(Morelia Spilota) Is an Important Cause of Mortality in a Free‐Living Koala

Home , Morelia (snake), Morelia spilota

Robbins Amy (Orcid ID: 0000-0001-9231-0636)

1 Carpet pythons: important predators of koalas Predation by carpet pythons (Morelia spilota) is an

important cause of mortality in a free-living koala

(Phascolarctos cinereus) population in South East

Queensland

Amy Robbins1*, Joanne Loader1, Deidre de Villiers1,

Hawthorne L. Beyer2 and Jonathan Hanger1

1 Endeavour Veterinary Ecology Pty Ltd, 1695 Pumicestone Road, Toorbul,

4510, Queensland, Australia

2 School of Biological Sciences, University of Queensland, St. Lucia, 4072,

Queensland, Australia

*corresponding author Amy Robbins

[email protected]

1695 Pumicestone Road, Toorbul, Queensland, 4510.

Abstract

This is the author manuscript accepted for publication and has undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/avj.12840

Koalas (Phascolarctos cinereus) are experiencing significant declines across

the northern part of their range. However, unbiased, population-level

estimates of mortality are rarely reported, as it’s difficult to quantify causes of

mortality robustly in this cryptic species. We aimed to determine the relative importance of carpet python (Morelia spilota) predation in a free-living koala

population and describe the characteristic pathological findings during

necropsy. In total, 503 koalas were captured, underwent veterinary examination and telemetric tagging, and were monitored after release over a

four-year period. Mortalities were detected when activity data reported by K-

Tracker® biotelemetry collars indicated low or zero activity, or during routine

field monitoring events. Experienced koala veterinarians performed thorough,

standardised necropsy examinations on retrieved carcasses to determine

causes of death. The three, sometimes subtle, cardinal signs used to

definitively diagnose carpet python-caused deaths of koalas were a U-shaped

primary bite site, slicking of the fur by python saliva (particularly around the

face), and diffuse, uniform pulmonary congestion. We found that carpet

pythons were important predators of koalas, second only to wild dogs

(dingoes and dingo hybrids (Canis familiaris dingo)), accounting for 11.6% of

predation deaths and 7.2% of total deaths. Less than half (38%) of the koalas

killed by carpet pythons were ingested. Though carpet pythons are known

predators of koalas, their relative importance as a cause of mortality hasn’t

previously been recognised. Population viability analyses and conservation

management plans benefit from robust cause-of-death data collected during longitudinal monitoring studies, requiring telemetry methods that facilitate rapid detection of mortalities.

Keywords

Koala; predation; python; pathology; ecology; population management

Introduction

Koalas (Phascolarctos cinereus) are experiencing significant declines across

the northern part of their range due to a variety of threats, including habitat

destruction, trauma from cars and predators, climate change and disease.1, 2,

3, 4, 5 Understanding the relative importance of the different causes of

mortality that are driving these declines is essential for informing conservation

and management, yet this has rarely been estimated at the population level.

Previous estimates of the common causes of mortality for koalas have largely

been derived from wildlife hospital records, which are inherently biased, with

over-representation of koalas displaying obvious signs of illness and those

found injured or dead near human habitation or roads. Predation is rarely

recorded as a cause of death in hospital records, other than with respect to

traumatic injuries caused by domestic dogs. A comprehensive and

mechanistic understanding of causes of mortality and their relative frequency

is important for informing management strategies that result in meaningful

outcomes for koala populations.6

The relative importance of predation on koala populations, particularly by

native predators, is largely unknown, because remains are rarely discovered

and predation is rarely observed. An analysis of 17 years of wildlife hospital

records from South East Queensland by Gonzalez-Astudillo et al.7 reported

that animal attack trauma occurred at a frequency of 5.2% among submitted

koalas (n=20,250). Most of these appear to have been attributed to dogs and livestock, but it was not clear whether domesticated or wild animals were implicated. There are, in fact, a wide variety of predators of koalas including domestic dogs (Canis familiaris), dingoes and dingo hybrids (Canis familiaris dingo), foxes (Vulpes vulpes), lace monitors (Varanus varius), carpet pythons

(Morelia spilota) and birds of prey such as wedge-tailed eagles (Aquila audax), powerful owls (Ninox strenua) and white-bellied sea eagles

(Haliaeetus leucogaster).8, 9, 10, 11, 12 Multi-year, population-scale studies of

free-living koalas in which robust datasets on causes of mortality have been

reported are lacking and are required if predation rates are to be accurately

quantified.

Although carpet pythons are recognised predators of koalas, python predation

has not previously been reported as an important cause of mortality for the

koala. The subspecies of carpet python occurring in South East Queensland

is M. spilota mcdowelli. Attaining a snout-vent length of >4m (mean 1.76m for females and 1.81m for males) and a weight of 11kg, it is the largest python

species in this region.13 A semi-arboreal, predominantly ambush predator, their varied diet consists of a range of prey species such as native and non- native birds, small mammals and reptiles, including domestic pets such as cats and ducks weighing 4 - 5kg in one study.13 Carpet pythons are regularly

observed in and around human habitation close to areas of remnant native

bushland, and their exploitation of non-native prey species suggests these

habitats are likely to support increased densities of carpet pythons (>6

snakes/ha).13, 14 It is therefore reasonable to anticipate that carpet pythons

could have important impacts on koala population dynamics in these habitats.

In this study, we undertook a multi-year, detailed demographic and veterinary

study of a koala population over a period of four years to quantify the causes

of mortality in that population and provide detailed diagnostic guidance to

assist in the identification of python predation in future field studies.

Materials and methods

Koala management program

A koala management program by the Queensland Department of Transport

and Main Roads was conducted between 2013 and 2017 to mitigate and

manage potential impacts of the construction and operation of the Moreton

Bay Rail (MBR) Project, a new 13 km rail line between Petrie and Kippa-Ring, just north of Brisbane, Queensland.12 The program involved the capture,

veterinary management (treating sick and injured koalas to improve welfare

and population health) and telemetric monitoring of a total of 503 free-living

independent and near independent koalas (217 male, 285 female and 1

intersex) inhabiting blocks of bushland intersected by the rail line. Ecological

characteristics varied significantly across the study site (1097 ha), with

vegetation communities ranging from open grasslands to various types of

open to closed eucalypt-dominated forests.

Detecting mortalities

Koalas weighing more than 3 kg were monitored using near-real-time

biotelemetry devices (K-Tracker® telemetry system (LX Group, Sydney,

Australia)). Global positioning system (GPS) locations and activity data was

uploaded to a server every 12 hours. Mortalities were generally detected

within 24 hours of death, facilitating rapid recovery and necropsy examination

of koala carcasses at the Endeavour Veterinary Ecology facilities at Toorbul,

Queensland.12 Koalas between 1 kg and 3 kg were monitored using very high frequency (VHF) radio telemetry collars and conventional field tracking.

Mortalities were detected during routine field monitoring events, which

occurred either daily or several times a week.12

Regulatory approvals

This work was conducted under animal ethics approvals issued by the

Queensland Department of Agriculture and Fisheries (approvals CA

2012/03/597, CA 2013/09/719, CA 2014/06/777, CA 2015/03/852, and CA

2016/03/950). Work with koalas was authorised by scientific purposes permits issued by the Queensland Department of Environment and Heritage

Protection (approvals WISP 11525212, WISP 16125415, WISP 13661313,

WITK 14173714 and WISP 17273716).

Results

Diagnosis of carpet python predation

Gross necropsy findings in definitive cases of carpet python predation, where the carcass of the koala was not ingested, included injuries caused by the primary bite from the python and the effects of constriction. The lesions were often subtle, and detection usually required careful examination of the carcass, including clipping of the fur. Three cardinal signs were relied upon to make a diagnosis of python predation: (1) identification of typical U-shaped bite wounds (Figure 1); (2) slicking of the fur of the koala on the head by carpet python saliva due to ingestion attempts (Figure 2); (3) diffuse, uniform pulmonary congestion in response to fatal constriction (Figure 3) (diffuse, uniform pulmonary congestion was never solely relied upon for definitive

diagnosis). Additional signs, observed in some cases, included injection of

subcutaneous vasculature, petechial and ecchymotic haemorrhages

associated with the bite site and/or constriction, and the finding of shed

python teeth in bite wounds (Figure 4).

Figure 1: U-shaped primary bite site on the limb of a koala due to carpet

python predation.

Figure 2: Slicking of the fur of a koala from carpet python saliva.

Figure 3: Lungs of a koala at necropsy examination demonstrating diffuse,

uniform pulmonary congestion following carpet python asphyxiation.

Figure 4: Python tooth retrieved from a primary bite site on a koala limb.

The primary bite site was identified on 10 of the 13 koalas (77%) that were

killed, but not ingested, by a carpet python. This was most commonly located

on the distal extremities, with U-shaped puncture marks detected on the limbs of 7 koalas (see Figure 1) and the head or body of 3 koalas. The primary bite site was not evident in 3 cases. Slicking of the fur by carpet python saliva was evident on the head of 10 koalas (77%) (Figure 2), and all 13 koalas had diffuse, uniform pulmonary congestion (Figure 3).

Circumstantial evidence in all cases of carpet python predation included the suddenness of the death in a previously well koala, with no premonitory signs,

and absence of evidence of predation by other common predators such as

canids. In all cases in which ingestion of the carcass had occurred (n=8), the

definitive evidence included the finding of a large, recently fed carpet python at the location of the strongest VHF radio signal for that koala’s radio telemetry device (Figure 5a), confirmation of ingestion of the tagged koala by

radiography (Figure 5b) and/or passing of the koala’s radio telemetry device in

the faeces of the carpet python.

Figure 5a: Recently fed carpet python after ingestion of a koala.

Figure 5b: Radiograph demonstrating an ingested koala wearing radio

telemetry devices inside a carpet python.

Importance of carpet pythons as a cause of mortality

In total 21 koalas (12 females: 9 males) were killed by carpet pythons during

the koala management program, representing 7.2% of all confirmed and

suspected (based on circumstantial evidence12) koala mortalities (n=293)

(Figure 6). Carpet pythons were responsible for 12.1% of all mortalities attributed to native predators (n=173) and were the second most common cause of death by predation after wild dogs (dingoes and dingo hybrids (Canis

familiaris dingo)). This figure excludes mortalities attributed to domestic dogs

(Canis familiaris), of which there were 8 cases (4.4% of predation mortalities),

and mortalities of untagged dependent joeys whose mothers were killed

(n=3).

Figure 6: Proportionate causes of mortality of telemetry monitored koalas during the Moreton Bay Rail koala management program. ‘Other causes’ include: vehicle strike, intra-specific fighting, old-age related causes, and misadventure.

Ingestion of carcasses

Koalas were ingested in only 8 of the 21 cases (38%) of confirmed carpet python predation deaths. In the remainder of cases (62%, n=13), the koala was killed by constriction with evidence of attempted ingestion, but the attempt was aborted and the carcass abandoned. Those that were ingested had an age range of 11.5 months – 14 months, and approximate weights of

1.3 kg – 2.9 kg (based on their most recent veterinary examination prior to death). Those that were not ingested varied significantly in age and size, with an age range of 10.5 months – 7 years and approximate weights of 1.3 kg –

7.2 kg. As expected, younger and therefore smaller koalas were far more likely to be ingested after being killed by a carpet python, with 6 of the ingested koalas (75%) weighing less than 2.6 kg. Conversely, only 3 koalas

(14%) weighing less than 2.6 kg were not ingested after being killed by a carpet python. Demographic and koala weight data are summarised in Table

1, below.

Table 1: Demographics of carpet python predations during the Moreton Bay

Rail koala management program.

Name Ingested Approx. Age Sex Area Season weight (years) (kg) Geisha N 5.17 12.9 Female West Summer Slick* N 1.27 0.97 Male West Summer Tait N 6.07 3.39 Male West Summer Lisa* N 1.59 0.88 Female West Summer Mud N 2.97 1.07 Female West Summer Spartan N 7.19 7.02 Male Central Autumn Peaches N 3.65 1.3 Female West Autumn Varney N 5.12 2.29 Male West Autumn Madso N 2.46 1.04 Male West Summer Miles N 2.62 1.06 Male East Summer Kate G N 5.35 5.51 Female Central Summer Ned* Y 2.04 0.98 Male East Summer Red N 4.8 2.09 Female West Summer Baron Jenny B N 5.49 4.38 Female West Spring K-ski* Y 1.36 0.96 Male West Summer Felicia Y 2.84 1.15 Female West Autumn Faith Y 2.09 0.98 Female West Summer Mattaeus Y 1.87 0.96 Female West Spring Silk Y 2.92 1.14 Female West Summer Hugo Y 2.28 1.06 Male West Summer Inara Y 1.72 1.01 Female West Summer * Dependent joeys

Temporal and spatial associations

Mortalities due to carpet python predation were significantly associated with

location, with 17 of the 21 koalas (80.9%) killed by carpet pythons residing in

the western extent of the rail corridor, in bushland sites close to freshwater

riparian vegetation. Seasonal variations in carpet python predations were

marked, with 15 koalas (71.4%) killed by carpet pythons during summer

months (December to February), and none during winter months (June to

August) (Figure 7). All carpet python predations occurred when maximum

daytime temperatures exceeded 24 degrees Celsius (range 24.7 – 31.6

degree Celsius).

Figure 7: Number of koalas killed by carpet pythons during the Moreton Bay

Rail koala management program over time.

Predation of dependent young

Of approximately 377 joeys produced during the four-year koala management program, at least 3 dependent joeys were known (n=1) or highly likely (n=2) to

have been killed when carpet pythons killed their mothers. An additional 31

dependent joeys who were emergent from the pouch (age range 6 months –

10 months) were lost during the program. These animals were either too

young to be telemetrically tagged, or had not yet been captured for tagging,

and so their fate is unknown. We can speculate, based on the evidence

above, that some of these were victims of carpet python predation.

Non-fatal attempted predation

One sub-adult female koala was recovered after falling from a tree whilst being constricted by a carpet python. Two bite wounds were located on the right distal extremities and the bite site on the right forelimb contained two python teeth that had been shed during the encounter. The koala recovered uneventfully and was released back into the wild a few days later.

Discussion

The Moreton Bay Rail koala management program allowed for the collection of an unprecedented quantity of longitudinal monitoring and mortality data from a large, free-living koala population. The K-Tracker ® biotelemetry system (LX Group, Sydney, Australia) facilitated the rapid detection of koala mortalities (often within 24 hours), ensuring prompt recovery of carcasses, so that diagnostic necropsy examinations were performed on most of the deceased koalas. Hence, quantitative mortality data were recorded for the population, enabling accurate determination of proportionate causes of mortality. We found that carpet pythons were an important cause of premature death, accounting for 7.2% of mortalities of telemetry monitored koalas. In addition, a large proportion of near-independent and back-rider offspring were lost during the four-year program, some of which are speculated to have been preyed upon by carpet pythons. Although incidental

predation of koalas by pythons has been previously reported,15 this study

represents the first population-level estimate of python predation.

In this study, only 38% of the koalas killed by carpet pythons were

successfully ingested, with the remainder found deceased, but intact, with

minimal or no overt indication of the cause of death. Pythons kill their prey by

securing it with a primary bite and rapidly wrapping their body around it,

constricting incrementally until it is killed by asphyxiation. The prey is then

released, and generally the head is sought as the first point of ingestion.16

We suggest that the three cardinal signs that should be investigated to

diagnose python-induced deaths of koalas are: (1) identification of the primary

bite site, as a U-shaped or two opposing U-shaped lines of fine punctures; (2) the presence of slicking of the fur, particularly on the head, by python saliva during unsuccessful attempts to ingest the carcass; and (3) diffuse, uniform pulmonary congestion (as opposed to patchy congestion, which is a common non-specific finding). These lesions are consistent with those reported for other species killed by pythons.17, 18

The diagnostic lesions described above are often subtle and very careful examination of the carcasses was required to avoid them being overlooked.

Due to the koala’s thick pelage, the fine puncture marks resulting from a

python bite can often be difficult to detect, and the carcass must be thoroughly

This article is protected by copyright. All rights reserved. Carpet pythons: important predators of koalas 15 shaved to locate the primary bite site. Furthermore, the characteristic slicking of the fur on the face may not be apparent in all cases, particularly when the post-mortem interval is long, post-mortem decomposition is advanced, or there has been significant scavenging or insect damage to the carcass.

Extreme weather conditions, such as high temperatures and excessive rainfall, accelerate post-mortem changes and rapidly obscure those lesions associated with carpet python predation, further impeding an accurate diagnosis of the cause of mortality. Such weather conditions often coincide with peak periods of python feeding,13, 19 with more than 70% of the python predations in this study occurring during the summer months (December to

February). If quantifying causes of death is an important objective in a study, telemetry methodologies must be designed to facilitate rapid detection of mortality events and retrieval of the carcass.

To our knowledge, this is the first study to report such ‘wastage’ of prey that are killed but not ingested by wild carpet pythons. It is not clear why some carpet pythons attacked and killed koalas that were far too large for them to consume. There may be considerable risk to a python when it attacks an adult koala, as they can bite in self-defence, inflicting substantial injury. Our observation of the koala and carpet python falling several metres to the ground during an attack highlights another potential source of injury. Given

the risk of injury and the energy wasted in killing a prey item that cannot be

consumed, there would appear to be an important fitness cost associated with

this behaviour that pythons should wish to avoid. We speculate that some

attacks on large koalas may occur in situations where the snake is unable to

ascertain the size of the koala prior to their attack, which seems reasonable

for an opportunistic, ambush predator. Alternatively, they may reflect a lack of

availability of other prey species.

Based on our findings, python predation should be included in the differential

diagnosis of sudden death of koalas in the wild when other definitive causes of mortality are excluded. Over most of the koala’s geographic range, the

most likely python to cause mortality of koalas is the carpet python (Morelia

spilota) in its various forms (including diamond pythons and jungle pythons),

and possibly the amethystine python (Morelia kinghorni) in the far northern

extent. Carpet pythons are semi-arboreal, predominantly ambush hunters,

that are more active nocturnally and spend significantly more time on the

ground during warmer months.13, 14 Koalas are also predominantly active

nocturnally, often descending to the ground to move between trees.8 We

could speculate this is when they are most susceptible to predation by carpet

pythons, although we observed one instance of a carpet python attacking a

koala in the canopy during the daytime. Carpet python densities are likely to

vary substantially across the koala’s range so their importance as a source of

mortality and role in koala population dynamics is likely to be site-specific.14

As koala populations become more intensively managed to avert local

extinctions, the identity of predators also becomes more relevant to inform

decisions about where to allocate limited predator control efforts in time and

space. If targeted actions are taken to reduce predation rates using, for

example, intensive wild dog control programmes, accurate apportioning of

predator-associated mortalities is useful for improving the cost-effectiveness and biological outcomes of control efforts. As carpet pythons disproportionally depredate smaller, juvenile koalas, their impact on population dynamics could be substantial in some populations. Hence, when investigating potential translocation recipient sites, we suggest consideration of resident python densities to maximise the success of management actions. Whilst we do not suggest that koala conservation management actions should be targeted at carpet pythons, the potential for carpet python predation may be an important consideration for koala management planning.

Conclusion

Carpet python predation, with or without ingestion, has not previously been documented as an important cause of natural mortality in koala populations, because few free-living koala populations have been monitored and studied

This article is protected by copyright. All rights reserved. Carpet pythons: important predators of koalas 18 intensively enough for the causes of mortality to be accurately quantified at the population level. Both population viability analyses6 and development of conservation management plans benefit from robust cause-of-death data collected during longitudinal monitoring studies. These can ensure that conservation management measures are targeted to the most important threats. Collection of such data requires that telemetry methods facilitate rapid detection of mortality in tagged animals and that python predation is considered as a differential diagnosis of sudden death. As koala populations continue to decline, it is vital that limited resources dedicated to the conservation of the species are utilised most effectively and gain the greatest benefit.

Acknowledgements

We thank the Queensland Government (Department of Transport and Main

Roads), specifically the Moreton Bay Rail project team, for their financial support of this research.

References

1 Melzer A, Carrick F, Menkhorst P et al. Overview, critical assessment, and conservation implications of koala distribution and abundance. Conserv Biol

2000;14(3): 619-628.

2 Seabrook L, McAlpine C, Baxter G et al. Drought-driven change in wildlife distribution and numbers: a case study of koalas in south west Queensland.

Wildlife Res 2011;38: 509-524.

3 McAlpine C, Lunney D, Melzer A, et al. Conserving koalas: A review of the contrasting regional trends, outlooks and policy challenges. Biol Conserv

2015;192: 226-236.

4 Rhodes JR, Beyer HL, Preece H et al. South East Queensland koala population modelling study. UniQuest: Brisbane. 2015.

5 Adams-Hosking C, McBride M, Baxter G et al. Use of expert knowledge to elicit population trends for the koala (Phascolarctos cinereus). Divers Distrib

2016: 1–14.

6 Beyer HL, de Villiers D, Loader J et al. Management of multiple threats achieves meaningful koala conservation outcomes. J Appl Ecol 2018: 1-10. doi:10.1111/1365-2664.13127

7 Gonzalez-Astudillo V, Allavena R, McKinnon A et al. Decline causes of koalas in South East Queensland, Australia: A 17-year retrospective study of mortality and morbidity. Sci Rep-UK 2017;7: 42587. doi:10.1038/srep42587

8 Martin RW, Handasyde, KA, editors. The koala. Natural history, conservation and management. UNSW Press, Sydney, 1999.

9 Ramsay S. The ecology and dispersal patterns of juvenile koalas,

Phascolarctos cinereus, in fragmented habitat. PhD, School of Biological

Sciences, University of Sydney, Sydney, 1999:234.

10 Melzer A, Tucker GM, Hodgon J et al. A note on predation on koalas

Phascolarctos cinereus by raptors, including wedge-tailed eagles Aquila audax, in Queensland. Queensl Nat 2003;41(13): 38-41.

11 Lunney D, Gresser SM, Mahon P et al. Post-fire survival and reproduction of rehabilitated and unburnt koalas. Biol Conserv 2004;120: 567-575.

12 Hanger J, de Villiers D, Forbes N et al. Moreton Bay Rail Koala

Management Program: Final technical report for Queensland Department of

Transport and Main Roads. Endeavour Veterinary Ecology, Toorbul, 2017.

13 Fearn S, Robinson B, Sambono J et al. Pythons in the pergola: the ecology of ‘nuisance’ carpet pythons (Morelia spilota) from suburban habitats in south-eastern Queensland. Wildlife Res 2001;28: 573-579.

14 Shine R, Fitzgerald M. Large snakes in a mosaic rural landscape: The ecology of carpet pythons Morelia spilota (Serpentes: Pythonidae) in coastal eastern Australia. Biol Conserv 1996;76: 113-122.

15 Queensland Government. Koala Hospital Data: KoalaBase. 2017. https://data.qld.gov.au/dataset/koala-hospital-data/resource/7c6f7da8-ef7a-

48e4-bf4e-c449a885e46d. Accessed May 2018.

16 Loop MS, Bailey LG. The effect of relative prey size on the ingestion behaviour of rodent-eating snakes. Psychon Sci 1972;28(3): 167-169.

17 Staker L, editor The complete guide to the care of macropods. Matilda’s

Publishing, Armidale, 2006.

18 MacCormick A, editor. The Mammoth Book of Maneaters. Constable and

Robinson, London, 2011.

19 Slip DJ, Shine R. Habitat use, movements and activity patterns of free-

This article is protected by copyright. All rights reserved. Carpet pythons: important predators of koalas 22 ranging Diamond Pythons, Morelia spilota (Serpentes: Boidae): a radiotelemetric study. Aust Wildl Res 1988;15: 515-531.

Conflicts of interest

One of the authors, Jonathan Hanger, was involved in the development of the

K-Tracker ® biotelemetry system.