Pearse et al., 2012). They attributed the lack of variation to the relatively short lifetime of the tumour in comparison to standard mammalian mutation rates as well as the relatively small number of loci sampled.

Although some regions of the Devil’s genome show no variation, an assertion that “Devil Facial Tumour cells are genetically identical Ujvari et al., 2012)” is clearly not the case.

Devils can have two or more tumours each with a different genotype. When these tumours occur in more than one animal, it is likely they have been passed on in biting. But when they occur uniquely in one animal, it suggests they may have arisen within the founding cancer cell line of that animal (Murchison et al., 2012).

Possible causal genes

The search for genes that might have had a role in causing the Devil facial tumour relies heavily on knowledge of the genetic associations in cancer in mice and humans, the two most-studied organisms in this regard.

Mutations that would lead to a single amino acid substitution have been found in three Devil tumour genes, ANTXR1 in the one sequenced tumour (Miller et al., 2011) and FANCD2 and RET in the two sequenced tumour cell lines (Murchison et al., 2012). The mutations in the latter two genes were absent in two normal genomes (Murchison et al., 2012). ANTXR1 is thought to regulate a gene encoding a protein involved in the regulation of the cell cycle and important in the suppression of tumours. FANCD2 is a gene involved in chromosome stability and DNA repair. And RET encodes a protein important in transmitting signals across the cell membrane that in nerve cells are important for normal development.

6/05/2016 8:41 AM 178

A deletion has been inferred in two genes, MAST3 and a BTNL9-like gene (Murchison et al., 2012).

A balanced translocation has also been inferred for one gene, PDGFA (Murchison et al., 2012).

Variation in the copy number of genes are sometimes involved in cancer. Copy number variations are either duplications or deletions in the genome that larger than 1 kb in size. So far, the Devil’s tumour genome has been found to contain 12 autosomal genes present in only one copy and three genes present in three copies, the normal number being two (Deakin et al., 2012: table 2).

Mutations close to genes involved with cancer are also of interest as they may be involved in the regulation of those genes. Four such genes with hemizygous deletions have been identified in the Devil’s tumour: APC, MLH1, MYC and NF2 (Deakin et al., 2012; Deakin and Belov, 2012).

MicroRNAs (miRNA) are short sections of RNA involved in regulating the expression of messenger RNA (mRNA). MicroRNAs are encoded by DNA. Many miRNAs are associated with cancers in humans. Devil facial tumour researchers have found high levels of several miRNAs commonly up-regulated in tumours and a miRNA linked with tumour immune evasion. They also found that the tumour expresses low levels of two miRNAs that may suppress tumours (Murchison et al., 2012).

A recent review of genomics in relation to the Devil’s facial tumour disease tabulated two other genes, ARHGAP26 and CDK6, in which possible cancer-related mutations had occurred in the Devil’s tumour (Deakin and Belov, 2012: table 2). However, these two genes could not be found in the work cited (Murchison et al., 2012).

The principal researcher in the group that sequenced the two Devil tumours has said, "Sequencing the genome of this cancer has allowed us to catalogue the mutations that caused this cancer to arise and to persist in the Tasmanian devil population (E. Murchison, in Wellcome Trust Sanger Institute Media Release, 16 February 2012)." This is over-egging the cake, but it is a start.

Maintenance genes

Normal cell division leads to a gradual shortening of telomere (“see Chromosomes” above) length, which eventually leads to an inability to replicate further. Tumour cell lines, however, have overcome this limitation by maintaining telomere length slightly above the “stall” limit. One way they do this is by up-regulating the enzyme telomerase, which contributes to telomere lengthening. Telomere lengthening, in turn, is controlled by shelterin, a structural protein involved in protecting telomeres from lengthening too much. Hence, telomere length can be maintained by the complementary effects of telomerase activity and shelterin activity.

Immunogeneticists recently examined the expression of two genes, TERT and TINF2, involved in this equilibrium. TERT encodes a subunit of telomerase, and TINF2

6/05/2016 8:41 AM 179

encodes a component of shelterin. Devil’s tumour cells show a significant elevation in the expression of the two genes in comparison to normal spleen cells, approximately 15 and 38 fold, respectively. What the expression levels are in normal nerve tissue, the tissue line from which the tumour was derived, is unknown.

Furthermore, the expression of each gene appears to have increased over time (Ujvari et al., 2012: fig. 6; but see comment about possible degradation of samples below, “Tumour Cytology”). These gene responses could be among a suite of cell changes that confers immortality on the Devil facial tumour cell line.

Additionally, telomerase, as measured directly, is said to show a “positive temporal shift” in living tumour cell lines established in 2003, 2006, 2007, 2010 and 2011, a statement that seems to be supported by the accompanying relevant figure (Ujvari et al., 2012: fig. 7). However, a redetermination of the Spearman rank correlation for the data in the original figure returns a + 0.9 but non-significant (P = 0.070) correlation instead of the reported -0.9 and significant (P = 0.037) correlation (Ujvari et al., 2012). The difference in sign in the correlation value may be due to a typographical error while the smaller P value suggests the test was one sided instead of, more appropriately, two sided.

How DFTD evades the immune response

When it became apparent that the Devil could reject reciprocal skin grafts, focus shifted from the Devil’s “low” genetic diversity to the characteristics of the tumour in explaining why the tumour evoked no response from the Devil’s immune system.

The tumour was known to have the MHC genes that would allow it to be recognised as foreign (Siddle et al., 2007a). It was also known these genes were transcribed into mRNA (Siddle et al., 2007b). From this, the researchers inferred that the mRNA was translated into protein and these MHC proteins made it to the cell surface. Otherwise, why would they have written, “We demonstrate that altered MHC expression, a common cause of immune evasion by tumours …, is not responsible for a lack of immune response to DFTD, and we suggest that low MHC diversity in the devil has enabled natural transmission of tumour cells between individuals (Siddle et al., 2007b).”

In fact, MHC molecules do not appear on the surface of DFTD cells. The first indication in this revised understanding came in an abstract at an April 2012 conference (Siddle et al., 2012) followed by the complete paper published online in March 2013 (Siddle et al., 2013). This research showed that MHC gene proteins are absent from the surface of DFT cells derived from both cell culture lines and fresh tissue biopsy (Siddle et al., 2013). The structural genes for these proteins are present and are transcribed into messenger RNA, as previously demonstrated (Siddle et al., 2007b). The MHC proteins that would be translated from this mRNA, however, are not found on the cell surface. This appears to be due, at least in part, to the low expression of other genes required to process and transport the antigen proteins to the cell surface. Importantly, however, the expression of these genes can be enhanced in vitro by the application of a histone deacetylation (below) inhibitor, trichostatin A and an endogenous immune system modulator, interferon-gamma. This suggests the genes have been only down-regulated in DFTs and not structurally changed. It also suggests

6/05/2016 8:41 AM 180

the regulation is mediated by non-genetic factors, that is, the regulation is epigenetic instead of genetic (See “Epigenetics” below).

An intriguing residual issue, is why DFTD cells do not attract a response from natural killer cells, which have the capacity to attack and kill cells lacking surface MHC molecules (Siddle et al., 2013).

As part of the new understanding, the researchers also found that in some cases, cells near the edge of a tumour produce surface antigens, possibly in response to stimulation from host lymphocytes (Siddle et al., 2013). Both sets of observations have implications for the development of a vaccine (see “A Vaccine Against DFTD” below).

Curiously, the first and last authors of the April 2012 abstract and the March 2013 paper published a separate paper online in November 2012 describing the results of the work but not the actual research (Siddle and Kaufman, 2013, the date of the hardcopy publication). Why these two authors published this article, replete with references to “we,” but with no mention of the earlier April 2012 abstract or the impending March 2013 paper and with no acknowledgment of the eight other authors on the abstract or published paper is unclear.

It is also interesting that a researcher in the Australian contingent of co-authors of the March 2013 paper was subsequently reported as having said that “… his team had worked out how to “turn on” devil genes to trigger an immune response to the cancer (Glaetzer, 2014b, emphasis added; see also Hobart Mercury Editorial, 24 November 2014).” However, according to the actual paper, the Australian authors’ input was solely the “contribution of new reagents/analytic tools.” They contributed nothing to the research design, performance of the research, data analysis or writing of the paper (Siddle et al., 2013, “Authors contributions”).

For completeness in this section, mention should be made of a rearrangement in NOD1, a gene that encodes a protein receptor that recognises intracellular bacterial molecules and initiates an immune response (Murchison et al., 2012). Considering, however, the gene’s function involves the detection of prokaryotic cells, it seems unlikely it would have a role in recognising foreign eukaryotic cells.

Epigenetics

On 7 November 2012 online and in early 2013 in hard copy, geneticists published the first research on the epigenetics of the Devil’s tumour. They examined the mean number of full methylation and hemi-methylation sites and, the sum of the two, the mean number of combined methylation sites. Methylation of two in-tandem base pairs, cytosine and adenine, in DNA can regulate gene expression. They found that in tumour samples collected in years 2005, 2006, 2007 and 2010, there was no difference among years in mean number of full methylation sites, but the mean number of hemimethylation sites was significantly lower in 2010 than in the preceding years (Ujvari et al., 2013: fig. 2).

The geneticists did not discuss their data in this way, but if one looks at potential trends in their numbers, it could be said that over the interval from 2005 to 2010, the

6/05/2016 8:41 AM 181

mean number of full methylation sites trended downward and the mean number of hemimethylation sites trended upward. And if the mean number of methylation sites in normal nerve cells can serve as a guide to conditions in the first tumour cell, then the mean number of full methylation sites may have been initially much reduced but recently trended “back” toward normal nerve cell levels, whereas the mean number of hemimethylated sites may have been initially much higher and have since trended to now be even lower than normal nerve cell levels (actual values can be read or inferred from Ujvari et al., 2013: table 1 and fig. 2). Each of the two data sets have a Spearman rank order P = 0.08 (pers. obs.), the lowest value possible with the existing data, which suggests, at the very least, the value of adding more years to the analysis as they become available.

The researchers said that tumour cells “have similar levels of methylation to peripheral nerves (Ujvari et al., 2013),” but as shown above, this may not have been the case for either (relevant) component of combined methylation viewed on a trend basis.

The researchers also found that in comparison to normal nerve tissue, in the period 2007/08 expression levels of four genes associated with methylation were higher in one gene and lower in three others but in the period 2010/12, the expression levels had risen in all four genes (Ujvari et al., 2013: table 2). These data allow speculation that since the origin of the tumour, the one gene increased its expression, whereas the other three first decreased and then increased their expression.

The researchers emphasised that their research showed a “significant increase in hypomethylation in DFT samples over time (Ujvari et al., 2013).” But the decrease in hemimethylation is the main change and hence the key to understanding the epigenetic importance, if any, in the results. Furthermore, if the “trends” are confirmed with additional data, then the possible reciprocal relationship between the two types of methylation could prove to be important. It may also be important that some genes involved in methylation appear to have decreased their expression and then increased it. It has to be emphasised, however, that what role, if any, these epigenetic changes may have had on the tumour’s biology is unknown.

While the cytogenetics and genetics of the Devil’s tumour are extremely interesting, especially in understanding the long-term changes in cancer, it is important to bear in mind that this information is unlikely to play an important role in Devil conservation anytime soon. One has only to look at the disparity between what is known about the genetics of human cancer and the therapies available to realise the Devil’s future will probably be decided long before genetics plays a decisive therapeutic role.

Tumour Cytology

Initial examination of facial tumour cells gave no clue as to their origin, because the cells are dedifferentiated (anaplastic), can assume a variety of new forms (pleomorphic) and lack distinctive ultrastructure features (Loh et al., 2006b; Pyecroft et al., 2007). Light and transmission electron microscopy could only resolve the tumours as soft tissue neoplasms (Loh et al., 2006b).

6/05/2016 8:41 AM 182

The tumour cells are round to spindle shaped with a central round nucleus and a high nucleus to cytoplasm ratio (1:1.2) (Loh et al., 2006a; fig. 3; Pye et al., 2015c: fig. 5).

In 2004, veterinary pathologists determined that the tumour was likely to have had its origin in one cell type (Darby, 2005; DPIWE, 2005a; Loh et al., 2005; Loh et al., 2006a). Using immunohistochemical techniques, they soon discovered that the cell type was most likely of neural crest origin, possibly even of neuroendocrine origin (DPIWE, 2005a-b; Loh et al., 2006b, 2007; Pyecroft et al., 2007). There was also one prescient suggestion that it was a “nerve cell type thing (S. Pyecroft, on ABC Stateline Tasmania, 9 October 2004).”

In 2010, genetic analysis followed by gene product analysis indicated that the original tumour cell was a Schwann cell or Schwann cell precursor. Schwann cells sheath the nerve fibres in the peripheral nervous system and are derived embryologically from the neural crest. The researchers, however, could not categorically exclude the possibility that the tumour might have elements of both Schwann cell and neuroendocrine cell types (Murchison et al., 2010; see also Tovar, 2012).

Parenthetically, the researchers doing the epigenetics work discussed above also claimed to have found “additional support for the clonal origin of the tumour from a Schwann cell or Schwann cell precursor” on the basis of “overall methylation patterns in tumour and nerves … [being] remarkably similar, but significantly higher than in other tissues (Ujvari et al., 2013).” The claim as stated was incomplete, however, because it lacked a test of significance between the results for nerves and the “other tissues,” brain, heart, kidney and liver (Ujvari et al., 2013: table 1). The claim would also not be supported if the trends in the data are as discussed above.

Interestingly, Schwann cells may be susceptible to turn cancerous. They can dedifferentiate and then redevelop after injury. Therefore, they are open to perturbations in the normal regulation these changes in growth and maturation. Schwann cells also already have low levels of MHC-I proteins on their cell surfaces, and reduced expression of these proteins is one means by which DFTs evade the hosts immune response (see discussion in Tovar, 2012; see also Murchison et al., 2010; Siddle et al., 2013).

On 29 August 2012, a research group published evidence that the Devil’s tumour cells have significantly shorter telomere lengths, as estimated by telomere copy number (TCN), than do cells of lung and spleen (Ujvari et al., 2012). This difference was interpreted as gradual shortening of telomere length during the many cell divisions the tumour has undergone since its origin in 1996, although the shortening is hypothesised to have stabilised and possibly even reversed due to a concomitant increase in an enzyme promoting telomere formation and a structural protein that helps prevent the telomere from lengthening indefinitely (Ujvari et al., 2012). However, the possibility that cells in various tissues differ in telomere length (see “Chromosomes” above) raises the possibility that the shorter telomere length in the Devil’s tumour cells may just reflect the normal telomere length in its original tissue type (nerve) rather than its tumourous nature. Data on telomere length in normal Devil nerve tissue is needed.

6/05/2016 8:41 AM 183

The evidence for telomere lengthening in the tumour cell lineage was based on pair- wise comparisons between three years, 2006 (N = 30), 2007 (N = 13) and 2010 (N = 8). Two of the comparisons, 2006 vs 2007 and 2006 vs 2010, were significant, although the former was just so (Kruskal Wallis H = 3.73, P = 0.054; pers. obs.) and one, 2007 vs 2010, was not (Ujvari et al., 2012: fig. 4). A Kruskal Wallis test of the data, as estimated from the original figure (Ujvari et al., 2012: fig. 3), confirms a significant difference among the three groups (H = 7.19, P = 0.027; pers. obs.).

Regression analysis of log TCN (original values estimated from Ujvari et al., 2012: fig. 3) and date of collection (data courtesy of DPIPWE) reveals a significant regression (F = 5.15, P = 0.028). When the two strongest outliers are omitted, the regression becomes more significant (F = 7.02, P = 0.011). However, just 10 and 13 percent of the variance in the TCN is explained by the regressions (r2 = 0.097 and 0.13, respectively; pers. obs.; Fig. 16). The passage of time, therefore, accounts for only a small part of the variance in the TCN, leaving the other contributors to be discovered.

Relative Telomere Copy Number and Specimen Collection Date By Sex

0 13/01/2006 26/08/2006 8/04/2007 19/11/2007 1/07/2008 11/02/2009 24/09/2009 7/05/2010

Fig. 16. Relative telomere copy length in primary tumours and date of collection of sample. Males, blue triangles; females, pink circles. TCN estimated from Ujvari et al, 2012: fig. 3 and date of collection and sex data courtesy of DPIPWE, Tasmania.

It may also be of interest that the variance of the untransformed TCN data is significantly higher for tumours growing in males than for tumours growing in females (F = 0.43, P = 0.02; pers. obs.). The reason for this difference also remains to be determined.

Interestingly, on 25 September 2012, a second group, using a direct visualisation technique of telomere length, published their results for several DFT culture cell lines and concluded that “telomere length has been maintained in the tumour over several

6/05/2016 8:41 AM 184

generations (Bender et al., 2012).” It is not clear if the difference between the two groups’ conclusions about DFT’s telomere length over time is due to a difference in degree of resolution, a time difference or is real.

Tests to Identify DFTD Tumours

Although advanced primary Devil Facial Tumours can be identified fairly confidently by visual inspection, it is more difficult to identify accurately both early primary DFTs and secondary (internal) metastases and to discriminate the occasional non-DFT lesion from a DFT. In the early years, researchers indentified DFTs through the lesion’s histological characteristics and karyotype (above). These, however, were time consuming. In 2011, researchers developed a third and more rapid test based on an immnohistochemical technique. The test uses an antibody-based stain for the protein periaxin, which as noted above is a product of DFTs original precursor cell type, a Schwann cell, and is still produced by DFT cells (Murchison et al., 2010; Kreiss et al., 2011a; Tovar et al., 2011; Tovar, 2012; Pye et al., 2015c).

Tumour and Devil Co-evolution

The work on the tumour strains and genetics makes it clear that the tumour shows heritable variability and that this variability can lead to evolutionary change. The Devil is also an evolving entity. Consequently, the two lineages - the tumour and the Devil - can respond evolutionarily to each other and therefore co-evolve (M. Jones, ABC Science Show, 3 September 2011). What the outcome of this co-evolution may be is unclear. Modelling offers no clear indication of the possible outcome in the Devil’s case, in part because critical parameters are still too poorly known (Hamede et al., 2012a).

The many examples of successful co-evolutionary relationships give hope for the Devil. The only other transmissible cancer, an ancient venereal cancer in dogs transmitted during copulation, is now benign, although how it behaved when it first arose some hundreds or possibly even thousands of years ago is unknown.

There are many instances as well, however, in which two “species” suddenly come together and one goes extinct. Many of these involve the introduction of widespread mainland predators or pathogens onto small islands where populations are small and locally adapted and consequently have neither the numbers nor genetic variability on their side.

The co-evolutionary interaction between the tumour and the Devils means that when there is an indication of possible resistance in a Devil population, the role each entity has to be ascertained. Has the tumour become less virulent or the Devil more resistant or both (K. Belov, in Coghlan, 2012)?

In mid-2015, a group of Devil researcher wrote, “the expected outcome of the facial tumour epidemic in devils based on other host/pathogen interactions is evolution of reduced pathogen virulence leading to coexistence (Hollings et al., 2015b; emphasis added). This was a surprisingly more optimistic view of the Devil’s future than the implication of extinction without human intervention that had characterised most previous assessments. But whether it was based on some new understanding of the

6/05/2016 8:41 AM 185

diseases’ virulence, or was an inference based on the Devil’s “unexpected” persistence (see “Annual Statewide Spotlight Surveys,” below) is unclear. Still, the researchers could not be too optimistic, writing: “… devil populations are likely to remain at very low levels for decades (Hollings et al., 2015b).”

Directed Evolution

Some Devil workers have suggested that humans should actively select Devils for certain traits to help the animals combat the disease. When it was realised that the disease might be due, at least in part, to the Devil’s own genetics, it was suggested they could be selectively bred in order to spread any resistant genes in the population (Mather, 2006; G. Woods, in Darby 2007b; DPIW, 2007a). This was subsequently discounted, because it might compromise the goal of saving as much of the Devil’s genetic diversity as possible (PHVA, 2008). Indeed, STDP policy seems to be that “insurance population devils will be bred so as to maximise and retain the current genetic diversity, rather than selective breeding for possible disease resistance (DPIPWE, 2010d).” Recently, however, some researchers in the STDP have begun talking about the possibility of selective breeding (Beeton and McCallum, 2011b: AFP, 2012), but whether or not this represents a change in official policy is unclear.

More recently, it was suggested that genetically less aggressive Devils might be selectively bred in order to decrease the aggressiveness that seems to facilitate the spread of the disease (Hamede et al., 2013).

The problem with both initiatives is that selective breeding for a specific trait can have unanticipated effects on other parts of the Devil’s genome and hence its biology. In the wild, natural selection leads to a well-integrated genome. But selective breeding in captivity, itself a source of artificial selection, could lead to a genome totally unsuitable for the wild.

In March 2013, a University of Sydney website dedicated to soliciting research funds from the public noted that “once conditions are stable, the … [aim is] to re-populate the wild with the most genetically robust population of devils (University of Sydney Inspired website, “Saving the Tasmanian Devil”).” This statement must have been either written or signed off by the principal university researcher. Whatever, it contains a remarkable assumption, namely, that once DFTD has passed from the scene, some researcher(s) will somehow determine what a “genetically robust population” is and then release it into the wild. And if this theoretically robust population should prove to be less robust than expected in the wild, where will the fault lie, with the researcher(s) or the Devil?

What Caused the Disease?

In all likelihood, the Devil’s facial tumour had its origin as a single mutation in a single cell in a single Devil (McCallum et al., 2007). Parenthetically, the original or “index” Devil was probably a female as no Y chromosome or vestige has been detected in the tumour karyotype (E. Murchison, Guardian video: Fighting contagious cancer, 29 September 2011; Darby, 2011; Deakin et al., 2012; Murchison et al., 2012; Pye et al., 2016b). But what caused that original mutation is unknown and as a unique historical event, it is likely to remain unknown (Bender, 2010; O’Neill, 2010; Jones

6/05/2016 8:41 AM 186

and McCallum, 2011b). Nonetheless, in the early stages of investigation into the disease, various parties suggested possible causes, and researchers made at least preliminary investigations into most of the suggestions.

Virus

An early thought was that a virus might have caused the cancer (Bevilacqua and Wood, 2003; B. Chadwick, in Wood, 2003c; Clark, 2003; Darby, 2003b; Jones, 2003b; Ladds et al., 2003; McManus, 2003; Pattie, 2003; Whinnett, 2003a; Wood, 2003a; Harington et al., 2006; McGlashan et al., 2006), as they are known to cause other types of cancer. However, attempts to find virus particles in tumour tissue or fluid derived from tumour tissue were unsuccessful (DPIWE, 2005a; A. Hyatt, in Loh et al., 2006; D. Hayes, in Pyecroft et al., 2007), and a search for virus DNA in the tumour genome revealed nothing (Murchison et al., 2012).

It is still possible, of course, that a virus did initiate the Devil’s tumour and the viral DNA lies undetected in the tumour genome. In that a virus is the only cause that is likely to have left a definitive signature, it would be worthwhile to continue surveying the tumour genome as opportunity permits. Interestingly, the STDP lists as two of its projects for 2012/13 a “search for tumour virus sequences in [the] devil genome and DFTD tumour genome databases” and the “validation of a devil virus isolation protocol (STDP Annual Report, 2012/13).”

Parenthetically, as mentioned above (“Genome sequence”) an early media story said that a preliminary study had found a healthy Devil’s DNA laden with a retrovirus’ DNA (Bevilacqua, 2006g, 2007). Nothing further seems to have come of this specific report. Interestingly, however, in May 2013, the head of the DPIPWE’s Animal Health Laboratory’s Wildlife Disease Research unit and co-workers submitted a 3189 base pair sequence of a DFTD-associated Devil retrovirus RNA to GenBank (KC 894732.1). The sequence was apparently submitted in anticipation of a then as yet unpublished paper (Ciu et al., unpublished ). The significance of this retrovirus for DFTD remains to be elucidated.

Toxins

Poisons and pollutants were also suggested as possible initiators of DFTD (Denholm, 2006a; Rehmeyer, 2013). As omnivores, Devils must frequently encounter such toxins through feeding, explaining perhaps why the primary tumour occurs almost always on the face, muzzle and oral cavity (Harington et al., 2006; McGlashan et al., 2006; Obendorf and McGlashan, 2008).

As to poisons, strychnine was widely used between the 1860s and early 1950s to control both native and exotic species (Mooney, 1992; Obendorf and McGlashan, 2008); it was even used as late as 1964 (Green, 1967). Although strychnine was banded long ago, it is unknown how much “old stock” might have been retained and used. Strychnine is unlikely to be relevant to DFTD, however, because although a potent poison, it is not mutagenic. Parenthetically, Devils feeding on strychnine-laced baits rarely died close enough to the baited carcass to be found, suggesting the poison took long enough to work to allow a Devil to move off some distance (Green, 1967).

6/05/2016 8:41 AM 187

Sodium monofluoroacetate (1080) took over from strychnine in 1952 (Obendorf and McGlashan, 2008). Devils, however, are fairly resistant to this naturally occurring plant compound, as measured directly by their tolerance levels and indirectly by their increase in population coincident with its introduction and use (data from RSPCA, cited in DPIPWE 2010; see also Ross, 2008). In addition, there seems to be no evidence that 1080 is mutagenic.

Organochlorines, of which DDT is the best-known example, were widely used to control insect pests up until the 1970s, and although no longer used today, some of the breakdown products, such as DDE, are long lasting in the environment. Organochlorines have a number of detrimental biological effects, but they are not generally recognised as mutagens. A survey of Tasmanian wildlife in 1975-1977 found organochlorine residues in Devils to be usually ≤ 0.01 mg/kg. But one of nine Devils had DDE residues of 0.50 mg/kg and another had 0.10 mg/kg. Both animals were from nearby localities in the central highlands (Bloom et al., 1979: figs 2-3, table 2). In a survey in 2003-2007, one of 32 animals had a level of 0.47 mg/kg and another 0.8 mg/kg (Ross, 2008). The relevance, if any, of the few animals with higher values is unclear, but it is interesting to note the only slight drop in maximum values over an interval of 26 years, during most of which organochlorines were no longer used.

The organophosphate insecticide mevinphos (product name: Phosdrin) was used illegally between 1991 and 1998 to poison Devils directly before being withdrawn from sale (Bevilacqua, 2005e; McGlashan et al., 2006; Obendorf and McGlashan, 2008). Mevinphos, like other organophosphates, is a mutagen and disrupter of chromosomes (Obendorf and McGlashan, 2008). Organophosphates biodegrade quickly in the environment, but this would not be relevant were an unlucky Devil to come into contact with a still-active compound.

A second-hand reference to the illegal poisoning of “hundreds” of Devils with a variety of poisons including mevinphos in northeast Tasmania in the mid-1990s, that is, about the time Devil facial tumour disease first appeared (S. Cronin, in Bevilacqua, 2005e), is intriguing but seems to have never been followed up in detail. The reasons for this are unclear.

As to pollutants, a variety of industrial chemicals and their breakdown products have been considered. Devils scavenge in rubbish tips, or at least they used to, before the tips were “tidied up” and more rigorously controlled, and rubbish tips were, at least were up until recently, the single biggest cumulative terrestrial sink for the by- and end-products of modern industrial society.

Several industrial pollutants have been analysed in both diseased and healthy Devils, and none occur in concentrations currently thought to be important. But most of the standards for mammals are based on placentals, not marsupials, so there remains the possibility that the latter react differently (Ross, 2008).

Dioxins are a widespread and long lasting class of pollutants that are stored cumulatively in fat and become more concentrated in animals higher in the food chain. They disrupt reproduction and development in vertebrates, but they are not carcinogenic. The dioxin levels in Devils, which are of course at the top of their food

6/05/2016 8:41 AM 188

chain, were well within the designated safe standards for vertebrates and showed no differences between healthy and stricken animals (Ross, 2008).

Organobromides were, and some continued to be, used as flame retardant agents in material goods, especially plastics and textiles. They are many different varieties and some a known to be toxic and others carcinogenic.

Two classes of organobromides have been examined in Devils, polybrominated biphenyls (Vetter et al., 2008) and polybromianted diphenyl ethers (Moore, 2008; Ross, 2008). Both classes of materials occur in many varieties, disperse widely, bioaccumulate (mainly in lipids) and have debilitating effects on the development and physiology of animals. Some forms of each class of materials are known, or suspected, to be carcinogenic at least in some animals.

The use of polybrominated biphenyls (PBBs) has been regulated in western countries since the 1970s. They were never manufactured in Australia and probably entered the country embedded in imported products.

Total PBB levels, of which PBB 153 comprised the major part in each Devil, ranged 340-0600 pg/g lipids in eight healthy and eight diseased Devils. There was no significant difference in mean levels between the two groups; indeed the two Devils with the highest (by far) levels were both healthy. These overall levels are well below internationally recognised thresholds for toxic effects (Ross, 2008; Vetter et al., 2008), although apparently the National Measurement Institute, which did the tests, was reported as describing them, at least initially, as “high (News.com.au, 22 January 2008).”

This reassuring outcome aside, the results raise two other interesting issues. First, it is surprising that levels in Devils are so high considering PBBs were never manufactured in Australia. This raises the question of where the PBBs came from and what the levels may be in other Australian species and in humans. And second, the results for individual Devils range over two orders of magnitude. This raises the question of why some Devils have such high levels. What might they have gotten into?

Total polybrominated diphenyl ethers (PBDEs) ranged 570-54800 pg/g lipids in the same Devils mentioned above. There was no significant difference in mean levels between the two groups; indeed the highest value (by far) was, again, from a healthy Devil. These overall levers are well below internationally recognised thresholds for toxic effects (Ross, 2008; Vetter et al., 2008), although apparently the National Measurement Institute, which did the tests, was reported as describing them, at least initially, as “ ‘reasonably high (News.com.au, 22 January 2008).’ ”

Four chemical elements that usually only occur in low concentrations in nature but that can be concentrated to high levels through human activity have also been examined in Devils: arsenic, cadmium, lead and mercury. Arsenic and cadmium are carcinogens; mercury is a possible carcinogen in some compounds, but lead is unlikely to be so. In the case of arsenic, some Devils had levels that would be considered high in dogs, another carnivore, albeit a placental one, and the difference between healthy and diseased Devils just failed to reach significance (P < 0.07) (Ross,

6/05/2016 8:41 AM 189

2008). The other three elements, however, were within concentrations thought not to be important from studies in humans and other animals. There was also no significant difference between healthy and diseased Devils (Ross, 2008).

After the discovery that DFTD almost certainly started as a mutation in a single cell in a single Devil, that is, was a unique historical event instead of an on-going one, focus turned away from the cause, virus or chemical, to how to deal with the disease.

In November 2014, however, two social science academics suggested the role of chemicals, especially atrazine and 1080, in the origin and spread of the disease should be re-opened as a line of research. They even intimated there was a collusion of interests in avoiding any possible implication of chemicals used in agricultural and forestry (Warren, 2013; Warren and Martin, 2014). The authors also had an issue with the meagre evidence for direct transmission from one Devil to the other, a scepticism fed by some experimental results purporting to demonstrate direct transmission but never published (Pyecroft, 2007; Pyecroft et al., 2007; also see “Direction of transmission”).

Their proposal met with widespread criticism on the grounds they were either uninformed or ignorant of the genetics of the tumour, especially the signature cytogenetics and similar genetics of the original tumour cell and its descendents.

Various commentators on the original article (comments in Warren and Martin, 2014) and a second article written in reply to it (Woods, 2014) laid out convincingly the case for the tumour’s unique origin and its direct transmission.

It is interesting to stop a moment and think what kind of experiments might address the question of the role of chemicals in the origin of the disease. Most likely, it would require establishing cultures of precursor Schwann cells (the cell in which DFTD- causing mutation seems to have occurred) and then applying various chemicals of interest in varying concentrations and looking for a massive chromosome rearrangement similar to that found in DFTD.

Even if a potential causal chemical were identified, however, its continued use would still most likely be subject to a cost benefit analysis, including the likelihood of its transmissibility, before it was banned.

The Role of Humans

Although we will probably never know the cause of the initial mutation that led to the Devil’s facial tumour, the circumstances of the disease suggest that on the balance of probabilities, humans had a role (N. Mooney and A.-M. Pearse, in Denholm, 2005; Hobart Mercury, 18 November 2009; N. Mooney, in Denholm, 2010e). The temporal association between the origin of the disease and the peak period of dumping and spread of mutagenic products of modern western industrial society is simply too striking to think otherwise. The alternative would be that this rare form of disease arose precisely at the time in the Devil’s 11 million year evolutionary history when modern technological society had just become well-established Tasmania.

6/05/2016 8:41 AM 190

Knowing how the facial tumour got its start would no doubt be interesting. But even if a single controllable cause, chemical or otherwise, could be identified, the fact that it was almost certainly a one-off occurrence might not be enough to lead to any change in practice.

Tumour Transmission

Almost as soon as it was realised Devil facial tumour disease was spreading, it was assumed that it was probably being transmitted by one Devil biting another (Wood, 2003a; N. Mooney, ABC Science, 1 August 2003; Whinnett, 2003a-b; Darby, 2003a). Comments recorded at the time do not indicate why this was thought to be the case. It may have been due to the initial belief that the tumour was caused by a virus, probably a retrovirus, and the Devil’s well-known propensity for biting each other would have sprung to mind an obvious form of transmission for such a virus.

There was also early recognition that if the disease were caused by a virus, it could also be transmitted through the air. It was only when the case for direct transmission of the tumour as an allograft was convincingly made that this thought ceased to guide quarantine and isolation measures.

Tumour as an allograft

The cytogeneticists who discovered that the tumour cells had a highly altered chromosome complement also noticed that while the karyotype of the tumour cells always had two similar pair five chromosomes, it obviously differed from the constitutional karyotype of one infected animal that had a pericentric inversion in pair five (now recognised as pair six).

This led the researchers to propose that the tumour did not arise anew within the cells of affected Devils but was apparently a cell line that started in one Devil and was transmitted to other Devils, probably through biting. Instead of being rejected as “non-self” by the bitten Devils’ immune system, the tumour cells were somehow accepted as “self” and allowed to grow uncontrolled.

In other words, although the transmitted cells are actually a graft from another individual (an allograft), they are accepted as if they were a graft, say of skin, from one part of an individual’s body to another part of that individual’s body (an autograft). But why the graft was accepted rather than rejected was unclear. This understanding of the tumour and its transmission came to be known as the “allograft theory” (Pearse and Swift, 2006; see also Gramling, 2006).

This research was published in a one page article and remains one of the most elegant pieces of research and inferential thinking to come out of the entire effort to understand the disease. And it helped re-direct thinking about the cause of the disease. Almost overnight, for example, it put paid to the idea a virus caused the disease (Jones, 2012).

Support for the allograft theory was soon forthcoming. In February 2007, the preliminary, orally-presented results of an amplified fragment length polymorphism (AFLP) analysis were in line with the theory. Levels of polymorphism were an order

6/05/2016 8:41 AM 191

of magnitude lower between tumours than between tumours and host and between Devils (Fox, 2007). It is not clear why this pilot project was not carried to completion.

In September 2007, immunologists reported that when lymphocytes from two healthy Devils were combined, they failed to stimulate each other to proliferate as might be expected. This result was consistent with the allograft theory’s basic inference that the Devil’s immune system could not recognise cells from another Devil as foreign (DPIW, 2007a; Woods et al., 2007; also G. Woods, pers. comm. in Pearse and Swift, 2006). This result was soon shown to be wrong (below) and as such became irrelevant to the allograft theory.

In October 2007, partially to test the allograft theory, geneticists genotyped four polymorphic microsatellite loci and three MHC in 20 tumours, 15 of which were matched, as well as five additional unmatched tumours and 11 healthy animals. They found that the genotypes at these loci were identical in all tumours examined (N = 20 from eight localities, but no MHC data for one tumour) but different from their matched hosts (N = 15 from seven localities, but no MHC data for one match) and the healthy Devils (N = 11, but no microsatellite data for one and no MHC data for four), which also mostly differed among themselves (Siddle et al., 2007b: tables 1 and SI table 3). Subsequently, an analysis of 14 microsatellite loci got a similar result (Murchison et al., 2010). The genetic similarity of all the tumours strongly supports the idea that DFTD arose in a single Devil.

If tumour cells are transmitted, the question arises as to when a tumour becomes capable of releasing cells that might be transmitted. When the tumour first becomes noticeable to the naked eye, the overlying skin or lining (mouth) is still intact, and it seems likely the potentially infectious tumour cells remain contained for some presumably short period prior to the tumour opening up. If so, then this “intact tumour” period provides a brief window in which a Devil could be identified as affected but not infectious (DPIPWE website: “Nipping DFTD in the bud,” 26 August 2010).

Interestingly, a STDP project for the year 2012/13 is a “review of the evidence for [the] allograft theory (STDP Annual Report, 2012/13).” It is not clear what this means, but it could be interpreted to mean there may be some reason to think the theory is wrong.

Direction of transmission

Prior to 2012, it was widely assumed that the predominant direction of transmission was from the biter to the bitten. The primary tumours of the head are friable and pieces dislodge easily. Swabs from animals with external tumours weeping into the mouth or tumours in the mouth itself “… show high levels of free floating tumour cells in the saliva … (Wells, 2010).” Hence it seemed likely that when Devils bite each other in fighting or mating, small pieces of tumour might break off and become inserted into the sub-dermal (skin) or sub-epithelial (mouth) tissue at the wound site (Pearse and Swift, 2006; McGlashan et al. 2007).

The Devils long canines seemed especially likely to embed tumour cells into another animal. Indeed, an unpublished study apparently showed that the canines were the

6/05/2016 8:41 AM 192

“only teeth anatomically capable of creating deep penetrating wounds into the soft tissue of another Devil (Obendorf et al., submitted, as cited in Obendorf and McGlashan, 2008, italics in original).” Further, DFT cells can be harvested by brush scraping the canines (Woods, 2010).

An experiment to test the direct transmission from a diseased Devil to a healthy one began sometime prior to July 2006 (Sunday Tasmanian, 23 July 2006; see also Harington et al., 2006; McCallum and Jones, 2006; McGlashan et al. 2007; Pyecroft, 2007; Pyecroft et al., 2007). It involved injecting 16 healthy captive Devils with tumour “cells derived from cultured cell lines and natural tumours,” which then grew into tumours in the test animals (Pyecroft, 2007; Pyecroft et al., 2007; Hamede et al., 2008; Obendorf and McGlashan, 2008). The results of this work have not been published, but they have been cited at least once (Bender, 2010). Curiously, in late 2014 the results were described as “in prep for publication” (S. Pyecroft as comment in Woods, 2014). In April 2008, the STTDP’s Biosecurity Guidelines noted, “… the actual method of transmission from one animal to another is not confirmed … (DPIW, 2008).”

In a public lecture in 24 May 2010, the lead immunologist working of DFTD noted that over “a couple” of experiments seven out of seven Devils inoculated with live DFT cells developed tumours (Woods, 2010). This may have been a reference to a PhD thesis in which seven Devils were injected with live tumour cells and at least six were explicitly identified as having developed a tumour at the site of injection (Brown, 2013).

The only other insight into the direction of transmission involved the incidental observation that when Devils were injected with live tumour cells to see if they would develop antibodies, tumours developed at the injection sites (see “Distribution of Tumours on the Body” and “Cedric and Clinky … and Christine”). Thus, there have been at least nine and possibly ten examples of direct transmission of tumour cells by injection.

In 2012, field work returned some startling results that caused researchers to re-think the direction of transmission. One implication of the idea that transmission is from biter to bitten is that healthy animals with more fresh bites should be more likely to develop primary tumours than animals with fewer fresh bites. But when tested with 600 observations of 107 adults in a marked population, West Pencil Pine, the opposite was found. Healthy animals with fewer bites were more likely to develop primary tumours over the following six and nine months than animals with more bites. Furthermore, most of the primary tumours were inside the mouth as opposed to outside. The researchers reasoned that dominant animals were probably more likely to bite than be bitten and hence more likely to bite into a tumour, once in the mouth, the tumour cells could implant in any pre-existing wound received in feeding (bone fragments and echidna spines) or fighting (Hamede et al., 2013).

Intriguingly, the bitten to biter direction of transmission explains why, otherwise inexplicably, primary tumours rarely appear elsewhere other than the head (pers. obs.). But reciprocally, tumours on the body, rare as they apparently are, as well as tumours on the head but well away from the mouth, suggest that in some cases at least, transmission may occur from biter to bitten

6/05/2016 8:41 AM 193

An important epidemiological implication of the bitten to biter theory is that as a dominant animal becomes subordinate due to the growth of its own facial tumour, it shifts from being a receiver to a transmitter. And an interesting evolutionary implication is that there should be increased selection pressure against aggressiveness in Devils (Hamede et al., 2013).

At present, the only means of independently assessing aggression/dominance in Devils other than the number of bites they receive, or do not receive, is their behaviour when handled after trapping (Hamede et al., 2013). And while it would be reasonable to assume that age or size might be an indicator, this has never been tested. In any event, the researchers found no effect of age, at least, on their results (Hamede et al., 2013).

In thinking about these counter-intuitive results, it is worth remembering that the population in which they were made is the only one in which there is evidence of some degree of innate resistance to DFTD or lesser virulence in the tumour (Hamede et al., 2012a). Consequently, the question arises of any possible relationship between the number of fresh bites and resistance to the disease, such as more bites and an accompanying intake of tumour cells boosts some immune response.

In any event, a useful follow up of this work during repeated surveys would be to map the fresh DFT lesions on individually marked Devils and when new lesions are subsequently found, determine if their position corresponds to a recent past lesion.

Other modes of transmission remain possible as well. Devils eating pieces of tumour bitten off dead or living Devils that had recently died from the disease might pick up live tumour cells through open sores in their mouths (Jones et al., 2007a; Hamede et al., 2008; McCallum et al., 2009; Hamede et al., 2012a-b). Devils might also pick up live cells from the saliva and froth (fomites) left in a carcass by a diseased Devil (Jones et al., 2007a; Hamede et al., 2008; McCallum et al., 2009; McCallum and Jones, 2010; Hamede et al., 2012a-b, 2013). If any of these modes of transmission were common, however, Devils with more open wounds in and around the mouth would be expected to be at greater risk of picking up DFT cells. But this is apparently not the case as the “bitten to biter” data show.

Why Aren’t the Tumour Cells Rejected?

The proposal that the tumour cells are transmitted from one Devil to the other as an allograft immediately raised the issue of why the Devil’s immune system did not reject the tumour cells as “other” or “foreign” (Pearse and Swift, 2006). There were two possible reasons. Either the Devil’s immune system was inherently unable to recognise the tumour cells, or the tumour had the capacity to suppress or evade the immune system’s surveillance system (A.-M. Pearse, on ABC Science Show, 4 February 2006).

Research addressing this issue quickly returned results suggesting the problem might lie with the Devil’s immune system. Genetic research, for example, showed that diversity in the Devil’s major histocompatibility complex, that part of the genome involved in the immune system’s recognition of self and non-self, is low (Siddle et

6/05/2016 8:41 AM 194

al., 2007b). The low diversity in this part of the Devil’s genome fit well with the earlier finding of the Devil’s generally low genetic diversity (Jones et al., 2004a; Siddle et al., 2007b). Indeed, as early as September 2003, a researcher had suggested that an earlier population decline might have reduced the Devil’s genetic diversity and made it susceptible to the disease (Wood, 2003c). This result suggested that the Devil lacked the ability to respond to the tumour.

Geneticists also showed that the tumour cells’ surface antigen genes were being expressed (Siddle et al., 2007b), although it remained to be determined whether the antigens were actually appearing on the surface of the cell (Murchison, 2009). As it turned out, they were not (Siddle and Kaufman, 2013; Siddle et al., 2013).

Immunologists measured the in vitro proliferation of lymphocytes in response to a variety of standard antigens. These tests indicated that the Devil’s immune system was fully competent, although there was wide variation in the degree of response among, and even within, individuals (Woods et al., 2007).

The logical next step was to test lymphocytes’ reactivity to each other. Lymphocytes from 15 Devils of unstated provenance were mixed together in pair-wise comparisons, but “there was a failure of the … lymphocytes to proliferate” (Stimulation Index mean = two; Woods et al., 2007: fig. 4; STTDP website: “A contagious cancer,” 18 April 2007; see also Kreiss and Woods, 2007; Menzies Research Bulletin 52).

Subsequently, a second mixed lymphocyte experiment was done but this time individual Devils were compared not with each other but against a pool of mixed Devils. Lymphocytes from nine eastern Devils were tested against a pool of lymphocytes from 30 eastern Devils, and “no mixed lymphocyte responses were observed … (SIs ranged 0.3 – 4.6; Siddle et al., 2007b: table 4)” Later, however, the response would be called “low (Kreiss et al., 2011b).” These results suggested that the lymphocytes from the different individuals were not different enough to be recognised as “non-self” (Woods et al., 2007; Siddle et al., 2007b).

Interestingly, however, the results of these two experiments were seemingly opened to question shortly after they were published, because the researchers wrote in their annual report for 2007 that “mixed lymphocyte reactions among eastern and western devils were performed, and some experiments (especially between West and East devils) showed high reactions … (Menzies Research Institute Annual Report 2007).”

This apparent ambiguity lay dormant for more than three years until mid-2011 when the researchers published the results of what was apparently a third mixed lymphocyte experiment using 24 Devils from a variety of locations including both the east (N = 15) and west (N = 9) and presumably similar (identical?) protocols as in the earlier two experiments. The results of this experiment showed that the pair-wise lymphocyte reactions varied between “weak” and “strong” (overall SIs ranged 0.5-128, medians ranged c. 2-32; Kreiss et al., 2011b: fig. 2). The researchers explained the different results between the second and third experiments (they did not mention the first experiment) by suggesting, “it is likely that different (or rare) MHC antigens were ‘diluted out’ and not present in enough quantity to stimulate proliferation of the responder cells (Kreiss et al., 2011b).”

6/05/2016 8:41 AM 195

The researchers also compared the lymphocyte reactions among groups in the east (n = 5), groups in the west (N = 4) and between a combined east vs east group, west vs west group and an east vs west group. Two results were of particular interest. First, it was noteworthy that the reactions of two individuals within one group could vary almost as much as any two individuals in all groups. The two Devils making up the Granville “group” in the west for example, had replicate SIs that were as broad as the entire range from all other eight groups, 0.5 to almost 128.

Second, non-parametric pair-wise comparisons within and among the groups revealed significant differences in all comparisons, but the “greatest … responses” were between Devils in the east vs west group, presumably as judged by the relative strength of the pair wise statistical comparisons with the east vs east group and the west vs west group (Kreiss et al., 2011b). Pairwise comparisons are not the most robust form of statistical analysis in this case, however, and it unclear what a multiple comparisons analysis would have returned.

Nonetheless, the researchers interpreted their mixed lymphocyte reaction results as suggesting, “… that unaltered allogeneic cells should initiate an immune response in most recipient devils, irrespective of the origin of the animal (Kreiss et al., 201b).”

It is not clear, however, that the lymphocyte reaction analysis, at least as carried out on the Devil to date, has any significance at all. The original PhD thesis underlying this work, when finally made generally available in March/April 2015, reported the data for the only duplicate experiment for the lymphocyte reaction analyses (Kreiss, 2009: tables 5.8 - 5.9), and there was no significant correlation between the two results (rs = 0.43, P = 0.11; pers. obs.). In other words, the results were not reproducible. This suggests an unidentified variable was affecting the results.

The revised results of the lymphocyte experiments were not published as stand-alone research, but as part of a larger and more significant experiment on skin grafts. A call for reciprocal skin graft experiments had been made as early as October 2007 (Siddle et al., 2007b); funding ($ 23 000) was available by early January 2008 (UTas media release, 7 January, 2008; ABC News, 7 January 2008); grafts were done sometime in mid-2008 (ABC News, 11 July 2008, and results were expected by the end of that year (ABC Hobart, 2008). The results first became available in an unpublished PhD thesis submitted in September 2009, which reported, “… skin grafts between unrelated devils were immunologically rejected …” (Kreiss, 2009; made generally available online in March/April 2015). They were made known to a specialist audience at a conference in July 2010 (Woods et al., 2010a) but not published broadly until July 2011 (Kreiss et al., 2011b, received by the journal on 27 May 2011). The results showed that a skin graft from one Devil would be rejected by another Devil. Seven Devils from eastern Tasmania received eight reciprocal grafts and the five that engrafted successfully were subsequently rejected (Kreiss et al., 2011b). Clearly, the Devil’s immune system has no trouble indentifying normal cells from another Devil as foreign. In the PhD thesis, the significance of the result was mentioned three times: “Importantly, skin grafts between unrelated devils were immunologically rejected, indicating that spreading of DFTD is not simply due to a lack of MHC diversity in this species;” “…it is unlikely that lack of MHC diversity alone accounts for the transmission of DFTD tumour cells across the MHC barrier,” and “importantly, skin

6/05/2016 8:41 AM 196

graft surgery between unrelated devils… showed that the functional MHC diversity of this species is enough to prevent the transplantation of foreign tissues.”

As soon as the result of the skin graft experiment were available, presumably sometime no later than late in 2008, the results must have been circulated rapidly among at least the Devil immunological and genetic community, if not among Devil researchers generally. Why, therefore, was the “low genetic diversity” explanation for DFTD’s spread not put to rest until the results were formally published in July 2011 (manuscript received on 27 May 2011)? As the researchers closest to this work wrote in their annual report for that year, “previously it was thought that the transmitted tumour cells weren’t rejected because of a lack of genetic variation within the Tasmanian devil population. But the latest research [the reciprocal skin grafts] suggests that this is not the explanation. Something is missing from the tumour cells and the spread of DFTD is due to the nature of the tumour rather than a lack of genetic diversity within the devil population (Menzies Research Institute Annual Report, 2011, p. 8).”

Parenthetically and paradoxically, even after the results of the skin graft experiments were published, researchers and administrators continued to evoke the Devil’s low genetic diversity as contributory to the disease’s spread (G. Woods, Science Alert, 25 July 2011; G. Woods, in STDP website: “Unlocking the secrets behind the spread of DFTD,” 9 August 2011; G. Woods, on ABC Rural Bush Telegraph, 20 September 2011; Newsletter, September 2011; Cheng et al., 2012a-b; Deakin et al., 2012; Hamede et al., 2012a-b; K. Belov, in Gill, 2012; Lane et al., 2012; Newsletter, September 2012; Tucker, 2012) even though there was no new evidence to support this belief. Even as late as mid-2014, one research group wrote, “… limited levels of genetic variation across all of these MHC loci … contribute to delayed allorejection responses in devils (…Kreiss et al., 2011 …)” (Cheng and Belov, 2014); however, the relevant reference actually states, “the onset of rejection occurred between seven and 14 days, which is within the normal range … (Kreiss et al., 2011b).”

The issue seems to have been laid to rest, however, when one of the strongest proponents of this view wrote on or before 29 May 2012, “We originally believed that DFTD was able to spread because of low diversity in the MHC of devils [Siddle et al., 2007a], and we predicted that devils would not be able to “see” DFT cells as foreign because the devils and tumors share the same MHC antigens. However, recent studies suggest that the tumor itself is capable of immune evasion [Kreiss et al., 2011b] (Deakin and Belov, 2012; see also Belov, 2012).”

It is also salutary to remember that the STTDP’s Senior Scientist once wrote, “This [DFTD]is the first case anywhere in the world in which it has been unequivocally demonstrated that lack of genetic diversity is the key problem in an emerging wildlife disease (McCallum 2008b, emphasis added).”

As far as can be determined from the public record, the last research article to evoke “low genetic diversity” as a factor in the spread of DFTD was published, online, on 3 September 2012 (Hamede et al., 2013). This was also the latest “received for publication” date, 21 July 2012, for the concept as well.

6/05/2016 8:41 AM 197

Interestingly, the Devil’s “low” genetic diversity is still evoked with regard to its “tumultuous history of population crashes (Morris et al., 2013).”

The skin graft experiment and the revised mixed lymphocyte experiment caused researchers to turn their attention to the tumour’s ability to counteract the host’s immune response (Kreiss et al., 2011b; ABC, World Today, 22 July 2011; Science Alert, 25 July 2011; Belov, 2012).

As mentioned, there are two general ways the tumour could counter the immune response. The tumour could have evolved a means of suppressing the host immune system’s ability to recognise or counter those antigens (Woods et al., 2007; G. Woods, as reported in Denholm, 2010b; E. Murchison, in Wylie, 2010; K. Belov, in Gill, 2012). This can be called the suppression hypothesis.

Alternatively, the tumour’s own antigens could have evolved to be unrecognisable as “other” by the host’s immune system (Siddle et al., 2007b; Woods et al., 2007; O’Neill, 2010; E. Murchison, in Wylie, 2010; Menzies Research Institute Annual Report 2011, p.8; Pearse et al., 2012;Woods, 2012b). This can be called the evasion hypothesis. On present evidence (below), the tumour seems to be evading rather than suppressing the Devil’s immune response.

Although the mid-2011 report of the skin graft experiments was first clear indication that the tumour’s ability to spread was due to the tumour and not to the Devil’s immune system, as shown below, some researchers were apparently thinking about the tumour’s role and making observations relevant to suppression/evasion hypotheses before this time.

Suppression

With regard to the suppression hypothesis, in the first half of 2007, it was noted that supernatant from a Devil facial tumour culture did not inhibit the ability of lymphocytes to proliferate when challenged with an antigen in vitro (Siddle et al., 2007b). This was, effectively, a test of the suppression hypothesis.

In August 2009, a researcher was given a grant to “… test the theory that the Devil’s facial tumour produces cytokines, which are a type of protein suspected of suppressing the devil’s immune response against the invading tumours (UTas media release, 21 August 2009; STTDP website: “Public generosity funds further devil research,” 19 October 2009).” This would have been a test of the suppression hypothesis. Cytokines are often mentioned as a possible suppressor of the Devil’s immune response (E. Murchison, in Giles, 2012). Nothing, however, ever seems to have been published on this research. Another research group subsequently showed that cytokines probably play no role in allowing DFT cells to avoid the immune response (Morris and Belov, 2013, below).

In early 2010, research threw up another intriguing result with relevance to the suppression hypothesis The Devil’s facial tumour expresses a precursor to the hormone that controls cortisol levels in the blood and Devils with tumours had levels of blood cortisol that were just short of significantly higher than in healthy Devils (P = 0.06). Cortisol has immunosuppressive activity. This was the first indication of how

6/05/2016 8:41 AM 198

the tumour might be counteracting the host’s immune system (Murchison et al., 2010). Nothing further, however, seems to have come of this observation

In 2012, a large three year grant was awarded to test, in part, “… if Devil Facial Tumour Disease reduces the effectiveness of the devil's immune system (Menzies Research Institute News Archive, “National funding success for Menzies,” 2012).” Although the exact scope of this is unclear, it could be taken to include research into the suppression hypothesis.

At a public lecture in June 2012, the head of the immunology lab working on the Devil’s immune capabilities reported on an honours project which found that the cancer-suppressing cytokines interlukin 10 and TGF-beta were detected in 16 of 18 tumours in the former case and 10 of 18 tumours in the latter but were clearly not universal (G. Woods, 2012b: slide at time 56:12).

And still on the issue of cytokines, in early 2013, researchers examined the expression of four cytokine genes in tumours and healthy controls and found no increased expression in tumours (Morris and Belov, 2013; paper first received in mid-October 2012). This result suggests, therefore, that DFT cells are not using a method that many cancers use to suppress the host’s immune system. This work, however, was called into question for not having using assay methods sensitive enough to detect the cytokines (Siddle and Kaufman, 2015).

This same lab, however, seemed to continue to believe suppression was important, because in a paper received in early November 2012, the researchers wrote, “… the tumour …must also be suppressing the immune response (Belov et al., 2013).”

Evasion

With regard to the evasion hypothesis, in February 2010, a grant was awarded to “explore why the Devil’s immune system fails to recognise the tumour and mount an immune response.” The focus of the research was the “… specialised immune molecules found on the outside of the tumour cells. If there is something wrong with these molecules or they are absent, this would explain why the devil immune system does not respond to the tumour (H. Siddle, in The Examiner, 18 February 2010).” In 27 March 2011, a media report said that researchers in this same lab had discovered “when the disease invades a host devil, it introduces a gene that lacks a protein that would identify that cell as coming from another animal” or as one of the researchers said “the receptor immune system doesn't see any enemy flags (Cart, 2011).” In late 2012, the researchers published their results but not the actual research. They reported that tumour cells do not produce the surface molecules that would allow them to be recognised as “other” (Siddle and Kaufman, 2013 but published online on 30 November 2012). This was the first clear indication that the tumour’s ability to evade the Devil’s immune system.

In a paper accepted for publication in August 2011, other immunologists suggested, based on the Devil’s functioning immune system but no immune response to DFTD, that the tumour cells had “… a capacity to evade the host antitumour response (Brown et al., 2011).”

6/05/2016 8:41 AM 199

In February 2012 one of the principal immunologists working on the Devil project told a journalist “…we discovered we could activate lymphocytes from a healthy devil, mix them with tumour cells in culture and they’ll kill the tumour cells quite well, so we know that the devil’s immune system has the capacity to kill the tumour cells (G. Wood, in Nogrady, 2013).” This was possibly the research referred to in late in 2012 when the researcher stated in an oral presentation, “… we now have evidence that under the right conditions, the immune cells of the Tasmanian devil can respond to DFTD cancer cells (Woods, 2012b).” It is not clear what meaning this may have for either the suppression or evasion hypothesis.

Aside from the STDP’s all too frequent habit of announcing results first in the media, one wonders why this “activated lymphocyte” experiment was not done earlier, as it would have helped shift the focus from the Devil’s immune response onto the tumour. Perhaps the “activation” of the lymphocytes was a recent technological breakthrough allowing the discovery.

It is interesting to speculate about what might have happened if the revised mixed lymphocyte results had been published quickly instead of several years later, because they could have prompted, if not the complete shift of focus away from the immune system and onto the tumour, then at least a more balanced approach. Perhaps researchers inside the Devil research community knew of the revised results before they were published, but outsiders would not have known.

There is also another reason the “low MHC genetic diversity” hypothesis may have enjoyed such a strong run. The demonstration that MHC genes are expressed in DFT cells seems to have led researchers to assume that the encoded molecules actually made it to the DFTs cell surface. This assumption was succinctly summarised in one table as “DFT cells express MHC molecules” (Woods et al., 2007: table 2) and the work behind it cited in the text as “Siddle et al., unpublished.”

The work referred to has shown that MHC genes, both class I and II, are expressed at least as far a transcribing messenger RNA (Siddle et al., 2007b). From this, the researchers concluded, that while “the most common mechanism of immune evasion by tumours is down-regulation of classical cell surface MHC molecules … we show that this mode of immune escape does not occur (Siddle et al., 2007b).” This assumption was a step too far, as it would subsequently be shown that while the genes are expressed, the end products, MHC molecules, never make it to the cell surface (Siddle et al., 2012; Siddle et al., 2013; and Siddle and Kaufman, 2015). The first acknowledgement in the literature the original assumption might not be correct was in early 2012 (Belov, 2012; Siddle et al., 2012; see also Deakin and Belov, 2012).

Another possible reason for the “low MHC genetic diversity” hypothesis’ long run may have been the enthusiasm of one of its strongest proponents. Shortly after having read the 2006 paper suggesting the tumour acted as an allograft (Pearse and Swift, 2006) this researcher is said to have immediately thought, " ‘I know what it is!’ " and phoned the senior author of the paper and said “‘Hey, can I work on the project?’” (K. Belov, in Dow, 2010; see also STDP Annual Report, 2013/14).

6/05/2016 8:41 AM 200

Finally, it has to be said that it was strange that STDP researchers did consider the evasion hypothesis more seriously from the very beginning. Not only is this a common method for cancers to evade the hosts’ immune response, it has also been suspected as early as 1987 as being the method used by the only other known transmissible cancer in free-ranging animals, a venereal cancer in dogs (Yang et al., 1987). Early on, a leading researcher on the dog cancer (Murgia et al., 2006) contacted the relevant Tasmanian Government department offering help but was ignored (Duncan, 2006).

It is hard to escape the conclusion that a desire to keep DFTD research in Tasmania (and for Tasmanians) first and in Australia (and for Australians) second, along with the strong advocacy of one hypothesis by an Australian researcher almost certainly delayed by several years the discovery of the real reason for the lack of an immune response to DFTD.

When Does Transmission Occur?

Although other forms of transmission, such as eating diseased dead Devils or ingesting tumour-carrying fomites left in carcasses, it seems likely that transmission occurs primarily when one adult Devil bites another. The sociality and seasonality of biting, therefore, may help in understanding when transmission occurs.

Sociality of biting

Devils are largely solitary animals, but they come together when feeding on a large carcass and during the mating season. Encounters that occur during the mating season are known or likely to include, in decreasing order of observation-based actuality: mating, mate selection, male sequestering of females and combat between males for females.

Seasonality of biting

The two studies of seasonality in biting among adults indicate there is much variability both between and within years (Hamede et al., 2008: figs 5-6; Hamede et al., 2013: fig. 4). The first study, based on data from three seasons over three combined years on Freycinet Peninsula, reported that the proportion of penetrating bites to the head, the site of origin for virtually all DFTs, was significantly higher in the immediate post-mating season (April) compared to the dispersal (January) and pouch-young seasons (June-July) (Hamede et al., 2008: fig. 6).

The second study, however, based on four seasons over five years at one locality, West Pencil Pine, and over three years at the other locality, Wisedale, could only demonstrate “the number of bites in adult devils was greater at both sites in autumn and winter than in spring and summer, … ” and therefore presumably not significantly different between the post-mating (autumn) and winter seasons (Hamede et al., 2013: fig. 4).

The prevalence of “biting injuries” in adults has been described as “… typically around twice as common during the mating season (March) as in other seasons … (McCallum et al., 2009, citing Hamede et al., 2008).” In the cited reference, however,

6/05/2016 8:41 AM 201 the data referred to include all bites, including non-penetrating bites. Other data in the same paper indicate that the mean number of penetrating bites, that is, those most likely to be involved in transmission, is only 1.2-1.4 times higher (Hamede et al., 2008: fig. 6; see also Hamede et al., 2013: fig. 4 for more data on variability of mean number of bites by location and season).

Biting during the mating season

Despite the remaining ambiguity about the seasonality of biting in Devils, many researchers seem fairly certain the most significant biting for transmission occurs during the mating season (Jones et al., 2007a; McCallum et al., 2007; Hamede et al., 2008; Hawkins et al., 2008; Jones et al., 2008). Some of the claims are quite explicit: “… most injuries to devils occur during the synchronized … mating season (Jones et al., 2007a);” “… the majority of injurious biting … occurs between adults during the mating season … (Hawkins et al., 2008),” “… most penetrating biting injuries occur among adult males and females in the mating season … (Jones et al., 2008),” and “… a peak of biting contact to adults resulting in injuries (and thus probably disease transmission) is associated with the mating season ( Hamede et al., 2008; see also McCallum et al., 2009).

Some of the claims go a step further and suggest the most significant biting for transmission is during mating per se: “… mating [,] which is when transmission is most likely … (Jones et al., 2007a);” “… most of the biting happens at the time of mating (H. McCallum, in Johnston, 2007);” “… much of the total penetrating biting may occur during mating encounters” and “… with the implication that these injuries have occurred during mating encounters (Hamede et al., 2008).”

If biting frequency is higher during the mating season, it may simply be an artefact of greater proximity of adult Devils during this time of year. In this period, they would continue to feed at carcasses and also come together for mating. As a result, bites associated with mating would be additional to the bites associated with social feeding at carcasses, resulting in an overall increase in biting at this time of year. The mating season may indeed see an increase in the frequency of transmission but only because of a possible overall increase in the total number of bites occurring at this time, not mating per se.

There is a logical assumption that the few days of the year a male and female remain together in a mating den “may facilitate transmission (Hamede et al., 2009).” However, data from “proximity loggers” attached to Devils show no significant difference between the mating and non-mating seasons in either the number of encounters or the length of encounters between Devils (Hamede et al., 2009: fig. 3). This lack of difference is striking, because the meagre evidence from both captive and wild animals is that pairs can spend several days secluded in mating dens. Why this behavioural difference failed to show up in the data is unclear.

The only direct observations of the four aspects of sociality during the mating season come from captive animals, and these have to be treated cautiously, as they may reflect highly artificial conditions or may even be contrived. Bearing in mind this caveat, the few published observations of possible male-male combat indicate that

6/05/2016 8:41 AM 202

biting can be directed to the head, the location of virtually all primary tumours, and can be severe (Parer and Parer-Cook, 2004).

Likewise, the few published observations of interactions during mate selection, that is females biting males in rejecting them and males gaining dominance over a female, show that biting is also directed to the head and can be severe (Parer and Parer-Cook, 2004).

The many observations of actual mating suggest the male’s bites are concentrated primarily on the back of the female’s neck and are not particularly penetrating (Parer and Parer-Cook, 2004). Furthermore, primary tumours never seem to appear first on the neck (see “Distribution of Tumours on Body”).

The few observations of male sequestration of females, mainly biting and holding them to prevent their escape, indicate the bites are concentrated on the ears (Roberts, 1915) and neck and appear to be usually non- or only mildly penetrating (Parer and Parer-Cook, 2004). And again, primary tumours never seem to appear first on the neck, and they never occur on the ears.

On the basis of the available evidence, therefore, it appears as if only two of the four of the interactions associated with mating might contribute importantly to transmission during the mating season: mate selection and perhaps to a lesser extent, sequestration. But is the number of bites delivered during these two seasonally brief periods of social interaction high enough to support the inference that the mating season is the critical transmission time?

Biting at carcasses

When two or more Devils feed at the same carcase, they can appear very aggressive. They may vocalise strongly, display with open mouths and make feints. They only infrequently, however, make physical contact by grappling with open jaws and or by biting.

Two sets of observations at set carcasses provide some insight into the frequency of biting in the only situation in which Devils are likely to come together several times throughout the year. Observations at a staked carcass for 16.75 hours over ten nights in March (late mating season) and June (females with pouch young) at a location in northeast Tasmania returned few interactions involving actual physical contact and no instances of head biting (Pemberton, 1990; Pemberton and Renouf, 1993).

In contrast, observations at a staked carcass for 481.5 hours over 47 nights in April (late mating season) and August (females with large pouch young) at seven locations in central and northwest Tasmania returned 147 biting bouts, a bout including “multiple bites in close succession.” In all, there were 0.3 biting bouts per hour or 3.1 per night. Unfortunately, the severity of the biting could not be quantified (Hamede et al., 2008).

If these data can be combined and summarised in any useful manner, it may be that they indicate that from an individual Devil’s perspective, the chances of being bitten on any night are low but on a cumulative basis, perhaps almost inevitable over a year.

6/05/2016 8:41 AM 203

Transmission: Density or Frequency Dependent?

The long-term future of an infectious (or parasitic) agent with just one host is dependent on whether it spreads in a density or frequency dependent manner. An agent that spreads according to the density of infected individuals is likely to go extinct because at some point the density of infected individuals falls so low that the agent can no longer contact a host frequently enough to keep its population from declining.

In contrast, an infectious agent that spreads according to the frequency of infected individuals will only go extinct when its host goes extinct. This is especially the case when the infection occurs in a vital and hence actively sought contact, such as sex. Even when the density of infected individuals is low, healthy individuals will still seek a contact with them.

It can be difficult to determine if a disease is transmitted in a density or frequency dependent manner. Devil researchers, however, came early to the belief the disease spread in a frequency dependent manner (McCallum et al., 2007, 2009; Jones et al., 2008; Bender, 2010; Hamede et al., 2013; Hollings, 2013). This conclusion was based on the increased prevalence of bites both sexes seem to acquire during the mating season and the observation that the disease occurs at high frequencies even in populations with low density (McCallum et al., 2008; Jones et al., 2008; McCallum, 2008; McCallum et al., 2009; Hamede et al., 2013). As noted above, however, there are other explanations for the increased prevalence of severe bites during the mating season, and it remains possible that the threshold density for a density dependent transmission mode is low.

Although most researchers discussed frequency dependent transmission in qualified terms as the most likely mode of transmission, some writers stated categorically this was the means of transmission (Beeton, undated; Bender, 2010; Parkes, 2012, and Jones et al., 2014). Indeed, one group stated the disease is “strongly frequency dependent (Brüniche-Olsen et al., 2013).”

Occurrence of Tumours: Age and Sex

Age

Devil facial tumour disease usually occurs in adults, that is, animals two years of age and older (Loh et al., 2006a; McCallum et al., 2009; Beeton and McCallum, 2011). However, it can occur in animals as young as one year old (DPIWE, 2005a, d; Hawkins et al., 2006; Loh et al., 2006; Newsletter, March 2006; Lachish et al., 2007, 2010; STDP website: “The Disease,” 17 June 2010; Fox et al., 2010b; S. Fox, ABC Rural, Bush Telegraph, 20 September 2011: Deakin et al., 2012: fig. S4; Hamede et al., 2012a: fig. 1a-d; Huxtable et al., 2015), including animals as young as 13 months old (McGlashan et al., 2006: fig. 3; C. Hawkins, in Jones et al., 2007a; McCallum et al., 2007). One survey found that of 147 infected Devils, seven (4.7 percent) were juveniles (Hawkins et al., 2006).

6/05/2016 8:41 AM 204

Some field-estimates of prevalence in one-year olds exceed 15 percent (Hamede et al., 2012a: fig. 1b-d.), and a survey at Fentonbury in early 2007 returned a prevalence estimate of about 39 percent among one-year olds (McCallum et al., 2009: fig. 2b; Hamede et al., 2012a: fig. 1b). In late 2011, a researcher noted that in the Freycinet Peninsula population “a lot of the juveniles [subadults?] … also have the disease (M. Jones, ABC Science Show, 3 September 2011, emphasis added).”

In the well-studied Freycinet population, “… older individuals were the first in the population to be infected and succumb to DFTD; younger adults followed; and subadults only became infected once the majority of adults had died (Lachish et al., 2007).”

Parenthetically, the occurrence of the disease in one-year old animals provides an opportunity to test the idea that transmission occurs during behaviour associated with mating (see: Transmission: Density or Frequency Dependent?). If it is associated with mating, then sexually mature one-year olds should have a higher incidence of the disease than in immature one-year olds.

Sex

The tumour affects both sexes equally (Loh et al., 2006a; Hamede et al., 2009; McCallum et al., 2009). At an early stage of research into the disease, it was thought that males were affected first and then females (Clark, 2003; McManus, 2003; Bevilacqua, 2004b). It was also thought that the disease was more prevalent in males than in females (ABC, World Today, 24 July 2003; Millikin, 2003; Sutherland, 2003; Sydney Morning Herald, 13 November 2003; Owen and Pemberton, 2005). Neither belief turned out to be true, although the assertion that males seem to be first affected persisted on Tasmanian government websites long after it was known to be incorrect (STDP website: “About Devil Facial Tumour disease,” created 20 January 2010, accessed 1 October 2011; DPIPWE website: “Devil facial tumour disease,” published 8 October 2012, accessed 12 December 2012).

Distribution of Tumours on the Body

The Devil’s facial tumours are usually described as occurring on the head, inside of the mouth and the neck (Hawkins et al., 2006; Loh et al., 2006a – in abstract but text indicates only head; DPIW, 2007b; Lachish et al., 2007; Siddle et al., 2007b; Kreiss et al., 2008; Kreiss, 2009; Kreiss et al., 2009a; Murchison, 2009; Bender, 2010; STDP, 2010c-d; Phalen et al., 2013; Holz, 2012; Pearse et al., 2012; Morris and Belov, 2013; Siddle and Kaufman, 2013; Siddle et al., 2013; Howson et al., 2014; Scheelings et al., 2014; Pye et al., 2015c; Siddle and Kaufman, 2015; Woods et al., 2015; University of Southampton media release, “Researchers Aiming to Produce Vaccine to Save the Tasmanian Devil,” 21 July 2015). However, the tumour only seems to involve the neck when a tumour on the head grows into the neck. Indeed, “almost all” primary tumours are now understood to occur on the head (Holz, 2012; R. Hamede and M. Jones, in Hamede et al., 2013). The explicit assertion that “… lesions initially arise on the face and neck…” (Phalen et al., 2013, emphasis added) appears to be in error with regard to the neck.

6/05/2016 8:41 AM 205

Supporting evidence for the statement that “85% of initial [primary] tumours are on the head,” a lower frequency than supported by all other observations (above), can not be found in the only published reference cited (Hamede et al., 2008).

As to the distribution of primary tumours on the head, an early survey found the distribution of single primaries was 62 percent on the oral mucosa and 38 percent on the haired part of the face (N = 26; Loh et al., 2006). A recent survey produced a similar result. At two localities, the percentages of primaries clearly identified as inside and outside the oral cavity were 68 and 32 percent (N = 34) and 63 and 37 percent (N = 32), respectively (Hamede et al., 2013: fig. 3).

Although the primary tumours are said to occur only on the head and neck, Devils bite each other over other parts of the body (AusVet, 2005). For example, observations on 220 Devils trapped in a three-year study in northeast Tasmania, showed that 48.4 percent of wounds and scars were on the muzzle, 28.7 percent on the rump and tail and the rest (22.9 percent) were on the ears, shoulders, legs and back (Pemberton and Renouf, 1993). Observations on Devils in one population feeding at a carcass in winter showed that of the 147 bites recorded 87.8 percent were to the head, 9.5 percent to the body, 2.7 to the tail and none to the limbs (Hamede et al., 2008). Among the adults of one population over three years, the proportion of injuries to the head was 54.6 and 65 percent in males and adult females, respectively; injuries to the body were 30.6 and 21.6 percent, respectively, and injuries to the tail and limb were rare. In subadults, most injuries were to the limbs (52 percent) followed by the head (25.7 percent) (Hamede et al., 2008). And in one population in the central west over three years and nine months, 107 adults had 55 percent of all bites to the head and the remainder to the other parts of the body (Hamede et al., 2013).

In addition to these data, the DPIW apparently has “a very nice data set on wounds of wild devils all over the state” (ABC Online, 1 May 2009), which would presumably give the most comprehensive indication of the number and distribution of wounds over the entire body. An analysis of these data has yet to be published.

Tumours on the body appear to be rare. There is a report of a tumour on the scapula (S. Pyecroft, ABC GNT, 1 September 2004), that is, on the shoulder. It was not said how the tumour was identified, by visual inspection, which is prone to error, or by histology, which is definitive. Two images of DFTs on the body were presented at a public lecture in mid-June 2012 (Sharman, 2012b; image at time 1:13:42). One was on the rump and the other apparently on the base of the tail; both are large in diameter compared to their body positions.

The apparent disparity between the frequency of bite on the body and the occurrence of tumours on the body are intriguing. It suggests that either only the head provides the conditions for the transmitted tumour cells to “take,” or that transmission occurs, perhaps almost exclusively, from bitten to biter (see “Direction of transmission”).

A few Devils have been injected with live DFTD cells on the body, the shoulder and back, during experimental trying to stimulate an immune response. Of seven Devils receiving such injections, six were reported to have developed tumours and one was unreported. Of the six, two were reported to have developed the tumours at the

6/05/2016 8:41 AM 206 injection site and four were unreported (pers. obs. on observations in Brown, 2013). These results also show DFTs can form on the Devil’s body.

The small number of explicit cases of DFTs being induced on the body of Devils (N = 2) suggests that the data already in hand should be examined for other explicit outcomes and perhaps even additional experiments being undertaken. Such experiments might be justified ethically if they were performed on either Devils that were reproductively senescent or diseased to such an extent they would be unlikely to live to the next breeding season. The issue is relevant both for determining if there are constraints on the body surface where a primary can take and for the direction of transmission.

Tumour’s Secondary Effect on Individuals

Although the Devil’s facial tumour affects the animal’s ability to eat, drink, breath and sense their environment, it has surprisingly little effect on other aspects of its biology, at least to the limited extent such effects have been considered.

Morphologically, for example, there appears to be no effect on adrenal function (Keeley et al., 2012a; see also Table 13).

Behaviourally, there is said to be no difference in the movements of healthy and diseased animals (Brandenburg, 2010).

The disease also seems to have little effect on the Devil’s reproduction. In males, the mass of the testis, epididymis, prostate and first and second pairs of Cowper’s glands showed no significant differences between healthy Devils and three classes of DFTD- affected Devils (ANOVA on data in Keeley et al., 2012b: table 4; possible interactive effect seasonal factor excluded due to small sample sizes). There was also no disease effect on seasonal levels in the serum of either testosterone or cortisol (Keeley et al., 2012b: fig. 2). Indeed, a summary view is that “male devil reproduction does not appear to be … inhibited by DFTD (Keeley et al., 2012b).

In females, to judge from the Freycinet Peninsula population, the mean litter size does not differ significantly between healthy (mean = 3.42) and diseased (mean = 3.40) individuals (Lachish et al., 2009).

In fact, the only major influence on the Devil’s biology appears to be on the sex ratio of the young in females. Healthy females have sex ratios that do not differ significantly from parity (above), but diseased females have significantly more females than males (Lachish et al., 2009).

Time to Death

It is difficult to estimate the period between the time an animal is bitten or bites and becomes infected and the time it dies, because it is difficult to infer or observe the critical beginning and end points. It is useful to think of this period as consisting of two parts: the latency period and the infectious period.

Latency period

6/05/2016 8:41 AM 207

The latency period is the time between when an animal is bitten and the time a tumour first appears. There is evidence that the latency period in healthy Devils can be as short as five to six weeks and as long as 15 months. It is generally thought latency periods average about six months (R. Hamede, in Lane et al., 2012). DFTD-affected Devils can develop new primary tumours with two weeks.

Latency periods of three to 12 months are used in models of disease transmission (McCallum et al., 2007; McCallum et al., 2009; Beeton and McCallum 2011; Hamede et al., 2012a-b; Beeton and Forbes, 2012; see also STTDP, 2008a). And quarantine periods for the Insurance Population in zoos are usually 12 months (DPIWE, 2005c; Wade, 2005a; Sinn et al., 2010; Kingston et al., 2014). It needs to be emphasised, however, that the data on the latency period are scanty and often circumstantial.

The shortest latency period is based on an apparent natural transmission involving a Tasmanian zoo Devil (Marooka) that could have been infected around mid-May and then infected another zoo Devil (Nipper) on 26 June in the same year (Andrewartha and Campbell, 2006). Under this” most likely” scenario, the latency period could be within 46 days. This inference was highly circumstantial however, and other scenarios involved longer latency periods. It is worth noting that a very different account of this case, but sighting the official report (Andrewartha and Campbell, 2006), was published in 2008 (Obendorf and McGlashan, 2008). The differences between the two accounts have never been reconciled.

Two Devils injected experimentally with live tumour cells first showed tell-tale tumours after 81 days (11 and a half weeks) in one case (“Clinky,” below; Kreiss et al., 2015; see also Obendorf and McGlashan, 2008) and 154 days (22 weeks) in the other (“Cedric,” below; Kreiss et al., 2015; see also Foley, 2009; UTas media release, 1 September 2010; Woods, 2010). Whether the detected tumours had reached the infectious stage is unclear. If not, their latency periods would be longer than indicated.

A few years later, five other Devils were experimentally infected with 2.5x104 live DFTD cells in either the shoulder or back. In one case, three healthy Devils were injected twice with dead tumour cells that had been stimulated to induce an immune response and then 47 days later injected with the live tumour cells. All three developed tumours at the injection site in 37 days. In the other case, two DFTD- affected Devils were infected with live tumour cells stimulated to induce an immune response. Both Devils developed rapidly growing tumours at the injection site by 14 days (Brown, 2013). As all five Devils were checked daily, all the tumours were probably discovered before they reached a potentially infectious stage.

Other Devils have been injected with live tumour cells and the cells have taken (Pyecroft, 2007b; Pyecroft et al., 2007; Woods, 2010). The latency periods have never been published, however, despite the importance of this parameter in modelling work (Hamede et al., 2012a). These injection experiments probably involved the transmission of many more cells than what would be likely in a natural infection and hence may also provide some insight into the relationship between dosage and latency.

6/05/2016 8:41 AM 208

When a technique for indentifying a pattern of metabolites in the blood of symptomless Devils that would go on to manifest DFTD (see “Test for Disease at a Pre-symptomatic Stage”) was applied to samples taken during a mark and recapture study on the Forestier Peninsula, researchers found several latencies of six months and some of more than six months but all less than 12 months (Gathercole, 2012).

There are two reports of a latency period of ten months. In the first case, the Devil involved was described variously as a “post-weaning” juvenile brought into quarantine as part of the insurance population program (Bevilacqua, 2005a, d; McGlashan et al., 2006) and as a female with four pouch young used in trials to see if the disease could be passed from mother to young (Albion, 2006; see also Andrewartha and Campbell, 2006; McCallum et al., 2007, 2009; PHVA, 2008).

The second case “… was observed in a devil that developed DFTD 10 months after it had been placed into a wildlife park (personal communication from Nolan Fox) (Gathercole, 2012).” The circumstances of this Devil were different enough from the other Devil to suggest there may have been two animals with a ten-month latency period.

The longest recorded latency period is apparently 15 months, but the details are as yet unpublished (Huxtable et al., 2015).

There is a suggestion that the latency period may be “strongly” dependent on the mode of transmission from one individual to another (McCallum et al., 2009). This comment was in reference to a paper published in 2007 (Pyecroft et al., 2007). But the cited work does not support this inference, and no further details have been published

The latency period is also thought to be dependent on how many tumour cells were inoculated and the location of the bite (Jones et al., 2014), but the citation for this supposition (McCallum et al., 2009) contains no such claim.

Infectious period

The infectious period can be considered as the time between when a tumour first becomes obvious and the time the animal dies or at least is last seen. Operationally, however, it has to be realised that a tumour may have to grow large enough to become friable before it becomes infectious and some Devils may become too debilitated to bite effectively before they die. To a certain extent, these two factors may cancel each other out.

Trapping surveys now routinely gather data on the length of time over which an infected marked animal carries the disease. These data can provide a working estimate of the minimum length of the infectious period (Lachish et al., 2007). But no analysis of these data has been published as yet.

The most frequently quoted figure for survival time once a tumour becomes obvious is six months (Dennis, 2006; Lachish et al., 2007; see also Newsletter, March 2006). The most frequently quoted figure for maximum survival time is nine months (Hawkins et al., 2006; Lachish et al., 2007). There is, however, one record of eleven months (Newsletter, March 2006; see also Dennis, 2006) and another for 14 months

6/05/2016 8:41 AM 209

(R. Hamede, on ABC, World Today, 1 January 2009). The provenance of the animals holding the nine month and near one year records are not in the public domain, but the animal still alive at 14 months was from a population showing an apparent immunologically based resistance to the disease (see also M. Jones, ABC Science Show, 3 September 2011). Furthermore, as early as October 2005, a media report said that “several devils have been observed living with the disease for ten to 12 months, raising the possibility that some may be building resistance (Denholm, 2005).” The location of these Devils was not given. And recently, a veterinarian working with the STTDP wrote that “it was originally hypothesised that devils would die in about 6 months but there are indications some infected animals will live much longer … (Wells, 2010).” The provenance of these longer living individuals was not given.

A comment that “affected animals may die as little as 3 months after the first appearance of lesions” (Bender, 2010) is not supported by the reference cited (Hawkins et al., 2006). But recently, a field researcher did say that in eastern populations, “most Devils are dead within three months of a visible tumour arising (M. Jones, ABC Science Show, 3 September 2011).” This is not surprising considering a tumour can grow tenfold in this time (R. Hamede unpublished data, in Hamede et al., 2012b). Even more recently, there was mention of unpublished field data that indicated a tumour could take from six to 12 months to kill its host (R. Hamede, in Hamede et al., 2012b).

It would appear, therefore, that the infectious period, may be anywhere from three to 12 months.

Combining the data for the latency and infectious periods, the time to death, that is, between infection and death, could vary from just under four months (20 days + 3 months) to 24 months (10 months + 14 months). These data suggest that an adult female newly infected with DFTD might just live long enough to have two litters.

Population Decline

In the early days of DFT, before any careful population estimation had been attempted, it was believed that Devil population numbered in the hundreds of thousands (Sunday Tasmanian, 18 February 2001). More considered and researched estimates, however, brought the number down substantially.

Just prior to the outbreak of the disease, the entire Devil population was estimated to have been as high as 100 000 to 150 000 individuals, based on extrapolations from estimated densities in different habitats (Mooney, 1992, 2004; Jones and Rose, 1996; N. Mooney, in Bradshaw and Brook, 2005; M. Jones and N. Mooney in Hawkins et al., 2008: McCallum, 2008; N. Mooney, unpublished, in DPIPWE, 2010d; Newsletter, March 2011). An estimate of 65 000 to 75 000 adults just prior to the onset of the disease (DEWHA, 2009a-b; Beeton, undated) is not to be found in the document cited as the basis of the estimate (Jones and Rose, 1996), but an estimate of 60 000 has recently been suggested (Beeton, 2012, as cited in Hollings, 2013).

On one occasion, estimates of 130 000 to 150 000 Devils pre-disease were criticised as being too high on the basis of overly generous density estimates. No alternative estimate was provided (McCallum et al., 2007), but a figure of “up to 40 000” was

6/05/2016 8:41 AM 210

subsequently reported (attributed to H. McCallum, in Pickrell, 2007). A review article used the estimate of 130 000 to 150 000 (Siddle and Kaufman, 2015), but cited the reference (McCallum et al., 2007) that actually questioned it

In any event, at the outbreak of the disease, the population was probably the largest it had been since European settlement (Owen and Pemberton, 2005) or perhaps ever. Once the disease appeared, however, the numbers declined steadily (Hawkins et al., 2006).

Absolute number decline

At the end of 2004, the population was estimated to be between 60 000 and 90 000 individuals (C. Hawkins, in Hobday and Minstrell, 2008).

In early 2007, the estimate was about 20 000 at a minimum (attributed to H. McCallum, in Byrnes, 2007). In mid-2007, the size was said to be between about 15 000 and 20 000 (attributed to H. McCallum, in Pickrell, 2007). And at some unspecified time this same year, it was give as between 20 000 and 50 000 (DPIPWE unpublished, in DPIPWE, 2010d).

In 2008, the estimate was 10 000 to 25 000 (DPIW, as cited in DEWHA, 2009a).

In mid-2010, the estimate was between 10 000 and 15 000 (attributed to H. McCallum, in Courtney, 2010). In late 2010 it was between 17 000 and 42 000 (DPIPWE, 2010d).

In March 2011, the estimate was between about 17 000 and 42 000 (Newsletter, March 2011). In late 2011, one researcher “guessed” there were about 30 000 Devils left (H. McCallum, ABC Brisbane Online, 8 September 2011).

Despite their high variability, these estimates indicate that is the current population is probably no larger, and in all likelihood appreciably smaller, than at the time of European settlement when it was estimated to have been about 45 000 (see “Early decline”).

Mention should also be made of media reports in early 2011 of only 2000 Devils being left in the wild (Dobbin, 2011; Platt, 2011a; Toy, 2011). Such an estimate might be disregarded as a misunderstanding or mistake had it not reported by three different journalists. The source is unknown.

Percentage declines

Periodically the authorities and researchers estimate the percentage decline of the entire Devil population. In early 2004, it was 27 percent (Hawkins et al., 2008); in early 2006, 41 percent (Hawkins et al., 2008); by 2006, 53 percent (Hawkins et al., 2006; McCallum et al., 2007); early 2007, 53 percent (STTDP website: “Latest spotlighting data,” 21 May 2007); by 2008, 64 percent (C. Hawkins et al. unpublished, cited in Hawkins et al., 2008; McCallum et al., 2009); in March 2009, 70 percent (Newsletter, March 2009); by mid-2010, more than 80 percent, and in February 2011, 84 percent (STDP Annual Report 2010/11).

6/05/2016 8:41 AM 211

Curiously, however, two researchers closely involved with modelling the Devil’s population wrote in March 2011 that the overall population decline was at least 60 percent (Beeton and McCallum, 2011a; see also H. McCallum, on ABC Brisbane, online 8 September 2011) and a lab actively working on the disease could write in September 2011 that only “… 41 per cent of the population had been lost to … [the] disease (Menzies Research Institute, Projects in the area of devil facial tumour disease, 21 September 2011). It is unknown if the disparity in these estimates represent lapses or a difference of opinion about the total percentage decline.

Since late 2011, there seem to have been fewer official comments as to the percentage decline of the Devil population. However, some of the people involved in captive breeding have stated that in 2012 the decline had reached 90 percent or more. The Autumn 2012 Devil Ark newsletter said, “there is now less than 10% of the wild population left.” On 14 September 2012, an article on the Devil Art website (“Devilish exciting times”) said “about 10% of the wild population [was] left of Tasmanian devils.” In May 2013, a Devil Ark spokesperson said that "90 per cent of the wild population in Tasmania is gone (ABC News, 14 May 2013)." And in June 2013, the Director of Wildlife Conservation and Science at Zoos Victoria said there had been a 95 per cent reduction in the population (R. Lowry, on ABC, The World Today, 4 June 2013). It is unclear whether these were guestimates, internal STDP figures or lapses.

Although all affected populations have declined, no local population has yet gone extinct (Decline within Local Populations, below). It is difficult to know, however, how much immigration contributes to keeping the local numbers up. It is also difficult to know if decimated local populations have reached a steady-state or are continuing to decline, albeit more slowly. In any event, the largest declines for local population to date are between 90 (M. Jones, in Pearse et al., 2012) and 95 (Jones and McCallum, 2011b; see also M. Jones on ABC Science Show, 3 September 2011) percent for the Freycinet Peninsula and 97 percent for Mt William (STDP Annual Report 2010/11).

Although it is encouraging that local populations appear to be hanging on, there is concern that the smaller they become, the more susceptible they are to stochastic events, such as bushfires, that could lead to their extinction (Bradshaw and Brook, 2005).

Geographic Spread

Devils with facial tumours were first found in 1996 at Mt William National Park in the northeast part of Tasmania (Mooney, 2004). It is almost certain that the disease arose here and then spread south and west.

The inferred rates of spread vary from about 4 to 52 km/y (Jones, 2003b; Ladds et al., 2003; Marvanek, 2007; McCallum et al., 2007; McCallum, 2008c; STDP website: “Background,” 13 March 2010; DPIPWE, 2010d; Newsletter, December 2010; S. Fox, on ABC News, 9 November 2011; B. Wightman, in Milman, 2013a; Thalmann et al., 2014).

6/05/2016 8:41 AM 212

The disease does not spread evenly along a single front. Its spread has been described as “patchwork” (N. Mooney, in Bevilacqua, 2004b); emerging in some areas but not others (J. Jackson, as reported in Sayer, 2004b), and “… with spots of disease flaring ahead of the wider front (Darby, 2011).” Presumably, this pattern simply reflects the vagaries of Devil movements, but the possibility it represents areas of local resistance or immunity would seem worth considering.

There are also periodic updates on the proportion of the state that is affected by DFTD. The disease was said to have spread across 65 percent of the state as of December 2004 (DPIWE, 2005a-b); 51 percent by mid-2005 (Hawkins et al., 2006; see also Brennan, 2005); 56 percent by the end of 2006 (Brennan, 2006); 51 percent by early 2007 (Lachish et al., 2007); 59 percent by March 2007 (Hawkins, 2007; McCallum et al., 2007); “more than 60 percent” in early 2010 (STDP Website: “The disease,” 17 June 2010; also Menzies Research Institute: Projects in the area of devil facial tumour disease, 21 September 2011); 75 percent at the end of 2011 (STDP Annual Report 2010/11; Newsletter, March 2012) and “approximately two-thirds of the state,” that is, 67 percent in May 2013 (Newsletter, May 2013)..

Estimates of when the disease will have extended throughout the state are quite variable. In 2007, the projected dates varied from 2009 - 2010 (H. McCallum, in Pincock, 2007; Alumni News, November 2007) and 2012-2013 (H. McCallum, in McGlashan et al., 2007; Jones et al., 2007a) to 2012-2017 (McCallum et al., 2007); in 2008, from 2014 to 2019 (McCallum, 2008); in 2009, it was to be in about 2014 (“as little as five years”; McCallum on Radio Australia, Innovations, 18 May 2009; Lachish et al., 2009); in 2011, the year 2016 was implied (H. McCallum, on ABC Brisbane Online, 8 September 2011), and in early 2012, the earliest projected date was 2027 (Lees and Andrews, 2012). These dates indicate that the projected date for the disease to spread across the entire state has been extended with the passage of time, perhaps suggesting that the rate of spread has been slowing.

Estimates of the possible extinction dates in the wild have varied between 2017 and 2040, specifically in chronological order: 2017-2022 (ABC News, 4 March 2007; H. McCallum, in McGlashan et al., 2007), 2027 (Jones et al., 2007a; H. McCallum, in McGuirk, 2007; Paine, 2007); 2026-2031 (H. McCallum, on ABC Brisbane Online, 8 September 2011); 2027-2032 (H. McCallum, in Edwards, 2007); 2032 (STTDP, 2007a); 2030-2040 (H. McCallum, on ABC PM, 1 September 2010); 2029 -2049 (H. McCallum in Naidoo, 2009), 2034-2044 (“approved conservation advice” reported in Ford, 2013a-b) and 2037-2042 (Lees and Andrew, 2012).

Decline within Local Populations

In 2007, it was estimated that local populations would go extinct about 15 years after the arrival of the disease (McCallum et al., 2007; Jones et al., 2007a). In 2011, it was suggested that local extinctions could occur in one to two decades (Beeton and McCallum, 2011). To date, however, no population of Devils is known to have gone extinct (STDP Annual Report 2010/11; STDP spokesperson, W. Brennan, in Burgess, 2010; S. Fox, in Darby, 2011; STDP pers. comms, in Hogg et al., 2015, 2016; Huxtable et al., 2015; STDP pers. comm., in Wright, 2015). The pattern of decline seems to be a decaying exponential, that is, an initial steep decline grading into ever slower rates of decline (McCallum et al., 2009: fig. 3), leaving it to be determined

6/05/2016 8:41 AM 213

whether the declines continue to approach zero ever more slowing or reach and maintain a low plateau.

Periodic sampling of long-studied local populations appears to be continuing (Hawkins et al., 2006 and below), but oddly, the cumulative results have not been published for years beyond 2008.

Three DFTD-affected populations have been surveyed in both pre- and post-disease periods. These three populations, Mt William, the Freycinet Peninsula and Fentonbury are, therefore, the most illustrative of the disease’s effect on population size.

Mt William

The first recorded observation of Devil’s facial tumour disease was in 1996 at Mt William in northeast Tasmania, making this population, which is still extant, the oldest known diseased population. It thus forms, arguably, the most important site for monitoring the disease. Fortuitously, the site was studied extensively between May 1983 and August 1985, that is, more than 11 years before the disease was first noticed (Pemberton, 1990). The study made two potentially useful estimates of population size: absolute numbers and density.

Using three different survey methods, the study estimated population size in absolute terms ten times between May 1983 and August 1985. Not surprisingly, the different methods returned estimates that varied among themselves in any one survey period, and each method picked up seasonal variation. Each method returned a more than 3.8 fold difference in each estimate over the entire survey period, and all methods considered together suggested a maximum total population size, in May, of between 200 and 400, with 300 taken as the one representative number.

The study estimated the population density as being between 2.2 and 7.4 individuals per km2, a more than 3.3-fold difference.

The study did not make clear, however, the size or geographic boundaries of the area to which the population estimates applied. Subsequent surveyors wishing to compare their surveys with the original survey, therefore, would have to follow the original trapping procedures and protocols (Pemberton, 1990) if they wanted strictly comparable results. This seems to never have been done.

The later mention of the geographical size of the study area of the original study throws up a question about the density estimate in the original survey, which, as noted, did not mention the geographic size of the study area. The subsequent report, of which the original researcher was a co-author, gave the trapping area as 5.5 km2 (Hawkins et al., 2006). But using this area and the lowest estimated mean absolute population size, about 100, gives a density estimate of about 18 individuals per km2, a figure more than twice the size of the study’s maximum density estimate, 7.4.

The Mt William population was surveyed post-disease between mid-2004 and mid- 2007 (McCallum et al., 2009: fig. 5; McCallum, 2012: fig. 3). These data, reported as density estimates, show a steady decline from about 0.6 individuals per km2 in the

6/05/2016 8:41 AM 214

period mid-2004 to late 2005 and then, arguably, a levelling off over the remaining period at about 0.4 individuals per km2. It is unclear whether the population continued to be surveyed under the same protocol after mid-2007, but it seems unlikely as no such a survey has been mentioned subsequently. In comparison to the range of densities reported in the original survey, the density at the end of the survey period represents a reduction of between 82 and 95 percent. It should be noted, however, that the survey methods were almost certainly different from the original, pre-disease survey.

They are many estimates of the decline in the Mt William population since the original survey, but none has ever been accompanied by an explanation of the data in the original survey upon which they are based. For example, one group of researchers used an original density figure of 7.14 individuals per km2 (note the two decimal place accuracy), but gave no explanation how they chose this relatively high figure (which would have exacerbated the estimated decline) from within the original overall range of 2.2 to 7.4 (Lachish et al., 2007).

In addition to the plot data, there are several free-standing comments about the percentage decline in the post-disease period in the Mt William population. On the basis of absolute numbers (but which original number was use?), the decline was said to be about 75 percent in August 2004 (Hawkins et al., 2006; McCallum et al., 2007); about 87.5 in July 2006 (McCallum et al., 2007 on the basis of an absolute number credited to Lachish et al., 2007 but not occurring in that work), and about 90 percent in February 2007 (Hawkins, 2007).

On the basis of density (again, which original number was used?), the decline was said to be about 97.5 percent in early 2007 (data in Lachish et al., 2007; above).

In July 2010, a report stated that “the population at Mt William National Park in the north-east of the State has declined by 94% according to annual spotlighting surveys (Fox et al., 2010a; see also Newsletter, May 2013). This assertion is surprising, because maps of the spotlighting localities throughout the state (DPIPWE, 2013: fig. 1; also see Hollings et al., 2014: fig. 1) show none anywhere close to Mt William.

On no stated basis, the decline was given as at about 90 percent at the end of 2006 (Darby, 2006); about 94 percent by March 2009 (Newsletter, March 2009); remaining about 94 percent in early 2010 (W. Brennan, in Burgess, 2010), and reaching 97 percent by the middle 2011 (STDP Annual Report 2010/11).

Despite what has clearly been a big decline in the Mt William population, it was still extant in 2015, 19 years after the first case of DFTD was seen. This is four years longer than the initial 2007 estimate of 15 years for local population persistence (McCallum et al., 2007) and two years longer than the STTDP’s then Senior Scientist expected this particular local population to be extinct (McCallum, 2008c; see also S. Thalmann, in Brand Tasmania Newsletter, August 2011). Indeed, in October 2015, a Devil researcher declared noted that this population had “stabilised by 2008 (Jones, 2015b).”

6/05/2016 8:41 AM 215

Despite the apparent accuracy of these estimates, there is nothing in the public record to indicate that the Mt William population has apparently never been re-surveyed in a manner that allows robust comparison with the original survey. This is odd, considering the importance of the site in monitoring the disease.

Freycinet Peninsula

The population in the upper part of the Freycinet Peninsula on the east coast of the state is unusual in that it is one of only two populations that were being surveyed when DFTD arrived. Surveying began in mid-1999 and continued until at least mid- 2008. DFTD was first detected in the population in June or July 2001, depending on accounts (Lachish et al., 2007, 2009, 2010; McCallum et al., 2007; see also Brüniche- Olsen et al., 2013).

There are six plots of population size for this population covering part or all of the survey period. Three plots are for the whole population, two based on total numbers (Lachish et al., 2007: fig. 5; McCallum et al., 2007: fig. 3) and the other based on density (McCallum et al., 2009: fig. 5). There are also three plots based on total number of adults (Lachish et al., 2007: fig. 5; Lachish et al., 2010: fig. 3c; Hamede et al., 2012a: fig. 3c).

The two plots of total population size based on absolute numbers cover the period between 1999 and 2006. The two plots do not correspond exactly despite having been apparently generated in the same way, but in general they show an asymptotic decline to about half the pre-disease numbers (Lachish et al., 2007: fig. 5; McCallum et al., 2007: fig. 3.; 2009: fig. 5). The plot of the density of the total population between early 2004 and mid-2007, which covers only a post-disease period, shows a similar levelling off at about 0.4 individuals per km2 (McCallum et al., 2009: fig. 5). This is a level about 40 percent of the pre-disease level of about one individual per km2 (M. Jones, on ABC Science Show, 3 September 2011).

The three plots of the absolute number of adults correspond in general shape, but one plots smaller numbers of each survey period (Lachish et al., 2007: fig. 5). The most comprehensive of the three plots shows a gradual decline in the population with a possibility of approaching an asymptote at about 5 to 10 percent of the pre-disease level over the last three survey periods (Lachish et al., 2010: fig. 3c).

A population biologist working on Devils interpreted the data as showing “the number of adults [was] halving every year,” a rate which would lead this or any other similar population to extinction in 10 to 15 years after disease arrival (McCallum, 2008c).” For the Freycinet population, this suggests extinction would occur between 2011 and 2016. As of 2015, the population is apparently still extant.

The only information about population size since mid-2008 are a few general comments indicating the total population has fallen to between ten (M. Jones, in Pearse et al., 2012) and five (Jones and McCallum, 2011b; see also M. Jones on ABC Science Show, 3 September 2011) percent of its pre-disease levels or from about 150 to ten or fewer (M. Jones, on ABC Science Show, 3 September 2011).

6/05/2016 8:41 AM 216

The primary researcher at the Freycinet site has said that “where the disease went through, within 12 to 18 months we lost 100% of the resident adults living in those areas (M. Jones, on ABC Stateline Tasmania, 23 September 2005; see also ABC Catalyst, 19 October 2006). The population graphs for the area, however, do not show this level of adult mortality (Lachish et al., 2007: fig. 5; Lachish et al., 2010: fig 3c).

Considering Freycinet is the only site to have been surveyed both pre- and post- disease using the same protocol (Hawkins et al., 2006), it is regrettable that regular surveying seems to have ceased in mid-2008. In November 2015, however, the principal researcher at this site noted that the population had “stabilised by 2008 (Jones, 2015b).” She also asserted that she had been studying the population for 15 years (since 1999). If this involved regular surveys, the results have yet to be published.

In 2012, there were apparently plans to turn the peninsula into a large free-range enclosure (ABC News, 13 September 2012; Darby, 2012; Denholm, 2012f; Lord, 2012b). The status of these plans is unclear (see: De-populate/re-populate, below), but if they were to proceed, one of the best sites ever for monitoring the course of the disease (and testing predictions about it) would be irrevocably lost.

Fentonbury

Fentonbury is in the southcentral part of the state is the second of only two sites where monitoring was in progress when DFT was first found. Surveying began in June 2004 (STTDP website: “A facial tumour that appeared to shrink in an “interesting” devil,” 17 August 2007) and was carried out three times a years until late 2007. DFTD was first detected in early 2005. A plot of the density of individuals in the population between mid-2004 and late 2007 (McCallum et al., 2009: fig. 5a; McCallum, 2012: fig. 4) shows a near two fold increase prior to the arrival of DFTD and then a decline post-arrival with such variability at the end of the survey period that an interpretation of stabilisation could be supported. Note that the two plots referred to are identical, but the former is unqualified as to age class of the individuals whereas the latter is said to be for adults. A plot of the number of adults in the population between late 2004 and late 2007 shows the population declining steadily after the arrival of DFTD, with a possible approach to stability over the last two sampling periods (Hamede et al., 2012a: fig. 3b).

Narawntapu National Park

Although information is scanty, it appears Narawntapu National Park was surveyed in the pre-DFTD period from 1999 to 2006, the year before DFTD arrived. During this period the population was said to been constant around 80 individuals (“M.J. unpublished data”) (Brüniche-Olsen et al., 2013). It is not clear what surveys, if any, were done in intervening years, but in 2014 when the STDP said it started long-term trapping surveys at ten sites, 19 Devils were trapped at Narawntapu (STDP Annual Report, 2013/14: fig. on p. 13; de Salis, 2015). Thus after eight years of DFTD, the population had fallen by 75 percent. Whether the population has stabilised at this level is unknown due to lack of data in the public domain.

6/05/2016 8:41 AM 217

Four other populations, Bronte Park, Buckland, Forestier and Wisedale, were surveyed for a period after the disease had become established.

Bronte Park

Bronte Park is in central Tasmania. A photograph of an apparently DFTD-affected Devil at Bronte Park prompted monitoring at this site (DPIWE, 2005a). The site was surveyed 11 times between early 2004 and mid-2007. Between January and October 2004, the adult population declined from an estimated 51 to nine, an 82 percent reduction (DPIWE, 2005a). A plot of the total population density estimates deriving from the entire survey period, shows the density declined from about 1.0 per km2 to about 0.5 km2 or about 50 percent in the period between January and October 2004 and continued to decline until about mid-2006, then possibly increased to mid-2007 the last data point (McCallum et al., 2009: fig. 5). Overall, the plot could be interpreted as showing an initial decline followed by a stabilisation at a density of about one per 0.4 km2.

Buckland

Buckland is in southeast Tasmanian. Devils with DFTD-like physical symptoms were first discovered at the site in 1999 (M. Jones, unpublished data in McCallum et al., 2009). The locality was surveyed five times between early 2006 and late 2007. A plot of the population density can only be described as stable throughout the entire period at about 0.5 individuals per km2 (McCallum et al., 2009: fig. 5).

ForestierPeninsula

Forestier is the first of two tandem peninsulas on the east coast. DFTD was discovered in the area in June 2004 (Lachish et al., 2010; see also Brüniche-Olsen et al., 2013). The c. 150 km2 area was surveyed 11 times between late 2004 and the second half of 2008 (Hamede et al., 2012a). Although diseased Devils were culled in this period, the culling may have had little effect on the population size as it simply brought forward an inevitable mortality.

The only available plot of the number of adults in the population begins at about 120, falls over the next two sampling periods to about 80, then jumps to a high of about 150 then falls over the next four periods to about 45 and remains stable over the last three periods (Hamede et al., 2012a: fig. 3d). The population of adults, therefore, seemed to stabilise at about 30 percent of its highest recorded value (45/150).

The only density estimate available was about 1.6 per km2 at the beginning of 2007 (Jones et al., 2007a). This is interesting, because two and a half years after the arrival of the disease, the population density was considered “high” (Jones et al. 2007).

This population will be lost to further monitoring - if any monitoring is still occurring, when the population is removed, the diseased animals killed and healthy animals re- introduced (see “De-populate/re-populate” below).

Wisedale

6/05/2016 8:41 AM 218

Wisedale is in the northcentral part of the state. Surveying of the population began in May 2006, shortly after the first appearance of the disease in early 2006, and continued through eight trapping periods until February 2008 (McCallum et al., 2009: fig. 5; Hamede et al., 2013: figs 4-5). An early plot of the density of the population fluctuates around 0.9-1.1 (range: about 0.5-1.2) individuals per km2 throughout the survey period (McCallum et al., 2009: fig. 5). Oddly, however, a later plot of density data for Wisedale shows over the three years a slight linear decline from about 2.0 to 1.8 in individuals per km2 (Hamede et al., 2013). These densities are about twice those in the earlier report. It is unclear which estimate is the more accurate.

The plots of population size for all the populations discussed above extend no further than late 2007 or late 2008. It is not clear whether this is due to a cessation of monitoring or of reporting. In 2005, soon after the monitoring program was underway, the Tasmanian government stated “it will be vital to maintain intensive long term monitoring at replicated sites (DPIWE, 2005a).

In mid-2011, the Tasmanian government indicated that it had three long term monitoring sites in 2009/10 and 2010/11 and was planning six such sites in 2011/12 (DPIPWE Annual Report 2010/11). In fact, it ran five sites in 2011/12 (DPIPWE Annual Report, 2011/12, 2012/13 and 2013/14; STDP Annual Report, 2011/12). In mid-2011, the STDP mentioned it had three long-term monitoring sites (STDP Annual Report 2010/11) and in mid-2012, it mentioned five such sites (STDP Annual Report 2011/12). It foreshadowed ten such sites for 2012/13 (STDP Annual Report 2011/12) although its parent government department foreshadowed only six (DPIPWE Annual Report 2011/12). In mid-2013, the DPIPWE reported there had been only four long-term monitoring sites in 2012/13, due to funding uncertainty, and it planned five sites for 2013/14 (DPIPWE Annual Report, 2012/13). It also mentioned, for the first time, that the monitoring method was trapping. In mid-2013, the STDP said there had been three long-term monitoring sites in 2012/13, and it planned six sites for 2013/14 (STDP Annual Report, 2012/13). In October 2014, DIPWE said it had eight sites in 2013/14 and expected to have the same number in 2014/15 (DPIPWE Annual Report 2013/14). In December 2014, however, the STDP said it had 10 long-term monitoring sites in 2014/15 and planned for the same number of sites through to at least mid-2019 (STDP, 2014c).

None of the above reports named the long-term monitoring sites, but in November 2014, the STDP named eight sites: Bronte Park, Buckland, Fentonbury, Freycinet, Granville Harbour, Kempton, Mt William, Narawntapu, Takone and Woolnorth (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014). The STDP, however, did not say whether monitoring had been continuous at some of the “old” sites, such as Fentonbury, or when the monitoring began at the new sites such as Kempton. Trapping is apparently the method of observation (DPIPWE Annual Report 2013/14), but it was unclear whether the current trapping protocols allow comparison with the earlier protocols.

Much of the uncertainty and ambiguity in the previous reports was clarified in the STDP’s Annual Report for 2013/14 (published July 2015). The report said the monitoring program began in 2014 and would run for five years. It would cover ten sites, eight of which were the DPIPWE’s responsibility and two the University of Tasmania’s (STDP Annual Report, 2013/14). The DPIPWE would monitor Bronte

6/05/2016 8:41 AM 219

Park, Buckland, Fentonbury Kempton, Narawntapu, Takone, Granville and Woolnorth and the university, Mt William and Freycinet. Each site would be trapped for seven nights using 40 trap sites twice a year (STDP Annual Report, 2013/14). Eight of the sites are old sites and two new, one of the latter of which could be inferred to be Kempton. Some sites were chosen “to cover sites for which there is historical data to compare the trend across time;” presumably these are some or all of the “old” sites. Finally, the approximately six-year gap in monitoring some if not all the old sites (above) was apparently real.

Infra-red sentinel cameras have been deployed in DFTD monitoring since at least July 2004 (ABC News, 2 August 2004; see also DPIWE, 2004b) in order to keep track of populations with low densities or in difficult-to-access areas (STTDP website, “Remote sensor cameras,” 6 February 2007, as accessed on the LINC Tasmania website). The cameras were first deployed at Bronte Park in the Central Highlands (ABC News, 2 August 2004).

In early 2010, the STDP formally moved to infra-red sentinel camera for more routine monitoring (DPIPWE Annual Report 2010/11; STDP website: “Remote sensor cameras,” 22 February 2010, “Population monitoring,” 4 August 2011). One advantage of the cameras is they may “capture” the 25 percent of the population that appears to be trap-shy. Cameras, however, make it more difficult to be sure of the identity of individual animals, their age and sex, the presence of tiny external tumours or larger mouth tumours and the size of the tumours.

Unfortunately, there appears to have been no field trial that provides for a comparison of the old trap capture-release-recapture monitoring with the new camera monitoring.

If the STDP has been monitoring the sites discussed above in a comparable manner, it has to be asked why have they not released up-dated plots of the population size at those localities? Or if they ceased monitoring, why did they do so? This lack of information is especially frustrating as it coincides with what in many cases look like the beginnings of population stabilisation.

In 2012, the first independent reviewer of the Save the Tasmanian Devil Program pointedly said: “monitoring the fate of diseased populations is, in my opinion, the critical unknown without which no effective and rational STDP can be planned (Parkes, 2012).” Why the reviewer did not enquire specifically about such monitoring at the sites mentioned above is unclear.

Annual Statewide Spotlight Surveys

Annually since 1985, the Tasmanian Government has been conducting standard spotlight surveys along sections of road, each 10 km in length, throughout the state between 15 November and 31 January. The surveys actually began in 1975 but were not standardised, making them broadly comparable, until 1985. Between 1985 and 2001, 132 sections were survey and between 2002 and present, 172 sections have been surveyed (Hocking and Driessen, 1992; Driessen and Hocking, 1992; DPIPWE, 2013).

6/05/2016 8:41 AM 220

The results, as a graph, were first made public as part of the “State of the Environment Tasmania” for 2009 but only for the years 1985-2005. In mid-2013, the complete raw results for 2002 – 2012 were published online (DPIPWE, 2013: p. 9 of appendix 1; the document’s “Properties” indicate the document was created on 26 July 2013; see also graph on p. 4 of STDP Annual Report, 2012/13). In March 2015 in response to a request, I received the complete raw data for 1985-2013. In May 2015, the complete raw data for the years 2002 -2014 were made published online (DPIPWE, 2015b; created 26 May 2015). The results based on the raw data received from DPIPWE (see Acknowledgements) and in DPIPWE, 2015b are reproduced here as Fig. 17.

Mean Number of Devils per 10 Km Transect by Year, 1985-2014

0.90

0.80

0.70

0.60

0.50

0.40 Transect 0.30

0.20

Mean Number of Devils per 10 Km 0.10

0.00 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Year

Fig. 17. Plot of the mean number of Devils observed per 10 km road transect by year for the years 1985-2014, inclusive (data for 1985-2001 based on c. 132 transects from G. Hocking, DPIPWE and data for 2002-2014 based on c. 172 transects from DPIPWE, 2015b).

There are several interesting things about these data. First, it seems likely that most official estimates of the Devil’s decline throughout the state (see “Percentage decline,” above) are based on these data.

Second, and more importantly, the data show that the apparent stabilisations that were evident in several local populations just as the regular surveys of these populations were being curtailed, or at least their results being published, in 2007/2008 were in fact indicative of a more widespread stabilisation.

Third, the STDP was slow to acknowledge the possible stabilisation of these populations. In July 2012, one of the STDP’s primary researchers said in a public lecture that the surveyed eastern populations “… seem to have levelled out at a very low level …” and suggested that it might be due to dispersal, presumably from areas

6/05/2016 8:41 AM 221 of higher remaining density. But she also thought this might only be able “… to maintain those populations for just a little bit longer (Jones 2012).”

Another indication of the official thinking about the stabilisation of these populations came on New Year’s Day 2013, a day renowned for the public’s lack of interest in the news, when a media story noted that “populations in areas where the disease first appeared in 1999-2002 have unexpectedly stabilised (Mounster, 2013a, d; emphasis added).” The substance of this story and the basis for this sentence could only have come from a source within the STDP. Further, in February 2013, the STDP’s Director said, “… what we’re seeing is that populations are being far more persistent (H. Williams, in Nogrady, 2013, emphasis added).”

Even the public seems to have become aware of a stabilisation, possibly even an increase, in Devil numbers locally. On 16 December 2013, the proprietor of a “farm stay” adjacent to Mt William National Park, where the disease was first recorded, said “We’ve been seeing, I think, more and more of them [Devils] (L. Manley, on ABC Northern Tasmania, 16 December 2013).” Yet in September 2013, the STDP still maintained on its website, “There is no evidence to date of the decline in devils stopping or the prevalence of the disease decreasing (STDP website, “The Disease,” 29 September 2013).”

Eventually, however, the tide of events prevailed. In its 2012/13 annual report, published in March 2014, the STDP said, “the Program now knows that the Tasmanian devil is persisting in its range across the state” and consequently, “throughout the last 12 months, the Program has responded to the challenge presented by this unusual situation by developing a management strategy to address two possible scenarios - will devils in the wild become extinct (and the disease along with them), or will the devil and the disease persist in the wild (emphasis added)?” (see also, Pemberton, 2014; STDP Annual Report, 2013/14). At last, the STDP had admitted the “unexpected” might be occurring: the Devil might live on in the wild – without [despite?] the Save the Tasmanian Devil Industry.

Further to the probable stabilisation of the Devil population, in November 2014, a retired but still active Devil field worker said, “the wild population is still going down, but in the areas the disease has been the longest, population it [sic] appears to have stabilised to very, very low levels" (N. Mooney, in Day, 2014). And the minister for the environment said in a November 2014 press release that “Our best estimates are that the number of devils in the North East have remained fairly constant over the course of the last decade (Groom, 2014, emphasis added).”

In its new five-year plan published in December 2014, the STDP wrote, “the long- term (10+ years) goal is to improve the conservation status of the Tasmanian devil and maintain its role in the Tasmanian ecosystem through stabilising and recovering the devil population (STDP, 2014c).” It appears, therefore, that even though the Devil population appeared to have stabilised entirely through natural processes and having done so might expected to begin a recovery, the STDP could not resist intervening.

In July 2015, the STDP put the first numbers on the expected stable size of local populations, when the Program Director said, "It's inevitable that it [DFTD] will move

6/05/2016 8:41 AM 222

across the lot [all of Tasmania], but it also appears that devils will persist among it with around 10 to 20 per cent of the population (Darby, 2015a)."

A paper published in March 2015 said, “relatively consistent, very low devil numbers with reduced disease prevalence, have been trapped in areas where DFTD has long been present,” (Fancourt et al., 2015, citing a STDP pers. comm.; emphasis added). Hence not only have some populations stabilised, the disease prevalence has also apparently declined.

In July 2015, the STDP Program Manager seemed to provide the first confirmation that the worst was over when in the title of an abstract for an oral presentation he indicated his organisation had now entered the “post-epidemic” management phase (Pemberton, 2015).

In September 2015, the head of the Insurance Program said, “we still have not seen any localized extinction. They are persisting in very low populations (C. Hogg, in Mester 2015; emphasis added).”

In December 2015, there was even a comment that could be interpreted as indicating at least some Devil populations were on the increase, an experienced field worker said, “We are getting anecdotal reports that people are seeing devils again (M. Jones, in Marris, 2015).”

Prevalence in Local Populations

Within populations, it is clear that the prevalence of the disease increases in the first few years after the arrival of DFTD (McCallum et al., 2009: fig. 2; Hamede et al., 2012a: fig. 1). It is less clear, however, what happens to prevalence levels as the years go by, because as the number of older individuals declines, the confidence intervals of the estimates increase (see McCallum et al., 2009: fig. 2; Hamede et al., 2012a: fig. 1; McCallum 2012: fig. 4). Furthermore, there are no prevalence estimates in the literature for any time later than 2008 or a maximum of eight years after the disease arrived in a monitored population. Intriguingly, however, an article co-authored by two senior Devil researchers said “Once in a population, disease prevalence increases to at least 50 percent in adults and … is maintained at that level despite massive population declines (Lunney et al., 2008).” The species’ draft recovery plan (DPIPWE, 2010d) states, “in infected areas the disease prevalence increases to 30– 50% of devils in four years, and is maintained at that level (McCallum 2008b – an apparent lapse for Lunney et al., 2008).” Unfortunately, the populations and their “maintained” prevalence levels are not specified.

DFTD’s Effect on Population Structure, Genetics and Behaviour

Population structure

The arrival of tumour facial disease in a new area can have profound consequences for the local Devil population. The most obvious affect is a reduction in population size. This may be as much as 97.5 percent (Lachish et al., 2007), although no local population has gone extinct as yet (DPIPWE, 2010d; de Salis, 2015; STDP pers. comm., in Hogg et al., 2015).

6/05/2016 8:41 AM 223

There is also a profound change in population age structure with a severe reduction in the number of animals older than two years (DPIWE, 2005; Jones et al., 2008; Lachish et al., 2009).

There is also a profound shift in age at first breeding, with an increase in the number of females breeding at one year of age (ABC Stateline Tasmania, 9 October 2004; Bevilacqua, 2004b; Lachish et al., 2007; Jones et al., 2008; Lachish et al., 2009: fig. 4). Pre-disease, between 0 and 12.5 percent of one-year old females might have bred in a population. But post-disease this can rise to as high as 83 percent in some populations, such as Mt William, where the disease was first noticed and has presumably been established the longest. The youngest age at which a female in an affected population has been observed breeding is 14 months (Jones et al., 2008). Indeed, it appears as if a young female can breed within three months of being weaned, if she can reach a certain minimum body weight (Jones, 2012). The precise minimum weight, however, has not been reported.

This shift to earlier breeding is thought to be due to an increase in resources brought about by the smaller local population size instead of an evolutionary shift (Jones et al., 2008). Furthermore, breeding one year old females are probably faster growing and definitely slightly larger than non-breeding one year old females, suggesting that more rapid attainment of some size or developmental threshold contributes at least in part to their earlier breeding (Lachish et al., 2009).

Parenthetically, a major work on the Devil makes a confusing claim about reproduction due to an apparent typographical error. It says as a result of the disease, “males are breeding when much younger – so taking the place of older males dying while in their mating prime – but as a result are themselves dying after their first and only litter is weaned (Owen and Pemberton, 2005). This passage only makes sense in the context of what is actually known about Devils, if “males” is a lapse for “females.”

There was also an early report that at least at one site, Devils had extended their breeding season. Pre-disease, Devils at Mt William bred in a three-week period in March and April. But post-disease, the Devil researcher most familiar with this site, was reported as having found evidence that the animals were now breeding throughout the year (D. Pemberton, as reported by ABC Hobart, ABC News and Agence France Press, 21 July 2008). But nothing further has been published.

Arrival of the disease also leads to a decrease in female dispersal in terms of both distance moved and frequency of movement. This is probably due to less competition for local resources. There is, however, no significant change in male dispersal (Lachish et al., 2011a).

Some population parameters appear to be unaffected by the disease. For example, the sex ratio was not significantly different in three east coast populations (combined) before and after the disease had become established (P = 0.16, Fisher Exact Test; data in Lachish et al., 2011a), each sex ratio (before = 68/78 and after = 81/65) being not significantly different from parity (see also Lachish et al., 2007).

6/05/2016 8:41 AM 224

Population genetics

The disease also appears to have caused significant changes in the frequencies of some alleles, which suggests that evolution has occurred (see “Genetic evolution”). For example, in three east coast populations, one allele had a frequency > 0.34 pre- disease, but this declined to < 0.015 two to three generations post-disease (Lachish et al., 2011a). The function of the alleles is unknown; hence, it is unclear if the change is adaptive or not. Other workers concluded these shifts were more likely due to drift than selection (Brüniche-Olsen et al., 2016).

The disease was originally thought to have led to a greater genetic differentiation among populations (Lachish et al., 2011a), but subsequent work suggested this may not be the case (Brüniche-Olsen et al., 2013).

There has also been ambiguity about whether the disease has led to loss of genetic diversity within populations. Early work suggested there had been no significant loss of genetic diversity within populations (Lachish et al., 2011a; Brüniche-Olsen et al., 2013) and “… populations with and without DFTD all had similar genetic diversity— a pattern that remained unaffected over time in DFTD populations (Brüniche-Olsen et al., 2016).

In contrast, there were indications, as yet unpublished, there had been a “reduction in genetic diversity (Pemberton, 2015),” and the Narawntapu National Park population, specifically, had “undergone significant loss of genetic diversity (Dingwall, 2015; Lohberger, 2015b).” There were also related claims some populations were becoming inbred (Darby, 2015c).

The effective population size showed a slight but not significant decrease at three sites assessed pre- and post-DFTD: Freycinet, Little Swanport and Pawleena (Brüniche- Olsen et al., 2013).

There was some early genetic evidence for an increase in inbreeding within populations, which would have been an expected due to the smaller population size and reduced dispersal (Lachish et al., 2011a; STDP Annual Report 2010/11). This result, however, may have been due to the pooling of three populations in the analysis (Brüniche-Olsen et al., 2013).

A subsequent analysis of genetic samples from five populations collected in 1999 (all without DFTD) and 2006 or 2007 (two then with DFTD and three still without DFTD) found a consistent increase in inbreeding in all five, but the difference was significant in only three and only one of these was DFTD affected (Brüniche-Olsen et al., 2013) (Brüniche-Olsen et al., 2013). This study also suggested that comparisons of the genetics in populations before and after DFTD might be premature due to the small sample sizes and few generation turnovers in the period during which DFTD has been in local populations.

Among females, the proportion of adults (≥ 2 y) that breed remains high, varying between 81 and 100 percent (Lachish et al., 2009: fig. 4), which is even higher than in most pre-disease populations (above).

6/05/2016 8:41 AM 225

Survivorship, while strongly depressed, does not differ between the sexes (Lachish et al., 2007).

Finally, both healthy and diseased animals are trapped at similar rates (McCallum et al., 2009), suggesting that the foraging activity in the two groups remains similar.

Population Behaviour

In 2015, comments began appearing suggesting that at least in some populations, Devils were becoming less aggressive toward one another. The mechanism for this was thought to be that less aggressive Devils might have a greater chance of not catching the disease (Marris, 2015).

A Second Type of DFTD?

In December 2015, researchers announced they had found a second form of Devil Facial Tumour, designated DFT2, in five animals (but subsequently found in three others) from the D’Entrecasteaux Channel area of far southeast Tasmanian (University of Cambridge media release, “Second contagious form of cancer found in Tasmanian devils,” 28 December 2015; Pye et al., 2016b: fig. 1 but published online, 28 December 2015; MedicalResearch.com, 21 January 2016). The first animals with the disease were collected in December 2012, but the oddity of their disease was first noticed in March 2014 in one animal (Young, 2016).

In gross aspect, DFT2 lesions are similar to those of DFT1 and like DFT1 lesions, occur only on the facial skin and in the oral cavity. DFT2 is also similar to DFT1 in metastasising to other parts of the body, such as the lymph nodes, kidney and lung. DFT2 has a uniform cytogenetic and genetic profile that differs in both regards from its hosts (which differ among themselves), suggesting that like DFT1, the cancer is a transmissible clone.

DFT2 differs from DFT1, however, in a number of ways. Histologically, it presents as consisting of “amorphic to stellate and fusiform cells” distributed in broad sheets instead of round cells in discrete bundles (Pye et al., 2016: fig. 2), and it does not stain for periaxin. Cytogenetically and genetically, it is very different from DFT1. Cytogenetics and genetics also show that DFT2 arose in a male, not a female as in DFT1.

It has yet to be determined what cell line gave rise to DFT2; how it is transmitted and why it is not detected by the Devil’s immune system.

It is has also yet to be determined what relationship DFT2 may have with DFT1. That DFT2 only appeared as DFT1 was entering that part of Tasmania, suggests there may be a connection.

Conservation

Conservation Status

6/05/2016 8:41 AM 226

Devils first received their first legal protection on 3 May 1941, when they were listed as “wholly protected” and it became illegal to kill them without a permit (Launceston Examiner, 5 May 1941). Prior to that, it was open slather. In the early part of the 19th Century there was a bounty on Devils, set at half the value of the bounty on Thylacines, and as late as 1928, Devils were listed as “wholly unprotected” (Hibberd, 2006). The legal protection given to Devils in 1941 may have been one of the contributing factors to the noticeable increase in their numbers beginning in the 1960s.

Interestingly, in what might have been a response to the scarcity of the Devil by the 1920s (above) the Commonwealth prohibited its export overseas, both alive and as skins, in December 1921 (Hobart Mercury, 16 December 1921).

With the advent of its facial tumour disease, the Devil received higher conservation status with up-grades on state, national and international threatened species list. In May 2008 it was raised to “Endangered” under Tasmania’s Threatened Species Protection Act 1995. In October 2008, it was raised to “Endangered” on the Red List of the International Union for the Conservation of Nature (IUCN). And in May 2009, it was raised to “Endangered” under the Commonwealth’s Environment Protection and Biodiversity Act 1999 (STDP website: “Status,” 3 August 2011).

In September 2006, the Devil’s facial tumour disease was gazetted under Tasmania’s Animal Health Act as a List B notifiable disease (STDP website: “Tasmanian Devils,” 24 June 2010; “Status,” 3 August 2011).

A draft species recovery plan was completed in November 2010 (DPIPWE, 2010d). In December 2014, the STDP had “moved closer to finalising” the recovery plan (STDP Annual Report, 2013/14), and in late September 2015 the plan was “in progress,” STDP website: “Wild devil recovery,” 23 September 2015). A 2015 DPIPWE document refers to a draft recovery plan with a date of 2013 (DPIPWE, 2015c), but I have not been able to find this putative later version.

Objectives, Bureaucracy, Communication and Finances

Objectives of the STDP

The objectives of the STDP have changed over time, and at least one goal seems to have been articulated slightly differently on different occasions.

One of the main objectives relates to the kind of population of Devils that may be living in the wild in the future. The original version of this goal seems to have been articulated at a workshop on the disease in October 2003. It was “the maintenance of the Tasmanian devil population at a sustainable level (recognising natural populations fluctuations), and the maintenance of its ecological role (AusVet, 2005, emphasis added).” Subsequent statements about the goal were in a similar vein, to “maintain the Tasmanian devil population in the wild (STTDP, 2007a; DPIW Annual Report, 2008/09, emphasis added)” and to “maintain the population in the wild (DPIPWE Annual Reports, 2009/10, 2010/11, 2011/12 and 2012/13; STDP Annual Report 201/11).” These statements seemed to imply the maintenance of a continuous population of Devils in the wild.

6/05/2016 8:41 AM 227

Other statements about the desired future Devil population seemed to have aspired to something less than a continuous population in the wild. In 2005, the goal was articulated as to “maintain the Tasmanian devil as a sustainable and ecologically functional species in the wild (AusVet, 2005).” It was then tweaked slightly to “maintain the Tasmanian Devil as an ecologically functional species in the wild” (Jones et al., 2007a; McCallum, 2008c; PHVA Report, 2008). It was tweaked again to ensure “an enduring and ecologically functional population of Tasmanian Devils in the wild in Tasmania (STTDP, 2007a; DPIW, 2008a; STDP, 2010a, c-d; STDP Annual Reports 2010/11, 2011/12).” The current tweak is, “… an enduring and ecologically functional population of Tasmanian devils living wild in Tasmania” (STDP, 2014c). On one occasion, it was said that “maintain[ing] the devil as an ecologically functional species in the wild” was “the overriding purpose of the Devil Program (PHVA Report, 2008).”

These statements could be interpreted as allowing for extinction in the wild with re- establishment from captive-bred animals, which would have the same ecological “functionality” as a continuously breeding population of wild Devils. So far, the only functions explicitly mentioned are “cleaning up carcasses, controlling herbivore populations and suppressing introduced predators such as cats and foxes” (McCallum, 2008c; see also Fox et al., 2010a). Perhaps of greater interest would be to know what aspects of the Devil’s ecological functionality the STDP would be willing to forego. In any event, one of the most experienced field workers with Devils was of the opinion that “we know surprisingly little about the ecological functions of Devils (Mooney, 2012).”

In 2007, the STDP added two other goals: “to maintain the genetic diversity of the Tasmanian devil population and to manage the ecological impacts of a reduced Tasmanian devil population over its natural range (STTDP, 2007a; see also STDP Annual Report, 2010/11; DPIPWE Annual Reports, 2009/2010, 2011/12, 2012/13).”

Bureaucracy

The first move toward a coordinated response to the Devil’s facial tumour disease came on 14 October 2003 when Australian wildlife biologists met in Launceston, Tasmania to discuss the disease (ABC World Today, 14 October 2003; Brand Tasmania Newsletter, October 2003; Darby 2003, 2011; AusVet, 2005; DPIPWE, 2010d; Woods and Sharman, 2010; Tyler et al., 2011; STDP, 2014c). The meeting was closed to the public and the media, because, according to the DPIWE public relations officer, the scientists “wanted to discuss issues in confidence.” This was news to some of the scientists (Bevilacqua, 2003c).

In April 2004, the Tasmanian state government brought the response to the Devil’s facial tumour disease under one group. This group was initially referred to, variously, as the Devil Facial Tumour Disease Project Team (Pyecroft et al. 2004: DPIWE, 2005a; Albion, 2006; Edwards, 2006; Mann, 2007; Siddle et al., 2007b; Woods et al., 2007); Devil Disease Project Team (DEH, 2006a); Devil Facial Tumour Disease Program (Jones and McCallum, 2007); Devil Project (Hesterman, 2007); DFT research group (Pyecroft et al., 2007) or just Devil Disease Project Team.

6/05/2016 8:41 AM 228

Sometime in the financial year 2007/08, this group appears to have been formally named the Save the Tasmanian Devil Program (STTDP) (DPIW Annual Report 2007/08; DPIPWE, 2010a, d). This group is now “the official national response to the threat posed” by the disease (STDP website: “Save the Tasmanian Devil Appeal,” 3 June 2010). Virtually all work done on the conservation of the Tasmanian Devil is carried out under the auspices of, or in partnership with, the STDP. This program is billed as “currently the largest single species conservation project in Australia (STDP Annual Report, 2011/12).”

Parenthetically, in 2013, the STDP said, “The Save the Tasmanian Devil Program was established in 2003 following a national workshop of specialists on the decline of the Tasmanian devil due to Devil Facial Tumour Disease” (STDP Annual Report, 2012/13; see also Huxtable et al., 2015). As noted above, this may have been true in terms of the concept of the group but not the name.

In June 2013, the Director of the STDP said “the program has been tracking … since 1996 (H. Williams, in Milman, 2013a, emphasis added).” It is hard to know what “program” he had in mind; it was certainly not even an early version of the STDP.

In any event, in February 2006, a Program Manager was appointed to the STTDP. This position is still current.

The STTDP originally described itself as “an initiative of the Australian and Tasmanian governments in partnership with the University of Tasmania” (STDP website: “Program administration,” 24 June 2010; DPIPWE Annual Report for 2010; see also Newsletter, September 2010).

In February 2010, however, the STDP was said to consist of the Commonwealth and Tasmanian Governments as well as the University of Tasmania (UTAS), the Zoo Aquarium Association (ZAA) and the Australian Wildlife Health Network (AWHN) (DPIPWE, 2010a).

In March 2011, the University of Tasmanian, Zoo and Aquarium Association and the Australian Wildlife Health Network were dropped from STDP’s description of “Who we are.” STDP was described simply as “… a joint initiative of the Australian and Tasmanian Governments.” Apparently, UTAS had become just one of several supporting institutions (STDP website: “The program,” as modified on 9 August 2011).

By September 2011, however, the university was reinstated to “Who we are,” but again only as a partner. The STDP was “… a joint initiative of the Australian and Tasmanian Governments in partnership with the University of Tasmania (STDP Newsletter, March 2012).”

In September 2013, the university was again omitted. The STDP was “a joint initiative of the Australian and Tasmanian Governments (STDP Newsletter, September 2012).”

By April 2014, however, the STDP was said to be “… a partnership-based involving at its core, the Tasmanian and Australian governments, the University of Tasmania

6/05/2016 8:41 AM 229

and the Zoo and Aquarium Association. Other partners include Devil Island Project, and local wildlife parks and zoos throughout Australia (STDP Newsletter, April 2014).” Why the university’s apparent official relationship to the STDP has changed so many times is unclear.

The coordinating authority for the STDP is the Tasmanian government’s Department of Primary Industries, Parks, Water and Environment (DPIPWE website (accessed 4 July 2010; STDP website: “The program,” 9 August 2011; DPIPWE, 2010a). The STDP’s website is through the DPIPWE, and its street address is the same as the DPIPWE.

All Devils, captive and wild, are “owned” by the Tasmanian government (CBSG/DPIPWE/ARAZPA, 2009; Lees and Andrew, 2012). As a result, the Tasmanian government decides who does and does not get access to Devils and their tissues (Waterhouse, 2007). The state government originally vested this responsibility to the staff of an earlier version of DPIPWE (Waterhouse, 2007). When a number of overseas researchers with relevant expertise complained their “offers to help” were either going unacknowledged or being knocked back (Duncan, 2006), however, the government gave the responsibility for assessing research proposals to the Senior Scientist (Duncan, 2006; Waterhouse, 2006). But when the Senior Scientist resigned in mid-2009 (below), it became unclear who had the assessment responsibilities, and it remains unclear today. Suffice it to say that in some areas, especially in some of its immunological aspects, the general research seems remarkably parochial.

In the early days of work on the disease, the Tasmanian government received “many requests” for samples of the tumour DNA from overseas scientists and these requests are assessed by Tasmanian researchers on the basis of whether “they might help the plight of the devil” (Johnston, 2007). However, to date only two overseas research groups appear to have received tissues (see Genome sequence). And one of these prevailed only after a major sponsor threatened to withdrawn support. The sponsor’s spokesperson said at the time: “It just shows you the politics of it. … There's too much self-interest, too much red-tape (Ball, 2009; see also Bevilacqua, 2005c and Appendix 2).”

At the 14 October 2003 meeting of Australian wildlife biologists in Launceston, a steering committee was formed to “to prioritise resourcing and individual projects in an overall strategy to tackle the devil disease (then Environment Minister Bryan Green, in Bevilacqua and Wood, 2003; Wade, 2003a)." Initially, the steering committee was generally referred to in lower case letters (DPIPWE, 2005a), as if it lacked a formal name. At least once, however, it was called the “Devil Disease Response Steering Group (DPIWE Annual Reports, 2003/04, 2004/05);” “Tasmanian Devil Disease Steering Committee” (UTAS, issue 273, 16 March 2005); the “Steering Group” (DPIPWE, 2010a), and possibly even the “Scientific Advisory Committee (DPIWE Annual Report 2004/05, 2005c),” a name which was apparently resurrected later for a new subcommittee (below). The committee consisted of seven representatives, two from DPIPWE, one each from Tourism Tasmania, Tasmanian Parks and Wildlife Service, Tasmanian Farmers and Graziers Association and the University of Tasmania, and a private ecotourism operator (DPIPWE 2010a).

6/05/2016 8:41 AM 230

In July 2006, apparently in conjunction with the appointment of a Senior Scientist (below), a new steering committee was formed (DPIPWE, 2010a). The Committee was to meet quarterly (STTDP, 2008d) but appears to have stopped meeting after 21 April 2008 and did not meet again for just over two years. It resumed meeting in the middle of 2010 (STDP Website: “Steering committee,” 5 October 2010). It is not clear why the break occurred.

As of 30 June 2013, the Steering Committee (SC) consisted of seven or eight members according to two different sections of the STDP’s Annual Report 2012/13. The seven members mentioned in both places included three from the Tasmanian Government’s DPIPWE, including the Chair; one was from the University of Tasmania (Zoology Department); one from the Australian Wildlife Health Network based at Taronga Zoo in NSW; one from Taronga Western Plains Zoo in NSW, apparently representing the Zoo and Aquarium Association, and one from a Commonwealth Government department. The possible eighth member was another DPIPWE staff member. The SC met quarterly (STDP, 2010d; Boland, 2011; STDP Annual Report 2011/12, 2012/13).

Sometime in 2014, the STDP underwent a structural redesign and the Steering Committee settled in numbers and changed in composition. It consisted of seven members. These included one from DPIPWE as the Chair; two from the STDP, including the most senior Tasmanian Government bureaucrat responsible for the program; one each from the Australian Government Department of the Environment (DoE) and the Zoo and Aquarium Association (ZAA), and one each from the science community (University of Tasmania) and the conservation community (a retired DPIPWE officer). The new SC also began to meet less frequently, twice a year instead of four times (STDP, 2014c; STDP Annual Report, 2013/14).

Prior to 2014, the Steering Committee was apparently responsible for both policy and governance. It was “responsible for policy and resource allocation decisions essential to the delivery of the program outputs and outcomes. The Committee was also responsible for oversight of the program components in the Business Plan, including risk management (STDP, 2010d).” Another view, from the program’s then Science Manager (below), was that the SC’s purpose is to approve all STTDP strategies and “ensure … the program continues to achieve its targets (Boland, 2011).” Another view, from the program’s first external reviewer, was the “SC is largely responsible for program governance and for making sure the program delivers the three objectives stated in the Business Plan and Recovery Plan by ensuring resources are allocated appropriately within the program (Parkes, 2012).” From the ZAA’s perspective, the SC was “responsible for policy and resource allocation decisions essential to the delivery of the program outputs and outcomes” and was “also responsible for oversight of the program components in the Business Plan, including risk management (Hibbard).” Clearly, the SC was, and after the 2014 redesign remains, the most powerful committee in the STDP.

In 2014, the Steering Committee was described succinctly as, “… responsible for the strategy and oversight of the program (STDP, 2014c).”

Prior to 2014, the Steering Committee took advice from three “subordinate” committees. In July 2011, the STDP established a Scientific Advisory Committee

6/05/2016 8:41 AM 231

(SAC) “to assist the Program with the development of science strategy and projects, evaluate the quality of science in the program, and to provide advice on the capability required to support the program (UTas media release, 5 July 2011).” As of 30 June 2013, the SAC had ten members, four from Tasmania, four from interstate and two from overseas (NZ) (STDP Annual Report, 2012/13). The SAC first met in 2012 and apparently met at least twice a year. In additional to its advisory role, it also held workshops and informal meetings as required (STDP, 2012b).” This Committee, both in its brief and composition, presumably made a major contribution to the research and conservation of the Devil.

The Meta-population Advisory Committee (MAC) provides advice on the Insurance Population to the Steering Committee. This committee met first in February 2010, but how often it met subsequently is unclear. In 30 June 2013, the MAC consisted of seven members. Two, including the Chair, were from DPIPWE; two from the zoo industry (Zoo and Aquarium Association); one from an interstate university (small population management expert), one from an international conservation group, and one from a Commonwealth government department (STDP Annual Report 2010/11, 2012/13). Sometime after this, the MAC was reduced to six members. Two, including the Chair, are from the DPIPWE; two from the zoo industry; one from an interstate university, and one from a Commonwealth government department (STDP Annual Report, 2013/14).

Finally, there was a Stakeholders Reference Group (SRG), which currently “provides a channel for communication between the Steering Committee and the Program’s key stakeholders (STDP Annual Report, 1011/12).” This group seems to have been formed in 2006 (Newsletter, November 2006). It met quarterly until November 2008 when it seems to have ceased meeting (McClone, 2010) for reasons that are unclear. It was apparently re-formed in 2010 and was to meet quarterly (DPIPWE, 2010a). As of 30 June 2013, the SRG consisted of 13 members, representing animal welfare (3), conservation (5), primary industry (1), veterinary medicine (1) and the University of Tasmania (1) and three DPIPWE representatives, including the Chair (STDP Annual Report 2012/13). For a while, the SRG met quarterly as planned, but when this proved difficult for its members it was to meet on an “at least twice a yearly basis (STDP, 2012b).” In fact, in 2010/11, it met twice (DPIPWE Annual Report 2010/11), and in 2011/12, it met three times (DPIPWE Annual Report 2011/12). Subsequently, it seems to have returned to quarterly meetings (STDP Annual Report 2011/12).

After the 2014 redesign of the STDP, the number of the subcommittees reporting to the Steering Committee was reduced to two, one new and one old. Gone were the Scientific Advisory Committee and the Stakeholders Reference Group, an apparent net loss of scientific and community input. The slightly renamed Metapopulation Advisory Committee (MAC) remained and the Strategic Implementation Group (SIG) was new.

The MAC retained oversight of the Insurance Program, including the Ambassador Program (below). It is not clear how many members MAC now has, but its institutional representation remains roughly the same. How often MAC meets remains unclear.

6/05/2016 8:41 AM 232

MAC did gain a new subcommittee, the pre-existing Captive Research Advisory Group (CRAG) (below).

SIG “provides oversight of the program components in the Business Plan, implementation and coordination of projects (STDP, 2014c). Its standing members are the STDP Director, Program Manager and a Zoo and Aquarium Association (ZAA) representative, but it also includes “other technical experts as required.” The SIG meets four times a year (STDP, 2014c).

Sometime in the latter half of 2012 or the first half of 2013, the Molecular Research Advisory Committee (MRAC) was formed to assess applications to use the STDP’s tissue collections. MRAC consisted of three members, one from the STDP and two from “external institutions (STDP Annual Report, 2012/13).” It was not clear to whom the MRAC reported. In the 2014 redesign of the STDP’s structure, this committee was renamed the Molecular Research Advisory Panel (MRAP). Whether its membership remained the same was unclear (STDP, 2014c; STDP Annual Report, 2013/14); it was made clear, however, that MRAP reported to the STDP Management Group (STDP Annual Report, 2013/14).

In the same period, the Captive Research Advisory Group (CRAG) was also formed. It assessed applications for research on Devils in the Insurance Population. It originally consisted of three members, two from the STDP and one from the Zoo and Aquarium Association (ZAA). In the 2014 redesign of the STDP structure, CRAG became a subcommittee of MAC. Whether it retained the same members is unclear.

In July 2011, a Management Group (MG) was “reformed” to “provide representation of the three operational sub-programs along with the Science Manager” as well as “a more effective operational management forum (STDP, 2012b).” It is not clear what name or purpose of the implied earlier group had. The Management Group consisted of the “heads of each sub-program” and “program management staff.” The Management Group reported to the Program Manager who reported to a Program Director who reported to the Steering Committee (Parkes, 2012: fig. 1). The PMG met fortnightly (STDP, 2012b).

Sometime in the latter half of 2012 or first half of 2013, the Management Group had its name changed to the “Program Management Group” (STDP Annual Report, 2012/13) and its brief re-stated as “providing for the strategic planning and management of the STDP … (STDP Annual Report, 2012/13).”

The Program Management Group (PMG) consisted originally of eight members, all from the DPIPWE: the Program Director (Chair), the Program Manager, the Science Coordinator, the “heads” of the three operational sub-programs (Monitoring and Management, Insurance Meta-population and Diagnostic Services and Research) and two others, one from public relations (STDP Annual Report 2010/11; 2012/13; STDP, 2012b).

In the STDP’s “Business Plan 2014-2019 (published December 2014), the Program Management Group seems to have morphed into the “STDP Management Group.” Its brief was to “provid[e] coordination and planning for the operational management of the Program.” Its membership was reduced in size to an unstated number, but

6/05/2016 8:41 AM 233

included explicitly, the STDP Director, Program Manager and Sub-program leaders. There were three sub-programs, two old and one new. The two old sub-programs were the slightly renamed Insurance Population Sub-program and Monitoring and Management Sub-program, and the new sub-program was the Program Management Sub-program. The old Diagnostic Services and Research Sub-program was dropped, a seeming further reduction in the input of science in the STDP. The STDP Management Group meets fortnightly (STDP, 2014c).

The STDP’s Annual Report for 2013/14 (published July 2015) seems to have reverted to the name “Program Management Group,” and its composition reduced to seven, all from the DPIPWE. These were the Program Director, the Program Manager, Science Coordinator, Sub-program Leader – Insurance Population, Sub-program Leader – Monitoring and Management, the DPIPWE’s Manager of Wildlife Management and a public relations officer. The Program Management Sub-program seems to have been dropped. The PMG’s brief remained unchanged, and it continues to meet fortnightly (STDP Annual Report, 2013/14).

For a while, the STTDP had a Senior Scientist. He was appointed in June 2006 (ABC News, 26 June 2006; see also ABC News, 10 November 2005) and resigned in mid- 2009. This position was apparently abolished after the incumbent’s departure.

In December 2010, the newly created position of Science Manager was filled. The job of the Science Manager was to provide focus on the “application of research” in the STDP (Newsletter, December 2010) and to spread research results within the STDP in advance of formal publication (STDP, 2010d). Or as the Science Manager later saw it, “providing strategic leadership to over 50 research scientists from diverse scientific disciplines (Westjobs.com.au: resume 771991).”

Sometime in late 2012, the Science Manager left the job and a new position of “Science Coordinator” was created. Although from the timing and the name the new position looked as if it was a continuation of the old position, the Science Coordinator’s responsibilities were never stated publically. In any event, the position seems not to have lasted very long as it was last mentioned in the STDP’s Annual Report for 2012/13. The five-year business plan for 2014-2019 mentions no job that sounds like a Science Manager or a Science Coordinator (STDP, 2014c), so the job may have been discontinued or give to another position, perhaps the Program Manager. This appears to be a further reduction in the input of science in the STDP.

Communication

There are four official sources of information about the Save the Tasmanian Devil Program: the STDP website, the STDP Newsletter and the annual reports of the DPIPWE and the STDP.

The STDP website contains much interesting information and many useful links. It also often foreshadows developments, mostly “good news” stories, well before they are published more formally. It is not clear why the number of stories fluctuates so greatly from year to year (Fig. 18).

6/05/2016 8:41 AM 234

Number of Articles per Year

20 Number of Articles 10

0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year

Fig. 18. The number of new articles published each year on the STDP website.

The STDP also publishes a newsletter on its website. The first issue was published in March 2006. Between 2007 and 2010, the newsletter appeared quarterly. In 2011, however, there were only three issues; in 2012, there were two, the last being for September 2012; in 2013 there was one, for May but appearing in early June, and in 2014, there was one, for April. It now appears as if the newsletter is published annually.

It is not clear why the website has become less informative and the newsletter produced less often. Perhaps it is just coincidence, but the slow-down seems to have begun in the latter part of 2011 after the STDP received some scrutiny based on information in the public record (Greer, 2011a-b).

Both the DPIPWE and its subordinate STDP publish annual reports. The DPIPWE reports usually contain only a few concise paragraphs, whereas the STDP is, unsurprisingly, lengthier. The DPIPWE and its earlier equivalents published annual reports throughout the DFT saga. The STDP started publishing its reports in the year 2010/11 and switching to calendar year reporting in 2014, which therefore included the 18 month period July 2013 – December 2014.

Sometime in 2013, the STDP began producing a “more detailed” news bulletin, “Speak of the Devil,” for dissemination via email to “key stakeholders (STDP Annual Report, 2013/14; see also DPIPWE Annual Report, 2015). In the 18 months to July 2014, three copies had been sent out (STDP Annual Report, 2013/14).

Finances

There is insufficient information readily available in the public domain to determine precisely the sources and sums of money coming into the STDP program, and the recipients and sums going out. In general, up until the later half of 2013, funding for STDP came from, in decreasing order of largess, the Tasmanian government, the Australian government and the public. Public donors range from large international

6/05/2016 8:41 AM 235

corporations to individual children donating money they have earned or solicited (see stories on STDP website: “Kids Club,” published 9 December 2010). Indeed, a large part of the STDP website is dedicated to communicating with the public and encouraging public donations.

Between 2004 and 2013, the Tasmanian and Commonwealth governments allocated or committed $ 26.8 million ($16.8 and $10 million, respectively) to the STDP, and between 2004 and 2011, public donors gave a total of almost $ 1.6 million (University of Tasmania Annual Report, 2011). This is a total of at least $ 28.4 million received by the STDP.

As to the Tasmanian and Commonwealth governments’ allocations, general statements of the annual expenditures of the STDP are available for the financial years 2008/09, 2009/10, 2010/11 and 2011/12 (Table 18). The budget for the current year, 2012/13, has not been released. New funding must be sought beginning with the 2013/14 financial year.

Table 18. Expenditure, by sub-program and rounded, of the $ 5 million annual Tasmanian and Commonwealth government allocations to the STDP (STDP Annual Reports, 2010/11, 2011/12, 2012/13, 2013/14 and Parkes, 2012).

Sub-program 2008/09 2009/10 2010/11 2011/12 2012/13 2013/142 Program 0.93 1.00 0.66 0.93 0.88 0.63 Management Monitoring and 1.06 1.17 1.08 1.13 1.13 0.77 Management Insurance 2.17 2.10 2.41 1.71 1.89 1.19 Populations Diagnostics and 0.83 0.83 0.85 0.86 0.78 0.35 Research Total 5.00 5.00 5.00 4.641 4.68 2.95 1. Less c. 0.40 to be carried over for “infrastructure” projects. 2. Covers period from 1 July 2013 to 31 December 2014.

Two large changes in these expenditures are noteworthy. First, is a decline in the Insurance Population spend in the 2011/12 year, which may reflect the reduced capital costs of establishing the program. Second, is the large reduction in total expenditure in the 2013/14 financial year, which may reflect the maturity of the STDP’s management plans.

In mid-2013, the Commonwealth Government stopped its funding for the STDP (Burgess, 2013, 2014b; ABC News, 13 January 2014). It appears as if the Tasmanian Government cast its bid to the Commonwealth, not in terms of general expenditure as in previous funding requests, but in terms of a $ 4 million-over-four-years proposal to fence off parts of Tasmania in order to exclude DFTD (ABC News, 30 August and 1 September 2013; Poskitt, 2013). The Commonwealth Government rejected a subsequent Tasmanian Government proposal to re-direct c. $2 million from the state’s fox eradication project to the STDP (ABC News, 13 January 2014; Burgess, 2014a; Mounster, 2014). However, in early 2014, the Commonwealth Government allocated $3 or 3.3 million, depending on accounts, to “... help establish more disease-free areas

6/05/2016 8:41 AM 236

…. (ABC News, 23 January 2014; Burgess, 2014b).” The Tasmanian State Government continued its funding to the STDP for the year 2013/14, to the amount reported variously as $3 million (ABC News, 13 January 2014) or $2.6 million (ABC News, 23 January 2014; Mounster, 2014).

As to public donations, the first public appeal for funds to research the Devil’s facial tumour disease was launched on 13 November 2003 (Wade, 2003a; Brand Tasmania Newsletter, December 2003; Hobart Mercury editorial, 14 November 2003). The appeal was called, variously, the “Tasmanian Devil Disease Community Appeal” (Wade, 2003b; Brand Tasmania Newsletter, December 2003 and April 2004); the “Tasmanian Devil Research Trust appeal” (Whinnett, 2003b); the “Tasmanian Devil Appeal” (Brand Tasmania Newsletter, December 2003), and the “Tassie Devil Appeal” (Unitas, issue 273, 16 March 2005).

The funds were to be administered by the “University of Tasmania Foundation” (Burnie Advocate, 14 November 2003; Whinnett, 2003b) and would be “…used for University of Tasmania research into the facial tumours” (Wade, 2003b). The first research grants were awarded in 2005 (Hobart Mercury, 2 February 2005; ABC News, 8 March 2005; STDP website: “Save the Tasmanian Devil Appeal: grants and scholarships allocations,” 3 March 2011).

In June 2009, an agreement between the Tasmanian government and the University of Tasmania led to what appeared to be a modification of the STTDP’s public appeal program to include management initiatives as well as research. The modified appeal, called the Save the Tasmanian Devil Appeal (STDA), described itself as “the only official fundraising entity that directs funds in full to … [the Save the Tasmanian Devil Program]. So 100 % of the funds raised by the Appeal go to the research and management activities that have been prioritised as important to the long-term solution to devil facial tumour disease and the aim to keep Tasmanian devils sustainable in the wild” (UTas website: “Save the Tasmanian Appeal,” 3 June 2010; see also joint Tasmanian government and UTas media release, 5 June 2009; ABC News, 6 June 2009; Young, 2009). The STDP itself calls the STDA its “fundraising arm (STDP website: “Funding boost for the survival of the Tasmanian Devil,” 24 April 2013).”

The STDA is over-seen by the Save the Tasmanian Devil Appeal Committee (STDAC), which is part of the University of Tasmania Foundation. STDAC consists of seven members, two from the University of Tasmania, two from Tasmanian state government departments, two from private industry in Tasmania and one from a Commonwealth government department (STDP website: “Save the Tasmanian Devil Appeal Committee (STDAC),” 10 August 2011; STDP, 2010a). The STDAC met for the first time on 13 August 2009 (Hobart Mercury, 14 August 2009).

On the face of it, it appears that the STDA, like its predecessor, is officially part of the University of Tasmania, but it directs its funds in accordance with the STDP, which is, as noted above is “an initiative of the Australian and Tasmanian governments” but “coordinated” by the Tasmanian government’s DPIPWE. Indeed, DPIPWE has a full- time manager who assists the STDA committee (DPIPWE Annual Report 2009/10; see also STDP Annual Report 2010/11).

6/05/2016 8:41 AM 237

In 2010, STDA’s goal and benchmark of success was to raise “more than $ 1 million” (UTas website: “Save the Tasmanian Devil Appeal,” 3 June 2010) and $ 1.7 million in 2011 (STDP, 2010a). In the event, it raised $ 277 000 in 2010/11 (STDP Annual Report 2010/11) and $ 441 796 in 2011/12 (STDP Annual Report 2011/12) The STDA’s goal for 2012 was $ 1 million (STDA website, accessed 20 December 2012); it achieved “$ 0.5 million” in 2012/13 (STDP Annual Report, 2012/13). Its goal for 2013 was unstated, but it took in $ 508 595 in 2013/14 (STDP Annual Report, 2013/14).

Between 2003 and mid-2012, the STDA (and its precursors) had raised a total of $ 2.5 million (data in cited documents).

According to the Department of Primary Industry, Parks, Water and Environment’s 2012/13 annual report, DPIPWE initiated an independent review of the STDP Appeal’s “effectiveness” (DPIPWE Annual Report 2012/13). The results of this review never seem to have been placed in the public domain.

Funds donated by the public to the STDP via STDA are allocated by the Tasmanian Devil Research Advisory Committee (TDRAC). At the beginning of 2010, this committee consisted of seven members, six from the University of Tasmania and one from DPIPWE. In August 2011, it still consisted of seven members, five from the University of Tasmania, one from DPIPWE (the STDP Manager) and one without stated affiliation but possibly a University of Tasmania Honorary Research Associate in the School of Zoology (STDP website: “Grants & scholarships,” 11 August 2011). In May 2013, it had been reduced to five members, four from the University of Tasmania and one from the DPIPWE (the STDP Manager) (STDP website: “Grants & scholarships,” 20 May 2013). In May 2015, it was back to seven members, five from the University of Tasmania and two from the DPIPWE (STDP Director and Manager) (STDP website: “Grants & scholarships,” as modified on 16 May 2015).

Currently, TDRAC allocates funds to three major areas: research (including student research), management and community projects.

Between 2005 and 2013, the STDP distributed at least $ 1 149 905 for research projects on the Tasmanian Devil (N = 70). Of this, 72 percent went to the students and staff of the University of Tasmania and 73 percent went to these plus other researchers in Tasmania (data in STDP website: “Grants and scholarships. Current grants recipients,” published: 16 February 2010, modified: 24 May 2012).

Between 2005 and 2013, TDRAC or its predecessors distributed $ 246 641 for management projects. Sixty-eight percent of the funds went to the six Tasmanian projects and the remainder to three mainland projects.

Between 2011 and 2012, TDRAC distributed $ 61 670 for two community projects, both in Tasmania.

Overall, therefore, of the $1 458 216 publicly donated funds that TDRAC or its predecessors has dispersed, 73 percent have gone to projects based in Tasmania.

6/05/2016 8:41 AM 238

In 2009, the STTDP established the Tasmanian Devil Conservation Grants scheme. “The purpose of these grants is to increase the capacity and number of Australian institutions participating in the Insurance Population (STDP website: $ 492,000 to expand the Tasmanian devil Insurance Population, 3 December 2010).” The scheme is funded by the Commonwealth and Tasmanian governments (STDP website: $ 492,000 to expand the Tasmanian devil Insurance Population, 3 December 2010) and administered by the Zoo Aquarium Association’s Wildlife Conservation Fund (STDP website: $ 492,000 to expand the Tasmanian devil Insurance Population, 3 December 2010; STDP Annual Report, 2010/11). And “while the Grants Program is open to activities across Australia, a portion of the funds is reserved for projects in Tasmania (DPIPWE annual report for 2010).”

It is not clear how this scheme actually works. The application procedure, review process, reviewers and even the Tasmanian and Commonwealth government budgets from which the funds are allocated are not in the public domain.

In any event, in December 2010, the scheme awarded three grants, totalling $ 492 000, to three ZAA members, one mainland zoo to build a free-range enclosure ($ 350 000) and two existing Tasmanian zoos to incorporate their Devils into the Insurance Population program ($ 105 000 and $ 37 000; STDP website: “$ 492,000 to expand the Tasmanian devil Insurance Population,” 3 December 2010). In April 2011, $ 400 000 was budgeted for a second round of grants (STDP Annual Report, 2010/11), but this expenditure was deferred to the 2012/13 financial year (STDP Annual Report, 2011/12). In April 2013, the scheme awarded four grants, totalling $ 414 330, to four ZAA members, two on the mainland and two in Tasmania, mainly to expand the number of enclosures and enhance security (STDP website: “Funding for the survival of the Tasmanian Devil,” 24 April 2013). In summary, therefore, the Tasmanian Devil Conservation Grants scheme has awarded a total of $ 905 330 to five zoos, three in Tasmania and two on the mainland.

It is important to appreciate that not all funds expended in the cause of saving the Devil pass through the STDP. For the period 2004-2013, an additional $ 9.7 million was allocated or committed directly to individuals and facilities involved in Devil research and management. These funds included government research and development grants and private donations. For example, the Commonwealth government committed $ 3.6 million in research grants and fellowships; the NSW government committed $ 1.4 million to two NSW zoos (which not only have breeding programs but are also major tourist attractions in the state), Devils in Danger (an association of Tasmanian corporations) gave $2.5 million to develop a “Devil Awareness and Research Conservation Centre” in Launceston, and private donors gave to various zoos for their Devil breeding programs.

There are also other “appeals” besides the STDA. Most institutions in the Insurance Program have appeals specifically for their Devils. Even some researcher institutions have launched appeals. In July 2012, for example, the University of Sydney launched an appeal for funds called “Save the Tasmanian Devil” to “understand the genetic make-up of each individual Devil” in captivity in order to manage them for genetic diversity and appropriate pairings (University of Sydney Tassie Devil Fund-raising Flyer, July 2012; Gill, 2012). Media reports indicate the testing will be done on behalf

6/05/2016 8:41 AM 239

of the Zoo and Aquarium Association, the umbrella organisation for the captive breeders, and the Tasmanian government (Ryan, 2012; Gill, 2012).

In summary, between 2004 and 2013, it seems at least $ 38.1 million has been allocated or committed to Devil research and conservation in response to the Devil’s facial tumour disease. This is a very conservative estimate, however, because it omits funding of uncertain provenance; large grants of which Devil research was just a part; in kind contributions; foreign institutional funding, and private donations not in the public domain. These sources could easily amount to several million dollars. The total expenditure to date almost certainly exceeds $ 40 million.

In 2003, an overseas “expert” thought at least $ 1.4 million a year would be needed to get a handle on the disease (Bevilacqua and Wood, 2003). In 2007, however, the then Senior Scientist for the STTDP said that at least $ 10 million a year for five years, that is, $ 50 million, would be needed for any serious attempt to save the Devil (McCallum, in Paine, 2007). And in 2011, it was estimated, somewhat unrealistically based on what had already been spent or committed, $ 11 million would be required to save the Devil (Phillips, 2011).

Captive management costs

The cost to keep a Devil for one year in the Insurance Program varies according to the type of facility. The cost to keep one Devil in a zoo for one year is estimated at between $ 7 000 and $ 10 000 (H. McCallum in ABC News, 21 February 2007; McCallum, 2008b; J. Weigel, in Newsletter, September 2008; STTDP website: “Appeal news,” accessed 17 August 2012; P. Andrew in Coorey, 2009; Taronga Conservation Society website: “Devil dinners,” 5 June 2012; C. Hibbard, pers. comm. in Parkes, 2012; calculations from figures given by Director of Wildlife Conservation and Science, on ABC World Today, 4 June 2013). In 2016, it was said that in some facilities, the cost can even exceed $ 10 000 per year (Weigel, 2016).

The cost to keep a Devil for one year in one of the three smaller free-range enclosures has been estimated to be about $ 900 (Denholm, 2010e). In 2008, the cost to keep one Devil for one year at the large Devil Ark FRE was estimated to be between $700 and $900 (J. Weigel, in Newsletter, September 2008; then NSW Environmental Minister Robyn Parker in Kelly, 2011b-c), but in 2016, the actual cost to feed 180 Devils was about $ 1300 per Devil per year (Weigel, 2016). In 2012, the cost to feed a single Devil in the Devil Ark FRE was $ 2.00 per day (Alma Park website: “Raising money for Devil Ark,” 9 September 2012; Devil Ark website: “Feed a Devil day,” accessed 7 October 2012), a figure which annualises to $ 730, which is about half the total cost.

In 2012, the STDP estimated the average cost of keeping a Devil in the Insurance Population was $ 5800 (STDP Annual Report, 2011/12). In 2013, however, this had risen to $ 9500 (STDP Annual Report, 2012/13). Using the latter figure, the total cost for maintaining 1500 Devils (below) in captivity for one year is $ 14.25 million and the cost for maintaining 5000 Devils (below) for one year is $ 47.5 million. Over 50 years (below) that amounts to $ 712.5 million and $ 2.375 billion, respectively.

In 2013, the total annual cost of maintaining the Insurance Population, then comprising about 550 animals, was reported as $ 5 million (Poskitt, 2013). Using this

6/05/2016 8:41 AM 240

figure, the cost of maintaining 1500 and 5000 Devils in captivity for one year would be $ 13.6 million and $ 45.5 million respectively. Over 50 years this would amount to $ 680 million and $ 2.27 billion, respectively.

Obligations of grantees

Up until 2012, there were no indications in the public domain about what conditions, if any, were placed on successful applicants to the public funds generated by the Save the Tasmanian Devil Appeal. In late May 2012, however, guidelines for grants and scholarships were published, and these contained some specific obligations to which grantees had to agree (TDRAC Grant Guidelines).

One of these was that “UTAS must be involved to some degree in the project or that some demonstrable and tangible benefit must flow back to UTAS from the making of the grant.”

Another was that, for research grants, an applicant was “more likely to be successful” if the applicant made a commitment “… to confer rights to the research to UTAS.” Or stated another way, in general, “… successful applicants will be required to transfer the project outputs (including, as a general rule, intellectual property generated in the research) to UTAS.” Furthermore, “… the applicant should indicate clearly how the project’s outcomes will be shared, including ownership of intellectual property.”

In terms of scholarships, the only eligible students are those enrolled in University of Tasmania.

Finally, the Save the Tasmanian Appeal “… must approve any publicity relating to the grant or the project prior to the release of that publicity.”

The general effect of these requirements is that although the donations may come from any place in the world, the benefits flow preferentially to the University of Tasmania, its staff and students. This may inhibit both non-UTas researchers and non- Tasmanian donors. Non-UTas researchers and their institutions may be put off by the apparent ownership of any intellectual property rights, and inter-state and overseas donors may be put off by the fact that their donations will be allocated on several criteria, only one of which is merit.

There is also a clear implication of censorship in what researchers can say about their work and its implications. This too may impede inter-state and overseas researchers.

Insurance Population

When Tasmanian wildlife authorities realised the facial tumour disease was spreading rapidly and seemingly unstoppable, they decided to try to protect at least some healthy Devils from infected Devils through isolation (N. Mooney, in ABC Science, 1 August 2003, Squires, 2003 and Wood, 2003a; Wood, 2003c). The idea was that if Devils went extinct in the wild, taking the tumour cell line with them, healthy Devils could be released back into the wild. Under this scenario, the Devil’s future existence could depend entirely on this Insurance Population, as it eventually came to be known officially (STTDP, 2007b). Furthermore, in the early days before it was known how

6/05/2016 8:41 AM 241

the disease spread, the possibility that it might be airborne led the authorities to plan for isolation on the Australian mainland. When it became apparent that the disease was almost certainly transmitted through contact and not through the air, however, the isolation program was extended to include Tasmania (STTDP website: “Insurance Population,” 8 April 2008).

The Insurance Population program envisages two kinds of isolation. One involves moving healthy Devils to new locations and other involves fencing off still healthy populations in their original location. In the first case, Devils are shipped to zoos and wildlife parks (hereafter, zoos) on the mainland; put into large “naturalistic” enclosures in Tasmania and, more recently, on the mainland, or released onto offshore islands. These endeavours are well underway. In the second case, healthy Devil populations were to be fenced off, taking advantage of natural limitations to dispersal, such as isthmuses and inappropriate habitat. This effort is just beginning.

The first meeting with the organisation representing the receiving institutions, what is now the Zoo and Aquarium Association (ZAA), was in mid-December 2004 (ABC PM, 8 December 2004; Hobart Mercury, 8 December 2004). Sometime in 2005, the ZAA produced a captive management plan that had the imprimatur of the DPIWE (Lees, 2005). Parenthetically, this important planning document is not in the public domain.

The authorities started to collect Devils from the wild for the insurance population in early 2005 (Hogg et al., 2015). They collected 25 Devils from Narawntapu National Park in northcentral Tasmania and from a far southern locality in 2005; three from an undisclosed locality in 2006, and 15 and 79 from northwest Tasmania in 2007 and 2008, respectively (Hogg et al., 2015: fig. 1). At some later, unspecified date, ten orphans” from diseased females were included in the IP (Hogg et al., 2015; see below). Currently and in the future, additions to the IP will be female orphans (Hogg et al., 2015).

These animals were placed in quarantine for 12 months in some accounts (DPIWE, 2005c; Devils on the Verandah website: “Noah’s Ark for Devils,” undated; Sinn et al., 2010), 18 in others (Woods and Sharman, 2010; STDP website: “Background,” 13 March 2010) and two years in others (Jones et al., 2007a). In fact, the first shipment of Devils to the mainland occurred in early December 2006 (AAP, 13 December 2006; ABC 7:30 Report, 20 November 2006; Denholm, 2006b; Grenfell, 2006; Wray, 2006), suggesting a 12 month quarantine period. A much later report said the quarantive period had been for “at least 12 months” (Hogg et al., 2015).

All the so-called founders for the Insurance Population had ear biopsy samples taken, and these were genotyped between 2006 and 2008 using polymorphic microsatellite markers (Hogg et al., 2015).

There must have been some formal agreement or contract between the Tasmanian authorities and the precursor to the Zoo and Aquarium Association. And this agreement must have been concluded sometime between the two group’s first meeting in December 2004 and the appearance of the captive management plan in 2005, but certainly no later than the first shipment in December 2006. The details of this agreement are not in the public record, but the fact that virtually every mainland ZAA

6/05/2016 8:41 AM 242

institution, even those in only marginally appropriate climate zones, has been supplied with Devils, may indicate its scope. Therefore, when the original rationale for mainland captivity, that is airborne transmission, was effectively negated with the discovery, perhaps as early as the end of 2005 and no later than early 2006, that direct physical contact between an infected and a healthy Devil was the mode of transmission, there may have already been a binding contract to supply Devils to most, if not all, ZAA members on the mainland.

At least one part of the agreement between the DPIPWE and ZAA can be inferred from a media story that reported “the baby Tasmanian devils have just emerged from mum's pouch and are being hand-reared in an effort to accustom them to human handling. It is hoped they can be used in interactive experiences with visitors (Pierce, 2012).” Indeed, one zoo, Dreamworld, offered for a “limited time only” an opportunity to be photographed holding the zoo’s only joey for $ 69.00 (Dreamworld website: “Tasmanian Devil Experience;” accessed 8 June 2013). In that this non- breeding role may make the animals less suited for the breeding for release back into the wild program, it is reasonable to infer that the agreement may allow some Devils in the Insurance Program to be used for purposes other than strictly breeding.

For the most part, the authorities stoped taking Devils from the wild sometime in 2008, by which time they had introduced 139 animals into the Insurance Population (Hibbard et al., 2012; 123 in Hogg et al., 2015). Since 2010, “the only wild-caught founders coming into the program” were said, in early 2012, to have been from the “orphaned wildlife program” (Lees and Andrew, 2012; see also Hogg et al., 2015). It was not clear what exactly this program was.

Whatever their source or age, the authorities continue to take a few animals from the wild in order to seed new facilities and to introduce fresh genomes into the Insurance Population. For example, the Freycinet Peninsula free-range enclosure (below) was stocked with 18 animals taken from the west coast sometime in 2010 (ABC News, 30 April 2011). And the Devil Ark free-range enclosure was stocked, in part, with 15 Tasmanian Devils of unstated provenance, and hence possibly from the wild, in 2011 (ABC AM, 18 January 2011). In 2012, Devil Ark also received an animal from the northeast part of the state, where the disease started (ABC AM, 27 November 2012; Egan, 2012b; Gloucester Advocate, 27 November 2012; Keene, 2012). In any event, the option remains for Devils to be sourced, as needed, from the wild in areas where the disease has not yet arrived and from joeys of diseased mothers (Hibbard et al., 2012; Lees and Andrew, 2012).

Size of Insurance Population

From the beginning, the size of the Insurance Population was based on the criterion of no net loss of genetic diversity (Jones et al., 2007a) or at least maintaining genetic diversity at 95 percent. This would require at least 1500 (STTDP, 2007b; Newsletter, December 2007; PHVA, 2008; Lees and Andrew, 2012) to 1700 (Jones et al., 2007a) if these Devils were managed intensively at the level of individuals and at least 5000 (Jones et al., 2007a; STTDP, 2007b; PHVA, 2008; Lees and Andrew, 2012) if they were managed more generally at the level of the population.

6/05/2016 8:41 AM 243

It was envisaged that the Insurance Population might have to be maintained for at least 50 years (30 years in some accounts) to allow for extinction to occur and re- introduction to be accomplished (Jones et al., 2007a; Sinn et al., 2010; STDP website: “Insurance Population,” 12 October 2010; M. Jones, ABC Science Show, 3 September 2011; H. Williams, ABC News, 25 May 2012; Boness, 2012). Fifty years is equivalent to about 25 Devil generations and is longer than the likely working lifetimes of most of the people involved with the Devil when these calculations were being made.

The stated number of Devils in the Insurance Population at any one time varies greatly. In August 2012, the IP was said to contain over 500 animals (Belov and Hogg, 2012). In October 2012, it was said to contain 515 animals. Of these, 385 were in zoos - intensively managed facilities, and 130 were in free-range enclosures (below) - intermediately managed facilities. This same number (515) was still current “up to the end of 2012” (Hogg, 2013). As of 30 June 2013, one document gave the number variously as “well over 500,” “over 550,” “approximately 580” and “622,” including the 15 Devils on Maria Island (STDP Annual Report, 2012/13). In October 2013, a Tasmanian Government minister said there were “over 600” Devils in the IP (ABC Hobart, 21 October 2013). In November 2013, a media report said there were 622 Devils in the IP. In December 2013, another media report gave the number as 560 (Kelly, 2013b). An anonymous document covering the period from 2012 to 2013 said there were “over 600 devils in the program (Zoo and Aquarium Association, ZAA Conservation Activity Report 2012 to 2013; see also STDP Newsletter, April 2014).” In mid-2013, the Insurance Population was expected officially to grow to “approximately 650” Devils in the year 2013/14 (STDP Annual Report, 2012/13). In November 2014, the insurance population was said to stand at 600 (Day, 2014; Mather, 2014b), and at the end of 2014, the STDP reported the size of the Insurance population as 610 (STDP Annual Report, 2013/14).

With the 2012 breeding season, Australian zoos reached their capacity to house their part of the Insurance Population, and for the 2013 breeding season, the zoos planned to manage breeding “to ensure that those animals that have not bred before and who are three years of age, are given the opportunity to breed; as well as ensuring that those individuals who are least related to others in the population are given breeding opportunities” (STDP Annual Report, 2012/13; see also Hibbard et al., 2012; STDP Annual Report, 2013/14).

At some stage, the STDP had realised the Insurance Population would rapidly reach saturation point (Hibbard et al., 2012) and there would need to control reproduction. One solution was a long-term contraceptive implant. The Meta-population Advisory Committee approved a pilot project with this goal in May 2013.

In 2013, the STDP noted that the efficacy of contraceptive “implants” was being examined (STDP Annual Report, 2012/13), and in December 2014, it announced that trials at two “captive facilities” showed early promise (STDP website: “Tasmanian Devil contraception trial shows early promise,” 2 December 2014).” Work carried out at Taronga Western Plains Zoo, showed that contraception was effective in female Devils, although questions remained about minimum dosage (STDP Annual Report, 2013/14; see also, STDP website: “Tasmanian Devil contraception Trial shows early promise,” 2 December 2014). Further work was foreshadowed on female Devils in

6/05/2016 8:41 AM 244

unnamed FREs and on Maria Island (STDP Annual Report, 2013/14; STDP website: “Maria Island Translocation Monitoring Project Update, December 2014,” 23 December 2014). In 2014, prior to the 2015 breeding season, seven females on Maria Island were given contraceptive implants and none of the females conceived (STDP, 2015a).

Selective contraception, of course, is just another example of the artificial selection that is rife in captive breeding. On the face of it, the genes of less fecund females will have greater proportional representation in the next captive generation than they might have had otherwise.

Initially, the contraceptive work also included males, in the hope that implants might curtain their aggressive behaviour to one another and allow them to be grouped together. The implants, however, raised the males’ levels of testosterone (increasing their aggressive behaviour?), and the work was discontinued (STDP Annual Report, 2013/14).

In early 2015, the long-term zoo component of the Insurance Population was projected to house about 550 Devils in financial years 2014/15 and 2015/16, then drop to 500 in the financial year 2016/17 and fall further to 450 in the years 2017/18 and 2018/19 (STDP, 2014c).

The other two outlets for the overflow from the Insurance Population are the Tasmanian Devil Ambassador Program, where Devils surplus to requirements are sent to overseas zoos (see “Overseas facilities”) and research programs (Glaetzer, 2014b; also Hobart Mercury Editorial, 24 November 2014).

If the wild Devil population should start (continue?) to increase in numbers naturally, the question will eventual arise about what to do with the Insurance Population. If these animals are released into the wild in order to “empty” the breeding program, locally adapted populations will experience an influx of genes from disparate localities; novel allelic combinations arising from “unnatural matings,” and new mutations artificially selected under captive conditions.

Genetic diversity in the Insurance Population

The importance of maintaining the pre-disease level of genetic diversity of wild Devils in the Insurance Population might lead observers to believe this parameter is estimated directly by genetic typing the Devils in the IP. It appears, however, that genetic diversity is “inferred from pedigree analysis (Lees and Andrew, 2012).”

Or is it? In 2010, a review article said, “each animal entering an insurance colony is typed for its MHC alleles with the intent of maximising genetic diversity for these loci (Bender, 2010).” The reference cited as the basis of this assertion (“TDSC, 2007” = DPIWE, 2007b in this work), however, contains no such indication.

Further, the STDP Annual Report for 2011/12 noted that 20 percent of the Insurance Population had been genotyped and the target for 2012/13 was 90 percent. There was no indication, however, what the genotyping involved, who was doing it, or how the results related to the published estimates of genetic diversity in the IP.

6/05/2016 8:41 AM 245

The DPIPWE’s 2015 annual report noted that an annual survey of the IP’s genetic diversity is taken following the breeding season and is related to the genetic diversity of the founding population (DPIPWE Annual Report 2015). As this founding population was heavily biased toward the northwest part of the state, it may not be representative of the genetic diversity in the entire population.

As mentioned (see “Income,” above), in July 2012, the University of Sydney launched an appeal for funds for “a new project to sequence genomic DNA from all of the Tasmanian devils in captivity (Gill, 2012).” It is unclear, however, how this effort fits in with the genotyping program alluded to in the STDP Annual Report for 2011/12.

However genetic diversity is measured in the Insurance Population, the STDP believes “more than 99 per cent of genetic diversity has been retained within the … population (Boland, 2012; see also DPIPWE Annual Reports, 2010/11, 2011/12 and 2012/13, 2014 and 2015; STDP Annual Reports, 2010/11, 2011/12 and 2012/13; P. Andrew, in Lindert-Wentzell, 2012; Hibbard et al., 2012; STDP website: “Generosity to save the Tasmanian Devil,” first published 13 September 2012; “Funding boost for the survival of the Tasmania Devil, first published 25 April 2013; Newsletter, May 2013; Day, 2014; Mather, 2014b). No confidence intervals have been associated with any of the estimates. They are occasionally reported to two (Hibbard et al., 2012; Lees and Andrew, 2012; DPIPWE Annual Report, 2015; STDP Annual Report, 2012/13) or even three (Hogg, 2013) decimal places, however, suggesting whatever the method of estimation, it is very precise.

Eventually, it will be interesting to compare the results of the indirect (pedigree) and direct (sequence) methods of determining genetic diversity.

In 2012, a female Devil from the northeast corner of the state was introduced into the Insurance Population in order to add to the genetic diversity in the Insurance Population. Keepers claimed that she was “not related to any other Devil in the program (T. Faulkner, in the Gloucester Advocate, 27 November 2012; also A. Good, on ABC AM, 27 November 2012 and in Egan, 2012b).” Perhaps the female was the only Devil available from this population, but if there were genuine concern about the introduction of some of this population’s genes into the IP, a male would have been more efficient conduit.

It is worth noting that species-wide genetics diversity is a somewhat abstract concept, as it does not take into account the adaptive diversity of local populations. The goal of the Insurance Population seems to be to maintain overall species genetic diversity while effacing locally adapted genetic variation. Arguably, the latter aspect of genetic diversity is as important to the Devil’s ability to exist in the wild as is its overall genetic diversity. It remains to be seen whether Devils from a population with a “one- size-fits-all” genome can survive when re-introduced back into the wild.

Finally, in 2013, the STDP mentioned that “a small proportion of the Insurance Population devils have a chromosome instability syndrome (STDP Annual Report, 2012/13). No further details have been published.

6/05/2016 8:41 AM 246

Source areas and animals

Devils for the Insurance Population have been sourced from areas in the east, southwest and northwest, where populations were thought to be disease free. As additional precautionary measure, only “recently weaned” juveniles were taken from the wild and then held in quarantine for a period, variously reported as 12 to 24 months, prior to being distributed to an Insurance Population facility (DPIWE, 2005a- b; Darby, 2006; Jones et al., 2007a; C.E. Hawkins, pers. comm., in Department of Sustainability, Environment, Water, Population and Communities, 2010; Sinn et al., 2010). More recently, however, wild Devils of unstated age were transported to a large free-range enclosure (below) after a quarantine of only six months (ABC News, 30 April 2011).

The recently weaned Devils are said to be chosen for their health and strength (Brand Tasmania Newsletters, February and April 2009). These selection criteria, of course, represent another form of artificial selection inflicted on both the wild and captive populations.

Quarantine/captive management facilities

Devils have been quarantined at seven different locations in Tasmania. Originally, these sites were called “quarantine facilities,” but subsequently, they were called “captive management facilities.”

The oldest facility is the Animal Health Laboratories at Mt Pleasant in Launceston. This is the state’s general quarantine facility and was where some of the earliest work on diseased Devils was done. Devils entering this facility are never released.

The remaining six facilities seem to have been dedicated solely to Devils. Three facilities seem to have been set up to receive young Devils for the Insurance Population. Maria Island off the east coast was to receive only females from the eastern part of the state. A private property at Lauderdale received only the males from the east. And Taroona, a suburb of Hobart, received Devils of both sexes from the west coast (DPIWE, 2005b-c; Parks and Wildlife Service News Article, 16 May 2005; Hobart Mercury, 27 May 2005; Albion, 2006; Kempton, 2010).

The Maria Island facility was finished in May 2005 (DPIWE, 2005b) and appears to only been used in 2005, despite a confusing comment in the September 2012 Newsletter that “between 2005 and 2008, Maria Island played an important quarantine role for Tasmanian devils when the disease was first noticed [!] but not well understood.” Maria Island was mentioned specifically as having received Devils in DPIWE’s annual report for 2004/05 but was conspicuously absent in the 2005/06 report. Its last mention in the media was November 2006 (ABC 7:30 Report, 20 November 2006; see also Loh, 2006). The one shipment in 2005 appears to have consisted of either nine or 13 females, depending on accounts, from Narawntapu National Park and Southport (Darby, 2005b; DPIWE, 2005b; DPIWE Annual Report 2004/05, 2005c; Hobart Mercury, 27 May 2005; Kempton, 2010). Why Maria Island was apparently dropped as a quarantine facility was never explained. In any event, in 2012, Maria Island made a comeback of sorts when it held, very briefly, the animals to be released onto the island (Newsletter, September 2012; see Off-shore Islands).

6/05/2016 8:41 AM 247

The Lauderdale quarantine facility seems to have accommodated only eastern males, perhaps four, some or all of which may have been from Southport and possibly Narawntapu NP. Interestingly, in September 2005, at least three females from the Maria Island facility were sent to Lauderdale for breeding purposes (Albion, 2006). The Lauderdale quarantine facility, if it was that officially, seems to appear only once in the public record (Albion, 2006), with one exception. In the draft version of the Zoo and Aquarium Association’s 2012 “Annual Report and Recommendations,” which may or may not have been intended for public release (below), Lauderdale was listed as a current DPIPWE facility holding two male and two female Devils (Hibbard et al., 2012).

Construction of the purpose-built Taroona facility finished in May 2005 and remains active (DPIWE, 2005b; Denholm, 2010b; Kingston et al., 2014). In time, it became “the central location for exporting captive devils to participating institutions [in the Insurance Population program] on the mainland (Newsletter, September 2012).” It and two other facilities in the Hobart suburbs of Fern Tree and Richmond are also used to hold experimental animals used in research (Brown, 2013: tables 8.2-5). All three facilities belong to DPIPWE.

The DPIPWE’s research and demonstration station near Cressy in the northern Midlands has been a quarantine station for Devils since at least 2006, when it served as an extension of the quarantine facilities at the Animal Health Laboratories at Mt Pleasant (Albion, 2006). Subsequently, it began to “quarantine apparently disease-free devils brought in from the wild,” especially those Devils removed in “de-populate/re- populate” programs (below; Newsletter, 2012; also Sharman, 2012a; Kingston et al., 2014). Cressy also apparently serves as a captive breeding facility (STDP website: “DPIPWE/ZAA Tasmanian Devil Insurance Population Workshop, 25-26 November 2010,” 26 November 2010 and “One mining company’s special delivery to Tasmanian Devils,” first published 25 July 2012).

An incident in 2010 suggests that the quarantine and export protocols put in place to control the spread of the disease failed spectacularly on at least one occasion. In May 2010, two Tasmanian zoos outside the Insurance Population program, East Coast Natureworld and Tasmania Zoo, sent six young Devils to two mainland zoos, which were also not part of the program. Two young went to Hunter Valley Zoo in New South Wales and four to Peel Zoo in Western Australia. And in at least one case, there was a “management fee” of $ 5000 for each of four Devils. The Tasmanian government approved the export of the animals but when the case was brought to public attention, the government decided to henceforth limit export only to facilities in the program. Another three young Devils destined for Phillip Island Zoo, also not a participant in the Insurance Population program, apparently got caught up in the government’s change of mind (Duncan, 2010c-d; Towie and Duncan, 2010; The Advertiser, 26 May 2010).

Zoos: Australia

Zoos, because of their existing facilities and experience, were the first to receive Devils, and the first Devils in the program were shipped to four mainland zoos in December 2006 and January 2007 (Darby, 2006; Denholm, 2006b; Newsletter,

6/05/2016 8:41 AM 248

February 2007). In May 2013, 21 mainland and Tasmania zoos and quarantine facilities were said to hold Devils (Newsletter, May 2013; see also Boland, 2011; Newsletter, September 2011). And in December 2014, 36 zoos held Devils (STDP, 2014c).

It is difficult at any one time to reconcile the number of zoos that officially participate in the Insurance Population and therefore potentially hold Devils. There are only two official listings and they differ. The STDP website lists 26 zoos as being IP members (STDP website, “Insurance Population,” published 8 April 2008, last modified 10 June 2014 and accessed 12 February 2015), while the Zoo and Aquarium’s website lists 25 (ZAA website, “Tasmanian Devil Insurance Population,” accessed 12 February 2015, Kyabram Fauna Park is omitted). Amusingly, both lists mention “George Wildlife Park,” presumably in error for “Gorge Wildlife Park.”

In December 2014, the STDP said 36 zoos (including wildlife parks) were participating in the Insurance Population program (STDP, 2014c). Perhaps the difference between this number and the numbers 25 or 26 (above) can be accounted for the number of zoos has approved that for Devils and those have actually received them.

According to the STDP, the zoos currently comprising the Insurance Program and hence presumably holding Devils are: Altina Wildlife Park, Australian Reptile Park, Australian Walkabout Wildlife Park, Devils Ark, Featherdale Wildlife Park, Symbio Wildlife Park, Taronga Western Plains Zoo, Taronga Zoo in New South Wales; Australia Zoo, Currumbin Wildlife Sanctuary, Dreamworld and Lone Pine Koala Park in Queensland; Adelaide Zoo, Cleland Wildlife Park, Gorge Wildlife Park and Monarto Zoo in South Australia; Devils@Cradle, and Trowunna Wildlife Park in Tasmania; Ballarat Wildlife Park, Halls Gap Wildlife Park and Zoo, Healesville Wildlife Sanctuary, Kyabram Fauna Park, Melbourne Zoo and Moonlit Sanctuary and Wildlife Conservation Park in Victoria, and Peel Zoo and Perth Zoo in Western Australia (STDP website: “Insurance Population,” accessed 12 February 2015).

One mainland zoo holding Devils implied for a few years that it was part of the Insurance Program, when it was not (Duncan, 2010c; PerthNow, 26 July 2010; Mandurah Coastal Times, 13 July 2011; Weekend Courier, 15 July 2011; Zaw, 2012). It was not accepted into the program until 2013 (Murray Mail, 2 October 2013).

Three other zoos were slated be added to this list in the near future: Alma Park Zoo and Flying High Bird Sanctuary in Queensland; Tasmanian Zoo in Tasmania (Hibbard et al., 2012: table 4).

Zoos rarely give enough information about Devil numbers and enclosure sizes to calculate the densities at which they keep Devils. There are some insights, however. For example, the Altina facility apparently maintains at least three Devils in a 200 m2 enclosure, which equates to a density of 15 000 Devils per km2 (data in ABC News, 28 December 2012).

The first arrival of animals at a new zoo is often greeted with glowing publicity, as are most other subsequent “good news” and human interest stories such as a suspected incipient breeding (ABC Newcastle, 10 February 2011), births (ABC, World Today, 5

6/05/2016 8:41 AM 249

April 2007) and old-age retirements (Bigpond News, 21 December 2012; Abey, 2014). Some zoos even announce “hopes for breeding (ABC News, 17 February 2015) and “signs of mating” in their Devils (Clifton-Evans, 2012). And, of course, name-the-young Devil story is a favourite of zoo media departments (Zoos Victoria website, “Devil of a name,” 4 November 2011; White, 2013). There is, however, no information in the public domain to allow evaluation of how the individual animals in all the facilities are faring and hence determine how “well” this part of the program is working.

Animal keepers often use breeding success as an indication of how good their facilities and husbandry practices are. There is no detailed accounting of how successful individual facilities in the Insurance Population are with breeding. Most zoos just announce the total number of young in a breeding season and not the number of adult females they tried to breed or the average number of young per mother (ABC News, 14 and 22 December 2011; Gloucester Advocate 22 December 2011). Interestingly, however, as of the end of 2012 and to judge from media reports, of the 21 zoos with Devils in the Insurance Population, only 12 have had a successful breeding while nine appear to have no breeding at all as yet, although some of these have only recently joined the program. Of the 12 successful facilities, two or three appear to account for most of the annual breeding.

The main problem with zoos, of course, is that they expose the animals to a plethora of artificial conditions (Jones et al., 2007a; H. McCallum, on ABC World Today, 5 April 2007; H. McCallum, in Marks, 2007; H. McCallum, on ABC Radio, 2 June 2008; McCallum and Jones, 2010; M. Jones, ABC Science Show, 3 September 2011; Keeley et al., 2012d), which any school child with an interest in Devils could identify. These conditions can affect not only individual behaviour in the short term but also species behaviour in the long-term through artificial selection, that is, a kind of domestication (Jones et al., 2007b). In terms of short-term behaviour, zoos are commercial organisations with families as clients. Consequently, they have to ensure that their Devils are active during the day, eat only a narrow range of pre-killed food, do not fight to the point of serious wounding, live in densities and social structures unlikely to occur in the wild (Newsletter, March 2010). Indeed, some zoos seem to regard the Devil’s nocturnality as something that must be overcome. The Albuquerque Zoo said upon the arrival of four recently imported Devils, “zookeepers will be working on introductions between the animals, and because they are nocturnal, increasing daytime activities (ABQ News Seeker, 20 December 2013).”

In the wild, Devils are usually fearful of humans, “lying motionless when handled and not struggling when manipulated (STDP Annual Report, 2011/12).” In captivity, however, Devils quickly lose their fear of humans (Kingston et al., 2014). And why wouldn’t they, because their keepers are a source of food and are unlikely to traumatise them in any way? One keeper noted that the only aggressive individuals she had encountered had been hand-reared and therefore lacked a fear of humans (A. Anderson, in Owen and Pemberton, 2005). Another keeper noted that even Devils kept in large “free-range” enclosures become “less fearful, more aggressive and struggled more,” the more they were handled (Sinn et al., 2010; see also Siegel, 2012).

6/05/2016 8:41 AM 250

Zoos affect reproduction. Compared to wild-born Devils, zoo-born Devils produce fewer joeys per female, significantly so if the female is zoo-born (Hogg et al., 2015: fig. 2), and among zoo-born Devils, successive generations have fewer joeys per female (Hogg et al., 2015). For example, the mean number of joeys per female for F0, F1, F2 and F3 females was: 2.9, 2.4, 2.3 and 2.2, respectively (Hogg et al., 2015). The number of joeys per female is also significantly lower zoo females than wild females (Hogg et al., 2015).

As noted, some zoos even “hand-rear” at least some their young (Bevin, 2007; Edwards, 2012a; Adelaide Now, 14 January 2012; Herald Sun, 14 January 2012; Sunday Mail, 15 January 2012; Albert and Logan News, 19 October 2012; Hartigan, 2013; White, 2013; Gloucester Advocate, 12 March 2014; Griffiths, 2014). This disrupts the interaction between mother and young, which may perturb their respective their hormonal states. Even more importantly, it may deprive the young Devils of the compounds in their mother’s milk that provide immunological protection (Hewavisenti et al., 2014).

Some zoos keep all the weaned young from one breeding season together in a crèche until they become mature at an age of two years old (Taronga Conservation Society Australia website: “Tasmanian Devil breeding commences,” 25 February 2011). There is, however, no evidence that such aggregations occur in the wild.

Other zoos, as part of their commercialisation of Devils, have special behind the scenes tours, during which visitors enter the animals’ free range enclosures, watch them lured out with food (the tours are during the day) and handle young-of-the-year Devils (Devil Ark website: Devils in the wild tours announced; Newcastle Herald, 2 December 2013; Gloucester Advocate, 3 December 2013; ABC Newcastle, 13 December 2013).

In terms of long-term evolution, Devils in zoos receive health care and “second chances (Ellis, 2012).” Indeed, such interventions would be mandated by animal welfare legislation, to say nothing of public opinion. In addition, some zoos with part of the Insurance Population are in climates very different from those in Tasmania. These generally warmer climates will exert additional selection pressures despite attempts to mitigate them.

As an example of exigencies imposed by an inappropriate climate, Altina Wildlife Park, just outside Griffith in interior New South Wales, is planning ponds for animals to bathe in, overhead sprays and even packs of ice in dens and burrows (ABC News, 28 December 2012).

The director of Tasmania’s conservation division summarised the problem concisely when he said that zoo animals would be “… incompetent to survive in the wild … (Duncan, 2010d).” And an experienced Devil researcher has noted, "animals change when they live in captivity and they adapt to become not a domestic species but a captive species (M. Jones, on ABC News and World Today, 6 July 2011).”

On a hot November afternoon, the STDP’s overseeing Steering Committee (above) visited the government’s Cressy Wildlife Centre and saw Devils “playing, sleeping, swimming and showing a keen curiosity in their visitors (Newsletter, March 2012).

6/05/2016 8:41 AM 251

One wonders if the committee members understood the implication and appreciated the irony in the Devils’ diurnal activity and lack of fear of humans.

Another problem with zoos is that the species will lose their natural parasites and commensals and pick up exotic ones (Jones et al., 2007a; McCallum and Jones, 2010; I. Beverage, in Mounster, 2012a). Indeed, in the early days, at least, Devils being held in quarantine prior to being shipped to mainland zoos as part of the insurance program were “wormed (Denholm, 2005; see also Albion, 2006).”

Considering the high level of “artificiality” of captive conditions, it should come as no surprise that the bacterial composition of captive Devils differs significantly (statistically) from that of captive Devils, although the difference is less for free- ranging captivity than intensive captivity (Cheng et al., 2015).

One Tasmanian zoo has now kept Devils in captivity for 15 generations (Tasmanian Examiner, 24 July 2014; see also, STDP website: “Trowunna devils join Insurance Population,” 19 August 2011; Newsletter, September 2011; 14 generations is claimed in one publication – KOB.com, 29 September 2015). Such long genealogies have the potential to provide valuable insights into the effects on lineages of long-term captivity (Hobart Mercury, 14 February 2012).

Life on public display in zoos also has implications for Devils’ health. One experienced keeper noted years ago “Like most marsupials Devils cannot stand continuous exposure to full daylight. Ultimately this causes them to become blind (Fleay, 1935).” Another keeper noted they might also suffer from solar dermatitis, sunburn and heat stress, as well as “pre-mature loss of vision (Kelly, 1993).”

Zoos usually justify their existence on the basis of their contributions to conservation, education and research. With regard to the latter, in the entire history of published research on the Devil, that is between 1808 and 2012, there have been 22 research publications based on Devils in zoos. Thirteen publications had an Australian contributor; the remaining nine had no Australian contributor. Fourteen publications were on basic biology and eight on health, usually cause of death in captivity and hence were not contributions to the Devil’s biology per se. Since DFTD first became know publicly in early 2003 and large numbers of Devils being placed in zoos in late 2006, there has been only one zoo-based publication. It was from Australia and dealt with an aspect of the Devil’s reproductive biology (Keeley et al, 2012d).

Free-range enclosures

Large free-range enclosures (FREs) were started in order to mitigate the acknowledged effects of the highly artificial environment of zoos (Newsletter, June 2010, March 2011, June 2011, May 2013). In Tasmania, this program is called the Devil Island Project (STDP website: “Devil Island Projects check out the Program’s facilities,” 8 July 2011; DIP Newsletter, December 2012; Newsletter, May 2013) and on the mainland, in NSW, it is called Devil Ark.

In Tasmania, the first free-range enclosure opened at Bicheno in the central eastern part of the state in 2008 (STDP website: “Three new free-range enclosures,” 16 December 2010). It is 12 ha in total area (STDP website: “Three new free-range

6/05/2016 8:41 AM 252

enclosures,” 16 December 2010; Newsletter, March 2011), but 11.3 ha when holding pens are taken into account (described as 11 ha in Newsletter, May 2013). The number of adult animals held at the Bicheno facility has been variously reported as 10 (Newsletter, September 2012, May 2013); 11 (Newsletter: June 2008, December 2009; Denholm, 2009; Newsletter, December 2009; DPIPWE, 2010d; Sinn et al., 2010; STDP website: “More free-range enclosures,” 16 December 2010; STDP website: “Devil island pinkies,” 16 December 2010) and 14 (Newsletter: December 2008). Ten Devils on 12 ha equates to a density of 83 Devils per km2.

The Bicheno FRE was stocked with Devils in between 2008 and February 2009 (Sinn et al., 2010). In an early media report, the proprietor of the FRE implied that the stocked animals would include two females with their young from the Badger Island removal (Radio National, Innovations: Devil Island Project, 26 May 2008; see “Distribution” above). The first annual report on the project, however, said all the stocked Devils, ultimately five males and six females, were collected from the wild on the northwest coast of Tasmania as juveniles in early 2008 and quarantined for one year (Sinn et al., 2010).

The second FRE was built at Bridport in the northeast part of the state in 2010 (Newsletter, March 2011) and officially opened in April 2011 (Hobart Mercury, 30 April 2011; McShane, 2011). Its total area has been reported variously as 22 ha (Newsletter: June 2010, September 2010, March 2011, June 2011, September 2012, May 2013; STDP website: “Three new free-range enclosures,” 16 December 2010; Healthy devils to return to the north, 16 December 2010; STDP Annual Report, 2010/11), 23 ha (Brand Tasmania Newsletter, September 2010) or 24 ha (Hobart Mercury, 30 April 2011; McShane, 2011). Initially, it housed 18 (Newsletter, March 2011), 19 (STDP Annual Report, 2010/11; Newsletter, September 2012) or 21 (STDP website: “Founder devils home from mainland Insurance Population,” 9 December 2010) Devils but will eventually have 38 (McShane, 2011). As of May 2013, it held 22 Devils. Eighteen Devils on 22 ha equates to a density of 82 Devils per km2. The provenance of only six of the founding Devils, two adult females and four young, were revealed. The two adult females, now past their breeding prime, were among the first to be sent to the mainland to establish the Insurance Population and hence were probably originally wild-caught. The four young were captive bred in the zoo where the two females were held (STTDP website: “Founder devils home from mainland Insurance Population,” 9 December 2010; STDP Annual Report 2010/11; Newsletter, March 2011).

The third FRE was built in Coles Bay on the Freycinet Peninsula in 2010 (Newsletter, 2011) and officially opened in April 2011 (Hobart Mercury, 30 April 2011; McShane, 2011). Its total area has been reported variously as 22 ha (STDP website: “Insurance Devils settle into Freycinet FRE,” 24 February 2011, “Friends of the Tasmanian Devil Volunteer for Poo Pick-up,” 21 March 2014; Tasmanian government media release, 29 April 2011; Newsletter: March 2011, June 2011, September 2012, May 2013; STDP Annual Report, 2010/11), 23 ha (Brand Tasmania Newsletter, September 2010), 24 ha (Hobart Mercury, 30 April 2011; Lakin, 2011; McShane, 2011) or 27 ha (Newsletter: June 2010; STDP website: “Three new free-range enclosures,” 16 December 2010). The facility was originally intended to house 21 individuals (STDP website: “Founder Devils home from mainland Insurance Population,” 9 December 2010) but on opening it actually housed 18 Devils (STDP website: “Insurance Devils

6/05/2016 8:41 AM 253

settle into Freycinet FRE,” 24 February 2011; Tasmanian government media release, 29 April 2011; ABC News, 30 April 2011; Newsletter, March 2011; Devils to range free at Freycinet, Tasmanian government media release, 29 April 2011). In May 2013, it held 23 Devils (Newsletter, May 2013). Eighteen Devils on 27 ha equates to a density of 67 Devils per km2. The 18 founding Devils were wild caught on the west coast (ABC News, 30 April 2011).

A fourth FRE was foreshadowed in early 2011 for Tasmania Zoo at Riverside, Launceston in northeast Tasmania (McShane, 2011; Bolger, 2011; McShane, 2011). Its size was reported variously as 10 ha (Bolger, 2011; STDP website: “Zoo donates land for State’s fourth large scale FRE,” 4 May 2011; Newsletter: June 2011, May 2013 – as Launceston Zoo; STDP Annual Report, 2010/11) or 22 ha (Newsletter, September 2012). The FRE was originally hoped to open in September 2011 (McShane, 2011; Newsletter, June 2011); construction was “about to commence” in July 2011 (STDP website: “Devil Island Projects check out the Program’s facilities,” 8 July 2011; was “nearing completion” in September 2012 (Newsletter, November 2012) but actually opened in December 2014 (STDP Annual Report, 2013/14). It opened with seven male Devils being held until mid-2015 for release into the wild (STDP Annual Report, 2013/14). Initially, at least, its purpose seemed to be as a holding instead of a breeding facility.

At the end of 2011, the three established free-range enclosures in Tasmania were said to have “over 50” Devils (STDP Annual Report 2010/11) and at the end of 2012 they were said to have total carrying capacity of “approximately 60 (STDP Annual Report, 2011/12).”

On the mainland, the Devil Ark program is building a series of free-range enclosures at a 350 ha site within a 500 ha private property, Tomalla Station, at an elevation of 1300 m in the Barrington Tops in central New South Wales. When completed, this facility will have 13 to 40 enclosures ranging from 0.5 to 10 ha in size (Connell, 2010; Denholm, 2010d; Keene, 2011; ABC AM, 18 January 2011). It started with 45 to 48 Devils, with plans to have 360 Devils by 2016 (Denholm, 2010d; ABC, AM, 18 January, 2011; Kelly, 2011a; T. Faulkner, in Speight, 2012; Lees and Andrew, 2012; Mather, 2014a) and 900 or 1000, depending on accounts, by 2020 (Ball, 2009; Courtney, 2010; Newsletter, December 2010; Keene, 2011; Gloucester Advocate, 21 December 2011; T. Faulkner, in Speight, 2012). The first announced stocking rates envisaged six to eight Devils in enclosures of two to eight hectares (STDP Annual Report 2010/11). These rates represent a density of 75 to 400 animals per km2. In fact, as of October 2012, the facility had between seven and eight Devils in 3 ha enclosures and between four and seven Devils in 2 ha enclosures. These represent densities of 233 to 266 Devils and 200 to 350 Devils per km2 (data in Hibbard et al., 2012). These densities are higher than in other FREs by an order of magnitude. Apparently as many a 12 Devils can be kept in one enclosure (Devil Ark Blog, 5 February 2013).

A new FRE had been scheduled to open at Healesville in late 2012 (Hibbard et al., 2012), but as of mid-2013, it was still not opened. It will consist of “eight interconnected 0.5 ha enclosures” and will house 30 animals (Hibbard et al., 2012), a density equivalent to 750 animals per km2.

6/05/2016 8:41 AM 254

In considering the densities reported above, recall that in the wild, a “very high” density is consider to be ≥ 2 Devils per km2 (Hamede et al., 2008; above). Nonetheless, the STDP reports that “… devils can be housed in mixed sex age groups in large enclosures at densities up to 1.6 devils per hectare [= 160 Devils per km2] and still achieve successful breeding outcomes. At higher densities, breeding success rates decline (STDP Annual Report, 2012/13).”

Large free-range enclosures provide a more naturalistic environment than a zoo, but they are still highly artificial. For example, in the pioneering Bicheno enclosure, which is simply an enclosed grassland, the animals are fed, provided with dens, have little opportunity to catch and kill prey, are caught regularly for examination, and are maintained at densities two to three orders of magnitude higher than in wild populations. And at least some of the founding individuals already display changes in behaviour, such as being active during the day (Brown, 2009 online: accompanying video; STDP Website: “Maria Island translocation project,” 21 June 2010: photo; Newsletter, June 2010: photo; Sinn et al., 2010) and, in their interactions with humans, being less fearful, more aggressive and more difficult to handle (Sinn et al., 2010).

There is also some suggestion that the size of the enclosure and the stocking density at Bicheno may not be adequate to allow some animals to escape physical aggression (Sinn et al., 2010).

At the Devil Ark free-range enclosure, the Devils are treated for parasites, ticks and worms (Kelly, 2011c). These treatments are just another form of artificial selection. The treatment for worms is of particular interest, because presumably it will also kill any Dasyurotaenia robusta, an endangered, and therefore protected, species of tapeworm known only to occur in Devils.

Devil Ark hand-rears some of its young Devils and even seeks members of the community to “adopt” a joey (Devil Ark Blog, 13 September 2012; Kelly, 2013a).

Devil Ark groups juveniles together in crèches for a year so they “learn how to socialise and interact with each other (L. Vella, in Egan, 2012b).” But there is no evidence that juveniles live in groups in the wild. Similar forced crecheing also occurs in zoos (above).

Devil Ark also traps all its adult Devils four times a year for health checks (Devil Ark Blog, 27 July 2012), a procedure that involves handling the animals.

Considering all of the above, referring to the Devils in a FRE as a “wild population (Tyler et al., 2011)” seems like one of those increasingly frequent attempts to achieve success by redefinition.

Some important reproductive parameters for the functioning FREs are shown in Table 19.

Table 19. Reproductive parameters for three Tasmanian FREs and one mainland FRE based on a variety of public domain sources.

6/05/2016 8:41 AM 255

FRE Year Female Total Mean Total Weaned Participation Young1 Litter Size Bicheno 2009 3/6 = 50 % 8 2.7 7/8 = 88 % Bicheno 2010 3/3 = 100 % ? ? ? Bicheno 2011 ? ? ? ? Bicheno 2012 ? ? ? ? Bicheno 2013 1/4 = 25 % 4 4.0 ? Bridport 2011 ?/? = 55 % 19 ? 15/19 = 79 % Bridport 2012 6/11 = 55 % 16 2.7 ? Bridport 2013 7/12 = 58 % 23 3.3 ? Freycinet 2011 ?/? = 40 % 13 ? 6/13 = 46.3 % Freycinet 2012 2/9 = 22 % 6 3.0 ? Freycinet 2013 3/8 = 38 % 11 3.7 ? Devil Ark 2011 13/23 = 57 % 26 2.0 26/26 = 100 % Devil Ark 2012 14/23 = 61 % 40 2.9 ?

1. Number of young initially attached to teats, irrespective of their subsequent fate.

Overseas facilities

It was recognised early on there would be a problem housing the requisite number of Devils foreshadowed in the Insurance Population program. The number of Australian facilities is limited and so is their capacity. Therefore, either additional Australian facilities would have to be developed or the program would have to be expanded to include overseas facilities (Jones et al., 2007a; Marks, 2007; H. McCallum, in Glaetzer, 2008).

In 2007, media stories said that 15 to 20 zoos in Europe, North America and New Zealand might receive Devils (The Age, 22 September 2007; Australian Air Express News, August 2007; Darby, 2007a; Jones et al., 2007a; Marks, 2007; PVHA, 2008).

In late October 2008, the relevant Minister said he was still thinking about sending reproductively senescent Devils offshore as part of an awareness campaign (D. Llewellyn, in Glaetzer, 2008).

In January 2009, a very specific online news story reported that Albuquerque Zoo had just completed a $ 300 k 4900 sq. ft. exhibit to “… house a dozen Tasmanian Devils this summer (ABQTrib.com, 9 January 2009).” It also said the zoo “ordered” 12 Devils; didn’t have to “buy” them, but just had to fly them from Australia. Nothing further was heard of this initiative until August 2013. The cause of the apparent delay was never explained.

In August 2009, a media report said two zoos in New Zealand were seeking biosecurity approval to import Devils (Hobart Mercury, 28 August 2009), something they presumably would not have done without having been given some official hope of being successful. A Wellington (NZ) Zoo staff member may have been offering further insight into this initiative when he said his zoo and Auckland Zoo stood ready to take 20 post-reproductive Devils off Australia’s hands (ABC News, 28 September 2009).

6/05/2016 8:41 AM 256

Sometime prior to December 2009, the Vice-Director of the Copenhagen Zoo wrote that the owner of Trowunna Wildlife Park would “… continue his efforts to persuade the necessary people to start two populations outside Australia: one in Europe (based in Copenhagen) and one in the USA (“Toddy” on online ZooChat, 20 December 2009; more below).”

In 2012, reports about sending Devils overseas gathered pace. In June 2012, a media story reported Toronto Zoo’s executive director of conservation, education and wildlife as saying that Australia had indicated it wanted to set up a captive population again in North America to save the animals from extinction (Vincent, 2012). In September 2012 and somewhat more authoritatively, the STDP’s program manager was reported as indicating that shipment to overseas facilities was being considered not only to extend the Insurance Population but also to “allow us to raise awareness and funds overseas for other activities of the program (A. Sharman, in Lindert- Wentzell, 2012).” And the Zoo and Aquarium Association’s Annual Report and Recommendations for 2012 noted that the Association submitted to the Steering Committee a paper exploring the options for “globalising” the captive breeding program (Hibbard et al., 2012). By late September 2012, there was a plan to send two groups of ten Devils each to Orana Wildlife Park in Christchurch and Wellington Zoo in Wellington, New Zealand (Hibbard et al., 2012: table 4).

On 23 June 2013, the authorities made the first official announcement about sending Devils overseas. The program was initially called the “Ambassador Devils” initiative and then the Tasmanian Devil Ambassador Program (STDP Annual Report, 2013/14). Under the program, certain Devils from the Insurance Program that are no longer needed for the IP due to their age or adequate representation of their genetics are sent “on loan” to overseas zoos as “ambassadors” for the species. The intention of the program is implicitly “advocacy and awareness, not breeding (STDP Annual Report, 2013/14),” Initially, about 20 animals retired from breeding were to be sent on trial to up to five overseas zoos, three in New Zealand – Auckland Zoo, Christchurch’s Orana Wildlife Park and Wellington Zoo and two in the US - San Diego plus an unnamed zoo. Each zoo would get four animals.

If successful, this program would be expanded “over the coming years” to “up to 100 Devils” in “up to 10 zoos in North America and Europe, three in New Zealand and two in Japan” (Examiner, 23 June 2013; see also Milman, 2013a; Paine, 2013; Stewart, 2013; STDP website: Ambassador devils for overseas zoos, 24 June 2013; STDP Annual Report, 2013/14). This is a total of 15 zoos

The “ambassador Devils” are intended to raise awareness of Devil Facial Tumour Disease and to raise research funds. Interestingly, these research funds appear to have been targeted, initially at least, “to help find a vaccine or potentially a cure (B. Wightman, in Paine, 2013).”

On 30 August 2013, more details of the program’s rollout were forthcoming (Wightman, 2013; Poskitt, 2013). San Diego Zoo would receive four Devils in late October 2013; the second US zoo was named as Albuquerque Biopark in New Mexico, and this zoo and the three New Zealand zoos would get their four Devils each “later this year or early next year.” Eventually, ten zoos in North America and ten in Europe would get Devils, the timing “determined by the rate at which animals

6/05/2016 8:41 AM 257

become available from the insurance population.” It was also revealed that the Devils chosen for transport overseas would be either “genetically over-represented in their population [which population?]” or “at the end of their breeding life (Poskitt, 2013).” The former category suggests that at least some overseas zoos may be able to breed Devils. It was further revealed that the “international program” would contribute to the annual $ 5 million cost of maintaining the insurance populations (Poskitt, 2013).” This figure is interesting as it is the first public acknowledgement of the annual cost of the insurance populations.

In late September 2013, the San Diego Zoo received its four Devils, three males and one male (CBS.8com, 9 October 2013; see also Tassy, 2013). In return, the zoo made a $ 500 000 contribution to an STDP program (Imperial Valley News, 18 March 2014; Mounster, 2014; STDP Newsletter, April 2014).

In early December 2013, Wellington Zoo received its four Devils, either one male and three females (Voxy.co.nz, 4 December 2013) or three males and one female (New Zealand Herald, 5 December 2013; New Zealand Radio News, 12 January 2014), depending on reports. Parenthetically, Wellington Zoo last had Devils in the 1920s (Voxy.co.nz).

On 18 December 2013, Albuquerque Zoo received its four Devils, two males and two females, from the Healesville Zoo (ABQnews Seeker, 20 December 2013; Tassy, 2013). Interestingly, an Albuquerque zoo media release said it had been “preparing for the possible arrival of Tasmanian devils since 2002 (GardenNews.biz, 2 September 2013; see also ABQnews Seeker, 20 December 2013).” This was five years earlier than the first public indication that Devils were destined for any overseas zoo (above), and it is unlikely the Albuquerque administrators would have begun their planning without some official Tasmanian Government indication their efforts would be rewarded.

In April 2014, Auckland Zoo received its four Devils, three males and one female, from Healesville Zoo (3 News NZ, 18 April 2014; STDP website: “Auckland Zoo helps raise awareness of Tasmanian devils,” [published and accessed, 12 May 2014; DPIPWE Annual Report 2013/14).

Sometime prior to November 2014, the Copenhagen Zoo, which had received Devils prior to the ambassador program, was brought into the program (DPIPWE Annual Report, 2013/2014).

In January 2015, the Fort Wayne Childrens’ Zoo in Indiana announced it had been promised an unspecified number of Devils at an unspecified time pending completion of an appropriate enclosure (WLFI.com, “Fort Wayne Children’s Zoo to feature Tasmanian devils,” 28 January 2015; 13WHTR.com, “Fort Wayne Children's Zoo to feature Tasmanian devils,” 29 January 2015; Kilbane, 2015). In April 2016, a media report said the zoo hoped the Devils would arrive in August or September and that it would receive two or three Devils (Kilbane, 2016).

A further US zoo was officially added to the “ambassador devils” in September 2015, when it was announced that Toledo Zoo and Aquarium in Ohio would receive three Devils from the Insurance Program sourced from the Monarto Zoo in South Australia.

6/05/2016 8:41 AM 258

The Devils arrived on 14 October 2015 (Mester, 2015b) and consisted of two females, aged 2 and 3, and one male, 3. They were second generation captive bred (Mester, 2015a-b; Broadstock, 2015). The male had had a vasectomy to prevent breeding (Mester, 2015b, 2016).

In return for the Devils, the “zoo has agreed to fund a five-year, conservation research program in Tasmania monitoring devil populations in areas where the disease has decimated their numbers. The program also will examine the effect of low populations and of the introduction of new animals from the captive breeding program” (Mester, 2015b; see also Baines, 2015). The total amount over five years is $ 500 000 (Baines, 2015; STDP website, “International grant boost to Tasmanian devil monitoring,” 21 December 2015; ABC News online: “Popularity of Tasmanian devil cartoon character helps researchers raise funds to save species,” 23 December 2015: Mester, 2016).

On 9 December 2015, the Los Angeles Zoo, a facility not foreshadowed as receiving any Devils, got two three-year old male sibs from the Trowunna Wildlife Sanctuary in Tasmania (Jackson, 2015; Los Angeles Zoo and Botanical Gardens Press Release, “Tasmanian Devils are back at the L.A. Zoo after 20 years!” 14 December 2015; Los Angeles Daily News online, “2 Tasmanian devils on display at Los Angeles Zoo,” 14 December 2015; NPR, “Tasmanian Devils Return to Los Angeles Zoo,” 15 December 2015; 83.9 KPCC, “LA Zoo now hosts Tasmanian devils,” 18 December 2015).

In mid-April 2016, the St. Louis media revealed that on 23 March, the St. Louis Zoo received two two-year old sibling female Devils from the Dubbo Western Plains Zoo (Bock, 2016; Held, 2016; Nobbe, 2016). In return, “… the Zoo provides funds to Australia’s Zoo Aquarium Association Wildlife Conservation Fund supporting Tasmanian devil population monitoring and management (Bock, 2016). If this report is true, it would be the first time, apparently, that payment has been made to the ZAA instead of the STDP.

The terms of the deals to send Devils to overseas zoos are not available in the public domain. It is known, however, the Devils remain the property of the Tasmanian Government (Examiner, 23 June 2013; Paine, 2013; STDP Annual Report, 2013/14) and are only on “long-term loan” to the respective zoos (Perry, 2014). Perhaps most significantly, it is unclear what access overseas researchers have to these Devils or the terms and conditions of any access. It is also unclear what will happen to the Devils when they die. Local museums would be derelict in not asking for the bodies, but if successful, would the long-term loan continue or would ownership be ceded to them.

As mentioned, the “ambassador” Devils initiative serves two purposes. First, it provides an outlet for the Devils no longer needed in the Insurance Population, because either their genes are well represented in the IP or they are too old to breed (STDP Newsletter, April 2014).

Second, it provides a new source of private, no-questions-asked, funding for STDP projects, as can be seen from the “contributions/donations” of $ 500 000 each made by the San Diego and Toledo Zoos. These funding commitments are unlikely to have been spontaneous donations; rather, they were almost certainly negotiated as part of the deal to send the animals. The contributions were “timely,” as the Commonwealth

6/05/2016 8:41 AM 259

curtailed it long-term funding with the beginning of the 2013/14 financial year (ABC News, 30 August and 1 September 2013; ABC PM, 30 August 2013) to $ 3.3 million over three years (STDP Annual Report, 2013/14).

In late 2014, the projected number of zoos in the ambassador program was raised to 25 by mid-2019 (STDP, 2014c). At the ratio of 100 Devils in 15 zoos noted above, that comes to about six Devils per zoo or a total of 150 Devils in overseas zoos by mid-2019.

The “ambassador Devils” concept came back to Australia in mid-2015, when it was revealed that “Australia’s top luxury” hotel, a $1950+ per night lodge in Freycinet National Park, had installed, with STDP authorisation, an enclosure to house four Devils, at least one of which was six years old (Daily Mail Australia, 24 June 2015). Presumably, this local effort is designed to raise the awareness of the Devils plight among the hotel’s wealthy visitors, many of whom would no doubt be from overseas.

A retirement village for aging Devils out of the Insurance Population was also established on a private property in Tasmania (STDP website, “What a devil of a time!,”, 17 December 2015). Neither the location of the site nor the number of “retirees” it will house was made public.

Devils have previously been kept in zoos in Europe and North America, starting in 1866 in Scotland and ending in 2004 in the United States, then again from 2006 in the exceptional case of Denmark and .

The first Devil to be kept in captivity outside Australian appears to have been an animal imported into Great Britain in the first part of 1866 by Mrs Wombwell’s Great National Mammoth Menagerie, apparently a travelling animal show (The Caledonian Mercury, 5 February 1866). The first Devil kept in the Gardens of the Zoological Society in London apparently entered the collection on 21 June 1866 (Flower, 1931; see also Penny Illustrated Paper, 29 February 1868).

The last animal kept in captivity outside Australia was a male that died of cancer (not DFTD) at the age of seven and a half years on 18 May 2004 in the Fort Wayne (Indiana) Children’s Zoo in the United States (Rochester Sentinel, 19 May 2004; Mount Airy News, 20 May 2004; Fort Wayne News Sentinel, 20 May 2004; Fort Wayne Daily News, 15 February 2010; see also Associated Press, 13 September 2003; Cock, 2003; Kilbane, 2015).

Between 1981 and 1998, eight different Australian institutions sent 16 Devils to four different North American zoos - Chicago, Fort Wayne, San Diego and Toronto – although with inter-zoo transfers nine North American zoos had at least one Devil in this period (data in Koncza and Pryor, 2001).

For nearly two years, there were no Devils in overseas zoos. Then in April 2006, four Devils were sent to Copenhagen Zoo to celebrate the birth of the first child of the Crown Prince of Denmark and his Tasmanian wife (The Age, 17 October 2005; AP Newswires, 17 October 2005; Brand Tasmania Newsletter, November 2005; DPIWE Annual Report 2005/06; Sydney Morning Herald, 11 April 2006; Hobart Mercury, 12 June and 25 October 2006; Duncan, 2007; STDP website: “Trowunna devils join

6/05/2016 8:41 AM 260

Insurance Population,” 19 August 2011). Originally, only two Devils were to be sent, but the Danish authorities asked for two more in order to establish a breeding population (AFP, 10 April; Sydney Morning Herald, 11 April 2006; AFP, 24 October 2006). The Devils were sent as two “breeding pairs,” which means they were at least two years old when exported. At least two of the animals, and possibly all four, were captive bred (40º South Podcast, 1 July 2008).

The Australian press carried no further information about the four Danish Devils. There was, however, some news from local Danish sources (“Toddy” on online ZooChat, 20 December 2009 and associated comments). One female had one litter but the young died in the pouch. Both females showed teat damage (possibly from the early loss of suckling young) which may have made it difficult to support further young. Whatever the reason, neither female bred successfully. One female, reputedly aged eight years old and suffering from back problems, was euthanased sometime between August and December 2009, and a male appears to have been euthanased around the same time. The fate of the other two individuals is unclear, although they probably died sometime before July 2011.

On 11 October 2012, without warning, the Tasmanian government announced that four more Devils, two males and two females, were being flown to Denmark, apparently on that same day (Tasmanian government media release, 11 October 2012; Mounster, 2012a; Hobart Mercury Editorial, 13 October 2012; reported as being sent in November by ABC News, 14 June 2013; STDP Annual Report, 2012/13). The four Devils were captive bred (Tasmanian Government media release, 11 October 2012; ABC News, 14 June 2013; STDP Annual Report, 2012/13).

Interestingly, the government media release said nothing about the reproductive success or ultimate fate of the original four Devils, and the Hobart Mercury editorial writer even thought they were still alive. Only an ABC story noted the first four Devils had “failed to reproduce” and died of “old age” (ABC News, 11 October 2012). As late as mid-2013, the STDP was silent about the fate of the original four Devils, noting in its Annual Report for 2012/13 that “the animals follow a consignment of devils sent to Copenhagen in 2006,” without mentioning the fate of these animals.

More information on the first consignment of Danish Devils may have been contained in an August 2014 report on the immunohistochemical analysis of four neural systems in the brains of three Devils euthanased in the Copenhagen Zoo (Patzke et al., 2014). One Devil was euthanased due to a “prolapsed intervertebral disc” – possibly the Devil with the back problem (above); another for “metastic squamous cell carcinoma,” and the other for “senescence and dilated cadiomyopathy.” The Devils can have only been the ones Australia sent to Denmark in 2006.

The latest Copenhagen Devils reproduced in tune with the normal Tasmanian seasonal cycle. The two females mated in late February and gave birth to a total of seven young in April (Hobart Mercury, 2 April 2013; Kempton 2013a, c; The Examiner, 16 June 2013; STDP Annual Report, 2012/13; Glaetzer, 2014a). A Tasmanian media report claimed this was “the first successful breeding of the endangered animals outside Australia (Glaetzer, 2014a; see also Hobart Mercury, 2

6/05/2016 8:41 AM 261

April 2013; Kempton, 2013a),” but Devils had been bred in several North American and European zoos previously (see “Birth,” above).

Parenthetically, four Tasmanian media reports said two Devils had been sent to Denmark in 2006 and four more sent in 2009 (Hobart Mercury, 2 April 2013; Kempton, 2013a, c; The Examiner, 13 June 2013; Glaetzer, 2014a). The 2009 shipment is little mentioned in the media and might have been dismissed as a misunderstanding. However, a scientific paper that used Devils from this “first shipment” mentions that the three Devils it used were all male (Patzke et al., 2014). As one of the authors of this paper was the Copenhagen Zoo vet, it seems unlikely that one of the three Devils was a female mistakenly identified as a male. The fact of three males gives credence to there having been a second shipment, but if so, it received no media coverage at the time. In any event, it is within the capacity of the STDP to clarify this issue.

Interestingly, one of the original Danish Devils, “Montague,” was used in a subsequent genetics study (Miller et al., 2011) and as noted above, three were used in a brain study after having been euthanased for welfare reasons by the zoo (Patzke et al., 2014). Both studies, raise the issue of to what purposes any Devils sent overseas can be put and by whom. The Danish zoo vet, and subsequent co-author of both the genetics and brain studies, must have taken the samples (hair) from the Devil and sent them to the relevant lab in the US for genotyping and then arranged for the whole bodies or brains only to be sent to the relevant lab in South Africa for the brain study. Perhaps there is a clear agreement between the Tasmanian and the Danish governments regarding how these “gift” Devils will be used, but if so, it is not in the public domain.

The fact that the Danish Devils were “gifts” of the Tasmanian government (Tasmanian government media release, 11 October 2012), suggests they may be the only Devils not “owned” by the government and hence beyond the control it places on all other Devils and the research based on them.

Finally, it may be of interest to mention that Copenhagen Zoo, like most zoos, has a policy of killing animals that do not have the genetics deemed desirable for a species’ captive breeding program (Sydney Morning Herald, 9 February 2014; The Australian, 10 February 2014; Johnson, 2014; Milmo, 2014). Copenhagen Zoo apparently kills 20 to 30 animals a year for this reason (Keegan, 2014).

Fencing off natural populations

As yet, no healthy population has been isolated within its own habitat, mainly for reasons of cost and sustainability. The authorities have, however, considered fencing off three disease-free populations that inhabit large areas largely surrounded by sea: Woolnorth in the northwest; Robbins Island (connected to the main island by an intertidal flat) also in the northwest, and Cape Sorrell Peninsula on the west coast (McCallum and Jones, 2010). But the double fence required for the land part of the enclosure would be costly to build and maintain as it would have to be secure against potential breaches due to accidents (burrowing wombats, falling branches, flooding and erosion), cleverness (Devils) and malevolence (humans). It would also have to be acceptable to the locals.

6/05/2016 8:41 AM 262

At first, planning advanced most rapidly for Woolnorth, a 22 000 ha property in far northwest Tasmania with about 550 Devils (DPIPWE website: “Assessing the impacts of broad-scale fencing,” 26 August 2010; Darby, 2010a; Denholm, 2010e). The project’s feasibility study was submitted in March 2011 (STDP Annual Report 2010/11; Sharman, 2012b: time in video: 1:26:22) and work on the 11 km long fence was to begin in the 2011/2012 financial year (ABC News, 23 June 2011). In early November 2011, a decision was expected by the end of the year (A. Sharman, as reported in Darby, 2011). It then appeared as if the proposal had been abandoned (below; Darby, 2012a). But apparently not. In May 2012, a fund-raising initiative was dedicated to contributing to the fence’s $ 1.5 million dollar cost (Brand Tasmania Newsletter, June 2012). In mid-June 2012, the length of the planned fence was given as 17 km and the cost $ 1.2 million (Sharman, 2012b). At the end of 2012, the STDP was said to be “facilitating engagement with potential fund raising bodies (STDP Annual Report, 2011/12),” and the property’s owners still spoke as if negotiations were on-going (Van Diemen’s Land Company media release, 27 November 2012; Kempton, 2012b). In February 2013, the STDP’s Director said “… we’re working with the land-owner there to put a fence in place that hopefully will stop the spread of disease (Nogrady, 2013).” A little later, the STDP said work was “progressing (Newsletter, May 2013).” In July 2014, however, it was reported that the project was facing funding problems (Mather, 2014a). In January 2016, it was reported the company newly proposing to buy Woolnorth had been meeting with the state government regarding the creation of a "devil-proof fence" to protect the Devil from facial tumour disease (Jarvis, 2016). In February 2016, the new owner had given “in principle” approval to construct a fence (Clark, 2016; Denholm, 2016). When contacted for comment, however, the STDP is reported as having indicated that “the fence was not a priority (Denholm, 2016).”

In the STDP’s Annual Report for 2012/13, the Woolnorth project was first referred to as the “Woolnorth Isolation Project.”

An enduring problem with the Insurance Population strategies on Tasmania itself is the threat of a diseased animal getting into the isolated area, through either its own accord or being placed there maliciously. In the absence of a diagnostic test for the pre-symptomatic stage of the disease, were a diseased animal to be found in a captive group, it would compromise all the animals in the facility (N. Mooney, in Jaffe, 2007; McGlashan et al., 2007; Parkes, 2012; Huxtable et al., 2015). Or as two applied ecologists, addressing the issue of the most efficient use of fencing, wrote: “A fence breach would require the removal of every devil from the enclosure (because their infection status would be unknown) and their replacement by new, uninfected individuals, which would be sourced from uninfected areas or other insurance populations (Bode and Wintle, 2009).”

A breach into a facility designed to keep Devils “in,” and by implication other Devils “out,” is a not an imaginary concern. In 2006, the occurrence of facial tumour disease in three captive Devils at the Trowunna Wildlife Park was thought most likely to have been initiated when an amorous wild male got into the facility during the mating season and then slipped out again undetected (Andrewartha and Campbell, 2006).

6/05/2016 8:41 AM 263

Trowunna appears to have had a fourth case of DFTD among its captive Devils. In early 2009, a keeper discovered a mouth lesion in a six-year old female and this was subsequently confirmed as DFTD (ABC News, 2 April 2009; Scott, 2009; Warren, 2013). This case raised the important question of how the second inferred breach occurred, but no detailed investigation comparable to the first cases was forthcoming. For example, did the outbreak represent a recent breach of the zoo’s biosecurity, or if the infected animal was in the zoo at the time of the original breach, could it have experienced a very long latency period?

In 2012, a Devil that had been in Tasmania Zoo since June 2011 was detected with the disease in March 2012 (Dadson, 2012). Unless the animal had an undetected infection prior to its entry into the zoo, it would have contracted the disease while there, either from a resident Devil or an entering wild Devil. The details of this case, if ever investigated, have not been made public.

In fact, in May 2012, the STDP’s Science Manager mentioned parenthetically in a media interview that the proposal to fence-off Devils in Tasmania’s northwest corner, which presumably meant Woolnorth (above), had been abandoned because a single breach could compromise the whole plan (Darby, 2012a). However, the plan appears to have been re-instated since then, pending funding (above).

Finally, it is worth noting that once part of the natural landscape is fenced off, it not only impedes the movement of Devils but all other large mammals as well (N. Mooney, in The Commercial Appeal, 25 January 2009).

De-populate/re-populate

A variation on the idea of fencing off healthy populations of Devils is to fence off an area with a diseased population, remove all the Devils, kill the diseased animals and once the area is disease free put healthy Devils back, including any survivors or progeny of the removed healthy Devils. The re-population phase is sometimes called, euphemistically, “re-wilding” (Newsletter, September 2012; DIP Newsletter, December 2012; STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014; STDP Annual Report, 2013/14).

The first locality nominated for such a de-populate/re-populate program was the Freycinet Peninsula, an initiative that came to be called, informally, the Great Devil Wall (H. McCallum, in Denholm, 2009; see also ABC News and Radio Australia, 13 September 2012; Bolger, 2012; Lord, 2012b; McKay, 2013) and more formally the Freycinet Landscape Isolation Project, or FLIP (STDP Annual Report, 2011/12) or the Freycinet Isolation Project (STDP Annual Report, 2012/13). An early report said the required fence across the top of the peninsula would be six km in length (Bolger, 2012), but a later, presumably more authoritative, “spade ready” plan for the project called for a 4.8 km fence (DIP Newsletter, December 2012; see also Sharman, 2012b, where the length is given as 4 km). The reported maximum number of Devils to be held in the Freycinet enclosure varies from up to 150 (Bolger, 2012) to up to 200 (Denholm, 2014). In late November 2014, however, the Tasmanian Government announced it was delaying the Freycinet project “for further consideration at a later time” and “… pending a review of the Tasman (and Forestier) project” (Groom, 2014;

6/05/2016 8:41 AM 264

see also STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014; STDP, 2014b).

Whether the Freycinet Peninsula is still targeted for de-population, however, may have been reconsidered, as the peninsula is now apparently slated for a long-term monitoring program to “…reveal its underlying demographic and epidemiological mechanisms of survival (Comte et al., 2015).

Parenthetically, it appears the name “The Great Devil Wall of Tasmania” is being formally applied to both the Freycinet and Woolnorth fencing projects (DIP newsletter, December 2012, April 2013).

In 2012, the Forestier-Tasman Peninsula was apparently leap-frogged to the head of the de-pop/re-pop queue (ABC News, 25 May 2012; Newsletter, September 2012; see also McKay, 2013, but see Newsletter, May 2013), probably because the tandem peninsula already had many of its diseased Devils removed from its northern part, Forestier (see “Disease Suppression,” below).

The project was first named the “Tasman Landscape Isolation Project” in a Tasmanian Government media release in late November 2014 (Groom, 2014; STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014). By December 2014, however, the name had been changed to the “Peninsula Devil Conservation Project” (STDP Annual Report, 2013/14).

The two peninsulas were to be isolated from the main part of the island by three barriers. First, is the existing trans-isthmus Denison Canal at Dunalley. Second, is a “devil-proof” grid on the bridge over the canal (Mather, 2014b; STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014; “Peninsula Devil Conservation Project,” 23 July 2015). And third, is an 800 m long “devil-proof” fence erected along Annie Street about 400m south of the canal (Newsletter, April 2014; Kingston et al., 2014; Huxtable et al., 2015), which interestingly enough replaced an earlier barrier fence destroyed in the January 2013 Dunalley bush fires (STDP Annual Report, 2013/14). The fence will also be extended into the inter-tidal reaches of both sides of the isthmus (STDP website: “Peninsula Devil Conservation Project,” 23 July 2015 ).The STDP is also trialling deterrents such as dark barking recordings and air pressure guns to deter Devils from crossing the barrier area (STDP Annual Report, 2013/14). The planned completion period for this “buffer zone” was July 2015 (STDP Annual Report, 2013/14).

In May 2012, the STDP began removing all the Devils on the two peninsulas (Sharman, 2012b; Bangor – partners with the Save the Tasmanian Devil Program, accessed online, 15 March 2013; see also Kingston et al., 2014). Healthy Devils were quarantined until the peninsula was “clean,” while DFTD-affected Devils were killed (Kingston et al., 2014).

In mid-2015, the STDP was developing a statistical model to assess the likelihood that all the Devils have been removed from the peninsula (Huxtable et al., 2015). Once cleared, the STDP expected to start putting 50 healthy Devils back onto the peninsula in late 2015 (Manager of the STDP, on ABC, The World Today, 26 November 2013;

6/05/2016 8:41 AM 265

see also ABC News online, 8 July 2014; Groom, 2014; Kingston et al., 2014; Mather, 2014b; Ray, 2014; STDP, 2014c; Huxtable et al., 2015).

Late on 18 November 2015 and depending on reports, 39 (STDP website: “Forestier Peninsula poised for devil release,” 18 November 2015; “Tasmanian devil wild release monitoring update - December 2015,” 24 December 2015; Groom, 2015b; Richards, 2015a; Ryan, 2015; probably the official number) or 40 (Devil Ark Newsletter, Spring 2015; Noone, 2015b) Devils were released at four localities on the Forestier Peninsula (see also: ABC News online, 17 November 2015; STDP website: “Peninsula Devil Conservation Project,” 23 July 2015, “Devils free-ranging on Forestier,” 19 November 2015; Huxtable et al., 2015; Kelly, 2015). As had been foreshadowed, the released animals were partly the progeny of the indigenous Devils removed and held in quarantine and partly “… from selected genetic lines within the insurance population …” (Kingston et al., 2014; STDP, 2015b). Hence, the reconstituted population contained individuals that had undergone the artificial selection of captivity and had “blended” genetics.

Within eight days of their release, however, two (Anonymous, 2015; Glumac, 2015) and then four (Darby, 2015c; SBS News online, “Plea to drivers after more dead devils,” 26 November 2015) of the 39 Devils were reported as having been killed on the peninsula’s roads. Within a month, the toll had climbed to nine (ABC News online, “Captive-bred Tasmanian devil program loses quarter of released animals as roadkill,” 15 December 2015; on 14 December 2015, one report had the total at only three – Richards, 2015b), then to 11 (STDP website: “Tasmanian devil wild release monitoring update - December 2015,” 24 December 2015: Richards, 2015c) and finally 12 (Mather, 2016b; STDP website: “Tasmanian devil wild release monitoring update - January 2016,” 29 February 2016). This attrition rate, 12 out of 39 or 31 percent in less than a month, would be several orders of magnitude above that of a naturally occurring local population.

The Devils had been released “about 10 km from the nearest road” (Mather, 2016a), suggesting that the first four Devils killed had travelled at least 1.2 km per day (10 km/8 days). At least one of the Devils had travelled up to 15 km from its release site before being killed, and an STDP official was reported as saying that “… animals that travelled furthest from their release site may be more likely to die on roads (Darby, 2015c).”

In November 2015, the STDP revealed that a further ten Devils from the Insurance Program would be released on the Forestier Peninsula in “early 2016” (STDP website: “Peninsula havens provide natural barriers to devil disease,” 6 November 2015 and “Helping our devils fly home,” 11 November 2015; STDP, 2015b). In late February 2016, ten “juvenile” Devils were released at Bangor Farm. By 29 February, however, two of the ten (20 percent) had been road killed near Murdunna (STDP website, “Forestier Peninsula Release - Update 29 February 2016,” 29 February 2016; Richards, 2016).

A media report of 19 November (ABC News online, “Fence built to keep newly released Tasmanian devils safe from roads) stated the fence across the top end of the Forestier Peninsula would keep the released Devils “off the roads.” This was, however, never the fence’s purpose; its purpose is to help keep diseased Devils from

6/05/2016 8:41 AM 266

the north from entering the peninsula. The report also stated, uniquely, that at least some of the released Devils had been inoculated against DFTD. This was almost certainly incorrect, as none of the associated reports, including the Tasmanian Government’s media release (Groom, 2015b), mentioned it.

Planning for the de-and re-population of the D’Entrecasteaux Peninsula was to begin in 2012/13 (STDP Annual Report, 2011/12), but there has been little subsequent information on this part of the de-pop/re-pop program.

The STDP’s business plan for 2014-19 lists two “landscape isolation” projects for completion by mid-2017 (STDP, 2014c); presumably, these will be the Forestier/Tasman and Freycinet projects.

Off-shore islands

Another early idea for isolating healthy Devils from the disease was to introduce them to one or more of Tasmania’s 334 larger off-shore islands, an idea first mentioned publicly by wildlife ecologist Nick Mooney (Wood, 2003c) and soon thereafter supported by a politician, J. Rockliff, in the Tasmanian parliament (Martain, 2003; see also Jones, 2003b).

Apparently, there were two efforts to identify suitable offshore islands, the one in 2003 just mentioned and another one in 2008 (Huxtable et al., 2015).

From the very beginning, the island of choice seemed to be Maria Island, some 4 km off the east coast (N. Mooney, in MX (Australia), 2 September 2003 and Wood, 2003c) and 96.7 km2 in area (STDP, 2011b). In February 2005, the Tasmanian government seems to have formally committed to Maria (ABC News, 4 February 2005; Whinnett, 2005a-b; Hobart Mercury, 25 October 2006). And in 2006, the Tasmanian government had already done an environmental risk assessment on the introduction of Devils onto the island (Jones and McCallum, 2007). One of the attractions of the island was that it has only one owner, the Tasmanian government.

Initially, the program to move Devils to Maria had no formal name, but in June 2010, the STDP began calling it the Maria Island Translocation Project (MITP) (Newsletter, June 2010, September 2012; STTDP website, “Maria Island Translocation Project,” 9 June 2010).

The timeline for the translocation was been pushed back repeatedly. Initially, it was thought the animals could be translocated as early as mid-2007 (Byrnes, 2007; see also Jones and McCallum, 2007). The STDP’s formal proposal document envisaged beginning the translocation in December 2011 (STDP, 2011b). However, the public consultation process did not open until March 2012, with a closing date in late April 2012 (ABC News, 23 March 2012). The first release was anticipated to occur in the “coming months (Martin, 2012a),” and “in 2012 (STDP 2012a).” It actually took place on 14 November 2012 (Tasmanian government media release, 14 November 2012; ABC News and ABC Hobart, 14 November 2012; Darby, 2012b; Martin, 2012b-c; STDP Annual Report, 2012/13).

6/05/2016 8:41 AM 267

Maria Island has endangered species, and therefore the introduction of an animal not native to the island required a permit from the Commonwealth government. In January 2007, the Tasmanian government made a proposal to the Commonwealth to “capture Tasmanian devils to establish an ‘insurance population.’ ... [This] involved the capture of up to 30 newly independent juvenile Tasmanian devils … to establish a … population at Maria Island or another suitable location (Department of Environment and Heritage Annual Report, 2006-07, vol. 2).”

The Commonwealth apparently gave its approval some time not long before 5 March 2007, because on this date the then environment minister said, “We’ve recently approved a proposal by the Tasmanian government to capture 30 disease-free animals, ahead of a possible move to Maria Island (ABC, World Today, 5 March 2007).” Subsequently, the media reported widely the proposal to send 30 healthy Devils to Maria Island (ABC, World Today, 5 April 2007; Byrnes, 2007; Cordingley, 2007a; Marks, 2007; McQuirk, 2007; Pincock, 2007).

Despite this approval, in April 2008, the STTDP said it was “…currently preparing a proposal for submission to Federal and State Government approval processes (STTDP website: “Proposed Maria Island Translocation Project,” 8 April 2008).” Then in December 2011, the media began carrying stories about an application by the Tasmanian government to the Commonwealth government to translocate 50 Devils to Maria Island (ABC, World Today, 12 December 2011; Neales, 2011; Paine, 2011). Why this new proposal was necessary is not clear. Perhaps, the Commonwealth’s original approval had a time limit that could not be met, or perhaps the Tasmanian government wanted to increase the numbers. In any event, it turns out that the proposal was “deemed not required” (Newsletter, September 2012; see also STDP, 2012a).

The project to move Devils to Maria Island involved several steps (STDP website, “Maria Island Translocation Project,” 9 June 2010). In June 2010, field workers were already carrying out baseline surveys of those resident native species, such as the ground-nesting seabirds, that might be affected by the Devils. These were completed in 2011 (Jones and McCallum, 2011a), although in an online document created on 2 December 2011 “baseline data … [was] currently being established for key species that may be preyed upon by devils on Maria Island (STDP, 2011b).”

The number of Devils to be released was originally given as 30 (above) but later this was increased to 40 (Duncan, 2010b) or 50 (STTDP website: “Proposed Maria Island Translocation Project,” 8 April 2008; ABC, World Today, 12 December 2011; Neales, 2011; Paine, 2011; ABC News, 23 March 2012; Hobart Mercury, 13 August 2012; STDP, 2012a). The latter figure seemed official (STDP, 2011b; Newsletter, September 2012; Tasmanian government media releases, 12 August and 14 November 2012; ABC Hobart, 14 November 2012; Martin, 2012b-c).

The release was to occur in stages. Originally, it was reported that ten Devils would be released initially with the remaining 40 released later (Hobart Mercury, 13 August 2012). Officially, it emerged that “around” ten Devils would be released in November and the remainder “over the next two years (Tasmanian government media release, 12 August 2012; Newsletter, September 2012).”

6/05/2016 8:41 AM 268

In fact, on 14 November 2012, 14 or 15 Devils, depending on reports, were released initially on Maria. The number 14 was reported in some media stories (Tasmanian Government media release, 14 November 2012; Darby, 2012b; Martin, 2012b-c), while the number 15 was reported in others (ABC News and ABC Hobart, 14 November 2012; ABC News, 16 November 2012; Denholm, 2012f; Hibbard et al., 2012; Salis and Graham, 2012; STDP website: The first translocation of devils on Maria Island, 23 November 2012; Tasmanian government media release, 15 November 2012; DIP Newsletter, December 2012 – where the release date was said to be on 15 November 2012; Newsletter, May 2013, April 2014). Even well after the release, the reported number of Devils released varied, some media stories reporting 14 (Claridge, 2013) and others 15 (ABC Online, 20 April 2013; STDP Annual Report, 2012/13; Beniuk, 2013; Blackwood, 2013; Devil Island, 2013a-f; McKay, 2013; Zimmer, 2013; S. Peck, in Bugard, 2014; STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; Huxtable et al., 2015). On the total number of citations, 15 would appear to be correct.

Originally, the animals in the foundation group were to be sourced from the wild (Cordingley, 2007a; Pincock, 2007), specifically the west coast (Jones and McCallum, 2007). Subsequently, it was said they would come from captivity (Duncan, 2010b). Later and officially, it was said the foundation group would consist of both wild and captive animals (STDP, 2011b, 2012; see also Belov and Hogg, 2012). In the event, however, the Devils actually released came from captive breeding populations. In some accounts these facilities were said to be both in Tasmania and on the mainland (Tasmanian government media release, 14 November 2012; Hibbard et al., 2012; Martin, 2012b-c) while in other accounts the facilities were said to be just in Tasmania (Salis and Graham, 2012; DIP Newsletter, December 2012). One of the latter accounts said seven Devils came from Taroona, five from the Freycinet FRE and three from the Bridport FRE (DIP Newsletter, December 2012; see also STDP Annual Report, 2012/13; Newsletter, April 2014).

The composition of the foundation group of Devils changed over time. In 2007, the proposal was to translocate only juveniles (Jones and McCallum, 2007). In early 2010, media reports said it would consist of sterilised older males (ABC News, 19 March 2010; Brand Tasmania Newsletter, April 2010; Duncan, 2010b; Ecos Magazine, Issue 154, 2010). These would be monitored for two years for signs of disease before intact Devils would be allowed to breed (Brand Tasmania Newsletter, April 2010; Duncan, 2012b). Subsequently, the formal proposal said, “Ideally, the translocated population will be composed of an age structure typical of free-living populations,” although individuals older than four years would be excluded on the grounds of diminished adaptability and a lower contribution to reproduction; the sex ratio would be approximately 1:1 (STDP, 2011b; see also STDP, 2012).

At the time of release, most of the accompanying media reports did not give the age, sex or reproduction status of the released Devils (Tasmanian government media release, 14 November 2012; ABC News and ABC Hobart, 14 November 2012; Darby, 2012b; Martin, 2012b). One Tasmanian government media story did say, however, “most were young, one to two year old devils of both sexes but there were several three to four year old males (Salis and Graham, 2012; see also DIP Newsletter, December 2012).” Several months after their release, it was revealed that at least eight of the animals were females. The composition was ultimately stated,

6/05/2016 8:41 AM 269

presumably authoritatively as it was from an STDP source, as seven males and eight females between one and three years old (Newsletter, May 2013; Annual Report, 2012/13; Devil Island, 2013e).

Devils being considered for release onto Maria Island were screened for wild-like behaviour (J. Clarke, on ABC AM, 20 October 2012; Devil Island, 2013a), age and genetic diversity (C. Hibbard, on ABC AM, 20 October 2012). There were four tests for wild-like behaviour, or what might be more realistically described as tourist- avoidance behaviour: whether a Devil will be put off by a strange object (a beach ball) lying over a bowl of food; whether a Devil will avoid taking food from a human; whether a Devil will run from a human, and how aggressive a Devil is to its own image in a mirror, a very high level of aggression being undesirable (Devil Island, 2013a).

The Devils bound for Maria were also wormed (STDP, 2011b), ensuring that any of their potentially occurring species-specific worm parasites, such as the equally endangered tapeworm species, Dasyurotaenia robusta, (see “Cestodes,” above), were excluded from the ark.

As foreshadowed, the released Devils are monitored for their impact on the island, especially its other species, and for the island’s impact on the Devils (Ecos Magazine, Issue 154, 2010). The plan involved trapping the animals at fortnightly and then monthly intervals (P. Bell, on ABC Hobart, 14 November 2012; see also STDP website: “Devils breeding on new island home,” 4 April 2013).

Depending on accounts, either five (Salis and Graham, 2012; Newsletter, May 2013; STDP Annual Report, 2012/13) or six (STDP Annual Report, 2012/13) of the initial 15 animals released were fitted with radio transmitter collars and followed between November and March (STDP Annual Report, 2012/13). All the animals were microchipped (ABC News, 16 November 2012).

Monitoring returned some interesting results. Five months after their release, at least five of the eight females were reported as breeding, with between 2 and 4 (mean = 3) pouch young (ABC online, 20 April 2013; Beniuk, 2013). Later, this was revised to six females (STDP website: “Devils breeding on new island home,” 22 April 2013; Newsletter, May 2013) and 16 young (mean = 2.7) (Newsletter, May 2013). In late November 2013, a media report said there were “20 babies” on the island (ABC, The World Today, 26 November 2013). And in January 2014, it was revealed that “all the females bred successfully” producing a total of 24 young (S. Peck, in Bugard, 2014; see also Devil Island, 2013f; see also Newsletter, April 2014), for a mean of 3.0 young per female. In November 2014, the STDP reported that “all eight females” had bred successfully producing “about … 20 young” (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014; see also STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; Huxtable et al., 2015).

Monitoring also showed that over a four month period, a two-year old male raised in a large FRE seemed unusually mobile, moving widely over the near coastal parts of the island and averaging 10.7 km per night with a maximum of 22.4 km per night (Newsletter, May 2013: map; see “Movements,” above).

6/05/2016 8:41 AM 270

It also appears that one two year old male, not the one mentioned in the preceding paragraph, sired all the young produced in the first breeding season after release (S. Peck, in Devil Island, 2013e).

Three of the females born to the first cohort of females released on Maria bred in the breeding season after their own birth (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014; see also STDP, 2014a).

Of the original cohort of 15 Devils, there was one fatality and one disappearance, presumed dead. About four months after the initial release (Devil Island, 2013d), that is, about mid-April, a collared male (Manny) was found dead in a collapsed wombat burrow (Claridge, 2013; STDP Annual Report, 2012/13; Devil Island, 2013d-e; Bugard, 2014; see also Huxtable et al., 2015). Then in November 2014, another male (Jimmy) had not been seen for about 10 months and was presumed dead (STDP, 2014a).

In November 2013, the Manager of the STDP said that initially, at least, the zoo- raised (“intensively managed”) animals lost weight at a “higher rate” than the FRE- raised animals and approached people to a greater extent, but these differences were not long lasting (ABC, The World Today, 26 November 2013). More specifically, apparently two weeks after the release, there had been a “general mean weight loss across the animals sampled,” but a “stabilisation of weight loss” after four weeks. The initial weight loss was attributed to the potential for more exercise in the now free-to- range animals as well as a possible increase in stress levels. One animal had lost so much weight it had to be captured and confined for a short period for supplementary feeding (STDP Annual Report, 2012/13).

A second release on Maria Island was mooted to occur in early 2013 (H. Williams, ABC News, 16 November 2012) or sometime after May 2013 (STDP Annual Report, 2012/13 – where October was mentioned; Newsletter, May 2013) and was to involve 15 animals (H. Williams, ABC News, 16 November 2012). In fact, the second release didn’t occur until 6 November 2013, although a much later report said the release had occurred in two stages, the first in later October and the second in early November (STDP Annual Report, 2013/14).

Again, the number of animals released varied according to reports. Some accounts said it involved 11 animals consisting of seven males and four females (ABC News and ABC AM, 7 November 2013; Darby, 2013; Marrow, 2013; see also Zoos Victoria website: Devils on run, 2 October 2013; ABC News, 27 September 2013 and Zoo Aquarium website: Tasmanian Devils released to the “wild,” accessed 7 June 2014). Other accounts said it involved 14 animals consisting of eight males and six females (Tasmanian Government media release, 6 November 2013; Bolger, 2013; B. Wightman on video accompanying Dawtrey, 2013; see also Law, 2014). Two late account even said 13 animals (Newsletter, April 2014; implied also in, STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; STDP Annual Report, 2013/14). Presumably, the count of 14 is correct as it is the number represented in most of the Tasmanian Government accounts.

6/05/2016 8:41 AM 271

The second-release animals were aged between one and three years (Tasmanian Government media release, 6 November 2013). All but one had been captive bred on the mainland, at Healesville (Zoos Victoria website: Devils on the run, 2 October 2013) and Monarto (ABC News, 27 September 2013; see also Tasmanian Government media release, 6 November 2013; Darby, 2013; Dawtrey, 2013) zoos, with the one exception coming from a zoo on Tasmania (Newsletter, April 2014). The STDP’s annual report for 2013/14, however, said Tasmanian-sourced animals (note plural) came from Trowunna Wildlife Park and from the program’s FREs (note plural).

There appears to have been five females in the second release (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014).

Subsequent to the second release, information on how the population is progressing has included all the animals. In this regard, there were indications in July 2014 that 11 of the 21 females on the island had pouch young, including the three females breeding in their first year after being born to females in the first introduction (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014).

In November 2014, the media began carrying the first reports of the effect of Devils on other species (Shine, 2014b; see also STDP Annual Report, 2013/14). Effects noted were the apparent lack of successful breeding of Cape Barren Geese due to Devils predating the eggs of this ground-nesting species (STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; STDP Annual Report, 2013/14; STDP, 2015a); apparent predation on Little Penguins to judge from feathers in the Devils’ scats (STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; see also Mounster, 2015b), and apparently fewer Native Hen chicks (Shine, 2014b). Most of these impacts were anticipated (Jones and McCallum, 2007; STDP, 2011b, 2012a). Of these species, only the Little Penguin is native to Maria, and in this regard, it is interesting that the STDP decided to put “protective nesting boxes” and/or “artificial nesting burrows” close to the water’s edge (Shine, 2014b; see also STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014; STDP Annual Report, 2013/14; Mounster, 2015b). This would be, if not the first instance, then at least one of the few, in which wildlife authorities have meddled with a native species in order to accommodate an exotic species.

The total number of Devils to be released on the island was originally reported in the media as between 80 and 100 (Ecos Magazine, Issue 154, 2010), which is close to the estimated carrying capacity of between 80 and 120 (PHVA, 2008; Newsletter, May 2013; STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014). It was subsequently suggested, however, that about 50 animals would be introduced and this founder population would be allowed to grow to the carrying capacity of the island. The Tasmanian government believes 120 Devils could be maintained on Maria (ABC online, 20 April 2013), which is close to the official understanding of the island’s carrying capacity. A media report gave 2016 as the year in which the population should attain carrying capacity, and authorities would have to start removing Devils from the island (Law, 2014).

6/05/2016 8:41 AM 272

In November 2014, media reports placed the total population on Maria at 70 (Day, 2014; Mather, 2014b) while a December 2014 STDP report said 70 to 90 (STDP website: “Maria Island Devil Translocation Project Update, December 2014,” 23 December 2014) and a STDP report published in July 2015 but for a period ending 31 December 2014 said a “likely” 80 to 90 (STDP Annual Report, 2013/14)..

Now that Devils have become established on Maria Island, they join Black Rats, Brown Bandicoots, Barred Bandicoots, Eastern Bettongs, Brush-tail Possums, Common Wombats, Bennett’s Wallabies, Forester Kangaroos and Emus, Tasmanian Native Hens and Cape Barren Geese in the island’s exotic menagerie (Rounsevell, 1989; N. Mooney, in Wood, 2003c; Newsletter, June 2010; Scoleri et al., 2015).

Proponents of the Maria Island introduction said it will help define the Devil’s “functioning” role in the wild (Newsletter: June 2010). This might be especially true with regard to the Devil’s ability to suppress local populations of exotic species, such as cats, which are present on Maria (there are no foxes or rabbits on Maria; see “Re- introduction to the mainland” below; C. Johnson, in Mounster, 2012b).

In addition to Maria, a number of other islands have been considered as possible translocation sites for Devils. These include Bruny (353 km2; M. Jones, to Wood, 2003c; PHVA, 2008; N. Mooney, on ABC Bush Telegraph, 22 May 2012), King (1094 km2; PHVA, 2008; Paine, 2011), Prime Seal (12.2 km2; Paine, 2011; N. Mooney, on ABC Bush Telegraph, 22 May 2012) and Three Hummock (70 km2; Loh, 2006). Of these, only King Island would be large enough to support a population large enough to overcome the effects of in-breeding depression (650 km2 supporting 500 Devils; STDP, 2011b).

As noted, the idea that Devils should be translocated to one or more off-shore islands as part of the Insurance Population was in the air as early as 2003 (Jones, 2003b; Martain, 2003; McManus, 2003; Wood, 2003c) and was discussed seriously as early as February 2005 (ABC News in Science, 7 February 2005; Newsletter, May 2010; McCallum, 2007; McGlashan et al., 2007). But in that the effort to remove the 120- 130 individuals comprising the introduced Devil population on Badger Island, off the west coast of Flinders Island in Bass Strait, may not have ended until sometime shortly before August 2007 (N. Mooney, in Hawkins et al., 2008), the authorities appear to have been talking about taking Devils off one island at the same time they were talking about introducing them onto other islands.

The attempt to remove Devils from Badger Island was apparently made in response to a demand by the local Aboriginal owners (Wade, 2005b; PHVA, 2008; N. Mooney, on ABC Bush Telegraph, 22 May 2012). As noted above (Distribution), most of the removed Devils were released at various mainland sites and a few were sent to captive facilities at Bicheno.

There was an initial concern that the Badger Island population might have been infected with DFTD. There were no old males, and one Devil had what appeared, on visual inspection, to be typical facial disease type tumours (DPIWE, 2005a-b; Wade, 2005b; Hawkins et al., 2006: appendix 1). But no “gross sign” of the disease was evident in the 120 + animals collected for removal (N. Mooney, in Loh, 2006), and

6/05/2016 8:41 AM 273

presumably no animals would have been sent to the main island from this population if facial tumour disease had been evident in it.

Interestingly, the authorities were not able to remove all Devils from Badger Island. It was thought that two animals might have eluded capture (N. Mooney, in 2010 document cited by Jones and McCallum, 2011a; N. Mooney, on ABC Bush Telegraph, 22 May 2012). This suspicion seems to have been confirmed by photographs of Devil tracks take in early 2010 (N. Mooney, pers. comm. in Jones and McCallum 2011).

Captive Breeding Strategy

For the moment, the breeding strategy of the Insurance Population program is to maintain most (90 or 95 percent depending on report) of the existing genetic diversity in Devils. It is not to make “improvements” in the stock, even in its resistance to facial tumour disease, because selection for single characteristics, such as tumour resistance, may have detrimental consequences for other parts of the genome (PHVA, 2008; Siddle et al., 2010a).

The Insurance Population effectively constitutes a metapopulation, that is, a series of separate populations loosely connected, in this case by human agency (CBSG/DPIPWE/ARAZPA, 2009). As the population grows, records will be needed of each Devil’s location, health, breeding history and genetics in order to maximise each animal’s genetic utility within the metapopulation. And as the captive population has the potential to grow rapidly, even with modest reproductive success, decisions will have to be made about which Devils will be allowed to breed and which will not. There is an existing studbook for Devils, maintained by the Healesville Wildlife Sanctuary (DPIWE, 2005b; PHVA 2008; Srb, 2012), and presumably it is current for all the animals in the Insurance Population (J. Hockley, ABC Rural, Bush Telegraph, 20 September 2011).

Extensive logistics will also be required, as animals will have to be shipped between facilities. And all this record keeping will have to be maintained indefinitely, perhaps for the rest of the species existence.

Breeding for Resistance

Surprisingly, one zoo in Tasmania has recently advertised itself as “… attempting to selectively breed devils with resistance to DFTD (Tyler et al., 2011).” Considering that the STDP has only recently acknowledged the existence of apparently resistant Devils in the wild, this project by a local zoo seems incongruous. Where did they get their breeding animals; how do they know the individuals are resistant, and how will they determine whether they have bred resistant offspring?

Gamete Banking

The harvesting and preservation of gametes is a common conservation measure for preserving the genetic diversity of species, and experiments have begun on both electrojaculation for sperm (Keeley et al., 2012a) and preservation techniques for both the sperm and eggs of Devils (Ogilvie, 2011).

6/05/2016 8:41 AM 274

With regard to sperm, researchers have tried with various electrojaculation techniques but with only middling success to date (Keeley et al., 2012a). They have also tested various solutions for storing sperm over the short term and have achieved viability ratings of between 61 and 77 precent after 12 h (Keeley et al., 2011, 2012c; Platt, 2011b).

With regard to eggs, other reseachers have experimented with the vitrification of the Devil’s granulosa-oocyte complex. Vitrification involves the diffusion of cryoprotectants into the cell, followed by rapid cooling which leads to a solid, glass- like structure that allows long-term cold storage. The experiment had a success rate, defined as viability post-thawing, of about 70 percent (Czarny and Rodger, 2010).

In 2011, a joint venture between the reproductive research facility of the commercially-based Busch Gardens in the US and the Taronga Conservation Society Australia set up the first gamete bank for Devils at the Taronga Western Plains Zoo in Dubbo, New South Wales (Platt, 2011b). Media reports suggest the bank will begin by storing sperm (Jeter, 2009; Kempton, 2011). Interestingly, the facility was first mooted in mid-July 2009 (Jeter, 2009), but it has never been mentioned on the STDP website. A critical question associated with any gamete bank is who “owns” the gametes and off-spring produced from them. The arrangement in the case of the Devil’s gamete bank is not in the public domain.

A disease suppression experiment on the Forestier Peninsula that ran for years (Disease Suppression) apparently included “gamete rescue from the culled animals (Keeley et al., 2011).” But the precise purpose of the rescue and the fate of the “rescued” gametes are unclear.

Disease Suppression

An early idea about how to deal with the disease in natural populations where movement into and out of the population could be controlled was to remove the affected animals already in the population and prevent affected animals from entering it. The idea was that this could lower the prevalence of the disease if not eliminate it.

The population on Forestier Peninsula provided such an opportunity. The peninsula, actually the first of two tandem peninsulas (Forestier and Tasman), was originally connected to the mainland by a very narrow spit. But in 1905 the isthmus was breached by the narrow but deep and fast-flowing Denison Canal, and reconnected by a vehicular bridge. Today, therefore, virtually all terrestrial species can only enter or leave the peninsula over the bridge, unless they are reasonably good swimmers, which Devils are.

In June or July 2004, depending on accounts (Newsletter, March 2006; Jones et al., 2007a; Lachish et al., 2010), researchers began a pilot project to cull all Devils judged, by visual inspection, to be infected. In 2004 and 2005, the project involved two ten day trips per year (Lachish et al., 2010; STDP website: “Wild management,” 4 August 2011; Beeton and McCallum, 2011a; Hamede et al., 2012a). In 2006, this became four trips per year (Jones et al., 2007a; Lachish et al., 2010; STDP website:

6/05/2016 8:41 AM 275

“Nipping DFTD in the bud,” 24 August 2010; Beeton and McCallum, 2011a). In 2007, there were five trips and in 2008, three trips (Lachish et al., 2010).

The cull included “all devils seen to be overtly infected or to have characteristic signs considered to precede DFTD (Jones et al., 2007; see also Lachish et al., 2010).” This apparently involved animals as young as 1.5 y of age (Keeley et al., 2011) and presumably females with pouch young, even those too young to be hand-reared (pers. obs.).

Between 2004 and 2008, 145 Devils were culled from the population, but 15 of these Devils (10.3 percent) proved to be unaffected by DFTD upon autopsy (data in Lachish et al., 2010).

Initially, the removal program seemed to be working as evidenced by a drop in the prevalence rate from 15 percent in June 2006 to 8 percent in June 2007 and holding steady to June 2008 (S. Smith, in Denholm, 2006b; Dennis, 2006; DPIW Annual Report 2006/07, 2007/08; DPIW, 2007a; Jones et al., 2007a: fig. 2; H. McCallum, on ABC, World Today, 5 March 2007; McCallum, 2008b; Trofimov, 2008; Lachish et al., 2010: fig. 2; Fig. 19).

Prevalence of DFTD on Forestier Peninsula 2006-2010

16.0 15.0 14.0 13.8 12.0 10.0 9.1 8.0 8.0 8.0 6.0

Prevalence (%) 4.0 2.0 0.0 2006 2007 2008 2009 2010 Year

Fig. 19. Prevalence of DFTD on Forestier Peninsula at mid-year in the period 2006- 2010. Data in annual reports of DPIW 2005/06, 2006/07, 2007/08, 2008/09 and DPIPWE 2009/2010.

By April 2009, however, the researchers concluded the cull wasn’t working after all (Hamede et al., 2009; Lachish et al., 2010; but see Woods and Sharman, 2010). Nevertheless, they decided to keep going, in the hope that more frequent trapping might work and on the possibility a pre-disease diagnostic test would be forthcoming (Lachish et al., 2010). In May 2010, the trapping times were increased from tri- monthly to monthly in order to catch animals before the lesions broke through the skin or mouth lining and became infectious (STDP website: “Nipping DFTD in the bud,” 24 August 2010; Newsletter, September 2010). But the diagnostic test never came (see “Test for Disease at a Pre-symptomatic Stage,” below).

6/05/2016 8:41 AM 276

In August 2010, a media story said the program now appeared to be working (Denholm, 2010b; see also Bender, 2010), apparently on the criterion that the population was larger than modelling predicted it would be in the absence of culling (DPIPWE, 2010d). In November 2010, however, the program was reviewed again, and in December 2010, it was announced it would be stopped due to lack of effectiveness. The population was still declining (STDP website: “Disease suppression winds down,” 22 December 2010; Newsletter, 2011). The estimated cost of this suppression program was “in excess” of $ 200 000 per year (Beeton and McCallum, 2011a; see also ABC, World Today, 5 March 2007; Waterhouse, 2007). The entire program, therefore, cost more than $1 million dollars.

A practical problem with the program was that a certain proportion of the Devil population, possibly as much as 20-25 percent, never entered the traps and hence constituted a persistent reservoir for the disease (Pincock, 2007; Newsletter, June 2009, December 2009; Lachish et al., 2010; STDP website: “Remote sensor cameras,” 22 February 2010, “Alternative trapping methods,” 23 February 2010; “Pavlov’s Devil,” 17 February 2011; M. Jones, in Beeton and McCallum, 2011).

It was realised as early as May 2007 that some Devils might be “trap shy” (Johnston, 2007), and it came to be recognised that a Devil’s “trappablity” does not change appreciably throughout life (Jones et al., 2012).

Once researchers realised that some devils avoid traps, they began working on other ways of catching animals reluctant to enter the standard trap (Newsletter, June 2009, December 2009). The results of this work are unknown.

There were two general concerns about the suppression program. First, it was removing individuals that might actually live long enough to breed at least once (McCallum and Jones, 2006; Lachish et al., 2007). Second, it was removing animals with any sign of tumours, indiscriminately including any animals that might have had a slight genetically based resistance to the disease (ABC PM, 9 June 2008 see also R. Hamede, in Johnston, 2007). A counter-consideration, however, was that culling of diseased, that is, manifestly genetically susceptible animals might increase the chances of matings between any genetically resistant devils (Beeton and McCallum, 2011).

Interestingly, modelling of the program undertaken prior to late 2008 concluded that for the program to have had any chance of being effective, culling would have had to be carried out almost continuously and at a logistically unrealistic intensity. This result was presented orally at a conference in December 2008 (Goldenfein, 2008, 2009; AFP, 23 May 2009) but not published until three months after the program was stopped (Beeton and McCallum, 2011a, but received by the journal in March 2011; see also Beeton, 2012). The modelling suggested “almost all of the infectious devils (around 96 %) would need to be removed on a regular basis to stop the disease (Betton and McCallum, 2011b).” The study commented that as a disease management strategy for wildlife, culling should only be attempted once appropriate models have shown that it is likely to be effective (Beeton and McCallum, 2011a). Considering the conclusiveness of this mid-experiment modelling, it is likely that any pre-experiment modelling of the projected program, based on even less knowledge, would have given an even more pessimistic (conservative) chance of its likely success.

6/05/2016 8:41 AM 277

A second modelling exercise looked at the costs of capturing and culling diseased Devils and found that “disease eradication is overtly costly because the marginal cost of trapping sick devils becomes large as densities of infected devils decline (Horan and Melstrom, 2011). This work was done independently and although published only in 2011, was submitted in December 2009. This relationship between cost and benefit with a diminishing “resource” is very basic concept but seems not to have been considered when the culling program was being planned.

It is also interesting to note that if the disease does spread in a frequency dependent manner, which has been thought to be the case since 2008 (Hamede et al., 2008, 2009; McCallum et al., 2009), then “culling to reduce host density is not a viable management strategy (Hamede et al., 2012a).” Nonetheless, the culling program continued for two years after frequency dependent transmission was accepted as the most likely mode of spread.

Remarkably, shortly after the authorities ended the culling program on Forestier Peninsula was brought to an end, they began a program to remove all the Devils on the peninsula and replace them with healthy ones (ABC News, 25 May 2012; Newsletter, September 2012), entails catching all the Devils in the 190 km2 area; killing the diseased animals and quarantining the rest for some unstated period, and then putting healthy Devils back into the area.. The effort will cost $ 200 000 and take up to 18 months. The program manager said they had already removed a third of the Devils estimated to be in the area, but it was unclear if this was based on the population size before or after the curtailed culling program.

This project has most of the problems of the suppression program and some additional ones as well. The old problems include the removal of diseased animals that might nonetheless have some genetic resistance to the disease; the rising cost and diminishing return on effort as the number of remaining animals declines; no sure way to tell that every last Devil has been caught, and no sure way to prevent wild Devils re-entering the area, either by swimming across Denison Canal, crossing the bridge over the canal or being maliciously introduced.

The program also has at least two new problems. First, the Devils caught at the beginning of the program will be held in captivity for a maximum of 18 months, the estimated length of the program. And all Devils will have to be held for at least nine, possibly even 12, months to give time for any latent infection to become obvious. This long period in captivity may lead to behavioural changes in the animals and their young, depending on the quarantine conditions. Second, this population will no longer be part of the natural wild population. Instead, it will become part of the captive mega-population. In other words, it becomes just another captive population, which Devils are moved in and out of in order to meet the genetic requirements of the captive population as a whole.

Release “Back” into the Wild

If Devils were ever released back into the wild, it would be desirable to know how they will fare after having spent time in captivity. To test this, some (all?) of the pouch young of females euthanased in the removal experiment on the Forestier

6/05/2016 8:41 AM 278

Peninsula were hand-reared until nearly a year old and then returned to the general area where their mothers were captured (DPIPWE, 2010d). The “orphans” were fitted with electronic collars and micro-chipped so their movements could be traced (ABC News, 6 December 2007, 3 January 2008).

The first release of animals under this program was in two groups of four, one in early December 2007 and the other over the following few weeks (ABC News, 6 December 2007, 3 January 2008; Tyler et al., 2011). Other reports, however, implied a total release of at least ten (ABC News, 13 March 2008) or stated a total of 12 (ABC News, 3 January 2012). There was second release in May 2008 (STDP website: “Sobering roadkill trials,” 5 October 2010), but the number of animals released was not disclosed. There was a third release sometime in 2009, but again the number of animals involved was not disclosed. It is unclear if there have been any subsequent releases. It is also unclear just how many animals in total have been released, although as of June 2010 it was said to be “more than 20 (Newsletter, June 2010).”

Results from the orphan release program were not submitted for publication until November 2013 (Sinn et al., 2014). In the interval, however, media reports and STDP documents provided snippets of insight. Within four months of their release, three of the Devils in the 2007 release had been killed on roads; two had disappeared and were presumed dead, and three were still alive (STDP website: “Sobering roadkill trials,” 5 October 2010). Another media report, however, said three had succumbed to DFTD and seven had been road-killed (ABC News, 13 March 2008). A scientist involved in the release called the hand-reared orphans “naïve” and called on assistance of the public to put some fright into them (C. Pukk, on ABC News, 3 January 2008).

In June 2008, two males from an unspecified release had “done extremely well in terms of survival and weight growth and they've done much better than some of the animals that …were released (M. Jones on ABC Radio, 2 June 2008).”

In the first half of 2010, at least two females from an unspecified release were carrying pouch young (Newsletter, 2010; see also STDP, 2011b). Despite these mostly “good news” stories, however, to judge the effectiveness of the orphan release program, it would be necessary to know the fate of similar cohorts of animals left in the wild in the same area over the same period of time.

Official results of the orphan release program, or at least the first two years of it, were published in December 2013 online (Sinn et al., 2014). According to this report, eight approximately 20 month-old Devils were released between 29 November 2007 and 10 December 2008, and eleven 10-14 month old Devils were released between 11 January and 19 May 2008. As of July 2008, five of the eight (63 percent) 2006 cohort and six of the eleven (55 percent) 2007 cohort were either dead or presumed dead. By way of comparison, one of four wild Devils (25 percent) born in 2006 died. These mortality rates are not exceptional (see “Mortality and Survivorship,” above).

The two surviving females in the 2006 cohort reproduced, with four young each, in the breeding season following their release. None of surviving three females from the 2007 cohort bred, due most likely to their young age. By way of comparison, the two surviving wild-born 2006 females also produced four young each in the same season. Four is the maximum number of young a female can support.

6/05/2016 8:41 AM 279

Apparently, not all orphaned Devils are released back into the wild. Although it is the STTDP’s general policy to return orphaned animals to the wild (Bird, 2012a), the decision to release is made on a “case by case” basis (H. Williams, in Bird, 2012b). For example, five of the orphans in the 2006 cohort mentioned above, all of which were all hand-raised, were not released into the wild because they had developed a stereotypical pacing behaviour.

Release “back into the wild” in any area that had had the disease would have to be predicated on the virtually certainty that all the diseased Devils in the area had been removed, either by disease or human agency. If even one diseased Devil was over- looked, the release would be compromised and the time and money expended on it wasted.

Translocation

From time to time, the suggestion is made to translocate Devils from one part of their current range to another part. Translocations have been justified on two grounds. First, it was thought that if some populations were to prove resistant to the disease, some individuals from those populations or their captive-bred progeny could be transferred to populations susceptible to the disease in an effort to “strengthen” them (Lunney et al., 2008; McCallum, 2008b-c; A. Sharman and M. Jones, on ABC World Today, 1 January 2009; M. Jones, on ABC, Stateline Tasmania, 13 November, 2009, and ABC News, 14 November 2009; K. Belov, on ABC Science Online, 10 March 2010; Jones and McCallum, 2011b; Jones et al., 2014). Similarly, diseased Devils could be removed from an area and replaced with disease-free Devils (STDP Annual Report 2010/11).

When it became evident that wild populations were no longer declining and captive breeding populations had reached capacity, a second reason for translocating animals became to “augment” or “supplement” population numbers and genetic diversity in the wild. This was a key proposal in the so-called Wild Devil Recovery Project, announced on 23 November 2014 by the Tasmanian Minister for Environment, Parks and Heritage (Groom, 2014; see also Fox et al., 2015). The STDP explained that “the outcome … [of the Wild Devil Recovery Project] would be the return of the devil to the wild – sooner than expected (STDP Annual Report 2013/14),” overlooking apparently that the Devil was already persisting everywhere in the wild.

The official descriptions of the translocation aspect of the Wild Devil Recovery Project bear quoting. The Minister himself said it would, “… test ways of augmenting the wild population (Groom, 2014). The STDP’s 2013/14 Annual Report, said, “Releasing devils that are well represented genetically within the Insurance Population will enable the Program to supplement existing wild populations and bolster them both numerically and genetically.” In July 2015, an abstract, authored in part by two STDP officers, wrote: “The STDP is now in a position to consider releasing some …captive animals into the wild to supplement dwindling wild populations and boost genetic diversity levels” (Fox et al., 2015; see also ABC News Online, 23 September 2015). And in August 2015, a captive breeding associate of the STDP said releases into existing populations would “… provide new diversity to the

6/05/2016 8:41 AM 280 animals that are already existing at the release site (C. Hogg, on ABC, The World Today, 27 August 2015).”

These views are curious. First, as to numbers, if a population is persisting but for some strange reason not growing toward its previous numbers, what reason is there to believe that adding more Devils to the population would be effective? If some factor were suppressing the existing population, in all likelihood, that same factor would continue to affect the supplemented population. On the other hand, if the population was increasing naturally, why gainsay this natural process?

Second, as to bolstering genetic diversity, as there is no evidence there has been any significant loss of specific genes or overall genetic diversity at any diseased site (see “Population genetics”) why introduce novel genes and alleles, especially without knowledge of what those genes do? Again, if existing populations are increasing, or even just persisting, why not let natural selection shape the genetic diversity suitable for a particular site?

In any event, the Wild Devil Recovery Project’s first move to translocate Devils was made public first in May 2015, when the STDP awarded a grant for a pilot project to “re-wild” Devils in northeast Tasmania (STDP website: “Grants & Scholarships,” as updated on 18 May 2015) where there was an existing resident population. In an abstract of an oral presentation on 27 July 2015 the release was foreshadowed for September 2015. In September 2015, attendant with the actual release, the chosen site was revealed to be Narawntapu National Park in central-north Tasmania. Twenty captive-bred Devils were released on 25 September 2015 (ABC 7:30, 23 October 2015; Advocate, 26 September 2015; Atkin, 2015; Discovery News online, 28 September 2015; Gibson, 2015; Groom, 2015; San Diego Zoo media release, 28 September 2015; Shannon and Lehman, 2015). As discussed below (A Vaccine against DFTD?), 18 of the Devils had been immunised with an experimental vaccine that had elicited an immune response. Thus, the Devils were part of two experiments, one to supplement an existing population and the other to test the vaccine under wild conditions.

All the released Devils were microchipped (ABC 7:30, 23 October 2015; Atkin, 2015) and “had their bottoms bleached” to make them standout in sentinel camera images (Atkin, 2015).

The immediate follow-up of the release was not promising. Within seven and a half days, four of the 20 Devils (20 percent) had been killed on roads (ABC News online, 3 October 2015; Carter, 2015c; Darby, 2015b; Gramenz, 2015; Lohberger, 2015a-b; The Advocate, 30 September 2015 and 3 October 2015; The Australian, 3 October 2015; Wisbey, 2015). This is an attrition rate several orders of magnitude higher than what would be expected in a natural occurring population.

After the deaths became public, STDP spokespersons indicated a loss of 40 to 60 percent of released animals would be acceptable (D. Pemberton, in Bolger, 2015a; S. Fox, in Atkin, 2015; see also S. Fox in ABC News online, “Captive-bred Tasmanian devil program loses quarter of released animals as roadkill,” 15 December 2015 in regard to a similar release on the Forestier Peninsula).

6/05/2016 8:41 AM 281

The actual release site in the park was never stated, but it was probably just east of the park headquarters and south of the west end of Spinglawn Lagoon (inference based on pictures in de Salis, 2015 and Google Maps). The three accident sites (two Devils were apparently killed at one site) were identified in only general terms so only a rough estimate can be made of how far the Devils had moved from the release site.

Of the first two Devils reported killed, one was killed sometime between late Friday afternoon and the following Monday on Greens Beach Road near Beaconsfield (Gramenz, 2015; Lohberger, 2015a-b; The Advocate, 30 September 2015). This no less than 13 km from the release site, assuming, conservatively, “near Beaconsfield” is the closest point to the park on the southern half of the road and is in a different catchment.

The second Devil was found dead “on Bakers Beach Road” south of the park (Gramenz, 2015; Lohberger, 2015a-b; The Advocate, 30 September 2015), a location too imprecise to analyse.

The last two Devils reported killed were apparently found at one site, the “Frankford Hwy near the turn-off to the national park” and found on (The Advocate, 3 October 2015; Wisbey, 2015; STDP website: “Signs to watch to watchout for wildlife,” 14 October 2015). This is about 11 km from the release site. These Devils could have followed Bakers Beach Road south out of the park.

The released Devils probably all had names (see Atkin, 2015), but the names of the deceased Devils were not mentioned in any of the dispatches.

Well prior to the release, the STDP suggested that “tools to ‘dampen’ the dispersal of the released devils” would be used in the release (STDP Annual Report, 2013/14). Later, it was revealed that one of these “tools” (or should that be “stools”?), was Devil scats. Scats of the to-be-released Devils were scattered in the release area to familiarise the resident Devils with their odours, so when the new Devils were released they would not come as so much of a surprise. Reciprocally, the scats of the resident Devils would be introduced into the holding enclosures of the to-be-released Devils (STDP website: “Narawntapu National Park Tasmanian Devil Translocation September 2015,” 25 September 2015; de Salis, 2015). In the event, the initiative seems to have failed in the case of at least the four road-killed Devils.

Parenthetically, it would be interesting to know what the STDP had learned about the immediate of post-release movements of the Devils released on Maria Island in 2012 and 2013 (see “Off-shore islands”) and whether that knowledge might have been informative about the post-release movements of the Devils in the Narawntapu National Park release.

A few days later developments got even “curiouser.” The STDP’s Program Manager revealed that in the light of the road-kill loss of the four Devils, he had asked geneticist to look for any genes that might make Devils “want to explore,” adding “we’ll try anything (Bolger, 2015a).” Aside from the further source of artificial selection this idea involved, it also reveals that the STDP is willing to change the Devil’s genetics in order to solve what is essentially a management problem arising out of its own interventionist approach and has nothing to do with the long-term well

6/05/2016 8:41 AM 282 being of the resident Devil population. Indeed, the Program Manager’s assertion that “we’ll try anything,” in principle opens the way to genetically engineering a Devil that glows in the dark.

The STDP explained the release into Narawntapu as “…an exciting and important next step in re-establishing wild populations in the Tasmanian landscape (STDP website: “Narawntapu National Park Tasmanian Devil Translocation September 2015,” 25 September 2015). This seems to overlook (discount?) the fact that wild populations are still present everywhere in Tasmania they have ever been since European settlement.

In any event, more captive-bred Devils were due for release into Narawntapu National Park in November 2015 (ABC 7:30, 23 October 2015; Atkin, 2015).

A release of Insurance Population Devils into a second natural population was foreshadowed for a locality in “NE Tasmania (STDP, 2015a).” This site may have been Stony Head on the central north coast, subsequently mentioned to have Devils released into it sometime in 2016 (STDP website, “Forestier Peninsula Release - Update 29 February 2016,” 29 February 2016).

This is very likely to be the Mt William area, where the disease was first recorded. If this happens, three natural sites – the Forestier/Tasman Peninsula (see “De-populate, Re-populate” above), Narawntapu and Mt William – will all be irrevocably “contaminated” with exotic gene lines.

It difficult not to think the Wild Devil Recovery Project’s initiative to “augment” natural populations is a handy solution to the immediate problem of a burgeoning Insurance Population and a long-term solution to the ultimate redundancy of that population as the wild populations stabilise and presumably grow. Indeed, it is even possible to think that the STDP, faced with the recovery of the Devil entirely on its own, indeed in spite of the STDP, want to be able to claim some small part in the Devil’s final successful recovery in order to justify its interventionist approach and expenditure of $ 50 + million.

Returning to the problem of translocations in general, two of its main potential effects are the introduction of novel genes into locally adapted populations and the perturbation of patterns of geographic genetic variation. Novel genes may be detrimental in and of themselves as well as disruptive of pre-existing gene complexes. And novel genes, as well as non-local frequencies of existing genes efface patters of geographic variation, which is, after all, part of natural genetic diversity. There is nothing in the public domain to suggest the STDP has the least concern about these possible effects.

Some translocation of Devils has already occurred. In an earlier day, some “nuisance” Devils were no doubt trapped and removed far enough so they wouldn’t return. More recently, the authorities themselves translocated Devils from the introduced population on Badger Island back to various sites on the mainland, specifically Bridport, Epping Forest and Georgetown (STDP, 2011b), without knowing the founding individuals’ exact original provenance (see “Distribution”).

6/05/2016 8:41 AM 283

The STTDP also authorised the release of an orphan Devil from West Pencil Pine onto a Cradle Mountain property some 20 km to the east (Anthony, 2011).

Captive breeding in the Insurance Population, including those on islands such as Maria, also effaces locally adapted genomes, and most, if not all captive-bred Devils, when released into the wild, will have genomes with no specific relevance to local conditions. For example, it was originally envisaged that the Maria Island introduction would include Devils from both western and eastern parts of Tasmania (STDP, 2011b), and although when the two sequential introductions were finally made there was no mention of the geographic provenance of the animals, the circumstances of the introduction make suggest the new population will be genetically “blended.” Similarly, the 19 or 20 immunised captive-bred Devils released into Narawntapu National Park on 25 September 2015 (ABC News, 22 July 2015; Carter, 2015a; Darby, 2015a; Mounster, 2015a; ABC, The World Today, 27 August 2015; Gibson, 2015; Shannon and Lehman, 2015; Advocate, 26 September 2015), almost certainly involved blended animals.

The proposed programmatic translocation of Devils raises a question common to all translocation programs. Does the imagined gain of the proposed translocation outweigh the irrevocable disruption of the natural pattern of genetic variation (Lachish et al., 2011a), which is, after all, as much a part of what the Devil is as the geographic variation in its colour pattern and body weight? If we value Devils, at least in part, for the natural continuity with location-specific natural conditions, then translocation is a destructive process. But if we are simply interested in saving a continuous line of Devils regardless of their place of origin and continuous occupation, then translocation poses no problem. But then, the same effect could be achieved in a zoo. The STDP seems to have adopted translocation as a management tool and in doing so has revealed part of its values.

Re-introduction to the Mainland

From time to time, there are suggestions to “re-introduce” Devils mainland Australia, the forested and cleared agricultural regions in the southeastern part of the country (Hunter et al., 2015; Huxtable et al., 2015), Wilson’s Promontory and the Otways in Victoria and far southwest Western Australia having been suggested so far. Proponents give three reasons for such a re-introduction.

The first reason is that Devils might help suppress the many exotic species that wreak havoc on the remnant native fauna and flora. The main target species would be predators such as cats and foxes and herbivores such as rabbits (M. Jones, in Wood, 2003c; Johnson and Wroe, 2003; Wroe and Johnson, 2003; Johnson, 2006; Lunney et al., 2008; M. Jones, in Maloney, 2011 and on ABC World Today, 6 July 2011; Mercer, 2012; Ritchie et al., 2012; E. Ritchie, in Squires, 2012; Ritchie et al., 2012; Science Network Western Australia, 7 November 2012; C. Johnson, in Elder, 2013; Ritchie, 2013; Jones et al., 2014; Louys et al., 2014; Hunter et al., 2015).

There are initial indications of the resources needed to begin introducing Devils to the mainland. A feasibility study for the introduction has been costed at $3.4 million over five years (Milman, 2015a). And a permanent fenced enclosure on the mainland would require “at least 100 km2 (Jones et al., 2014).”

6/05/2016 8:41 AM 284

The second reason is a more general version of the first: re-introduction would “rewild” or “return areas to pre-settlement ecosystems (Elder, 2013).”

The third reason for re-introduction to the mainland is to help save the Devils themselves by giving them a refuge not only free of facial tumour disease but all the artificial aspects of even the largest free-range enclosures, to say nothing of zoos (Johnson, 2007; M. Jones, on ABC News, 6 July 2011; Science Network Western Australia, 7 November 2012; Jones, 2012; Louys et al., 2014).

There are two issues with the idea of re-introduction to the mainland. First, would Devils “take” if released on the mainland? The answer to this is related to what caused Devils go extinct on the mainland in the first place, because, if the causative factors were still operative, it would be pointless to try and “put Devils back.” Some have suggested releasing them in places with no dingoes and ongoing fox control (M. Jones, on ABC News, 6 July 2011), showing a belief that dingoes may have been at least partly responsible for the disappearance of Devils from the mainland in the first place. But the fact is, there is simply no clear understanding why Devils disappeared from the mainland (above).

The second issue is just how effective Devils would be in suppressing exotic species. The answer to this question lies in the effectiveness of Devils in limiting the abundance of exotic species in Tasmania.

Devils would almost certainly eat young rabbits in their nest as well as any adults they might catch. Indeed, at some localities, rabbits seem to be the major food item (Guiler, 1964). There are no studies addressing the issue of the Devils impact on rabbit numbers in Tasmania, and there is apparently no evidence, even anecdotal, of rabbit numbers increasing with the recent decline of Devils.

As to cats, which occur throughout Tasmania, it is perhaps relevant that in all the work on the diet of wild Devils, cats appear on only one list and for only one locality, Cape Portland in northeast Tasmania (Guiler, 1970a). Further, cats were apparently unrecorded in the diet at Granville in northwestern Tasmania, where Devils were “frequently caught (Guiler, 1970a).” It is assumed Devils would eat cats, especially kittens left in a den, but it might be useful to test the palatability of cats (roadkills or euthanased for other reasons) on captive Devils just to verify this assumption.

Devils would probably interfere with cats at large carcasses (STDP website: “Do devils suppress cats?,” as updated on 5 October 2010), but as cats do not appear to be particularly dependent on large carrion, this may not be an important interaction. Some spotlight survey evidence suggests that the number of cat observations have increased as Devil numbers have decreased, at least in some areas, but it is unclear if this is due to a decrease in the cat population or an increased furtiveness in their behaviour (See “Consequences of the Devil’s Extinction”).

There is also apparently image evidence of cats co-habitating peacefully with Devils (Mooney, 2015).

6/05/2016 8:41 AM 285

As part of its advance planning for the release of Devils on Maria Island (See “offshore islands” above), the STDP began a survey of the island’s cat population (STDP website: “Maria Island Translocation Project,” published 9 June 2010, revised 6 December 2010; Devil Island, 2013d-e). In a few years, therefore, the STDP should be able to provide some actual data on the question of the Devil’s impact on this particular exotic species.

How Devils might go with foxes is problematic, as foxes have a yet to build to large numbers in Tasmanian. Some observers believe the absence/low density of foxes, despite apparent repeated introductions since European settlement, is due to the presence of a robust Devil population. It seems likely Devils would eat fox cubs left in a den and compete strongly with adult foxes for dens and carrion. The assertion Devils kill adult foxes (Beeton, undated) is unsupported. Perhaps one of the large free-range enclosures planned for Devils, especially the one on the mainland, could be used to test some of the assumptions about Devil/fox interactions.

In terms of total effect, therefore, Devils might contribute in a small way to the suppression of at least these three common exotic species on the mainland, but whether it would be worth the effort is problematic.

The re-introduction of the Devil to the mainland for its own sake, even if it were to “take,” again raises questions about the point of conservation. If it is to maintain continuous breeding populations of a species in its natural habitat, then re- introduction to an area where it went extinct naturally doesn’t meet the criterion. But if it is simply to maintain continuous breeding populations anywhere in the world under any conditions, then it would be better than, say, a zoo.

In November 2015, it suddenly became known that there had been a proposal to trial a release of a group of same sex Devils in the Barrington Tops area of New South Wales. The Tasmanian Environment Minister apparently rejected the proposal on the reported grounds that “… it would detract from current initiatives (Noonan, 2015).” Of course, it would not have escaped the minister’s attention that if the Devil were ever to become established on the mainland, it would also detract from one of Tasmania’s unique tourist offerings.

Where Are All the Bodies?

The fact that the diseased Devils in the Forestier Peninsula removal experiment are euthanased (J. Hamilton, in McGlashan et al., 2006; Woods et al., 2007) raises the issue of what happens not only to these diseased animals but also to all the other dead Devils, both healthy and diseased, that pass through the hands of researchers, wildlife authorities and zookeepers.

There is little information on this issue in the public domain. At least some of the diseased Devils culled from Forestier Peninsula were buried (Trofimov, 2008), and at least one experimental animal, Cedric, was also buried (ABC News, 24 September 2010).

6/05/2016 8:41 AM 286

Some body parts (Keeley et al., 2012a-c) and tissue samples (Woods et al., 2007; Ujvari et al., 2012, 2013) are taken (below), but it is not clear what happens to the bodies from which the parts and samples came.

If the authorities truly believe that the Devil is headed for extinction, then it would seem prudent, if not mandatory, to provide for the culled and euthanased Devils to go into either existing museums or a special-built facility.

Surprisingly little is know about the ontogenetic, sexual and geographic variation in Devils, and healthy dead Devils, such as those from roadkills, might provide our last chance to assess this variation and to assess many other aspects of Devils’ biology. There is even more to be learned about the diet of Devils and their inferred methods of finding food, from the analysis of the stomach contents of healthy Devils. Diseased dead Devils would provide insight into the systemic effects of disease. Specimens that die in zoos can provide insights into any physical effects that may be caused by life in captivity.

Lineages kept in captivity for generations may show a reduction in brain size, which on the face of it might suggest an impaired ability to survive in the wild. Laudably, there is now a project in place to measure the effect, if any, of captivity on brain size in Devils and the word has gone out for the heads of any Devils that may die in captivity (Hanson, 2012; Science Alert, 14 February 2012; Research Matters, Victoria University, 27 February 2012; Hibbard et al., 2013; Victoria University News, 10 January 2013; Smiley, 2013).

How many Devils should be preserved? The answer must surely be: about the same number of Tasmanian Tigers we wish had been preserved when they were still common. This would surely be on the order of several hundred.

In this regard, a 1962 open letter from the Wild Life Preservation Society of Australia to the Tasmania government’s Animals and Birds Protection board regarding the bodies of Devils the Board had permitted to be killed is as germane today as it was then. The Society asked, among other things, “In the view of the scientific value of destroyed Tasmanian Devils, has the Board made provision to see that such specimens are not wasted? Is there a legal obligation on the part of permit holders to return destroyed animals to the Board? … (The [Wild Life Preservation Society of Australia] is vitally concerned with this question in view of the thoughtless wastage of destroyed thylacines in the past, and the consequent rarity of both skeletal and anatomical material, which, for the most part is housed in foreign institutions. It is strongly urged that such a situation should not be permitted to occur in respect of Sarcophilus) (p. 31, Australian Wild Life, June 1962).”

There is also a problem of what to do with Devils in the Insurance Population when they become too old to breed, a problem that will increase as the number of captive Devils increases. Initially, two facilities, Kyabram, a zoo in Victoria (ABC News, 18 January 2010), and Wild Life Sydney Zoo in New South Wales (Bigpond News, 21 December 2012), took some aged animals. More recently, there are plans to expand some of the participating zoos’ facilities to house such Devils (STDP website: “Funding boost for the survival of the Tasmanian Devil,” 24 April 2013). And in 2014, a 600+ ha enclosure was being built just north of Hobart for “… older members

6/05/2016 8:41 AM 287

of the Tasmanian Devil insurance population and those that are no longer required for breeding purposes” (STDP website: “The Appeal is thrilled to be ‘Taking Care of the Elders,” 12 February 2014; also Abey, 2014).

Another approach to the problem of too many animals in the Insurance Population is a contraceptive, and to this end in 2013, the STDP awarded a grant for “Developing a method of contraception for overrepresented animals in the Tasmanian devil insurance population” (STDP website: “Grants and Scholarships,” as modified on 12 June 2014). Any curtailment of breeding, however, means selecting some animals for reproduction and rejecting others, yet another form of artificial selection (see “Re- introduction to the mainland”).

Parenthetically, it is worth mentioning that providing aged care facilities for reproductively senescent Devils is one consequence of the STDP’s early decision to gain public support for the program by harnessing the public’s propensity to anthropomorphise about animals.

Another less publicised plan of older Devils is to use them “… for short term research projects whenever possible (Wells, 2010).”

Tissue Sample Collections

A variety of tissues has been collected from healthy Devils since 1986 and from diseased Devils since at least 1999. These collections, which seem to have been given the collective name of Tasmanian Devil Bio Archive, are held in at least two Tasmanian institutions, the Menzies Research Institute, University of Tasmania, Hobart and the Mount Pleasant Laboratories, DPIPWE, Kings Meadow (Launceston) (Tovar, 2012; STDP Annual Report 2013/2014).

As of the end of 2014, these collections consisted of the following: tissue, sera, blood – 15,500 samples; tumour cell culture – 2052 samples; blood and fibroblast culture – 2052 samples; ear biopsy and tumour (both primary and secondary) samples – 15 000, and a Bacterial Artificial Chromosome (BAC) library – the entire genome of two devils (STDP Annual Report, 2013/14). Unfortunately, the DNA in some ear biopsies may have degraded and fragmented due to deficient collection and storage protocols (Wright et al., 2015).

There is a curator for at least some, if not all, these collections, although where the curator is domiciled is unstated. Applications to access the collections are overseen by the Molecular Research Advisory Panel (MRAP). According to the STDP, “the principal criterion used by MRAP to assess applications is the project’s potential to contribute to devil recovery (STDP Annual Report, 2013/14).” In the case were the materials are not unduly limited, this seems unnecessarily restrictive.

The database for the collections of these samples, that is, tissue type, location, date of collection, sex, age, health status and the like, is not openly accessible, nor is the process for accessing this basic, seemingly non-proprietary, information.

Test for DFTD Infection at a Pre-symptomatic Stage

6/05/2016 8:41 AM 288

A diagnostic test for Devil facial tumour disease before it becomes obvious by visual inspection would be, of course, a great help in combating the disease. Such a test has been a goal of researchers since the beginning of the fight against the disease. But despite much promise and hype, a pre-clinical test remains elusive.

In July 2007, researchers at the University of Tasmanian were awarded $ 85 000 to search for blood serum protein markers that could be used for pre-symptomatic diagnosis (UTas media release, 23 May 2007; ABC News, 24 May 2007). The payoff seemed to come on 2 April 2009 when the University of Tasmania announced that researchers in the university’s Australian Centre for Research on Separation Science within the School of Chemistry had developed a blood test for the disease. The test was said to be “… fast, taking about three to four hours to produce a result,” “non- invasive,” requiring only “…one drop of blood from an ear prick,” and not requiring “a PhD to use it (UTas media release, 2 April 2009).” However, an additional “six months” would be required to “validate” the research (Newsletter, June 2009; see also Scott, 2009; ABC 7:30 Report, 21 April 2009). Not withstanding this delay, one of the principal researchers described the work as a “definite breakthrough (Tedmanson, 2009).” In mid-2009, Save the Tasmanian Devil Program manager wondered whether the test was “… indicative of just DFTD, or is it indicative of some other process going on with the devils (ABC Online. Stateline Tasmania, 1 May 2009)?” At the end of 2009, the test, said to be akin to a PSA test for prostate cancer, was reported as “still undergoing scientific validation” (Denholm, 2010a). In August 2010 it was still undergoing “validation” but with the hope of delivery “by the end of the year (Denholm, 2010e).” In November 2010, a STDP document listed the “lack of a DFTD pre-clinical diagnostic test” as a major challenge (STDP, 2010d). In 2012, the STDP reported, somewhat ambiguously, at least with regard to the university project, that “collaborative partnerships with external researchers are working towards the goal of developing such a test (STDP Annual Report, 2011/12, note plural).” As of the end of 2014, investigations into a pre-symptomatic diagnostic test were continuing (STDP Annual Report, 2013/14).

In October 2012, a graduate thesis, done in the same lab as the work described above, published a means of identifying DFTD in at least some samples for a period of up to at least six months before the source animals showed clinical signs of the disease, presumably small pink nodules. The work involved the mass spectrometry of metabolites in the blood, followed by a principal component analysis of the data from infected Devils and healthy Devils. Using the technique on blood samples from mark and recapture surveys, the best result was the correct identification of 95 percent of the pre-disease animals in a sample consisting of females from different localities. The result for males was less successful, although the percentage of correct identifications was not given. The researcher thought the less successful result for males might “… have been caused by the bias in the sampling locations of the male DFTD devils (Gathercole, 2012).” Apparently, the technique still requires “validation” before it can be of practical use (Gathercole, 2012).

In 2013, the STDP awarded a grant for the “Development of a test to diagnose DFTD before the appearance of tumours (STDP website: “Grants and Scholarships. Grant recipients updated to June 2013,”`as modified, 12 June 2014). No further details were available.

6/05/2016 8:41 AM 289

In March 2014, the STDP website carried a story of research into the possibility of using a serum protein, the tyrosine kinase RrbB3. The level of this enzyme is elevated with some cancers, and hence once baseline levels are established, elevated levels might be indicative of DFTD (STDP website: “Characterising DFTD - Research progressing well,” 20 March 2014).

In October 2014, a grant proposal foreshadowed the possibility of identifying DNA from the Devil’s facial tumour in the blood (Murchison and Rosenfeld, 2014).

In May 2015, the STDP announced a $ 33 000 grant for the “metabolomics for [the Tasmanian Devil/DFTD biomarker discovery (STDP website: “Grants & Scholarships,” as updated on 18 May 2015).” So six years after the first announcement that a blood-bourn, pre-clinical test for DFTD had been discovered, the research apparently continues.

One other attempt at a pre-clinical test involved the possibility that the tumour might affect the structure or chemical constituents of the Devil’s hair as is apparently the case with the hair of human breast cancer patients (Church et al., 2007; CSIRO media release, 4 June 2007; Radio Australia. Innovations, 13 August 2007; Newsletter, August 2007). The only result that came of this work, however, was a detailed description of the ultrastructure and chemical composition of Devil hair, with no word about any diagnostic potential (Church, 2009).

No matter when a pre-clinical test becomes available, the very early stage of the latency period is likely to remain undetectable. This is because it there will always be a lag between the time the cancerous cell becomes embedded in a healthy Devil and the time when the proliferating cancer cells release compounds in detectable amounts. At best, therefore, when such a test becomes available, it will reduce the quarantine period before an at-risk Devil can be considered disease-free (Gathercole, 2012).

Other Conservation Threats

With Devil numbers declining steadily from facial tumour disease, every surviving Devil becomes increasingly important to the future of the species. Two current concerns with other potential sources of mortality are baiting programs to kill wild canids, especially foxes, and road traffic.

Baiting for foxes

Currently, there is a broad scale program to bait introduced foxes. This involves inserting 3 mg of the poison 1080 into pieces of processed or kangaroo meat and spreading them across the landscape at a density of about 200 m apart. Devils, being scavengers, will of course find and eat these baits. Luckily, however, Devils are much less susceptible to 1080 than are foxes. The LD 50 for 1080 for Devils is 4.2 mg/kg whereas for foxes it is 0.10 mg/kg (data from RSPCA, cited in DPIPWE 2010; see also Ross, 2008). A Devil, therefore, needs to eat ten times the amount needed to kill a fox (Duncan, 2010a). The dosage and bait size are adjusted so that a fox will be killed by finding and eating only one bait, but a Devil would have to find and eat at least 10 baits every two or three days, the time it takes for the poison to be largely metabolised. 1080 baiting is unlikely to contribute significantly to Devil deaths

6/05/2016 8:41 AM 290

Roadkill

There are two concerns about the impact of roads on Devils. The first is the establishment of new roads into currently disease-free areas, such as the Tarkine. Devils move along roads, no doubt for the same reasons humans do – easier access, and hence a new road into a disease-free area is likely to increase the rate at which the disease moves into an area (Hawkins et al., 2005, as cited in Obendorf and McGlashan, 2008).

The second issue with roads is that when existing dirt roads are up-graded, that is, sealed, widened and straightened, vehicular speeds increase, leading to an increase in the number of Devils being killed. The authorities believe this source of Devil mortality can be mitigated, at least in some areas, by keeping the sealed part of the road narrow, laying a pale coloured seal (making Devils more conspicuous) and erecting signage (ABC News, 16 April 2011).

Habitat destruction

Although Devils arguably cope well, and may even benefit, from some forms of habitat destruction such as partial clearing for pastoralism and forestry, they remain susceptible to the destruction of their dens. Clearing may enhance opportunities for scavenging dead stock and wildlife, but indiscriminate clearing is also likely to lead to a net loss of dens, unless there are provisions to preserved denning habitat.

Mining is one form of habitat destruction that has no redeeming features for Devils. It destroys the Devils and their habitat in the immediate mining area and increases other risks such as to roadkill in the surrounding area. Furthermore, in disease free areas, such as the Tarkine, it enhances the spread of the disease via the newly created roads.

Interestingly, the Commonwealth government’s response to this problem, as shown by conditions on two open-cut iron ore mining approvals in the Tarkine is to require payment of what is effectively blood money both for the direct destruction of Devil habitat in the immediate area of the mining lease and for each individual mine-related death in the surrounding area.

For Shree Minerals’ 150 ha mine with a ten year life this involves an up-front payment of $ 380 000 (Macintosh, 2012; Commonwealth Government Media Release, 31 July 2013; Hobart Mercury, 31 July 2013; McCallum, 2013; Middleton, 2013; The Australian, 31 July 2013). For Venture Minerals’ 119 ha mine with a two year life the payment is $144 000 (Arup, 2013; The Australian, 5 August 2013). These amounts are to be paid to the Save the Tasmanian Devil Appeal (Macintosh, 2012; McCallum, 2013), from where they will find their way into the Save the Tasmanian Devil Program.

Developments in a relatively small area have a way of cascading as the first development, argued on the basis of economic necessity, sets a precedent for another and another until the number of developments degrade an area sufficiently for the argument to shift to here being little reason to preserve what remains. The process is exacerbated if the early developments are individually assessed as unlikely to “have a

6/05/2016 8:41 AM 291

very serious impact on devils” (N. Mooney, in Bird, 2013; see also ABC News, 30 July 2013). The Tasmanian Government is considering ten other mining proposals for the Tarkine, and presumably it and the Commonwealth governments will approve them all with most of the same conditions.

The payment conditions put the Save the Tasmanian Devil Program in an interesting ethical situation. As a joint Commonwealth and Tasmanian government initiative, the STDP can probably not openly criticise either the Tasmanian or the Commonwealth’s decisions about the mining approvals and their conditions. There is, however, a question about whether it could have quietly declined to be the beneficiary of these funds if it had serious concerns about these particular mines’ impact on the Devil or the precedent for future mines. Without explanation, it will remain unclear whether the STDP thought the decisions were truly benign as far as the Devil’s future was concerned or whether it was guided by political or financial expediency.

Parenthetically, it would also be interesting to know how the Commonwealth came up with the figure of $ 48 000 as the value of a Devil. Putting a dollar value on anything without a market is fraught, but the precision of the figure suggests the Commonwealth may have struck on a method that has escaped others. Having arrived at this figure for an accidental death, why could it not now be applied to the wilful killing of any Devil?

Animal Welfare Insights

Although observers occasionally express a belief about the intensity of a diseased wild Devil’s suffering, no one has described the basis for their belief. Human observers certainly find the Devil’s facial lesions horrific to look at, and it is easy to imagine the kind of physical pain they might cause in a human. But if any species raises an issue of whether all animals experience physical pain and suffering in the same way as humans they themselves would, it is the Devil. For not only is severe wounding in intraspecific interactions part of its normal life, Devils can behave in a seemingly quite normal manner, despite horrific wounds. It is possible, therefore, that Devils have a very high threshold for physical pain, which may serve them well in dealing with the disease.

Devils may also have a strong recuperative capacity. One keeper has described their healing powers as “remarkable (Kelly, 1993).”

The Save the Tasmanian Devil Program has a document called “Euthanasia Policy” (STDP Annual Report 2012/13), but it is not in the public domain. In 2008, it was apparently government policy to immediately remove and euthanase Devils with DFTD captured in the wild (Jones et al., 2012). This implies all infected Devils were killed. This policy seems to have been relaxed subsequently.

A 2010 DPIPWE document contains the following advice to members of the public who may trap Devils inadvertently. “Should a trapped Tasmanian devil with very severe advanced-stage DFTD be suffering in the extreme, it is permissible to euthanase that animal. Generally, only devils with very large tumours that restrict feeding may require euthanasia…. Before euthanasing a female devil, every effort must be made to check for the presence and growth-stage of any pouch-young or for

6/05/2016 8:41 AM 292

lactating teats. If a female devil has large lactating teats (young are likely denned), or fully-furred pouch young, euthanasia must not be carried out as the young may not yet be weaned (DPIPWE, 2010c, emphasis in original).” Additionally, a veterinarian associated with the program noted that “the current position is to euthanase infected males, but females must be treated on their merits (Wells, 2010; see also Darby, 2003b).”

In the period February 2004 to June 2011, it appears as if “moribund” animals were euthanased (Pearse et al., 2012).

In 2012, the then STDP manager wrote that animals removed in the de-populate/re- populate program (see: De-populate/re-populate) would be assessed and “those animals that have advanced symptoms of DFTD and the associated welfare issues will be humanely euthanized in line with current policy (Sharman, 2012a).”

It would be sensible to not euthanase on welfare grounds alone if an individual may be able to contribute to the next generation, say a male with small, non-infectious, tumours just before the mating season or a female with advanced tumours but with a chance of getting off her current litter of young. This seems to be the practical policy, at least with female Devils (Bevilacqua, 2004a; Kreiss et al., 2009a; Denholm, 2010b; Wells, 2010; STDP website: “6233 2006,” 31 May 2011).

Consequences of the Devil’s Extinction

There has been a great deal of speculation as to what would happen if the Devil were to go extinct in the wild. In essence, this is a question concerning the functional role of Devils in the ecosystems it inhabits.

Increase in carrion?

Perhaps the most noticeable immediate effect of the Devil’s serious decline and ultimate extinction would be the longer time the bodies of dead animals, both native and exotic, would remain in the landscape. Instead of being cleaned up over a few nights, they might linger for as long as they do on the mainland, because less efficient scavengers were to increase in numbers due to the additional resources. As early as October 2003, there were anecdotal accounts of increased numbers of carcasses where Devil numbers had declined (Young, 2003; Hawkins et al., 2006). And it is now estimated that each day, between 70 and 100 tonnes of carcasses a night that Devils would have otherwise eaten are added to the Tasmanian landscape (Mooney, 2005; N. Mooney, in Brennan, 2007; Newsletter, May 2007; DPIW, 2007a; STDP website: “Nick Mooney’s Devil tales – it’s ‘all creatures great and small’ gone wild,” 25 February 2011).

In terms of the impact on living species, some pressure might come off those species whose young the Devil is known or likely to prey on, including indigenous species such as the two quolls and introduced species such as cats, foxes and rabbits (Mooney, 1992; Hawkins et al., 2006; McCallum and Jones, 2006).

Some pressure might also come off the Devil’s competitors, such the indigenous forest ravens and the two quolls, as well as the introduced competitors, cats and foxes

6/05/2016 8:41 AM 293

(Hawkins et al., 2006; Mooney, 2012; Lazenby and Dickman, 2013; N. Mooney, in Circular Head Chronicle, 4 October 2013).

Increase in cats?

The first concern that the number of cats was increasing as Devils declined was apparently expressed at a closed workshop meeting on 7 October 2003 (Young, 2003). A similar concern was expressed in 2005 (D. Obendorf, as reported in Hobart Mercury, 29 March 2005).

In 2007, the concern increased (Mooney, 2007; N. Mooney, in Brennan, 2007 and Pincock, 2007) and the first hard data was published. This was in the form of a graph of annual spotlighting survey results taken over much of Tasmania in the period 1995- 2005 (Jones et al., 2007a: fig 3). The authors noted that “monitoring evidence suggests that an increase in cats has begun (Jones et al., 2007a). Statistical analysis of the original data, however, shows that the correlation, although negative, was not significant (rs = -0.55, P = 0.08; pers. obs.).

In a document created in mid-2013, the Tasmanian Government published graphs of the spotlighting surveys for a number of species, including Devils and cats, for the years 1985-2012, inclusive (DPIPWE, 2013, p. 10) and in mid-2015, it published graphs for the years 1985-2014, inclusive (DPIPWE, 2015: fig. 4). There was no analysis of the relationship between the number of Devils and cats observed in either report.

Out of personal interest, I obtained the original data did an initial Spearman rank correlation analysis of the Devil and cat data for the years 1985-2014, inclusive (data courtesy of G. Hocking, DPIPWE, for the 132 transects surveyed consistently over the years and the data used by DPIPWE) and using zero, one-, two- and three-year time lags for cats, found no significant relationship for any of the four cases.

A detailed regional analysis of the spotlighting surveys with regard to the relationship between Devils and feral cats was published online in September 2013 and in hard copy in early 2014. This analysis covered the survey period 1985-2008 and showed that in the northeast part of the state, where the disease has been present the longest, the number of cat spottings rose post-DFTD as the number of Devil spottings declined (Hollings et al., 2014: figs 2 and 4). In the central-east part of the state, however, there was no obvious relationship between the number of Devil and cat observations.

This analysis is important, but it is worth bearing in mind that the number of cat observations was trending upward in the northeast area before DFTD arrived and the interpretation of an affect on cats rests on an apparent upturn in this trend post-DFTD (Hollings et al., 2014: figs 2 and 4) in what is a rather variable data set. The authors wrote that there was “… a strong negative association of feral cats with devils …. (Hollings et al., 2014),” but the statistical test and value for the strength of this relationship is not evident to me.

Out of personal interest, I estimated the spotlighting values for both Devils and cats from the paper’s figures (Hollings et al., 2014: figs. 2 and 4) for the entire data set for the northeast, that is, the years 1985-2008, which includes both pre- and post-DFT

6/05/2016 8:41 AM 294

periods. A Spearman rank correlation analysis showed that the strongest association was with data employing a one year lag for cats (rs = -0.46; P < 0.03, N = 23). Note: the original data used in the article were not available, as they were subject to a project-specific intellectual property agreement with the Tasmanian Government (T. Hollings, pers. comm., email of 6 January 2015). Re-analysis with the actual data would be interesting.

A plot of these data with a fitted polynomial trendline (Fig. 20) allows visualisation of the data and suggests two further points of interest. When the number of Devil observations is in the 0.40 – 0.80 range the number of cat observations fluctuates around 0.05. It is only when the number of Devil observations falls below about 0.40 that the number of cat observations increases perhaps to fluctuate around 0.15, a threefold increase. If Devils are having an effect on the number or furtiveness of cats, it may be that the effect reaches saturation at about half the highest density of Devils. Parenthetically, this may be an example of the non-linearity of the interaction between a predator and its “prey” (Ritchie et al., 2012).

Number of Devil and Cat Observations in Northeast Tasmania in the Years 1985-2008, with a One Year Lag for Cats

0.30 R2 = 0.39, P <0.01 0.25

0.20

0.15

0.10 (One Year Lag) (One Year 0.05 Number of Observations Cat Number 0.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Number of Devil Observations

Fig. 20. Plot of the number of Devil and cat observations along road transects in northeast Tasmania in the years 1985 – 2008, with a one-year lag for cats. Data from Hollings et al., 2014: figs. 2, 4. See cited work for how original observation values were obtained.

A Spearman rank correlation analysis for the central-east part of Tasmania, where the disease arrived somewhat later returned an rs = 0.01. P = 0.97, that is, no significant correlation. The authors of the original paper suggest the “lack of clear effects [in this area] may be due to a lagged response to devil decline (Hollings et al., 2014).”

Parenthetically, careful readers may wonder why there were no data in the above spotlighting study for the years 2009-2011 even though the results were only first mentioned in an oral abstract in November 2011 (Hollings et al. 2011a; see also Hollings et al., 2011b) and then not submitted for publication until May 2012. The answer apparently is that “there was a change in methodology in 2009 (T. Hollings pers. comm. in email of 6 January 2015).” In reply to a query to DPIPWE about a

6/05/2016 8:41 AM 295

possible methodological change in the spotlighting survey in 2009, Greg Hocking replied that it involved the way the spot-lighting on the survey vehicle was attached (now largely abandoned), and he considered it to be a minor change (email, 23 March 2015).

Considering how often the Hollings et al., 2014 paper is cited in support of the introduction of Devils to the mainland it would seem incumbent to revisit the original analysis with all the data, that is, 1985-2014, before this time-consuming, expensive and problematic initiative gains traction.

There was also apparently another analysis of the spotlighting data that was completed no later than May 2009 (T. Hollings, pers. comm. in email of 6 December 2015). One of the principles of this analysis said, "Our measurements through spotlighting show cats going up in the perfect relationship with devils going down" (N. Mooney on ABC News, 25 May 2009). It is not clear whether this analysis included all or only some of the spotlighting data, but putting aside a “perfect relationship” as hyperbole, the result sounds extremely compelling. Unfortunately the analysis seems never to have been published.

A camera survey conducted at four sites over several weeks in 2009-2011 in southeastern Tasmania showed that the number of recordings of cats were significantly less frequent at localities where Devils were recorded than where Devils were unrecorded (Lazenby and Dickman, 2013).

The caveat in all the preceding work was that the apparent reciprocal relationship between the number of observations of Devils and cats may be due to a change in cat behaviour instead of cat numbers (STDP website: “Do devils suppress cats?,”as updated on 5 October 2010; Mooney, 2012; Hollings et al., 2014; N. Mooney, in Mounster, 2015c). Cats may avoid areas where they detect Devil presence, or they may just become more generally furtive. In this regard, it is interesting that when Devils were removed from Badger Island over a short period of time, cat tracks quickly became more common along the beaches (Mooney, 2012). It is also worth remembering that at an early study site in northwest Tasmania that had a robust population of Devils “feral cats [were] frequently caught (Guiler, 1970a, emphasis added).”

Subsequent work on the reciprocal relationship between the number of Devil and cat observations suggested it may be due, indeed, to an effect on cat behaviour instead of cat numbers. An unpublished 2012 PhD thesis apparently found “feral cat population trends did not appear to be negatively affected by devils in Southern Tasmania (Fancourt et al., 2015, in reference to Lazenby, 2012).”

An unpublished 2012 honours thesis apparently “found a strong positive association between cat occupancy and devil abundance in DFTD-free areas supporting high devil abundance in NW Tasmania (Fancourt et al., 2015, in reference to Saunders, 2012).

In a 2013 advertisement of a PhD research opportunity, there was a claim that, “in the Midlands Biodiversity Hotspot in Tasmania, vertebrate biodiversity is threatened by agricultural intensification, continued ecological simplification and disruption,

6/05/2016 8:41 AM 296

including an increase in feral cats following the catastrophic decline of the native top predator, the Tasmanian devil (Jones, 2013, emphasis added). The evidence for this claim, however, has never been published.

An unpublished 2014 PhD thesis apparently found, “no numerical or behavioural relationship between devils and cats across NE, NW and southern Tasmania (Fancourt et al., 2015, in reference to Troy, 2014).”

The published work citing the unpublished work described above examined the relationship between the number of Devils and cats recorded by sentinel camera at 12 widespread sites in a three-week period between July and November 2012. It found no relationship between the estimated number of Devils and cats at the camera sites (Fancourt, 2015a; Fancourt et al., 2015). It did find, however, “… evidence of temporal separation of cat and devil activity, with reduced separation and increasing nocturnal activity observed in areas where devils had declined the longest (Fancourt, 2015a; Fancourt et al., 2015).” [Parenthetically, the first result was based on an analysis that included one locality said to be “devil-free,” South Bruny Island – but see “Distribution” above. Re-doing the analysis excluding this locality still returns a non-significant result: rs = 0.28. P = 0.41; Fig. 21]

500 R2 = 0.21, Rs = 0.28, P = 0.41 450 400 350 300 250 200 150 100

Cat EstimatedCat Abundance 50 0 0 20 40 60 80 100 120 140 160 180 Devil Estimated Abundance

Fig. 21. The abundance of cats and Devils at 11 camera sites (based on Fancourt et al., 2015, supplementary data, but exclusive of South Bruny Island).

A mid-2015 abstract reported that since the introduction of Devils onto Maria Island, there had “…been significant changes in the total predator assemblage with devils having significant impacts on feral cat … activity …. (Scoleri et al., 2015).

Unfortunately, there are very few direct observations of interactions between Devils and cats, and those are equivocal. As previously mentioned, there is a video of a large cat moving off a carcass at the approach of a Devil (STDP website: “Do devils suppress cats?,”as updated on 5 October 2010), suggesting that cats may not care to contest a common food source with a Devil. On the other hand, another “video shows an adult domestic cat repeatedly entering a devil den under a house containing 6 month old devils, well within the prey range of cats, but not preying on those young,”

6/05/2016 8:41 AM 297

and “other photos show a very large feral cat freely mixing with multiple adult devils in a free-range enclosure while feeding, grooming and sleeping … (Moody, 2015).”

In summary, the hard evidence suggests that Devils may have no affect on actual cat numbers but may cause them to change their activity times. Therefore, assertions that Devils do or could suppress cat numbers (Boness, 2012; Ritchie, 2013; S. Legge, on ABC Melbourne, 13 October 2013; M. Appleby, on SBS News, 9 April 2015; M. Letnic, in University of New South Wales media release, 11 August 2015) seem premature at best.

Cats are not only predators of native species, they are also the definitive host of the parasite causing toxoplasmosis. Toxoplasmosis can cause disease in native mammals, but its impact on populations is unstudied. There is preliminary evidence that in areas where the number of Devils has decreased the number of cats has increased and the incidence of toxoplasmosis in both native herbivores and carnivores has also increased (Hollings et al., 2011a-b; Hollings et al., 2013).

Increase in foxes?

There have been sporadic reports of foxes in Tasmania for many years (I saw one in the central highlands in 1975), but there is no solid evidence they ever became established. Their failure to establish may be due partly to predation and competition by Devils (Hawkins et al., 2006). A rumoured deliberate introduction of foxes in 1999 caused many to think foxes may have become established and were in the early stages of population growth and spread (Saunders et al., 2006; Mooney, 2012; Sarre et al., 2012). A robust review of the evidence for this introduction and reputed establishment and spread, however, was highly critical and casts doubts on whether the status of foxes in Tasmania has changed at all (Gonçalves et al., 2014; Marks et al., 2014a-b).

Interestingly, it is thought that just as smaller predators such as cat, foxes and quolls might increase in numbers when relieved of the pressure from Devils, so too possibly did the Devil increase in numbers when Thylacines declined to extinction in the early part of the last century (N. Mooney, in Bevilacqua, 2003a; Mooney, 2012).

Native species

Some researchers believe that if the Devil goes extinct, other native species will do so as well due to increased predation from increasing numbers of cats and foxes (Jones et al., 2007a). The STTDP’s website asserts that the fox could endanger seven of its possibly 70 prey species in the Tasmanian landscape (Newsletter, May 2007; see also N. Mooney, in Brennan, 2007). And the STTDP’s Senior Scientist at the time suggested that four other, unnamed, species would follow the Devil into extinction (H. McCallum, on Radio National, Innovation, 18 May 2009). The native species named specifically as being under threat include the Tasmanian Native Hen, Eastern Quoll, Tasmanian Bettong, Eastern Barred Bandicoot and Tasmanian (or Red-bellied) Pademelon (Jones et al., 2007a; Johnson, 2007; M. Jones, in Gallasch, 2011; The Mercury, 27 May 2011; M. Jones, ABC Science Show, 3 September 2011).

Commercial consequences

6/05/2016 8:41 AM 298

From the human perspective, however, the extinction of the Devil would mean that Tasmania would lose its most widely recognised living animal. If tourists viewed the efforts to “save the Devil” as inadequate, the state’s “green” image might be tarnished further, adding to the calumny of the loss of the Thylacine in the last century (Hobart Mercury editorial, 2 September 2003).

Spread to Other Species?

Some researchers are concerned that the disease could spread to other species (A.-M. Pearse, on ABC News, 25 July 2008 and in Newsletter, September 2008; K. Belov, in Rex, 2008; PHVA, 2008; Belov, 2011; E. Murchison, on BBC News, 27 June 2011 and in Giles, 2012; Belov, 2012), the most likely candidates being, presumably, the Devil’s closest relatives in Tasmanian, the two species of quoll. This is only likely to happen if an infected Devil were to bite a quoll; the quoll survived; the tumour cells avoided detection by the quoll’s immune response, and the cells found a congenial location in an organ that would facilitate transmission from one quoll to another, such as in and around the mouth or perhaps more likely in this less socially “bitey” species, the reproductive organs. This seems like a big ask.

Relevant to the question of the possible spread of the disease to other species is the demonstration that live facial tumour cells do not “take” in immuno-competent mice in the same period (30 weeks) in which they developed into large tumours in immuno- compromised mice (Kreiss et al., 2012). Presumably, the normal mice mounted an effective immune response.

Lastly, it may come to a relief to some that DFTD is apparently not transmissible to humans. Apparently, there have been some needle stick injuries to people taking tissue samples from DFTD-affected Devils, but with no transmission of the disease (YouTube: “Devil Facial Tumour Disease – Smarter Every Day 140,” although audio is indistinct).

An Immune Response?

The best possible outcome in the fight against Devil Facial Tumour Disease, of course, would be to discover that some individuals or populations in the wild were resistant to the disease. Indeed, in September 2008, the STTDP’s Senior Scientist wrote, “It is clear that if some animals are resistant to the disease, this will totally change the overall management strategy for the disease and the prognosis for the species as a whole (STTDP, 2008c; see also A. Kreiss, on ABC World Today, 1 January 2009; H. McCallum, in Smith, 2010).”

Interestingly, it appears as if the STTDP had an inkling that at least some Devils might be resistant to the disease as early as late 2006 or early 2007 and that some Devils might even be able to generate antibodies to the disease as early as late 2006 but no later than early 2007 (see “Resistance in Wild Populations?,” below).

Beginning in late 2006 or early 2007, therefore, researchers would have had a clear incentive, if one had been needed, to begin working on the immunity of Devils to facial tumour disease. This work developed along three lines: efforts to stimulate an immune response; the genetic basis of the apparent different abilities of Devils to

6/05/2016 8:41 AM 299

mount an immune response, and field surveys to identify resistant individuals and populations.

Interestingly, the early observation that seven percent of 91 tumours examined had been infiltrated by lymphocytes (Loh et al., 2006a; not 32 percent as reported in Murchison, 2009 or 8 out 18 as stated in Morris and Belov, 2013 and attributed to Loh et al., 2006b where there is no mention of lymphocyte infiltration) failed to alert anyone as to the possibility that Devils might vary in their natural alibility to mount an immune response to the disease. Indeed, subsequent papers would say there was no lymphocyte infiltration of the tumours, citing the original paper as the basis for that statement (Kreiss et al., 2008; Brown et al., 2012). Seven percent may not seem like a large number, but in terms of the potential for natural selection, it is relevant.

This early but largely ignored early observation of lymphocyte infiltration of some tumours was corroborated in two 2014 studies using histochemical techniques. In one study, researchers found that in “rare cases,” stained biopsies showed that lymphocytes, either T cells of natural killer cells, were found within the connective tissues around tumours and seemed to have induced the up-regulation of MHC molecules in nearby tumour cells (Siddle and Kaufman, 2012; Siddle et al., 2014). Ten biopsies were examined and “some” showed signs of this infiltration. Although described as “rare,” if “some” is taken to mean at least two, then at least ten percent of the biopsies would have showed evidence of such infiltration.

In the second study, histochemical analysis of ten tumours found evidence for dendritic cells in all ten tumours and T cells in eight tumours, although not all examined subsets of the dendritic and T cells were found in all tumours. No B cells, however, were found in any tumour (Howson et al., 2014). Again, these results could be interpreted as evidence for an immune response, albeit a variable and currently ineffective one. Even such partial responses, however, could provide variation on which natural selection could work.

Finally, some researchers are fond of saying, in effect, that “No devil has ever recovered from the disease” (Woods, 2012b; see also Darcé, 2010; Morris and Belov, 2013). This may be true, but it misses the point. What is of interest is not so much those individuals that will succumb, but those populations that may show resistance. But to judge from its publicity, the STDP seems more concerned about applying its interventionist programs than detecting natural resistance.

Cedric and Clinky … and Christine

In early 2007, immunologists began the first experiment to test the immune response of healthy Devils by inoculating two individuals with tumour cells. For months, this highly publicised experiment would bolster the public’s hopes about the Devil’s long- term survival. But what began in hope, ended in tears.

The experiment involved two Devils named Cedric and Clinky (most Devils that come into contact with humans for any length of time get names; see Darby, 2011; Lane et al., 2012; Brown, 2013). They were half-brothers born in captivity. They had the same mother but different fathers (Rex, 2008). Their mother was from the

6/05/2016 8:41 AM 300

northwest part of the state (Woolnorth; Miller et al., 2011; Cheng et al., 2012a and subsequently, Kreiss, 2009).

Cedric’s father was also from the northwest. Originally, his provenance was given as Arthur River Region (University of Tasmania media release, 1 September 2010; Miller et al., 2011; Cheng et al., 2012a). Subsequently, however, it was revealed his father was one of two captive Devils from Arthur River (Cedric is TD 145 in Kreiss, 2009) in northwest Tasmania.

The provenance of Clinky’s father was originally reported as having “eastern” genes (Obendorf and McGlashan, 2008), suggesting he was from the east coast. Subsequently, however, it was revealed he was in his mother’s pouch when she was taken from the wild (Clinky is TD 146 in Kreiss, 2009, but only made publicly available online in March/April 2015), so his father must have been also from the northwest.

When the experiments to be described began, Cedric was two years old and Clinky was three (Kreiss, 2009). Although captive born, if their births followed the natural cycle, they were born in about March 2005 and March 2004, respectively.

It should be mentioned that an assertion that Cedric “came from the Cradle Mountain area” is apparently incorrect (Anthony, 2011). The further assertion that Cedric had the AC5 chromosome (see “Karyotype,” above) (Anthony, 2011) has never been corroborated.

Starting on 13 March and on three further occasions up to and including 14 May 2007, the two Devils were injected in the cheek with 1 x 108 freeze/thawed (3x) and then irradiated (1x) Devil facial tumour cells from Strain 2 (lines ½ Pea and 2112) in the “commercially available veterinary adjuvant” Montanide (Woods, 2010; see also ABC Online, 20 February 2007; ABC News, 24 March 2007; Woods, 2008; ABC World Today, 1 January 2009; Pearse et al., 2012).

Cedric developed “high titres of tumour-specific antibody” “within a couple of weeks,” presumably after the first injection, but Clinky did not (Obendorf and McGlashan, 2008; G. Woods, on ABC World Today, 1 January 2009; also Woods, 2010). This result was first announced in late November 2007 (UTas media release, 29 November 2007; Darby, 2007b). The researchers were reported as noting the number of cells injected was much higher than what would be delivered in one Devil biting another (Naidoo, 2008).

Just how the live DFTD cells for the first four injections were made non-viable was conflicting. In the original thesis, the protocol was as described above (Kreiss, 2009) and therefore assumed to be correct. On 24 May 2010 in a public lecture, however, one of the researchers described the protocol as consisting only of irradiation (Woods, 2010). Then when the work was finally published, the protocol was described as just freezing/thawing (Kreiss et al., 2015: table 2).

On 24 December 2007, researchers injected 2.5 x 104 live cells from Strain 2 (line 2112) into two sites on the head in both Cedric and Clinky (Woods, 2010; see also ABC AM, 31 March 2008; University of Tasmania media release, 2 June 2008;

6/05/2016 8:41 AM 301

Marks, 2008; Martain, 2008; Obendorf and McGlashan, 2008; Woods, 2008; ABC World Today, 1 January 2009; Kreiss et al., 2015; table 1). One of the researchers described the quantity of injected cells as “very few” (A. Kreiss, on ABC AM, 31 March 2008).

In late March 2008, two media reports indicated that Cedric and Clinky were still alive and showing no signs of a tumour (ABC AM, 31 March 2008; ABC News, 31 March 2008; Denholm, 2008a, published on 31 March 2008), and one (ABC News) reported that researchers had not detected an immune response (antibodies) in Clinky.

It is interesting that Clinky was reported to the public as being well in the last week of March 2008, because it appears from original sources that researchers had discovered two tumours no later than mid-March 2008. Clinky’s fatal challenge with live cells was on 24 December 2007, and he was discovered to have two tumours 84 days (Kreiss et al., 2015 as TD 1) or 12 weeks later (Woods, 2008; Obendorf and McGlashan, 2008; Kreiss, 2009 as TD 146). The period 24 December 2007 to 16 March 2008 is 12 weeks.

In early June 2008, a few days before the Tasmanian state budget was to be handed down, a University of Tasmania media release announced that Cedric had “showed an immune response” and remained tumour free whereas Clinky had not developed an immune response and had developed a tumour. A researcher close to the work said, “We are 90 to 100 percent certain that Cedric is resistant to the disease …” (University of Tasmania media release, 2 June 2008).

Clinky was treated with two different chemotherapies, although the timeline for these therapies was never revealed (ABC News, 2 June 2008; Killick, 2008; Radio Australia, Innovations, 30 June 2008; A. Kreiss, on ABC AM, 17 October 2008). The larger of his two tumours was surgically removed around mid-July 2008 (18 weeks after the tumours were discovered), but the smaller was too inaccessible for surgical removal (Kreiss, 2009). Sometime around early August 2008, he was euthanased (A. Kreiss, on ABC AM, 17 October 2008; Foley, 2009; Kreiss, 2009).

On 3 July 2008, researchers injected Cedric in the cheek with 2.5 x 104 live cells from Strain 3 (lines 2533 and 5464) (Woods, 2010; also Darby, 2008; Save the Tasmanian Devil Newsletter for June 2009; Foley, 2009; G. Woods, on Catalyst Online, 18 June 2009; Pearse et al., 2012), in order to test how long Cedric’s “immunity would last” according to one account (ABC, PM, 1 September 2010) or whether “his immunity would be for all strains” according to another (ABC AM, 21 April 2009).

This was a potentially inconclusive protocol, because if Cedric did develop a tumour how could researchers tell if it was due to something different in the second part of the experiment other than the different strain type? For example, Cedric received a fifth booster two and a half months before injection with live cells of Strain 2 but went seven months with no booster before injection with live cells of Strain 3 (data in Woods, 2010).

In any event, on 3 December 2008 researchers found that Cedric had developed two tumours, each about the size of a five cent piece or a pea, depending on reports, “at the site of his last injection” (Woods, 2010; see also Bevilacqua, 2008; UTas media

6/05/2016 8:41 AM 302

releases, 17 December 2008 and 26 February 2009; ABC Radio National, 17 December 2008; Obendorf and McGlashan, 2008; Foley, 2009). The tumours were surgically removed on 15 December 2008 (UTas media release, 17 December 2008). Subsequent analysis showed the tumours were from Strain 3 (UTas media release, 26 February 2009; Kreiss, 2009; Kreiss et al., 2015).

Parenthetically, this experiment provided additional insight to the latency period of DFTD infection. Cedric’s latency period was about 22 weeks (early July to early December) whereas Clinky’s was about 15 and a half weeks (late December to early April). Cedric’s longer period, however, could have been due to still-circulating antibodies due to earlier inoculations, whereas Clinky had no circulating antibodies.

On 24 May 2010, one of Cedric’s principle researchers said in a public lecture that they were “… still using him quite regularly (Woods, 2010).” What this further use amounted to became apparent when a 2013 PhD thesis went on line (Brown, 2013; accessed mid-June 2013) and when a paper was published in early (Kreiss et al., 2015) where Cedric is identified as CD 14 and TD 2, respectively.

Cedric’s last experiment was straightforward. He and another Devil were injected in the right shoulder over three successive months (early April, May and June 2010, as inferred from Kreiss et al., 2015: table 1) with sonically killed DFTD cells of Strain 3 along with an immune-stimulating adjuvant supplemented with CpG oligonucleotides. This was done to see if the combination would stimulate an immune reaction. The treatments produced both cytotoxicity and antibody responses (Brown, 2013; Kreiss et al., 2015). Parenthetically, a protocol tabular entry shows Cedric being injected first in about the first week of September 2009 (Kreiss et al., 2015: table 1) for a total of four injections. However, this injection does not appear in his result’s tabular entry, where only three injections are indicated (Kreiss et al., 2015: table 2).

It is surprising that this “last experiment” was never mentioned in what can only be called the eulogies following Cedric’s death. Perhaps it was just too ignoble for his, by then, heroic status. Perhaps even more remarkably, Cedric’s identity was not revealed when all the work on him was formally published (Kreiss et al., 2015) or in all the media coverage of this publication as “new research” (UTas media release, 25 February 2015; see also ABC Online, 25 February 2015; ABC PM, 25 February 2015; Caruana, 2015; Hansen, 2015). One would have thought that the researchers having created Cedric as a personality would have been mindful of this fan club’s probable interest in his final contribution to science.

The surprise announcement of Cedric’s death came in a 1 September 2010 media release, which announced that Cedric had been euthanased in the preceding week because an x-ray had shown his facial tumour had metastasised to his lung (UTas media release, 1 September 2010; AAP, 1 September 2010; Gelineau, 2010). Cedric was six years old (Darby, 2010a). There was no mention of the lung tumour’s strain type.

Cedric’s story seemed an open and shut case. He died, presumably, from the last injection of live cells, Strain 3, which was different from the Strain 2 to which he had seemed to form an immune response after injection with dead cells.

6/05/2016 8:41 AM 303

But on the day his death was announced, that is, 1 September 2010, the team vet offered a less tidy assessment of what had happened. He said, “There is some suggestion that even that that tumour right at the very early days, the ones that killed him now may well have been the metastatic from that early challenge two years ago [presumably live Strain 2 in December 2007] …. It is only with this recent vaccination [presumably, live Strain 3 on 3 July 2008] that he was given that it looks like the cells that were involved with that vaccine were not the ones that killed him even though they triggered a lump to form (ABC World Today, 1 September 2010).” This apparent uncertainty about which strain, 2 or 3, actually killed Cedric has never been otherwise acknowledged, let alone clarified.

One can also not help asking if the last experiment involving Cedric was what may have actually killed him? If he died from a lung metastasis from a primary tumour discovered and removed in December 2008, then the time between the appearance of the primary (December 2008) and manifest ill health (August 2010) was about 17 months. This seems like a very long time. If on the other hand, he died from the final experiment, the time of his four injections (in the period September 2009 to June 2010) to manifest ill health (August 2010) would have been between 11 and two months, a more plausible time frame.

Admittedly, there was no mention of a primary tumour (below), so presumably there was none. On the other hand, very little is known about happens when live DFT cells (presuming some of the sonicated cells were still alive) are injected into a Devil’s body other than on the head (see “Distribution of Tumours on the Body”). Certainly, in nature, Devils bite other Devils on the body but tumours on the body are very rare.

Cedric’s body was buried for biosecurity reasons (ABC AM, 24 September 2010; Richards, 2010).

It is important to emphasise the only immune response reported for Cedric was that associated with the injection of dead cells of Strain 2. It was described in the original media release a “strong” (UTas media release, 29 November 2007); subsequently by the project’s vet as “very strong” (ABC World Today, 1 September 2010), and by the two principal researchers as with “high titres” (G. Woods and A. Kreiss pers. comm. in Obendorf and McGlashan, 2008). It was described by journalists as “good” (Darby, 2007b) and “robust” (ABC World Today, 1 January 2009). But after Cedric developed his tumours, one of the principal researchers described the immune response as being “fairly weak” (G. Woods on ABC News, 26 February 2009).

There was a third Devil involved in this story as well. In mid-2008, a female named “Christine” was reported as having “the same immune response as Cedric (Woods, 2008; see also 60 Minutes, 24 July 2008),” although the protocol used was not revealed. And two months after showing an immune response, she was still disease free (Menzies Research Bulletin 56, Spring 2008). This Devil, first confusingly called a “second Devil” (Woods, 2008) and “another Devil” (after Cedric) (60 Minutes, 24 July 2008) was later revealed to be Cedric’s mother (Menzies Research Bulletin 56, Spring 2008; see also Rex, 2008).

The details of the experiment on Christine only became apparent in March/April 2015 when the relevant PhD thesis work (Kreiss, 2009) was finally made generally

6/05/2016 8:41 AM 304

available (online). Christine (as TD 111) was injected with 8 x 107 irradiated Strain 3 DFTD cells at week 0 and challenged with live DFTD cells (strain unspecified) at week 32. She was found to have two tumours at the injection site(s) on week 48. Boost injections of 0.7 to 1 x 108, presumably irradiated or dead DFTD cells were made with Strain 3 at week 51 and with Strains 1, 2 and 4 at week 53. The tumours were surgically removed at week 58 but recurred after approximately 16 weeks. Shortly afterward Christine was euthanased for an unrelated health issue and upon autopsy was found to have no DFTD metastases. At no time did Christine show an immune response, calling into question how she could have been described as having “the same immune response as Cedric (above)”.

The work on Cedric, Clinky and Christine suggested at least four important things about the Devil’s immune response to the disease. First, at least some Devils may be able to mount an effective immune response to the dead cells of at least one strain of the disease. Second, immunity raised by one strain may not be sufficient to confer immunity to another strain. Third, the latency period between the time of injection and the time of the first appearance of a tumour can be as long as 22 months. And fourth, the primary tumour can possibly metastasise early in its development and the secondaries can go undetected for up to 20 months.

Further to the issue of how long metastasis can go undetected once a primary is detected and removed, possibly as long as 20 months in Cedric’s case, it is worth noting that only two other Devils have had their early-stage tumours removed in order to prolong their lives. Two of the three Devils that developed the disease at a Tasmanian wildlife park (above) had their tumours surgically removed (Duncan, 2007), but there was no public follow-up of how they fared.

Other Devils

There were also media reports of research on a possible immune response in other Devils. In June 2008, in conjunction with the then good news of Cedric’s apparent resistance, it was reported that a “special six” Devils from the west coast and with DNA similar to Cedric’s were to be investigated “to determine if they are also resistant to infection” (University of Tasmania media release, 2 June 2008; see also ABC News, 2 June 2008; Luntz, 2008).

In December 2008, after the first bad news about Cedric had been released, it was reported that “dead cells would be injected into more than 20 other Devils to try to replicate the immune response” (Darby, 2008; see also ABC AM report, 31 March 2008; Bevilacqua, 2008; The Mercury, 27 February 2009; Ball, 2009; Brand Tasmania Newsletter, February 2009).

Some “results” of experiments on other Devils were hinted at in an interview with one of the researchers in December 2009. The researcher was reported as having indicated, “Only one out of six Devils mounted an immune response following recent vaccinations (Borrell, 2009).”

These six Devils were probably the ones used in the researcher’s just-submitted PhD thesis (Kreiss, 2009, submitted in September 2009. The abstract of this thesis says, “To investigate whether immunity against DFTD could be induced, six Tasmanian

6/05/2016 8:41 AM 305

devils were immunised with inactivated DFTD tumour cells. Most of these animals exhibited a very weak immune response against DFTD. Three of these devils were challenged with live DFTD tumour cells and succumbed to the disease. However, one devil had a more prominent antibody response against DFTD and was protected against a first challenge, suggesting that immunity was protective, but likely to be short-term (Kreiss, 2009).” The Devil with the “prominent antibody response” was probably Cedric. As to the three Devils that succumbed to the first challenge with live cells, however, only Clinky seems to be clearly identifiable. Christine may have been a second, but the identity of the third Devil is a mystery. Interestingly, when these thesis results were finally published in 2015, only two of the original six Devils were mentioned: TD 1 (Clinky) and TD 2 (Cedric) (Kreiss et al., 2015). Furthermore, the published account said, “Due to ethical considerations, it was only possible to challenge two devils [TD 1 and TD 2] (Kreiss et al., 2015).”

At a 20-22 July 2010 immunology conference, an abstract reported that “… some devils were immunized with irradiated tumour cells. A small proportion of these devils responded to the irradiated tumour cells and one was challenged with live tumour cells. This devil was initially protected against DFTD but when challenged 12 months later the tumour developed. It is likely that the immunisation could only confer short-term protection (Woods et al., 2010a; emphasis added).” The challenged Devil was probably Cedric, and the others were probably those mentioned in Brown et al., 2010 (below).

At this same conference 2010, immunologists also reported on three experiments involving the injection of dead DFTD cells (Brown et al., 2010). In the first experiment, two Devils of unstated provenance were injected four times with irradiated DFT cells in the adjuvant Montanide and “neither … produced cytotoxic nor antibody responses.” This result was later published in 2011, where one Devil was said to have produced no cytogenetic response and the other a weak response, and neither produced an antibody response (Brown et al., 2011, as Devils Td 1 and 2, respectively and later Devils CD 8 and CD 9, respectively, in the thesis of Brown, 2013).

In the second experiment, four Devils of unstated provenance were injected three times with irradiated DFT cells and an even more stimulatory adjuvant (Montanide and CpG) and “none … produced cytotoxic or antibody responses.” In the underlying thesis work, the result would be modified to a “moderate” cytological response in all four Devils at best after at least the first dose and “… no clear evidence of any antibody development in any of the four devils…” Brown, 2013 and as Devils 10-13). In the published work, the results would be rendered as, “… all four devils …developed cytotoxic responses and three of them developed antibody responses (Kreiss et al., 2015). The different interpretation between Brown, 2013 and Kreiss et al., 2015 may have been due to an additional test for antibodies (ELISA) and a different statistical test in the published work.

In the third experiment, two Devils that had been previously unresponsive to irradiated DFT cells were injected with sonically killed DFT cells in stimulatory adjuvant (Montanide and CpG), and both formed “… moderate cytotoxic and antibody responses” (Brown et al., 2010). In the underlying thesis work, both Devils had moderate cytotoxic responses having a where cytotoxic response for both Devils

6/05/2016 8:41 AM 306 remained at “moderate” while one (CD 10) had “no clear evidence for antibody development” and the other (CD 14 – Cedric) “provided evidence of a weak antibody response (Brown, 2013).”

In mid-June 2012, the head of the main lab working on a vaccine for DFTD gave a public lecture in which he reported work of a graduate student (G. Brown), which showed that healthy Devil lymphocytes can be activated to kill DFT cells in vitro. DFT cells were activated by exposure to the adjuvant Con A for 48 hrs in one experiment and to Poly I:C for 18 hours in another and then mixed with live DFT cells. Depending on the concentration of activated lymphocytes to DFT cells, the Con A-activated lymphocytes killed up to 70 percent of the DFT cells and the Poly I:C- activated lymphocytes killed up to about 50 percent of DFT cells (Woods, 2012a: slide presented on-screen at 55:09 time mark).

In November 2012, a published abstract reported the same work as described in the preceding paragraph. Killer cells, a type of lymphocyte, stimulated with adjuvant Con A, killed up to 70 percent of DFT cells in culture. Furthermore, when such activated cells were injected into live tumours, there was an increase in the number of other immunologically active cell types within the tumour (Brown et al., 2012). This work has not been published, but the first part was almost certainly the same experiment as discussed in the public lecture in June 2012 (preceding paragraph) and the second part of the work would be described in an online thesis (Brown, 2013.)

In summary, the 2013 thesis work dedicated largely to trying to stimulate an immune response against DFTD (Brown, 2013) carried out 15 series of injections of dead DFT cells and various adjuvants into 14 Devils (one Devil was used twice). The 15 injections led to cytotoxic reactions that were scored as weak to moderate in eight cases and negative in seven, and antibody reactions that were scored as weak in one, slight in one and negative in 13. It is interesting that this research was able to stimulate an antibody response in 15 Devils that was no better than slight to weak in two individuals, whereas the original experiment on Cedric was said to have elicited a “strong” response in early reports and his mother Christine was supposed to have had a similar response (above).

The 2013 thesis also reported an induced immune response that seems to have led to the reduction in the tumour (Brown, 2013; see also Brown et al., 2012). A four-year- old female Devil (CD 15; referred to as CD 16 twice on p. 62 but correctly thereafter) developed a tumour after an earlier experiment involving the injection of a DFT cell protein extract followed by a challenge with live DFT cells. The Devil was given three treatments, which led to the total regression of the tumour after 24 weeks. The treatments in the order given were: the Devil’s own killer cells, a type of lymphocyte, activated by having been place in the adjuvant Con A for 48 hrs; cultured DFT cells also “activated” in Con A; Con A supernatant rich in cytokines to induce expression of MHC I genes on the cells’ surfaces, and a final dose of activated cultured DFT cells. Discounting time lags, the greatest effect, that is when the tumour started to regress, was following the treatment with activated cultured DFT cells (Brown, 2013: fig. 5.1).

Because the activated DFT cells seem to be the most effective treatment for tumour regression, two other Devils (DD 11 and 18) with DFTD were treated with activated

6/05/2016 8:41 AM 307

DFT cells. However, there was no cytotoxic or antibody response, and the two Devils developed rapidly growing tumours at the injection site within 14 days (Brown, 2013: fig. 5.9).

In 2012, two large grants were awarded to the research on a vaccine. One grant for the years 2013-2015 was, in part, to “… test if activated immune cells can protect against this disease (Menzies Research Institute News Archive, “National funding success for Menzies, 2012,” emphasis added). The other grant for the years 2013-2016 was to “… develop an immune enhancing vaccine to protect Tasmanian devils against …” DFTD. This grant application stated, “Our team has shown in proof of concept that the cancer cells can be recognised by the devil immune system. This project will develop and test a vaccine against the tumour, which will ultimately protect devils in the wild (emphasis added).” What is interesting is that these two grants, totaling just over $ 790 k, seem to have been won largely on the basis of results of an elicited immune response in Devils that were unpublished at the time and only partially published for the first time in early 2015 (Kreiss et al., 2015).

Along these same lines, it is worth remembering that thousands of dollars of donations and grant money was almost certainly raised and expended largely on the strength of the media exposure of one Devil’s (Cedric) single immune response.

The immunisation work on Devils as of the latter part of 2012 was perhaps best summed up in the observation that “immunisation of Tasmanian devils with various formulations of tumour cells and adjuvants has met with limited success. But this limited success has provided evidence that some devils can produce a weak antibody response (Woods et al., 2012).”

In early 2013, a small grant was awarded that may reflect the progress made to date in developing any sort of vaccine against DFTD. The grant was “to develop a strategy to develop and produce an effective vaccine to protect Tasmanian devils against Devil Facial Tumour Disease (STDP website: “Funding boost for the survival of the Tasmanian devil,” 25 April 2013). This sounds alarmingly like a plan for a plan.

Finally, it may be worth asking if any of this research into stimulating an immune response against Devil facial cancer has advanced further than similar research on stimulating an immune response against mouse and humans cancers. If there is a breakthrough here, it should be flagged as such, but if it is just following early mouse and human cancer research, then one has to ask what are the chances of it ever vaulting ahead?

Comments on an immune response

Considering there has been some evidence for an immune response both in the wild (the early blood samples from West Pencil Pine that apparently had antibodies) and in the lab (the early antibody-eliciting experiment in Cedric and perhaps others), it is odd that researchers, even those close to the work, have commented as if Devils have no capacity for an immune response to DFT cells. For example, in early 2010, five researchers, including at least two who had long been working on Devils wrote, “Devils do not mount an immune response to DFTD (Siddle et al., 2010a).” In 2011,

6/05/2016 8:41 AM 308

one of these two would write, “There is no evidence of devils mounting an immune response to DFT (Belov, 2011).”

In October 2010, one geneticist closely involved in the work wrote, “there is no evidence of devils mounting an immune response to DFTD …,” but then also wrote, contradictorily, “… preliminary studies suggest that at least antibody responses can be generated to DFTD (Obendorf and McGlashan, 2008) (Belov, 2011).”

In a research article completed in May 2011, an immunologists wrote, “the infected devil succumbs to the disease without evidence of an immune response to the tumour (Brown et al., 2011),” but the cited reference for this comment (Woods et al., 2007) contains no such evidence.

In 2012, a geneticist working on the Devil wrote, “Anti-DFT antibodies have not been seen in affected adult Devils (Belov, 2012).”

Even in mid-2014, authors of review article wrote about the “lack of an immune response to DFTD (Siddle and Kaufman, 2015).”

A Vaccine Against DFTD?

The development of a vaccine has long been a core objective in the effort to save the Tasmanian Devil (PHVA, 2008), and the idea has been so actively promoted that it ranks high in the public’s expectations for the disease. This is despite the fact that the development of a vaccine against “cancer” in well-known animals such as humans and mice remains elusive despite billions of dollars of funding and decades of research. As the STTDP’s past and only Senior Scientist said succinctly, "Given the limited progress in developing vaccines against human cancers, despite huge investment in research, hoping a vaccine can be developed against [DFTD] seems optimistic (H. McCallum, in Darby, 2011; also Luntz, 2011)."

Undaunted, however, a past STDP Science Manager said, “Let’s say everything does absolutely perfectly, and in fours years’ time some of our great scientists have developed a successful vaccine against DFTD. Now a vaccine against a cancer is going to get worldwide news; cancer biologists and oncologists will surely prick up their ears (C. Boland, in Siegel, 2012).” Indeed.

The first indication in the public record that anyone was thinking about a vaccine against DFT was a media story on 25 January 2005, which noted, “it's not clear whether a vaccine will be possible, because the devil's immune system is poorly understood … (Gliddon, 2005).” This view must have come from someone close to the work on the disease at the time.

The first mention that a vaccine was an official goal of DFTD research was in 29 March 2005 media report, which said, “a diagnostic test, and ultimately, a vaccine are the scientific team's main aims (Peatling, 2005).”

The first indication work on a vaccine had begun was a 25 November 2006 media story that said, “embryonic work on a potential vaccine is also about to get under way (Denholm, 2006b).” There were no details, however, what that work involved.

6/05/2016 8:41 AM 309

By February 2007, researchers were “actively” pursuing a vaccine against the Devil’s facial tumour (McGlashan et al., 2007). Again, it was not clear what this work was, but at the time, researchers were beginning to inject healthy Devils with large doses of dead DFT cells in an effort to elicit an immune response (below).

What might actually be involved in making a vaccine has been described in various media reports. One researcher said we had “… to modify … [the] tumour cell or try to do something to make … [the] tumour cell look different to the Devil’s own immune system so that it will respond (G. Woods, on ABC News, 24 September 2010).” Another researcher said the work might immunise “… against a protein that is expressed by the tumour but not by normal devils (J. Graves, in Australian R&D Review, February 2010). Another researcher said she hopes to find, ''some fusion protein, two genes that have come together in the cancer, that might be specific to the cancer and not present in the individuals, that we could use to vaccinate the animals with (J. Deakin, in Canberra Times, 1 November 2010).'' And a student researcher said her lab was “… trying to find a marker, a protein or something else on the surface of the DFTD cells, which we can try to isolate and try to prepare into a vaccination, which will be targeted to induce an antibody response (G. Brown, in Lawes, 2010).”

The discovery in 2012 that DFT cells can be stimulated to express antigens on their surfaces and can even occasionally do this naturally under the apparent stimulation of near-by host lymphocytes (see “Immune response evasion genes,” above) opened another way to stimulate an immunological response. If any of these expressed antigens proved to be specific to all DFT cells and different from all healthy cells, there would be a theoretical basis for a vaccine. Healthy Devils could be injected with DFT cells stimulated to express the surface antigens. The unique antigens would then prime the immune system and thereby allow it to respond quickly when challenged by a few naturally transmitted DFT cells, if those cells expressed any “residual” surface cell antigens, as could be the case (Siddle and Kaufman, 2012; Siddle et al., 2013). This approach received three year funding in mid-2015 (University of Southampton media release: “Researchers Aiming to Produce Vaccine to Save the Tasmanian Devil,” 21 July 2015).

In November 2014, media stories reported that immunologists at the Menzies Research Institute had developed a “new immunisation technique” and would “now” field test it at Narawntapu (Day, 2014; Glaetzer, 2014b; Groom, 2014; Mather, 2014b; Hobart Mercury Editorial, 24 November 2014; STDP website: “New phase of the Save the Tasmanian Devil Program,” published 25 November 2014).

At the end of 2014, any vaccine was said to be at least two years away (G. Woods, in Clarke, 2014).

In January 2015, a media story said, “the first Tasmanian devil facial tumour immunisation … [would] be trialled at East Coast Natureworld in Bicheno this month (Clarke, 2015).” Nothing further was revealed about this trial.

In June 2015, a media story said that immunologists had inoculated 20 devils with tumour cells treated so the devil’s immune system could “hopefully recognise them.” The Devils would be released into a diseased area in the wild in September and then

6/05/2016 8:41 AM 310

be monitored by trapping to see if they remained free of the disease (Mounster, 2015; see also Waterworth, 2015). On 22 July 2015, the STDP said 19 Devils had been immunised and would be released at Narawntapu National Park in central-north Tasmania (STDP website: “Tasmanian devil vaccine field trials announced,” 22 July 2015). In an abstract of an oral presentation of 27 July 2015, however, it was said 17 Devils had been immunised and would be released (Pye et al., 2015a). Contemporary reports said all the Devils would be from the captive bred Devils in the Insurance Population (ABC News, 22 July 2015; Carter, 2015a; Darby, 2015a; Shannon and Lehman, 2015).

On 25 September 2015, 19 immunised Devils were released in Narawntapu National Park on 25 September 2015. The immunisation had not stimulated an immune response in one Devil, but it was released nonetheless (Carter, 2015b; Healthcanal online, 28 September 2015; Walter and Eliza Hall Institute of Medical Research media release, 26 September 2015).

Although 19 immunised Devils were released, several accounts said 20 Devils had been released, including the relevant Tasmanian Government minister’s press release (Discovery News, 28 September 2015; Gibson, 2015; Groom, 2015; Shannon and Lehman, 2015; San Diego Zoo media release, 28 September 2015; The Advocate, 26 September 2015;). This implies that one unimmunised Devil was also released (see The Advocate, 30 September 2015), which along with the one immunised Devil in which the immunisation did not elicit an immune response, makes a total of 18 successfully immunised Devils released. Some accounts, including that of the relevant Tasmanian Government minister, however, stated or implied that all 20 animals had been immunised (Carter, 2015c; Discovery News, 28 September 2015; Groom, 2015; Lohberger, 2015a; San Diego Zoo media release, 28 September 2015; Shannon and Lehman, 2015; The Advocate, 26 September 2015). All 20 animals had been microchipped (Darby, 2015b).

All the media reports citing the sex reported the total number of Devils released as 20 (Discovery News, 28 September 2015; Gibson, 2015; Groom, 2015; San Diego Zoo media release, 28 September 2015; The Advocate, 26 September 2015), and of these eleven were males from Launceston (presumably the Tasmania Zoo FRE) and nine females from the Bicheno FRE (The Advocate, 26 September 2015).

“Devastatingly,” within eight days, four of the released Devils had been road killed, remarkably, all outside the park. One site where two Devils died, the Frankford Highway near the turn-off to the park” (The Advocate, 3 October 2015; Wisbey, 2015), was at least 9 km from the nearest park boundary, and the other site, “Greens Beach Road near Beaconsfield” (Gramenz, 2015; Lohberger, 2015b), was at least 6 km from the nearest park boundary (assuming “near Beaconsfield” is anywhere south of the halfway point between Beaconsfield and Greens Beach on Greens Beach Road). That the Devils are moving such large distances quickly and extending well outside the government’s land tenure does not bode well for the recapture of the remaining released Devils.

From an immunological perspective, the primary purpose of the release was “…to effectively test the immunisation response against DFTD and to help refine and develop more effective vaccination techniques.” (STDP website, “Tasmanian devil

6/05/2016 8:41 AM 311

vaccine field trials announced,” 22 July 2015; “Narawntapu National Park Tasmanian devil translocation September 2015,” 25 September 2015; “Wild devil recovery,” c. 25 November 2015; see also Pye et al., 2015a). It is unclear why this could not have been done more effectively by keeping the animals in captivity.

A secondary purpose of the release was to “… provide some protection against any natural DFTD challenge these devils might face once released and thus mitigate this threat (Pye et al., 2015a).” Depending on infection rates in the indigenous animals in the population, the statistical assessment of the efficacy of the vaccine will be compromised by the deaths of at least two and possibly as many as four effectively immunised animals, And this is assuming all released and living animals can be trapped.

Few details of the vaccine, the protocol involved in the immunisation or the immune response of the 19 Devils released at Narawntapu were made available. In an abstract, the researchers referred to “previous immunization trials” involving “all four captive devils” (Pye et al., 2015a), but it is not clear which of several trials (Kreiss et al., 2009; Brown et al., 2012; Brown, 2013) were meant. In any event, from the public record, the injected cells were grown in culture and then killed (de Grooyer, 2015), and adjuvants were added to the injectate to help stimulate an immune response (Darby, 2015a; de Grooyer, 2015; Pye et al., 2015a). The Devils received five injections (Bolger, 2015b; Darby, 2015a) subcutaneously on a monthly basis (Pye et al., 2015a). The immunisations began on 9 February 2015 and preliminary results were expected in May (Pye et al., 2015a). The strength of the immune response was indicated only indirectly when the lead researcher, anticipating the follow-up field surveys, said they wanted to see if the “… immune response stays high …” (G. Woods, in Bolger, 2015b, emphasis added).

Despite the uncertainty of a vaccine, in December 2014, the STDP noted in the last four years of its new five-year plan, 30, 20, 60 and 60 “immunised” Devils would be “rewilded” each year (STDP, 2014c).

There have been various suggests as to how an injectable vaccine might be deployed. One suggestion was that healthy Devils from the Insurance Population could be injected prior to releasing them into the wild (Cadell, 2013; Jenkins, 2014). Another was that wild Devils could be injected during the breeding season to enhance the number of Devils breeding and raising their young (Kreiss et al., 2015).

Despite all the talk about immunising Devils, however, only one Devil has been demonstrated to have achieved immunity, that is, survived a challenge with live cells, after being immunised with dead DFTD cells, and this was to just one strain of the disease. This Devil (“Cedric,” above) subsequently succumbed to a second strain (below).

In a recent STDP fund-raising video an immunologist said of his vaccine research, “Recent breakthroughs in the laboratory have rekindled hopes of saving this iconic species in the wild” (G. Woods, on STDP website, “Devil vaccine research,” 6 February 2016). Perhaps this is the case among immunologist, but one would have to think that the most hopeful recent development in the Devil DFTD story is the stabilization and perhaps even signs of an increase in some local populations.

6/05/2016 8:41 AM 312

Immunotherapy

An immunotherapy - a therapeutic stimulus of the immune system - against DFTD is largely of theoretical interest for the development of a vaccine, because unlike a preventive vaccine, which could be deployed therapeutically across a large number of healthy Devils, immunotherapy would only be relevant for a small number of already diseased animals.

Although some DFTs have regressed naturally (see “Tumour Characteristics”), only one tumour has been stimulated to regress in response to stimulation of the immune system (Brown, 2013). The Devil in question was a female that developed DFTD when injected with some live DFTD cells to see if it had developed a protective immune response when vaccinated experimentally. When it developed a tumour, it became a candidate for immunotherapy.

The immunotherapy involved three sequential treatments: injection of mitogen activated live mononuclear cells (MAK treatment); injection of Concanavalin A supernatant-treated live DFTD cells, and intratumoural cytokine injection. The tumour seemed to stop growing and start regressing shortly after the second treatment. To test if this treatment alone might arrest tumour development, two diseased Devils were injected. However, they developed new tumours at the injection sites. Taken together, these experiments were interpreted to mean that all three treatments were necessary for tumour regression Brown, 2013).

It would have been interesting to know where the successfully treated Devil came from and what its genetics were to see if its genetics may have had something to do with its ability to respond to treatment. Presumably, the relevant information and tissues are available for future analysis.

Other Therapies

Researchers have trialled other therapies besides experimental vaccines. In October 2008, a private consulting company announced online that the Tasmanian government had funded a collaborative project (with University of Sydney and DPIW researchers) to investigate “chemotherapy agents [note plural] effective against the Tasmanian Devil Facial Tumour,” especially those used successfully in a transmissible cancer in dogs (Veterinary Oncology Consultants, 21 October 2008). This work began in August 2008 and was published in 2013. It showed that the injectible drug vincristine, which is effective in a wide range of cancers in a wide variety of animals, was not effective in Devils (Phalen et al., 2013).

A 2008 honours thesis apparently trialled a number of standard cancer chemotherapies with “all results suggesting DFTD cells were found to be highly resistant to chemotherapy in vivo (Brown, 2008, as reported in Brown, 2013).

In late 2008, immunologists mentioned that cytotoxic drugs could kill the tumour cells, but “…when trialled on diseased Devils the chemotherapy was [only?] partially effective (Menzies Research Institute Annual Report, 2008). This work appears not to have been published.

6/05/2016 8:41 AM 313

A March 2010 media story based on a visit and interviews with the staff at the quarantine facility at Taroona on the outskirts of Hobart mentioned in passing that “chemotherapy trials are being carried out at other sites [note plural] (Smith, 2010).”

In June 2010, the head of the Animal Health Laboratories in Tasmania was reported as saying his facility was investigating several possible chemotherapeutic treatments (Martain, 2010). One of these was a drug, EBC-46 based on an extract from the seed of a tropical Queensland rainforest tree, Hylandia dockrilili, a euphorb. In July 2013, media reports said the drug had been trialled on four diseased Devils. It had been injected directly into and at the edge of the animals’ primary tumours. This had resulted in “… either regression or palliation of the tumours in all animals (S. Pyecroft, in Money, 2013.)” Another account of the result was more dramatic, "Within 20 minutes, or half an hour, of that treatment you could see that there was a change in the tumours and within 24 hours at a local site you would see that the tumour had basically been killed (S. Pyecroft, on ABC News, 1 July 2013)." The drug has been described a working by cutting off the tumour’s blood supply (Mounster, 2013) and by stimulating the immune response (ABC Hobart, 1 July 2013; (STDP website: Anti-cancer drug treatment trial for DFTD completed, 1 July 2013; ABC online, 3 July 2013). It was not clear if the drug actually totally eliminated the primary cancer or only slowed or arrested it. Because the drug has to be injected, it would only be useful on external primary tumours and not on internal metastases. It was also likely that the drug would only be practical in captive animals, say genetically important females carrying pouch young. The paper describing the results was to have been published before the end of 2013 (S. Pyecroft, ABC News, 1 July 2013).

The DPIPWE’s annual report for 2010 noted that “chemotherapy drugs have proven to be ineffective in treating tumours in devils and this work is close to completion,” but “[t]rials will commence in 2010 on new therapeutic drugs which have proved highly effective in treating cancers in dogs.” It is likely this was a reference to the drug EBC-46.

In the STDP’s annual report for 2010/11, one of the projects listed to be undertaken in the 2011/12 financial year was “chemotherapy/treatment trials for devils with DFTD.”

The STDP’s annual report for 2011/12 mentioned several drug initiatives. One involved “two, or possibly, three” potentially orally delivered drugs that can apparently kill DFTD cells in vitro. A trial of this treatment, apparently in live animals, is proposed (STDP Annual Report, 2011/12). Such a drug could possibly be delivered in a bait in the wild.

Another treatment involves an injection directly into the tumour, causing the tumour to die and slough off. A trial of this treatment is apparently currently underway (STDP Annual Report, 2011/12). This was possibly a reference to EBC-46.

In July 2013, the past head and principal pathologist of diagnostic services at the Tasmanian government’s veterinary laboratories revealed that “we had tried all the usual treatment[s] for animal carcinoma (S. Pyecroft, in Money, 2013; also ABC Hobart, 1 July 2013; ABC online, 3 July 2013).”

6/05/2016 8:41 AM 314

Perhaps two of these drugs were the chemotherapeutic compounds doxorubicin and carboplatin, which were reported as ineffective in treating DFTD in late 2015 (Phalen et al., 2015).

A personal communication indicated that DFTD cells are “irradiation resistant” (G. Woods, pers. comm. in Siddle and Kaufman, 2015), but as DFTD cells are routinely killed by irradiation in the lab, this must mean that DFT tumours are irradiation resistant.

In 2010, immunologists succeeded in getting Devil facial tumour cells to grown in immuno-deficient lab mice (Kreiss et al., 2011a). As a control, the injected live tumour cells in healthy mice and found, as expected, that the cells did not grow (Kreiss, et al., 2011a). In 2014, they injected dead DFTD cells into healthy mice, and found all the expected immune responses (Pinfold et al., 2014). These results showed the mouse will provide a useful model in which to test DFTD cells that have been manipulated in a quest for a vaccine or an immunotherapy.

Genetics of the Immune Response

It is not clear when work actually began on the Devil’s major histocompatibility complex. In 2004 researchers suggested for the first time that the genetic differences between Devil populations in the northwest and those further east (Jones et al., 2004a) might hold the key to resistance to the disease (Loh et al., 2006b). And the late 2006 or early 2007 observational and experimental evidence showing that at least some Devils from the northwest were able to mount an immune response would re-enforced this thought.

In any event, in May 2007, geneticists were able to submit two articles suggesting that it was lack of diversity in the MHC that was probably facilitating the spread of the disease (Siddle et al., 2007a-b). Critically, however, the researchers had no specimens from the northwest population (ABC News and ABC PM, 3 October 2007).

This lack of material was apparently quickly overcome, however, because the minutes of the 23 October 2007 meeting of the steering noted that “Committee members discussed recent research that show that a percentage of western devils in Tasmanian express some genes that are not found in eastern devils or the devil facial tumour, and that these differences may correspond to some form of resistance to the deadly tumour (Communiqué – Steering Committee Meeting, 23 October 2007).” In November 2007, researchers reported by media release that “… some western devils are genetically distinct from eastern devils and the devil facial tumour that threatens it” (University of Tasmania media release, 21 November 2007; see also Darby 2007) and that “Cedric” was “one of the western devils that have been found to have slightly different MHC … gene expression compared to the tumour itself (University of Tasmania media release, 29 November 2007).”

In March 2010, the geneticists formally published the results underlying the informal reports. They showed that Devils have two groups of nucleotide sequences in the Class I region of the MHC part of their genome, and while eastern Devils and the facial tumours have both, northwest Devils often have only one or the other. It was

6/05/2016 8:41 AM 315

hypothesized that having only one group might allow an animal to recognise the “non-selfness” conferred by the other group in the tumour and mount an immune response to it (Siddle et al., 2010a-b; also Kreiss et al., 2011b). Unfortunately, the result was due to a technical error, which rendered it void (Cheng et al., 2012a; Lane et al., 2012). The population continues to show resistance, but its basis is unknown (Lane et al., 2012).

Resistance in Wild Populations?

Researchers may have first realised that some wild Devils might be at least somewhat resistant to the disease in June 2005, when field workers found one individual that had lived with the disease much longer than other Devils, nearly one year (Dennis, 2006). And this view would have been reinforced by the subsequent discovery of two more individuals described as “partly resistant,” presumably on the same criterion as the earlier animal, longevity (Dennis, 2006). Where these animals were found and what became of them was never mentioned again.

Sometime in the last half of 2008, the Australian Research Council, a Commonwealth Government agency, gave $ 400 000 to researchers to initiate, what seemed, on the face of it, a wider survey for a natural immune response. As one of the principal researchers said: “We're going to start surveying animals in the wild just to see where there's any evidence of any animals who have been exposed to the tumour and produced immune response. So we're going to collect serum samples from a range of devils just to see if there's any animals out there that may have responded (ABC AM, 17 October 2008; see also ABC News, 15 October 2008).” To date, no results of this survey appear to have been published. Presumably, all results have been negative, because positive results would be “good news,” capable of profoundly influencing the course of future funding, research and conservation, in addition to being of intense public interest.

West Pencil Pine

The first clear public indication there might be a population with natural resistance was at a forum to discuss DFTD in February 2007. A field researcher reported that populations at two unnamed localities where the disease had recently arrived were being sampled, and “the proportion of susceptible and infected individuals seems to remain relatively stable after 10 months of data collection. No evidence on shift of age structure related to presence of DFTD has been observed up to date (Hamede, 2007).” At least one of these localities was almost certainly West Pencil Pine, near Cradle Mountain (below). It was subsequently revealed that researchers had begun “formally tracking” this population in May 2006 (Keim, 2011; Hamede et al., 2012a; Lane et al., 2012; but see Hamede et al., 2015, where the first data points are for August 2006). This report suggests researchers may have suspected there was a resistant population in late 2006 or early 2007. Interestingly, the second unnamed population being sampled was never made clear.

The observations made at the February 2007 forum may have been the basis of a comment later in 2007 that “there are very early signs that there may be a few resistant animals out there (H. McCallum, in DPIW, 2007a).”

6/05/2016 8:41 AM 316

The first public acknowledgement of the unusual site came in an April 2008 article on the STTDP website (“Mapping the disease,” published 8 April 2008), which mentioned that “one population in the Central Highlands … seem[s] to be less susceptible to the disease,” and “research is being carried out to investigate why this may be the case.”

On 22 May 2008, the relevant minister was quoted as saying, "Some devils in the western half of the state have survived contact with the disease and have been found to have antibodies in their systems (Clark, 2008).”

On 29 May 2008, a Tasmanian Government media release mentioned publically for the first time the presence of antibodies to DFTD in a wild population, almost certain West Pencil Pine. The release said, “several wild caught devils from … [a western] area have already developed antibodies to the tumour (Tasmanian Government media release, 29 May 2008).”

The specific location of the quasi-mysterious population was narrowed further in June 2008, when the media reported that although the disease had been in the “Pencil Pine” area for several years, it was “unusual that not all the Devils had been infected (ABC News online: Search for immune gene in Tas Devils, 9 June 2008).”

Researchers themselves first explicitly named West Pencil Pine as the location of the resistant population and the presence of antibodies in some of its animals at a specialists’ meeting in early July 2008 (PHVA, 2008). A working group that included two immunologists wrote, “Three devils in West Pencil Pine have generated antibody responses to DFTD, suggesting they have been exposed to the disease and seroconverted (PHVA 2008).”

Apropos of when researchers first became aware of the presence of antibodies to DFTD in a natural population, on 1 January 2009, the field researcher who was sampling the Devils from the “Cradle Mountain area” said, “In one of … [the biopsy] samples we sent [to the immunologists in Hobart] a couple of years ago …there were a couple of animals that had been tested with antibodies of the DFTD…. In other words, they were exposed to the pathogen, but they didn’t show clinical symptoms. So three years after that those animals never showed clinical symptoms (R. Hamede, on ABC World Today, 1 January 2009; see also Hamede et al., 2012a).” This account suggests that if the researchers had processed the samples promptly, they would have found antibodies in a wild population as early as late 2006 and no later than early 2007 (a “couple of years” prior to the likely late 2008 date of this interview).

In October 2008, a media story reported a Devil tumour immunologist as saying that in early 2009 a state survey would start looking of other disease-resistant animals by testing for an immune response (Heath, 2008).

A PhD thesis submitted in September 2009, but not released generally until five and a half years later (March/April 2015), provided the first data for anti-DFTD antibodies in a wild population. The work tested for anti-DFTD antibodies in a sample of 17 Devils from West Penicil Pine in March. Three Devils “had evidence for the presence of anti-DFTD antibodies,” but only one, a female with pouch young, had titres high

6/05/2016 8:41 AM 317

enough to be convincingly above possible natural variation. This Devil was still healthy when recaptured a year later (Kreiss, 2009).

In late 2009, TDRAC gave a grant to test “whether DFTD tumours are evolving to overcome genetic resistance in some Tasmanian devil populations” (STTDP website: “Public generosity funds further devil research,” first published 19 October 2009) or “whether the DFTD tumours are evolving to overcome natural genetic resistance in some devil populations (UTas media release, 21 August 2009; emphasis added; note plural “populations”). An existing genetic resistance to the DFTD is clearly implied. The application for this grant would have been submitted no later than 31 May 2009.

In March 2010, researchers confirmed the apparent resistance in the West Pencil Pine population when they wrote, “… disease prevalence and population impacts appear to be lower than expected at West Pencil Pine (Siddle et al., 2010).”

Further insight into this population came in August 2010, when one of the principal immunologists involved with the “Cradle Mountain area” population said, “We have screened about 40 devils and at least three of them and possibly five – two results are not as strong – look as if they have antibodies (G. Woods, in Denholm, 2010e).” The details of these important observations were never published.

In a mid-March 2011 interview discussing the work just mentioned, a researcher noted that in the three years DFTD had been in West Pencil Pine, a field survey had found only 26 diseased Devils in the population (K. Belov, on ABC Science Online, 10 March 2010). As the number of apparently unaffected animals was not mentioned, no estimate could be made of just how resistant this population may have been in that period.

In mid-2011, researchers began speaking publicly about a possible resistance in some northwest populations. A research application made public on 3 May 2011, spoke of the “underlying apparent resilience” to the disease of the northwest population (DPIPWE website: “Application for scientific research permit,” 3 May 2011). And in June 2011, two major Devil researchers wrote that in the northwest devil population the tumour “is not causing any population decline;” “disease prevalence remains low; and infected devils … survive with the cancer much longer (Jones and McCallum, 2011b)” (see also M. Jones, video on The Australian website, 7 September 2011).

In July, 2011, the STDP website mentioned there was “information on the apparent resistance to DFTD” in “a percentage” of the population in the Cradle Mountain area (STDP website: “Grant to fund facility upgrades for “Devils@Cradle,” 12 July 2011). Presumably, this meant the West Pencil Pine population.

In August 2011, the STDP made the fullest disclosure on the West Pencil Pine population when it noted that the population had been studied for five years and “… abundance throughout the population has not plummeted, the full age structure is still intact and tumours appear to grow slower, regress and take longer to kill the animal.” It went on to say that “the STDP, in collaboration with UTas, will begin trapping a neighbouring and similar population to determine whether this atypical response occurs in more than one population and whether it is the disease or the devil

6/05/2016 8:41 AM 318

population that is different (STDP website, “Mapping the disease,” published 18 August 2011, revised 31 August 2011).”

In September 2011, one researcher noted that at a study site “near Cradle Mountain” they were “not seeing population decline… or loss of older animals out of the population;” prevalence is very low,” and animals seem to be “surviving much longer with the tumours” (up to 12 months after first signs of the disease) (M. Jones, ABC Science Show, 3 September 2011). Presumably, this was another reference to the West Pencil Pine population.

In the same month, a researcher said that “some animals might be genetically resistant” to the disease (H. McCallum, on ABC Brisbane Online, 8 September 2011). Without an accompanying geographical qualification, it is reasonable to assume he was thinking of West Pencil Pine.

In a paper submitted for publication in January 2011, revised and then accepted for publication in June 2011, published online in October 2011 and published in hardcopy in 2012, researchers provided the first full analysis of the West Pencil Pine population. Their succinct summary was “ … at West Pencil Pine disease prevalence did not increase rapidly in any age class and there were no changes in population size or age structure or significant declines in population growth rate (Hamede et al., 2012a).” The paper also noted that “the strain at West Pencil Pine is tetraploid (A.M.P. unpublished) (Hamede et al., 2012a; emphasis added).”

In a public lecture in July 2012, a leading Devil researcher noted that at West Pencil Pine in the period from May 2006 to just before August 2011, there had been “… no population decline, no change in age structure, [and] very low prevalence [of the disease]” in the population (Jones, 2012; see also Fig. 11). She also revealed that Devils with tumours at West Pencil Pine had been living longer than at more eastern sites, Fentonbury being the one site named specifically. Some WPP diseased Devils had apparently lived as long as 18 months. She also noted that in the last two or three years (that is, previous to July 2012) at WPP, there had been five Devils in which tumours had regressed and at least three became DFTD free (Jones, 2012). One complete regression occurred between February and August 2009 and another between February and May 2012 (Jones, 2012). Tumours of both diploid and tetraploid origin had regressed.

It appears, however, that conditions deteriorated at West Pencil Pine in August 2011, just following the optimistic report on the site (Hamede et al., 2012a). In the July 2012 public lecture, the speaker announced that starting in August 2011, “prevalence [of the disease at the site] went up very high and is now in excess of 50 percent, and the population is starting to decline and starting to look like an eastern site” (Jones, 2012; see also Hamede et al., 2015, which also almost certainly pertains to this cite). She noted that prior to mid-2011, the predominant strain(s) of DFTD was a, presumably slower growing, tetrapod strain and that in mid-2011 the site was invaded by a presumably faster growing diploid strain, almost certainly Strain 1 to judge from her accompanying image (Jones, 2012: time marker 57:40, reconstructed here as Fig. 22). However, to judge from the image, it would be a close call to say that a tetraploid strain(s) had been predominant at the site prior to mid-2011. Recall also that it had been said that “the strain” at the site prior to mid-2011 was tetraploid (Hamede

6/05/2016 8:41 AM 319

et al, 2012a, above).” Furthermore, the diploid Strain 1 was already at the site in 2006 (Pearse et al., 2012: table 1), the year the WPP surveys began, and the only other diploid strain at the site, Strain 2 has been there since 2008 (Jones, 2012; Pearse et al., 2012; Fig. 22). Another interpretation of the data is that rather than an invasion of Strain 1, Strain 1 was already at the site and either it increased in virulence or Devils lost resistance to it.

Number of New Cases of DFTD by Strain Type per Year at West Pencil Pine

15 S1 S2 10 T1 of DFTD 5 T2

Number of New Cases 0 2007 2008 2009 2010 2011 2012 Year

Fig. 22. The number of new cases of DFTD by strain type per year at West Pencil Pine between 2007 and 2012. Figure legend: S1 – diploid Strain 1; S2 – diploid Strain 2; T1 – tetraploid form of Strain 1, and T2 – tetraploid form of Strain 2. Redrafted based on data in Jones, 2012: figure at time marker 57:40. Note that data for 2012 include only the first two of the four sampling periods carried out each year.

In the latter part of 2012, the STDP noted for the first time in print that at West Pencil Pine, not only did some tumours regress, some had disappeared altogether (STDP Annual Report, 2011/12). A popular account of the Devil published in May 2014 noted specifically that tumours had regressed and disappeared in five Devils (Rehmeyer, 2014). In September 2012, the STDP reported for the first time in print that “there has been a recent increase in infection rates” in the population (Newsletter, September 2012). A 2014 review noted that “five years after disease arrival the prevalence increased to 50% and the population declined (Siddle and Kaufman, 2015),” but the citation given for the assertion (Hamede et al., 2012a) contains no such observation.

In August 2012, the STDP noted that “recent evidence” suggests that, contrary to previous general belief, the apparent resistance in the West Pencil Pine population was not due to the Devils but to the strain of the disease in the population (STDP, 2012b). They were undoubtedly referring to the previous high prevalence of [slow- growing] tetraploid strains of the disease.

It should be mentioned that it was the West Pencil Pine population in which immunogeneticists thought they might have found a genetic difference that might account for the resistance to DFTD (Siddle et al., 2010). This result, however, was later revealed to have been due to a technical error (Cheng et al., 2012a; Lane et al., 2012).

6/05/2016 8:41 AM 320

In a paper received for publication in January 2015, accepted in April and actually published in June 2015, immunologists discounted the earlier report of “antibodies of the DFTD” (above) by saying this “… still needs confirmation, as it is unknown whether these Abs were specific for DFTD cells or merely cross-reactive (Woods et al., 2015)”.

This confirmation was seemingly forthcoming in an abstract of an oral presentation made in July 2015. It was reported that flow cytometry had revealed “… evidence for a naturally occurring immune response in wild devils from a site where DFTD first arrived in 2006 [almost certainly West Pencil Pine]. Remarkably, a serum antibody response has been detected in a small number of these devils. This preliminary evidence of a naturally occurring immune response to DFTD … is note-worthy (Pye, 2015; emphasis added).” Use of the word “remarkably” suggests that the researchers did not entertain the possibility of an immune response despite having an indication of one; otherwise how to explain the at least eight year delay in pursuing confirmation of that first indication?

Despite the most recent, seemingly definitive, comment on anti-DFTD antibodies in the West Pencil Pine population, it would be useful to have a review of this research that would include a description of the antibody tests, sample sizes, health histories and timelines as well as any accompanying genetic analysis.

Resistance in other populations?

In addition to the West Pencil Pine population, there are other populations that have been mentioned briefly that raise the question of resistance within them. In August 2011, the proprietor of a zoo at Cradle Mountain, 20 km east of West Pencil Pine, noted that after five years (2006-2011) of observing and monitoring in and around his country property wrote, “we see a healthy population of devils year round and very little sign of DFTD (Anthony, 2011).” Arguably, this population could be considered part of a greater West Pencil Pine – Cradle Mountain population.

In the latter part of 2011, the STDP’s annual report mentioned an apparently disease- free population but did not give its location. “Recently cameras deployed within an area thought to be diseased provided an indication of a small population of devils which appears to be disease free. Further investigation through trapping supported this. It appears this population has been isolated from the disease by landscape features such as large rivers, mountain ranges and even urban development combining to prevent diseased devils from moving into the area (STDP Annual Report, 2010/2011).” In the latter half of 2012, the STDP identified the site as the Snug Tiers area near Hobart and suggested again that its disease-free status was due to its physical isolation (STDP Annual Report, 2011/12). It is concerning that the STDP is apparently assuming that geographic isolation is solely responsible for the lack of disease in this population and has no plan to assess it for possible resistance to the disease. No further information about this site has been forthcoming.

In the latter part of 2012, the STDP announced that researchers would set up a “replicate site” for the West Penicil Pine population at “Takone,” northwest of West Pencil Pine (STDP Annual Report, 2011/12). Whether this was because the site

6/05/2016 8:41 AM 321

shared similarities in resistance in its Devils was not made clear, but it is likely this was the case. In November 2014, this site was named as one of ten long-term monitoring sites (STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014), but no further information about this site has been forthcoming.

In April 2014, the STDP revealed that the population on Tasman Peninsula had apparently remained disease free (Newsletter, April 2014; see also STDP website: “New phase of the Save the Tasmanian Devil Program,” 25 November 2014, “Peninsula havens provide natural barriers to devil disease,” 6 November 2015 and “Helping our devils fly home,” 11 November 2015; STDP Annual Report, 2013/14; Hunt and Groom, 2015; Huxtable et al., 2015; Lazenby et al., 2015b; STDP, 2015b) despite the population on the Forestier Peninsula just to the north “being heavily infected with the disease”- STDP Annual Report, 2013/14).

Whether this population’s disease-free status was due to its relative isolation - the Tasman Peninsula is separated from the Forestier Peninsula by only a narrow sand spit, or was indicative of some inherent resistance is unknown and apparently unresearched.

The estimated size of the Tasman Peninsula population varied with reports. At the end of 2013, it was estimated at about 19 Devils (STDP Annual Report, 2013/14). In April 2014, it was reported as “20 to 30” Devils (STDP website: “Latest Program newsletter,” 30 April 2014; Newsletter, April 2014). At the end of 2014, it was given as 43, consisting of 23 adults and 20 juveniles (STDP Annual Report, 2013/14). And in mid-2015, it was said trapping in 2014 resulted in the capture of 18 adults and 2 juveniles and an estimate of 23 adults and 20 juveniles (STDP website: “Peninsula Devil Conservation Project,” 23 July 2015; see also Groom, 2015b; Hunt and Groom, 2015; Huxtable et al., 2015).

In March 2015, a study that involved trapping Devils at two localities, Cradoc and Judbury in southern Tasmania, in a 26-month period up to July 2013 reported that despite the disease being in the region for at least eight months previously there were no recorded cases of the disease (Fancourt et al., 2015).

There are thus six, perhaps even seven, populations that are possibly resistant to DFTD: Cradoc, Judbury, West Pencil Pine, Snug Tiers, Takone, Tasman Peninsula and Cradle Mountain (if this population is distinct from the WPP population).

Evidence of resistance in any population raises the question of whether those few Devils continuing to live in decimated eastern populations may owe their survival to either a pre-existing or developing resistance to the disease (McGlashan et al., 2007). This is especially true of Devils three years of age and older. Why these animals have seemingly never been checked for antibodies against DFTD is an enduring mystery, as they could be extremely informative about the potential for the Devil to “save itself” from DFTD without human intervention.

Furthermore, there is no evidence of any local population going extinct. Even long- affected local populations appear to be holding their own, albeit in much reduced

6/05/2016 8:41 AM 322 numbers (STDP website: “Populations persist at ‘ground zero,’ ” 1 June 2011; S. Fox, in Mounster, 2016).

Early indications based on the rate of decline within populations suggested local populations might die out within 15 years of the appearance of the disease (McCallum et al., 2007). But the oldest affected population at Mt William, where the disease was first noticed in 1996, is surviving (STDP Annual Report 2010/11). If the date of first notice is taken, conservatively, as the date of onset of the disease, then under the 15- year time frame, it should have gone extinct in 2011. On the most recent information, the population is down by over 90 percent (Hamede et al., 2012a).

Similarly, the disease first arrived in the population on Freycinet Peninsula in 2001 (Jones and McCallum, 2011b), but in the last year for which there are public data, 2007, it seemed to be holding its own (data in McCallum et al., 2007: fig. 3, 2009: fig. 5). In 2011, however, this population was said to be down by 90 to 95 percent (M. Jones, in Pearse et al., 2012; Jones and McCallum, 2011b; M. Jones on ABC Science Show, 3 September 2011)

As noted, the STTDP’s then Senior Scientist wrote in September 2008 that “it is clear that if some animals are resistant to the disease, this will totally change the overall management strategy for the disease and the prognosis for the species as a whole (STTDP, 2008c).” Despite there having been some good evidence of “animals resistant to the disease” even in September 2008, there has not as yet been any indication of a change of strategy. In fact, the first public call for research specifically on the “innate immunity” to the disease only came in October 2010 (Belov, 2011), 14 years after the disease first came to notice.

Public comments about resistance

Considering researchers almost certainly had some thoughts about the possibility of both resistance and antibody production in the West Pencil Pine population no later than early 2007 and subsequently found at least some Devils whose tumours regressed completely, some subsequent comments by some researchers are difficult to understand. For example, in spring of 2008, an immunologist wrote, “As far as we knew [when the experiments on Cedric started in March 2007], no devil in the wild had shown any evidence of [an] immune response to DFTD (Woods, 2008).”

In September 2008, the STTDP’s Senior Scientist commented, “as yet, there is no clear evidence that any animals are resistant to the disease” (STTDP, 2008c). He also said sometime in 2008, “…there is currently no firm evidence that any animals are either totally or partially resistant to disease (McCallum, 2008).”

In the latter half of 2009, five Devil researchers wrote, “… no recovery, any immunity, or resistance to DFTD [has been] observed Lachish et al., 2010).”

In the same year, the group that discovered the antibodies in the two wild Devils from what was almost certainly the West Pencil Pine population wrote, “There has been no evidence of resistance against DFTD in the wild population (Menzies Medical Institute Annual Report, 2009).

6/05/2016 8:41 AM 323

The Devil’s October 2010 management plan stated, “no devils with resistance or natural immunity have been found (DPIPWE, 2010d).”

Two modellers wrote in March 2011, “There is no evidence thus far of acquired or innate resistance to infection (Beeton and McCallum, 2011).”

Another researcher on the disease would write sometime in 2012 that “no resistance has been found in the wild (Kreiss, 2012).”

In a public lecture in mid-June 2012, one immunologist made repeated assertions about a lack of resistance: “Resistant animals. As yet none identified – therefore unlikely;” “No resistant animals;” 100 percent mortality;” any Devil that gets the disease will die from the disease;” there is no resistance” (Woods, 2012a).

In October 2015, however, an experienced Devil field researcher noted that, “in the last three years, healthy adult devils have started to appear in the longest-diseased areas of the Tasmanian East Coast. Seven devils at our study sites [which ones?] have recovered from tumours and some of these animals lived to a healthy old age (Jones, 2015b).”

In 2015, however, the immunologists were still claiming the cancer was invariably fatal: “No devils has survived infection …” (Kreiss et al., 2015); “… 100 % mortality …” (Pinfold et al., 2015); “… invariably fatal cancer …” and “… causes mortality in all affected animals …” (Pye et al., 2015c), and “… always fatal …” (Woods et al., 2015).

In light of the apparent ability of some individuals in the wild to resist the disease, it is remarkable so little effort as been devoted to researching resistance in the wild generally. A 2013 PhD thesis examined some wild-caught Devils with DFTD for evidence of cytotoxic and antibody production. It found eight Devils (DD 1-8) from the Forestier Peninsula (N = 5) and West Pencil Pine (N = 3) had no evidence of a cytotoxic response. It also found that either 12 (text, p. 65) or 13 (fig. 3.2; DD 1-13) Devils, from the Forestier Peninsula (N = 10) and West Pencil Pine (N= 3), showed no evidence of antibody production (Brown, 2013: figs 3.1 and 3.2, respectively).

Such surveys should be expanded to other affected populations and should concentrate not only diseased Devils, whose lack of resistance is already evident, but on older healthy Devils in diseased populations, whose exceptional longevity is intriguing and perhaps indicative.

The discovery of naturally resistant individuals or populations would be the most promising development in the history of the disease and would certainly be the best outcome – at least for the Devils. But what would happen to the now substantial “Save the Tasmanian Devil Industry” and the rivers of gold in public donations, government grants and entrance fees that support that industry if it was discovered the Devils were capable of overcoming the disease by themselves?

Genetics and “Resistance”

6/05/2016 8:41 AM 324

To date, the only observation about a possible relationship between the genetics of Devils and possible resistance to DFTD is the observation that DFTD is not yet known in Devils with some of the 25 identified MHC types (Saddle et al., 2010). However, the sample sizes were small and other aspects of the animals studied, such as age (too young to be susceptible?), were not given, making it impossible to test the possibility, statistically. Hence it would be premature to draw any inferences about MHC type and susceptibility to DFTD.

Devils vary geographically in several genetic features, MHC types (Siddle et al., 2010), mitochondrial haplotypes (Miller et al., 2011; Murchison et al., 2012), and the deletion in the MHC Class Ia gene Saha-UA (Cheng et al., 2012). Variants in all these markers are unusual in prevalence or unique in kind in the extreme northwestern part of the state where it seems the spread of the disease is slow or perhaps even arrested. This suggests that the Devils in this area could be resistant to some degree. This fact, plus the variability in the other parts of the state where the disease is widespread, suggests strongly that in any affected areas, Devils older than two years of age should be screened for their suite of variable genetic features in order to identify variants that might be associated with resistance. It is not clear if there is any plan to pursue such research, but it is as strongly indicated as is that for the screening of antibodies in older animals in affected areas.

Interventionism and the Devil

There never seems to have been a question that humans would intervene to try and stop the spread of the Devil’s facial tumour disease. This decision was no doubt made on the basis of the disease’s uncertain cause (natural or human, the latter implying more responsibility); rapid spread; apparent 100 percent mortality; presumed density independent transmission; absence of any preventative or cure, and the likely secondary impact on other species, both native and exotic. The state government would have also been acutely aware of the opprobrium that would follow if, after having just lost the world’s previous largest living carnivorous marsupial, the Thylacine, it also lost the world’s then second and now largest living marsupial carnivore within the space of 100 years (McCallum and Jones, 2006; Editorial, Hobart Mercury, 14 November 2003). Furthermore, the gross nature of the tumour seems to have aroused human sensibilities to a pitch that an equally rampant and deadly, but less obvious disease such as tuberculosis, ever would have (N. Mooney, in Pincock, 2007 and on ABC Rural Bush Telegraph, 20 September 2011). And these sensibilities have, no doubt, been stoked further by some researchers’ recent allusions, qualified as they may be, to the possibility of a transmissible cancer arising in humans (K. Belov, in Cart, 2011; Jones and McCallum, 2011b; M. Stratton, in Wellcome Trust Sanger Institute media release, 16 February 2012).

Interventionism is currently the dominant conservation philosophy. Indeed, it is now a major industry. But it has one major hazard. Being the dominant philosophy, it can unconsciously or otherwise discount natural solutions. Evidence for this discounting in the Devil’s case is evident in two aspects of the Save the Tasmanian Devil Program. It is quick to report potential successes of interventionism, but it is slow to report both the failures of interventionism and, more importantly, the potential for natural solutions.

6/05/2016 8:41 AM 325

Examples of slow or inadequate reporting of interventionist efforts that fell short of initial promises include the lack of follow-up of the pre-symptomatic blood test for the facial tumour and the lack of publication of the immunisation experiments on Cedric, Clinky, Christine and other Devils. Related to this is the apparent lack of review of how it came to be that at least three projects announced with great fanfare were subsequently curtailed for reasons that should have been apparent from the outset. These programs include the culling of diseased animals on Forestier Peninsula, which subsequent (!) modelling showed was doomed from the beginning; the plan to fence off of Woolnorth in the northwest part of the state, which floundered when the burrowing abilities of wombats and the effect of gravity on trees were appreciated, and the stocking of mainland zoos with Devils that were very likely going to be unfit for re-introduction to the wild after at least 25 generations under artificial selection.

Examples of the inadequate reporting of promising natural solutions include the lack of regular updates of recent results for continuously monitored sites; no enquiry into the persistence of local populations, even in areas with the longest history of the disease; the delayed publication of the protocol for detecting circulating antibodies to the tumours, and no results from any field surveys for antibodies to the disease.

Another way in which interventionism is inimical to natural solutions is that when interventionists become involved, they naturally think first of interventionist solutions, relegating potential natural solutions to the background. Furthermore, if interventionists are given authority, less well-informed observers will assume their recommendations are considered and correct.

There are many statements to the effect that the Devil is doomed without intervention. Here are some, in chronological order: "It doesn't look like there's any way the devils could survive without active human intervention (K. Belov, in Leung, 2007)."

“The future of the Tasmanian devil hangs in the balance. In a race against time, only science can save this Aussie icon (Australian Government media release, 1 July 2011).”

“It's inevitable that the disease will wipe out the wild Tasmanian devil population (A. Good of Devil Ark, in Egan, 2012a).”

“Without human intervention the only way that DFTD can be eliminated is through the loss of the host species,” and “intervention is necessary to protect the species (Woods, 2012b).”

“Without human intervention the only way that DFTD can be eliminated is through the loss of the host species (B. Ujvari, in Lord, 2012a).”

“… zoos are the only viable solution to saving the Tasmanian Devil through the Insurance Population, a captive breeding program in partnership with the Tasmanian Government (Australian Zoo and Aquarium website: “Australian Zoos and Aquariums Unite to Save Our Species,” 21 August 2013).”

6/05/2016 8:41 AM 326

“In this [PhD] project, I wanted to contribute to what I believe is the only reliable option to save the Tasmanian devil: creating a vaccine or immunotherapy that could deal with the disease … (Brown, 2013).”

“Ultimately the disease will wipe out devils in the wild …. (K. Belov, in University of Sydney News, 26 February 2014).”

“If the Tasmanian devil is to survive in the wild, research into a vaccine remains critical (STDP website, “Short story competition supports the Tasmanian devil,” 29 February 2016).”

Such categorical and seemingly authoritative statements are likely to make a decisive impression on policy makers, funding agencies and private donors.

In addition, the on-going refrain that the Devil’s low genetic diversity is related to the spread of a virtually unprecedented transmissible cancer clearly implies that the Devil is somehow deficient and in need of human intervention. This attitude prevails despite a clear, if belated, demonstration of the Devil’s robust ability to reject skin allografts; its obvious ability to deal with all sorts of pathogens in its carrion diet despite frequent open wounds in its mouth; the extensive scarring in older individuals, belying the many wounds that are part of a Devil’s life, and its rapid wound healing.

Interventionism can also make natural solutions more difficult. In the Devil’s case, the capture of healthy animals for the Insurance Population is unlikely to have been “at random,” considering that there seems to be a proportion of any population that doesn’t enter traps. The culling of diseased animals removes prematurely any latent genetically based resistance to the disease, thereby making it just that much less likely resistance will evolve. The foreshadowed translocation of individuals from one part of the Devil’s range to another part will disrupt locally adapted genetic complexes. Translocation to new locations will set the introduced population onto a novel evolutionary path, making later translocation back into the original distribution area that much more problematic. And directed selection is a leap in the dark as to what kind of animal will result. In these examples, interventionism is just another human perturbation of the natural processes that created the actual Devil for which humans now profess concern.

Indeed, it is difficult to see how interventionism can be justified without concurrent exhaustive and continual monitoring of the potential for natural resistance to the disease, so that, at the very least, any intervention will not interfere with any resistant processes. There is no evidence this has been a guiding precept in the STDP.

In this regard, it is worrying that now that some Devil populations at least appear to have stabilised, the “STDP is developing practical on-ground management measures to ensure that wild populations are maintained and can recover (DPIPWE Annual Report, 2013/14; see also Pemberton, 2015).” In other words, now that at least some Devil populations have stabilised in their own right, the STDP has effectively said, “We’re from the government and we’re here to help you.” At least one of these assistance programs involves something described for the moment at least as “dispersal dampening protocols (Fox et al., 2015).” This new “post-epidemic” management effort will no doubt prolong the life of the Save the Tasmanian Devil

6/05/2016 8:41 AM 327

industry, but its impact on Devil populations as naturally evolving entities is problematic.

Finally, in terms of philosophy, interventionism bases its case largely on utilitarian outcomes. In the Devil’s case, the most strongly argued rationales for intervention all relate to some human benefit, be it aesthetic (Devil as icon), conscience (not a second Thylacine!), economic (tourist attraction; paddock hygienist) or scientific (largest living marsupial carnivore). No one has made the non-utilitarian argument that values Devils as co-participants in the second most remarkable phenomenon in the Universe after the origin and evolution of the Universe itself, the origin and evolution of life. Is this value held anywhere inside the STDP?

How We Know What We Know about the Devil

Publication by media

In reading the literature about the Tasmanian Devil, the one thing that stands out strongly is just how much of the information about the species, especially its disease and its conservation, appears first, sometimes only, in the media, that is, stories in newspapers, TV and radio and, of course, the web. Much of this information never seems to get published in peer-reviewed scientific journals. Consequently, many aspects of the Devil’s story can only be pieced together from what appears in the media.

The hazards of relying on these secondary and tertiary sources is that they may involve, on the part of the protagonist, an inadvertently poor expression of an idea or opinion, and on the part of the recorder, a misunderstanding or even mis-transcription of what has been said.

At first thought, the best way to deal with this “low quality” information might seem to be to just put it all to one side and forget about it. On the other hand, there is so much information and insight to be gained from this literature, that it can not be neglected if the goal is a complete and up to date account. Indeed, the STDP’s policy of providing a constant flow of information to the public about Devil research and conservation as well as its on-going drive for donations invites scrutiny of that information as to its timeliness, accuracy, completeness and delivery on created expectations.

At least one major part of the overall STDP is remote from public scrutiny. The Zoo and Aquarium Association is an official STDP “partner” and is largely responsible for managing the Insurance Population, that is, the captive breeding program. Its official planning and management documents, however, are not in the public domain. The most important ZAA documents would appear to be its foundation captive management plan for the Insurance population (Lees, 2005) and its “Annual Report and Recommendations” (ZAA, 2010; Hibbard et al., 2011 and Hibbard et al., 2012; a 2009 report is mentioned in Lees and Andrew, 2012). The latter is “the principal document that guides the management of the Insurance Population (STDP, 2011b).” The captive management plan is “not a publicly available document” (email from C. Hogg, 21 February 2012) and the annual reports, which are produced each December (STDP Annual Report, 2013/14), are not for release to non-[ZAA] members (emails

6/05/2016 8:41 AM 328

from C. Hogg, 21 February 2013 and R. Frederickson, 2 April 2012, respectively). Interestingly, although the ZAA’s “Annual Report and Recommendations” are not available to the public, they are cited by both the DPIPWE (DPIPWE Annual Report, 2015) and the STDP (STDP, 2011b) in public documents. Indeed, one published paper (Hogg et al., 2015) even directed readers to one of these private publications: “Further details of annual program recommendations can be found in the annual report and recommendations produced by the ZAA (Hogg et al. 2013).”

The original captive management plan would be of interest because it presumably laid the foundation for the Insurance Population, and the annual reports, which the STDP identifies as “… the principal document that guides the management of the Insurance Population … ( STDP, 2011b, 2012a; Parks, 2012)” and which “… reports against the goals of the captive management plan (STDP Annual Report, 2013/14),” would be of interest for insight into how well the Insurance Population is working.

Interestingly, a draft version of the ZAA’s annual report for 2012 (Hibbard et al., 2012) is available publicly. The draft was made available not by the ZAA but on the website of an individual zoo. This draft noted that the two active Tasmanian FREs obtained a total of 65 percent weaning success, 79 and 43.2 percent individually, whereas the one mainland FRE achieved nearly 100 percent (Hibbard et al., 2012). These sorts of differences invite enquiry.

The ZAA is essentially an association representing the best interests of it members, zoos and wildlife parks, all of which are commercial enterprises. At the same time, it has been given the responsibility for making decisions that are, presumably, in the best interest of the Devil. These decisions include the distribution to its members of public donations solicited by the Save the Tasmanian Devil Appeal.

The Tasmanian Government’s DPIPWE also has a policy that allows internal documents (DPIPWE, 2010e) to be cited in its public documents but withheld from public scrutiny (e.g., STDP, 2011b)(email from J. McGee, 16 December 2014).

The New Uncertainty Principle

When scientist do research on the Devil’s biology or use existing data, there is the implicit assumption that the information reflects conditions before the influence of humans. For example, the comparison of brain weight on body weight among dasyuromorphians, is based on the assumption that the weights in all species are unaffected by human influence, such as a steady diet of processed food scraps or prolonged periods in captivity. But the advent of facial tumour disease now means there is uncertainty about any data derived from any animal from an area with the disease. What component of the result reflects the Devil’s healthy state and what component might represent the diseased state? And an unaffected population does not necessarily provide a benchmark, due to the potentially confounding effect of geographic variation, which is itself under threat from impending translocations. If any quandary about the influence of the disease or any other human influence can not be answered by reference to the few preserved remnants of the Devil’s pre-disease state, namely, preserved specimens, tissue samples, photographs, recordings, field and lab notes or publications, then it will probably go unresolved. Of course, this problem goes beyond the Devil. It now affects our understanding of all species.

6/05/2016 8:41 AM 329

Materials, Methods, Conventions and Abbreviations

Materials

This account is based entirely on the information that is in the public domain, that is, scientific articles, media stories and videos. I have no direct experience with Devils other than having spent a few weeks one summer in Tasmania and seeing and hearing Devils as they foraged unperturbed around several public campsites. Like so many other features of the natural environment in Australia, those days are now gone forever, replaced with ever more management of an ever-dwindling number of remnants.

Methods

The information on Devils as whole organisms comes from two sources: the wild and captivity. When information on any particular topic is available from both, I lead with the field data and follow with the captive data. I do this in the belief that it is important to know what the animals are actually capable of in a natural situation before looking at what they are doing in a highly artificial one.

The chapters on the biology of the Devil are based only on disease free conditions, either from disease-free populations or from disease-free captive animals.

Conventions

Decimals are usually rounded to the nearest 0.1 or 0.05. The integer “0.5” is rounded up if preceded by an even integer and down if preceded by an odd integer.

In reporting dispersion statistics, I indicate when the value is the standard deviation (SD) but let the standard error stand-alone.

Individual entries on the Save the Tasmanian Devil Website usually carry a date of publication and a date of modification. For example, an item may have been published first in 2008, modified in 2009 and again in 2010. I cite the date of modification of the document I refer to and assume that all information in the entry is correct up to and including this date.

When citing figures or tables in the text, those in published texts begin with a lower case letter (e.g., fig. 1) whereas those in the current text begin with an upper case letter (e.g., Fig. 1).

In quoted material, I have corrected misspellings and typos.

For articles published first online and then in hard copy, I cite the year of the hard copy, because this is the last and hence definitive version - the “version of record.”

A statistical analysis of data, which is followed by the phrase “data in Bloggs, 2010,” means I have analysed the raw data in the cited reference.

6/05/2016 8:41 AM 330

Abbreviations

Con A. Concanavalin A, an adjuvant or mitogen that induces expression of MHC I proteins on a cell’s surface

dg. Decigram; deci is 10-1

DEH. [Commonwealth] Department of the Environment and Heritage.

DEWHA. [Commonwealth] Department of the Environment, Water, Heritage and the Arts.

DFT. Devil facial tumour.

DFTD. Devil facial tumour disease.

DIP. Devil Island Project.

Dmin. Double minute chromosome.

DPIW. [Tasmanian] Department of Primary Industries and Water.

DPIWE. [Tasmanian] Department of Primary Industries, Water and Environment.

fg. Femtogram; femto is 10-15

fL. Femtoliters.

ka. Kiloannus, one thousand years.

Ma. Megaannum, one million years.

MHC. Major Histocompatibilty Complex

Newsletter. This refers to the “quarterly” Newsletter of the STDP, available online.

pg. picogram; pico is 10-12

PHVA. Reference notation for the Tasmanian Devil Population Habitat Variability and Assessment Final Report. The report of a workshop held in Hobart in 3-6 July 2008.

REM. Rapid eye movement sleep.

rs. Spearman rank order correlation coefficient.

STDAC. Save the Tasmanian Devil Appeal Committee

STTDP. Save the Tasmanian Devil Program.

6/05/2016 8:41 AM 331

STDP. Save the Tasmanian Devil Program. The STDP introduced this shortened abbreviation for STTDP without notice toward the end of 2009 or the beginning of 2010 and then retrospectively applied it to all its earlier documents. The longer abbreviation continued to be used, however, in some Tasmanian government and STDP documents after 2009 (DPIPWE, 2010d; DPIPWE advertisement for four field officers, 11 February 2011; Newsletter, September 2011), even in an online document that was modified in December 2012 (DPIPWE, 2010d).

SWS. Slow wave sleep.

TDRAC. Tasmanian Devil Research Advisory Committee.

UTAS. University of Tasmania ybp. Years before present. y. Years

Acknowledgements

For bibliographical and library assistance, I thank Fran Smith and Fiona Simpson of the Australian Museum.

Paul Feindt of Competence Center for non-textual Materials (CNM), German National Library of Science and Technology (TIB), Department for Research and Development, Welfengarten 1B, 30167 Hannover.

Elizabeth Stephenson of TAFE Mudgee.

Jane McGee, Annie Philips, Catherine Stephens and Toni Venettacci of the Tasmanian Department of Primary Industries, Parks, Water and Environment.

Steve Johnson of the Tasmania Parks and Wildlife Service.

For date of collection data for samples analysed for telomere copy number by Ujvari et al., 2012 (fig. 3), I thank Drs P. Bell and B. Lazenby of DPIPWE, Tasmania.

For data on the number of Devil and cat observations in spotlighting surveys between 1985 and 2013, I thank G. Hocking and H. Williams of DPIPWE, Tasmania.

I also benefited from email correspondence with Alex Baynes on dates for Devil fossils.

6/05/2016 8:41 AM 332

Appendices

Appendix 1

Developmental Events in Relation to Age and Length for Young Devils

Recorded total lengths and the inferred ages for pouch young are: c. 32 mm at five weeks (Fleay, 1935); c. 70 mm at seven weeks (Fleay, 1935), c. 120 mm at 15 wks (Guiler, 1970b):c. 275 mm at 20 weeks (Fleay, 1935), and c.305 at 22 weeks (Fleay, 1935).

Recorded head and body lengths are: 2.8 mm (Pemberton, 1990) and c. 6.0 mm at birth (Hughes, 1982); c. 102 mm at 11 weeks (Fleay, 1935); c. 140 mm at 15 weeks (Fleay, 1935); c. 200 at 20 weeks (Fleay, 1935), and 305 at 22 weeks (Fleay, 1935).

Growth in mean grown-rump length in four new-born Devils is given in Fig. 23.

Growth in Crown-rump Length in Four New-born Devils

140

120

100

20 Crown-rump Length (mm) Length Crown-rump 0 3 9 16 23 30 37 44 51 58 65 72 78 85 93 100 107 115 Age (Days)

Fig. 23. Growth in mean crown-rump length in four new-born Devils. Data from Phillips and Jackson, 2003: table 2.

Recorded head lengths are: c. 6.0 mm at birth (Hughes, 1982).

Growth in mean head length in four new-born Devils is given in Fig. 24.

6/05/2016 8:41 AM 333

Growth in Head Length in Four New-born Devils

120

100

Head Length (mm) Length Head 20

3 16 30 44 58 72 85 100 115 128 142 155 170 184 198 212 220 Age (Days)

Fig. 24. Growth in mean head length in four new-born Devils. Data from Phillips and Jackson, 2003: table 2.

Growth in mean head width length in four new-born Devils is given in Fig. 25.

Growth in Head Width in Four New-born Devils

Head Width (mm) Width Head 10

23 37 51 65 78 93 107 121 135 148 163 177 Age (Days)

Fig. 25. Growth in mean head width in four new-born Devils. Data from Phillips and Jackson, 2003: table 2.

Recorded weights are 0.030 g for young < 24 h old (Hughes and Hall, 1988; Hughes and Rose in Rose, 1989: table 1); 0.18-0.29 g (Guiler, 1970b), c. 0.244 g (Hughes, 1982) for “new born” young; c. 4 g at 15 +/-2 days (Guiler, 1970b); c. 70 g at c. 80 days (Guiler, 1970b), and c. 200 g at c. 105 days (Guiler, 1970b).

Growth in mean weight in four new-born Devils is given in Fig. 26.

6/05/2016 8:41 AM 334

Growth in Weight in Four New-born Devils

1600 1400 1200 1000 800 600 Weight (g) Weight 400 200 0 142 148 155 163 170 177 184 191 198 205 212 218 Age (Days)

Fig. 26. Growth in mean weight in four new-born Devils. Data from Phillips and Jackson, 2003: table 2.

External sexual characteristics. The external sexual characteristics are said to be evident at birth (Phillips and Jackson, 2003).

Pigmentation. The young are pink until c. 58 or 60 days according to one observer (Guiler, 1970b) and seven weeks according to another (Fleay, 1935).

One observer reported pigment first appeared on the ears (Fleay, 1935; see also Macey, 2008; fig.; Evershed, 2009: fig.), and another noted that it develops from front to back (Guiler, 1992).

One report noted the ears start to show pigment at 44 days (Phillips and Jackson, 2003).

At nine weeks (63 days) pigment appears on the tail tip according to one report (Fleay, 1935) but appears on the tail at 58 days according to another (Phillips and Jackson, 2003).

The white body patches are first evident at 93 days (Phillips and Jackson, 2003).

Pigment forms over those areas where the fur will be dark and fails to form where the fur will be white (Guiler, 1970b).

Nose. The nasal swellings are extremely prominent at birth (Hughes and Hall, 1988).

Rhinarium. The rhinarium, the naked area around the nostrils in many mammals, becomes evident at c. 7 days and is heavily pigmented at c. 12 days in one account (Guiler, 1970b) and 30 days in another (Parer and Parer-Cook, 2004).

Pelage. There are different dates for the first appearance of fur. A photo of pouch young said to be about five weeks (35 days) old shows no fur except for possibly on the snout (DPIPWE, 2010d: fig. 2). Otherwise, one observer gives the age at the first

6/05/2016 8:41 AM 335

appearance of fur as c. 49 days (Guiler, 1970b); another gives c. nine weeks (c. 63 days) (Fleay, 1935) and two others 58 days (Phillips and Jackson, 2003).

The fur develops from the front to the back (Roberts, 1915). One observer recorded it as appearing first on the head and fore-quarters and growing progressively backwards (Fleay, 1935). And another observer said fur appears first on the end of the snout and “progressively appears over the rest of the body, reaching the rump by c. 90 days”, although noting that the tail becomes furred before the rump (Guiler, 1970b, 1992).

The head, body and tail but not the rump are “well furnished with black hair” and the rump “is sparsely covered with pale hair” at a head + body length of 150-157 mm (N = 4; Anonymous, 1914).

The body hair is sparse, c. 2 mm in length and most prominent on head and anterior part of back at a total length of c. 160 mm (Green, 1967; plate 4).

It is longest on head and neck, about 10 mm at a total length of 230 mm and a weight of 190 g (Green, 1967).

Vibrissae. No vibrissal primordiae are evident at birth (Hughes and Hall, 1988).

Of the facial vibrissae, the mystacials erupt first, but reports differ on the time of first appearance: 17 +/- 2 days and a body weight of 4.0 g, but with one individual still without them at 27 day, in one account (Guiler, 1970b) and 37 days in another (Phillips and Jackson, 2003).

The infraorbitals, interramals, submentals and supraorbitals erupt between 28 +/- 2 and 37 +/- 2 days, with the submentals usually last, in one account (Guiler, 1970b) and at 44 days in another (Phillips and Jackson, 2003).

The mystacials grown to a length of 22 to mm, the infraorbitals (genals) to 22 mm, the supraorbitals to 18 mm and the interramals to 16 mm at total length of c. 160 mm (Green, 1967). And the mystacials grow to 31 mm, the infraorbitals (genals) to 33 mm, the supraorbitals to 25 mm and the interramal to 18 mm at a total length of c. 230 mm at a weight of 190 g (Green, 1967).

The ulnar carpal vibrissae appear about the same time as the mystacials (Guiler, 1970b).

Eyelashes. The eyelashes appear at 50 +/- 2 days (Guiler, 1970b).

Eyes. At birth, the eye primordia are barely visible; the retinal pigment is absent, and there are no eyelids (Hughes and Hall, 1988). At 16 +/- 2 days the eye slit appears as a lightly pigmented line. The eyelids are separated and the eyes open at 115 days in one report (Phillips and Jackson, 2003). But the eyes are said to be open at about 90 +/-2 days in another report (Guiler, 1970b) and at 15 weeks (105 days) in another (Fleay, 1935: plate 8).

The eyes are open, perhaps recently, at a total length of c. 160 mm (Green, 1967).

6/05/2016 8:41 AM 336

Ears. At birth, the ear primordial are barely visible (Hughes and Hall, 1988). One observer noted the pinnae become evident at c. 15 +/- 2 days and remain closely applied to the head until about 76 +/- 2 days when they become free and erect (Guiler, 1970b). Two other observers, however, noted the pinnae become free at 30 days (Phillips and Jackson, 2003).

The auditory meatus is fully open at 107 days (Phillips and Jackson, 2003).

The ears are open, perhaps only recently, at a total length of c. 160 mm (Green, 1967).

Mouth. The oral shield is “extensive and complex” at birth and there is only slight definition of the mandible. There is no indication of taste buds (Hughes and Hall, 1988).

The mouth edge remains round until about 20 days when the lips appear (Guiler, 1970b); the mouth can be opened at c. 80 days (Guiler, 1970b), and the lips are fully separated at 115 days (Phillips and Jackson, 2003).

The young are able to let go of the teat at c. 100 days (Guiler, 1970b) or c. 105 days (15 weeks; Fleay, 1952) and are “detached” from the teat at 121 days (Phillips and Jackson, 2003).

Dentition. Upper canines erupt at 142 days (Phillips and Jackson, 2003).

Lower incisors erupt at 142 days (Phillips and Jackson, 2003).

Lower canine erupt at 148 days (Phillips and Jackson, 2003).

One pair of upper incisors in both the upper and lower jaws just starting to erupt at a total length of c. 230 mm and a weight of 190 g (Green, 1967).

Vocalisation. The young Devil makes its first vocalisation while still attached to the teat. One study reports this as a “slight squeaking” noise at c. 49 days (seven weeks; Fleay, 1952); another as a squeaking noise at c. 58 days (Guiler, 1970b), and another as a whimper/squeak at 65 days (Phillips and Jackson, 2003).

“Yapping” may occur at 16 weeks (Kelly, 1993; see also Fleay, 1952).

Growling and hissing occur at about 20 weeks (Kelly, 1993).

Limbs. The front limb is very well developed but the rear limb is an undifferentiated paddle at birth (Hughes and Hall, 988: fig. 2.1). The feet can grasp at nine weeks (Fleay, 1935).

Claws. Claws are present on the front limb but absent on the barely discernable rear limb at birth (Hughes and Hall, 1988: fig. 2.1). They are stout and c. 3 mm long at a total length of c. 160 mm (Green, 1967).

Tail. The tail becomes distinct from body at 30 days (Phillips and Jackson, 2003).

6/05/2016 8:41 AM 337

Pouch exodus. A full complement of pouch young, four, can still be carried completely concealed in the pouch at an age of 15 weeks (105 days; Fleay, 1935). But a full complement can not fit in the pouch at 107 days (Phillips and Jackson, 2003).

The young first leave the pouch completely and hence may have left the teat for the first time at just over 15 weeks (Fleay, 1935).

Behaviour. Young can ride on their mother’s back at 121 days (Phillips and Jackson, 2003).

The mother leaves young in nest by themselves at 135 days (Phillips and Jackson, 2003).

Young drag materials into the nest at 140 days (Phillips and Jackson, 2003).

In contrast to the general trend in most young mammals, young Devils have a more pointed snout than adults (Fleay, 1935).

Young chewing solid food at 142 days (Phillips and Jackson, 2003).

Young in posturing and threatening displays at 148 days (Phillips and Jackson, 2003).

Young with fully competent walking gait at 155 days (Phillips and Jackson, 2003).

Young playing and tugging at food at 163 days (Phillips and Jackson, 2003).

Young playing and fighting at 170 days (Phillips and Jackson, 2003).

Young weaned at 232 days (Phillips and Jackson, 2003).

6/05/2016 8:41 AM 338

Appendix 2

Analysis of the Contributions to Devil Articles in the Years 2004-2015

Despite the international interest in the Devil both as the largest surviving marsupial carnivore and as the host to a rare self-perpetuating tumour cell line in nature (D. Obendorf, in Bevilacqua, 2005e; Duncan, 2006; B. Brewster, in Ball, 2009), research into the Devil rests largely in the hands of only a few researchers.

Between 2004 and 2015, inclusive, 315 contributors (authors) made 697 contributions to 127 peer-reviewed articles in which the Devil’s facial tumour disease was given as at least a partial rationale for the research (Fig. 27).

Cumulative Number of Devil Articles in Peer-reviewed Journals in the Years 2004-2015

140

120

100

40 Number of Articles 20

0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year

Fig. 27. The cumulative number of Devil articles published in peer-reviewed scientific journals in the years 2004-2015, inclusive.

Nine of the 315 contributors (2.9 precent) made 191 of the 697 contributions (27 percent). One contributor contributed to 35 percent of all articles (43/127).

Two hundred and sixteen of the 315 contributors (69 percent) made only one contribution (Fig. 28).

6/05/2016 8:41 AM 339

Number of Contributors and the Number of their Contributions to Devil Articles in the Years 2004-2015

200 180 160 140 120 100 80 60 40 Number of Contributors Number 20 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 Number of Contributions

Fig. 28. The number of contributors and contributions to peer-reviewed articles in scientific journals between 2004 and 2015 in which the Devil and its facial tumour disease was the primary focus.

Of the 697 contributions made to articles on Devils, 553 (79 percent) were from Australia and 144 (21 percent) were from other countries. Of the 553 contributions from Australia, 302 (55 percent) were from Tasmania and 251 (45 percent) from the other states and territories (Fig. 29).

Number of Contributions to Devil Articles and their State of Origin in Years 2004-2015

350

300

250

200

150

100 50 Number of Contributions Number 0

UK VIC SA WA NT TAS NSW USA ACT QLD OTHER State of Origin

Fig. 29. The number of contributions by state of origin in peer-reviewed scientific articles between 2004 and 2015 in which the Devil and its facial tumour disease was the primary focus.

6/05/2016 8:41 AM 340

The predominance of contributions from Australia becomes even more apparent when it is realised the increase in the number of non-Australian contributions in 2010, 2011 and 2012 (Fig. 30), was due to the fact that 9, 20 and 48 of such contributions were to single articles in each of those years (Murchison et al., 2010, N = 18; Miller et al., 2011, N= 30, and Murchison et al., 2012, N = 55). These are the three most multi- contributed articles in Devil research so far. All but four of these 103 total contributions were unique.

Number of Contributions and their State of Origin for Devil Articles in Years 2004-2015 70

50 TAS A US 40 UK 30 USA 20 OTHER

Number of Contributions Number 10

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year

Fig. 30. The number of contributions, by state of origin and year, in peer-reviewed scientific articles between 2004 and 2015 in which the Devil and its facial tumour disease was the primary focus. “Australia” means all jurisdictions in Australia other than Tasmania.

Tasmanian Devil Bibliography

Allen Greer Visit: sugargum.wix.com/tasmaniandevil

Abbie, A.A. 1940. The excitable cortex in Perameles, Sarcophilus, Dasyurus, Trichosurus and Wallabia (Macropus). Journal of Comparative Neurology 72(3):469- 487.

Abernathy, K. 2016. Tasmanian Devils Are Cuter and More Clever Than You Think National Geographic (online), 14 January 2016.

Abey, D. 2014. Ageing devils retire in style. Sunday Tasmanian, 2 February 2014.

Adams, J. 2011. Hope for the devil? Science World 68(2): 9-11.

Albion, I. 2006. Devil enrichment paper – Crikey who moved that branch? Proceedings of the 1st Australasian Regional Environmental Enrichment Conference

6/05/2016 8:41 AM 341

1st – 2nd November, 2006, Royal Melbourne Zoological Gardens, Melbourne, Victoria.

Alderton, G.K. 2010. Wildlife cancer: The details are not so devilish. Nature Reviews Cancer 10: 79.

Alford, K., D.L. Obendorf, T.M. Fredeking, E. Haehling and G.L. Stewart. 1998. Comparison of the inflammatory responses of mice infected with American and Australian Trichinella pseudospiralis or Trichinella spiralis. International Journal for Parasitology 28(2):343-348.

Alwaluddin, A.B. and S. McLean. 1985. Conjugation of benzoic acid in marsupials. Australian Journal of Zoology 33(5):693-698.

AMMRF, 2010. A devil of a disease. AMMRF News 10:1. (June 2010).

Anderson, A. 2008. Cold Spring Harbor, 454 Life Sciences join forces to fight Tasmanian Devil disease. GenomeWeb News, 30 January 2008.

Andrewartha, R. and R. Campbell. 2006. Report into the diagnosis of devil facial tumour disease at Trowunna Wildlife Park. Biosecurity and Product Integrity Division. Department of Primary Industry, Water and Environment; 15 pp.

Anonymous. 1914. Natural history note. Victorian Naturalist 31:72.

Anonymous. 1966. The Tasmanian devil. Tasmanian Naturalist 7: 3.

Anonymous. 1969. Cancer rate high in Tasmanian devils. Science News 96:115.

Anonymous, 2007. UTAS helping save our Tassie devil. University of Tasmania Alumni News. Issue 33:4-6 [November 2007].

Anonymous. 2015. Roadkill claims two devils that were part of a ‘re-wilding project’ on the Forestier Peninsula. Hobart Mercury, 25 November 2015.

Anthony, W. 2011. Devils of the alpine. Field monitoring project – 2011; 7 pp. Online. [Document created 19 August 2011; accessed 8 July 2014].

Antonius, O. 1933. Einige Schönbrunner Neuerwerbungen der Jahre 1932/ 1933. Der Zoologische Garten NF 6: 244-247.

6/05/2016 8:41 AM 342

Archer, M. 1976. The dasyurid dentition and its relationships to that of didelphids, thylacinids, borhyaenids (Marsupicarnivora) and peramelids (Paramelina: Marsupialia). Australian Journal of Zoology, Supplement 39; 34 pp.

Archer, M. 1978. Quaternary vertebrate faunas from the Texas Caves of southeastern Queensland. Memoirs of the Queensland Museum 19:61-109.

Archer, M. 1979. The status of Australian dasyurids, thylacinids and myrmecobiids. Pp. 29-43, in M.J. Tyler, ed., The Status of Endangered Australian Wildlife. Royal Zoological Society of South Australia, Adelaide; 210 pp.

Archer, M. 1982. A review of Miocene thylacinids (Thylacinidae, Marsupialia), the phylogenetic position of the Thylacinidae and the problem of apriorisms in character analysis. Pp. 445-476, in M. Archer, ed., Carnivorous Marsupials. Royal Zoological Society, Mosman; 804 pp.

Archer M. and A. Baynes. 1972. Prehistoric mammal faunas from two small caves in the extreme south-west of Western Australia. Journal of the Royal Society of Western Australia 55:80-89. Arup, T. 2013. Second mine cleared with some conditions. The Age, 6 August 2013.

6/05/2016 8:41 AM 343

Ashwell, K.W.S. 2008. Encephalization of Australian and New Guinean marsupials. Brain, Behavior and Evolution 71(3):181-199.

ASMP. 2013. Tasmanian Devil (Sarcophilus harrisii) Annual Report. Australasian Species Management Program, Zoo and Aquarium Association, Sydney. [Not seen, cited in Jakob-Hoff et al., 2015.]

Atkin, M. 2015. Vaccinated Tasmanian devils receive first check-up in fight against fatal facial cancer. ABC News online, 23 October 2015.

Atkinson, H.D. 2001. The Woodpecker Papers - Talks with a Naturalist. Tales of Tasmania in the late 1800's. Mary Hume, Perth; 296 pp.

Attard, M.R.G., U. Chamoli, T.L. Rogers and S. Wroe. 2011. Skull mechanics and implications for feeding behaviour in a large marsupial carnivore guild: the thylacine, Tasmanian devil and spotted quoll. Journal of Zoology 285(4):292- 300.

Attard, M.R.G., W.C.H. Parr, L.A.B. Wilson, M. Archer, S.J. Hand, T.L. Rogers and S. Wroe. 2014. Virtual reconstruction and prey size preference in the Mid Cenozoic thylacinid, Nimbacinus dicksoni (Thylacinidae, Marsupialia). Plos One April 09, 2014DOI: 10.1371/journal.pone.0093088.

AUSVET. 2005. Tasmanian Devil Facial Tumour Disease Response. Technical Workshop. 29-31 August 2005. Final Report to Department of Primary Industries, Water and Environment, Tasmania. AusVet Animal Health Services Pty Ltd; 19 pp.

Azvolinsky, A. 2015. Second contagious cancer found in Tasmanian Devils. The Scientist, 29 December 2015

Bain, A. 2010. Devil’s last stand. Outback 72:59.

Baines, R. 2015. Tasmanian devil monitoring program given $500,000 by US zoo amid federal funding concerns. ABC News online, 20 December 2015

Ball, C. 2009. Focus shifts to ‘free-range’ devils. Brand Tasmania Newsletter, Issue 93, April 2009.

Barnett, C.H 1954. The structure and functions of fibrocartilages within vertebrate joints. Journal of Anatomy 88(3):363-368.

Barnett, C.H. and Napier. 1953. The form and motility of the fibula in metatherian mammals. Journal of Anatomy 87(2):207-213.

Barnes, B. 2007. Spatial behaviour and landscape use in a Tasmanian devil population. Unpublished thesis, University of Tasmania; 66 p. Not seen.

Bartels, H., K. Riegel, E. Kleihauer and E.M. Lang. 1966. Comparative studies of the respiratory functions of mammalian blood – II. Marsupials: great grey kangaroo and Tasmanian devil. Respiratory Physiology 1(2):145-150.

6/05/2016 8:41 AM 344

Bartholomai, A. and L.G. Marshall. 1973. The identity of the supposed dasyurid marsupial, Sarcophilus prior De Vis, 1883, with comments on other reported ‘Pliocene’ occurrences of Sarcophilus. Memoirs of the Queensland Museum 16(3):369-374. Bauer, J. 1961. Protection that doesn’t protect. Australian Outdoors, November 1961, pp. 36-41. As reproduced in Owen and Pemberton, 2011.

Baverstock, P.R., M. Archer, M. Adams and B.J. Richardson. 1982. Genetic relationships among 32 species of Australian dasyurid marsupials. Pp. 641-650, in M. Archer, ed., Carnivorous Marsupials, Royal Zoological Society of New South Wales, Mosman; 804 pp.

Baverstock, P.R., M. Krieg and J. Birrell. 1990. Evolutionary relationships of Australian marsupials assessed by albumin immunology. Australian Journal of Zoology 37:273-287.

Baynes, A. 1987. The original mammal fauna of the Nullarbor and southern peripheral regions: evidence from skeletal remains in superficial cave deposits. Pp. 139-152, in N.L. McKenzie and A.C. Robinson, eds, A Biological Survey of the Nullarbor region South and Western Australia in 1984. Department of Environment and Planning, Adelaide: # pp.

Baynes, A., D. Merrilees and J.K. Porter. 1976. Mammal remains from the upper levels of a late Pleistocene deposit in Devi’ls Lair, Western Australia. Journal of the Royal Society of Western Australia 58: 97–126. Beeby, R. 2010. A ray of hope for Tassie’s devils. Sunday Canberra Times, 11 July 2010, p. 18.

Beddard, F.E. 1911. Contributions to the anatomy and systematic arrangement of the Cestoidea II. On two new genera of cestodes from mammals. Proceedings of the Zoological Society of London 81(4):994-1018.

Beddard, F.E. 1912. Contributions to the anatomy and systematic arrangement of the Cestoidea. V. On a new genus (Dasyurotaenia) from the Tasmanian Devil (Dasyurus ursinus), the type of a new family. Proceedings of the Zoological Society of London 82(3):677-695.

Beddard, F.E. 1915. Contributions to the anatomy and systematic arrangement of Cestoidea. XVI. On certain points in the anatomy of the genus Amabilia and Dasyurotaenia. Proceedings of the Zoological Society of London 85(2):175-191.

Beeby, R. 2010. A ray of hope for Tassie’s devils. Sunday Canberra Times, 18 July 2010, p. 18.

Beeton, N. 2012. Population and disease modelling in the Tasmanian devil. PhD Thesis, University of Tasmania.

Beeton, N.J. and L.K. Forbes. 2013. Dynamical systems analysis of a model describing Tasmanian Devil Facial Tumour Disease. ANZIAM Journal 54:89-107.

6/05/2016 8:41 AM 345

Published online: 18 March 2013; available at CJO2013. doi:10.1017/S1446181113000011.

Beeton, N. and H. McCallum. 2011a. Models predict that culling is not a feasible strategy to prevent extinction of Tasmanian Devils from facial tumour disease. Journal of Applied Ecology 48(6):1315-1323. Published online: 5 October 2011.

Beeton, N. and H. McCallum. 2011b. Saving the Tasmanian devil: if not by selective culling, then how? The Conversation (online), 19 October 2011.

Beeton, R.J.S. Undated. Advice to the Minister for the Environment, Heritage and the Arts from the Threatened Species Scientific Committee (the Committee) on Amendment to the list of Threatened Species under the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act). Publisher unstated. [Available online.]

Bell, L. 2001. The hormonal control of birth behaviour in the Tasmanian devil, Sarcophilus harrisii. B.Sc. Honours Thesis, University of Tasmania.

Bell, C.J., R.V. Baudinette and S.C. Nicol. 1983. Cutaneous water loss at rest and exercise in two species of marsupials. Australian Journal of Zoology 31(1):93–99.

Bell P. 2014. Management prescriptions for the Tasmanian devil (Sarcophilus harrisii): Background information. Unpublished report. [Not seen.]

Belov, K. 2011. The role of the Major Histocompatibility Complex in the spread of contagious cancers. Mammalian Genome 22(1/2):83-90.

Belov, K. 2012. Contagious cancer: lessons from the devil and the dog. Bioessays 34(4):285-292.

Belov, K. and C. Hogg. 2012. Captive breeding last hope of Tasmanian Devil. ABC Science Online Opinion, 27 August 2012.

Bender, J., A. Alsop, A.-M. Pearse and J.A.M. Graves. 2007. Devil facial tumour disease: cytogenetic clues to transmission and development. P. 12, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Bender, H.S., J.A.M. Graves and J.E. Deakin. 2014. Pathogenesis and molecular biology of a transmissible tumor in the Tasmanian Devil. Annual Review of Animal Biosciences 2: 165-187.

Bender, J.S., E.P. Murchison, H.A. Pickett, M.A. Strong, C.W. Greider, T. Ezaz,. A.- M. Pearse, P.R. Reddel, J.A.M. Graves. 2009. Telomere length differences between homologous chromosome in dasyurid marsupials – a parental effect. Genetics Society of Australasia 2009. Abstract.

Bender, H.S. 2010. Devil Facial Tumour Disease (DFTD): Using genetics and genomics to investigate infectious disease in an endangered marsupial. Pp. 499-515 in

6/05/2016 8:41 AM 346

J.E. Deakin, P.D. Waters, J.A.M. Graves, eds, Marsupial Genetics and Genomics. Springer, Dordrecht; 519 pp. [Marsupial Genetics and Genomics 9: 499-515.]

Bender, H.S., E.P. Murchison, H.A. Pickett, J.E. Deakin, M.A. Strong, C. Conlan, D. A. McMillan, A.A. Neumann, C.W. Greider, G.J. Hannon, R.R. Reddel, J.A.M. Graves. 2012. Extreme telomere length dimorphism in the Tasmanian Devil and related marsupials suggests parental control of telomere length. PLoS ONE 7(9): e46195. doi:10.1371/journal.pone.0046195.

Benga, G., D. Porutiu, I. Ghiran, P.W. Kuchel, C.H. Gallagher and G.C. Cox. 1992. Scanning electron microscopy of red blood cells from eleven species of marsupial. Comparative Haematology International 2(4):227-230.

Benga, G., B.E. Chapman, C.H. Gallagher, N.S. Agar and P.W. Kuchel. 1993. NMR studies of diffusional water permeability of erythrocytes from eight species of marsupial. Comparative Biochemistry and Physiology 106A(3):515-518.

Beniuk, D. 2012. Aggression earns Tas devils tumour disease. AAP, 4 September 2012.

Beniuk, D. 2013. Tassie devils revel on island home. Sun Herald, 19 April 2013.

Bensley, B.A. 1903. On the evolution of the Australian Marsupialia; with remarks on the relationships of the Marsupialia. Transactions of the Linnean Society (2)9(3):83- 217.

Beveridge, I. 1984. Dasyurotaenia robusta Beddard, 1912, and D. dasyuri sp. nov., from carnivorous Australian marsupials. Transactions of the Royal Society of South Australia 108(4):185-195.

Beveridge, I. and M.K. Jones. 2002. Diversity and biogeographical relationships of the Australian cestode fauna. International Journal for Parasitology 32(3): 343-51.

Beveridge, I., M.D. Rickard, G.G. Gregory and B.L. Munday. 1975. Studies on Anoplotaenia dasyuri Beddard, 1911 (Cestoda: Taeniidae), a parasite of the Tasmanian devil: observations on the egg and metacestode. International Journal for Parasitology 5(3):257-267.

Beveridge, I. and D.M. Spratt. 2003. Parasites of carnivorous marsupials. Pp 383-396, in M. Jones, C.R. Dickman and M. Archer, eds, Predators with Pouches. The Biology of Carnivorous Marsupials. CSIRO Publishing, Collingwood, Victoria; 486 pp.

Beveridge, I. and D.M. Spratt. 2014. Biodiversity and parasites of wildlife: Helminths of Australasian marsupials. Trends in Parasitology. DOI: http://dx.doi.org/10.1016/j.pt.2014.10.007. Available online, 26 November 2014. Publication stage: In Press Corrected Proof

Bevilacqua, S. 2001. Devils may not only be from Tassie. Hobart Mercury, 19 August 2001.

6/05/2016 8:41 AM 347

Bevilacqua, S. 2003a. Credible sightings keep hopes alive. Hobart Mercury, 6 July 2003.

Bevilacqua, S. 2003b. Sensitive side to tough little devils. Sunday Tasmanian, 14 September 2003.

Bevilacqua, S. 2003c. Science issues to get a private hearing. Hobart Mercury, 14 October 2003.

Bevilacqua, S. 2004a. Shocking pain and suffering. Sunday Tasmanian, 15 February 2004.

Bevilacqua, S. 2004b. Devilish frustrations. Sunday Tasmanian, 14 November 2004.

Bevilacqua, S. 2005a. Devil disease heartbreak. Sunday Tasmanian, 10 July 2005.

Bevilacqua, S. 2005b. Devil deaths despair. Sunday Tasmanian, 14 August 2005.

Bevilacqua, S. 2005c. Research too secret, critics say. Hobart Mercury, 21 August 2005.

Bevilacqua, S. 2005d. Confusion over what was reported and when. Sunday Tasmanian, 28 August 2005.

Bevilacqua, S. 2005e. Bush spraying blamed. Sunday Tasmanian, 20 October 2005.

Bevilacqua, S. 2005f. Change in fortune for the hapless Lucky. Sunday Tasmanian, 6 November 2005.

Bevilacqua, S. 2006a. Top docs check on virus link. Sunday Tasmanian, 23 July 2006.

Bevilacqua, S. 2006b. Silence of the labs. Sunday Tasmanian, 12 November 2006.

Bevilacqua, S. 2006c. What the devil? Top research offer snubbed. Sunday Tasmanian, 12 November 2006.

Bevilacqua, S. 2006d. New devil knockback. Now UK geneticist reports Tassie snub. Hobart Mercury, 14 November 2006.

Bevilacqua, S. 2006e. New push for devil aid uni plea to involve expat US biologist. Hobart Mercury, 15 November 2006.

Bevilacqua, S. 2006f. State urged to embrace foreign help. Sunday Tasmanian, 19 November 2006.

Bevilacqua, S. 2006g. Focus turns to devils’ genes. Sunday Tasmanian, 10 December 2006.

6/05/2016 8:41 AM 348

Bevilacqua, S. 2007. Slow trip to top US lab. Sunday Tasmanian, 25 March 2007.

Bevilacqua, S. 2008. Blow to immune Devil hope. Hobart Mercury, 14 December 2008.

Bevilacqua, S. and D. Wood. 2003. It’s time for urgent action and money. Hobart Mercury, 2 November 2003.

Bevin, E. 2007. Devil of a burden. The Daily Telegraph, 18 May 2007.

Bininda-Emonds, O.R.P., M. Cardillo, K.E. Jones, R.D. E. MacPhee, R.M.D. Beck, R. Grenyer, S.A. Price, R.A. Vos, J.L. Gittleman and A. Purvis. 2007. The delayed rise of present-day mammals. Nature 446:507-512.

Bird, I. 2012a. Devil order a `death row' sentence. The Examiner Newspaper, 21 February 2012.

Bird, I. 2012b. Orphaned devil to remain with zoo. The Examiner, 12 March 2012.

Bird, I. 2013. Mine impact minimal on devils. The Examiner, 3 August 2013.

Blackwood, F. 2013. Endangered devils breeding on Maria Island. ABC News, 19 April 2013.

Blake, C. 1992. Tasmanian devils: is there still a mainland population? B.Sc. Honours Thesis, La Trobe University. Abstract only seen.

Blanchard, W. 1991. Idiopathic meningitis/meningeoencephalitis in a Tasmanian Devil (Sarcophilus harrisii). Wildlife Disease Association, Australian Section Scientific Meeting Proceedings, 1991. Not seen, as cited in Obendorf, 1993.

Bland, D.K. and R.A.B. Holland. 1977. Oxygen affinity and 2,3-diphosphoglycerate in blood of Australian marsupials of differing body size. Respiration Physiology 31(3): 279-290. . Bloom H., W. Taylor, W.R. Bloom and G.M. Ayling. 1979. Fish, wildlife and Estuaries. Organochlorine pesticide residues in animals of Tasmania, Australia -- 1975-77. Pesticides Monitoring Journal 13(3):99-108.

Bock, J. 2016. New to the St. Louis Zoo: Tasmanian devils. St. Louis Post-Dispatch, 21 April 2016.

Bode, M., C. Hawkins, T. Rout and B. Wintle. 2009. Efficiently locating conservation boundaries: searching for the Tasmanian devil facial tumour disease front. Biological Conservation 142:1333-1339.

Bode, M. and B. Wintle. 2009. How to build an efficient conservation fence: Contributed paper. Conservation Biology 24(1):182-188.

6/05/2016 8:41 AM 349

Boitard, P. 1841. Le Jardin des Plantes descriptions et moeurs des mammifères de la Mènagerie et du Muséum d’Histoire Naturelle, précédé d’une introduction historique, descriptive et pittoresque par M.J. Janin. J.J. Dubochet et Cie; 472 pp.

Boland, C. 2011. Devilishly clever team. Hobart Mercury, 27 September 2011.

Bolger, R. 2011. Launceston will become home to a new free-range sanctuary for the endangered Tasmanian devil. The Examiner, 21 April 2011.

Bolger, R. 2012. New projects help fight to save devil. The Examiner, 12 September 2012.

Bolger, R. 2013. Devils return to populate Maria Island. The Examiner, 6 November 2013.

Bolger, R. 2015a. Tasmanian devil researchers turn to genetics to help captive animals survive in the wild. ABC News online, 8 October 2015.

Bolger, R. 2015b. Save the Tasmanian Devils program to focus on vaccinating wild devils after road kill losses. ABC News online, 9 October 2015.

Boness, L. 2012. I will survive. Science Illustrated online, 16 April 2012 (but originally published in the March/April 2011 issue). Booth, R. 1994. Medicine and husbandry: dasyurids, possums and bats. Pp. 423-442 in D.I. Bryden, ed., Wildlife. The TG Hungerford Refresher Courses for Veterinarians. Proceedings 233. The Post Graduate Committee in Veterinary Science, University of Sydney; 675 pp.

Borrell, B. 2009. Hopes of a tumour test for Tasmanian Devils. Nature News (online), 31 December 2009.

Bostanci, A. 2005. A Devil of a Disease. Science 307:1035.

Bourne, G. 1936. The phylogeny of the adrenal gland. American Naturalist 70(727):159-178.

Bowdler, S. 1984. Hunters Hill, Hunter Island. Terra Australis 8; 148 pp.

Bradbury, R. 2015. A disease from the devil? Could Baylisascaris tasmaniensis cause disease in humans? Australian Journal of Medical Science 36(1):15-17. Bradshaw, C. and B. Brook. 2005. Disease and the devil: density-dependent epidemiological processes explain historical population fluctuations in the Tasmanian devil. Ecography 28:181-190.

Brandenburg, H. 2010. Analysis of movement of Tasmanian Devils (Sarcophilus harrisii) between traps to understand the spread of DFTD. Unpublished doctoral thesis, University of Utrecht; # pp. Only abstract seen. Brennan, W. 2005. Tasmanian Devil endangered! Tasmania 40º South No. 39:43-46.

6/05/2016 8:41 AM 350

Brennan, W. 2006. Devil facial tumour disease. Tasmanian Conservationist No. 309: 8-9.

Brennan, W. 2007. The cascade effect. Tasmanian Conservationist No. 311:11

Brereton, R.N., Taylor, R.J. and Rhodes, M. 1996. Vertebrates of the Domain, an urban bushland remnant in Hobart. Tasmanian Naturalist 118: 31-40.

Brice, S. (ed). 2006. Husbandry guidelines for Tasmanian devil, Sarcophilus harrisii. Australasian Regional Association of Zoological Parks and Aquaria, Sydney; 36 pp.

Brix, L., M.J. Hansen, A. Kelly, M.F. Bertelsen and A.M. Bojesen. 2015. Occurrence of Pasteurellaceae bacteria in the oral cavity of the Tasmanian Devil (Sarcophilus harrisii). Journal of Zoo and Wildlife Medicine 46(2):241-245.

Broadstock, A. 2015. Monarto Zoo Tassie Devils bound for Toledo, Ohio The Advertiser, 15 October 2015.

Brown, D. 2009. Devils reveal dark secrets. Hobart Mercury, 8 October 2009.

Brown, G.K. 2008. Genetic diversity, cytotoxic T lymphocyte responses and anti- cancer therapy in Tasmanian devils. Honours Thesis, Biotech, University of Tasmania, Hobart. [Not seen, as reported in Brown, 2013]

Brown, G.K. 2013. Manipulating the immune response of Tasmanian Devils to target Devil Facial Tumour Disease. PhD Thesis, Menzies Research Institute, University of Tasmania; 231 pp.

Brown, G., A. Kreiss and G. Woods. 2010. Cytotoxic ability of devil lymphocytes and implications for Devil Facial Tumour Disease. 2010 ANCCART Conference, 20-22 July 2010, Hobart, Tasmania, p. 33.

Brown, G.K., A. Kreiss, A.B. Lyons and G.M. Woods. 2011. Natural Killer Cell mediated cytotoxic responses in the Tasmanian Devil. PLoS ONE 6(9): e24475. doi:10.1371/journal.pone.0024475.

Brown, G.K., A. Kreiss, A.B. Lyons and G.M. Woods. 2012. Activating the Tasmanian devil's immune response to target DFTD. Abstract, 6th Australian Health and Medical Research Congress 2012, 27 November 2012.

Brown, G.K., C. Tovar, A.A Cooray, A. Kreiss, J. Darby, J.M. Murphy, L.M. Corcoran, S.S. Bettiol, A.B. Lyons and G.M. Woods. 2016. Mitogen activated Tasmanian devil blood mononuclear cells kill devil facial tumour disease cells. Immunology and Cell Biology. Published online, 19 April 2016; doi: 10.1038/icb.2016.38

6/05/2016 8:41 AM 351

Brown, O.J.F. 2006. Tasmanian devil (Sarcophilus harrisii) extinction on the Australian mainland in the mid-Holocene: multicausality and ENSO intensification. Alcheringa: An Australasian Journal of Palaeontology 30(1):49-57.

Brüniche-Olsen, A. 2011. NCCARF travel grant report. 21 October 2011. Accessed online on 29 August 2012.

Brüniche-Olsen, A. and J. Austin. 2014. Tasmanian devils survived two big falls in numbers but now need help. The Conversation, 5 November 2014.

Brüniche-Olsen, A., J.J. Austin, M.E. Jones, B.R. Holland and C.P. Burridge. 2016. Detecting selection on temporal and spatial scales: a genomic time-series assessment of selective responses to Devil Facial Tumor Disease. PLoS ONE 11(3): e0147875. doi:10.1371/journal.pone.0147875.

Brüniche-Olsen, A., C.P. Burridge, J.J. Austin and M.E. Jones. 2011. Impacts of devil facial tumour disease (DFTD) on genetic diversity, effective population size and migration rates among Tasmanian devil populations. Postgraduate poster night University of Tasmania, Australia.

Brüniche-Olsen, A., C.P. Burridge, J.J. Austin and M.E. Jones. 2013. Disease induced changes in gene flow patterns among Tasmanian devil populations. Biological Conservation 165:69-78.

Brüniche-Olsen, A., M.E. Jones, J.J. Austin, C.P. Burridge and B.R. Holland. 2014. Extensive population decline in the Tasmanian devil predates European settlement and devil facial tumour disease. Biology Letters 10(11): 20140619; 5 pp. Published online, 5 November 2014.

Bryant, S.L. 2014. Carnivorous Mammal Monitoring 2014: Five Rivers Reserve, Bronte. Tasmanian Land Conservancy, Sandy Bay, Tasmania.

Bryant, S. 2015. Tasmanian Devil monitoring and where it leads you! 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p.10.

Bryant, S. and J. Keble-Williams. 2014. The value of mammal monitoring on the Five Rivers Reserve, Bronte. Tasmanian Naturalist 136:168-173.

6/05/2016 8:41 AM 352

Buchman, O.L.K. and E.R. Guiler. 1977. Behaviour and ecology of the Tasmanian Devil, Sarcophilus harrisii. Pp. 155-168, in B. Stonehouse and D. Gilmore, eds, The Biology of Marsupials. MacMillan, London; 486 pp.

Bugard, M. 2014. Doing the work for the devil. Sydney Morning Herald, 23 January 2014.

Burgess, J. 2010. Devil efforts ‘succeeding.’ The Examiner Newspaper. Online, 28 January 2010.

Burgess, G. 2013, Devil program hit by $2m funding cut. The Examiner, 19 December 2013.

Burgess, G. 2014a. Blame game begins over devil funding. The Examiner, 15 January 2014 .

Burgess, G. 2014b. Devil to get disease funding. Examiner, 23 January 2014.

Butler, W.H. 1969. Remains of Sarcophilus the “Tasmanian” Devil (Marsupialia, Dasyuridae) from coastal dunes south of Scott, River, Western Australia. Western Australian Naturalist 11(4):87-88.

Byrnes, M. 2007. Battle to save Tasmanian devil from extinction. Reuters, 15 March 2007.

Cadell, C. 2013. Tassie devil vaccine may be on the horizon. Australian Geographic, 14 March 2013.

Calaby, J. and D.J. Lewis. 1977. The Tasmanian Devil in Arnhem Land rock art. Mankind 11(2):150-151.

Calaby, J.H. and C. White. 1967. The Tasmanian Devil (Sarcophilus harrisii) in northern Australia in Recent times. Australian Journal of Science 29(12):473-475.

Cameron, T.W.M. 1931. On a species of Trichostrongyle from the Tasmanian Devil. Journal of Helminthology 9(3):153-156.

Canfield, P.J. and A.A. Cunningham. 1993. Disease and mortality in Australasian marsupials held at London Zoo, 1872-1972. Journal of Zoo Wildlife Medicine 24(2):158-167.

Canfield, P.J., W.J. Hartley and G.L. Reddacliff. 1990. Spontaneous proliferations in Australian marsupials – a survey and review. 2. Dasyurids and bandicoots. Journal of Comparative Pathology 103(2):147-158. Capels, J. 2003. State's $1.8m move to ensure Tasmanian devils won't die out. The Examiner, 13 November 2003.

6/05/2016 8:41 AM 353

Cart, J. 2011. Drumming up support for Tasmania’s unsexy beast. Los Angeles Times, 27 March 2011. Accessed online.

Carter, H.J. 1933. Australia’s dwindling wildlife. Australian Quarterly 5(17):106-113.

Carter, P. 2015a. Devils put vaccine to test. Hobart Mercury, 22 July 2015.

Carter, P. 2015b. Tasmanian devil release marks turning point for species survival. Hobart Mercury, 26 September 2015.

Carter, P. 2015c. Two more Tasmanian devils road-killed in vaccine trial takes death toll to four. Sunday Tasmanian, 3 October 2015.

Caruana, P. 2015. Study finds devil tumour vaccine possible. The Examiner, 25 February 2015.

CBSG/DPIPWE/ARAZPA. 2009. The Save the Tasmanian Devil Program: Strategic Framework for an Insurance Meta-population. DPIPWE, Tasmania, Australia. [DPIW, 2007b is the forerunner of this work]

Cerabona, R. 1997. Sanctuary in the hunt for a devil of a name. Canberra Times, 15 April 1997. Chatin, J. 1876. Sur les glandes anales du Sarcophile ursien (Sarcophilus ursinus, Fr. Cuvier). Bulletin de la Société Philomathique de Paris (6)13:54-55.

Cheng, Y. 2012. Characterization of the Tasmanian devil histocompatibility complex. PhD Thesis, University of Sydney.

Cheng, Y. and K. Belov. 2012a. Isolation and characterisation of 11 MHC-linked microsatellite loci in the Tasmanian devil (Sarcophilus harrisii). Conservation Genetics Resources 4(2):463-465.

Cheng, Y. and K. Belov. 2012b. Evolution of the major histocompatibility complex: insights from characterisation of marsupial genes. In: eLS, John Wiley & Sons, Ltd; Chichester. DOI:10.1002/9780470015902.a0023982. Published online: 17 December 2012.

Cheng, Y. and K. Belov. 2014. Characterisation of non-classical MHC class I genes in the Tasmanian devil (Sarcophilus harrisii). Immunogenetics 66(12):727-735.

Cheng, Y., S. Fox, D. Pemberton, C. Hogg, A.T. Papenfuss and K. Belov. 2015. The Tasmanian devil microbiome - implications for conservation and management. Microbiome 3: 75.

Cheng, Y., H.V. Siddle and K. Belov. 2010. The Major Histocompatibility Complex. Pp. 339-356, in J.E. Deakin, P.D. Waters, J.A.M. Graves, eds, Marsupial Genetics and Genomics. Springer, Dordrecht; 519 pp. [Marsupial Genetics and Genomics 9: 339- 356.]

6/05/2016 8:41 AM 354

Cheng, Y., A. Stuart, K. Morris, R. Taylor, H.V. Siddle, J.E. Deakin, M. Jones, C.T. Amimiya and K. Belov. 2012a. Antigen-presenting genes and genomic copy number variations in the Tasmanian devil MHC. BMC Genomics 13:87-#.

Cheng, Y., C. Sanderson, M. Jones and K. Belov. 2012b. Low MHC class II diversity in the Tasmanian devil (Sarcophilus harrisii). Immunogenetics 67:525-533.

Church, J.S., D.J. Evans and A.L. Woodward. 2009. The characterization of Tasmanian Devil Sarcophilus harrisii pelage fibres and their associated lipids. Acta Zoologica 90(4):380-389.

Church, J.S., A.L. Woodhead and J.N. Varghese. 2007. Analysis of keratin fibres for the pre-clinical detection of DFTD. P. 27, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Claridge, A. 2013. Devils' island of hope. Hobart Mercury, 21 April 2013.

Clarke, C. 2014. Devil population improving. The Examiner, 23 November 2014.

Clarke, C. 2015. Immunisation on devil facial tumours trialled. The Examiner, 21 January 2015.

Clark, N. 2003. Call from the top to save our little devil. Hobart Mercury, 24 July 2003.

Clark, N. 2008. Devils’ danger status upgraded. Hobart Mercury, 22 May 2008.

Clark, N. 2009. Breeding program for devils. Hobart Mercury, 4 February 2009.

Clark, N. 2016. $280m Van Diemen’s Land Company sale decision. Hobart Mercury, 24 February 2016.

Clifton-Evans, L. 2012. Devils at Pearcedale Sanctuary in mood for love. Cranbourne Leader, 8 March 2012.

Cock, A. 2003. US in devil’s thrall. Herald Sun (Melbourne), 1 October 2003.

Coghlan, A. 2012. ‘Immortal’ Tasmanian devil brings vaccine hope. New Scientist, 17 February 2012.

Coldham, T. 2004. The detection and characterisation of Helicobacter species in Australian marsupials. PhD Thesis, University of New South Wales. Not seen; as reported in Ladds, 2009).

Coleman, J.D. and D.M. Forsyth. 2006. Review of research into alternatives to the use of 1080 for management of browsing damage by mammals in Tasmania. Tasmania. Dept. of Primary Industries and Water.

6/05/2016 8:41 AM 355

Collins, L.R. 1973. Monotremes and Marsupials. A Reference for Zoological Institutions. Smithsonian Institute Press, Washington, D.C.; # pp.

Colliver, F.S. 1938. Cave hunting in Victoria. Victorian Naturalist 54(10):152- 155. Comte, S., R. Hamede, S. Carver and M. Jones. 2015. There and back again … a devil’s tale. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 10.

Connellan, I. 2005. The devils and Dr Jones. Australian Geographic 80:17.

Conservation Breeding Specialist Group. 2008. The Tasmanian Devil PHVA. Final Report. IUCN/SSC Conservation Breeding Specialist Group, Apple Valley, Minnesota; 144 pp.

Cook, D.L. 1960. Some mammal remains found in caves near Margaret River. Western Australian Naturalist 7(4)107-108.

Cook, D.L. 1963a. The fossil vertebrate fauna of Strong’s Cave, Boranup, Western Australia. Western Australian Naturalist 8(7):153-162.

Cook, D.L. 1963b. Thylacincus and Sarcophilus from the Nullarbor Plain. Western Australian Naturalist 9(2):47-48. Cooper, C.E. and P.C. Withers. 2011. Comparative physiology of Australian quolls (Dasyurus; Marsupialia). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology 180(6): 857-868.

6/05/2016 8:41 AM 356

Corbett, L. 1995. The Dingo in Australian and Asia. University of New South Wales Press, Sydney; # pp.

Cordingley, G. 2007a. Conservation plan a devil of a problem. Canberra Times, 31 March 2007.

Cordingley, G. 2007b. Tassie devil cancer spread fear over. Australian Associated Press, 17 October 2007.

Correy, M. 2009. Australian zoo plays role of guardian angel to Tasmanian devil. Vancouver Sun, 30 January 2009.

Cosgrove, 2014. Tasmanian archaeology and reflections on modern human behaviour. Pp. 175-188, in R. Donnell and M. Porr, Southern Asia, Australia and the Search for Human Origins. Cambridge University Press, #; 342 pp.

Coupland, C., and W. Anthony. 2007. Devils of the Alpine Project Field Monitoring Program. Tasmanian Naturalist 129:65-81.

Courtney, R. 2010. Devils get sanctuary in NSW. Australian Geographic, 15 July 2010.

Cox, T.E., P.J. Murray, G.P. Hall and X. Li. 2010. Pest responses to odors from predators fed a diet of target species conspecifics and heterospecifics. Journal of Wildlife Management 74(8):1737-1744. Crabb, P.L. 1982. Pleistocene dasyurids (Marsupialia) from south-western New South Wales. Pp. 511-516, in M. Archer, ed., Carnivorous Marsupials. Royal Zoological Society, Mosman, New South Wales; 804 pp.

Cramb, J., S. Hocknull and G.E. Webb. 2009. High diversity Pleistocene rainforest dasyurid assemblages with implications for the radiation of the Dasyuridae. Austral Ecology 34(6):663-669.

Crane, C. 2008. Poor devils. Science World 64(11): 19.

CRES. 2006. Sarcophilus harrisii (Tasmanian Devil). In: S.J. O’Brien, J.C. Menninger and W.G. Nash, eds. Atlas of Mammalian Chromosomes. New York: Wiley; 30 p.

Cribb, T.H. and J.C. Pearson. 1993. Neodiplostomum spratti n. sp. (Digenea: Diplostomidae) from Antechinus spp. (Marsupialia: Dasyuridae) in Australia, with notes on other diplostomids from Australian mammals. Systematic Parasitology 25(1): 25-35.

Cubby, B. 2008. Devil of a challenge to save species. Sydney Morning Herald, 14 February 2008.

Cui, J. 2015. Variation of the Toll-like receptors in two marsupial species, Tasmanian devils, Sarcophilus harrisii, and Koalas, Phascolarctos cinereus. Master of Science in Veterinary Science, University of Sydney; # pp.

6/05/2016 8:41 AM 357

Cui, J., Y. Cheng and K. Belov. 2014. Characterisation of Toll-like receptors in two bottlenecked species, Tasmanian devil and Koala. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 33.

Cui, J., Y. Cheng and K. Belov. 2015. Diversity in the Toll-like receptor genes of the Tasmanian Devil (Sarcophilus harrisii). Immunogenetics 67:195-201. Published online, 7 January 2015.

Cui, X., Y. Wang, B. Huaa, W. Miller, Y. Zhao, H. Cui, X. Kong. 2016. Sex determination by SRY PCR and sequencing of Tasmanian devil facial tumour cell lines reveals non-allograft transmission. Biochemical and Biophysical Research Communications #(#):#-#. Published online 13 April 2016.

Cui, X., Y. Wang, B. Hua, Y. Zhao, H. Cui and X. Kong. Unpublished. Devil facial tumour disease associated retrovirus and herpesviruses originated from avian leukosis virus and Marek’s disease virus. Unpublished. Citation from: Tasmanian devil retrovirus DRV 2010 polymerase (pol) and envelope protein (env) genes, partial cds. GenBank, 19 May 2013.

Cuttell, L., S.W. Corley C.P. Gray, P.B. Vanderlinde, L.A. Jackson and R.L. Traub. 2012. Real-time PCR as a surveillance tool for the detection of Trichinella infection in muscle samples from wildlife. Veterinary Parasitology 188(3-4):285-293.

Czarny, N.A. and J.C. Rodger. 2010. Vitrification as a method for genome resource banking oocytes from the endangered Tasmanian devil (Sarcophilus harrisii). Cryobiology 60(3):322-325.

Dadson, M. 2012. Facial tumour disease discovered in zoo devil. The Examiner, 24 April 2012.

Darby, A. 2003a. Search for what in the Tasmanian devil is killing them. Sydney Morning Herald, 13 October 2003.

Darby, A. 2003b. Scourge of the night terrorists. Sydney Morning Herald, 13 November 2003.

Darby, A. 2004. Feisty or sleepy. Devils are defenceless to tumours. Sydney Morning Herald, 6 March 2004.

Darby, A. 2005. Tasmania moves to save devil from cruel disease. The Age, 12 May 2005.

Darby, A. 2006. Last-ditch effort to save Tasmanian devils for the wild. The Age, 1 December 2006.

Darby, A. 2007a. Escape from devil’s island: the plan to save a species. The Age, 22 September 2007. Also as: Zoos build gene pool to save devil. Sydney Morning Herald, 22 September 2007.

6/05/2016 8:41 AM 358

Darby, A. 2007b. Plucky little Devil called Cedric offers hope for his species’ future. Sydney Morning Herald, 28 November 2007.

Darby, A. 2008. Tassie devil at risk as experiment fails. The Age, 18 December 2008.

Darby, A. 2010a. Cedric's fight raises hope beyond death. The Age, 2 September 2010.

Darby, A. 2010b. Devil genome map to aid cancer fight. The Age, 17 September 2010.

Darby, A.2011. Poor devils. Sydney Morning Herald, 9 November 2011, online.

Darby, A. 2012a. Island may be devil’s last chance. The Age, 10 May 2012.

Darby, A. 2012b. Parks become a haven for devils. Sydney Morning Herald, 15 November 2012.

Darby, A. 2013. Mainland devils get a second chance at ancestral home. Sydney Morning Herald, 6 November 2013.

Darby, A. 2015a. ''Huge'' step forward with devil cancer vaccine. Sydney Morning Herald, 22 July 2015.

Darby, A. 2015b. Saviours shattered after immunised Tasmanian devils die on Narawntapu roads. Sydney Morning Herald, 30 September 2015.

Darby, A. 2015c. Tasmanian devils escape disease, to be killed by cars. Sydney Morning Herald, 27 November 2015.

Darcé, K. 2010. Local scientists map Tasmanian devil's genome to save it. San Diego Union-Tribune, 19 September 2010. [Also as, “Using genetics to save Tasmanian devil,” SignOnSanDiego, 19 September 2010.]

Davis, G., R. Hamede, E. Cameron and M. Jones. 2015. Relating behaviour in response to handling to intra-specific behaviour in wild devils. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 6.

Dawson, L. 1982. Taxonomic status of fossil devils (Sarcophilus, Dasyuridae, Marsupialia) from later Quaternary eastern Australian localities. Pp. 517-525, in M. Archer, ed., Carnivorous Marsupials. Royal Zoological Society, Mosman, New South Wales; 804 pp.

Dawtrey, A, 2013. Maria Island receives a new batch of disease-free Tasmanian devils to boost the healthy population. Hobart Mercury, 7 November 2013.

Day, L. 2014. Hopes healthy Tasmanian devils will soon be returned to the wild after being on brink of extinction. ABC News, 23 November 2014.

6/05/2016 8:41 AM 359 de Grooyer, K. 2015. Scientists ready to test Tasmanian devil facial tumour vaccine in the wild. ABC News online, 22 July 2015. de Salis, S. 2015. Tasmanian Devils returned to the wild at Narawntapu. DPIPWE, Natural and Cultural Heritage, News, Save the Tasmanian Devil Program Share, 1 October 2015 (online). de Salis, S. and S. Graham. 2012. First translocation of devils on Maria Island. Tasmanian Regions, Summer, 2012: 2-4.

Deakin, J.E. 2010. Marsupial breeding. Pp. 1-3, in J.E. Deakin, P.D. Waters, J.A.M. Graves, eds, Marsupial Genetics and Genomics. Springer, Dordrecht; 519 pp. [Marsupial Genetics and Genomics 9:1-3.]

Deakin, J. 2013. Devil Facial Tumor Disease Has The Potential To Provide Insight into Human Cancers. Abstract W 366. Plant and Animal Genome XXI Conference, 11-16 January 2013, San Diego, California.

Deakin, J.E. and K. Belov. 2012. A comparative genomics approach to understanding transmissible cancer in Tasmanian devils. Annual Review of Genomics and Human Genetics 13:207-222.

Deakin, J.E., H.S. Bender, A.-M. Pearse, R. Willem, P.C.M. O’Brien, M. A. Ferguson-Smith, Y. Cheng, K. Morris, R. Taylor, A. Stuart, K. Belov, C.T. Amemiya, E.P. Murchison, A.T. Papenfuss and J.A.M. Graves. 2012. Genomic restructuring in the Tasmanian Devil Facial Tumour: chromosome painting and gene mapping provide clues to evolution of a transmissible tumour. PLoS Genetics 8(2): e1002483. [For corrections to this paper, see The PLoS Genetics Staff, 2014]

Deakin, J.E., Y. Zhang, G. Sakthivel, J. Schoning, R. Taylor and A.-M. Pearse. 2014. Evolution of devil facial tumour disease chromosomes. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 80.

Denholm, M. 2005. Sympathy for the Devil. The Australian Magazine, 22 October 2005.

Denholm, M. 2006a. Devils in decline as disease bites hard. The Australian, 1 July 2006. Denholm, M. 2006b. Devil shipment insurance against extinction. Week-end Australian, 25-26 November 2006, p.11.

Denholm, M. 2007. Genetically superior devils to save species. The Australian, 22 February 2007, p. 3.

Denholm, M. 2008a. Hope of cure for dying Tasmanian devils. The Australian, 31 March 2008.

6/05/2016 8:41 AM 360

Denholm, M. 2008b. Vaccine promises to protect some devils from tumour. The Australian, 2 June 2008, p.7.

Denholm, M. 2009. Move to fence off Devil for its survival. The Australian 10 September 2009, p. 7.

Denholm, 2010a. Breakthrough in Tasmanian devil facial cancer fight. The Australian, 1 January 2010.

Denholm, M. 2010b. Genetic study raises hope for Tasmanian devils hit by facial tumour disease. The Australian, 10 March 2010.

Denholm, M. 2010c. ‘Special’ devils may save species from extinction in wild. The Australian, 11 March 2010.

Denholm, M. 2010d. Mainland home offers hope for endangered Tasmanian devil. The Australian, 14 July 2010.

Denholm, M. 2010e. Race for the Devil. Weekend Australian, 21-22 August 2010.

Denholm, M. 2012f. Devil island offers hope, sanctuary. The Australian, 15 November 2012.

Denholm, M. 2014. Tasmania plans devil-proof fence across Freycinet Peninsula. The Australian, 28 March 2014.

Denholm, M. 2016. Tasmanian devils ‘in danger’ from VPL dairy land clearing. The Australian, 27 February 2016.

Dennis, C. 2006. Endangered species: Time to raise the devil. Nature 439:530 [1 February]

Department of Primary Industries, Parks, Water and Environment. 2010a. Communication strategy. Stage 3:2008-2010. [18 February 2010; document properties indicate author to be K.M. Nolan]

Department of Primary Industries, Parks, Water and Environment. 2010b. Effects of 1080 fox baiting on Spotted-tailed Quolls and Tasmanian Devils. [19 May 2010] Department of Primary Industries, Parks, Water and Environment. 2010c. Code of practice for trapping and destruction of wallabies and brushtail possums in Tasmania for crop protection and commercial purposes. Department of Primary Industries, Parks, Water and Environment. [May 2010]

Department of Primary Industries, Parks, Water and Environment. 2010d. Draft Recovery Plan for the Tasmanian Devil (Sarcophilus harrisii). 51 pp. [18 October 2010]

6/05/2016 8:41 AM 361

Department of Primary Industries, Parks, Water and Environment. 2010e. Biodiversity Branch discussion paper regarding the Proposed Experimental Introduction of Tasmanian devils to Maria Island. Unpublished internal document, DPIPWE, Hobart, Tasmania. [Not available from DPIPWE; as reported in STDP, 2011b.]

Department of Primary Industries, Parks, Water and Environment. 2013. Annual Statewide Spotlight Surveys Tasmania 2012/13. Regional Summary: Priority Harvested Species. [Document properties indicate document was created on 26 July 2013, modified on 17 September 2013 and was authored by J. Dawson.]

Department of Primary Industries, Parks, Water and Environment. 2015a. The recovery plan for the Tasmanian devil (Sarcophilus harrisii). Department of Primary Industries, Parks, Water and Environment, Tasmania.

Department of Primary Industries, Parks, Water and Environment. 2015b. Annual State-wide Spotlight Surveys Tasmania 2014/15. Regional Summary: Priority Harvested Species.

Department of Primary Industries, Parks, Water and Environment. 2015c. Survey guidelines and management advice for development proposals that may impact on the Tasmanian Devil (Sarcophilus harrisii). A supplement to the Guidelines for Natural Values Surveys - Terrestrial Development Proposals. Dept of Primary Industries, Parks, Water and Environment, Hobart.

6/05/2016 8:41 AM 362

Department of Primary Industries and Water. 2007a. Threatened Species. 15 pp. [24 September 2007]

Department of Primary Industries and Water. 2007b. Wildlife Health in Tasmania. 42 pp. Online. [Author: A. Philips]

Department of Primary Industries and Water. 2008a. Species Information Sheet for the Tasmanian Devil. Unpublished report to the Department of Environment, Water, Heritage and the Arts.

Department of Primary Industries and Water. 2008b. Tasmanian Devils: Fact Sheet. [27 May 2008]

Department of Primary Industries, Water and Environment. 2004a. Latest information from the devil project. 2pp. [October 2004]

Department of Primary Industries, Water and Environment. 2004b. Information on Devil Facial Tumour Disease. 2 pp. [November 2004]

Department of Primary Industries, Water and Environment. 2004c. Latest information from the devil project. 4 pp. [December 2004]

Department of Primary Industries, Water and Environment. 2005a. Research into the Tasmanian Facial Tumour Disease (DFTD). Progress report; 28 pp. [January 2005]

Department of Primary Industries, Water and Environment. 2005b. Tasmanian Devil Facial Tumour Disease (DFTD). Disease Management Strategy. 21 pp. [February 2005]

Department of Primary Industries, Water and Environment. 2005c. Tasmanian Devil Facial Tumour Disease update; 5 pp. [June 2005]

Department of Primary Industries, Water and Environment. 2005d. Devil Facial Tumour Disease update; 5 pp. [August 2005] Department of Sustainability, Environment, Water, Population and Communities. 2010. Sarcophilus harrisii in Species Profile and Threats Database, Department of Sustainability, Environment, Water, Population and Communities, Canberra.

Department of the Environment and Heritage. 2006a. Sarcophilus harrisii – conservation advice. [26 May 2006]

Department of the Environment and Heritage. 2006b. EPBC Act Policy Statement 3.6 Tasmanian Devil (Sarcophilus harrisii). [July 2006] Department of the Environment, Water, Heritage and the Arts. 2009a. Advice to the Minister for the Environment, Water, Heritage and the Arts from the Threatened Species Scientific Committee (the Committee) on Amendment to the List of Threatened Species under the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act); 12 pp.

6/05/2016 8:41 AM 363

Department of the Environment, Water, Heritage and the Arts. 2009b. Approved Conservation Advice for Sarcophilus harrisii (Tasmanian Devil) (s266B of the Environment Protection and Biodiversity Conservation Act 1999); 4 pp.

Devil Island. 2013a. Mission Maria Island. Series 1, Episode 1; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 1 February 2014.

Devil Island. 2013b. Devil Island. Devil D-Day. Series 1, Episode 2; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 8 February 2014.

Devil Island. 2013c. Devil Island. Meet the Neighbours from Hell. Series 1, Episode 3; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 22 February 2014.

Devil Island. 2013d. Devil Island. Summer on Devil Island. Series 1, Episode 4; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 1 March 2014.

Devil Island. 2013e. Devil Island. Manny Goes Missing. Series 1, Episode 5; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 8 March 2014.

Devil Island. 2013f. Devil Island. Devil Dynasty. Series 1, Episode 6; 360 Degree Films, Brunswick, Victoria; as show on ABC IVIEW on 15 March 2014.

Dewar, E. 2013. Understanding Behaviour, Stress and Disease in Tasmanian Devils (Sarcophilus harrisii): Implications for Selective Adaptation. Honours Thesis, University of Tasmania. Not seen.

Dingwall, D. 2015. Devils dying on roads. The Advocate, 4 October 2015.

Dobbin, M. 2011. Sanctuary breeding hope for devils. The Age, 17 January 2011 (online).

Dobson, A.P. 2007. Sympathy for the Devil. EcoHealth 4(3):241-243.

Doman, C. 2014. Uncovering clues to save Tasmanian devil. Monitor Online (University of Canberra), 19 June 2014.

Doran, G. 1975. A note on tooth anomalies in the Tasmanian Devil, Sarcophilus harrisii. Australian Mammalogy 1(4):375.

Douglas, A.M., G.W. Kendrick and D. Merrilees. 1966. A fossil bone deposit near Perth, Western Australia, interpreted as a carnivore’s den after feeding tests on living Sarcophilus (Marsupialia, Dasyuridae). Journal and Proceedings of the Royal Society of Western Australia 49(3):88-90.

Dow, S. A saint among the devils. Sun Herald, 21 March 2010.

6/05/2016 8:41 AM 364

Driessen, M.M. and G.J. Hocking. 1992. Review and analysis of spotlight surveys in Tasmania 1975-1990. Scientific Report No. 92/1. Department of Parks, Wildlife and Heritage, Tasmania; 75 pp.

Dubois, G. and L.M. Angel. 1972. Strigeata (Trematoda) of Australian birds and mammals from the helminthological collection of the University of Adelaide. Transactions of the Royal Society of South Australia 96(4):197-215.

Duncan, P. 2006. Biologist gets DNA nod. Hobart Mercury, 22 November 2006.

Duncan, P. 2007. Ailing devil almost went to Denmark. Hobart Mercury, 12 January 2007.

Duncan, P. 2010a. Fox blitz across the state. Hobart Mercury, 14 March 2010.

Duncan, P. 2010b. Island home hope for devils. Hobart Mercury, 21 March 2010.

Duncan, P. 2010c. Devils export sales lash. Hobart Mercury, 23 May 2010.

Duncan, P. 2010d. Parks face devil of a dilemma. Hobart Mercury 30 May 2010.

Dunnett, G.M. 1970. Siphonaptera (Fleas). Pp. 647-655, in D.F. Waterhouse, ed., Insects of Australia. Melbourne University Press, Melbourne; # pp.

Dybas, C. 2013. Tasmanian devils: Will rare infectious cancer lead to their extinction? National Science Foundation press release, 13 November 2013.

Eastley, T. and A. Blanck. 2010. Wild diets for captive Tasmanian Devils: Is this the way to improved health and management of this species? Australasian Society of Zoo Keeping, Adelaide, Australia.

Ebrington, M. 2008. Cuddling the Devil. Mountain Views Mail, 30 October 2008.

Edwards, A. 2012a. Devils come out to play up at Barrington. The Newcastle Herald, 6 August 2012.

Edwards, A. 2012b. Little devils emerge on Tops in record numbers. Newcastle Herald, 26 September 2012.

Edwards, L. 2006. A devil's occupation. Proceedings of the 1st Australasian Regional Environmental Enrichment Conference 1st –2nd November, 2006, Royal Melbourne Zoological Gardens, Melbourne, Victoria.

Edwards, L. 2007. Mystery lesion makes rescue bid devilishly tricky. The Age, 12 October 2007. Effron, M., L. Griner and K. Bernirschke. 1977. Nature and rate of neoplasia found in captive wild mammals, birds and reptiles at necropsy. Journal of the National Cancer Institute 59(1):185-194.

6/05/2016 8:41 AM 365

Egan, J. 2012a. 40 Tassie devil babies boost ark population. Australian Geographic, 26 September 2012.

Egan, J. 2012b. Genetic diversity gives hope to Tassie devils. Australian Geographic, 3 December 2012.

Eisenberg, J.F., L.R. Collins and C. Wemmer. 1975. Communication in the Tasmanian Devil (Sarcophilus harrisii) and a survey of auditory communication in the Marsupalia. Zeitschrift für Tierpsycholgie 37:379-399.

Eisenberg, J.F. and I. Golani. 1977. Communication in Metatheria. Pp. 575-599, in T.A. Sebok, ed., How Animals Communicate. Indiana University Press; 1128 pp.

Eisenberg, J.F. and P. Leyhausen. 1972. The phylogenesis of predatory behaviour in mammals. Zeitschrift für Tierpsychologie 30:59-93.

Elder, J. Prom may be in for devil of a time. Brisbane Times and Sydney Morning Herald, 21 July 2013.

Eldridge, M.D.B. and C.J. Metcalfe. 2006. Order Dasyuromorphia. Pp. 21-30, in S.J. O’Brien, J.C. Menninger, W.G. Nash, eds, Atlas of Mammalian Chromosomes. Wiley, New York; 714 pp. Ellis, S. 2012. Sympathy for the devils. Brisbane Times, 26 July 2012.

Environmental Evidence Australia. 2013. Final Report – End of Program Review for Save the Tasmanian Devil Program; 117 pp.

Evershed, N. 2009. Lucky Devils put on a brave face. Sydney Morning Herald 2 July 2009, p. 5.

Ewer, R.F. 1968. The Ethology of Mammals. Logos Press, London; 418 pp.

Ewer, R.F. 1969. Some observations on the killing and eating of prey by two dasyurid marsupials: the Mulgara, Dasycercus cristicauda, and the Tasmanian Devil, Sarcophilus harrisii. Zeitschrift für Tierpsycholgie 26:23-38.

Fabian, T.J. 2010. Observations of Wounding in Wild Populations of Tasmanian Devils (Sarcophilus harrisii): Injury Rates Do Not Covary Linearly with Population Density. Save the Tasmanian Devil Program.

Fain, A. and R. Domrow. 1974. Two new parasitic mites (Acari: Sarcoptidae and Atopomelidae) from Tasmanian marsupials. Proceedings of the Linnean Society of New South Wales 98(3):122-130.

Fain, A. and B.R. Lawrence. 1975. Satanicoptes armatus, n. gen., n. sp. (Astigmata: Sarcoptidae), a new mite producing mange in the Tasmanian Devil (Sarcophilus harrisii Boitard). Journal of Medical Entomology, Honolulu 12(4):415-417.

6/05/2016 8:41 AM 366

Fancourt, B.A. 2015a. Drought, Disease or Devil declines? Identifying the cause of decline of the eastern quoll, Dasyurus viverrinus. Implications for conservation and management. PhD Thesis, University of Tasmania, Hobart; 190 pp. April 2015.

Fancourt, B. 2015b. Avoiding the subject: the implications of avoidance behaviour for detecting predators. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 24.

Fancourt, B.A., C.E. Hawkins, E.Z. Cameron, M.E. Jones and S.C. Nicol. 2015. Devil declines and catastrophic cascades: is mesopredator release of feral cats inhibiting recovery of the Eastern Quoll? PLoS ONE 10(3): e0119303. doi:10.1371/journal. pone.0119303. Published online, 11 March 2015.

Farmer, W.D. 2006. Conservation genetics of the Tasmanian Devil (Sarcophilus harrisii). B.Sc. Thesis, University of Tasmania.

Faulkner, T. 2014. Love is in the air and breeding season is in full swing at Devil Ark. Devil Ark website, Winter Update 2014, 10 June 2014 (accessed 31 July 2014).

Finch, M.E. and L. Freedman. 1988. Functional-morphology of the limbs of Thylacoleo carnifex Owen (Thylacoleonidae: Marsupialia). Australian Journal of Zoology 36(3): 251-272.

Finlay, E. 2015. Tasmanian devil at Australian Reptile Park has strong bond with the keeper who raised her from the start. Daily Telegraph, E. 5 August 2015.

Finnie, E.P. and D.E. Thomas. 1982. Management of monotremes at Taronga Zoo, Sydney. Pp. 1-3, in D.D. Evans, The Management of Australian Mammals in Captivity. Proceedings of the Scientific Meeting of the Australian Mammal Society, Healesville, Victoria, February 1979. The Zoological Board of Victoria; 188 pp.

Firestone, K.B., M.S. Elphinstone, W.B. Sherwin and B.A. Houlden. 1999. Phylogeographical population structure of tiger quolls Dasyurus maculatus (Dasyuridae: Marsupialia), an endangered carnivorous marsupial. Molecular Ecology 8: 1613–1625.

Firestone, K.B., B.A. Holden, W.B. Sherwin and E. Geffen. 2000. Variability and differentiation of microsatellites in the genus Dasyurus and conservation implications for the large Australian carnivorous marsupials. Conservation Genetics 1: 115–133.

Flannery, T. 1994. The Future Eaters. Reed Books, Kew, Victoria; 423 pp.

6/05/2016 8:41 AM 367

Fleay, D. 1935. Notes on the breeding of Tasmanian Devils. Victorian Naturalist 52:100-105.

Fleay, D. 1946. On the trail of the marsupial wolf. Victorian Naturalist 63: 129- 135, 154-159 and 174-177.

Fleay, D. 1952. The Tasmanian or Marsupial Devil – its habits and family life. The Australian Museum Magazine 10(9):274-280. Fleay, D. 1990. The devil's brood. Wildlife Australia, Autumn 1990:3-5. [Largely a repeat of Fleay, 1935.]

Flood, J. 1973. Pleistocene human occupation and extinct fauna in Cloggs Cave, Buchan, south-eastern Australia. Nature 246:303.

Flood, J. 1983. Archaeology of the Dreamtime. Collins, Sydney; # pp. Flood, J. 2007. Late survival of megafauna in Gippsland: dated faunal sequences from Cloggs and other Buchan caves. Pp 51-54, in M.L. Cupper and S.J. Gallagher, eds, Climate change or human impact? Australia's megafaunal extinction, Selwyn Symposium of the GSA Victoria Division, September 2007, Geological Society of Australia. Extended Abstracts No. 79; 63 pp.

Florez, P.A.R.D. 2004. Conservation endocrinology of the large dasyurids in Tasmania. B.Sc. Honours Thesis, University of Tasmania.

Flower, S.S. 1931. Contributions to our knowledge of the duration of life in vertebrate animals. – Mammals. Proceedings of the Zoological Society of London 101(1):145- 234.

Flynn, T.T. 1910. Contribution to a knowledge of the anatomy and development of the Marsupialia. Note i. The genitalia of Sarcophilus satanicus (♀). Proceedings of the Linnaean Society of New South Wales 35(4):873-887.

Flynn, T.T. 1911. Notes on marsupialian anatomy. II. On the female genital organs of a virgin Sarcophilus satanicus. Proceeding of the Royal Society of Tasmania 1911:144-156.

Flynn, T.T. 1922. Notes on certain reproductive phenomena in some Tasmanian marsupials. Annals and Magazine of Natural History (9)10:225- 231. Flynn, T.T. 1939. Note on the egg of the rare marsupial, Sarcophilus. Proceedings of the Royal Irish Academy. Minutes of Proceedings Session 1938-39 45:16-19.

Foley, M. 2009. ‘Demoralizing’ setback in fight to save Tasmanian devil. The New York Times, 7 January 2009.

Ford, S. 2013a. Devils detail key to Tarkine mine approval. The Advocate, 23 July 2013.

Ford, S. 2013b. Butler's fact-finding mission on Tarkine. The Advocate, 24 July 2013.

6/05/2016 8:41 AM 368

Fox, N. 2007. Investigation of the allograft hypothesis using amplified fragmented length polymorphism genetic markers. P. 11, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Fox, S., D. Lee, D. Sinn, S. Thalmann and P. Wise. 2010a. Evaluating functional extinction in Tasmanian devils – what, how and when? Australian Mammal Society 56th Meeting, 4-9 July 2010. Abstract.

Fox, S., D. Pemberton, C. Grueber, E. Reid-Wainscoat and R. Pye. 2015. Returning devils to the wild – not such a straight forward affair!! 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 27.

Fox, S., P. Wise, D. Lee, S. Thalmann, D. Sinn and S. Huxtable. 2010b. Locating the western extent of devil facial tumour disease 2009: monitoring and management field team. Department of Primary Industries, Parks, Water and Environment, Hobart; 11 pp.

Fraser, S. 2005. What the Devil? Current Science 91(6):4-5.

Gallasch, R. 2011. Eco disaster devil fear. The Examiner, 5 September 2011.

Gallus, S, B.M. Hallström, V. Kumar, W.G. Dodt, A. Janke, G.G. Schumann and M.A. Nilsson. 2015a. Evolutionary histories of transposable elements in the genome of the largest living marsupial carnivore, the Tasmanian devil. Molecular Biology and Evolution, doi: 10.1093/molbev/msv017. Published online, 28 January 2015.

Gallus, S., A. Janke, V. Jumar and M.A. Nilsson. 2015b. Disentangling the relationship of the Australian marsupial orders using retrotransposon and evolutionary network analyses. Genome Biology and Evolution, Advance Access, doi:10.1093/gbe/evv052. Published online, 18 March 2015.

Garvey, J. 2007. Surviving an ice age: the zooarchaeological record from southwestern Tasmania. Palaios 22: 583-585.

Gathercole, J.L. 2012. Biomarker discovery for pre-clinical diagnosis of Tasmanian Devil Facial Tumour Disease. PhD Thesis, University of Tasmania.

Gelineau, K. 2010. Famed Tasmanian Devil euthanized after tumor found. Associated Press, 2 September 2010.

Geoffroy Sainte-Hilaire, E. 1810. Description de deux espèces de dasyures (Dasyurus cyanocephalus et Dasyurus ursinus). Annals du Muséum de Histoire Naturelle 15:301-306

6/05/2016 8:41 AM 369

Georgihiou, P.R., T.F. Mollee and M.H. Tilse. 1992. Pasteurella multocida infection after a Tasmanian Devil bite. Clinical Infectious Diseases 14(6):1266-1267. Gerrard, E. 1862. Catalogue of the Bones of the Mammalia in the Collection of the British Museum. Taylor and Francis, London; 316 pp.

Gerdtz , W.R. 2001. McCoy and Sarcophilus harrisii Boitard, 1842 - A diabolical relationship. The Victorian Naturalist 118( 5): 231- 233.

Gerdtz, W. and N. Archbold. 2003a. Glaucodon ballaratensis (Marsupialia, Dasyuridae), a Late Pliocene 'devil' from Batesford, Victoria. Proceedings of the Royal Society of Victoria 115(2):35-44.

Gerdtz, W.R and N.W. Archbold. 2003b. An early occurrence of Sarcophilus laniarius harrisii (Marsupialia, Dasyuridae) from the early Pleistocene of Nelson Bay, Victoria, Proceedings of the Royal Society of Victoria 115(2):45-54.

Gewalt, W. 1977. Tasmanian Devil twins (Sarcophilus harrisii) born at Duisburg zoo. International Zoo News 24(7):12-13.

Gewalt, W. 1980.Űber einige seltenere Naschsuchen im Zoo Duisberg, 2. Beutelteufel (Sarcophilus harrisii). Zool. Gart. N.F., Jena 50(2-3): 138-154.

Gibson, S. 2015. Captive Tasmanian devils to be set free to boost wild population. ABC News online, 23 September 2015.

6/05/2016 8:41 AM 370

Gilbert, C., P. Waters, C. Feschotte and S. Schaack. 2013. Horizontal transfer of OC1 transposons in the Tasmanian devil. BMC Genomics 14:134 doi:10.1186/1471-2164-14-134. [Published online 17 February 2013.]

Giles, C. 2012. The detail is in the devil. Wellcome Trust Blog, 17 February 2012.

Gill, E.D. 1953. Distribution of the Tasmanian Devil, Tasmanian Wolf and the dingo in S.E. Australia in Quaternary time. Victorian Naturalist 70(5):86-90.

Gill, E.D. 1955. The problem of extinction, with special reference to Australian marsupials. Evolution 9(1):87-92.

Gill, E.D. 1971. The Australian aborigines and the Tasmanian Devil. Mankind 8:59-60.

Gill, K. 2012. Sydney Science Forum: Sympathy for the Devil. University of Sydney News, 24 July 2012.

Gillespie, R., D.R. Horton, P. Ladd, P.G. Macumber, T.H. Rich, R. Thorn and R.V.S. Wright. 1978. Lancefield Swamp and the extinction of the Australian megafauna. Science 200:1044-1048.

Glaetzer, S. 2014a. Denmark's Copenhagen Zoo plans to lead way in Tasmanian devil care. Sunday Tasmanian, 5 January 2014.

Glaetzer, 2014b. Vaccine hope to protect devils from devastating facial tumour disease. Hobart Mercury, 24 November 2014. Gollan, K. 1984. The Australian dingo: in the shadow of man. Pp. 921-927, in M. Archer and G. Clayton, eds, Vertebrate Zoogeography and Evolution in Australasia: Animals in Space and Time. Hesperian Press, Carlisle, #; # pp.

Gonçalves, J., C.A. Marks, D. Obendorf, A. Amorima and F. Pereira. 2014. The risks of using ‘‘species-speciﬁc’’ PCR assays in wildlife research: The case of red fox (Vulpes vulpes) identiﬁcation in Tasmania. Forensic Science International: Genetics 11(#):e9-e11.

Gould, J. 1863. The Mammals of Australia. London; 416 pp Glauert, L. 1912. Fossil marsupial remains from Balladonia in the Eucla Division. Records of the Western Australian Museum and Art Gallery 1(2):47-65.

Glauert, L. 1914. The Mammoth Cave. Records of the Western Australian Museum and Art Gallery 1(3):244-251.

Glauert, L. 1948. The cave fossils of the south-west. Western Australian Naturalist 1(5):100-104.

Gliddon, 2005. Things that grow lumps in the night. The Bulletin 143(5), 25 January 2005.

6/05/2016 8:41 AM 371

Goldenfein, J. 2008. Experts losing battle to save devils. Cosmos Online, 4 December 2008.

Goldenfein, J. 2009. Experts losing battle to save devils. Cosmos 25:12.

Graham, S. 2013. Solving a devil of a problem . Public Administration Today 36: 45- 47.

Gramenz, E. 2015. Two of 20 immunised Tasmanian devils released into wild killed on road days after re-introduction. ABC News online, 30 September 2015.

Gramling, C. 2006. Poor Devils: Critters' fights transmit cancer. ScienceNews 169(5):67.

Graves, J.A.M. 2010a. Genomics to the rescue of the Devil. ARDR, 1 February 2010.

Graves, J.A.M. 2010b. Marsupial genome sequencing. Pp. 117-119, in J.E. Deakin, P.D. Waters, J.A.M. Graves, eds, Marsupial Genetics and Genomics. Springer, Dordrecht; 519 pp.Marsupial Genetics and Genomics 9: 499-515.

Gray, J.E. 1841. Contributions towards the geographical distribution of the Mammalia of Australia, with notes on some recently discovered species, in a letter addressed to the Author. Appendix C, pp. 397-414, in G. Grey, Journals of Two Expeditions of Discovery in North-West and Western Australia, During the Years 1837, 38 and 39, Under the Authority of Her Majesty’s Government. Describing many newly discovered, important, and fertile districts, with observations on the moral and physical conditions of the aboriginal inhabitants, &c, &c. T.&W. Boone, Volume 2; 482 pp. [see p. 400]

Green, B. 1984. Composition of milk and energetics of growth in marsupials. Symposium of the Zoological Society of London 51:369-387.

Green, B. 1997. Field energetics and water fluxes in free-living macropods. Pp. 141- 162, in N.R. Saunders and L.A. Hinds, eds, Marsupial Biology. Recent Research, New Perspectives. University of New South Wales Press, Sydney; 413 pp.

Green, B. and I. Eberhard. 1979. Energy requirements and sodium and water turnovers in two captive marsupial carnivores: the Tasmanian devil, Sarcophilus harrisii, and the Native Cat, Dasyurus viverrinus. Australian Journal of Zoology 27:1–8. Green, R.H. 1967. Notes on the Devil (Sarcophilus harrisii) and the Quoll (Dasyurus viverrinus) in north-eastern Tasmania. Records of the Queen Victoria Museum 27:1- 13.

Green, R.H. 1973. Mammals of Tasmania. Foot and Playsted Pty Ltd, Launceston; 76 pp. Privately printed.

Green, R.H.. 1983. An Illustrated Key to the Skulls of the Mammals in Tasmania. Queen Victoria Museum and Art Gallery, Launceston; # pp.

6/05/2016 8:41 AM 372

Greenwood, A.W. 1923. Marsupial spermatogenesis. Quarterly Journal of Microscopical Science 67:203-218.

Greer, A.E. 2007. Wollemi Pine. Review, Australian Financial Review, 30 March 2007, pp. 1-2, 10.

Greer, A.E. 2009. Cloning the Thylacine. Quadrant 53(7-8):28-39.8 July 2009.

Greer, A.E. 2011a. Vale Cedric. The Life and Death of a Tasmanian Devil. Quadrant 55(4):91-93.

Greer, A.E. 2011b. The Devil’s in the detail. The Hobart Mercury, 16 September 2011.

Gregory, G.G. 1972. Taenia ovis in the Tasmanian Devil. Australian Veterinary Journal 48(7):423-424.

Gregory, G.G. 1976. The role of Tasmanian Devil in the epizootiology of ovine cysticercosis and echinococcosis. Australian Veterinary Journal 52(10):471- 472.

Gregory, G.G., B.L. Munday, I. Beveridge and M.D. Rickard. 1975. Studies on Anoplotaenia dayuri Beddard, 1911 (Cestoda: Taeniidae), a parasite of the Tasmanian Devil: life cycle and epidemiology. International Journal of Parasitology 5(2):187-191.

Grenfell, A. 2006. Devil’s advocates. Gold Coast Sun, 6 December 2006.

Griffiths, P. 2014. Tiny ‘devil’ born outside of breeding season gets lucky. Daily Telegraph, 11 March 2014.

Griner, L.A. 1979. Neoplasms in Tasmanian Devils (Sarcophilus harrisii). Journal of the National Cancer Institute 62(3):589-595.

Griner, L.A. 1983. Family Dasyuridae. Pp. 269-281, in Pathology of Zoo Animals: A Review of Necropsies Conducted over a Fourteen-year Period at the San Diego Zoo and San Diego Wild Animal Park. Zoological Society of San Diego, San Diego, California; # pp.

Groom, M. 2014. Saving our Devils' Wild Future. Tasmanian Government media release, 23 November 2014.

Groom, M. 2015a. Bringing devils back into the Tasmanian wild. Tasmanian Government media release, 26 September 2015.

Groom, M. 2015b. Saving the devil: more healthy devils released. Tasmanian Government media release, 18 November 2015

6/05/2016 8:41 AM 373

Grueber, C.E., K. Belov and C.J. Hogg. 2014. Studying immunogenetic diversity to aid the conservation of threatened species. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 32.

Grueber, C.E., C.J. Hogg, J.A. Ivy and K. Belov. 2015a. Impacts of early viability selection on management of inbreeding and genetic diversity in conservation. Molecular Ecology 24:1645-1653.

Grueber, C.E., E. Peel, R. Gooley and K. Belov. 2015b. Genomic insights into a contagious cancer in Tasmanian devils. Trends in Genetics 31(9):528-535.

Grzelewski, D. 2002. Give the devil his due: blame Bugs Bunny and a nasty yawn for the Tasmanian devil's bad rap. Smithsonian 32(11):68-75.

Grzelewski, D. 2003. Dining with the devil. Australian Geographic 70 (April-June): 79-91.

Guiler, E.R. 1957. The present status of some Tasmanian mammals in relation to the fur industry of Tasmania. Papers and Proceedings of the Royal Society of Tasmania 91:117-128.

Guiler, E.R. 1960. Marsupials of Tasmania. Tasmanian Museum and Art Gallery, Hobart; 31 pp.

Guiler, E.R. 1961. The former distribution and decline of the thylacine. Australian Journal of Science 23(7): 207-210.

Guiler, E.R. 1964. Tasmanian Devils. Australian Natural History 14(11):360-362.

Guiler, E.R. 1970a. Observations on the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). I. Numbers, home range, movements and food in two populations. Australian Journal of Zoology 18:49-62. Guiler, E.R. 1970b. Observations on the Tasmanian devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). II. Reproduction, breeding, and growth of pouch young. Australian Journal of Zoology 18:63-70.

Guiler, E.R. 1970c. Tasmanian Devils and agriculture. Tasmanian Journal of Agriculture 42(2):134-137.

Guiler, E.R. 1971. The Tasmanian Devil, Sarcophilus harrisii, in captivity. International Zoo Yearbook 11:32-33.

Guiler, E.R. 1972. Some thoughts on man and the mammals in the Central Plateau. Pp. 125-134, in The Lake Country of Tasmania. A Symposium Conducted by the Royal Society of Tasmania at Poatina, Tasmania, November 11-12, Papers and Proceedings of the Royal Society of Tasmania 1972.

Guiler, E.R. 1976. Albinism in the Tasmanian devil. Papers and Proceedings of the Royal Society of Tasmania 110:155-156.

6/05/2016 8:41 AM 374

Guiler, E.R. 1978. Observations on the Tasmanian devil, Sarcophilus harrisii (Dasyuridae: Marsupialia) at Granville Harbour, 1966-75. Papers and Proceedings of the Royal Society of Tasmania 112:161-188.

Guiler, E.R. 1982. Temporal and spatial distribution of the Tasmanian Devil, Sarcophilus harrisii (Dasyuridae: Marsupialia). Papers and Proceedings of the Royal Society of Tasmania. 116:153-163.

Guiler, E.R. 1983a. Tasmanian Devils - Australia’s ugliest marsupials. Australian Natural History 21(3):117-119.

Guiler, E.R. 1983b. Pp. 27-28, in R. Strahan, ed., The Australian Museum Complete Book of Australian Mammals. Angus and Robertson Publishers, North Ryde; 530 pp.

Guiler, E.R. 1985. Thylacine: the Tragedy of the Tasmanian Tiger. Oxford University Press, Melbourne; 207 pp.

Guiler, E.R. 1992. The Tasmanian Devil. St David’s Park Publishing, Hobart; 28 pp.

Guiler, E.R. 2001. The Enthusiastic Amateurs: The Animals and Birds Protection Board 1929-1971. E.R. Guiler, Sandy Bay; 122 pp.

Guiler, E.R. and P. Godard. 1998. The Tasmanian Tiger: a Lesson to be Learnt. Abrolhos Publishing Pty Ltd; 256 pp.

Guiler, E.R. and R.W.L. Heddle. 1970. Testicular and body temperatures in the Tasmanian Devil and three other species of marsupial. Comparative Biochemistry ad Physiology 33:881-891.

Guiler, E.R. and R.W.L. Heddle. 1974a. The eruption and growth of teeth in the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae).Papers and Proceedings of the Royal Society of Tasmania 108:137-140.

Guiler, E.R. and R.W.L. Heddle. 1974b. Body temperatures in the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). Comparative Biochemistry and Physiology 47A:981-989.

Gunn, R.C. 1852. A list of mammals indigenous to Tasmania. Proceedings of the Royal Society of Tasmania 2(1): 77-90.

Gunn, R.C. and J.A. Gray. 1838. Notices accompanying a collection of quadrupeds and fish from Van Diemen’s Land. With notes and descriptions of new species. Annals of Natural History; or, Magazine of Zoology, Botany, and Geology I:101-111.

Gutman, N., M.J. Hansen, M.F. Bertelsen and A.M. Bojesen. 2016. Pasteurellaceae bacteria from the oral cavity of Tasmanian Devils (Sarcophilus harrisii) show high MIC values towards aminoglycosides and clindamycin. Letters in Applied Microbiology, published online 7 January 2016.DOI: 10.1111/lam.12545.

6/05/2016 8:41 AM 375

Hale, H.M, 1956. The first hundred years of the museum – 1856-1956. Records of the South Australian Museum 12:1-225.

Hale, H.M. and N.B. Tindale. 1930. Notes on some human remains in the Lower Murray Valley, South Australia. Records of the South Australian Museum 4:145-218.

Hall, J. and R. Jones. 1990. Palaeoscatology and taphonomic implications at Rocky Cape, northwestern Tasmania. Pp. 219-229, in S. Solomon, I. Davidson and D. Watson, eds, Problem Solving in Taphonomy. Archaeological and Palaeontological Studies from Europe, Africa and Oceania. Tempus 2.

Hallam, J.F., R.A.B. Holland and T.J. Dawson. 1995. The blood of carnivorous marsupials: low haemoglobin oxygen affinity. Physiological Zoology 68(2):342-354.

Hamede, R. 2004. Patterns of contact in wild Tasmanian devils and the implications for postulated transmission dynamics of DFTD. Honours Thesis, School of Zoology, University of Tasmania, Australia. [Not seen.]

Hamede, R. 2007. Epidemiology and transmission dynamics of DFTD. P. 29, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Hamede, R. 2004. Patterns of contact in wild Tasmanian devils and the implications for postulated transmission dynamics of DFTD. Honours Thesis, School of Zoology, University of Tasmania, Australia.

Hamede, R. 2012. The ecology and epidemiology of Devil Facial Tumour Disease. PhD thesis, School of Zoology, University of Tasmania, Australia.

Hamede, R.R., J. Bashford, J.M. Jones and H. McCallum. 2012b. Simulating devil facial tumour disease outbreaks across empirically derived contact networks. Journal of Applied Ecology 49:447-456.

Hamede, R.K., J. Bashford, H. McCallum and M. Jones. 2009. Contact networks in a wild Tasmanian Devil (Sarcophilus harrisii) population: using social network analysis to reveal seasonal variability in social behaviour and its implications for transmission of devil facial tumour disease. Ecology Letters 12(11):1147-1157.

Hamede, R., S. Lachish, K. Belov, G. Woods, A. Kreiss, A.-M. Pearse, B. Lazenby, M. Jones and H. McCallum. 2012a. Reduced effect of Tasmanian Devil facial tumor disease at the disease front. Conservation Biology 26(1):124-134.

Hamede, R.K., H. McCallum and M. Jones. 2008. Seasonal, demographic and density-related patterns of contact between Tasmanian devils (Sarcophilus harrisii): Implications for transmission of devil facial tumour disease. Austral Ecology 33(5):614–622.

Hamede, R.K., H. McCallum and M. Jones. 2013. Biting injuries and transmission of Tasmanian devil facial tumour disease. Journal of Animal Ecology 82(1):182-190.

6/05/2016 8:41 AM 376

Hamede, R., A.-M. Pearse, K. Swift, L. Barmuta, E. Murchison, N. Beeton, S. Carver and M. Jones. 2015. What’s in a strain? Competition in tumour genetic lineages and host population response in Tasmanian Devil Facial Tumour Disease. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 33.

Hamede, R.K., A.-M. Pearse, K. Swift, L.A. Barmuta, E.P. Murchison and M.E. Jones. 2015. Transmissible cancer in Tasmanian devils: localized lineage replacement and host population response. Proceedings of the Royal Society, B 282:20151468. Published online, 2 September 2015.

Hamilton, D., R. Hamede, M. Jones and E. Cameron. 2015. Social networks and transmission of facial tumour disease. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 14.

Hamilton, P.B., J.R. Stevens, J.Gidley, P. Holz and W.C.Gibson. 2005a. A new lineage of trypanosomes from Australian vertebrates and terrestrial bloodsucking leeches (Haemadipsidae). International Journal of Parasitology 35: 431–443.

Hamilton, P.B., J.R.Stevens, P. Holz, B. Boag, B.Cooke and W.C. Gibson. 2005b. The inadvertent introduction into Australia of Trypanosoma nabiasi, the trypanosome of the European rabbit (Oryctolagus cuniculus), and its potential for biocontrol. Molecular Ecology 14:3167–3175.

Hansen, A. 2006. Devils in distress. Alternatives Journal 32(4/5):6.

Hansen, B. 2015. Progress on vaccine for Tasmanian devil tumours. The Advocate, 26 February 2015.

Hanson, J. 2012a. How captivity affects the brain size of animals. Australian Geographic, 17 February 2012.

Hanson, J. 2012b. Tasmanian devil genome offers cancer clues. Australian Geographic, 20 February 2012.

Harington, J.S., D.L. Obendorf and N.D. McGlashan. 2006. Researching the Tasmanian devil facial tumour. Australian Veterinary Journal 84(5):N 26.

Harington, J.S. and N.D. McGlashan. 2005. The Tasmanian Devil Facial Tumour Disease (DFTD) – a problem unresolved. Annals of the Australasian College of Tropical Medicine 6(2):34-35. [Conference extract.]

Harris, G.P. 1808. Description of two new species of Didelphis from Van Dieman’s Land. Transactions of the Linnean Society of London 9:174-178.

Harris, T. 2010. Australia moving cancer-hit Tasmanian devils to new islands. Sydney Morning Herald, 21 January 2010.

6/05/2016 8:41 AM 377

Hartigan, J. Devil of a visit to Gresford. Dungog Chronicle, 17 December 2013.

Hartz, G. 1966. Notes on some recent mammal breeding successes at Lincoln Park Zoo, Chicago. International Zoo Yearbook 6:206-207.

Hawkins, C. 2007. Figuring out how to mitigate Devil facial tumour disease through monitoring and modelling. P. 19, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Hawkins, C., C. Baars, H. Hesterman, G. Hocking, M. Jones, B. Lazenby, D. Mann, N. Mooney, D. Pemberton, S. Pyecroft, S. Restani and J. Wiersma. 2006. Emerging disease and population decline of an island endemic, the Tasmanian devil Sarcophilus harrisii. Biological Conservation 131:307-324.

Hawkins, C.E., M.E. Jones, N. Mooney et al., 2005. Epidemiological features of a new disease in the Tasmanian Devil. P. 58, in Proceedings of the Wildlife Disease Association, Cairns, Australia; # pp.

Hawkins, C.E., H. McCallum, N. Mooney, M. Jones and M. Holdsworth. 2008. Sarcophilus harrisii. In, IUCN 2010, IUCN Red List of Threatened Species. Version 2010.1. www.jucnredlist.org

Hayashi1, K., M. Sugisaki, K. Kino, T. Ishikawa, S. Kawashima and T.Amemiya. 2014. Morphological Characteristics of the temporomandibular joint in the pouch young of the Tasmanian Devil. Anatomia, Histologia, Embryologia 43(2):157-160. Article first published online: 8 May 2014. DOI: 10.1111/ahe.12118

Hayashi, K., M. Sugisaki, K. Kino, T. Ishikawa, M. Sugisaki and S. Abe. 2013. Absence of the articular disc in the Tasmanian Devil temporomandibular joint. Anatomia, Histologia, Embryologia 42(6):415-419. [Published online 26 February 2013.]

Hayden, E.C. 2009. Genome scan may save Tasmanian Devils from cancer. Nature News (online). 3 March 2009.

Hayman, D.L. and P.G. Martin. 1974. Mammalia I: Monotremata and Marsupialia. Pp, #-#, in B. John, ed., Animal Cytogenetics. Gebriider Borntraeger, Berlin and Stuttgart; # pp.

Hayward, J.S. and P.A. Lisson. 1992. Evolution of brown fat: its absence in marsupials and monotremes. Canadian Journal of Zoology 70(3):171-179.

Heath, M. 2008. Australia searches for tumor-resistant devils to save species. Bloomberg, 17 October 2008.

Held, K.S. 2016. Tasmanian devil habitat opening at Saint Louis Zoo.

6/05/2016 8:41 AM 378

Fox2Now (St Louis), 20 April 2016.

Henty, K., R.M.G. Wells and B. Thomas. 2007. Characterization of the hemoglobins of the adult brushtailed possum, Trichosurus vulpecula (Kerr) reveals non-genetic heterogeneity. Comparative Biochemistry and Physiology - Part A: Molecular and Integrative Physiology 148(3): 498-503.

Hesterman, H. 2007. Establishment and management of captive insurance populations. P. 32, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Hesterman, H. 2008. Reproductive physiology of the Tasmanian devil (Sarcophilus harrisii) and spotted-tailed quoll (Dasyurus maculata). PhD Thesis, University of Tasmania, Hobart; # pp.

Hesterman, H. and S.M. Jones. 2009. Longitudinal monitoring of plasma and fecal androgens in the Tasmanian devil (Sarcophilus harrisii) and the spotted-tailed quoll (Dasyurus maculatus). Animal Reproduction Science 112(3-4):334-346.

Hesterman, H., S.M. Jones and F. Schwarzenberger. 2008a. Reproductive endocrinology of the largest dasyurids: Characterization of ovarian cycles by plasma and fecal steroid monitoring. Part I. The Tasmanian devil (Sarcophilus harrisii). General and Comparative Endocrinology 155(1):234-244.

Hesterman, H., S.M. Jones and F. Schwarzenberger. 2008b. Pouch appearance is a reliable indicator of the reproductive status in the Tasmanian devil and the spotted- tailed quoll. Journal of Zoology 275(2):130-138.

Hett, G.S. and H.G. Butterfield. 1909. The anatomy of the palatine tonsils. Journal of Anatomy and Physiology 44(1):35-55.

Hewavisenti, R., K. Morris, D. O’Meally, Y. Cheng and K. Belov. 2014. The identification and characterisation of immune genes in the milk transcriptome of the Tasmanian devil (Sarcophilus harrisii). Genetics Society of Australasia Conference 6- 9 July 2014, University of Sydney. Abstract # 31.

Hewavisenti, R.V., K.M. Morris, D. O’Meally, T. Faulkner, L. Vella et al. 2014. The Identification of immune genes in the milk transcriptome of the Tasmanian devil (Sarcophilus harrisii). Submitted manuscript. Cited in Morris et al., 2015.

Hibbard, C., C. Srb, J. Hockley, C. Hogg and C. Ford. 2011. Tasmanian devil conservation program annual report and recommendations. Zoo Aquarium Association unpublished report. [Not seen.]

Hibbard, C., C. Srb, J. Hockley and C. Hogg. 2012. Annual Report and Recommendations. Full Report. Tasmanian Devil Conservation Program. Zoo and Aquarium Association. Draft. Accessed online, 14 December 2012.

6/05/2016 8:41 AM 379

Hibberd, R. 2006. Tasmanian Devil. The Companion to Tasmanian History, Centre for Tasmanian Historical Studies; 1 p.

Hildebrandt, M. 1977. Analysis of asymmetrical gaits. Journal of Mammalogy 58(2):131-156.

Hird, D. 1995. Vertebrates of Lambert Park / Skyline Reserve on Mt Nelson in southern Tasmania. Tasmanian Naturalist 117: 52-66.

Hird, D. 2000. Terrestrial mammals of Denne’s Hill, North Bruny Island. Tasmanian Naturalist 122: 9-14.

Hird, D. and T. Hammer. 1995. Mammals of Gumtop spur in the northwest Wellington Park with comments on a new habitat type for the barred bandicoot. Tasmanian Naturalist 117: 32-38.

Hobday, A.J. 2010. Nighttime driver detection distances for Tasmanian fauna: informing speed limits to reduce roadkill. Wildlife Research 37:265-272.

Hobday, A.J. and M. Minstrell. 2008. Distribution and abundance of roadkill on Tasmanian highways: human management options. Wildlife Research 35 (7):712-726.

Hocking, G.J. and M.M. Driessen. 1992. Spotlight Survey Manual. A Set of Instructions for Conducting Spotlight Surveys in Tasmania. Department of Primary Industries, Parks and Water and Environment. Hobart, Tasmania; 10 pp.

Hocking, G.J. and E.R. Guiler. 1983. The mammals of the Lower Gordon River Region, south-west Tasmania. Australian Wildlife Research 10(1):1-23.

Hoffmann, M., C.E. Hawkins and P.D. Walsh. 2008. Action needed to prevent extinctions caused by disease. Nature 454:159.

Hogg, C.J. 2013. Preserving Australian native fauna: zoo-based breeding programs as part of a more unified strategic approach. Australian Journal of Zoology 61:101-108.

Hogg, C.J. and J. Hockley. 2013. DPIPWE/ZAA husbandry guidelines for Tasmanian devil, Sarcophilus harrisii. Third Edition. Australia: Zoo and Aquarium Association.

Hogg, C.J., C. Srb and J. Hockley. 2013. Annual Report for the DPIPWE-ZAA Tasmanian Devil Insurance Population. Zoo and Aquarium Association, Mosman, NSW. Not seen.

Hogg, C.J., J.A. Ivy, C. Srb, J. Hockley, C. Lees, C. Hibbard and M. Jones. 2015. Influence of genetic provenance and birth origin on productivity of the Tasmanian devil insurance population. Conservation Genetics. First available online, 10 July 2015; published online, 4 January 2016.

Hogg, C.J., C.E. Grueber, D. Pemberton, S. Fox, A.V. Lee, J.A. Ivy and K. Belov. 2016. “Devil tools & tech”: a synergy of conservation research and management practice. Conservation Letters #(#):#-#. Published online, 9 December 2015.

6/05/2016 8:41 AM 380

Hoggett, C. 2012. Tiger snake versus devil. Hobart Mercury, 27 April 2012.

Hollings, T. 2013. Ecological effects of disease-induced apex predator decline. The Tasmanian Devil and Devil Facial Tumour Disease. PhD Thesis, University of Tasmania; 193 pp.

Hollings, T., G. Hocking, M. Jones, N. Mooney and H. McCallum. 2011a. Ecosystem impacts following disease-induced top predator decline: The Tasmanian Devil and DFTD. Abstract. Wildlife Society’s 18th Annual Conference, Waikoloa, Hawaii, 5-10 November 2011.

Hollings, T., G. Hocking, M. Jones, N. Mooney and H. McCallum. 2011b. Top predator decline, mesopredator release, and disease transmission: the case of the Tasmanian devil, feral cat and toxoplasmosis. Abstract. International Congress for Conservation Biology, Auckland, New Zealand, 5-9 December 2011.

Hollings, T., M. Jones, N. Mooney and H. McCallum. 2013. Wildlife disease ecology in changing landscapes: mesopredator release and toxoplasmosis. International Journal for Parasitology: Parasites and Wildlife 2:110-118.

Hollings, T., M. Jones, N. Mooney and H. McCallum. 2014. Trophic cascades following the disease-induced decline of an apex predator, the Tasmanian Devil. Conservation Biology 28(1):63-75. Published online 11 September 2013.

Hollings, T., M. Jones, N. Mooney and H. McCallum. Submitted. Ecosystem impacts of disease induced apex predator decline: the Tasmanian devil and devil facial tumour disease (DFTD). [As cited in Hollings et al., 2013]

Hollings, T., M. Jones, N. Mooney and H. McCallum. 2015a. Disease-induced decline of an apex predator drives invasive dominated states and threatens biodiversity. Ecology 96(10): #-#. (doi: 10.1890/15-0204.1)

Hollings, T., H. McCallum, K. Kreger, N. Mooney and M. Jones. 2015b. Relaxation of risk-sensitive behaviour of prey following disease-induced decline of an apex predator, the Tasmanian devil. Proceedings of the Royal Society B, published online 17 June 2015.

Holtcamp, W. 2007. Sympathy for the Devil. Scientific American 296(3):14.

Holtcamp, W. 2008. Tasmania’s Devil of a problem. National Wildlife 46(4):#-#.

Holz, P. 1992. Immobilization of marsupials with tiletamine and zolazepam. Journal of Zoo and Wildlife Medicine 23: 426-428.

Holz, P. 2002. Restraint and anesthesia of Dasyurids (Dasyuromorphia). Pp. #-#, in D. Heard, ed., Zoological Restraint and Anesthesia. International Veterinary Information Service, Ithaca, New York; # pp.

6/05/2016 8:41 AM 381

Holz, P.H. 2008. Dasyurids. Pp. 359-382, in L. Vogelnest and R. Woods, eds, Medicine of Australian Mammals. CSIRO Publishing, Collingwood, Victoria; 686 pp.

Holz, P.H. 2012. Tasmanian Devil Facial Tumor Disease. Pp. 386-389, in R.E. Miller and M.E. Fowler, eds, Zoo and Wild Animal Medicine, Volume 7. Elsevier, St. Louis.

Holz, P.H. and P.B. Little. 1995. Degenerative leukoencephalopathy and myelopathy in dasyurids. Journal of Wildlife Diseases 31:509-513.

Hope, J.H. 1972. Mammals of the Bass Strait islands. Proceedings of the Royal Society of Victoria 85:163-196.

Hope, J.H. 1978. Pleistocene mammal extinctions: the problem of Mungo and Menindee, New South Wales. Alcheringa 2:65-82.

Hope, J.H., R.J. Lampert, E. Edmonson, M.J. Smith and G.F. Van Tets. 1977. Late Pleistocene faunal remains from Seton rock shelter, Kangaroo Island, South Australia. Journal of Biogeography 4:363-385.

Horan, R.D. and R.T. Melstrom. 2011. No sympathy for the Devil. Journal of Environmental Economics and Management 62(3):367-385.

Horton, D. 1977. A 10,000-year-old Sarcophilus from Cape York. Search 8(10):374- 375.

Howitt, W. 1855. Land, Labour and Gold; or Two Years in Victoria: With Visits to Sydney and Van Dieman’s Land. Vol. II. Longman, Brown, Green and Longmans, London; 398 pp.

Howkins, A.B. 1966. The role of Macropodidae in Tasmania as intermediate hosts in hydatid disease. Australian Veterinary Journal 42(7):240-241.

Howson, L.J. 2011. Tumour Immune Escape Mechanisms in DFTD. Batchelor of Medical Research Honours Thesis. University of Tasmania, Hobart: # pp. [Not seen].

Howson, L.J., K.M. Morris, T. Kobayashi, C. Tovar, A. Kreiss, A.T. Papenfuss, L. Corcoran, K. Belov and G.M. Woods. 2014. Identification of dendritic cells, B cell and T cell subsets in Tasmanian Devil lymphoid tissue; evidence for poor immune cell infiltration into Devil facial tumors. Anatomical Record 297(5):925-938. Published online: March 2014.

Hughes, C., R. Gaffney and C.R. Dickman. 2011. A preliminary study assessing risk to Tasmanian devils from poisoning for red foxes. Journal of Wildlife Management 75(2):385–392.

Hughes, R.L. 1965. Comparative morphology of spermatozoa from five marsupial families. Australian Journal of Zoology 13(4):533-543.

6/05/2016 8:41 AM 382

Hughes, R.L. 1976a. Preliminary observations on reproduction and embryonic- development in marsupial devil, Sarcophilus harrisii. Journal of Reproduction and Fertility 46(2):504-505.

Hughes, R.L. 1976b. Reproduction in the male marsupial Devil Sarcophilus harrisii with particular reference to the ultrastructure of epididymal spermatozoa. Proceedings of the VIIIth International Congress of Animal Reproduction 4: 908-910.

Hughes, R.L. 1980. Embryonic loss during early pregnancy in the marsupial devil Sarcophilus harrisii. Journal of Anatomy 130:211. Abstract.

Hughes, R.L. 1982. Reproduction in the Tasmanian Devil Sarcophilus harrisii (Dasyuridae, Marsupialia). Pp. 49-63, in M. Archer, ed. Carnivorous Marsupials. Royal Zoological Society, Mosman, New South Wales; 804 pp.

Hughes, R.L. and L.S. Hall. 1988. Structural adaptations of the newborn marsupial. Pp. 8-27, in C.H. Tyndale-Biscoe and P.A. Janssens, eds, The Developing Marsupial. Models for Biomedical Research, Springer Verlag, Berlin; 245 pp.

Hughes, R.L., A.-M. Pearse, S.W. Cooper, J.M.P. Joss, P.G. Johnston and M.K. Hughes. 1993. The genetic basis of marsupial gonadogenesis and sexual phenotype. Pp. 17-48, in K.C. Reed and J.A.M. Graves. Harwood Academic Publishers, #: # pp.

Hulbert, A.J. and R.W. Rose. 1972. Does the Devil sweat? Comparative Biochemistry and Physiology 43A:219-222.

Hunt, G. and M. Groom. 2015. Healthy devils returned to the wild will give others a chance. Commonwealth and Tasmanian Governments joint media release, 18 November 2015.

Hunter, D.O., T. Britz, M. Jones and M. Letnic. 2015. Reintroduction of Tasmanian devils to mainland Australia can restore top-down control in ecosystems where dingoes have been extirpated. Biological Conservation 191:428-435.

Hunter, D. and M. Letnic. 2015. Bringing devils back to the mainland could help wildlife conservation. The Conversation, 24 August 2015.

Huxtable, S., B. Brown, S. Fox and D. Pemberton. 2015. The removal, reintroduction and reinforcement of a diseased population of Tasmanian devils. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Symposium abstract, p. 3.

Huxtable, S.J., D.V. Lee, P. Wise and the Save the Tasmanian Devil Program. 2015. Metapopulation management of an extreme disease scenario. Chapter 11, pp. 141- 154, in D. Armstrong, M. Hayward, D. Moro and P. Seddon, eds, Advances in Reintroduction Biology of Australian and New Zealand Fauna, CSIRO Publishing, Clayton South, Victoria, Australia; 320 pp.

Hyatt, A., D. Hayes, V. Olsen and D. Boyle. 2007. In search of viruses. An electron microscopical study. P. 16, in Devil Facial Tumour Disease. Senior Scientist’s

6/05/2016 8:41 AM 383

Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Isaacks, R., S. Nicol, J. Sallis, R. Zeidler and H. D. Kim. 1984. Erythrocyte phosphates and hemoglobin function in monotremes and some marsupials. American Journal of Physiology 246: R236-R241.

Izzard, L. 2010. Rickettsiales and rickettsial diseases in Australia. PhD Thesis, Murdoch University; # pp.

Izzard, L., S. Graves, E. Cox, S. Fenwick, N. Unsworth and J. Stenos. 2009. Novel rickettsia in ticks, Tasmania, Australia. Emerging Infectious Diseases 15(10):1654- 1656.

Jackson, H. 2015. No, this isn’t ‘Looney Tunes': Pair of Tasmanian devils comes to LA Zoo. MyNewsLa.com, 14 December 2015.

Jackson, S. 2003. Australian Mammals: Biology and Captive Management. CSIRO Publishing, Melbourne; # pp.

Jaffe, E. 2007. Tasmanian Tailspin. Smithsonian.com, 1 June 2007.

Jakob-Hoff, R., D. Harley, M. Magrath, M. Lancaster and G. Kuchling. 2015. Pp. 201-215, in D. Armstrong, M. Hayward, D. Moro and P. Seddon, eds, Advances in Reintroduction Biology of Australian and New Zealand Fauna, CSIRO Publishing, Clayton South, Victoria, Australia; 320 pp.

Jarvis, C. 2016. Moon Lake commits to working with community, govt on VDL.The Advocate, 14 January 2016.

Jenkins, D.J. 2006. Echinococcus granulosus in Australia, widespread and doing well! Parasitology International 55:S203 – S206.

Jenkins, J. 2014. The Devil’s details. The Scientist, 1 November 2014.

Jiang, T. and L.RG. DeSantis. 2014. Dental microwear texture analysis of the Tasmanian Devil: assessing variability among teeth. Young Scientist. A High School Research Journal May 2014. [Accessed online, 13 July 2014]

Johnson, C.N. 2006. Australia’s Mammal Extinctions. Cambridge University Press, Melbourne, Vic.; 278 pp.

Johnson, C.N. 2007. Rebalancing Australia’s ecology? Online Opinion, 2 October 2007.

Johnson, C.N. 2014. The rise and fall of large marsupial carnivores. Pp. 13-26, in A.S. Glen and C.R. Dickman, eds, Carnivores of Australia. Past, Present and Future. CSIRO Publishing, Collingwood, Victoria; 448 pp.

6/05/2016 8:41 AM 384

Johnson, C.N. and S. Wroe. 2003. Causes of extinction of vertebrates during the Holocene of mainland Australia: arrival of the dingo, or human impact? The Holocene 13(6):941-948.

Johnston, T. 2007. Cancer is killing Tasmanian devils. International Herald Tribune, 17 February 2007

Johnston, T.H. 1913. Cestoda and Acanthocephala. Report of the Australian Institute of Tropical Medicine 1911:75-96.

Johnston, T.H. and P.M. Mawson. 1940. New and known nematodes from Australian marsupials. Proceedings of the Linnean Society of New South Wales 65(5-6):468- 476.

Jones, E., Y. Cheng, D. O’Meally and K. Belov. 2014. Novel defensin peptides of the Tasmanian Devil (Sarcophilus harrisii). Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 38.

Jones, F.W. 1923. The Mammals of South Australia. British Science Guild (South Australian Branch), Adelaide; 458 pp. [Not used in text]

Jones, M. 1993. Road mortality of eastern quolls and devils on the Cradle Mountain tourist road. Report for Department of Transport and Works and Department of Environment and Land Management, Parks and Wildlife Service, August 1993.

Jones, M. 1994. Dining with the Devil. Australian Natural History 24(11):32-41.

Jones, M. 1995. Guild Structure of the Large Marsupial Carnivores in Tasmania. PhD Thesis, University of Tasmania; 143 pp.

Jones, M. 1995. Tasmanian Devil. Pp. 82-84, in R. Strahan, ed., The Mammals of Australia. Second Edition. Reed New Holland, Sydney; 756 pp.

Jones, M. 1996. Clash of the carnivores. Nature Australia 25(5):46-55.

Jones, M. 1997. Quolls and Tasmanian Devil. P. 69, in Reader’s Digest Encyclopedia of Australian Wildlife; 624 pp.

Jones, M. 1997. Character displacement in Australian dasyurid carnivores: size relationships and prey size patterns. Ecology 78(8):2569-2584.

Jones, M.E. 1998. Function of vigilance in sympatric marsupial carnivores: the eastern quoll and the Tasmanian devil. Animal Behaviour 56:1279-1284.

Jones, M.E. 2000. Road upgrade, road mortality and remedial measures: impacts on a population of eastern quolls and Tasmanian devils. Wildlife Research 27:289-296.

Jones, M.E. 2001. Large marsupial carnivores. Pp. 814-817, in D.W. MacDonald, ed., The New Encyclopedia of Mammals. Oxford University Press; 930 pp.

6/05/2016 8:41 AM 385

Jones, M. 2003a. Convergence in ecomorphology and guild structure among marsupial and placental carnivores. Pp. 281-292, in M.E. Jones, C.R. Dickman and M. Archer, eds, Predators with Pouches. The Biology of Carnivorous Marsupials. CSIRO Publishing, Collingwood, Victoria; 486 pp.

Jones, M. 2003b. Disease decimates the devil. Wilderness News 164:14-15.

Jones, M. 2006. Conservation management of wild Tasmanian devils. Proceedings of the 2006 Conference of the Australian Association of Veterinary Conservation Biologists, Hobart, Tasmania, May 2006.

Jones, M. 2007. Tasmanian Devil. P. 202, in C.R. Dickman, ed., A Fragile Balance. The Extraordinary Story of Australian Marsupials. Craftsman House, Fishermans Bend, Victoria; 246 pp.

Jones, M. 2008. Tasmanian Devil Sarcophilus harrisii (Boitard, 1841). Pp. 78-80, in S. Van Dyck and R. Strahan, eds. The Mammals of Australian. Reed Books, Sydney; 887 pp. Third Edition.

Jones, M. 2010. A novel disease threatens Tasmania’s top predator: an integrated approach to research for managing the entire ecosystem. Australian Mammal Society 56th Meeting, 4-9 July 2010. Abstract.

Jones, M. 2012. Will the Tasmanian devil follow the Tasmanian tiger to extinction? Royal Society of Tasmania Winter Series – Lecture 2, 17 July 2012. [Available online]

Jones, M. 2013. PhD project opportunities on ecological restoration of mammals and birds in Tasmania. The project: Restoring resilience in wildlife populations. Online. Accessed 13 November 2015.

Jones, M. 2015a. The ecological role of the Tasmanian Devil. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Symposium abstract, p. 10.

Jones, M. 2015b. Understanding the evolution of devils and Devil Facial Tumour Disease (DFTD). Devil’s Advocate Newsletter, December 2015.

Jones, M., M. Oakwood, C.A. Belcher, K. Morris, A.J. Murray, P.A. Woolley, K.B. Firestone, B. Johnson and S. Burnett. 2003. Carnivore concerns: problems, issues and solutions for conserving Australasia’s marsupial carnivores. Pp. 422-434, in J. Jones, C. Dickman and M. Archer, eds, Predators with Pouches. The Biology of Carnivorous Marsupials. CSIRO Publishing, Collingwood, Victoria; 486 pp.

Jones, M. and L.A. Barmuta. 1998. Diet overlap and abundance of sympatric dasyurid carnivores: a hypothesis of competition? Journal of Animal Ecology 67:410-421.

Jones, M.E. and L.A. Barmuta. 2000. Niche differentiation among sympatric Australian dasyurid carnivores. Journal of Mammalogy 81(2):434-447.

6/05/2016 8:41 AM 386

Jones, M.E, S. Burnett, A. Claridge, B. Fancourt, G. Körtner, K. Morris, D. Peacock, S. Troy and J. Woinarski. 2014. Australia’s surviving marsupial carnivores: threats and conservation. Pp. 197-240, in A.S. Glen and C.R. Dickman, eds, Carnivores of Australia. Past, Present and Future. CSIRO Publishing, Collingwood, Victoria; 448 pp.

Jones, M.E., A. Cockburn, R. Hamede, C. Hawkins, H. Hesterman, S. Lachish, D. Mann, H. McCallum and D. Pemberton. 2008. Life-history change in disease-ravaged Tasmanian devil populations. Proceedings of the National Academy of Sciences 105(29):10023-10027.

Jones, M. E., G. Grigg and L. Beard. 1997. Body temperatures of Tasmanian devils (Sarcophilus harrisii) and eastern quolls (Dasyurus viverrinus) through an alpine winter. Physiological Zoology 70(1):53-60.

Jones, M., P. Jarman, C. Lees, H. Hesterman, R. Hamede, N. Mooney, D. Mann, C. Pukk, J. Bergfeld, and H. McCallum. 2007a. Conservation management of Tasmanian devils in the context of an emerging, extinction-threatening disease: Devil Facial Tumor Disease. EcoHealth 4(3):326-337.

Jones, M., R. Hamede and H. McCallum. 2012. The devil is in the detail: conservation, biology, animal philosophies and the role of animal ethics committees. Pp. 79-88, in P. Banks, D. Lunney and C. Dickman, eds, Science Under Siege: Zoology Under Threat. Royal Zoological Society of New South Wales, Mosman, New South Wales; # pp.

Jones, M.E and H. McCallum. 2007. Environmental Risk Assessment - Impact of the introduction of Tasmanian devils to Maria Island on the natural values of the island. DPIPWE Internal Document. This is Appendix 3 of STDP, 2012a.

Jones, M. and H. McCallum. 2011a. Natural values (fauna) summary report – Maria Island National Park. Tasmanian National Parks and Wildlife Service, November 2011. This document consists of Appendices 4a and b of STDP, 2012a

Jones, M. and H. McCallum. 2011b. The Devil’s Cancer. Scientific American 304(6):54-59.

Jones, M., J. Moore, T. Rout, C. Wilcox, Y. Ben-Yaim and H. Possingham. 2007b. Which is the best strategy for establishing wild insurance populations given risks of disease invasion, costs of eradication, quarantine and surveillance? P. 7, in B. Wintle, Coordinator, Optimal and robust monitoring for conservation and natural resource management. An AEDA workshop report. 6th August - 10th August 2007. Freycinet Lodge, Freycinet National Park, Tasmania; 15 pp.

Jones, M.E., M. Oakwood, C. Belcher, K. Morris, A. Murray, P.A. Woolley, K.B. Firestone, B. Johnson and S. Burnett. 2003a. Carnivore concerns: problems, issues and solutions for conserving Australasia’s marsupial carnivores. Pp. 418-430, in M. Jones, C. Dickman and M. Archer, eds, Predators with Pouches. The Biology of Carnivorous Marsupials. CSIRO Publishing, Melbourne; 486 pp.

6/05/2016 8:41 AM 387

Jones, M.E., D. Paetkau, E. Geffen and C. Moritz. 2003b. Microsatellites for the Tasmanian Devil (Sarcophilus laniarius). Molecular Ecology Notes 3:277-279.

Jones, M.E., D. Paetkau, E. Geffen and C. Moritz. 2004. Genetic diversity and population structure of Tasmanian devils, the largest marsupial carnivore. Molecular Ecology 13:2197-2209. [This is Jones et al., 2004a]

Jones, M., D. Pemberton and R.K. Rose. 20##. Sarcophilus harrisii. Mammalian Species. In press?

Jones, M., C. Pukk, D. Mann and H. McCallum. 2007. The Tasman disease suppression trial. P. 28, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Jones, M. and R.K. Rose 1996. Preliminary assessment of distribution and habitat associations of the spotted-tailed quoll (Dasyurus maculatus maculatus) and eastern quoll (D. viverrinus) in Tasmania to determine conservation and reservation status. Report to the Tasmanian RFA (Regional Forest Agreement). Environment and Heritage Committee, November, 1996. Public Land Use Commission, Hobart.

Jones, M.E, G.C. Smith and S.M. Jones. 2004b. Is anti-predator behaviour in Tasmanian eastern quolls (Dasyurus viverrinus) effective against introduced predators? Animal Conservation 7(2):155-160. [This is Jones et al., 2004b]

Jones, M.E. and D.M. Stoddart. 1998. Reconstruction of the predatory behaviour of the extinct marsupial thylacine. Journal of Zoology 246:239-246.

Jones, S.M., T.J. Lockhart and R.W. Rose. 2005. Adaptation of wild-caught Tasmanian Devils (Sarcophilus harrisii) to captivity: evidence from physical and plasma cortisol concentrations. Australian Journal of Zoology 53:339-344.

Joyce, N. 2011. Little devils delight. Sunshine Coast Daily, 19 August 2011.

Louys, J. R.T. Corlett, G.J. Price, S. Hawkins and P.J. Piper. 2014. Rewilding the tropics, and other conservation translocations strategies in the tropical Asia-Pacific region. Ecology and Evolution 4(22): 4380-4398.

Kabat, A.P., R.W. Rose and A.K. West. 2003. Non-shivering thermogenesis in a carnivorous marsupial, Sarcophilus harrisii, in the absence of UCP1. Journal of Thermal Biology 28:413-420.

Kabat, A.P., R.W. Rose and A.K. West. 2004. Molecular identification of uncoupling proteins 2 and 3 in a carnivorous marsupial, the Tasmanian Devil (Sarcophilus harrisii). Physiological and Biochemical Zoology 77(1):109-115.

Katsineris, S. 2015. Tasmanian Devil roadkill fatalities. Guardian (Sydney) No 1674, 25 Feb 2015.

6/05/2016 8:41 AM 388

Keble, R.A. 1946. The Sunklands of Port Philip Bay and Tasmania. Memoirs of the National Museum of Victoria 14(2):69-122.

Keeley, T. and T. Eastley. 2012. Evaluation of the reproductive potential of 1-year- old female Tasmanian devils and their prospective contribution to the captive breeing program. Program, 17th International Congress on Animal Reproduction. Vancouver Convention Centre, Vancouver, British Columbia, Canada; July 29 – August 2, 2012; 64 pp [Abstract]

Keeley, T., M. Harris, P.D. McGreevy, D. Hudson and J.K. O’Brien. 2012a. Development and evaluation of electroejaculation techniques in the Tasmanian devil (Sarcophilus harrisii). Reproduction, Fertility and Development 24(7):1008-1018.

Keeley, T., P.D. McGreevy and J.K. O’Brien. 2010. Reproductive and adrenal hormones and their relationships to successful breeding in captive Tasmanian Devils. Australian Mammal Society 56th Meeting, 4-9 July 2010. Abstract.

Keeley, T., P.D. McGreevy and J.K. O’Brien. 2011. Characterization and short-term storage of Tasmanian Devil sperm collected post-mortem. Theriogenology 76(4):705- 714.

Keeley, T., P.D. McGreevy and J.K. O’Brien. 2012b. The effects of season and devil facial tumour disease on the reproductive physiology of the male Tasmanian devil (Sarcophilus harrisii). Reproduction, Fertility and Development 24(7): 999-1007.

Keeley, T., P.D. McGreevy and J.K. O’Brien. 2012c. Cryopreservation of epididymal sperm collected postmortem in the Tasmanian devil (Sarcophilus harrisii). Theriogenology 78(2):315-325.

Keeley, T. J.K. O’Brien, B.G. Fanson, K. Masters and P.D. McGreevy. 2012d. The reproductive cycle of the Tasmanian Devil (Sarcophilus harrisii) and factors associated with reproductive success in captivity. General and Comparative Endocrinology 176:182-191.

Keene, N. 2011. Having a Devil of a time. The Daily Telegraph, 19 January 2011.

Keim, B. 2011. Tasmanian Devils might survive cancer scourge. Wired Science, 25 October 2011.

Kelly, A. 1993. Management of Tasmanian Devils (Sarcophilus harrisii) at the Tasmanian Wildlife Park. Pp. 79-84, in M. Roberts, J. Carnio, G. Crawshaw and M. Hutchins, eds, The Biology and Management of Australian Carnivorous Marsupials. Metropolitan Toronto Zoo and AAZPA, Ontario, Canada; # pp.

Kelly, A. 2006. Husbandry guidelines for Tasmanian Devil, Sacrophilus harrisii. Australasian Regional Association of Zoological Parks and Aquaria, Mosman, NSW; 36 pp. Not seen. Kelly, M. 2011a. 30 lucky Devils first to board conservation ark. Newcastle Herald, 18 January 2011.

6/05/2016 8:41 AM 389

Kelly, M. 2011b. Top time for randy Devils. Newcastle Herald, 22 September 2011.

Kelly, M. 2011c. Santa might bring a devil to Hunter homes. Newcastle Herald, 10 December 2011.

Kelly, M. 2013a. Gallery: Devilishly cute babies. Southern Cross, 15 August 2013.

Kelly, M. 2013b. Tassie Devils entertain kids in hospital. Newcastle Herald, 17 December 2013.

Kelly, M. 2015. Devils bred in the Barrington Tops to be returned to Tasmania.

Maitland Mercury and Newcastle Herald, M. Kelley, 17 November 2015.

Kempton, H. 2011. Devil sperm bank created. Hobart Mercury, 23 November 2011.

Kempton, H. 2012a. Wild devils say hello. Hobart Mercury, 28 February 2012.

Kempton, H. 2012b. Devil-proof fence plan floated. Hobart Mercury, 27 November 2012.

Kempton, H. 2013a. Hopes high for devil history. Hobart Mercury, 3 April 2013.

Kempton, H. 2013b. Boost for devil rescue fund. Hobart Mercury, 25 April 2013.

Kempton, H. 2013c. Danes' baby devil delight. Hobart Mercury, 11 June 2013.

Keene, N. 2012. Princess Mary is a new breed of devil. The Daily Telegraph, 29 November 2012.

Kershaw, J.A. 1912. The Tasmanian Devil in Victoria. Victorian Naturalist 29(5):75- 76.

Kilbane, K. 2015. Memorable animals part of 50 years of fun at zoo. News~Sentinel, 16 April 2015.

Kilbane, K. 2016. UPDATED: Renovated Australian Outback ready when Fort Wayne Children's Zoo opens Saturday. News~Sentinel, 19 April 2016

Killick, D. 2008. Devil success but scientists plead for cash. Hobart Mercury, 3 June 2008.

King, J. and R. Best. 1996. Husbandry guidelines for Tasmanian Devil Sarcophilus harrisii at Taronga Zoo. Taronga Zoo, Sydney, Australia. Not seen.

Kingston, E., S. Izzard, A. Smith and C. O’Halloran. 2014. Observations of behavioural changes in wild-caught Tasmanian Devils brought into intensive captive management as part of the Forestier Peninsula depopulation project. Department of Primary Industries, Parks, Water and Environment, Prospect, Tasmania. [According to the Document Properties, created on 17 June 2014.

6/05/2016 8:41 AM 390

Kinnear, J.E. and G.D. Brown. 1967. Minimum heart rates of marsupials. Nature 215:1501.

Kirsch, J.A.W. and M. Archer. 1982. Polyphyletic cladistics, or, when parsimony’s not enough: the relationships of carnivorous marsupials. Pp. 595-619 pp, in M. Archer, ed., Carnivorous Marsupials, Royal Zoological Society of New South Wales, Mosman, New South Wales; 804 pp. Sarcophilus clusters consistently with Thylacinus.

Klaer, N.L. 2010. Remote camera observations of Tasmanian devils and other mammals at Franklin. Tasmanian Field Naturalist 132:81-84.

Knadler, C.E. 2012. An agent-based model of the spread of Devil Facial Tumor Disease in an isolated population of Tasmanian Devils. Agent Technologies and Systems 4(4):1-16.

Koller, P.C. 1936. The genetical and mechanical properties of the sex chromosomes. II. Marsupials. Journal of Genetics 32(3):451-472.

Kolkert, H. 2005. Devil Facial Tumour Disease (DFTD): Links with Endocrinology and Reproductive Function in the Tasmanian Devil Sarcophilus harrisii. B.Sc. (Hons) Thesis, University of Tasmania; 162 pp. Not seen.

Koncza, E. and W. Pryor. 2001. Regional studbook. Tasmanian Devil Sarcophilus harrisii. Zoological Society of Florida, Miami, Florida; 101 pp.

Krajewski, C., A.C. Driskell, P.B. Baverstock and M.J. Braun. 1992. Phylogenetic relationships of the Thylacines (Mammalia: Thylacinidae) among dasyuroid marsupials: evidence from cytochrome b DNA sequences. Proceedings of the Royal Society of London B 250:19-27.

Krajewski, C., J. Young; L. Buckley, P.A. Woolley and M. Westerman. 1997. Reconstructing the evolutionary radiation of dasyurine marsupials with cytochrome b, 12S rRNA, and protamine gene tress. Journal of Mammalian Evolution 4:217-236.

Krajewski, C., G.R. Moyer, J.T. Sipiorski, M.G. Fain and M. Westerman. 2004. Molecular systematics of the enigmatic 'phascolosoricine' marsupials of New Guinea. Australian Journal of Zoology 52(4):389-415.

Krajewski, C., J. Painter, L. Buckley and M. Westerman. 1994. Phylogenetic structure of the marsupial family Dasyuridae based on cytochrome b DNA sequences. Journal of Mammalian Evolution 2(1):25-35.

Krajewski, C., J. Painter, A.C. Driskell, L. Buckley and M. Westerman. 1993. Molecular systematics of New Guinean dasyurids (Marsupalia: Dasyuridae). Science in New Guinea 19(3):157-166.

6/05/2016 8:41 AM 391

Krajewski, C. and M. Westerman. 2003. Molecular systematics of Dasyuromorphia. Pp. 3-20, in M. Jones, C. Dickman and M. Archer, eds, Predators with Pouches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melbourne; 486 pp.

Krajewski, C., P.A. Woolley and M. Westerman. 2000. The evolution of reproductive strategies in dasyurid marsupials: implications of molecular phylogeny. Biological Journal of the Linnean Society 71(3):417-435.

Krajewski, C., S. Wroe and M Westerman. 2000. Molecular evidence for the pattern and timing of cladogenesis in dasyurid marsupials. Zoological Journal of the Linnean Society 130(3):375–404.

Kreiss, A. 2009. The immune responses of the Tasmanian devil and the Devil Facial Tumour Disease. PhD Thesis, University of Tasmania, Hobart. But not made generally available (online) until March/April 2015.

Kreiss, A. 2012. An update on devil facial tumour disease in Tasmanian devils. Online, 22 November 2012. [Document properties indicate document was created on 2 March 2012.]

Kreiss, A., G.K. Brown, C. Tovar, A.B. Lyons and G.M. Woods. 2015. Evidence for induction of humoral and cytotoxic immune responses against devil facial tumor disease cells in Tasmanian devils (Sarcophilus harrisii) immunized with killed cell preparations. Vaccine 33(26):3016-3025. First published online, 20 February 2015.

Kreiss, A., Y. Cheng, F. Kinble, B. Wells, S. Donovan, K. Belov and G.M. Woods. 2011b. Allorecognition in the Tasmanian Devil (Sarcophilus harrisii), an endangered marsupial species with limited genetic diversity. PLoS ONE 6(7): e22402. doi:10.1371/journal.pone.0022402. (Published online, 21 July 2011)

Kreiss, A., N. Fox, J. Bergfeld, S.J. Quinn, S. Pyecroft and G.M. Woods. 2008. Assessment of cellular immune responses of healthy and diseased Tasmanian devils (Sarcophilus harrisii). Developmental and Comparative Immunology 32:544-553.

Kreiss, A., D.L. Obendorf, S. Hemsley, P.J. Canfield and G.M. Woods. 2009a. A histological and immunohistochemical analysis of lymphoid tissues of the Tasmanian Devil. The Anatomical Record: Advances in Integrative Anatomy and Evolutionary Biology 292(5):611-620.

Kreiss A., C. Tovar, D.L. Obendorf, K. Dun and G.M. Woods. 2011a. A murine xenograft model for a transmissible cancer in Tasmanian devils. Veterinary Pathology 48(2):475-481. (Published online, 22 September 2010).

Kreiss, A., B. Wells and G.M. Woods. 2009b. The humoral immune response of the Tasmanian devil (Sarcophilus harrisii) against horse red blood cells. Veterinary Immunology and Immunopathology 130(1-2):135-137.

Kreiss, A. and G.M. Woods. 2007. The immune responses of the Tasmanian devil (Sarcophilus harrisii) and the Devil Facial Tumour Disease, Proceedings of the 8th

6/05/2016 8:41 AM 392

International Veterinary Immunology Symposium, 15-19 August, Ouro Preto, Brazil. [Conference Extract; not seen]

Krefft, G. 1865. On the Vertebrata of the Lower Murray and Darling. Read before the Philosophical Society of New South Wales, 10 September 1862.

Krefft, G. 1868. Notes on the fauna of Tasmania. Monthly Notices of Papers and Proceedings of the Royal Society of Tasmania 1868: 90-105.

Kühme, W. 1983.Beutelteufel (Sarcophilus harrisii) im Kölner Zoo. Zeitschrift des Kölner Zoo 26(3): 79-83.

Ladds, P.2009. Pathology of Australian Native Wildlife. CSIRO Publishing, Collingwood, Victoria; 640 pp.

Ladds, P., R. Loh, M. Jones and P. Tucker. 2003. Probable lymphosarcoma in the Tasmanian Devil (Sarcophilus harrisii). Proceedings of the Australian Society for Veterinary Pathology Annual Meeting, Abstracts of the Scientific Sessions.

Lachish, S. 2007. The impact of Devil facial tumour disease on the demography and population dynamics of Tasmanian devils. P. 22, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Lachish, S., M. Jones and H. McCallum. 2007. The impact of devil facial tumour disease on the survival and population growth rate of the Tasmanian devil. Journal of Animal Ecology 76(5):926–936.

Lachish, S., H. McCallum and M. Jones. 2009. Demography, disease and the devil: life-history changes in a disease-affected population of Tasmanian devils (Sarcophilus harrisii). Journal of Animal Ecology 78(2):427–436.

Lachish, S., H. McCallum, D. Mann, C.E. Pukk and M.E. Jones. 2010. Evaluation of selective culling of infected individuals to control Tasmanian Devil Facial Tumor Disease. Conservation Biology 24(3):841-851.

Lachish, S., K. J. Miller, A. Storfer, A.W. Goldizen and M.E. Jones. 2011a. Evidence that disease-induced population decline changes genetic structure and alters dispersal patterns in the Tasmanian devil. Heredity 106:172-182.

Lachish S., A. Passmore and M. E. Jones. 2011b. A new PCR assay for reliable molecular sexing of endangered Tasmanian devils (Sarcophilus harrisii) from non- invasive genetic samples. Conservation Genetics Resources 3:279-281.

Ladds, P., R. Loh, M.E. Jones and P. Tucker. 2003. Probable lymphosarcoma in the Tasmanian devil (Sarcophilus harrisii). Australian Society for Veterinary Pathology Annual Conference Proceedings, Elizabeth Macarthur Agricultural Institute, Menangle, New South Wales, Australia, 12-13 April, 2003.

Lakin, A. 2011. Jan Cameron. The Examiner, 17 July 2011.

6/05/2016 8:41 AM 393

Lane, A., Y. Cheng, B. Wright, R. Hamede, L. Levan, M. Jones, B. Ujvari and K. Belov. 2012. New insights into the role of MHC diversity in devil facial tumour disease. PLoS ONE 7(6): e36955. doi:10.1371/journal.pone.0036955.

Lang, C.J., A.D. Postle, S. Orgeig, F. Possmayer, W. Bernhard, A.K. Panda, K.D. Jürgens, W.K. Milsom, K. Nag and C.R. Daniels. 2005. Dipalmitoylphosphatidylcholine is not the major surfactant phospholipid species in all mammals. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 289: R 1426-R1439.

Lang, E.M. 1958 (1959). Einige seltenere Zuchterfolge im Basler Zoo. Zoologische Garten, Leipzig 24(3/4):282-283.

Larkins, D. 2012. The secret love-life of Tasmanian devils. ABC Local, 14 February 2012.

Laurent, M.M., Bazin (de Bassenville), Hollard, Coste and P. Gervais. 1839. Annales françaises and etrangères d’Anatomie et de Physiologie appliquées à la Médicine et à l’Histoire naturelle. Volume 3. Pitois-Levrault et Cie, Paris; 384 pp.

Law, B. 2014. The Devils of Maria Island. Brochure distributed with the Saturday Newspaper newspaper on 22 March 2014.

Lawes, A. 2010. She plays the devils' advocate. Sydney Morning Herald, 9 October 2010.

Lawrence, C. and C. Donnelly. 2011. Save the Tasmanian Devil Roadkill Project. Save the Tasmanian Devil Program; available online.

Lazell, J.D., Jr. 1984. A population of Devils Sarcophilus harrisii in north-western Tasmania. Records of the Queen Victoria Museum No. 84: 53-56.

Lazenby, B.T. 2012. Do feral cats affect small mammals? A case study from the forests of southern Tasmania. PhD Thesis, University of Sydney, New South Wales, Australia.

Lazenby, B.T. and C. Dickman. 2013. Patterns of detection and capture are associated with cohabiting predators and prey. PLoS ONE 8(4): e59846. doi:10.1371/journal.pone.0059846

Lazenby, B., D. Pemberton, S. Fox and the Save the Devil Program. 2015a. Status and trends in wild devils in the aftermath of Devil Facial Tumour Disease. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Symposium abstract, p. 11.

Lazenby, B.T., N.J. Mooney and C.R. Dickman. 2015. Detecting species interactions using remote cameras: effects on small mammals of predators, conspecifics, and climate. Ecosphere 6(12), Article 266; 18 pp. Published online, 15 December 2015.

6/05/2016 8:41 AM 394

Lees, C., ed. 2005. ASMP/DPIW Captive Management plan for Tasmanian Devils, Sarcophilus harrisii. Australasian Regional Association of Zoological Parks and Aquaria, Mosman, Australia.

Lees, C. 2007. Managed breeding. P. 31, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Lees, C. and P. Andrew. 2012. The Tasmanian Devil Insurance Meta-population: 2012 Evaluation and Review. IUCN/SSC Conservation Breeding Specialist Group. Apple Valley, Minnesota; 66 pp. [Dated: 17 April 2012]

Le Souef, A.S. and H. Burrell. 1926. The Wild Animals of Australasia. Harrap and Co., London; 388 pp.

Letnic, M., M. Fillios and M.S. Crowther. The arrival and impact of the dingo. Pp. 53- 67-26, in A.S. Glen and C.R. Dickman, eds, Carnivores of Australia. Past, Present and Future. CSIRO Publishing, Collingwood, Victoria; 448 pp.

Leung, C.C. 2007. Extinction warning for Tassie devils, 4 October 2007.

Lewis, D. 1988. The Tasmanian Devil in Arnhem Land rock art. Australian Archaeology 27:121-123.

Lewis-Hasteley, H. 2011. Sympathy for the devils. New Statesman 140 (5074):14.

Lightoller, G.S. 1940. The comparative morphology of the platysma. A comparative study of the sphincter colli profundus and the trachelo-platysma. Journal of Anatomy 74(3):390-396.

Lim, K., T. Keeley and S. Hemsley. 2010. Preliminary investigation of spermatogenesis in the Tasmanian devil (Sarcophilus harrisii). Pp. 29-30, in Wildlife Disease Association (Australasia) Conference, Dover, Tasmania. Wildlife Disease Association, Hobart.

Lindert-Wentzell, A. 2012. Extinction showdown. Business in Focus Magazine, 25 September 2012.

Lockhart, T.J. 2000. Effects of captivity on the reproduction of the Tasmanian devil, Sarcophilus harrisii. Thesis, University of Tasmania; 256 pp.

Loh, R. 2003. Tasmanian Devil (Sarcophilus harrisii) facial tumour (DFT). Paper prepared for the Devil Facial Tumour Disease Workshop, Sir Raymond Ferrall Centre, University of Tasmania, 14 October 2003, p. 2.

Loh, R. 2004. Devil facial tumour disease: devastating the Tasmanian icon. Pp. 126- 127, in R. Woods, ed., Proceedings of the 2004 Conference of the Australian Association of Veterinary Conservation Biologists. Australian Association of Veterinary Conservation Biologists, Canberra.

6/05/2016 8:41 AM 395

Loh, R. 2006. The Pathology of Devil Facial Tumour Disease in Tasmanian Devils (Sarcophilus harrisii). Masters Thesis, Murdoch University; 173 pp.

Loh, R., D. Hayes, AA. Mahjoor, M. O’Hara, S.B. Pyecroft and S. Raidal. 2005. Devil facial tumour disease (DFTD) in Tasmanian Devils (Sarcophilus harrisii) – The gross, histological, ultrastructural and immunohistochemical characteristics of the neoplasm. Proceedings of the Wildlife Disease Association International Conference, Cairns; p. 266. Poster.

Loh, R., J. Bergfeld, D. Hayes, A. O’Hara, S. Pyecroft, S. Raidal and R. Sharpe. 2006a. The pathology of devil facial tumor disease (DFTD) in Tasmanian devils (Sarcophilus harrisii). Veterinary Pathology 43:890-895.

Loh, R., D. Hayes, D. Mahjoor, A. O’Hara, S. Pyecroft and S. Raidal. 2006b. The immunohistochemical characterization of devil facial tumor disease (DFTD) in the Tasmanian devil (Sarcophilus harrisii). Veterinary Pathology 43:896-903.

Loh, R., B.D. Hayes, A. Mahjoor, A. O’Hara, S. Pycroft and S. Raidal. 2007. The pathology of devil facial tumour disease in Tasmanian devils. P. 8, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Loh, R., M. O’Hara, S. Raidal, S. Pyecroft and R. Sharpe. 2004. Devil facial tumour disease. What is happening to our Devils?

Lohberger, L. 2015a. Tasmanian devils in groundbreaking vaccine trial killed by motorists. Hobart Mercury, 30 September 2015.

Lohberger, 2015b. Tasmanian researchers to track remaining devils in Narawntapu National Park after road deaths. Hobart Mercury, 2 October 2015.

Lombard, L.S. and E.J. Witte. 1959. Frequency and types of tumours in mammals and birds at the Philadelphia Zoological Garden. Cancer Research 19:127-141.

Longman, H.A. 1921. Abstract of proceedings, 31st October, 1921. Proceedings of Royal Society of Queensland 33: xiv.

Longman, H. 1925. Fossil marsupials from Marmor. Memoirs of the Queensland Museum 8(2):109-110.

Longman, H. 1945. Fossil vertebrates from Gore Quarries. Memoirs of the Queensland Museum 12(3):164.

Lord, C.E. 1919a. Notes on the mammals of Tasmania. Papers and Proceedings of the Royal Society of Tasmania 1918:16-52.

Lord, C.E. 1919b. A trip to the National Park of Tasmania at Mount Field. Emu 19:98-106.

6/05/2016 8:41 AM 396

Lord, C.E. 1928. Existing Tasmanian marsupials. Papers and Proceedings of the Royal Society of Tasmania 1927:17-24.

Lord, C.E. and H.H. Scott. 1924. A Synopsis of the Vertebrate Animals of Tasmania. Oldham, Beddome and Meredith, London; 340 pp.

Lord, E. 2012a. Devil cancer invincible. Hobart Mercury, 31 August 2012

Lord, E. 2012b. Cooking up a future for devils. Hobart Mercury, 13 September 2012.

Lowry, D.C. and J.W.J. Lowry. 1966. Discovery of a Thylacine (Tasmanian Tiger) in a cave near Eucla, Western Australia. Helictite 5:25-29.

Lucas, A.H.S. and W.H.D. Le Souef. 1909. The Animals of Australia. Mammals, Reptiles and Amphibians. Whitcombe and Tombs Ltd, Melbourne; 327 pp.

Luckett, W.P. 1993. An ontogenetic assessment of dental homologies in therian mammals. Pp. 182-204 [284?], in F.S. Szalay, M.J. Novacek and M.C. McKenna, eds, Mammal Phylogeny; Mesozoic Differentiation, Multituberculates, Monotremes, Early Therians and Marsupials, Springer, New York; # pp.

Lundelius, E.L. 1960. Post Pleistocene faunal succession in Western Australia and its climatic interpretation. Proceedings of the 21st International Geological Congress 21(4):142-153.

Lundelius, E.L. 1963. Vertebrate remains from the Nullarbor Caves, Western Australia. Journal and Proceedings of the Royal Society of Western Australia 46(3):75-80.

Lundelius, E.L. 1966. Marsupial carnivore dens in Australian caves. Studies in Speleology 1(4):174-180.

Lundelius, E.L. and W.D. Turnbull. 1978. The mammalian fauna of Madura Cave, Western Australia. Part III. Fieldiana. Geology 38; 120 pp.

Lunney, D., M. Jones and H. McCallum. 2008. Lessons from the looming extinction of the Tasmanian Devil. Pacific Conservation Biology 14(3):151-153.

Luntz, S. 2008. Hope for Tassie devils. Australasian Science 29(6):6.

Luntz, S. 2011. Culling won’t save devil. Australasian Science, 15 December 2011.

Lydekker, R. 1887. Catalogue of Fossil Mammalia in the British Museum (Natural History). Part V. Marsupialia. Taylor and Francis, London; 345 pp.

6/05/2016 8:41 AM 397

Lyne, A.G. 1959. The systematic and adaptive significance of the vibrissae in the Marsupialia. Proceedings of the Zoological Society of London 133(1):79-133.

Maake, C. 2011. Being the advocatus diaboli: Photodynamic therapy to combat Tasmanian devil facial tumors. Photodiagnosis and Photodynamic Therapy 8(2):139.

Macalister, A. 1870. On the myology of the Wombat and the Tasmanian Devil. Annals of Natural History (4)5:153-173.

Macalister, A. 1872. Further observations on the myology of Sarcophilus ursinus. Annals of Natural History (4)10:17-20.

Macey, R. 2008. Little Devils may be saviours yet. Sydney Morning Herald 23 July 2008, p. 12.

Macintosh, N.W.G., K.N. Smith and A.B. Bailey. 1970. Lake Nitchie skeleton – unique aboriginal burial. Archaeology & Physical Anthropology in Oceania 5(2):85- 101.

Macintosh, N.W.G. 1971. Analysis of an Aboriginal skeleton and a pierced tooth necklace from Lake Nitchie, Australia. Anthroplogie 9(1):49-62.

Mackenzie, W.C. 1916. The shape and peritoneal relationships of the spleen in Monotremes and Marsupials. Journal of Anatomy 51(1):1-18.

Mackenzie, W.C. 1917. Further studies on the peritoneum and intestinal tract in Monotremes and Marsupials. Journal of Anatomy 51(3):278-292.

Mackenzie, W.C. 1918. The Liver, Spleen, Pancreas, Peritoneal Relations and Biliary System in Monotremes and Marsupials. Critchley Parker Pty Ltd, Australia; 157 pp.

Mackenzie, W.C. 1921. The Action of Muscles. P.B. Hoeber, New York; 267 pp.

MacKenzie, W.C. and W.J. Owen. 1919. The genito-urinary system in monotremes and marsupials. Pp. #-#, in The Comparative Anatomy of Australian Mammals. Part IV. Jenkin, Buxton and Co., Melbourne; # pp.

MacMillen, R.E. and J.E. Nelson. 1969. Bioenergetics and body size in dasyurid marsupials. American Journal of Physiology 217(4):1246-1251.

Macdonald, D. 1927. Nature notes and queries. The Argus, 25 February 1927.

Madsen, O. 2009. Marsupials (Metatheria). Pp. 459-461, in S.B. Hedges and S. Kumar, eds, The Timetree of Life. Oxford University Press, Oxford; 551 pp.

Mahony, D.J. 1912. On the bones of the Tasmanian Devil and other animals, associated with human remains near Warrnambool: with a note on the dune sand. Victorian Naturalist 29(3):43-46.

6/05/2016 8:41 AM 398

Macintosh, A. 2012. Approval for a devil of an issue in the Tarkine. Crickey, 19 December 2012

Maloney, M. 2011. Devil's advocate. The Examiner, 11 September 2011.

Mann, D. 2004. Documenting demographic changes in Tasmanian devil populations associated with the arrival of Devil Facial Tumour Disease. B.Sc. Honours Thesis, University of Tasmania.

Mann, D. 2007. Standard Operating Procedure for Trapping and Handling Wild Tasmanian Devils. Tasmanian Government STORS Archive, online; 2 pp.

Mann, D. 2015. Whose den is that? Identification and management of threatened mammals in the Tasmanian forestry landscape: a case study. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 21

Mallick, S. 2003. Translocation of Tasmanian devils Sarcophilus harrisii (now laniarius) to Tasmania’s offshore islands – a proposed measure to quarantine an infection-free population from Devil Facial Tumour (DFT) Disease: Selection of potential islands and preliminary cost/benefit assessment. Department of Primary Industries, Parks, Water and Environment Internal Document.

Marks, C.A., D. Obendorf, F. Pereira, I. Edwards and G.P. Hall. 2014a. The dispersion and detection patterns of mtDNA-assigned red fox Vulpes vulpes scats in Tasmania are anomalous. Journal of Applied Ecology: DOI: 10.1111/1365- 2664.12278 [first published online, 6 June 2014].

Marks, C.A., D. Obendorf, F. Pereira, I. Edwards and G.P. Hall. 2014b. Opportunistically acquired evidence is unsuitable data to model fox (Vulpes vulpes) distribution in Tasmania. Wildlife Society Bulletin: DOI: 10.1002/wsb.448 [first published online: 24 June 2014].

Marks, K. 2007. Devil of a problem. The Age, 21 October 2007.

Marks, K. 2008. Cedric is only hope as disease hits Tasmanian devil. The Independent, 1 April 2008.

Markus, N. 1996. Methodology of enrichment and applications on Tasmanian devils. Thylacinus 21(3):3-8.

Marris, E. 2015. Can Tasmanian Devils Beat Cancer of the Face? National Geographic, 17 December 2015. Published online.

Marrow, M. 2013. Tasmanian Devils fly to their new home. Stawell Times-News, 8 November 2013.

6/05/2016 8:41 AM 399

Marshall, B. and R. Cosgrove. 1990. Tasmanian Devil (Sarcophilus harrisii) scat- bone: signature criteria and archaeological implications. Archaeology in Oceania 25(3):102-113.

Marshall, L.G. 1973. Fossil vertebrate faunas from the Lake Victoria region of southwestern New South Wales. Memoirs of the National Museum of Victoria 34:151-171.

Marshall, L.G. 1974. Late Pleistocene mammals from the “Keilor Cranium site,” southern Victoria, Australia. Memoirs of the National Museum of Victoria 35:63-86.

Marshall, L.G. and R.S. Corruccini. 1978. Variability, evolutionary rates and allometry in dwarfing lineages. Paleobiology 4(2):101-119.

Marszalek, J. 2007. Tasmanian Devils welcomed in rare birth. Brisbane Times, 5 April 2007.

Martain, T. 2003. Island colony idea debated. Hobart Mercury, 2 October 2003.

Martain, T. 2004. Research boost by fellow battler. Hobart Mercury, 17 February 2004.

Martain, T. 2008. Inching back from the abyss. Hobart Mercury, 23 August 2008.

Martain, T. 2010. Drug lifts hope for devil cure. Hobart Mercury, 21 June 2010.

Martin, H. 2012a. Devil island to the rescue. Hobart Mercury, 4 March 2012.

Martin, H. 2012b. Devils board 'Noah's Ark.' Hobart Mercury, 14 November 2012.

Martin, H. 2012c. Operation devil hop. Hobart Mercury, 15 November 2012.

Martin, P.G. and D.L. Hayman. 1967. Quantitative comparisons between the karyotypes of Australian marsupials from three different superfamilies. Chromosoma 20(3):290-310.

Marvanek, S. 2007. Application of GIS to visualising DFTD distribution. P. 21, Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Mather, A. 2006. Breeding breakthrough. Hobart Mercury, 22 May 2006.

Mather, A. 2008a. Tassie Devil disease mutating. The Advertiser 25 July 2008.

Mather, A. 2008b. Massive devil setback. Hobart Mercury, 26 July 2008.

Mather, A. 2010. Genetic clue to saving species. Hobart Mercury, 11 March 2010.

Mather, A. 2014a. Fight for Tasmanian devils as critical breeding program faces financial crisis. Hobart Mercury, 31 July 2014.

6/05/2016 8:41 AM 400

Mather, A. 2014b. Good news at last as Tassie’s little devils come back from the brink. Hobart Mercury, 23 November 2014.

Mather, A. 2016a. Virtual barrier along roads to help protect Tasmanian devils. Sunday Tasmanian, 10 January 2016.

Mather, A. 2016b. West Coast mining company MMG digs deep for Tasmanian devil sound barrier. Sunday Tasmanian, 14 February 2016.

Matson, P.L., P.A. Eden, S.D. Vitali, C.A. Lawrence and H.M. Robertson. 2005. Sperm morphology with the testis and cauda epididymis of a Tasmanian Devil (Sarcophilus harrisii). Society for Reproductive Biology Thirty-sixth Annual Scientific Meeting. Abstracts.

Mawson, P.M. 1968. Two species of Nematoda (Spirurida: Spiruridae) from Australian dasyurids. Parasitology 58:75-78.

McAloose, D. and A.L. Newton. 2009. Wildlife cancer: a conservation perspective. Nature Reviews 9:517-526.

McCallum, H. 2007. Translocations of Tasmanian devils to islands. P. 30, Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

McCallum, H. 2008a. Tasmanian devil facial tumour disease: lessons for conservation biology. Trends in Ecology and Evolution 23(11):631-637.

McCallum, H. 2008b. The devil in decline. Higher Education Supplement. The Australian, 28 May 2008, p. 30.

McCallum, H. 2008c. Extinction real risk. Hobart Mercury, 29 October 2008.

McCallum, H. 2012. Disease and the dynamics of extinction. Philosophical Transactions of the Royal Society. B. 367:2828-2839.

McCallum, H. 2013. Tarkine mines could be last straw for Tasmanian devils. The Conversation, 15 January 2013

McCallum, H. and M. Jones. 2006. To lose both would look like carelessness. Tasmanian devil facial tumour disease. Public Library of Science 4(10):1671-1674.

McCallum, H. and M. Jones. 2010. Sins of omission and sins of commission: St Thomas Aquinas and the devil. Australian Zoologist 35(2):307-314. [This paper should be read in the understanding that although the paper was published in 2010, it was written sometime before July 2008]

McCallum, H, M. Jones, C. Hawkins, R. Hamede, S, Lachish, D.L. Sinn, N. Beeton, B. Lazenby. 2009. Transmission dynamics of Tasmanian devil facial tumor disease may lead to disease-induced extinction. Ecology 90:(12) 3379-3392.

6/05/2016 8:41 AM 401

McCallum, H., D.M. Tompkins, M. Jones, S. Lachish, S. Marvanek, B. Lazenby, G. Hocking, J. Wiersma and C.E. Hawkins. 2007. Distribution and impacts of Tasmanian Devil Facial Tumor Disease. EcoHealth 4(3):318-325.

McCoy, 1882. Sarcophilus ursinus (Harris sp.) Prodromus of the Palaeontology of Victoria 7:11-13. Geological Survey of Victoria, Melbourne.

McCarthy, F.D. 1976. The Tasmanian Devil in Aboriginal rock art. Mankind 10(3):181-182.

McClone, P. 2010. Save the Tasmanian Devil Program in need of urgent reforms. Tasmanian Times (online), 17 May 2010.

McDonald, K. 2007. Sympathy for the devil. Australian Life Scientist, 29 November, 2007.

McDonald-Madden, E., W.J.M. Probert, C.E. Hauser, M.C. Runge, H.P. Possingham, M.E. Jones, J.L. Moore, T.M. Rout, P.A. Vesk, and B.A. Wintle. 2010. Active adaptive conservation of threatened species in the face of uncertainty. Ecological Applications 20(5):1476-1489.

McEwan, S.R. 2000. Monitoring reproduction in the common Wombat (Vombatus ursinus) and Tasmanian Devil (Sarcophilus harrisii) through faecal steroid analysis. MSc Thesis Veterinary Science, University of Sydney, Sydney; 206 pp.

McGlashan N.D, D.L. Obendorf and J. Harington. 2006. Aspects of the fatal malignant disease among the Tasmanian Devil population (Sarcophilus laniarius). European Journal of Oncology 11(2):95-102.

McGlashan, N.D., D.L. Obendorf and J.S. Harington. 2007. Update on the devil facial tumour in Tasmania. European Journal of Oncology 12 (2):75-80.

McGovern, M. 2005. Spatial analysis of Tasmanian Devil Facial Tumour Disease. Thesis, University of Southern Queensland; 84 pp.

McIlroy, J. C. 1981. The sensitivity of Australian animals to 1080 poison. II. Marsupial and eutherian carnivores. Australian Wildlife Research 8(2):385-399.

McIntyre, D. 2015. 'Virtual fence' shows promise in reducing road toll of Tasmanian devils and other wildlife. ABC News online, 9 December 2015.

McKay, B. 2013. Devil breeding success. The Examiner, 20 April 2013.

McKeag, M., P.R. Brown and J.D. Sallis. 1981. A comparative study of the whole blood nucleotide profiles of the Tasmanian devil and other marsupials using high performance liquid chromatography. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry 70(3):541-547.

6/05/2016 8:41 AM 402

McKenzie, N.L., A. A. Burbidge, A. Baynes, R. N. Brereton, C. R. Dickman, G. Gordon, L. A. Gibson, P. W. Menkhorst , A. C. Robinson, M. R. Williams and J. C. Z. Woinarski. 2007. Analysis of factors implicated in the recent decline of Australia's mammal fauna. Journal of Biogeography 34(4):597–611.

McLean, S.; H. Galloway; S. Butler; D. Whittle and S.C. Nicol. 1983. Acetylation of sulphanilamide in four marsupials and a monotreme. Xenobiotica 13(2):81-85.

McManus, T. 2003. Tasmanian Devil Facial Tumour. Pp. 24-26, in Wildlife Disease Association. Australasian Section. 2003 Annual Conference, Healesville, Victoria, 1- 5 December 2003; 91 pp.

McQuirk, R. 2007. Tasmanian devil rescue plan sparks worry. Associated Press, 10 April 2007.

McShane, B. 2011. Another two devil facilities opened. The Examiner, 30 April 2011.

Medlock K.M. and D. Pemberton. 2010. Historical evidence of the existence of Tasmanian devils on Bruny Island in the late 1800’s. Internal report DPIPWE, Hobart.

Megirian, D., P.F. Murray, P.K. Latz and K.A. Johnson. 2002. The Mygoora Local Fauna: a late Quaternary vertebrate assemblage from central Australia. The Beagle: Records of the Museums and Art Galleries of the Northern Territory 18: 77–93.

Menzies, B.R., M.B. Renfree, T. Heider, F. Mayer, T.B. Hildebrandt and A.J. Pask. 2012. Limited genetic diversity preceded extinction of the Tasmanian Tiger. PLoS ONE 7(4): e35433. doi:10.1371/journal.pone.0035433.

Mercer, D. 2012. SW could offer safe haven for devils. The West Australian, 20 November 2012.

Mercer, P. 2008. Hope over Tasmanian Devil cancer. BBC News, 1 April 2008.

Meredith, [Mrs.] C. 1853. My Home in Tasmania or Nine Years in Australia. Bunce and Brother, New York; 370 pp.

Meredith, L.A. 1880. Tasmanian Friends and Foes: Feathered, Furred and Finned. A Family Chronicle of Country Life, Natural History, and Veritable Adventure. J. Walch & Sons, Hobart Town; 259 pp.

Meredith, R.W., M. Westerman, J.A. Case and M.S. Springer. 2008. A phylogeny and timescale for marsupial evolution based on sequences for five nuclear genes. Journal of Mammalian Evolution 15(1):1–36.

Meredith, R.W., J.E. Janecka, J. Gatesy, O.A. Ryder, C.A. Fisher, E.C. Teeling, A. Goodbla, E. Eizirik, T.L.L. Simão, T. Stadler, D.L. Rabosky, R.L. Honeycutt, J.J. Flynn, C.M. Ingram, C. Steiner, T.L. Williams, T. J. Robinson, A. Burk-Herrick, M. Westerman, N.A. Ayoub, M.S. Springer and W.J. Murphy. 2011. Impacts of the

6/05/2016 8:41 AM 403

Cretaceous Terrestrial Revolution and KPg Extinction on mammal diversification. Science. 334:521-524. doi:10.1126/science.1211028.

Mester, A. 2015a. Toledo Zoo joins effort to save Tasmanian devils. The Blade, 6 September 2015.

Mester, A. 2015b. Toledo Zoo’s devil trio greets first guests. New arrivals from Australia shriek, yowl, growl, chirp. The Blade, 28 October 2015.

Mester, A. 2016.Toledo Zoo bolsters effort to save Tasmanian devils. $500,000 to be contributed over 5 years. Toledo Blade, 10 January 2016.

Michael, S.A. and C.R. Sanster. 2010. Pulmonary mycobacteriosis caused by Mycobacterium intracellulare in a Tasmanian devil (Sarcophilus harrisii). Australian Veterinary Journal 88(7):280-282.

Middleton, A. 2013. Tarkine mine re-approved in devil territory. Australian Geographic, 31 July 2013.

Miles, A.E.W. and C. Grigson. 1990. Colyer’s Variations and Diseases of the Teeth of Animals. Cambridge University Press, Cambridge; 672 pp.

Miller, W., V.M. Hayes, A. Ratan, D.C. Petersen, N.E. Wittekindt, J. Miller, B. Walenz, J. Knight, J. Qi, F. Zhao, Q. Wang, O.C. Bedoya-Reina, N. Katiyar, L.P. Tomsho, L.M. Kason, R.-A. Hardie, P. Woodbridge, E.A. Tindall, M.F. Bertelsen, D.Dixon, S. Pyecroft, K.M. Helgen, A.M. Lesk, T.H. Pringle, N. Patterson, Y. Zhang, A. Kreiss, G.M. Woods, M.E. Jones and S. Schuster. 2011. Genetic diversity and population structure of the endangered marsupial Sarcophilus harrisii (Tasmanian Devil). Proceedings of the National Academy of Sciences 108(30):12348-12353.

Millikin, D. 2003. Disease threatening Tasmania's devils likened to HIV and SARS. Agence France Presse, 30 October 2003

Mills, C. 2011. Sympathy for the devils. Physorg online, 24 October 2011.

Milman, O. 2013a. Tasmania hopes foreign zoo breeding program will help save devils. The Guardian (Australian Edition), 24 June 2013.

Milman, O. 2013b. Tasmanian devils to be released back on to mainland. The Guardian (Australian Edition), 23 August 2013.

Milman, O. 2014. Tasmanian devils’ decline ‘driven by climate change’, new research shows. The Guardian (Australian Edition), 5 November 2014.

Milman, O. 2015a. Zoos Victoria unveils $30m plan to save 20 Australian species from extinction. The Guardian (Australian Edition), 24 March 2015.

Milman, O. 2015b. Tasmanian devils on mainland would reduce feral cats and foxes, study finds. The Guardian, 11 August 2015

6/05/2016 8:41 AM 404

Moeller, H. 1970. Vergleichende Untersuchungen zum Evolutionsgrad der Gehirne grosser Raubbeutler (Thylacinus, Sarcophilus und Dasyurus). I. Hirngewicht. II. Hirnform und Furchenbild. Zeitschrift für zoologische Systematik und Evolutionsforschung 8(1):69-88.

Moeller, H. 1972a. Sarcophilus harrisii (Dasyuridae). Gebrauch der Vorderbeine bei Beuteerwerb und Fressen. Film F 1834 des Institut für den Wissenschaftlichen Film, Göttingen.

Moeller, H. 1972b. Sarcophilus harrisii (Dasyuridae). Beutefang und Fressen. Film E 1835 des Institut für den Wissenschaftlichen Film, Göttingen.

Moeller, H. 1972c. Sarcophilus harrisii (Dasyuridae). Fressen von Eiern. Film E 1836 des Institut für den Wissenschaftlichen Film, Göttingen.

Moeller, H. 1974. Bewegungsweisen, Beutefang und Fressen beim Beutelteufel, Sarcophilus harrisii. Film C 1105 des Institut für den Wissenschaftlichen Film, Göttingen.

Mollison, B.C. 1961. Birds and mammals poisoned on game control operations. Special Report to Animals and Birds Protection Board No. 2.

Money, L. 2013. Drug trial offers hope for decimated devil population. Sydney Morning Herald, 1 July 2013.

Montali, R.J. 1980. An overview of tumors in zoo animals. Pp. 531-542, in R.J. Montali and G. Miyaki, eds, The Comparative Pathology of Zoo Animals. Smithsonian Institute Press; Washington, DC; # pp.

Mooney, N. 1992. The Devil you know. Leatherwood: Tasmanian Journal of Discovery 1(3):54-61.

Mooney, N. 2003a. Deadly diseases – Tasmanian Devils. Brochure published in March 2003 by DPIW but seen as published in Land for Wildlife News in March 2004, vol. 7, p. 3.

Mooney, N. 2003b. Tassie devil under threat. Hobart Mercury, 1 September 2003.

Mooney, N. 2004. The Devil’s new hell. Nature Australia 28(3): 34-41.

Mooney, N. 2005. The Devil I know. Tasmanian Conservationist No. 303: 8-9.

Mooney, N. 2007. Flow-on effects of DFTD on evasive and other species. P. 33, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Mooney, N. 2012. Tasmanian without devils: environmental impacts and implications for feral species. Royal Society of Tasmania Winter Series – Lecture 2; 17 July 2012. [Available online]

6/05/2016 8:41 AM 405

Mooney, N. 2014. Devil survival needs shared resolve. Hobart Mercury, 19 April 2014.

Mooney, N. 2015. Tasmanian devil and cat interactions; all may not be as presumed. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Symposium abstract, p. 13.

Moore, M.R. 2008. Opinion on a chemical aetiology for facial tumour development in the Tasmanian Devil. Letter to Hamish McCallum, 27 February 2008.

Moore, R. and M. Tizard. 2007. Determination of tumour origin via profiling of microRNA expression patterns. P. 15, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Morris, K. 2013. Devil is in the DNA: [Ancient DNA is helping shed light on why the Tasmanian devil is being driven to extinction.] Australasian Science 34(2):31-33.

Morris, K., J. Austin and K. Belov. 2011. When did Tasmanian Devils lose MHC diversity? Faculty of Veterinary Science’s Postgraduate Forum – 2011, University of Sydney, p. 51.

Morris, K., J. Austin and K. Belov. 2013. Low major histocompatibility complex diversity in the Tasmanian devil predates European settlement and may explain susceptibility to disease epidemics. Biology Letters 9(1):20120900. [Published online, 5 December 2012.]

Morris, K. and K. Belov. 2013. Does the devil facial tumour produce immunosuppressive cytokines as an immune evasion strategy? Veterinary Immunology and Immunopathology 153(1-2):159-164. Published online, 13 February 2013.

Morris, K.M. and K. Belov. 2014. Characterisation of the Tasmanian devil immunome and identification of SNPs within devil immune genes. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 36.

Morris, K.M., Y. Cheng, W.C. Warren, A.T. Papenfuss and K. Belov. 2015a. Identification and analysis of divergent immune gene families within the devil genome. Immunome Research 11(1); 12 pp. Published online, 10 February 2015.

Morris, K.M., B. Wright, C.E. Grueber, C. Hogg and K. Belov. 2015b. Lack of genetic diversity across diverse immune genes in an endangered mammal, the Tasmanian devil (Sarcophilus harrisii). Molecular Ecology 24:3860-3874. Published online, 28 June 2015. doi: 10.1111/mec.13291

Morrison, P.R. 1965. Body temperatures in some Australian mammals. IV. Dasyuridae. Australian Journal of Zoology 13(1):173-187.

Mounster, B. 2010. Biologists find vital signs of healthy life. Hobart Mercury, 11 February 2010.

6/05/2016 8:41 AM 406

Mounster, B. 2012a. Devil delivery to Denmark. Hobart Mercury, 12 October 2012.

Mounster, B. 2012b. Curse of the feral cats. Hobart Mercury, 30 October 2012.

Mounster, B. 2013a. Devils edge back from brink. Hobart Mercury, 1 January 2013.

Mounster, B. 2013b. Tide may be turning in battle against deadly devil disease. Hobart Mercury, 1 January 2013.

Mounster, B. 2013c. Caring for their future. Hobart Mercury, 1 January 2013.

Mounster, B. 2013d. Devils step up in fight to survive. Tasmanian Country, 4 January 2013.

Mounster, B. 2013e. Tumour drug hope. Hobart Mercury, 2 July 2013.

Mounster, B. 2014. US zoo and Sydney University join forces to help save wild devils. Hobart Mercury, 26 February 2014.

Mounster, B. 2015a. Honour rewards Professor Greg Woods’ devil tumour obsession. Perth Now, 1 June 2015.

Mounster, B. 2015b Little penguins bunker down to evade Tasmanian devils. Hobart Mercury, 6 June 2015.

Mounster, B. 2015c. Wildlife expert doubts fall in devil numbers has led to cat surge. Hobart Mercury, 22 June 2015.

Mounster, B. 2016. Devils show signs of resistance to facial tumour disease.Hobart Mercury, 13 January 2016.

Mueller, J.F. 1937. A repartition of the genus Diphyllobothrium. Journal of Parasitology 23(3):308-310.

Mulvaney, D.J. 1960. Archaeological excavations at Fromm’s Landing on the lower Murray River, South Australia. Proceedings of the Royal Society of Victoria 72(2):53-85.

Mumbray, T. 1992. A note on the analysis of a series of dasyurid scats. Tasmanian Naturalist 109:1-4.

Munday, B.L. 1966. Laboratory Investigation of Specimens from Diseased and Apparently Normal Free-Living Wild Animals in Tasmania. Tasmanian Department of Agricultural Research Bulletin No. 5; 29 pp.

Munday, B.L. 1972. A serological study of some infectious diseases of Tasmanian wildlife. Journal of Wildlife Diseases 8(2):169-175.

6/05/2016 8:41 AM 407

Munday, B.L. 1988. Marsupial diseases. Pp. 299-365, in D.I. Bryden, ed., Australian Wildlife. The John Keep Refresher Course for Veterinarians. Proceedings 104. Post Graduate Committee in Veterinary Science, University of Sydney; 952 pp (exclusive of index).

Munday, B.L. and G.G. Gregory. 1974. Demonstration of larval forms of Baylisascaris tasmaniensis in the wombat (Vombatus ursinus). Journal of Wildlife Diseases 10:241-242.

Munday, B.L. and R.W. Mason. 1978. Sarcocystis and related organisms in Australian wildlife: I. Survey findings in mammals. Journal of Wildlife Diseases 14(2):417-433.

Munday, B.L., R.W. Mason, W.J. Hartley, P.J.A. Presidente and D. Obendorf. 1978. Sarcocystis and related organisms in Australian wildlife: 1. Survey findings in mammals. Journal of Wildlife Diseases 14:417-433.

Munday, B.L., R.J. Whittington and N.J. Stewart. 1998. Disease conditions and subclinical infections of the platypus (Ornithorhynchus anatinus). Philosophical Transactions of the Royal Society of London B 353:1093-1099. Read online.

Murchison, E. 2006. Pledge to use funding for devil cure. Sunday Tasmanian, 12 November 2006.

Murchison, E. P. 2009. Clonally transmissible cancers in dogs and Tasmanian devils. Oncogene 27:S19-S30.

Murchison, E. and Rosenfeld. 2014. Early detection of cancer using circulating tumour DNA: feasibility and field trial using the unique Tasmanian Devil transmissible cancer. Online. Accessed 7 October 2014. [Grant proposal].

Murchison, E.P. C. Tovar, A. Hsu, H.S. Bender, P. Kheradpour, C.A. Rebbeck, D. Obendorf, C. Conlan, M. Bahlo, C.A. Blizzard, S. Pyecroft, A. Kreiss, M. Kellis, A. Stark, T.T. Harkins, J.A.M. Graves, G.M. Woods, G.J. Hannon and A.T. Papenfuss. 2010. The Tasmanian Devil transcriptome reveals Schwann cell origins of a clonally transmissible cancer. Science 327: 84-87.

Murchison, E.P. and 45 co-authors. 2012. Genome sequencing and analysis of the Tasmanian Devil and its transmissible cancer. Cell 148:780-791.

Murgia, C., J.K. Pritchard, S.Y. Kim, A. Fassati and R.A. Weiss. 2006. Clonal origin and evolution of a transmissible cancer. Cell 126: 477-487.

Naidoo, M. 2008. This devil could save his species. Hobart Mercury, 1 April 2008.

Naidoo, M. 2009. Extinction threat on march. Hobart Mercury, 22 May 2009.

Natural History New Zealand. 1999. The Devil’s Playground. Natural History New Zealand, Dunedin, New Zealand. Video.

6/05/2016 8:41 AM 408

Neales, S. 2011. Devil of a disappointment as tumour disease spreads to isolated pocket. The Australian, 10 December 2011.

Nicholls, H. 2009. Time to sequence the ‘red and the dead.’ Nature News, 14 April 2009.

Nicol, S.C. 1978. Rates of water turnover in marsupials and eutherians: a comparative review with new data on the Tasmanian Devil. Australian Journal of Zoology 26(3):465-473.

Nicol, S.C. 1982. Physiology of the Tasmanian Devil Sarcophilus harrisii (Dasyuridae, Marsupialia). Pp. 273-278, in M. Archer, ed, Carnivorous Marsupials. Royal Society of New South Wales, Sydney; 804 pp.

Nicol, S.C. and M. Maskrey. 1980. Thermoregulation, respiration and sleep in the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). Journal of Comparative Physiology B140(3):241-248.

Nicol, S. and M. Maskrey. 1986. Arterial blood acid-base regulation in the Tasmanian Devil, Sarcophilus harrisii, during exercise. Physiological Zoology 59(2):212-219.

Nielsen, A.S. 2009 The potential impact of Tasmanian devil (Sarcophilus harrisii) decline on prey behaviour. Honours thesis, University of Tasmania, Hobart, Australia.

Nilsson, M.A. 2016. The devil is in the details: transposable element analysis of the Tasmanian devil genome. Mobile Genetic Elements 6(1):#-#.

Nilsson, M.A., U. Amason, P.B. Spencer and A. Janke. 2004. Marsupial relationships and a timeline for marsupial radiation in South Gondwana. Gene 340(2):189-106.

Nilsson, M.A., G. Churakov, M. Sommer, N.V. Tran, A. Zemann, J. Brosius and J. Schmitz. 2010. Tracking marsupial evolution using archaic genomic retroposon insertions. PLOS Biology 8(7); 8 pp.

Nilsson-Janke, M.A.N., A. Janke, E.P.Murchison, Z. Ning and B.M. Hallström. 2012. Expansion of CORE-SINEs in the genome of the Tasmanian devil. BMC Genomics 13:172.

Nobbe, L. 2016. St. Louis Zoo welcomes Tasmanian devils. KDSK.com, 19 April 2016.

Noetling, F. 1910. The food of Tasmanian Aborigines. Papers and Proceedings of the Royal Society of Tasmania 1910:279-305.

Nogrady, B. 2012. Devil of a disease. Ecos Magazine, 25 February 2013. [As seen on Science Alert online.]

Noone, R. 2015a. Tasmanian devils: Release bid for NSW blocked by our island state

6/05/2016 8:41 AM 409

Daily Telegraph, 2 November 2015.

Noone, R. 2015b. Devil Ark brings Tasmanian devils back from brink of extinction Daily Telegraph, 19 November 2015.

Northwood, C. 1990. A taphonimic analysis of Sarcophilus Harrisii in Australian archaeological sites. B.A. Honours Thesis, La Trobe University.

Novak, R.M. and J.L. Paradiso. 1983. Walker’s Mammals of the World. 4th Ed. Vol. 1. Johns Hopkins University Press, Baltimore, Maryland.

Obendorf, D.L. 1992. Pathology seen in dasyurid marsupials. Proceedings of the Canadian and American Zoo and Aquarium Park Conference. Toronto, Canada; 114 pp.

Obendorf, D.L. 1993. Diseases of dasyurid marsupials. Pp. 39-46, in M. Roberts, J. Camio, G. Crawshaw and M. Hutchins, eds, The Biology and Management of Australian Carnivorous Marsupials. Metropolitan Toronto Zoo and the Monotreme and Marsupial Advisory Group of the American Association of Zoological Parks and Aquariums, Toronto, Ontario; # pp.

Obendorf, D.L., J.H. Handlinger, R.W. Mason, K.P. Clarke, A.J. Forman, P.T. Hooper, S.J. Smith and M. Holdsworth. 1990. Trichinella pseudospiralis infection in Tasmanian wildlife. Australian Veterinary Journal 67(3):108-110.

Obendorf, D.L. and N.D. McGlashan. 2008. Research priorities in the Tasmanian Devil facial tumour debate. European Journal of Oncology 13(4):229-238.

Obendorf, D.L. and S.J. Smith. 1989. A re-examination of Dithyridium cyanocephali Ransom 1905, a metacestode parasite from the Thylacine Thylacinus cyanocephalus. Papers and Proceedings of the Royal Society of Tasmanian 123:133-136.

Obringer A., A. Lund, W. Pryor, E. Crichton, N. Loskutoff and E. Kirchner. 2002. Sex and the devil: Monitoring reproductive activity in the Tasmanian devil (Sarcophilus harrisii). Abstract in Program of the Indiana Academy of Science 118th Annual Meeting: p. 106.

Ogilvie, F. 2011. Devil sperm collected in bid to save species. ABC News, 2 December 2011.

Oglesby, R. 1978. Notes on breeding the Tasmanian devil in captivity. Thylacinus 3 (1-4):17-20.

O’Neal, T. 2011. Animal Enrichment Strategies for Promoting Natural Behaviors in Captive Populations of Tasmanian Devils (Sarcophilus harrisii). ISP Collection. Paper 1129.

O’Neill, I.D. 2010. Tasmanian devil facial tumor disease: insights into reduced tumor surveillance from an unusual malignancy. International Journal of Cancer 127:1637- 1642.

6/05/2016 8:41 AM 410

O’Neill, M. 2014. Scientists resign 'living dead' species to extinction, call for triage debate. ABC News, 19 March 2014.

Owen, D. and D. Pemberton. 2005. Tasmanian Devil. A Unique and Threatened Animal. Allen & Unwin, Crows Nest, New South Wales, Australia; 225 pp. [The 2011 “edition” seems to be essentially the same as this edition.]

Owen, R. 1838. Fossil remains from Wellington Valley, Australia. Pp. 359-369 (Appendix), in T.L. Mitchell, Three Expeditions into the Interior of Eastern Australia, with Descriptions of the Recently Explored Region of Australia Felix, and of the Present Colony of New South Wales. Two Volumes. T. and W. Boone, London; # pp.

Owen, R. 1840-1845. Odontography; or a Treatise on the Comparative Anatomy of the Teeth; their Physiological Relations, Mode of Development, and Microscopic Structure, in the Vertebrate Animals. Vol. 2. Atlas. Hippolyte Bailliere, London.

Owen, R. 1841. The osteology of the Marsupialia. Transactions of the Zoological Society of London 2:379-408.

Owen, R. 1877. Researches on the fossil remains of the extinct mammals of Australia with a notice of the extinct marsupials of England. J. Erxleben, London; 522 pp.

Packer, W.C. 1966. Wheel-running behavior in the marsupial Sarcophagus harrisii. Journal of Mammalogy 47(4):698-701.

Paddle, R. 2012. The thylacine’s last straw: epidemic disease in a recent mammalian extinction. Australian Zoologist 36(1):75-92.

Paine, M. 2006. Disbelief at devil aid rebuff. Hobart Mercury, 13 November 2006.

Paine, M. 2007. Devil disease spreading. Hobart Mercury, 20 October 2007.

Paine, M. 2011. Devils heading offshore. Hobart Mercury, 13 December 2011.

Paine, M. 2012. Farmers call for action. Hobart Mercury, 3 January 2012.

Paine, M. 2013. Zoos rush to save devils. Hobart Mercury, 24 June 2013.

Papenfuss, A.T., A. Hsu and M. Wakefield. 2010. Marsupial sequencing projects and bioinformatics challenges. Pp. 121-132, in J.E. Deakin, P.D. Waters, J.A.M. Graves, eds, Marsupial Genetics and Genomics. Springer, Dordrecht; 519 pp.

6/05/2016 8:41 AM 411

Pardoe, C. 1995. Riverine, biological and cultural evolution in southeastern Australia. Antiquity 69:696-713.

Parer, D. and L. Parer-Cook. 2004. Terrors of Tasmania. ABC Natural History and National Geographic. Video. M. Jones, Scientific Consultant.

Parkes, J. 2012. Review of the Save the Tasmanian Devil Program. Landcare Research, Lincoln, New Zealand; 37 pp. Report prepared for DPIPWE, January 2012.

Parks and Wildlife Service Tasmania. 2005. Devils on Maria Island. Parks and Wildlife Service Tasmania, May 2005.

Parks and Wildlife Service Tasmania. 2006. Living with wildlife. Tasmanian Devils and Quolls. Parks and Wildlife Service Tasmania, January 2006

Parks and Wildlife Service Tasmania. 2009. Tasmanian Devil, Sarcophilus harrisii. Parks and Wildlife Service Tasmania, online, version of 30 March 2009.

Parrott, M.L., R. Nightingale, T. Eastley, M. Lancaster and E.G. Ritchie. 2015. The devil is in the detail: maintaining natural behaviours and socialisation in Tasmanian Devil breeding programs. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 59.

Parsons, F.G. 1899. The joints of mammals compared with those of man. Journal of Anatomy and Physiology 34(1):41-68.

Parsons, F.G. 1900. The joints of mammals compared with those of man. Part II. Joints of the hind limb. Journal of Anatomy and Physiology 34(3):301-323.

Parsons, R.S. and E.R.Guiler. 1972a. Nonprostatic formalin-resistant acid phosphatase in the Tasmanian devil. Clinical Chemistry 18 (8):867.

Parsons, R.S. and E. Guiler. 1972b. Observations on the blood of the marsupial tiger cat (Dasyurops maculatus). Comparative Biochemistry and Physiology 43A:935-939.

Parsons, R.S., E.R. Guiler and R.W.L. Heddle. 1971. Comparative studies on the blood of monotremes and marsupials – II. Electrolyte organic constituents, proteins, gas analysis and enzymes. Comparative Biochemistry and Physiology 39B:209-217. [Parsons et al., 1971a]

Parsons, R.S., R.W.L. Heddle, W.G. Flux and E.R. Guiler. 1970. Studies on the blood of the Tasmanian Devil. Comparative Biochemistry and Physiology 32B:345-351.

6/05/2016 8:41 AM 412

Parsons, R.S., R.W.L. Heddle and E.R. Guiler. 1971. The distribution of acid phosphatase in the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). Comparative Biochemistry and Physiology 39B:219-226. [Parsons et al., 1971b]

Patchett, A.L., R. Latham, K.H. Brettingham-Moore, C. Tovar, A.B. Lyons and G.M. Woods. 2015. Toll-like receptor signaling is functional in immune cells of the endangered Tasmanian devil. Developmental and Comparative Immunology 53:123- 133. Published online, 13 July 2015.

Pattie, I. 2003. Virus could be the trigger for devil downfall. The Examiner, 24 July 2003.

Patzke, N., M.F. Bertelsen, K. Fuxe and P.R. Manger. 2014. Nuclear organization of cholinergic, catecholaminergic, serotonergic and orexinergic systems in the brain of the Tasmanian devil (Sarcophilus harrisii). Journal of Chemical Neuroanatomy 61- 62C:94-106. Published online, 20 August 2014.

Payne, A. Strategies for the devil. Wildlife Australia 49(2):18-23.

Pearse, A.-M. 1981. Aspects of the biology of Uropsylla tasmanica (Rothschild) (Siphonaptera). M.Sc. Thesis, University of Tasmanian.

Pearse, A.-M. 2005. The cytogenetics of the Tasmanian Facial Tumour DFTD. Cairns, CSIRO Publishing. [as cited in Loh, 2006]

Pearse, A.-M. and K. Swift. 2006. Transmission of devil facial-tumour disease. Nature 439:549.

Pearse, A.-M., K. Swift, P. Hodson, B. Hua, H. McCallum, S. Pyecroft, R. Taylor, M.D.B. Eldridge and K. Belov. 2012. Evolution in a transmissible cancer: a study of the chromosomal changes in devil facial tumor (DFT) as it spreads through the wild Tasmanian devil population. Cancer Genetics 205(3):101-112.

Pearse, A-M., K.S. Swift and K. Nutter. 2007. Inverted chromosome 5 and DFTD evolution. P. 10, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Pearson, J. 1940. Notes on the blood system of the Marsupialia. Papers and Proceedings of the Royal Society of Tasmania 1939:77-94.

Pearson, J. 1945. The female urogenital system of the Marsupialia with special reference to the vaginal complex. Papers and Proceedings of the Royal Society of Tasmania 1944:71-98.

6/05/2016 8:41 AM 413

Pearson, J. 1950. Placentation of the Marsupialia. Proceedings of the Linnean Society 161(1):1-9.

Pearson, J. and J.M De Bavay. 1953. The urogenital system of the Dasyurinae and the Thylacininae (Marsupialia, Dasyuridae). Papers and Proceedings of the Royal Society of Tasmania 87:175-199.

Peatling, S. 2005. Tasmanian Devils decimated by mystery cancer. National Geographic News, 29 March 2005.

Peck, S., R. Corkrey, R. Hamede, M. Jones and P. Canfield. 2015. Hematologic and serum biochemical reference intervals for wild Tasmanian devils (Sarcophilus harrisii). Veterinary Clinical Pathology 44(4):519-529. Published online 27 November 2015.

Peel, E., Y. Cheng, C. Wilson, J. Djorjevic, S. Chen, T. Sorrell and K. Belov. Cathelicidins in the Tasmanian devil (Sarcophilus harrisii). Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 37.

Pemberton, D. 1990. Social organisation and behaviour of the Tasmanian devil, Sarcophilus harrisii. PhD Thesis, University of Tasmania, Hobart, Tasmania; 247 pp.

Pemberton, D. 2014. The Tasmanian devil and DFTD in the wild – a unique management issue. Abstract. NRM Regions Knowledge Conference, Launceston; 18 March 2014.

Pemberton, D. 2015. Road to recovery: from a case definition to post epidemic in situ management. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 61.

Pemberton, D. and N. Gales. 1991. Field immobilisation of Tasmanian Devils (Sarcophilus harrisii) with ketamine hydrochloride and xylazine hydrochloride. Wildlife Research 18:695-698.

Pemberton, D. and D. Renouf. 1993. A field study of communication and social behavior of the Tasmanian devil at feeding sites. Australian Journal of Zoology 41:507-526.

Pemberton, D., S. Gales, B. Bauer, R. Gales, B. Lazenby and K. Medlock. 2008. The diet of the Tasmanian Devil (Sarcophilus harrisii) as determined from analysis of scat and stomach contents. Papers and Proceedings of the Royal Society of Tasmania 142(2):13-21.

Pennisi, E. 2010. Genomics researchers upset by rivals' publicity. Science 329:1585.

Perry, T. 2014. San Diego Zoo joins a heck of an effort: to save the Tasmanian devil. Los Angeles Times, 21 March 2014.

6/05/2016 8:41 AM 414

Phalen, D.N., A.E. Frimberger, S. Peck, S. Pyecroft, C. Harmsen, S. Lola and A.S. Moore. 2015. Doxorubicin and carboplatin trials in Tasmanian devils (Sarcophilus harrisii) with Tasmanian devil facial tumor disease. The Veterinary Journal, published online 13 October 2015.

Phalen, D.N., A. Frimberger, S. Pyecroft, S. Peck, C. Harmsen, S. Lola, B. de Mello Mattos, K.M. Li, A. J. McLachlan and A. Moore. 2013. Vincristine chemotherapy trials and pharmacokinetics in Tasmanian Devils with Tasmanian Devil Facial Tumor Disease. PLoS ONE 8(6): e65133. doi:10.1371/journal.pone.0065133

Phillips, B.T. and S.M. Jackson. 2003. Growth and development of the Tasmanian devil (Sarcophilus harrisii) at Healesville Sanctuary, Victoria, Australia. Zoo Biology 22(5):497-505.

Phillips, N. 2011. Survival of the cheapest. Sydney Morning Herald, 11 August 2011.

Phillips, N. 2014. Tasmanian devils hit by climate change before facial tumour devastation, study finds. Sydney Morning Herald, 5 November 2014.

Pickrell, J. 2007. Tasmanian Devil epidemic: cause isolated? Cosmos Magazine, Cosmos Online, 27 June 2007.

Pierce, J. 2012. Devil of a good day out at Dreamworld as new arrivals emerge. Herald Sun, 11 October 2012.

Pincock, S. 2007. The devil to pay. Financial Times Weekend Magazine, 12 May 2007.

Pinfold, T.L., G.K. Brown, S.S. Bettiol and G.M. Woods. 2014. Mouse model of Devil Facial Tumour Disease establishes that an effective immune response can be generated against the cancer cells. Frontiers in Immunology 5, 10 pp. [Published online, 27 May 2014]

Pinfold, T., G. Brown, A. Kriess and G. Woods. 2013. Cancer from Hell: Devil Facial Tumour Disease. Front. Immunol. Conference Abstract: 15th International Congress of Immunology (ICI). doi: 10.3389/conf.fimmu.2013.02.00604

Pinfold, T., G. Brown, A.L. Patchett, C.D. Tovar, S.S. Bettiol and G.M. Woods. 2015. Curse of the Devil: Devil Facial Tumour Disease. Australasian Cytometry Society 38th Annual Scientific Meeting, October 11-14, 2015, Perth, Australia . [Conference Extract]

Platt, J.R. 2011a. Has an infectious cancer doomed Tasmanian Devils to extinction? Scientific American online, 18 January 2011.

6/05/2016 8:41 AM 415

Platt, J.R. 2011b. Sperm bank and reproductive research could help save Tasmanian Devils from extinction. Scientific American online, 14 December 2011.

The PLOS Genetics Staff (2014) Correction: Genomic Restructuring in the Tasmanian Devil Facial Tumour: Chromosome Painting and Gene Mapping Provide Clues to Evolution of a Transmissible Tumour. PLoS Genet 10(10): e1004840. doi:10.1371/journal.pgen.1004840

Pocock, R.I. 1914. On the facial vibrissae of Mammalia. Proceedings of the Zoological Society of London 1914(3): 889-912.

Pocock, R.I. 1926. The external characters of Thylacinus, Sarcophilus and some related marsupials. Proceedings of the Zoological Society of London 1926:1037- 1084.

Porter, M. and S. Jennings. 2012. Cost-effectiveness analysis of options for building the Tasmanian Devil effective insurance population. Draft. Unpublished report prepared for STDP and DPIPWE. [Not seen.]

Poskitt, L. 2013. Devils off to San Diego. The Examiner, 30 August 2013.

Prowse, T.A.A., C.N. Johnson, C.J.A. Bradshaw and B.W. Brook. 2013. An ecological regime shift resulting from disrupted predator-prey interactions in Holocene Australia. Ecology. http://dx.doi.org/10.1890/13-0746.1

Pukk, C.E. 2005. The habitat use of Tasmanian devils Sarcophilus laniarius harrisii across natural and pastoral mosaics. B.Sc. Honours Thesis, University of Tasmania; 95 pp.

Pye, R. 2015. Demonstration of induced and naturally occurring immune responses against Devil Facial Tumour Disease in Tasmanian devils. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 65.

Pye, R., A. Kreiss, D. Pemberton and G. Woods. 2015a. Devil facial tumour disease (DFTD) immunisation trial on Tasmanian devils to be released in initial phase of rewilding programme. Abstract of oral presentation, # 4, 64th International Conference of the Wildlife Disease Association 2015; 27 July 2015.

Pye, R. J., D. Pemberton, C. Tovar, J.M.C. Tubio, K.A. Dun, S. Fox, J. Darby, D. Hayes, G.W. Knowles, A. Kreiss, H.V.T. Siddle, K. Swift, A.B. Lyons, E.P. Murchison and G.M. Woods. 2015b. A second transmissible cancer in Tasmanian devils. Proceedings of the National Academy of Sciences 113(2):374-379. Published online, 28 December 2015.

Pye, R.J., G.M. Woods and A. Kreiss. 2015c. Devil Facial Tumor Disease. Veterinary Pathology, published online, 13 December 2015, doi: 10.1177/0300985815616444

6/05/2016 8:41 AM 416

Pyecroft, S.B. 2007a. Chemicals: their role in DFTD. P. 17, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Pyecroft, S.B. 2007b. Transmission trials: Devil facial tumour disease. P. 18, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Pycroft, S.B., M. Jones, C. Hawkins and N. Mooney. 2004. Tasmanian Devil Handling Safe Work Practices. Version 2. July 2004; 9 pp. Tasmanian Government STORS Arckive online.

Pyecroft, S.B., A.-M. Pearse, R. Loh, K. Swift, K. Belov, N. Fox, E. Noonan, D. Hayes, A. Hyatt, L. Wang, D. Boyle, J. Church, D. Middleton and R. Moore. 2007. Towards a case definition for Devil Facial Tumour Disease: what is it? EcoHealth 4(3):346-351.

Quammen, D. 2008a. Contagious cancer: the evolution of a killer. Harpers April, 2008, pp. 32-39.

Quammen, D. 2008b. Contagious cancer. Review. Australian Financial Review 2 May 2008, pp. 1-2, 10-12.

Quammen, D. 2008c. What’s killing the Tasmanian Devil? Yale Environment 360. Online.

Rae, C., K.J.E. Sweeney, A.K. Krockenberger, N.S. Agar, C.H. Gallagher and P.W. Kuchel. 1993. Comparison of the 1H and 31P NMR spectra of erythrocytes and plasma from some Australian native animals: bandicoot, echidna, koala, little penguin, tammar wallaby, Tasmanian devil, tree kangaroo and wombat. Comparative Haematology International 3(2):71-80.

Ratcliffe, H.L. 1933. Incidence and nature of tumours in captive wild mammals and birds. American Journal of Cancer 17:116-135.

Ray, L. 2014. Baby Tasmanian devils have first health check, pics, video. Newcastle Herald, 8 July 2014.

Read, G. 2000. Roadkill ‘scent test.’ Sunday Tasmanian, 4 June 2000.

Reed, E.H. and S.J. Bourne. 2000. Pleistocene fossil vertebrate sites of the South East region of South Australia. Transactions of the Royal Society of South Australia 124(2): 61-90.

Rehmeyer, J. 2014. The Immortal Devil. Discover 35(4):42-49.

6/05/2016 8:41 AM 417

Reppas, G., P. Nosworthy, T. Hansen, M. Govendir and R. Malik. 2010. Inguinal panniculitis in a young Tasmanian devil (Sarcophilus harrisii) caused by Mycobacterium mageritense. Australian Veterinary Journal 88(5):197-200.

Rex, E. 2008. For Tasmanian Devils, hope against a wily cancer. The New York Times, 18 November 2008.

Richards, B. 2010. Cedric gave life to species. Hobart Mercury, 2 September 2010.

Richards, B. 2014. Deadly facial tumour disease takes a back seat as cars take their toll on populations. The Sunday Tasmanian, 12 January 2014.

Richards, B. 2015a. Tasmania starts second healthy Devil haven boosting survival hopes. Hobart Mercury, 18 November 2015.

Richards, B. 2015b. Disease-free devil project suffers another blow. Hobart Mercury, 14 December 2015.

Richards, B. 2015c. Roadkill toll for captive-bred and released Tasmanian devils rises to 15. Hobart Mercury, 31 December 2015.

Richards, B. 2016. Tasmanian devil roadkill toll hits 18 as more than one in four released animals hit by cars. Hobart Mercury, 1 March 2016.

Ride, W.D.L. 1964. A review of Australian fossil mammals. Proceedings of the Royal Society of Western Australia 47:97-131.

Ride, W.D.L. 1970. A Guide to the Native Animals of Australia. Angus and Robertson, Melbourne; 249 pp.

Ritchie, E. 2013. Should we move Tasmanian Devils back to the mainland? The Conversation (online), 5 November 2013.

Ritchie, E.G., B. Elmhagen, A.S. Glen, M. Letnic, G. Ludwig and R.A. McDonald. 2012. Ecosystem restoration with teeth: what role for predators? Trends in Ecology and Evolution 27(5):265-271.

Roberts, B. 1988. The Tasmanian Devil Who Couldn't Eat Meat: a Fable for Children of All Ages. Rainbow Books, Lauderdale, Tasmania; 76 pp.

Roberts, M.G. 1915. The keeping and breeding of Tasmanian Devils (Sarcophilus harrisii). Proceedings of the Zoological Society of London 1915(4):575-581.

Robertson, H. 2005. Devilish decline. Current Biology 15(21): R858-859.

Robertson, K. 2012. Three Tassie devils on the loose from Peel Zoo, near Mandurah. PerthNow, 17 October 2012.

6/05/2016 8:41 AM 418

Robinson, K.W. and P.R. Morrison. 1957. The reaction to hot atmospheres of various species of Australian marsupial and placental animals. Journal of Cellular and Comparative Physiology 49(3):455-478.

Robson, S.K. and W.G. Young. 1990. A comparison of tooth microwear between an extinct marsupial predator, the Tasmanian Tiger Thylacinus cynocephalus (Thylacinidae) and an extant scavenger, the Tasmanian Devil Sarcophilus harrisii (Dasyuridae: Marsupialia). Australian Journal of Zoology 37(5): 575–589.

Rofe, R. and D. Hayman. 1985. G-banding evidence for a conserved complement in the Marsupialia. Cytogenetics and Cell Genetics 39: 40–50.

Rose, R.W. 1989. Embryonic growth rates of marsupials with a note on monotremes. Journal of Zoology 218:11-16.

Rose, R.W., L. Bell and G. Shaw. 2006. Hormonal control of birth behavior in the Tasmanian devil Sarcophilus harrisii. Hormones and Behavior 50(3):417-423.

Rose, R.W., C.M. Nevison and A.F. Dixon. 1997. Testes weight, body weight and mating systems in marsupials and monotremes. Journal of Zoology 243(3):523-531.

Ross, T. 2008. Persistent chemicals in Tasmanian Devils. Report for the “Save the Tasmanian Devil Program; 24 pp.

Rounsevell, D.E. 1989. Managing offshore islands for nature conservation in Tasmania. Pp. 157-161, in A. Burbidge, ed., Australian and New Zealand Islands: Nature Conservation Values and Management. Proceedings of a Technical Workshop, Barrow Island, Western Australia, 1985. Occasional Paper 2/89. Department of Conservation and Land Management, Perth; 233 pp.

Rounsevell, D.E., R.J. Taylor and G.J. Hocking. 1991. Distribution records of native terrestrial mammals in Tasmania. Wildlife Research 18(6): 699-717.

Rudolphi, C.A. 1819. Entozoorum Synopsis. Berlin; # pp.

Ruiz-Aravena, M., R. Hamede, S. Carver and M. Jones. 2015. Devil-tumour physiological relationships: is there room for coevolution between devil and tumour? 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 27.

Ryan, J. 2015a. More healthy Tasmanian devils released into wild as part of tumour disease program. ABC New online, 18 November 2015.

Ryan, K. 2012. Researchers have devil of a time taking blood from Tasmanian devils at Ballarat Wildlife Park. Herald Sun, 27 July 2012.

Saey, T.H. 2011. Genes & cells: Tasmanian devil diversity detailed: First genetic blueprints don't explain cancer vulnerability. Science News 180(3):8.

6/05/2016 8:41 AM 419

Sallis, J.D. and E.R. Guiler. 1977. Unusual plasma acid phosphatase activity in two species of Tasmanian marsupials. Pp. 327-335, in B. Stonehouse and D. Gilmore, eds, The Biology of Marsupials. MacMillan, London; 486 pp.

Sallis, J.D. and E.R. Guiler. 1977. Acid phosphatase in the Tasmanian Devil: characterization through immunoelectrophoresis and thin-layer isoelectric focusing in polyacrylamide gel. Comparative Biochemistry and Physiology 56B:189-193.

Sallis, J.D., S.C. Nicol, P. Perrone and P.R. Brown. 1984. High performance liquid chromatographic profiles of nucleosides, bases and tryptophan in the plasma of the Tasmanian Devil and four other marsupial species. Comparative Biochemistry and Physiology 79B(3):391-394.

Sallis, J.D., R.S. Parsons and E.R. Guiler. 1973. Isolation and some properties of an acid phosphatase from the plasma of the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). Comparative Biochemistry and Physiology 44B:971-980.

Sandars, D.F. 1957a. A new strigeid trematode from an Australian marsupial. Journal of Helminthology 31(4):257-264.

Sandars, D.F. 1957b. Redescription of some cestodes from marsupials. I. – Taeniidae. Annals of Tropical Medicine and Parasitology 51:317-329.

Sanderson, K.J., L.J. Pearson and J.R. Haight. 1979. Retinal projections in the Tasmanian Devil, Sarcophilus harrisii. Journal of Comparative Neurology 188:335- 346.

Sarre, S.D., A.J. MacDonald, C. Barclay, G.R. Saunders and D.S.L. 2012. Foxes are now widespread in Tasmania: DNA detection defines the distribution of this rare but invasive carnivore. Journal of Applied Ecology 50: 459–468.

Saunders, A. 2012. The occupancy of native and introduced Tasmanian carnivores in intact and fragmented landscapes. BSc Honours thesis, University of Tasmania. Not seen; as mentioned in Fancourt et al., 2015.

Saunders, G., C. Lane, S. Harris and C. Dickman. 2006. Foxes in Tasmania: a report on an incursion of an invasive species. Invasive Animals Cooperative Research Centre, University of Canberra, Canberra, ACT.

Sayer, L. 2004a. NW sanctuary hopes for devil. Hobart Mercury, 12 November 2004.

Sayer, L. 2004b. Deadly disease spreads west. Hobart Mercury, 18 November 2004.

Schaap, D. 2002. Breeding action plan for the Tasmanian Devil, Sarcophilus harrisii. Zoological Parks Board of New South Wales; 12 pp.

Schaap, D. 2002. Enriching the devil: the Tasmanian devil. The Shape of Enrichment 11(1):1–4.

6/05/2016 8:41 AM 420

Scheelings, T.F., E.C. Dobson and C. Hooper. 2014. Cutaneous T-cell lymphoma in two captive Tasmanian Devils (Sarcophilus harrisii). Journal of Zoo and Wildlife Medicine 45(2):367-371.

Schiburr, R. and L. Muschketat. 1983. Beutelteufel (Sarcophilus harrisii Boitard) im Zoo Rotterdam. Zeitschrift des Kölner Zoo 26(3):84-89.

Schliebs, M. 2005. Only a mum could love them. Daily Telegraph, 30 November 2005.

Scoleri, V., M. Jones, C. Johnson, E. Woehler and J. Lewis. 2015. Unleashing the Devil in the “Garden of Eden.” 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 29

Scott, L. 2009. Breakthrough test for devil facial tumour. The Examiner, 2 April 2009.

Shah, S.K.H. and S.C. Nicol. 1989. Cephalic vasculature and distribution of blood flow through the cranial arterial circle of the Tasmanian Devil, Sarcophilus harrisii. Journal of Mammalogy 70(1):123-131.

Shah, S.K.H., S.C. Nicol and R. Swain. 1986. Functional morphology of the cranial vasculature and the nasal passage in the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae): a marsupial carotid rete. Australian Journal of Zoology 34:125-133.

Shannon, L. and R. Lehman. 2015. Release of captive bred Tasmanian devils hailed as turning point in fight against disease. ABC News online, 26 September 2015.

Sharland, M.S.R. 1940. A naturalist’s notebook. Sydney Morning Herald, 2 May 1940.

Sharland, M. 1962. Tasmanian Wild Life. A Popular Account of the Furred Land Mammals, Snakes and Introduced Mammals of Tasmania. Melbourne University Press, Melbourne; 86 pp.

Sharland, M. 1967. Devil’s name. Tasmanian Naturalist 8: 4.

Sharland, M. 1975. Reminder of old Battery Point Zoo. The Tasmanian Naturalist 41: 8.

Sharman, A. 2012a. Forestier depopulation and isolation project. STDP, online.

Sharman, 2012b. Managing devils for species survival. Royal Society of Tasmania Winter Lecture Series 2012. Stanley Burbury Theatre, University Centre, Churchill Avenue, Sandy Bay campus; 19 June 2012.

6/05/2016 8:41 AM 421

Sharman, G.B., R.L. Hughes and D.W. Cooper. 1990. The chromosomal basis of sex- differentiation in marsupials. Australian Journal of Zoology 37(2-3):451- 466.

Shedlock, A.M. and S.V. Edwards. 2009. Amniotes (Amniota). 2009. Pp. 375-379, in S.B. Hedges and S. Kumar, eds, The Timetree of Life. Oxford University Press, Oxford; 551 pp.

Shine, T. 2014a. Tasmanian devils' low genetic diversity, extinction risks date back to ice age. ABC News, T. Shine, 5 November 2014

Shine, T. 2014b. Bid to save birds from predatory Tasmania devils on Maria Island haven. ABC 7.30 Tasmania, 28 November 2014.

Siddle, H.V. and J. Kaufman. 2013. A tale of two tumours: comparison of the immune escape strategies of contagious cancers. Molecular Immunology 55(2):190- 193 (published online 30 November 2012).

Siddle, H.V. and J. Kaufman. 2015. Immunology of naturally transmissible cancers. Immunology 114:11-20. First published online, 4 September 2014.

Siddle, H.V., A. Kreiss, M.D.B. Eldridge, E. Noonan, C.J. Clarke, S. Pyecroft, G.M. Woods and K. Belov. 2007b. Transmission of a fatal clonal tumor by biting occurs due to depleted MHC diversity in a threatened carnivorous marsupial. Proceedings of the National Academy of Sciences 104(41):16221-16226.

Siddle, H., A. Kreiss, C. Tovar, Y.Y. Cheng, K. Belov, K. Swift, A.-M. Pearse, K. Skjodte, G. Woods and J. Kaufman. 2012. Immune escape strategies of a contagious cancer, devil facial tumour disease. Abstract from 7th International EMBO Workshop on Antigen Presentation and Processing, Amsterdam, The Netherlands, April 24-27 2012. Molecular Immunology 51(1): 30.

Siddle, H.V., A. Kreiss, C. Tovar, C.K. Yuen, Y. Cheng, K. Belov, K. Swift, A.-M. Pearse, R. Hamede, M.E. Jones, K. Skjødtf, G.M. Woods and J. Kaufman. 2013. Reversible epigenetic down-regulation of MHC molecules by devil facial tumour disease illustrates immune escape by a contagious cancer. Proceedings of the National Academy of Sciences 110(13):1501-1508. (Published online 11 March 2013)

Siddle, H.V., J. Marzec, Y. Cheng, M. Jones and K. Belov. 2010a. MHC gene copy number variation in Tasmanian devils: implications for the spread of a contagious cancer. Proceedings of the Royal Society B 277:2001–2006.

Siddle, H.V., J. Marzec, M. Jones and K. Belov. 2010b. Major histocompatibility genetics of a clonal contagious cancer in the Tasmanian Devil. Abstracts 24th European Immunogenetics and Histocompatibility Conference. 17th Annual Meeting of the Italian Society for Immunogenetics and Transplantation Biology May 15-18, 2010. Florence, Italy. Tissue Antigens 75:468.

6/05/2016 8:41 AM 422

Siddle, H.V., C. Sanderson and K. Belov. 2007a. Characterization of major histocompatibility complex class I and class II genes from the Tasmanian devil (Sarcophilus harrisii). Immunogenetics 59:753-760.

Siegel, M. 2012. Dance with the Devil. Sydney Morning Herald Good Weekend, 26 May, 2012. pp. 22-25.

Simmonds, M.J., O.K. Baskurt, H.J. Meiselman, M. Pyne, M. Kakanis, E. Brenu, J. Keane, R. Christy ad A.M. Marshall-Gradisnik. 2011. Haemorheology of the eastern grey kangaroo and the Tasmanian Devil. Australian Journal of Zoology 59(1):26-34.

Sinn, D.L., L. Cawthen, S.M. Jones, C. Pukk and M.E. Jones. 2014. Boldness towards novelty and translocation success in captive-raised, orphaned Tasmanian devils. Zoo Biology 33:36-48. Published online, 28 December 2013.

Sinn, D.L., K. Macnab and A. Sharman. 2010. Save the Tasmanian free-range project, annual report January-December 2009. Save the Tasmanian Devil Program website.

Slater, G. 1993. Husbandry strategies for the Tasmanian Devil and Brush-tailed Phascogale at the Healesville Sanctuary. Pp. 73-78, in M. Roberts, J. Carnio, G. Crawshaw and M. Hutchins, eds, The Biology and Management of Australasian Carnivorous Marsupials. Metropolitan Toronto Zoo and AAZP Monotreme and Marsupial Advisory Group, Ontario, Canada; # pp.

Smeller, J.M., M.B. Bush and R.W. Custer. 1977. The immobilization of marsupials. Journal of Zoo Animal Medicine 8(3):16-20.

Smiley, S. 2013. Researchers look into whether captivity makes Tasmanian devils dumber. ABC News, 26 August 2013.

Smith, G. 1909. A Naturalist in Tasmania. Clarendon Press, Oxford; 152 pp.

Smith, H. 1993. Breeding of Tasmanian Devils (Sarcophilus harrisii) at the Adelaide Zoo. Pp. 85-90, in M. Roberts, J. Carnio, G. Crawshaw and M. Hutchins, eds, The Biology and Management of Australian Carnivorous Marsupials. Metropolitan Toronto Zoo and AAZPA, Ontario, Canada; # pp.

Smith, L. 2010. Secret life at devils’ domain. Hobart Mercury, 9 June 2010.

Smith, T. 2010. Tasmanian Devil colony shows immunity to cancer. The Associated Press; 10 March 2010.

Smith, T., A. MacFayden and R. Rose. 2001. Hormonal control of birth behaviour in the bandicoot (Perameles gunni). Physiology and Behavior 72:527-532.

Sobbe, I.H. 1990. Devils on the Darling Downs – the tooth mark record. Memoirs of the Queensland Museum 27(2):299-322.

6/05/2016 8:41 AM 423

Solomon, S. 1985. People and other aggravations: taphonomic research in Australia. B.A. (Hons) Thesis, University of New England, Armidale, NSW. [Not seen; as cited in Hall and Jones, 1990]

Speight, J. 2012. Video: Little devils are 'dream come true.' Newcastle Herald, 25 July 2012.

Spratt, D.M., I Beveridge and E.L. Walter. 1991. A catalogue of Australasian monotremes and marsupials and their recorded helminth parasites. Records of the South Australian Museum Monograph Series 1:1-105.

Sprent, J.F.A.1970. Baylisascaris tasmaniensis sp.nov. in marsupial carnivores: heirloom or souvenir? Parasitology 61:75-86.

Sprent, J.F.A., J. Lamina and A. McKeown. 1973. Observations on migratory behaviour and development of Baylisascaris tasmaniensis. Parasitology 67(1):67-83.

Springer, M.S., C.W. Krajewski and R.W. Meredith. 2009. Marsupials (Metatheria). Pp. 467-470-474, in S.B. Hedges and S. Kumar, eds, The Timetree of Life. Oxford University Press, Oxford; 551 pp.

Squires, M. 2012. Bring Tassie devils to Otways. Geelong Advertiser, 26 March 2012.

Squires, N. 2003. Tasmanian devil faces extinction in virus riddle. The Telegraph, 16 August 2003.

Srb, C. 1999. Tasmanian devil captive management group. 1999 review. [Tasmanian Devil Studbook]. Australasian Species Management Program, Healesville Sanctuary, Healesville, Victoria.

Srb, C. 2012. Tasmanian devil Studbook. Healesville Sanctuary on behalf of the Zoo and Aquarium Association.

Stalder, K. J. 2013. Herpesviruses in Australian marsupials. Masters Research thesis, Faculty of Veterinary Science, The University of Melbourne. Not seen.

Stalder, K, P.K. Vaz, J.R. Gilkerson, R. Baker, P. Whiteley, N. Ficorilli, L. Tartarczuch, T. Portas, K. Skogvold, G.A. Anderson and J.M. Devlin. 2015. Prevalence and clinical significance of herpesvirus infection in populations of Australian marsupials. PLoS ONE 10(7): e0133807. doi:10.1371/journal.pone.0133807

Stannard, H.J., P. Thompson, B.M. McAllan, D. Raubenheimer. 2016. Hematology and serum biochemistry reference ranges of healthy captive Tasmanian devils (Sarcophilus harrisii) and their association with age, gender and seasonal variation.

Mammalian Biology - Zeitschrift für Säugetierkunde #(#):#-#. Published online 13 April 2016.

6/05/2016 8:41 AM 424

Stapleton, J. 2008a. Zookeepers thrilled at birth of four female Tasmanian devils. The Australian, 18 October 2008.

Stapleton, J. 2008b. Young she devils give hope for dying breed. The Week-end Australian, 18-19 October 2008.

STDP. 2010a. Save the Tasmanian Devil Program Appeal. [5 May 2010]

STDP. 2010b. Save the Tasmanian Devil Program Teacher Kit. [July 2010. Apparently written by K.M. Nolan]

STDP. 2010c. Strategic plan. [29 July 2010. Document properties indicate author was K.M. Nolan. This is an updated version of DPIW, 2007a] STDP. 2010d. Business plan, 2010-2013. November 2010.

STDP. 2011a. Save the Tasmanian Devil monitoring strategy; 3 pp. March 2011.

STDP. 2011b. Translocation proposal. Proposal to establish a managed population of Devil Facial Tumour Disease (DFTD)-free Tasmanian Devils (Sarcophilus harrisii) to Maria Island; 67 pp. November 2011.

STDP. 2012a. Development Proposal and Environmental Management Plan. The establishment of a disease-free population of Tasmanian Devils (Sarcophilus harrisii) on Maria Island National Park. March 2012. With Appendices 1 (November, 2011), 2 (August 2011), 3 (2007) and 4 a and b (November, 2011, both). Appendix 1 is STDP 2011b; Appendix 3 is Jones and McCallum, 2007 and Appendix 4a and b is Jones and McCallum, 2011a. Appendices 1, 3-4 are cited specifically.

STDP. 2012b. Response to the Save the Tasmanian Devil Program Mid-term External Review. August 2012; 5 pp.

STDP. 2014a. Maria Island Devil Monitoring Trip, 2014 Summary Report; 3 pp.

STDP. 2014b. EAA Recommendations Response Table; 2 pp.

STDP. 2014c. Business Plan 2014- 2019. Version 1.1; 17 pp. December 2014.

STDP, 2015a. Save the Tasmanian Devil Program-Maria Island Project Summary. (Document created, 2 September 2016. S. Wise indicated as author in “Document Properties.”]

STDP. 2015b. Peninsula devil conservation project. Creating a wild devil recovery project. 2 pp. [Document created, 14 December 2015.]

Stedman, M. 2007. More money needed for devils to survive. The Examiner, 4 January 2007.

Steel, R. 2010. Recent Australian tick rickettsioses discoveries and the threatened Tasmanian devil (Sarcophilus harrisii). Australian Veterinary Journal 88(3):63–63.

6/05/2016 8:41 AM 425

Stefan, C. 1999. Tooth enamel structure of some Australian carnivorous marsupials. Alcheringa 23:111-132.

Stephenson, N.G. 1963. Growth gradients among fossil monotremes and marsupials. Palaeontology 6(4):615-624.

Stewart, M. 2013. Wellington Zoo in line to get four devilish 'ambassadors.' Dominion Post, 26 June 2013.

Stewart, N.J., S.S. Bettiol, A. Kreiss, N. Fox and G.M. Woods. 2008. Mitogen- induced responses in lymphocytes from platypus, the Tasmanian Devil and the eastern barred bandicoot. Australian Veterinary Journal 86(10):408-415.

Stirton, R.A. 1957. Tertiary marsupials from Victoria, Australia. Memoirs of the National Museum 21:121-134.

Stokes, K. 2015. Meet Monarto Zoo’s two new Tasmanian devil babies. The Advertiser, 25 August 2015.

Stratham, M. 2006. The development of 1080 use for rabbit control in Tasmania. Papers and Proceedings of the Royal Society of Tasmania 139:1-6.

STTDP. 2007a. Strategic plan. 2 pp. [24 July 2007]

STTDP. 2007b. Insurance population strategy. 4 pp. [24 July 2007]

STTDP. 2007c. Business Plan 2007-2008. 33 pp. [15 November 2007]

STTDP. 2008a. Biosecurity guidelines. 3 pp. [21 April 2008]

STTDP. 2008b. Risk categorisation. Guidelines for relocations of captive Tasmanian Devils. Version 2.2. [21 April 2008]

STTDP. 2008c. Save the Tasmanian Devil program Scientific Research Strategy. Version 1A; 10 pp. [1 September, 2008; document properties identify H. McCallum as author]

STTDP. 2008d. Policy and process for review of research and management proposals relating to Tasmanian devils and requests for collaboration or samples. [1 September 2008] Sutherland, L. 2003. Tasmanian devil faces deadly bug. Daily Telegraph, 15 October 2003.

Swift, K. 2007. Evidence for direct infection. P. 9, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Taggart, D.A., G.A. Shimmin, C.R. Dickman and W.G. Breed. 2003. Reproductive biology of carnivorous marsupials: clues to the likelihood of sperm competition. Pp. 358-375, in M. Jones, C.R. Dickman and M. Archer, eds, Predators with Pouches. The

6/05/2016 8:41 AM 426

Biology of Carnivorous Marsupials. CSIRO Publishing, Collingwood, Victoria; 486 pp.

Tancred, E. 1981. The distribution and sizes of ganglion cells in the retinas of five Australian marsupials. Journal of Comparative Neurology 196: 585–604.

Tassy, E. 2013. Ready to meet the public. Albuquerque Journal News, 28 December 2013.

Tate, G.H.H. 1947. On the anatomy and classification of the Dasyuridae (Marsupialia). Bulletin of the American Museum of Natural History 88:101-155.

Taylor, J. 1920. “Zoo.” Studies from Nature. James Taylor, Sydney; 118 pp.

Taylor, P. 2011. The Devil we have to have. Australian Forest Grower 34(1):19.

Taylor, R., R. Mesibov, R. Brereton and K. Bonham. 1997. Terrestrial fauna of Cataract Gorge Reserve, Launceston. Tasmanian Naturalist 119: 46-58.

Taylor, R., M. Jones, R. Hamede, J. Bowman and M. Ambrose. The microbial ecology of transmission sires of Devil Facial Tumour Disease. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 31.

Taylor, R.J. 1985. Identification of the hair of Tasmanian mammals. Papers and Proceedings of the Royal Society of Tasmania 119:69-82.

Taylor, R.J. 1986. Notes on the diet of the carnivorous mammals of the Upper Henty Region, Western Tasmania. Papers and Proceedings of the Royal Society of Tasmania 120:7-10.

Taylor, R.J., S.L. Bryant, D. Pemberton and T.W. Norton. 1985. Mammals of the Upper Henty River Region, western Tasmania. Papers and Proceedings of the Royal Society of Tasmania 119:7-14.

Taylor, R.J. and P.B. McQuillan. 1994. Fauna of Mount Wellington. Tasmanian Naturalist 116: 2-19.

Taylor, R.J. 1989. Mammals of Tasman Peninsula and their interaction with Europeans. Pp. 81-87, in A.J. Smith, ed., Is History Enough? Past, Present and Future Use of Resources of Tasman Peninsula. Royal Society of Tasmania, Hobart.

Tedmanson, S. 2009. Test may protect Tasmanian devil from extinction. The Times Online, 3 April 2009.

Temminck, C.J. 1827. Monographies de Mammologie. Dufour & d’Ocagne. Paris; 352 pp.

6/05/2016 8:41 AM 427

Thalmann, S. 2013. Evaluation of a degradable time-release mechanism for telemetry collars. Australian Mammalogy 35(2):241-244.

Thalmann, S., S. Peck, P. Wise, J. Potts, J. Clarke and J. Richley. 2015. Translocation of a top-order carnivore: Tracking the initial survival, spatial movement, home range establishment and habitat use of Tasmanian devils on Maria Island. Australian Mammalogy #.

Thalmann, S., P. Wise and S. Huxtable. 2014. Sentinel camera traps monitor the emergence of infectious disease in Tasmanian Devils. Pp. 27-33, in P. Meek, P. Fleming, G. Ballard, P. Banks, A. Claridge, J. Sanderson and D. Swann, eds, Camera Trapping: Wildlife Management Research. CSIRO Publishing: 392 pp.

Thomas, A. and P. Willis. 2001. Devils get down and dirty. ABC Science. Scribbly Gum. 3 April 2001.

Thomas, O. 1887. On the homologies and succession of the teeth in the Dasyuridae, with an attempt to trace the history of the evolution of the Mammalian teeth in general. Philosophical Transactions of the Royal Society B 128:443-462.

Thomas, O. 1903. Note on the technical name of the Tasmanian Devil. Annals and Magazine of Natural History 7(11): 289.

Thomas, O. 1912. The technical name of the Tasmanian Devil. Proceedings of the Biological Society of Washington 25:116.

Thomas, R.H., W. Schaffer, A.C. Wilson and S. Pääbo. 1989. DNA phylogeny of the extinct marsupial wolf. Nature 340:465-467.

Thompson, C.K., S.S. Godfrey and R.C.A. Thompson. 2014. Trypanosomes of Australian mammals: A review. International Journal for Parasitology: Parasites and Wildlife 3(2):57-66.

Thompson, J., R. Yang, M. Power, J. Hufschmid, I. Beveridge, S. Reid, J. Ng, A. Armson and U. Ryan. 2008. Identification of zoonotic Giardia genotypes in marsupials in Australia. Experimental Parasitology 120(1): 88-93.

Thomson, S.C. 1940. Studies of the anatomy of the extra-hepatic biliary tract in Mammalis. Zoological Series. Field Museum of Natural History 22(6):415-449.

Thornly, M. 2004. Devils advocates. Australian Veterinary Journal 82(10):584.

Tindale, N.B. 1957. Culture succession in south-eastern Australia from Late Pleistocene to the present. Records of the South Australian Museum 12:1-49.

Todd, M. 2008. Devil of a time for Symbio’s little Tassies. Illawarra Mercury, 26 July 2008.

6/05/2016 8:41 AM 428

Tompkins, D.M. 2007. Modelling the impact of DFTD. P. 20, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Tovar, C. 2012. Insights into the nature and immune escape of a transmissible cancer in Tasmanian Devils. PhD Thesis, Menzies Research Institute, University of Tasmania; 231 pp.

Tovar, C., D. Obendorf, E.P. Murchison, A.T. Papenfuss, A. Kreiss and G.M. Woods. 2011. Tumour-specific diagnostic marker for transmissible facial tumours of Tasmanian Devils: immunohistochemistry studies. Veterinary Pathology 48(6):1195- 1203.

Towie, N and P. Duncan, 2010. Devils go west to give species future hope. Hobart Mercury, 16 May 2010

Toy, M. 2011. Hopes fading for Tassie Devil. Herald Sun, 16 January 2011.

Travers, T., C. Johnson and M. Jones. 2015. Predator diversity, habitat heterogeneity and the landscape of fear for small mammals. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Poster presentation abstract, p. 32.

Tribe and Middleton. 1988. Marsupial diseases. Pp. 789-814, in D.I. Bryden, ed., Australian Wildlife. The John Keep Refresher Course for Veterinarians. Proceedings 104. Post Graduate Committee in Veterinary Science, University of Sydney; 952 pp (exclusive of index).

Triggs, B. 1996. Tracks, Scats and Other Traces: A Field Guide to Australian Mammals. Oxford University Press, Melbourne; 340 pp.

Trofimov, Y. 2008. Tasmanian Devils get reprieve in Australia’s old penal colony. The Wall Street Journal, Asia Edition, 28 May 2008.

Troughton, E. 1938. Australian mammals. Their past and future. Journal of Mammalogy 19(4):401-411.

Troughton, E. 1967. Furred Animals of Australia. Angus and Robertson, Sydney; 384 pp.

Troy, S.N. 2014. Spatial ecology of the Tasmanian spotted-tailed quoll. PhD thesis, University of Tasmania. Not seen; as mentioned in Fancourt et al., 2015.

Tucker, A. 2012. What is killing the Tasmanian Devil? Smithsonian, May 2012.

Turner, K. 1970. Breeding Tasmanian Devils, Sarcophilus harrisii, at Westbury Zoo. International Zoo Yearbook 10:65.

6/05/2016 8:41 AM 429

Twin, J.E. and A.M. Pearse. 1986. A malignant mixed salivary tumour and a mammary carcinoma in a young wild eastern spotted native cat Dasyurus viverrinus (Marsupialia). Journal of Comparative Pathology 96(3): 301-306.

Tyler, J., W. Boyles and L. Wilson. 2011. Devil in crisis. Fourtysouth 58:23-28.

Ujvari, B. and K. Belov. 2015. Characterization of antibody V segment diversity in the Tasmanian devil (Sarcophilus harrisii). Veterinary Immunology and Immunopathology 167(3-4):156-165. Published online, 14 August 2015.

Ujvari, B., A.T. Papenfuss and K. Belov. 2012. Devil facial tumour disease: evolution of a fatal cancer. Abstract, 6th Australian Health and Medical Research Congress 2012, 28 November 2012.

Ujvari, B., A.T. Papenfuss and K. Belov. 2016. Transmissible cancers in an evolutionary context. Inside the Cell 1(1):17-26. Published online, 7 July 2015.

Ujvari, B., A.-M. Pearse, M. Eldridge and K. Belov. 2011. The evolution of Devil Facial Tumour Disease (DFTD). Abstract. Genetics Society of AustralAsia Annual Conference 2011. Melbourne, July 10-13th, 2011, p. 104.

Ujvari, B., A.M. Pearse, S. Peck, C. Harmsen, R. Taylor, S. Pyecroft, T. Madsen, A.T. Papenfuss and K. Belov. 2013. Evolution of a contagious cancer: epigenetic variation in Devil Facial Tumour Disease. Proceedings of the Royal Society B 280(1750):#-#. Published online, 7 November 2013.

Ujvari, B., A.-M. Pearse, K. Swift, P. Hodson, B. Hua, S. Pyecroft, R. Taylor, R. Hamede, M. Jones, K. Belov and T. Madsen. 2014a. Anthropogenic selection enhances cancer evolution in Tasmanian devil tumours. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 84.

Ujvari, B., A.-M. Pearse, K. Swift, P. Hodson, B. Hua, S. Pyecroft, R. Taylor, R. Hamede, M. Jones, K. Belov and T. Madsen. 2014b. Anthropogenic selection enhances cancer evolution in Tasmanian devil tumours. Evolutionary Applications 7:260-265.

Ujvari, B., A.-M. Pearse, R. Taylor, S. Pyecroft, C. Flanagan, S. Gombert, A.T. Papenfuss, T. Madsen and K. Belov. 2012. Telomere dynamics and homeostasis in a transmissible cancer. PloS ONE 7(8):e44085.

Ujvari, B., L. Piddington, A.-M. Pearse, S. Peck, C. Harmsen, R. Taylor, S. Pyecroft, M. Kowarsky, T. Madsen, A.T. Papenfuss and K. Belov. 2014c. Devil Facial Tumor Disease, a potential model of the cancer stem-cell process? GSTF International Journal of Veterinary Science 1(1): 43-49. van der Kraan, L.E. 2011. Daring to be different: The evolution of NK receptors in the endangered marsupial, Tasmanian devil. Semester 1 Honours & Graduate Diploma Seminar Abstracts, pp. 17-18.

6/05/2016 8:41 AM 430 van der Kraan, L.E., E. S.W. Wong, N. Lo, B. Ujvari and K. Belov. 2013. Identification of natural killer cell receptor genes in the genome of the marsupial Tasmanian devil (Sarcophilus harrisii). Immunogenetics 65(1):25-35.

Venditti, C. A. Meade and M. Pagel. 2011. Multiple routes to mammalian diversity. Nature online, 17 November 2011.

Vetter, W., R. von der Recke, R. Symons and S. Pyecroft. 2008. Determination of polybrominated biphenyls in Tasmanian devils (Sarcophilus harrisii) by gas chromatography coupled to electron capture negative ion tandem mass spectrometry or electron ionization high-resolution mass spectrometry. Rapid Communications in Mass Spectrometry 22:4165-4170.

Vilcins, I.-M.E., J.M. Old and E. Deane. 2009. Detection of a hepatozoan and spotted fever group Rickettsia species in the common marsupial tick (Ixodes tasmani) collected from wild Tasmanian Devils (Sarcophilus harrisii), Tasmania. Veterinary Parasitology 162:23-31.

Vinales, P. 2016. Push to end Tasmanian devil road deaths with 'negative conditioning' trial. ABC News online, 2 March 2016.

Vincent, D. 2012. Zoos ‘phasing’ out some animals. Toronto Star, 1 June 2012.

Vogelnest, L. and G. Allan. 2015. Radiology of Australian Mammals. CSIRO Publishing, Clayton South, Victoria; # pp.

Vrolik, M.W.1852. Annals and Magazine of Natural History (2)9:245-246.

Wade, R. 2003a. Libs on board as Labor injects $1.8 million. Hobart Mercury 13 November 2003.

Wade, R. 2003b. Tumour research to be tackled by Tassie academics, Hobart Mercury, 14 November 2003.

Wade, R. 2005a. Who’s a lucky devil. Hobart Mercury, 19 February 2005.

Wade, R. 2005b. Island animals carry hope. Hobart Mercury, 25 February 2005.

Wakefield, N.A. 1964a. Recent mammalian sub-fossils of the basalt plains of Victoria. Proceedings of the Royal Society of Victoria 77(2):419-425.

Wakefield, N.D. 1964b. Mammal remains. Proceedings of the Royal Society of Victoria 77(2):494-498.

Walch, G. 1890. Huddart, Parker and Co’s Guide to Tasmania. Huddart, Parker and Company, Melbourne; 120 pp.

Walker, E.P. 1964. Mammals of the World. Vol. 1. Johns Hopkins Press, Baltimore; 644 pp.

6/05/2016 8:41 AM 431

Walshe, K. 1994. Tasmanian Devils, Sarcophilus harrisii, and human occupation. Estimating the cultural component. Pp. 375-380, in M. Sullivan, S. Brockwell and A. Webb, eds, Archaeology in the North, Proceedings of the 1993 Australian Archaeological Association Conference. North Australian Research Unit, ANU, Darwin; 415 pp.

Walshe, K. 1998. Taphonomy of Mungo B assemblage: indicators for subsistence and occupation of Lake Mungo. Archaeology in Oceania 33:200-206

Walshe, K. 2000. Carnivores, taphonomy and dietary stress at Puntutjarpa, Serpent’s Glen and Intitjikula. Archaeology in Oceania 35:74-81.

Wang, L. 2007. Identification of infectious pathogens via molecular biological approaches. P. 13, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Ward, B. 2011. Devilish fun all round at zoos. Hobart Mercury, 24 August 2011.

Ward, S.J. 2003. Molecular systematics of Dasyuromorphia. Pp. 376-382, in M. Jones, C. Dickman and M. Archer, eds, Predators with Pouches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melbourne; 486 pp.

Warner, G. 2001. New dish on menu to catch cunning fox. Hobart Mercury, 24 July 2001.

Warren, J.V. 2013. The Devil Undone. The Science and Politics of Tasmanian Devil Facial Tumour Disease. PhD Thesis, University of Wollongong; 337 numbered pages + 15 un-numbered pages of appendices.

Warren, J. and B. Martin. 2014. What’s killing Tassie devils if it isn’t a contagious cancer? The Conversation, 17 November 2014.

Waterhouse, B. 2015. Pride of Australia: Professor gives devils a chance. Hobart Mercury, 19 July 2015.

Waterhouse, C. 2007. Million-dollar rescue. Hobart Mercury, 4 January 2007.

Waterhouse, G.R. 1846. A Natural History of the Mammalia. Vol. 1. Marsupialia, or Pouched Mammals. Baillière, London; 553 pp.

Waterworth, B. 2015. Pride of Australia: Professor gives devils a chance Hobart Mercury, 19 July 2015.

Weigel, J. 2016. Devil Ark. Giving hope to the Tasmanian Devil. Australian Wildlife, vol. 2, 2016.

Weiss, M. and P. G. Richards. 1971. Adrenal steroid secretion in the Tasmanian devil (Sarcophilus harrisii) and the eastern native cat (Dasyurus viverrinus). A comparison

6/05/2016 8:41 AM 432

of adrenocortical activity of different Australian marsupials. Journal of Endocrinology 49:263–275.

Wells, B. 2010. Welfare issues with Tasmanian Devil research. 2010 ANCCART Conference, 20-22 July 2010, Hobart, Tasmania, p. 13.

Wells, R.T. 1978. Fossil mammals in the reconstruction of Quaternary environments with examples from the Australian fauna. Pp. 103-124, in D. Waler and J.C. Guppy, eds, Biology and Quaternary Environments. Australian Academy of Science, Canberra, 264 pp.

Werdelin, L. 1986. Comparison of skull shape in marsupial and placental carnivores. Australian Journal of Zoology 34(2) 109-117.

Werdelin, L. 1987. Some observations on Sarcophilus laniarius and the evolution of Sarcophilus. Records of the Queen Victoria Museum, Launceston. No. 90; 27 pp.

Westerman, M., C. Krajewski, B.P. Kear, L. Meehan, R.W. Meredith, C.A. Emerling and M.S. Springer. 2015. Phylogenetic relationships of dasyuromorphian marsupials revisited. Published online, 10 November 2015. DOI: 10.1111/zoj.12323.

Westerman, M, and P.A. Woolley, 1989. Cytogenetics of some New Guinean dasyurids and genome evolution in the Dasyuridae (Marsupialia). Australian Journal of Zoology 37(5) 521-531.

Whinnett, E. 2003a. All-out effort to save an icon. Hobart Mercury, 3 September 2003.

Whinnett, E. 2003b. VDL Company finds funds for research. Hobart Mercury, 16 December 2003.

Whinnett, E. 2005a. Devil rescue. Hobart Mercury, 5 February 2005.

Whinnett, E. 2005b. Tassie creates its own devils’ island. Herald-Sun, 5 February 2005.

White, A. 2013. Sunday Herald Sun readers can help name Healesville Sanctuary's newest little devil. 28 July 2013.

White, L. 2015. Cute little devils brighten day for children at Lismore. Northern Star, 27 August 2015.

Widowson, H. 1829. Present state of Van Diemen's Land; comprising an account of its agricultural capabilities, with observations on the present state of farming, &c. &c. pursued in that colony: and other important matters connected with emigration. S. Robinson, Chapter-House Passage, St. Paul’s; W. Joy, St. Paul’s Church Yard; J. Cross, Holborn; and J. Birdsall, Northampton, London; 200 pp. Available online.

Wild, M. and K. Argent. 2007. Ruby Roars. Allen and Unwin, Sydney; 32 pp.

6/05/2016 8:41 AM 433

Wilkie, A.A. and A.R. Osborn. 1920. Almost human. Reminiscences from the Melbourne Zoo. Told by A.A.W. Wilkie; written by Mrs. A.R. Osborn. Whitcombe & Tombs Ltd, Melbourne; 237 pp.

Wylie, F. 2010. Detail in the devil. Australian Life Scientist 7(3):44, 46.

Williams, B. 1986. Mandibular glands in the endoparasitic larva of Uropsylla tasmanica Rothschild (Siphonaptera: Pygiopsyllidae). International Journal of Insect Morphology and Embryology 15(4):263-268.

Williams, B. 1991. Adaptations to endoparasitism in the larval integument and respiratory system of the flea Uropsylla tasmanica Rothschild (Siphonaptera: Pygiopsyllidae). Australian Journal of Zoology 39:77-79.

Williams, D.L.G. 1980. Catalogue of Pleistocene vertebrate fossils and sites in South Australia. Transactions of the Royal Society of South Australia 104(5-6):101-15.

Williams, D.L.G. 1981. Genyornis eggshell (Dromornithidae; Aves) from the Late Pleistocene of South Australia. Alcheringa 5(2):133–140.

Wintle, S.H. 1886. The fossil mammalian remains of Tasmania compared with those of the Australian mainland. Victorian Naturalist 1886:27-34, 40-45 and 65-68.

Wisbey, M. 2015. Two more vaccinated devils killed. The Examiner, 3 October 2015

Wood, D. 2003a. Devil of a plague, Devil of a problem for top predator. Hobart Mercury, 29 March 2003.

Wood, D. 2003b. The devil you know may be down below. Hobart Mercury, 15 July 2003.

Wood, D. 2003c. Terrible disease stalks an icon. Hobart Mercury, 1 September 2003

Wood, D. 2003d. The devil and the deep too-hard basket. Pp. 497-506, in P. Craven, ed., The Best Australian Essays 2003. Black Inc., Melbourne; 557 pp.

Woods, G. 2008. Cedric – a different devil. Tasmania 40º South No. 50:32-34.

Woods, G. 2010. Devil Facial Tumour Disease: a Truly Transmissible Cancer. RiAus Cutting Edge Lecture, 24 May 2010 [online video].

Woods, G. 2012a. The nature of Tasmanian Devil Facial Tumour Disease. Royal Society of Tasmania Winter Lecture Series 2012. Stanley Burbury Theatre, University Centre, Churchill Avenue, Sandy Bay campus; 19 June 2012.

Woods, G. 2012b. How the devil did it happen? Australian Society of Anaesthetists. Abstract. P. 124, in 71st National Scientific Congress, 28 September- 2 October 2012, NSC Handbook, Hobart, Tasmania.

6/05/2016 8:41 AM 434

Woods, G. 2014. Tassie devil facial tumour is a transmissible cancer. The Conversation, 19 November 2014.

Woods, G. 2015. Devil Facial Tumour Disease – a transmissible cancer of ignorance. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 78.

Woods, G.M., G.K. Brown, T. Pinfold, C.D. Tovar and A. Kreiss. 2012. Devil facial tumour disease and recent progress. Abstract, 6th Australian Health and Medical Research Congress 2012, 28 November 2012.

Woods, G.M., L.J. Howson, G.K. Brown, C. Tovar, A. Kreiss, L.M. Corcoran and A.B. Lyons. 2015. Immunology of a transmissible cancer spreading among Tasmanian Devils. Journal of Immunology 195(1): 23-29.

Woods, G.M. and A. Kreiss. 2007. A vaccine? P. 10, in Devil Facial Tumour Disease. Senior Scientist’s Scientific Forum, 20-22 February 2007, Sandy Bay Campus, University of Tasmania; 34 pp. Abstract.

Woods, G.M., A. Kreiss, K. Belov, H.V. Siddle, D.L. Obendorf and H.K. Muller. 2007. The immune response of the Tasmanian Devil (Sarcophilus harrisii) and devil facial tumour disease. EcoHealth 4(3):338-345.

Woods, G.M., A. Kreiss, C. Tovar and G. Brown. 2010a. Devil facial tumour disease. 2010 ANZCCART Conference, 20-22 July 2010, Hobart, Tasmania, p. 31. Abstract.

Woods, G.M., A. Kreiss, C. Tovar and G. Brown. 2010b. Devil facial tumour disease: A transmissible cancer that avoids immune detection. 14th International Congress of Immunology Kobe, Japan, 23-27 August 2010. Abstract. International Immunology Meeting Abstracts (2010) 22 (Supplement 1, Part 3): iii144-iii172. doi: 10.1093/intimm/dxq269

Woods, R. and A. Sharman, 2010. Update on the Save the Tasmanian Devil Program. Animal Health Surveillance 15(1):2-3.

Wray, A. 2006. Ark all part of a devilish plan. Brisbane Courier-Mail, 5 December 2006.

Wright, B., R. Hamede, B. Ujvari, U. Cheng, A. Lane, J. Donald, M. Jones and K. Belov. 2011. Evaluating genetic variation of infection status of Devil Facial Tumour Disease in a population of Tasmanian Devils. Abstract. Genetics Society of AustralAsia Annual Conference 2011. Melbourne, July 10-13th, 2011, p. 177.

Wright, B., K. Morris, C. Willet, C. Hogg, R. Hamede Ross, M. Jones, C, Wade and K. Belov. 2014. A SNP-based approach to determining pedigrees and assessing genetic diversity in the Tasmanian devil insurance population. Genetics Society of Australasia Conference 6-9 July 2014, University of Sydney. Abstract # 26.

6/05/2016 8:41 AM 435

Wroe, S., M. Ebach, S. Ahyong, C. de Muizon and J. Muirhead. 2000. Cladistic analysis of dasyuromorphian (Marsupialia) phylogeny using cranial and dental characters. Journal of Mammalogy 82(4):1008-1024.

Wroe, S. and C. Johnson. 2003. Bring back the Devil? Nature Australia 27(11):84.

Wroe, S. and B.S. Mackness. 2000. A new genus and species of dasyurid from the Pliocene Chinchilla Local Fauna of south-eastern Queensland. Alcherigna 24:319- 326.

Wroe, S., C. McHenry and J. Thomason. 2004. Bite club: comparative bite force in big biting mammals and the prediction of predatory behaviour in fossil taxa. Proceedings of the Royal Society B 272:619-625.

Wylie, F. 2010. Detail in the Devil. Australian Life Scientist 7(3): 44, 46.

Wynwood, S.J., M. Burns, G.C. Graham, D.B. McKay, S.B. Craig and S. Peck. 2015. Leptospirosis in Tasmanian devils. 61st Annual Scientific Meeting of the Australian Mammal Society, University of Tasmania, Hobart, Tasmania; 6-10 July 2015. Abstract, p. 79.

Yang, T.J., J.P. Chandler and S. Dunne-Anway. 1987. Growth stage dependent expression of MHC antigens on the canine transmissible venereal sarcoma. British Journal of Cancer 55(2):131-134.

Young, E. 2003. Facial tumours kill a third of all Tasmanian devils. New Scientist, 25 October 2003 (2418): 9.

Young, E. 2005. Tasmanian Devil’s bite is the kiss of death. New Scientist 9:14. (6 July 2005).

Young, E. 2016. The curious case of a contagious cancer, The Atlantic, 7 January 2016.

Young, G.J., J.A.M. Graves, I. Barbieri, P.A.Woolley, D.W. Cooper and M. Westerman. 1982. The chromosomes of dasyurids (Marsupialia). Pp. 783-795, in M. Archer, ed., Carnivorous Marsupials. Sydney: Royal Zoological Society of New South Wales; 804 pp.

Young, K. 2009. Boost for Tasmanian devil. Hobart Mercury, 6 June 2009.

ZAA. 2010. Annual Report and Recommendations for 2010. [Not seen.]

Zaw, Y. 2012. Shy, sassy devils are good news. The West Australian, 29 August 2012.

Zemann,A., G. Churakov, S. Donnellan. F. Grützner, F. Zhao, J. Brosius and J. Schmitz. 2013. Ancestry of the Australian termitivorous Numbat. Molecular Biology and Evolution 30(5):1041–1045.

6/05/2016 8:41 AM 436

Zimmer, C. 2013. Raising Devils in seclusion. New York Times, 21 January 2013.

6/05/2016 8:41 AM