Journal of the Royal Society of Western Australia
ISSN 0035-922X
CONTENTS Page Distribution of Westralunio carteri Iredale 1934 (Bivalvia: Unionoida: Hyriidae) on the south coast of southwestern Australia, including new records of the species M W Klunzinger, S J Beatty, D L Morgan, A J Lymbery, A M Pinder & D J Cale 77
Original distribution of Trichosurus vulpecula (Marsupialia: Phalangeridae) in Western Australia, with particular reference to occurrence outside the southwest I Abbott 83
Grasstree stem analysis reveals insufficient data for inference of fire history • B P Miller, T Walshe, N J Enright & B B Lamont 95
Ichthyoplankton assemblages associated with pink snapper (Pagrus auratus) spawning aggregations in coastal embayments of southwestern Australia N B Breheny, L E Beckley & C B Wakefield 103
Importance of Lake MacLeod, northwestern Australia, to shorebirds: a review and update D Bertzeletos, R A Davis & P Horwitz 115
Volume 95 Part 2 July 2012 The Royal Society of Western Australia
To promote and foster science in Western Australia Patron Her Majesty the Queen Vice-Patrons His Excellency Mr Malcolm ] McCusker, AC CVO QC Governor of Western Australia Professor Lyn Beazley AO FTSE MA PhD
COUNCIL 2011-2013
President P O'Brien Hon Librarian Immediate Past President L Milne Journal Manager Vice-President H Bekle Councillors Vice-President K M Trinajstic A W R Bevan Hon Secretaries L B Collins Council L Milne C Florides Membership J Rosser M Gallegos Proceedings W A Loneragan Hon Treasurer P O'Brien M Menz Hon Editor A E Cockbain S Toby
EDITORIAL BOARD 2011-2013
Editor-in-chief K A Haskard, Data Analysis Australia A E Cockbain, South Perth P G Ladd, Murdoch University Associate Editors D Laird, Murdoch University H Bekle, Notre Dame University K M Trinajstic, Curtin University A W R Bevan, Western Australian Museum M Van Keulen, Murdoch University R A Davis, Edith Cowan University K V Wright, Curtin University
The Royal Society of Western Australia was founded in 1914. The Society promotes exchange among scientists from all fields in Western Australia through the publication of a journal, monthly meetings where interesting talks are presented by local or visiting scientists, and occasional symposia or excursions on topics of current importance. Members and guests are encouraged to attend meetings on the third Monday of every month (March - November) at 7 pm, Kings Park Board offices, Kings Park, West Perth, WA 6005, or as advertised, in the RSWA Proceedings, Diary of Events. Individual membership subscriptions for the 2012/2013 financial year are $75 for ordinary Members, $40 for Student Members and $40 for Associate Members. For Library, company and institution subscriptions see the Society’s website http://www. royalsocietyofwa.com. Prices include GST. For membership forms, contact the Membership Secretary, PO Box 7026, Karawara, Western Australia 6152, or visit tile website http://vvvvw.royalsocietyofwa.com. The Journal of the Royal Society of Western Australia was first published in 1915. Its circulation exceeds 650 copies. Nearly 100 of these are distributed to institutions or societies elsewhere in Australia. A further 300 copies circulate to more than 40 countries. The Society also has over 350 personal members, most of whom are scientists working in Western Australia. The Journal is indexed and abstracted internationally. Cover design: The four subjects symbolize the diversity of sciences embraced by the Royal Society of Western Australia. Mangles' kangaroo paw (Anigozanthos numglesii) and the numbat (Myrmecobius fnsciatus) are the floral and faunal emblems of Western Australia, and stromatolites are of particular significance in Western Australian geology (artwork: Jan Taylor). The Gogo Fish (Mcttamaraspis knprios) is the fossil emblem of Western Australia (artwork: Danielle West after an original by John Long). Journal of the Royal Society of Western Australia, 95: 77-81,
Distribution of Westralunio carteri Iredale 1934 (Bivalvia: Unionoida: Hyriidae) on the south coast of southwestern Australia, iricludingnew. records of the species
M W KLUNZINGER1*, S J BEATTY1, D L MORGAN1, A J LYMBERY1, A M PINDER2 & D J CALE2
1 Freshwater Fish Group & Fish Health Unit, Murdoch University, Murdoch, WA 6150, Australia. 2 Science Division, Department of Environment and Conservation, Woodvale, WA 6026, Australia. * Corresponding author [email protected]
Westralunio carteri Iredale 1934 is the only hyriid in southwestern Australia. The species was listed as 'Vulnerable' by the IUCN, due to population decline from dryland salinity, although the listing was recently changed to 'Least Concern'. The Department of Environment and Conservation lists the species as 'Priority 4', yet it lacks special protection under federal or state legislation. Accuracy in species accounts is an important driver in determining conservation status of threatened species. In this regard, discrepancies in locality names and vagary in museum records necessitated the eastern distributional bounds of W. carteri to be clarified. Here we present an updated account of the species' distribution and describe two previously unknown populations of W. carteri in the Moates Lake catchment and Waychinicup River, resulting in an eastern range extension of 96-118 km from the Kent River, formerly the easternmost river where W. carteri was known. For mussel identification, samples (n = 31) were collected and transported live to the laboratory for examination and internal shell morphology confirmed that the species was W. carteri. Due to an acute salinity tolerance of 3.06 g/L (LD^), the species is unlikely to exist in catchments east of the Waychinicup River where salinities are known to be as high as 45 g/L. The current study clarifies the eastern distribution of W. carteri and will allow monitoring of its range decline to be made in the future.
KEYWORDS: freshwater mussel, geographic distribution, Moates Lake, species conservation, Waychinicup River, Westralunio carteri
INTRODUCTION a highly seasonal rainfall and climate pattern similar to that of the Mediterranean, with hot dry summers and Freshwater mussels (Bivalvia: Unioniformes) are found cool wet winters (Pen 1999). Many of the formerly in freshwater environments on every continent except freshwater streams, rivers, lakes and wetlands have Antarctica (Bogan & Roe 2008). The life history of these become affected by dryland salinity through a legacy of bivalves is generally characterised by their parasitic land clearing dating back to the late 1800s (Wood 1924; larvae that usually require fish as hosts (Bauer & Beresford el al. 2001). Wachtler 2001; Strayer 2008). Parasitic larvae of the Unionoidea (Hyriidae, Margaritiferidae and Unionidae) Westralunio carteri Iredale, 1934 is the only unioniform are known as 'glochidia', which are morphologically to inhabit the South West Coast Drainage Division of distinct from the 'lasidia' or 'haustoria' of Etherioidea Australia (McMichael & Hiscock 1958; Walker 2004). (Etheriidae, Iridinidae and Mycetopodidae) (Bauer & Kendrick (1976) first noted the disappearance of W. Wachtler 2001), although Graf & Cummings (2006) carteri from the Avon River, and proposed that this placed the glochidia of the Hyriidae in the Etherioidea resulted from secondary salinisation of its formerly based on phylogenetic evidence. freshwater habitat. Later, results of a biodiversity study (Pinder et al. 2004) suggested that W. carteri had The Southern Hemisphere is dominated by the disappeared from parts of its range within the Hyriidae with 76-83 species occurring in Australasia and agricultural zone, strengthening Kendrick's (1976) South America (Walker et al. 2001; Graf & Cummings assessment which resulted in the species being listed as 2006; Bogan & Roe 2008). Australia currently has 18 'Vulnerable' on the International Union for the nominal species (McMichael & Hiscock 1958; Ponder & Conservation of Nature's Red List of Threatened Species Bayer 2004; Walker 2004), although recent genetic (IUCN 1996 in Kohler 2011), but was not listed under investigations have confirmed the existence of additional state or federal legislation. However, the species has species, which have not yet been described (Baker et al. recently been assessed as 'Least Concern' '...as it is 2003, 2004; Hughes et al. 2004). widespread in Western Australia, is a habitat generalist, The high concentration of endemic flora and fauna and is resistant to organic pollution. Although previously unique to southwestern Australia contributes to its listing assessed as Vulnerable (under version 2.3 of the Red List) as one of the world's biodiversity hotspots (Myers et al. under the revised Categories and Criteria this species no 2000). The region is comprised of 19 drainage basins with longer qualifies for a threatened listing.' (Kohler 2011). Westralunio carteri has recently been shown to have an © Royal Society of Western Australia 2012 acute salinity LD5(| of 3.06 g/L (Klunzinger et al. 2010).
77 Journal of the Royal Society of Western Australia, 95(2), July 2012
Stewart (2009) proposed two aquatic bioregions for the MATERIALS AND METHODS South Coast Region of Western Australia based largely on salinity and differences in macroinvertebrate The south coast of Western Australia has been heavily compositions, presumably with different salinity cleared since the 1950s and the land is used primarily for tolerance thresholds. In the list of individual species used dryland cropping with some plantation forestry and for the assessment, 'Westralunio spp.' was reportedly some national parks and nature reserves (Mayer et al. found in greater abundance in the Eastern South Coast 2005). We surveyed regional water bodies for W. carteri where salinities are much higher than in the Western by tactile searches in wadeable areas of river and lake South Coast. Inland freshwater lakes of the south coast habitats from 76 sites on the south coast from the Kent were also surveyed for macroinvertebrates by Edward et River to the Pallinup River during March 2011 and al. (1994), but no W. carteri was reported in the study area January 2012 (Figure 1). Where mussels were found, from Kent River to Esperance. patch densities were determined by counting the number of mussels in localised quadrats (1 m2). Records from the Western Australian Museum, Perth (WAM) and the Australian Museum, Sydney (AMS) for Benthic habitat characteristics, including the dominant W. carteri on the south coast east of the Kent River are substrate type and presence/absence of plant debris were sparse and locality descriptions are vague, but include also noted. Specimens were randomly collected from the following: AMS_126151, King George Sound (ca three sites (n = 32) and transported live to the laboratory 1877); WAM_34290, South Stirlings, Calvin River (ca in a 15 L plastic bucket of river water. Shells were 1963); WAM_34289, Esperance Town Beach (ca 1978), but measured as per McMichael & Hiscock (1958) and no coordinates for these localities are available. In the Walker (2004), subsequently euthanased in 0.01% context of accurate assessment of species conservation, benzocaine solution, and dissected to determine sex and here we present updated and new records for W. carteri reproductive status. Gills were examined for the presence on the south coast of Western Australia. (females) or absence (males) of specialised brooding
Figure 1 Distribution of Westralunio carteri Iredale 1934 in the south coast region of southwestern Australia. Thick black lines represent river basin boundaries and the dashed line represents the boundary between the eastern and western south coast bioregions as proposed by Stewart (2009).
78 Klunzinger et al.: Westralunio carteri in southwestern Australia
chambers within the demibranchs of the gills with patch density rapidly decreasing beyond the mouth ('marsupia') and whether they contained glochidia by the of the Goodga River (Figure 2). The greatest densities of presence or absence of thickened, orange to red coloured mussels were found in slower flowing areas around river marsupia as described for other hyriids (Jones et al. 1986; bends near the banks, nestled amongst submerged tree Byrne 1998). Internal shell morphology was examined roots. Where mussels were found, the substrate was using McMichael & Hiscock (1958) as a guide. Shells and dominated by coarse sand with some woody debris and soft tissues were then preserved in 100% ethanol for leaf litter. Above the Goodga River gauging station, future study. benthic habitats consisted primarily of soft sludge and detritus. Species distributions from Edward et al. (1994), Pinder et al. (2004) and this study were mapped using ArcGIS™ Very few mussels were found in the small area of Desktop 10. Linear hydrography and river basin Waychinicup River, but four individuals were located boundary data were obtained under license from the near a submerged log in fine sand and silt. All of the Western Australian Department of Water. Coastal and sites that contained W. carteri had characteristic tannin- administrative boundaries as well as population centres stained water and were fringed by well-vegetated river were mapped from the 'Global Map Australia 1M 2001' banks lined with thick shrubs, and shading trees. dataset (Geoscience Australia 2004).
DISCUSSION RESULTS To our knowledge, this study represents the first account Freshwater mussels were found in the lower Kent River of W. carteri from the Moates Lake catchment and and new populations were discovered in the Moates Waychinicup River of the south coast, which together are Lake catchment in Two Peoples Bay Nature Reserve, probably the only remaining populations within the located -30 km east of Albany and in the Waychinicup Albany Coast Basin, although neighbouring streams such River, -22 km east of Moates Lake on the south coast of as the Angove River may also support the species. We Western Australia, but no mussels were found in any of suspect that the current existence of W. carteri on the other waterways sampled (Figure 1). All freshwater Esperance Town Beach (WAM_34289, ca 1978) is unlikely mussels were identified as W. carteri. Marsupia were given that mean salinities of rivers near Esperance are present in the inner demibranchs of females, a greater than 5 g/L, and have been since the late 1970s characteristic feature of the Hyriidae. None of the females (Mayer et al. 2005). Westralunio carteri was recorded from examined contained glochidia. the King George Sound locality in the late 1800s, but the Westralunio carteri was found to occupy a 2.4 km exact location is not available from Australian Museum stretch of Goodga River, but was not recorded above the records. Although a record for the species exists in the fishway at gauging station AWRC 602199 (Figure 2). In 'Calvin River', there is no river by that name in Western Moates Lake, the species was only found within a 300 m2 Australia and was probably made in error. The record area extending beyond the mouth of the Goodga River, probably refers to the Kalgan River, but this has not been but no live mussels were found beyond the river mouth confirmed. If indeed the species did exist in this locality, (Figure 2). Densities of W. carteri ranged from 1 to 15/m2, the new information from this study possibly represents
Figure 2 Distribution of Westralunio carteri Iredale 1934 in the Moates Lake catchment of southwestern Australia. Mussel density symbols represent the number of mussels per lm2 quadrat.
79 Journal of the Royal Society of Western Australia, 95(2), July 2012
a range extension of ~8-30km east from Kalgan River. 2003; Morgan & Beatty 2006). If host fishes could not However, our recent surveys failed to record the species deliver attached glochidia upstream of the gauging in the Kalgan River, thereby placing the nearest station and if aging mussels above the weir were unable population 96-118 km to the west in the Kent River. to expose their glochidia to potential hosts, they could Waychinicup River represents the easternmost limit of have died out from recruitment failure. If we knew the the distribution of endemic freshwater fishes (Morgan et longevity of adult W. carteri in this system, had al. 1998, 2011), which is probably true for W. carteri as information on suitable host fishes, knew tolerance well. Very few, if any perennial rivers east of thresholds for NH3 and whether we could find juvenile Waychinicup are fresh (Mayer el al. 2005; Stewart 2009), W. carteri, we would be able to test these hypotheses. with a majority of the rivers in the region having Although host fishes in these systems are unknown, salinities above the known acute tolerance of 8.2-14.6 g/L Nannoperca vittata (Castelnau 1873) and Pseudogobius for several endemic freshwater fishes (Beatty et al. 2011), olprum (Sauvage 1880) which are known to inhabit both and 3.06 g/L for W. carteri (Klunzinger et al. 2010), so the Goodga and Waychinicup Rivers (Morgan & Beatty 2006; likelihood of finding freshwater mussels in salinised Morgan el al. 1998, 2011) are likely candidates given that rivers in the Eastern South Coast is probably low. The they are known to support glochidia of W. carteri species data from Binder el al. (2004), presented in Figure elsewhere in its range (Klunzinger et al. in press). The 1, lends support to this presumption. Although Stewart Goodga River contains Australia's only Critically (2009) reported ‘Westmlunio spp/ occurring in many of Endangered fish, Galaxias truttaceus (Cuvier & the salinised rivers of the south coast, the species Valenciennes 1846) as well as Galaxias maculatus (Jenyns identification was probably erroneous and was most 1842) (Morgan 2003; Stewart 2011; Morgan & Beatty likely Fluviolanatus subtorta (Dunker 1857), based on 2006; Morgan et al. 1998, 2006, 2011), both of which may photos provided by the former author (B Cook pers. be hosts for glochidia. comm. 2010). This trapezoid has migrated inland from The accuracy of species distributions is important for estuaries to salinised rivers of the region such as Avon conservation assessments, particularly at the broad River (Kendrick 1976) and Blackwood River (Pen 1999). regional scale (Anderson & Martinez-Meyer 2004). This We have also observed this species and Mytilus (Mytilus) study adds to the current state of knowledge on the planatus Lamarck 1819, a marine/estuarine mytilid, biogeography of W. carteri, namely extent of occurrence, inhabiting the salinised Kalgan River nearly 40 km which is a key criteria used for conservation assessment inland from the estuary. under the Western Australian Wildlife Conservation Act Besides salinity, sedimentation may be causing 1950, the federal Environment Protection and Biodiversity localised declines based on our observations of the Conservation Act 1999 and the IUCN Red List of species elsewhere, which is supported by Jones & Byrne Threatened Species. Results from this study will (2010) who found that sedimentation is a limiting factor undoubtedly be useful in future conservation assessments for eastern Australian hyriids. Reservoir dewatering and of the status of the species. Factors influencing the rainfall reductions appear to be having a negative effect occurrence of this species in relation to conservation on W. carteri populations (M W Klunzinger et al. unpubl. criteria should be the focus of ongoing investigations. data). The region's climate is changing, most notably since the 1970s with rainfall, surface flow and groundwater reductions, amplifying the effect of dryland ACKNOWLEDGEMENTS salinity and increasing the vulnerability of freshwater fishes to population loss with the trend expected to This work was funded by a grant from the Holsworth continue into the future (Morrongiello et al. 2011). Wildlife Research Endowment and an International PhD Scholarship from Murdoch University. A License to Take Due to limited time and resources during the survey, Fauna for Scientific Purposes (Regulation 17 of the we did not collect an adequate number of mussels to Wildlife Conservation Act 1950) was obtained from the draw any conclusions from shell measurements about Department of Environment and Conservation and a population structure. Because growth rates, host fishes Research Permit (under exemption from the Fish and reproductive biology of W. carteri in these systems Resources Management Act 1994) from the Department have not been quantified, we are reluctant to infer age, of Fisheries. We especially thank James Keleher and Iain reproductive status or recent recruitment patterns. Future Foster for their help during field sampling. We thank assessments of the population structure in these systems two anonymous reviewers for constructive criticism should utilise a more rigorous sampling design to during review of earlier versions of this publication and maximise the probability of collecting the full range of Pierre Horwitz for helpful comments and suggestions size classes and in greater quantity for statistical power. during revision of the manuscript. The absence of W. carteri above the gauging station weir in Goodga River could be due to the build-up of detritus and fine sediments which have been known to REFERENCES cause hypoxia/anoxia and a build-up of un-ionised ammonia (NH3) during summer months and non-flow or Anderson R P & Martinez-Meyer E 2004. Modeling species' slow-flow periods which is particularly lethal to juvenile geographic distributions for preliminary conservation assessments: an implementation with the spiny pocket mice freshwater mussels elsewhere (Bauer & Wiichtler 2001; (Heteromys) of Ecuador. Biological Conservation 116, 167-179. Strayer 2008). Prior to the fishway being opened in April Baker A M, Bartlett C, Bunn S E, Goudkamp K, Sheldon F & 2003, upstream migration of fishes was prevented by the Hughes J M 2003. Cryptic species and morphological gauging station weir and no fish of any species existed plasticity in long-lived bivalves (Unionoida: Hyriidae) from above the weir before the fishway was opened (Morgan inland Australia. Molecular Ecology 12, 2707-2717.
80 Klunzinger et al.: Westralunio carteri in southwestern Australia
Baker A M, Sheldon F, Somerville J, Walker K F & Hughes J M McMichael D F & Hiscock I D 1958. A monograph of the 2004. Mitochondrial DNA phylogenetic structuring suggests freshwater mussels (Mollusca: Pelecypoda) of the Australian similarity between two morphometrically plastic genera of Region. Australian Journal of Marine and Freshwater Research 9, Australian freshwater mussels (Unionoida: Hyriidae). 372-507. Molecular Pln/logenetics and Evolution 32, 902-912. Morgan D L 2003. Distribution and biology of Galaxias trutlaceus Bauer G & Wachtler K 2001. Ecology and evolution of the (Galaxiidae) in south-western Australia, including first freshwater mussels Unionoida. Springer-Verlag, New York. evidence of parasitism of fishes in Western Australia by Ligula intestinalis (Cestoda). Environmental Biology of Fishes Beatty S J, Morcan D L, Rashnavadi M & Lymbery A J 2011. 66, 155-167. Salinity tolerances of endemic freshwater fishes of south¬ western Australia: implications for conservation in a Morgan D L & Beatty S J 2006. Use of a vertical-slot fishway by biodiversity hotspot. Marine and Freshwater Research 62, 91- galaxiids in Western Australia. Ecology of Freshwater Fish 15, 100. 500-509. Beresford Q, Bekle H, Phillips H & Mulcock J 2001. The Salinity Morgan D L, Beatty S J, Klunzincer M W, Allen M A & Crisis. Landscapes, communities and politics. University of Burnham Q F 2011. A field guide to the freshwater fishes, Western Australia Press, Crawley. crayfishes and mussels of south-western Australia. SERCUL and Freshwater Fish Group & Fish Health Unit, Murdoch Bogan A H & Roe K J 2008. Freshwater bivalve (Unioniformes) University, Perth. diversity, systematics, and evolution: status and future directions. Journal of the North American Benthological Society Morgan D L, Chapman A, Beatty S J & Gill H S 2006. 27, 349-369. Distribution of the spotted minnow (Galaxias maculatus (Jenyns, 1842)) (Teleostei: Galaxiidae) in Western Australia Byrne M 1998. Reproduction of river and lake populations of including range extensions and sympatric species. Records of Hyridella depressa (Unionacea: Hyriidae) in New South Wales: the Western Australian Museum 23, 7—11. implications for their conservation. Hydrobiologin 389, 29-43. Morgan D L, Gill H S & Potter I C 1998. Distribution, Edward D H D, Gazey P & Davies P M 1994. Invertebrate identification and biology of freshwater fishes in south¬ community structure related to physico-chemical parameters western Australia. Records of the Western Australian Museum of permanent lakes of the south coast of Western Australia. Supplement 56, 1-97. Journal of the Royal Society of Western Australia 77, 51-63. Morrongiello J R, Beatty S J, Bennett J C, Crook D A, Ikedife D Geoscience Australia 2004. Global Map Australia 1M 2001 E N, Kennard M J, Kerezsy A, Lintermans M, McNeil D G, Product User Guide (3rd Edition). Geoscience Australia, Pusey B J & Rayner T 2011. Climate change and its Canberra. implications for Australia's freshwater fish. Marine and Graf D L & Cummings K S 2006. Paleoheterodont diversity Freshwater Research 62, 1082-1098. (Mollusca: Trigonoida + Unionoida): what we know and Myers N, Mittermeier R A, Mittermeier C G, Da Fonseca GAB what we wish we knew about freshwater mussel evolution. & Ken J 2000. Biodiversity hotspots for conservation Zoological Journal of the Linnaean Society 148, 343-394. priorities. Nature 403, 853858. Hughes-J M, Baker A M, Bartlett C, Bunn S, Goudkamp K & Pen L J 1999. Managing our rivers. A guide to the nature and Somerville J 2004. Past and present patterns of connectivity management of the streams of south-west Western Australia. among populations of four cryptic species of freshwater Water and Rivers Commission, East Perth. mussels Velesunio spp.(Hyriidae) in central Australia. Pinder A M, Hai.se S A, McRae J M & Shiel R J 2004. Aquatic Molecular Ecology 13, 3197-3212. invertebrate assemblages of wetlands and rivers in the wheat Jones H A & Byrne M 2010. The impact of catastrophic channel belt region of Western Australia. Records of the Western change on freshwater mussels in the Hunter River system, Australian Museum Supplement 67, 7-37. Australia: a conservation assessment. Aquatic Conservation: Ponder W F & Bayer M 2004. A new species of Lortiella Marine and Freshwater Ecosystems 20, 18-30. (Mollusca: Bivalvia: Unionoidea: Hyriidae) from northern Jones H A, Simpson R D & Humphrey C L 1986. The reproductive Australia. Molluscan Research 24, 89-102. cycles and glochidia of freshwater mussels (Bivalvia: Stewart B A 2009. Two aquatic bioregions proposed for the Hyriidae) of the Macleay River, northern New South Wales, South Coast Region, Western Australia. Journal of the Royal Australia. Malacologia 27, 185-202. Society of Western Australia 92, 277-287. Kendrick G W 1976. The Avon: faunal and other notes on a Stewart B A 2011. Assessing the ecological values of rivers: an dying river in south-western Australia. The Western application of a multi-criteria approach to rivers of the South Australian Naturalist 13, 97-114. Coast Region, Western Australia. Biodiversity and Conservation 20, 3165-3188. Klunzinger M, Beatty S & Lymbery A 2010. Acute salinity tolerance of the freshwater mussel Westralunio carteri Iredale, Strayer D L 2008. Freshwater mussel ecology: a multifactor approach 1934 of south-west Western Australia. Tropical Natural to distribution and abundance. University of California Press, History Supplement 3, 112. Berkley. Klunzinger M W, Beatty S J, Morgan D L, Thomson G J & Walker K F 2004. A guide to the provisional identification of Lymbery A J (in press). Glochidia ecology in wild fish the freshwater mussels (Unionoida) of Australasia. CRC populations and laboratory determination of competent host Freshwater Ecology Identification Guide 51 (Murray Darling fishes for an endemic freshwater mussel of south-western Freshwater Research Centre, Albury). Australia. Australian Journal of Zoology. Walker K F, Byrne M, Hickey C W & Roper D S 2001. Freshwater mussels (Hyriidae) of Australia. In: Bauer G & Wachtler K Kohler F 2011. Westralunio carteri. In: IUCN Red List of Threatened Species. Version 2011.2.
81
Journal of the Royal Society of Western Australia, 95: 83-93, 2012
Original distribution of Trichosurus vulpecula (Marsupialia: Phalangeridae) in Western Australia, with particular reference to occurrence outside the southwest
I ABBOTT
Department of Environment and Conservation, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia. IS1 [email protected]
Trichosurus vulpecula, 'common brushtail possum', is currently considered to have never occurred in -20% of Western Australia. This paper reports the results of a survey of historical sources, showing that the species was widely known to Aborigines and was once more broadly distributed, and may have occurred across almost all of Western Australia in 1829, the year of first settlement by Europeans. A remarkable contraction in geographical range commenced ca 1880, caused by an epizootic, and continued from ca 1920-1960 as a result of depredation by foxes {Vulpes vulpes). Some 220 Aboriginal names from outside the southwest were discovered. These are tabulated, and most can be reduced to seven regional names. Several or all of these names appear suitable for adoption into regional use in Western Australia. The current vernacular name is now inaccurate (the species is no longer common anywhere in Western Australia and is not closely related to the Virginia opossum, Didelphys virginiana, which belongs to a different family of marsupials), and should be replaced by one or more Aboriginal names.
KEYWORDS: Aboriginal names, biogeography, brushtail possum, ecological history, mammal, regional extinction, Trichosurus vulpecula
INTRODUCTION 1. Specimens held in the collection of the Western Australian Museum (Kitchener & Vicker 1981) Trichosurus vulpecula, known as the 'common brushtail possum' (Kerle & How 2008), until recent times occurred 2. Scientific reviews and reports by Collett (1897), across Australia but was apparently absent from parts of Helms (1897), Thomas (1904), Dahl (1927), Lundelius Western Australia (Figure 1) and South Australia due to (1963), Youngson et al. (1981), How (1983), Kerle (1983, extinction both before and after European settlement. 2001), McKenzie & Youngson (1983), Baynes (1987,1990), This species was once the fourth most widely distributed Boscacci et al. (1987), Burbidge et al. (1988), How et al. native mammal species in Australia (as well as Western (1988), Kerle et al. (1992), Bavnes & Jones (1993), Jones Australia) after the short-beaked echidna Tachyglossus (2004), Start et al. (2007) and Abbott (2008). aculeatus, Gould's wattled bat Chalinolobus gouldii, and 3. Journals and documents written by explorers and lesser long-eared bat Nyctophilus geoffroyi (Burbidge et al. others, namely Exploration Diaries (1827-1871), Grey 2008; Van Dyck & Strahan 2008). (1841), Eyre (1845), Austin 1854, (see Abbott 2009a), T. vulpecula has become extinct since European Lefroy (1863), Clarkson (1864), Hunt (1864), Martin settlement over much of inland Australia (Kerle & How (1864), Forrest (1875), Gregory & Gregory (1884), Lindsay 2008). However, intriguingly, there are two large areas in (1893), Carnegie (1898), R T Maurice (Murray 1904), Giles Western Australia where there are no records of (1889), Hill 1899-1900 (see Abbott 2009b), Davidson occurrence, and three other areas where the species is (1905), Gribble (1905), F Wittenoom (Lefroy 2003), F known only from subfossil or cave deposits (Figure 1) Hann (Donaldson & Elliot 1998), Russel (1899), F M (Kerle & How 2008). Nonetheless, perusal since 1999 of House (Brockman 1902), Basedow (1904) and Easton Western Australia colonial, state and other newspapers, (1922). exploration diaries, and other early records has revealed 4. Records published in newspapers: Perth Gazette, that T. vulpecula was more widely distributed than Inquirer [Perth], The West Australian, The Australasian indicated on the maps published by How & Kerle (1995) [Melbourne], The Western Mail [Perth], Kalgoorlie Miner and Kerle & Flow (2008). and The Geraldton Guardian. I bis paper documents the occurrence of T. vulpecula in Western Australia, based on all available records. 5. Aboriginal knowledge, particularly their names for Reasons for its decline since the European settlement of the species, recorded by Curr (1886), Helms (1896), Western Australia are discussed. Yabaroo (1899), Bates (nd), Mathews (1910), Brown (1913), Mjoberg (1915), Alexander (1920), Ngabidj & Shaw (1981), Burbidge et al. (1988), and Abbott (2001). METHODS Nearly 130 Aboriginal language groups ('tribes') originally occurred in Western Australia (Tindale 1974). Distribution records were obtained from seven kinds of sources. 6. Archival documents (1912-13) from the Colonial Secretary's Department (1913) held by the State Records Office of Western Australia. In 1909 the Western © Crown Copyright Australia Government declared the first ever closed
83 Journal of the Royal Society of Western Australia, 95(2), July 2012
season to protect T. vulpecula from hunting (Abbott 2008). Comparison of Figures 1 and 2 indicates that T. The 1912-13 reports from police districts were intended vulpecula was more widely distributed in Western to provide the necessary information as to whether Australia than previously thought. The extent of the area populations had recovered sufficiently to allow the in which the species is not known to occur has been closed season to be repealed. reduced considerably, raising the possibility that at least 7. Recollections by oldtimers (Richards & Short 1996, some of the blank areas in Figure 2 may represent under Abbott 2008). sampling rather than true absence.
RESULTS DISCUSSION All of the localized records found of T. vulpecula are mapped in Figure 2. These show that T. vulpecula was Original geographical range once widely distributed in Western Australia. However, T. vulpecula is unknown as subfossil in the region most of the records provide little more information than adjoining Shark Bay (Baynes 1990). Futhermore, no the geographical location. Those that present additional Aboriginal term for this species has been recorded by information are detailed in Table 1. Several of the records Austin (1992a-f) for the Yinggarda (Shark Bay, Gascoyne from the interior of Western Australia allude to the rarity River north to Wooramel River, and inland to Red Hill of T. vulpecula before exotic predators, particularly Vulpes and Gascoyne Junction), Warriyangga (upper Gascoyne vulpes (fox), colonised. More than 220 names used by River), Thargari (Lyons River), Thalanyji (Ashburton Aboriginal people in Western Australia for T. vulpecula River), and Jiwarli (upper Lyons River) people. Apart were located during my review of documents (Table 2). from an Aboriginal name recorded for the Watchandie
Figure 1. Current understanding of the past, recent and present distribution of Trichosurus vulpecula in Western Australia (after Kerle & How 2008). Scale in kilometres
84 Abbott: Trichosurus vulpecula distribution in WA
112*E 114*E 116*E 118*E 120*E 122*E 124*E 128’E 130°E
LEGEND
■ Aboriginal name
M ft
J) to a t12*E 124*E 128*E 130*E
Figure 2. Localised records of the occurrence of Trichosurus vulpecula, in Western Australia, based on Aboriginal names for the species, observations and museum specimens.
85 Journal of the Royal Society of Western Australia, 95(2), July 2012
Table 1 Historical records of occurrence of Trichosurus vulpecula in Western Australia that provide additional information about abundance, ecology, or Aboriginal usage*.
Year Record Reference
1837 Opossum remains found around campfires and many hollow trees with signs of having been smoked by Aborigines to capture possums, Kimberley Grey 1841 vol. It 240-241 1841 Opossums 'abound', Eucla-Eyre district Eyre 1845 vol. 1: 351 1845 Dexterity of an Aboriginal capturing a possum in an inaccessible tree, Murray River Inquirer 15.1.1845: 2 1854 Great quantity of opossum fur at Aboriginal campsite, near Fraser Rocks Austin 1855 (Abbott 2009a: 160) 1861 Aboriginal headdress made of grass and possum fur, Oakover River Gregory 1862: 408 1864 Opossums being eaten by Aborigines, near Mt Monger Hunt 1864: 62 1864 Opossum skins noted, and twine made of opossum hair. Roebuck Bay Perth Gazette 5.8.1864: 3^1 1882 Aborigines 'live wretchedly, there being very little game up here. Even the The Australasian opossum is scarcely ever found' [A Crawford, Moorrarrie station, Murchison region] 28.1.1882: 570-1 ca 1886-1887 '[The Aborigines'] condition has been aggravated during the last five or six years J P Brooks in The West by a vast array of domestic cats, which has spread everywhere. Opossum, tammur, Australian 4.7.1892: 3 kangaroo rats, wallaby, murnine, rock wallaby, bandicoots, pigeon, mallee hen, squeakers, and many other birds and animals [mammals] which formerly formed the chief food of the natives, are almost, if not quite, annihilated.' [Balbinya, c. 70 km SW of Balladonia] ca 1874-1890 'O'possums were not plentiful' Lefroy 2003: 38 1891 Aborigines cooking opossum Lindsay 1893: 23 1894 A 'full-sized opossum' found in the stomach of a carpet snake c. 6'6" in length. The Inquirer & Commercial Canning River News 5.1.1894: 3 1894 A 'full grown opossum' disgorged by a captured carpet snake c. 8' in length, The Inquirer & Commercial Mt Bakewell News 7.12.1894: 22 1895-1896 Common at Roebuck Bay, but 'nowhere numerous'; 'much in request by the natives Collett 1897: 329 for food'. 1896 Tassels hung on string of opossum and human hair Carnegie 1898: 244 1896 Aboriginal party preparing their evening meal, which included opossums. Carnegie 1898: 258 Afterwards saw their tracks on a large eucalypt nearby. [27.9.96] 1897 Aboriginal neckband of opossum wool Carnegie 1898: 391 1897 Aboriginal woman hunting; her catch included 'several 'possums' Carnegie diary 15.5.97 1897 Aborigines had just eaten an opossum Russel 1899: 66 1901 Opossums 'somewhat rare in this country' Murray 1904: 14 1901 'not numerous apparently. I saw only one, which 1 found curled up in a F M House in Brockman "chatterer's" nest' 1902: 51 1901-1902 Possum fur used by Aborigines as string during initiation ceremonies D Bates in White 1985: 166, 168 1905 'a full-grown silver-grey opossum' trapped: The first of its kind I have known to be Kalgoorlie Miner 4.7.1905: 4 seen or caught in these parts. Others were incredulous until they had seen it. The natives were greatly interested, and did not know what it was at first. Afterwards one of them had an excited 'wongi' with his dusky brethren. On being appealed to he said 'That one gammo,' but vouchsafed no other information. [Feysville] 'The only opossum I knew having been found on the Eastern Goldfields was got by The West Australian Mr. Wm. Connaughton just near the Cheapside Gold Mine, and was a white one with 24.12.1927: 5 pink eyes.' ca 1910? 'a coal black male...The Italian charcoal burners got him in the bush at Kensington, The West Australian I think it was...brought him in to...Yundagi [Youndaga], four miles from Menzies, and 24.11.1928: 5 he was on view for some time' ca 1910? Young Gascoyne Aboriginal girls occasionally wore a band of possum fur around D Bates in White 1985: 127 their heads 1921 'Several opossums were observed, and judging by the markings on some of the Easton 1922: 45 trees, are very numerous in certain localities' (north Kimberley region) When a boy, used to get possums out of burrows in the ground, using a dog to locate The West Australian tenanted ones, ?Ka tanning 26.9.1925: 11 Trapped possums at a rabbit warren, ?Narrogin The Wcsf Australian 3.10.1925: 15 Ground possums, no locality The West Australian 10.10.1925: 15 Possums in rabbit burrows, Narrogin Tlw Wes/ Australian 22.5.1926: 9 1937 Two 'mangrove opossums' from Yarraloola [Robe River] loaded on a ship at Onslow The Northern Times for transport to Perth Zoo 17.11.1937: 2 1948 While woodcutting near Ivanhoe station [Kimberley region]. Aborigines found a The Northern Times possum in a hollow log. 'O'possums have rarely, if ever, been seen in the Kimberleys. 25.11.1948: 5 Even the natives admitted that it was a stranger to them'
* See also Abbott (2008 pp. 72-84) for further information relating to this species in southwest Western Australia.
86 Abbott: Trichosurus vulpecula distribution in WA
Table 2 Names used by Aboriginal people in Western Australia for Trichosurus vulpecula*.
Language group/ Geographical location Name(s) Reference
Ord River guman Ngabidj & Shaw 1981: 129 Worora (opposite Augustus Island) bul'kumba Love nd Derby Nan goo in, Kanderal Bates nd 52/132 Sunday Island [?Derby] Lang'orr Bates nd 53/40 Lower Fitzroy River langura Bates nd 51/30 Fitzroy River telegraph station Long o Bates nd 53/1 Fitzroy Chambindie Bates nd 53/92 Halls Creek Nan'goo'in, tchamin'dee, kanderal, nungwin Bates nd 52/132 Halls Creek nunguin Bates nd 53/87 Halls Creek nunguin Mathews 1910: 221 Turkey Creek Nan'gooin Bates nd 53/13 Jaru (Broome) )an8 Anon. 1992 Between Broome & Willie Creek Langoor Bates nd 52/1 Broome Lang'er, lang'ur Bates nd 52/61 Noonkanbah; St George Range gundaman; chambigin Mjoberg 1915 Bardi (N Dampierland peninsula) langgoor, langgooroo Aklif 1999 Yulparija (Bidyadanga) wartawirrkarl, winkuma Anon. 2006 Northern Nyangumarta (Anna Plains) langkurr, winkuma McKelson 1989 Nganumarta (Anna Plains to langkurr Hanson 2007 (McKelson & Warrwagine & Bidyadanga) Dodd 2007) Port Hedland Wall-um-burry Bates nd 50/106 Ngurla (mouth of De Grey River) wallambine Curr 1886 vol. It 292 Banaka-Palyeri (Turner River, Walambari Brown 1913: 163, 165, 189 Sherlock River, Chanjara clan) De Grey River mouth Wallambine Bates nd 50/119 Weedookarry (Shaw River) koobalyie Curr 1886 vol. 1: 294 Shaw River Koobalyie Bates nd 50/123 Pilbara (Koongan) Kooberdee Bates 48/24 Pilbara (Warran/Wandarree) Nyardee, walarra Bates 48/38 Pilbara (Nullagine) wal'lamberree Bates nd 49/45 Nickol Bay wolumberree Curr 1886 vol. 1: 300 Ngalooma (Roebourne) walambarri Hall 1971 Roebourne Wallumberree Bates nd 48/1 Roebourne Kulbin Bates nd 49/142 Roebourne district (Karralla station) Wallamberre Bates nd 49/1 Roebourne district (Tambrey station) Warr-kie Bates nd 50/47 Roebourne Wallumberrie Bates nd 50/101 Roebourne district Walumberri Yabaroo 1899 Roebuck Bay (interior) Parrudine Perth Gazette 5.8.1864: 3-4; Martin 1864: 88 Fortescue Wallambung Bates nd 50/117 Nyamal (H of Marble Bar) marlarlparra Anon. 2007 Watjarri (lower Murchison River & wayurta Douglas 1981 Sanford River) Sanford River (Murchison) Wai'oor'da Bates nd 54/52 Payungu (Minilya & Lyndon Rivers) wayurd Austin 1992e Kakarakala (North West Cape-30 miles marajaak Curr 1886 vol. 1: 304 S of Gascoyne River) Inland Ashburton & upper Hamersley walyi Anon. 2008 Range, Pilbara Jiwarli (between Ashburton River & wayurta Burgman 2006a source of Lyndon River) Ashburton district (Naroo station) Kailburn Bates nd 48/96 Ashburton district Kulyiban Yabaroo 1899 Kurrama (Hamersley Range) kupiti Burgman 2006b Lyons River Wyulda Alexander 1920 Lyons River Gunamoongoo Bates 48/135 Wandajee, Gascoyne & Minilya marra jugga, waioorda Bates nd 48/149 Gascoyne Wyardiar Bates nd 50/1 Carnarvon Wai-oorda Bates nd 48/48 Weld Range (Cue) Wai'oorda Bates nd 54/74 Nyangara (Sandstone) Dharlda Bates nd 54/105 Yooldhura (Morarie & NW of Nannine) Wai'oorda Bates nd 54/117 SW of Peak Hill (Meekatharra) Wai'oorda Bates nd 54/140
87 Journal of the Royal Society of Western Australia, 95(2), July 2012
Table 2 (cont.)
Language group/ Geographical location Name(s) Reference
NW of Peak Hill (Wamgun & Yalladnoolgu) Jangala Bates nd 54/167 Peak Hill (Murchison district) Jang'ala Bates 54/200 Yuin (Murchison district) wiurda Bates nd 57/99 Yooin waioorda Bates nd 58/86 Yuin-Murgoo district Wyurda F Wittenoom in Lefroy 2003: 39 Annean station (Murchison district) Wy-a-da Bates nd 57/101 Yuin Creek via Annean Gnu-err, y-ar-do Bates nd 57/102 Mt View (Murchison district) Wi-yau-dar Bates nd 57/105 Duketon (Mt Magnet district) Wy-a-doo, Gin-yi-ga Bates nd 58/73 Yoweragabbie (Mt Magnet district) waiadu L A Wells in Lindsay 1893: 78 Watchandie [?Wooramel River, Shark Bay] We-urda, We-zue-da Oldfield 1865 Cheangwa (Murchison district) wioora Bates nd 58/86 Gullewa (Murchison district) waioorda Bates nd 58/86 Shark Bay yoorda Curr 1886 vol. 1: 308 Mouth of Murchison River weurda Curr 1886 vol. 1: 312 Nhanda (Kalbarri - Port Gregory) wiyarda Blevins 2001 Northampton Wai'oor'da Bates nd 54/24 Oakabella waioora Bates nd 58/86 Champion Bay [Geraldton] Wy-a-der Bates nd 58/86 Victoria Plains (Dongara) Wai'oor'da Bates nd 54/1 Pinjarrah (EAA Fawcett) Coumal Bates nd 43/97 Pinjarrah (W Hymus) Koom-marl Bates nd 43/98 Karatjari langkurr, winkuma Burbidge el al. 1988 Kartutjarra nyalumpara, nyuwilpa, wayurta, winkuma Burbidge et al. 1988 Kartujarra (W of Lake Disappointment) wayurta Anon, nd Kukatja kitu, kuringka, marlalparra, ngarlumpa, Burbidge et al. 1988 nuywilypa, tjampinti, tjanganpa, wartawirrka, wayurta, wina-wina Kukatja (Balgo) tjaatutu, tjapiyinti, tjawi-tjawi, kumputinti, Valiquette 1993 malaparra, mulalpara, mungawayuru, pirtikarra, tatikara, tjanpiyinti, tjanganpa, wartangkakawirrpa, wayurta Luritja numta, wayurta, yimurra Burbidge el al. 1988 Pitjantjatjara (Kata Aria Winjelinna) wayuta Anon. 1987 Mangala langkurr, tjinapapi, winkuma Burbidge et al. 1988 Manytjilytjarra marlalparra, nyalumpara, partawirrka, Burbidge et al. 1988 wayurta, winkuma Ngaanyatjarra mungawayurru, tjaparlpa/tjapalpa, tjiwangu, Burbidge et al. 1988 wayurta Ngaanyatjarra & Ngaatjatjarra (near Northern wayurta Glass & Hackett 2003 Territory/South Australia border) Ngaatjatjarra marliyara, mungawayurru, tjaparlpa/tjapalpa. wayurta Burbidge et al. 1988 Nyamal wanitjara, wantjinara Burbidge et al. 1988 Nyangamarta marlalparra, wanitjara, wantjinara, winkuma Burbidge et al. 1988 Nyiyapali kaparti, marlalparra, nyalumpara, turrumanta Burbidge et al. 1988 Parti langkarr/langkurr, walanpara? [sic] Burbidge et al. 1988 Pintupi katatjurta, marlalparra, marliyarra, mungawayurru, nyunta, parntawirrka, pirtikarra/pirtikarratja, tjanganpa, wayurta. wina-wina, yurltukarra/yurltukarratja Burbidge et al. 1988 Putitjarra Wayurta Burbidge et al. 1988 Tjaru Tjanganpa Burbidge et al. 1988 Walmatjari kitu, tatikarra, tjanpitjin, tjanpiyinti, winkuma, Burbidge et al. 1988 wayurta Wangkatjungka kitu, marlalparra, ngarlumpa, nuywilypa, tatikarra/talikarra, tjanganpa, wartawirrka. wayurta, wina-wina, winkuma Burbidge et al. 1988 Warnman marlalparra, winkuma Burbidge et al. 1988 Warnman Papumalu Anon. 1991 Western Desert (Kalgoorlie, Ceduna SA, wayurta, tjaparlpa Douglas 1988 Wiluna, Alice Springs NT) Lake Way waiurda Bates nd 58/86
88 Abbott: Trichosurus vulpecula distribution in WA
Table 2 (cont.)
Language group/ Geographical location Name(s) Reference
Norseman, Southern Cross, Wamba, womba, Jammar Bates nd 47/10 E goldfields district Lawlers (East Murchison district) waioordoo, dhoogalleeda Bates nd 58/86 NE Laverton waiurda Bates nd 58/86 E of Laverton waiurda, tugalida Bates nd 58/86 Laverton waioordoo Bates nd 58/86 Kookyne (Menzies district) kinnika, wyadoo Bates nd 58/44 Mulline (Menzies district) waioorda Bates nd 58/86 Canegrass (Jaurdi) Jammunmaaroo, Jammun Bates nd 58/86 Coolgardie Jamma Bates nd 47/44 Feysville gammo Kalgoorlie Miner 4.7.1905: 4 Between Southern Cross & Tchamma, jamma Bates nd 47/68, 47/12 Mt Jackson, Burracoppin Israelite Bay Janna Bates nd 58/86 Fraser Range Damma Helms 1896: 322 Balladonia jamma Bates nd 58/86 Wonunda Meening (Eyre Sand Patch [Point pilda Curr 1886 vol. 1: 398 Culver to 140 miles E and inland for 30-40 miles]) Eyre, Eucla Bilda Bates nd [71912-14] 58/86 Twilight Cove Bilda Bates nd [71912-14] 58/86 Yircla Meening (Eucla to 100 miles E, bilda Curr 1886 vol. 1: 404 40 miles W and N to Nullarbor Plain) Nullarbor Plain (South Australian sector) bilda D M Bates in The Australasian 10.8.1918: 279 Karura (Eucla district) Bilda, Jamma Bates nd [71912-14] 58/86
* Not included in this Table are 67 Noongar (southwest Aboriginal) names for T. vulpecula (Abbott 2001)
(?Wooramel River) by Oldfield (1865) and the Jiwarli one Western Australia island is intriguing. Barrow Island (upper Lyons River) by Burgman (2006a), no post- is a large island (23 950 ha) that was never visited by European settlement records of this species were found Aborigines. Dirk Hartog Island is about three times around the coastline of Shark Bay. larger than Barrow Island. At its southern end it is very close to the mainland, and Aborigines may have lived on Kerle & How (2008) showed T. vulpecula as present on and travelled to and from the island until the final the Nullarbor Plain as recent subfossil material, evidently connection to the mainland was breached relatively based on the accounts of Lundelius (1963) and Baynes recently (~5 ka BP). Although able to swim. Aborigines (1987). No post-European settlement records of this apparently ceased to visit by 3.5 ka BP, as no subfossil species were found actually in the Western Australia remains of the dingo Canis lupus dingo have been sector of the Nullarbor Plain, only to the north and south. discovered on Dirk Hartog Island (Baynes 1990). All However, there is a sight record from the South accounts of the island between 1697 and 1858 indicate Australian sector of the Nullarbor Plain (Figure 2). that Aborigines did not occur there (1 Abbott unpubl. Presumably this species on the Nullarbor Plain sheltered data). Although rafts and logs were in use by Aborigines mainly in warrens made by the boodle, Bettongia lesueur. in large rivers of the nearby Gascoyne and Murchison T. vulpecula is too large an animal to have been regions (Bowdler 1995), no rivers occur near Dirk Hartog captured by barn owls Tyto alba. The masked owl T. Island, and none of the early European visitors recorded novaehollandiae occurs in the northern Kimberley and watercraft in use in Shark Bay. One of the attractions for lower southwest regions of Western Australia but nests Aborigines to visit this island may have been marine in tree hollows, not caves (Johnstone & Storr 1998). T. turtle breeding grounds situated at the northern end. As vulpecula is known to have sheltered in caves and under well as tire absence of T. vulpecula, this island has other breakaways, where these are available (Burbidge & anomalies, particularly the absence of macropodid Fuller (1979). Caution should therefore be exercised in mammals that have persisted on the smaller but more presuming that local absence, particularly in arid regions, remote Bernier and Dorre Islands (Baynes 1990). These of subfossil material of this species necessarily indicates absences may have resulted from predation by that possums did not occur. Aborigines, as hypothesised more generally for many Australian islands by Abbott (1980). T. vulpecula is unknown as subfossil on Western Australia's largest island (Dirk Hartog Island: Burbidge Tire minimum area of islands elsewhere in Australia & George 1978). Therefore, the record of 'a small that are known to support a natural population of T. opossum' in 1822 (King 1827 p. 183) is likely to be a vulpecula is -600-700 ha (Abbott & Burbidge 1995). In the misidentified mammal of similar size, most probably mesic Kimberley region of Western Australia there are Bettongia lesueur. The absence of T. vulpecula from all but many islands of this and larger area. Some of these are
89 Journal of the Royal Society of Western Australia, 95(2), July 2012
Table 3 Apparent extinction of Trichosums vulpecula in various localities in Western Australia, based on replies received to a circular letter from the Police Commissioner asking officers to ascertain if the closed season has had the desired result of increasing the number of possums in the various police districts (from Colonial Secretary's Department Aborigines and Fisheries file 652/108/5a/992).
Police district Date Comments
Broome -.5.1912 Extinct for many years Yalgoo 3.7.1913 None in this sub-district. Old residents say that they have never seen any here Nannine 18.6.1913 'not indigenous to this Sub-district & there is no record of them having been ever met with in the [?sub-district], not even by natives' Sandstone 9.5.1912 None present & there never has been any Sandstone 23.6.1913 'so far as I can leam there never has been any' Wiluna 12.5.1912 None present Wiluna 15.6.1913 None in this sub-district Peak Hill 13.5.1912 Do not exist in this district & have never heard of any resident having seen any here Meekatharra 9.5.1912 None present (interviewed station employees, prospectors, & Aborigines) Meekatharra 30.6.1913 None present in this sub-district (Police Constable asked old prospectors and Aborigines) Mt Magnet 7.5.1912 None exist in this sub-district Mt Magnet 11.6.1913 Enquired of Dowden Bros (c. 40 miles E), James Bros (c. 18 miles W) & W Watson (local store owner): None seen for at least 16 years [1897] or nor have they heard of any being caught Youanmi 8.5.1912 None ever present (old residents interviewed) Youanmi 15.6.1913 None in district Cue 14.6.1913 None seen in district - last seen > 9 years ago [< 1904] at McGrath's station, Kalli [NW of Cue] Carnarvon 20.5.1912 None in any part of Gascoyne District Carnarvon 13.5.1913 'there are no opossums in the Gascoyne District' Bullfinch 11.6.1913 There has never been any here, so far as can be ascertained Marvel Loch 9.6.1913 None in the sub-district Southern Cross 9.6.1913 None in the sub-district Menzies 4.5.1912 None in this district Kalgoorlie 11.6.1913 None
Table 4 Other records of apparent extinction of Trichosums vulpecula in Western Australia, additional to those published in Abbott (2006, 2008).
Locality Date Comments Reference
East of Malara Between 'In 1860 I was out the other side of Malara cutting sandalwood, and The Western Mail [27°35’, 115°43'] 1860 & 1897 the country abounded with opossums, but now they are extinct - 24.9.1897: 47 through the droughts, the natives say' Burracoppin 'all dead' Bates nd 47/12 1901 'some virulent disease, such as has decimated opossums' The Inquirer & Commercial News 1.2.1901: 11 Karura (Eucla <1912-1914 'none at Eucla' Bates nd [71912-14] 58/86 district) Near Coomberdale 1915 One found recently: 'it is stated that it is thirty years [1885] since The Geraldton Guardian (Moora district) opossums have been found so far east' 20.2.1915: 2 composed (at least partially) of fertile volcanic soils and Abbott 2008), and in large termitaria in spinifex country should support woodland with nutrient-rich foliage, a (Burbidge et al. 1988). On Barrow Island, which lacks habitat characteristic known to facilitate high densities of trees, T. vulpecula lives in caves, rock piles and termite arboreal marsupials (Braithwaite et al. 1983). However, all mounds (Butler 1970; Morris & Burbidge 2002). of these islands were accessible to Aboriginals, so that the The wide distribution of T. vulpecula demonstrated by unexpected absence of T. vulpecula may have resulted from the historical approach of this paper aligns with the unsustainable predation over a long period of -5 ka. occurrence of T. vulpecula in all but one (Central In many of the early accounts located and reviewed, Kimberley) of the 26 bioregions in Western Australia T. vulpecula was not recorded even though this was well (Burbidge et al. 2008). This study also demonstrates that a before the arrival of the fox in Western Australia in 1912. distribution map based solely on museum specimens This is suggestive of a very low density of occurrence (OZCAM 2012) is misleading. and sparse dispersion across the landscape in the interior of Western Australia, perhaps linked to the paucity of watercourses. It has been demonstrated for the Northern Extinction of populations Territory and South Australia that T. vulpecula is (or was) The most recent map available of the geographical most abundant in rocky ranges and outcrops and along distribution of T. vulpecula in Western Australia (Kerle & watercourses (Kerle et al. 1992). Nevertheless, T. vulpecula How 2008, redrawn here as Figure 1) shows that this can live without trees and is known to have sheltered in species has become extinct since European settlement warrens made by Bettongia lesueur (Burbidge et al. 1988; over a large portion of its Western Australia range. Most
90 Abbott: Trichosurus vulpecula distribution in WA of this area receives low annual rainfall, so that densities ACKNOWLEDGEMENTS of T. vulpecula were probably very low most of the time. I thank Phil Bianchi for the extracts from the diary of In the 1880s an epizootic disease swept through the David Carnegie, Aaron Rivers for preparing the maps, western portion of the interior (Abbott 2006) and this Judy Dunlop for providing information about the Lorn species, one of many, declined and in many places Glen translocations, Andrew Burbidge and Neil Burrows became locally extinct. Kerle (2004) also cited records of for commenting on a draft, and John Dell and an disease causing almost local extinction of this species in anonymous reviewer for their comments. Queensland in 1881 and 1888, and also in New South Wales in 1950. Local extinction caused by 'some eruptive skin disease' in 1881-1882 of T. vulpecula along the REFERENCES Darling River, in the district between Broken Hill and Wilcannia, and in 1884-1885 along the Barwon River, Abbott I 1980. Aboriginal man as an exterminator of wallaby and kangaroo populations on islands round Australia. New South Wales, has also been placed on record (Royal Oecologia 44, 347-354. Commission 1890 pp. 41, 84, 96; The Western Mail 8.9.1900 Abbott I 2001. Aboriginal names of mammal species in south¬ p. 5). west Western Australia. CALMScience 3, 433—186. Unlike the wheatbelt of Western Australia and similar Abbott I 2006. Mammalian faunal collapse in Western Australia, agricultural landscapes elsewhere in Australia (Kerle 1875-1925: the hypothesised role of epizootic disease and a 2004), habitat change (as tree removal) is obviously not a conceptual model of its origin, introduction, transmission, and spread. Australian Zoologist 33, 530-561. relevant factor in causing extinction of T. vulpecula throughout the deserts of Western Australia. Abbott I 2008. Historical perspectives of the ecology of some conspicuous vertebrate species in south-west Western The presence of exotic predators, particularly the fox, Australia. Conservation Science Western Australia 6, 1-214 has also greatly impacted on the distribution and (Corrigendum 7, 633). abundance of T. vulpecula (Burrows & Christensen 2002; Abbott I 2009a. Appendix H. Fauna. In: Bridge P J, Epton K, Orell 2004). Kerle et al. (1992) have proposed a conceptual Hercock M & Milentis S (eds) The Finest Gold Fields in the World, the Austin Expedition to the Murchison in 1854, pp. 157- model for the arid zone that links drought and fox 166. M G Creasy & Hesperian Press, Carlisle. predation together in explaining the widespread Abbott 1 2009b. Appendix 25. Fauna. In: Chambers M & Bridge extinction of T. vulpecula in the arid zone of Australia. P J (eds) Desert, Drought, and Death, the Border Exploration Following successful fox control via aerial deployment of Syndicate Expedition to the Rawlinson Range 1899-1900 by poison baits, translocation of 95 animals from southwest Henry William Hill, pp. 124-125. M G Creasy & Hesperian Western Australia to Lorna Glen, a former pastoral Press, Carlisle. property, has been accomplished successfully during the Abbott 1 & Burbidce A A 1995. The occurrence of mammal period 2007-2009 (Dunlop & Morris 2009; Miller et al. species on the islands of Australia: a summary of existing 2010). The long-term success of this translocation is yet to knowledge. CALMScience 1, 259-324. be determined. Aklif G 1999. Ardiyooloon Bardi Ngaanka One Ann Point Bardi Dictionary. Kimberley Language Resource Centre, Halls Creek. Aboriginal names for Trichosurus vulpecula in Western Alexander W B 1920. Aboriginal names of the animals of the Australia Lyons River district. Journal and Proceedings of the Royal Society of Western Australia 6, 37-40. It is evident from Table 2 and Abbott (2001) that there are strong regional groupings of Aboriginal names. The Anon. nd. Kartujarra Wordlist. Wangka Maya Pilbara Aboriginal Language Centre, South Pledland. vernacular name used by contemporary zoologists is 'common brushtail possum' (Kerle & How 2008). This is Anon. 1987. A Basic Pitjantjatjara/Yankunytjatjara to English Dictionary. Institute for Aboriginal Development, Alice no longer appropriate for Western Australia as the Springs. species is now nowhere common in Western Australia. Anon. 1991. Warnman Wordlist. Wangka Maya Pilbara Even in the more mesic southwest of Western Australia Aboriginal Language Centre, Port Hedland. this species is clearly now much reduced in abundance Anon. 1992. Jaru Dictionary. Kimberley Language Resource compared with the numbers that were hunted or snared Centre, Halls Creek. for their pelts in the period 1899-1910 (Abbott 2008; The Anon. 2006. Yulparija. Wangka Maya Aboriginal Language West Australian 17.7.1925 p. 10). Centre, South Pledland. The term 'possum' is also inappropriate, being derived Anon. 2007. Nyanial Dictionary. Wangka Maya Pilbara from the North American (Algonquian) word Aboriginal Language Centre, South Hedland. wapathemwa, meaning 'white animal'. This refers to the Anon. 2008. Yinhawangka Cictionary English-Yinhawangka Virginia opossum, Didelphys virginiana, which belongs to Wordlist and Topical Wordlists. Draft 1. Wangka Maya Pilbara a different family (Didelphidae) of marsupials. An Aboriginal Language Centre, Port Hedland. Aboriginal word would seem more preferable to a North Austin P 1992a. A Dictionary of Yinggarda, Western Australia. La Trobe University, Bundoora. American indigenous name. Kerle & How (2008) do not list any Aboriginal names for T. vulpecula. Austin P 1992b. A Dictionary of Warriyangga, Western Australia. La Trobe University, Bundoora. For the above two reasons, it seems more appropriate Austin P 1992c. A Dictionary of Thargarri, Western Australia. La that T. vulpecula be called the langkurr or nunguin in the Trobe University, Bundoora. Kimberley region, wallamberri in the Pilbara region, Austin P 1992d. A Dictionary of Thatanyji, Western Australia. La wayurta in desert regions, koomal in southwest Western Trobe University, Bundoora. Australia, jarrtma in thetgoldfields, and bilda along the Austin P 1992e. A Dictionary of Payungu, Western Australia. La southeast coast. Trobe University, Bundoora.
91 Journal of the Royal Society of Western Australia, 95(2), July 2012
Austin P 1992f. A Dictionary of Jiwarli, Western Australia. La Pioneering and Exploration in Western Australia. Pearson, Trobe University, Bundoora. London.
Basedow H 1904. Anthropological notes made on the South Clarkson B D 1864. Journal of an expedition. The Inquirer & Australian Government North-West prospecting expedition, Commercial Neivs 21.9.1864, 2-3. 1903. Transactions of the Royal Society of South Australia 28, Collett R 1897. On a collection of mammals from north and 12-51. north-west Australia. Proceedings of the Zoological Society of Bates D M nd (? 1904-12). [Manuscript or typescript London 1897, 317-336. vocabularies] Section XII. Part 2. Language: Grammar and Colonial Secretary's Department. Aborigines and Fisheries 1913. Vocabularies. Western Australia (folio 365/39/1-237, 43/97- Microfiche 652/108/5a/992. State Records Office of Western 100, 47/1-128, 48/1-216, 49/1-133, 50/1-122, 51/1-83, 52/1-199, Australia, Perth. 53/1-111, 54/1-232, 55/1-159, 56/1-194, 57/1-126, 58/1-109). National Library of Australia, Canberra. Curr E M 1886. The Australian Race: its origin, languages, customs, place of landing in Australia, and the routes by which it spread Baynes A 1987. The original mammal fauna of the Nullarbor itself over the continent. Volume 1. Government Printer, and southern peripheral regions: evidence from skeletal Melbourne. remains in superficial cave deposits. In: McKenzie N L & Robinson A C (eds) A Biological Survey of the Nullarbor Region Dahl K 1927. In Savage Australia, an account of a hunting and South and Western Australia in 1984, pp. 139-152. Government collecting expedition to Arnhem Land and Dumpier Land. Allan, Printer, Adelaide. London. Baynes A 1990. The mammals of Shark Bay, Western Australia. Davidson A A 1905. Journal of Explorations in Central Australia. In: Berry P F, Bradshaw S D & Wilson B R (eds) Research in Government Printer, Adelaide. Shark Bay. Report of the Trance-Australe Biccnlenanj Committee, Donaldson M & Elliot I (Editors) 1998. Do Not Yield to Despair. pp. 313-325. Western Australian Museum, Perth. Frank Hugh Hann's Exploration Diaries in the Arid Interior of Baynes A & Jones B 1993. The mammals of Cape Range Australia, 1895-1908. Hesperian Press, Victoria Park. peninsula, north-western Australia. Records of the Western Douglas W H 1981. Watjarri. In: Dixon R M W & Blake B J (eds) Australian Museum Supplement 45, 207-225. Handbook of Australian Languages, pp. 196-272. Australian Blevins J 2001. Nhanda an Aboriginal Language of Western National University Press, Canberra. Australia. University of Hawai'i Press, Honolulu. Douglas W H 1988. An Introductory Dictionary of the Western Boscacci L J, McKenzie N L. & Kemper C M 1987. Mammals. In: Desert Language. Western Australian College of Advanced McKenzie N L & Robinson A C (eds) A Biological Survey of Education, Perth. the Nullarbor Region South and Western Australia in 1984, pp. Dunlop J & Morris K 2009. Into the wild: restoring rangelands 103-107. Government Printer: Adelaide. fauna. Landscape 24(4), 52-58. Bowdler S 1995. The excavation of two small rockshelters at Easton W R 1922. Report of the North Kimberley district of Western Monkey Mia, Shark Bay, Western Australia. Australian Australia. Government Printer, Perth. Archaeology 40, 1-13. Exploration Diaries 1827-1871. Vols 1-6, typescript. Battye Braithwaite L W, DudzNski M L & Turner J 1983. Studies on the Library, Perth. arboreal marsupial fauna of eucalypt forests being harvested Eyre E J 1845. Journals of Expeditions of Discovery into Central for woodpulp at Eden, N.S.W. II. Relationship between the Australia and overland from Adelaide to King George's Sound. T fauna density, richness and diversity, and measured & W Boone, London. variables of the habitat. Australian Wildlife Research 10, 231- 247. Forrest J 1875. Explorations in Australia. Sampson Low, Marston, Low & Searle, London. Brockman F S 1902. Report on Exploration of North-West Kimberley 1901. Government Printer: Perth. Giles E 1889. Australia Twice Traversed...from 1872 to 1876. Sampson Low, Marston, Searle & Rivington, London. Brown R 1913. Three tribes of Western Australia. Journal of the Royal Anthropological Institute of Great Britain and Ireland 43, Glass A & Hackett D 2003. Ngaanyaljarra & Ngaatjatjarra to 143-194. English Dictionary. IAD Press, Alice Springs.
Burbidce A A & Fuller P J 1979. Mammals of the Warburton Gregory A C & Gregory F T 1884. Journals of Australian region. Western Australia. Records of the Western Australian Explorations. Government Printer, Brisbane. Museum 8, 57-73. Gregory F T 1862. Expedition to the north-west coast of Burbidce A A & George A S 1978. The flora and fauna of Dirk Australia, journal of the Royal Geographical Society of London Hartog Island, Western Australia. Journal of the Royal Society 32, 372-429. of Western Australia 60, 71-90. Grey G 1841. Journals of Txvo Expeditions of Discovery in North- Burbidce A A, Johnson K A, Fuller P J & Southgate R I 1988. West and Western Australia, during the Years 1837, 38, and 39. T Aboriginal knowledge of the mammals of the central deserts & W Boone, London. of Australia. Australian Wildlife Research 15, 9-39. Giubble J B 1905. Dark Deeds in a Sunny Land or Blacks and Burbidce A A, McKenzie N L, Brennan K E C, Woinarski J C Z, Whites in North-West Australia. Daily News, Perth. Dickman C R, Baynes A, Gordon G, Menkhorst P W & Hall H A 1971. A Partial Vocabulary of the Ngalooma Aboriginal Robinson A C 2008 Conservation status and biogeography of Tribe. Australian Institute of Aboriginal Studies, Canberra. Australia's terrestrial mammals. Australian Journal of Zoology 56, 411-422. Helms R 1896. Anthropology [of the Elder Scientific Expedition]. Transactions of the Royal Society of South Australia 16,: 237-332. Burgman A 2006a. Jiwarli Dictionary. Wangka Maya Pilbara Aboriginal Language Centre, South Hedland. Helms R 1897. East Kimberley. Journal of the Bureau of Agriculture of Western Australia 4, 1283-1287, 1299-1303, 1306-1311. Burgman A 2006b. Kurrama-English Dictionary. Wangka Maya Pilbara Aboriginal Language Centre, South Hedland. How R A 1983. Common brushtail possum Trichosurus vulpecula. In: Strahan R (ed.) The Australian Museum complete book of Burrows N D & Christensen P E S 2002. Long-term trends in Australian mammals, pp. 147-148. Angus & Robertson native mammal capture rates in a jarrah forest in south¬ Publishers, Sydney. western Australia. Australian Forestni 65, 211-219. How R A, Dell J & Muir B G 1988. Vertebrate fauna. The Butler W H 1970. A summary of the vertebrate fauna of Barrow biological survey of the eastern goldfields of Western Island, W.A. Western Australian Naturalist 11, 149-160. Australia. Part 4. Lake Johnstone-Hyden study area. Records Carnegie D W 1898. Spinifex and Sand. A Narrative of Five Years' of the Western Australian Museum Supplement 30, 44-94.
92 Abbott: Trichosurus vulpecula distribution in WA
How R A & Kerle J A 1995. Common brushtail possum McKenzie N L & Youngson W K 1983. Mammals [of the Great Trichosurus vulpecula. In: Strahan R (ed.) The mammals of Sandy Desert]. Wildlife Research Bulletin Western Australia 12, Australia pp. 273-275. Reed Books, Chatswood. 62-93. Hunt C C 1864. Journal of Exploration Eastward of York in 1864 Miller E, Dunlop J & Morris K 2010. Rangelands restoration: (typescript of Field Book). Battye Library, Perth. fauna recovery at Lorna Glen Western Australia. Progress Report August 2008- June 2010. Department of Environment Johnstone R E & Storr G M 1998. Handbook of Western Australian and Conservation Library, Kensington. Birds. Vol. 1. Western Australian Museum, Perth. MjOberg E 1915. Among Wild Animals and People in Australia Jones B 2004. Tire possum fauna of Western Australia: decline, (translated by M Luotsinen & K Akerman, 2012). Hesperian persistence and status, in: Goldingay R L & Jackson S M (eds) The Biology of Australian Possums and Gliders, pp. 149- Press, Carlisle. 160. Surrey Beatty & Sons, Chipping Norton. Morris K & Burbidce A 2002. Bountiful Barrow. Landscope 17(3), 18-24. Kerle A 2001. The Brushtails, Ringtails and Greater Glider. University of New South Wales Press, Sydney. Murray W R 1904. Extracts from journals of explorations by R. T. Maurice, Fowler's Bay to Rawlinson Ranges and Fowler's Kerle A 2004. A cautionary tale: decline of the common brushtail Bay to Cambridge Gulf. Parliamentary Paper South Australia possum Trichosurus vulpecula and common ringtail possum Pseudocheirus peregrinus in the woodlands of the western 43, 1-39. slopes and plains of New South Wales. In: Goldingay R L & Ngabidj G & Shaw B 1981. My Country of the Pelican Dreaming. Jackson S M (eds) The Biology of Australian Possums and Gliders, The Life of an Australian Aborigine of the Gadjerong... 1904-1977. pp. 71-84. Surrey Beatty & Sons, Chipping Norton. Australian Institute of Aboriginal Studies, Canberra. Kerle J A 1983. Northern brushtail possum Trichosurus Oldfield A 1865. On the Aborigines of Australia. Transactions of amhemensis. In: Strahan R (ed.) The Australian Museum the Ethnological Society of London 3, 215-298. Complete Book of Australian Mammals, pp. 149. Angus & Orell P 2004. Fauna monitoring and staff training: Western Robertson Publishers, Sydney. Shield review - February 2003. Conservation Science Western Kerle J A, Foulkes J N, Kimber RG & Papenfus D 1992. The Australia 5, 51-95. decline of the brushtail possum, Trichosurus vulpecula (Kerr OZCAM 2012. Online map of museum specimen records of 1798), in arid Australia. Rangelands journal 14, 107-127. Trichosurus vulpecula.
93 4- Journal of the Royal Society of Western Australia, 95: 95-102, 2012
Grasstree stem analysis reveals insufficient data for inference of fire history
B P MILLER r, T WALSHE 2, N J ENRIGHT 3 & B B LAMONT 4
1 Kings Park and Botanic Gardens and Parks Authority, Fraser Avenue, West Perth,WA 6005, Australia. 2 School of Botany, University of Melbourne, Vic 3010, Australia. 3 School of Environmental Science, Murdoch University, WA 6150, Australia. 4 Department of Environment and Agriculture, Curtin University, GPO Box U1987, Perth, WA 6845, Australia. ' Corresponding author El [email protected]
Grinding back dead leaf bases on the stems of arborescent grasstrees (Xanthorrhoea spp.) reveals a pattern of horizontal bands that has been interpreted as a record of the fire history experienced by the plant. The validity of this fire history record has previously been assessed through comparison of 100 grasstree stems sampled from shrubland near Eneabba in Western Australia against a 30 year fire history determined from satellite imagery. This analysis showed that the two records matched more than would be expected by chance, but concluded that tire interpretation of the grasstree record as a fire history was not warranted as most of the grasstree fire records did not match satellite fire records. A second analysis of the same two sets of records, published in this journal, also showed that the records matched more than by chance, but concluded that the interpretation of grasstree banding as fire history was valid, though it failed to quantify the strength of this agreement. Here we examine the approaches and interpretations of the two previously published studies, and provide new analyses to refine estimates of the amount of fire- related data present in the grasstree record. We show that only -20% of grasstree 'fire' records may be attributable to fire. With eight out of ten of records not attributable to fire, we confirm that the grasstree record in its current form cannot be interpreted as fire history, and therefore claims of the grasstree technique to support management actions are untenable.
KEYWORDS: fire history, fire management, grasstrees, Xanthorrhoea spp.
INTRODUCTION some of the grasstree fire records could be attributed to fire (as indicated by the satellite image record), most Quantified recent and historic fire histories are useful in could not, and further, that this rate of error increased the interpretation of modern ecological processes and into the past. That is, the grasstree record tended to patterns, as well as for understanding ecological history overestimate the frequency of fire in the distant past and inferring past cultural practices. In turn, this more so than in the recent past. We concluded that while knowledge may be used to construct management plans the grasstree record did contain some information about aimed at conservation or restoration of ecosystems and fire, its interpretation as a fire history was not warranted biodiversity values. An approach to determining because of the dominance of non-fire signals. Subsequent landscape fire histories has been described that analyses re-analysis of the same data also found that the records the pattern of lighter and darker bands revealed on the do match more than by chance (Ward 2009). However, stems of grasstrees (Xanthorrhoea spp.) after their leaf failing to quantify the strength of this agreement, this bases have been ground back (Ward et al. 2001). While second analysis made the unsupported conclusion that the results and interpretation of this technique have the interpretation of grasstree banding as fire history is proven controversial on several grounds (Lamont et al. valid. 2003; Wardell-Johnson et al. 2004; Enright et al. 2005a,b, 2006; Gill 2006; Ward 2006, 2009; Miller et al. 2007), they Here we compare the findings and approach of the continue to be promoted as support for a high frequency two papers (Miller et al. 2007; Ward 2009) and examine of managed fire in many southwest Australian the latter's major criticisms of the former. Primarily, ecosystems (Ward 2009, 2011). As no tests of the validity Ward asserted that in Miller et al. (2007) we: (i) of the technique had been made (apart from showing misaligned the grasstree and satellite records and failed that the colour changes have an anatomical basis: to take into account the inherently lower precision and Colangelo et al. 2002), we conducted a study comparing potential systematic bias in grasstree record dating; (ii) the grasstree technique against a fire history derived inappropriately excluded data; (iii) unduly relied upon from satellite imagery for the Eneabba region (Miller et an inappropriate statistic; and (iv) failed to contemplate al. 2007). In that study we compared the fire history or analyse variation in fire intervals. The remaining record from an annual sequence of satellite images for criticisms of Ward (2009) were either not made explicit the period 1973-2002 with one derived from a sample of (e.g. on extrapolating findings from shrublands to 100 grasstrees in the same area. We found that while woodlands or forests), or were issues that have been addressed previously that Ward returned to, but added no new arguments. An example is the criticism that we © Crown Copyright failed to account for fine-scale variability and patchiness
95 Journal of the Royal Society of Western Australia, 95(2), July 2012 in fires: Ward (2009) repeated the argument that WARD'S METHODOLOGICAL grasstrees are lit or extinguished individually by various CRITICISMS agents. We clearly addressed this issue in Miller et al. (2007), stating that, firstly there is no evidence for this variability in the 1973-2002 study period, and secondly Misalignment of the records and failure to consider that (if so) it fundamentally invalidates any extrapolation grasstree record precision of the grasstree results to the landscape scale—arguments Ward (2009) provided a number of justifications for re¬ that Ward overlooks. analysing the data. One of which was that our analysis As in Miller et al. (2007), and previous papers (Enright was of non-matching census years. In fact, his et al. 2005b, 2006), we re-affirm our view that grasstree representation of the annual period that we consider (and banding does appear to record valuable ecological clearly define) was mistaken (Figure 1). In our analysis, information (especially in relation to growth rates and we labelled the year of the satellite image so that, for plant age: Lamont et al. 2004), and that some fraction is instance, a year bounded by images from September 1981 likely to represent fires. In Miller et al. (2007) we and August 1982 is labelled "1982". However, our provided an estimate of rates of true positive fire records statement that 'fire year here is assigned as the 12-month in the grasstree data. However, we also noted that some period from September to August' (Miller et al. 2007 p. positive grasstree records would be expected to match 910) was overlooked by Ward who believed we used a satellite fire records by chance alone, even if the grasstree calendar year for the census period. He also assumed record was random. Here, we address Ward's criticisms that the grasstree year is equivalent to the calendar year of our earlier paper, and extend our analyses to quantify although, representing seasonal changes in phenology more clearly the fraction of grasstree bands which may (Lamont et al. 2004), it is more likely to correspond to our be interpreted as associated with fire. fire year (Figure 1).
Figure 1 Comparison of census years and precision windows for grasstree and satellite records considered by Miller et al. (2007) (A- G), versus those of Ward (2009) (H- I). Satellite image dates (A) define the satellite census year (B) but vary with availability around August- September (arrows delimit one sample year). Miller et al. analysed varying degrees of temporal precision by resizing a symmetric window of agreement around each fire record to 1, 3, 5, 7 and 9 years (B-F). While the duration of grasstree annual banding is not known, in southwest Australia grasstree phenology changes from a slow to a fast growth phase in late winter/early spring (Lamont et al. 2004, Korczynskyj & Lamont 2005), suggesting a grasstree growing year (G) similar to the satellite year. Ward implicitly assumed grasstree year = calendar year (H) and takes this as the census year. As each of these overlap two satellite years. Ward (2009 p. 262) allowed '[grasstree fire dates] a tolerance of a calendar year either way, and fires dated from satellite images a year's extension backwards' —an offset asymmetrical agreement window of four years (1).
96 Miller et al.: Grasstrees insufficient for fire history
Ward also stated that re-analysis of the data was Ward additionally claimed that we omitted 12 required because we overlooked problems with the particular records of fire - although his table 1 (Ward known imprecision of grasstree records. In fact, we 2009) showed only 10 new records. It is not clear what explicitly recognised these issues, devoting one quarter these instances represent: we have rechecked our original of our paper, including a section titled 'Varying data files and they match all the way through to our precision', solely to their consideration. These results analyses. The omissions are described only from the four were plainly and clearly shown in that paper's figure 3 sites that burnt in 2002 and were therefore recently burnt and table 4, and also were considered in detail in its at the time of survey in 2004. It is possible that the abstract and discussion. excluded records lie outside of our comparison period, which we constrained to the three-decade period 1973- We responded to the inaccuracy of dating of fire scars 2002, as it corresponded with the satellite fire-history on grasstree stems by varying the window across which record that we constructed in 2003. Although we did not we analysed agreement with the satellite record from 1 to have satellite imagery for the period between the date of 9 years (i.e. from an exact match to a match somewhere the last image in 2002 and the time of the grasstree within +4 years), and we discussed how increasing survey in 2004, we know that there had been no fires in window size would increase the likelihood of finding the study area and so we could have extended (and agreement between records if they were distributed by thereby hybridised) the record accordingly. However, we chance. Allowing a 9 year window for matches would did not do this for several reasons: (i) to facilitate almost guarantee agreement if fire records had an average comparison among decades; (ii) to ensure data interval of 9 years. In the data analysed, the interval consistency; (iii) to minimise analytical problems with the averaged 15 years, but clearly there must be some tradeoff bookending of census data with precision windows; (iv) between window size and accuracy. We analysed and to enforce some level of 'blind sampling' in the grasstree discussed this somewhat intractable issue in Miller et al. survey (while evidence of recent fires is easily observable (2007) and settled on a 5 year window (i.e. ± 2 years). in the field and it is impossible to prevent practitioners Ward failed to acknowledge this part of our study in from making their own conclusions about fire history his re-analysis, claiming that his asymmetric, 4 year from field observation, we thought that excluding a few analysis window is 'wider and more reasonable' (Ward years would help to remove the most obvious part of the 2009 p. 262). Ward also failed to provide numerical visible fire history); and (v) as we did not feel that results from his 4 year window re-analysis. However, comparison of grasstree records from the year of our results clearly show that when a 5 year window is observation and its preceding year with known fire allowed, 54% of the positive fire records in the grasstree histories was an adequate test when we were interested record are false (they do not match fires in the satellite in interpretation in the deeper past. If 12 'omitted' record), and when a 3 year window is used, 63% are records are incorporated as Ward described them, and false (Miller et al. 2007). we allow a 5 year precision window, the overall rate of false positives in the grasstree record is 45%, i.e. still Omission of data poor. A third justification given by Ward for revisiting the analysis is a suggestion that we did not use all of the Statistics of agreement data available: that we omitted two clusters of five Ward (2009) criticised our use of the simple and widely grasstrees from a total sample of 105 grasstrees, together employed Kappa statistic to assess the degree of with several other specific fire records. The rejection of similarity between the two records. The critical paper one grasstree cluster, which Ward (2009) called no. 8 (we cited by Ward (Allouche et al. 2006) recommended use of gave it no number), was described in our paper. We the true skill statistic (TSS) as an appropriate alternative rejected this cluster as it lay too close to one fire test: it alleviates the problem of varying record boundary to satisfactorily determine its satellite fire Prevalence. In our case Prevalence is not an issue and record, and its omission is the reason for the subsequent TSS (0.105) is almost identical to Kappa (0.104). These differences in number schemes between the two papers. closely related indices are interpreted in the same way, Concerning the second 'omitted' cluster. Ward correctly and both range from 0 (indicating no agreement) to 1 identified an error in our figure 2 (Miller el al. 2007) that (complete agreement). Neither assesses whether datasets showed 95 and not 100 grasstree samples as we had differ significantly, but merely quantify the relative intended.The five missing plants (his group 18, our degree to which they can be said to agree. Recognising group 17) had no fire signal in either record and the the limitations of the tests, we stated that 'In broad terms, blank space that should have represented these a value for the Kappa statistic below 0.2 indicates poor individuals in the chart was mistakenly lost in formatting agreement and a Kappa above 0.8 indicates very good (but is shown in Figure 2 herein). Its omission from agreement (Landis & Koch 1977)' (Miller el al. 2007 p. calculations would alter rates of true and false negatives 911). We presented Kappa as just one of a suite of tests, (as all of its records are true negatives) but not rates of including a bootstrapping test that explicitly identifies true and false positives (as it contains no positive the level of agreement. records). However, the omission was restricted to the offending figure and it did not influence our calculations. While Ward partly justified his re-analysis on the basis of Fire interval our omission of these clusters, he repeated and extended Ward (2009) criticised us for focussing on matching the the same omissions in his analyses. His tables 1 and 2 dates of fires, when fire interval is an ecologically more (Ward 2009) omit both of the clusters described above, important variable. It is certainly true that fire interval is the second of which we did not omit from analysis. ecologically important. However, our approach was to
97 Journal of the Royal Society of Western Australia, 95(2), July 2012
Year T5 CTU O OO 2000 ooooo- e
o 1995 o o o o 1990 ob
LJ o 1985 : o „ o o o 1980 OO o o OO o OO OO OO o 1975 o OO O o °o ° t i i i i i oT i i i i i i i i i i i i i i i i r 1 I I I I I I I I T I I I T 13 14 15 16 17 18 19 20 Satellite record O Grasstree record Grasstree cluster (each of 5 individuals)
Figure 2 A comparison of grasstree and satellite records for 8 (of 20) analysed grasstree clusters, illustrating elements of the data and analysis, (a) The five precision windows employed by Miller el al. (2007) illustrated here for 1990 with (from left): a 1 year window (i.e. 'exact' precision) with a false positive grasstree record (a grasstree fire not matched by a satellite fire within the window); 3 year (i.e. ±1 years) and 5 year windows (±2 years), both true negatives (no grasstree fire in 1990 + no satellite fire in the window); a 7 year window (± 3years) showing a false negative (no grasstree fire but a positive satellite record within the window not otherwise matched by a closer grasstree fire); and a 9 year window (±4 years)—a true positive (a grasstree fire + an otherwise unmatched positive satellite record in the window), (b) The cluster of five grasstrees mistakenly omitted from Miller et al. (2007) figure 2. (c) Clusters of five grasstrees with contrasting ratios of grasstree-.satellite fire incidence: c, with 11 v 5, and c2 with 7 v 20. (d) A pair of matching 8 year inter-fire periods: the first of three in the grasstree record (d,), and the first of five in the satellite record (d2). (e) The only example of spatial aggregation (i.e. among individual grasstrees in a cluster) in positive agreement between the satellite and grasstree records. first verify that the grasstree record actually was and (iii) a test of their 'agreement' in 5-15 subsets of the indicating fire before considering analysis of the intervals data. The longer temporal context shows simply that its records present. If it does not indicate fire, there is no shrubland grasstree stems reveal the same pattern (i.e. value in analysing its interval distribution. shortening intervals into the past) as grasstree stems sampled from forest and woodlands elsewhere in southwestern Australia, but it does not say anything NEW ANALYSES about the validity of the data. Similarly, Ward's analysis of fire intervals also says Ward (2009) presented three new analyses: (i) a longer nothing about the validity of the grasstree data. It infers temporal context for the grasstree sequence; (ii) a from the coincidence in the mean (or median) observed comparison of fire intervals in the two sets of records; inter-fire interval of the two records that these arise from
98 Miller et al: Grasstrees insufficient for fire history
the same process — fire. In Miller et al. (2007) we noted tests show arbitrarily more agreement than chance to be that the total incidence of positive fire records is matched 0.515 or p = 0.00003. While this value seems extremely in both grasstree (n = 202) and satellite (n = 200) records. convincing, it only shows that the records agree to some Because both records are derived from the same sample of undetermined (but potentially very small) extent. It is individuals and years (N = 100 x 30), their mean between- created by exploiting the very spatial non-independence fire interval is also similar, and it would be surprising if of the grasstree sample that purportedly justified its re¬ their median interval differed. We previously concluded analysis. that 'the close overall agreement across all stems and the Literally interpreted. Ward's concern for spatial whole period may be coincidental' (Miller et al. 2007 p. autocorrelation implies a belief that between-record 914). Comparing the two hundred or so positive fire agreement (not just fire-incidence) might be influenced incidences between the two records shows that they do by the spatial arrangement of the sample. The grasstree not coincide in space - within the same individuals - or dataset does have a unique spatial arrangement: each time in the two records. That the two records have the analysed stem occurs within one of 20 clusters of exactly same fire-incidence rates merely reflects the fact that sites five neighbouring grasstrees, and these clusters are with more satellite fires than grasstree fires are separated by up to several kilometres (Figure 3c). We numerically balanced by sites with the reverse ratio selected this design so that we could examine the (Figure 2 note c). Similarly, there are many between-fire correspondence of fire record within neighbouring intervals of the same size in the two records (Figure 2 note grasstree stems (expected to be quite good, but actually d), but few of them are on the same grasstree and at the not: Figure 2; Miller et al. 2007 figure 2). Because each same time. Ward compared grasstree and satellite fire cluster contains five grasstrees, the sampling is not intervals across all grasstrees, but it makes no sense to biased by unbalanced spatial sampling intensity at that compare aggregated mean or median fire intervals across scale. However, if concern persisted that spatial non¬ grasstrees when the crucial correspondence is within independence of agreement between the two records grasstrees. Unfortunately, nothing can be concluded from among sampled grasstrees within clusters might bias Ward's aggregated analysis of fire interval in relation to results, then analysis of just one individual from each the validity of the grasstree record. cluster would be a reasonable approach to remove the Finally, Ward (2009) addressed the 'agreement' problem. However, it is clearly not sensible to suggest between the datasets in a new analysis. His approach that five samples of 19 grasstrees, each made up of one was to use contingency tables to determine whether there individual sampled from 19 identical locations, with was more or less agreement between the two records (in every sampled individual just metres from one the asymmetric 4 year window) than would be expected individual in each of the other four samples, could be given the marginal totals of these agreement tables independent - especially if used in a test that was overly (Figure 3a), and then to create, and test, a large number sensitive to small differences, and in a system with even of these tables to determine a probability of them all a small amount of aggregation in agreement at the cluster having more agreement than by chance (Figure 3b). As scale (Figure 2). Multiplying the probability of more we have already shown that there is more agreement agreement-than-chance occurring in one sample, by the than chance between the records (Miller et al. 2007), this probability of the same event occurring in a sample with analysis adds nothing new to our collective an identical spatial pattern and repeating this for understanding. But because Ward's presentation of this multiple identical samples merely exploits the spatial analysis is rather complex, and also to more clearly pattern of the sample to inflate the calculated overall justify the preceding statement, we address this analysis level of significance. in some detail. While this test might assess for more agreement-than- First, the absolute difference in the number of by-chance, it does not quantify the extent of the observed and expected records that Ward tabulated to agreement. Its low p-value simply reports that the co¬ assess agreement can be extremely small, and may not be occurrence of fire incidence in the grasstree and satellite significant in any statistical or practical sense. His table 2 records was higher than would be obtained by a random includes examples of 0.2 observations out of a total of 190 scattering of the records in space and time. A statistically (0.1%) as being sufficient to count as more 'agreement' significant departure from randomness does not imply than chance, and the average difference across his 15 that the grasstree record is a faithful record of fire comparisons represents only 2.9% of observations. If the history. The necessary question is 'how much better than question is whether grasstree and satellite record chance is it'? In Miller et al. (2007), we answered this agreement is more or less than chance, then there is only question through a bootstrap analysis, finding that 46% one sample with which to test this outcome—the entire of the 202 positive grasstrees fire records matched a dataset. However this approach gives only an absolute, positive satellite fire record (within a 5-year window) but binary outcome —'more' or 'less' agreement than that on average, 27% of matches would be expected by expected by chance. It does not say how much more, nor chance. However, we did not explicitly follow this with whether it is significant (although both are calculable via the observation that the difference between these values bootstrapping). Instead, to assess significance. Ward should indicate the percentage of grasstree records that created 15 separate agreement tests by splitting the data were both true positives and not attributable to chance. longitudinally into five sets of 19 grasstrees and then That is to say, on average, 19% of grasstree fire records these again into three decades (Figure 3b). The can actually be attributed to fire. The remainder either do justification that Ward provided for this data splitting is not match a satellite fire record in the 5 year window 'to avoid spatial autocorrelation within sites' (Ward 2009 (54%), or do, but would be expected to do so by chance p. 263). He assessed the cumulative probability that all 15 alone (27%).
99 Journal of the Royal Society of Western Australia, 95(2), July 2012
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 18 19 20 Grasstree cluster (each of 5 individuals)
Figure 3 The Ward (2009) approach to assessing agreement among the grasstree and satellite records. (A) Counts of fire presence (+) and absence (-) observed in the grasstree G and satellite S records were tabulated in a two-way table. The difference between these observed (o) and expected (e) counts, calculated from marginal totals (n), indicates whether there is more agreement (a + d) or disagreement (b + c) than expected for the table. (B) To create more tables, the data were divided into three decades (f, g, h), each with 95 individual grasstrees sampled (vertical grey bars; the grasstrees were arranged in 20 clusters of 5 individuals: Ward excluded the 17th cluster as it had neither more nor less agreement). These decadal data were then further divided longitudinally, with one single individual from each cluster per group. This creates 15 tables (here numbered diagonally), where 1 = the 1st grasstree in each cluster in the 1st decade, 2 = the 2nd grasstree in each cluster in the 1st decade ... to 15 = the 5th grasstree in each cluster in the 3rd decade. (C) The spatial distribution of the 20 grasstree clusters around the Eneabba sandplain, overlain on the satellite fire history (dark shades indicate more fires).
DISCUSSION group of these tests considers the number and historical distribution of positive fire records. Ward's chart of A number of tests of the grasstree/satellite record pattern of grasstree fire records with time (Ward 2009 p. 265) comparison are presented here, as well as in Ward (2009) is simply illustrative; it is silent on the data's validity. and in Miller et al. (2007). As the variety and relatedness We previously noted the close correspondence in number of these and the detail to which they are discussed may of fire incidences in the two records, but the incongruent cause some confusion about their overall interpretation, distribution of these records among grasstrees and we here list (italicised below) and summarise their decades shows that it does not indicate a common outcomes for testing and quantifying the relationship process. Ward's comparison of mean and median inter-fire between the grasstree and satellite fire records. The first interval between the two records is effectively a
100 Miller et al.: Grasstrees insufficient for fire history
restatement of the correspondence of overall fire issues such as plant longevity, population structure and incidence, it reflects the same data, and has the same the effect of growing conditions on growth rates. The fire interpretation. signal in grasstree bands is worthy of further study, but its presently interpreted signal is weak and increasingly The second group of tests relates to quantification of unreliable the further we go back in time. 'agreement' between the two records and are all based on two-way tables (also discussed as contingency tables Our major concern is that because 8 out of 10 dark and confusion matrices - that have a specific grasstree stem bands cannot be attributed to fire, it is terminology). These are analysed with exact, and broad, wrong to interpret the whole grasstree record as a precision windows and interpreted via three qualitative reliable fire history. This issue would represent only an indices and two quantitative tests. The qualitative indices academic disagreement if the interpretation was not are the Kappa and TSS indices—which consider the exact extended to support fire management procedures that match of records in the same year, here providing threaten biodiversity by burning at intervals shorter than identical results indicating 'poor agreement'—and the most plant species require to accumulate self-replacement fraction of grasstree fire records that are false positives capacity in seed- and bud-banks (Enright et al. 2011, (together with false negatives, etc). We report the fraction Burrows et al. 2008). As rates of grasstree error increase of false positive grasstree records in precision windows with time into the past and as this error overstates fire of 1-9 years: 54% of the grasstree records do not match a frequency, use of the grasstree fire record from the satellite fire record within a 5 year (±2 years) window. distant past to inform 'appropriate' contemporary fire This index provides a suggestion of the scale of the regimes will elevate conservation risks associated with agreement between the records but does not test its overburning. As it stands, the grasstree banding method likelihood. We previously reported a bootstrapping cannot be considered useful for reconstructing fire approach to test tills likelihood, finding that the records histories, and it should have no role in determining do agree more than by chance, but here emphasise that modern fire-management practices. only 19% of the positive grasstree fire records match positive satellite records (within ±2 years) and cannot be attributed to chance. Ward's test of manipulated agreement ACKNOWLEDGEMENTS probabilities is simply the absolute difference between We thank Journal reviewers Neil Burrows, Michael observed and expected counts of agreement applied to Calver, John Henstridge and Stephen Hopper for their different subsets of the data. We have already shown constructive comments. that there is more agreement between the records than would be expected by chance alone: Ward's test says no more about the data than simply confirming this result. REFERENCES A theme of this discussion is the extent to which Allouche O, Tsoar A & Kadmon R 2006. Assessing the accuracy observed agreement can be considered adequate. Does it of species distribution models: prevalence, kappa and the matter if a TSS of 0.105 is different to a Kappa score of true skill statistic (TSS). Journal of Applied Ecology 43, 1223- 0.104, or that neither provides an actual test of 1232. significance? Both are simply indices of agreement and Burrows N D, Wardell-Johnson G & Ward B 2008. Post-fire in this case exactly where one draws the upper boundary juvenile periods and regeneration syndromes of plants in SW for 'poor agreement' is immaterial: both tests range from Australian forests and implications for fire management. 0 to 1 (no agreement to complete agreement) and no-one Journal of the Royal Socieh/ of Western Australia 91, 163-174. would consider 0.1 as 'very good' or even 'good' Colancelo W I, I.amovi B B, Jones A S, Ward D J & Bombardteri agreement. Similarly with Ward's analysis of agreement S 2002. The anatomy and chemistry of the colour bands of grasstree stems (Xanthorrhoea preissii) as a basis for plant age in two-way tables: should a difference of 0.2 observations and fire history determination, Annals of Botany 89, 605-611. among 190, or an average difference of 2.9% be counted Enright N J, Fontaine J B, Lade J, Miller B P & Westcott V C as an important difference between observed and 2011. Fire interval effects on persistence of resprouter species expected counts? The answer depends on the number of in Mediterranean-type shrublands. Plant Ecologt/ 212, 2071- independent trials. The outcome of Ward's analysis 2083. indicates that the agreement between the satellite and Enright N J, Groeneveld J, Lamont B B & Miller B P 2005a. grasstree records is better than might be expected from a Grasstree reconstructions of pre-European fire regimes and random coin toss, but not how much better. Our results models Banksia (Proteaceae) species co-existence in the show broadly how much better than random the northern sandplain shrublands of southwestern Australia: grasstree record is at predicting fire history, with are they compatible? In: Walshe T V & Williams J E (eds) Variability in fire interval. How important is it and how do we bootstrap analysis suggesting that around 20% of manage for it?, pp. 11-21. Proceedings of a workshop held 26 grasstree fire indications may actually be due to fire. September 2003, Albany Western Australia. University of Western Australia, Crawley. Enright N J, Lamont B B & Miller B P 2005b. Anomalies in CONCLUSIONS grasstree fire history reconstructions for south-western Australian vegetation. Austral Ecology 30, 668-673. We acknowledge that the grasstree record does appear to Enright N J, Lamont B B & Miller B P 2006. Commentary: include information about fire history; that some dark Anomalies in grasstree fire history reconstructions for banding in grasstree stems may represent a fire response. southwestern Australian vegetation: Reply from Enright N. We believe that analysis of the lighter, annual banding of J., Lamont B. B. and Miller B. P. Austral Ecology 31, 792-793. grasstrees may be sufficiently accurate for determining Gill A M 2006. Dating fires from Balga stems: the controversy their annual growth rates (provided the colour bands are over fire histories based on south-western Australian distinct enough) and hence provide insights on important Xanthorrhoea. Report to the Western Australian Department
101 Journal of the Royal Society of Western Australia, 95(2), July 2012
of Environment and Conservation, Perth August 2006
Received 14 February 2012; accepted 6 June 2012
102 Journal of the Royal Society of Western Australia, 95:103-114, 2012
Ichthyoplankton assemblages associated with pink snapper (Pagrus auratus) spawning aggregations in coastal embayments of southwestern Australia
N B BREHENY1, L E BECKLEY1* & C B WAKEFIELD2
1 School of Environmental Science, Murdoch University, South Street, Murdoch, WA 6150 Australia. 2 Western Australian Fisheries and Marine Research Laboratories, Department of Fisheries, PO Box 20, North Beach, WA 6920, Australia. * Corresponding author [email protected]
Pink snapper (Pagrus auratus) form spawning aggregations during the austral spring/summer in three adjacent, sheltered, coastal embayments in southwestern Australia (Cockburn Sound, Warnbro Sound and Owen Anchorage). Larval fishes were sampled in these embayments, as well as in the more exposed adjacent waters of Five Fathom Bank, to ascertain which teleost species utilised these areas for spawning concurrently with P. auratus. Obliquely towed bongo nets (500 pm mesh) were used to collect icthyoplankton in November 2007 over three days during the new moon period when spawning of P. auratus is known to peak. A total of 13 270 larvae from 30 teleost families was collected with an overall mean larval fish concentration of 1.91 in 3 (± s.e. 0.28), Larval fish assemblages were significantly different in each of the three embayments and Five Fathom Bank, reflecting the degree of shelter, water-circulation patterns and associated benthic habitats. The highest larval fish concentrations were recorded in Cockburn Sound (3.69 nr3 ± s.e. 0.05) and the lowest along Five Fathom Bank (0.16 nr3 ± s.e. 0.02). P. auratus larvae were only present in samples from the three embayments. The most abundant larvae were those of the Australian anchovy (Engraulis australis), which occurred predominantly in Cockburn Sound and, to a lesser extent, Owen Anchorage. The most ubiquitous larvae included the longspine dragonet (Callionymus goodladi) and leatherjackets (Monacanthidae). This study provides circumstantial evidence that eggs and larvae of P. auratus and other teleosts are retained in the sheltered waters of Cockburn and Warnbro Sounds during spring/summer.
KEYWORDS: coastal embayments, Cockburn Sound, fish eggs, fish larvae, ichthyoplankton, Pagrus auratus, southwestern Australia.
INTRODUCTION Despite P. auratus being widely distributed in Western Australia, very few spawning and nursery areas have Dispersal of the pelagic eggs and larvae of teleost fishes been located (Wakefield 2006; Jackson 2007; Lenanton et by meso-scale and regional physical processes has been al. 2009). Between Kalbarri (28°S) and Geographe Bay the pervasive paradigm in the literature, although recent (34°S), the only identified locations for spawning studies have led to the reconsideration of the scales of aggregations of large adults (i.e. >70 cm total length, TL) larval transport (Pineda el al. 2007). In particular, the and regular annual occurrences of 0+ juveniles are the complex flows in shallow coastal waters, where three adjacent embayments of Cockburn Sound, Warnbro bathymetry, topography, wind-forcing and other factors Sound and Owen Anchorage, located near Fremantle combine, can result in smaller dispersal kernels and (32°S) (Lenanton 1974; Wakefield 2006; Lenanton et al. retention of larvae (Gawarkiewicz et al. 2007). Likewise, 2009). the spawning behaviour of fishes can contribute to the extent of the dispersal kernels of their offspring, and a Cockburn Sound, Owen Anchorage and Warnbro particularly interesting example is when adults swim to Sound are prominent geographic features along the specific sites and aggregate for spawning. Well-known linear, lower west coast of Australia and, to the west of aggregation spawners include species of the families the embayments, the more exposed waters of Five Serranidae, Sciaenidae, Lutjanidae and Sparidae and, as Fathom Bank are located (Figure 1). These embayments their spawning aggregations are temporally and spatially are generally shallow (~20 m maximum depth), occur consistent (predictable), there are many examples of within a microtidal zone (D'Adamo 1992) and are subject heavy fishing exploitation on such spawning to moderate seas and swell from the south to southwest aggregations (Colin et al. 2003; Sadovy & Cheung 2003). direction (Lcmm et al. 1999). Swell waves generally lose energy when crossing the coastal reefs, banks and sills Pink snapper (Pagrus auratus Forster 1801) is a species and it has been estimated that oceanic swell height is of Sparidae that is widely distributed throughout the reduced to 10% at the shorelines of Owen Anchorage temperate waters of Australia and New Zealand (Paulin and Warnbro Sound and to about 5% in southern 1990). It forms large spawning aggregations in the Cockburn Sound (Department of Environmental sheltered embayments of Port Phillip Bay (Coutin et al. Protection 1996; blegge et al. 1996). The geomorphology 2003), Spencer Gulf and Gulf St Vincent (Fowler & of Cockburn and Warnbro Sounds and the prevailing Jennings 2003) and Shark Bay (Jackson & Cheng 2001). southwesterly winds during spring/summer result in a counterclockwise gyre in each of these two semi-enclosed © Royal Society of Western Australia 2012 embayments (Steedman & Craig 1983; Gersbach 1993).
103 Journal of the Royal Society of Western Australia, 95(2), July 2012
104 Breheny et ale. Ichthyoplankton and P. auratus spawning aggregations
The spawning periodicity of P. auratus aggregations in temperature and salinity were recorded from a bucket of Cockburn Sound is correlated with water temperature surface water using a WTW 315i conductivity meter with and lunar phase (Wakefield 2010) and also coincides with a WTW tetracon®325 conductivity cell. Mean hourly the formation of the seasonal wind-driven gyres within wind speed (km/h) and direction (°) for the 14 days Cockburn and Warnbro Sounds. Particle-tracking model preceding sampling were obtained from the Garden projections based on depth-averaged water movement Island meteorological station upon request from the have suggested a high probability of retention of P. Australian Bureau of Meteorology. auratus eggs and larvae in Cockburn Sound (Doak 2004; Pagrus auratus eggs were identified from the plankton Wakefield 2006). The occurrence of the early life stages of samples according to Wakefield (2006), separated and other fish species utilising the retentive dynamics in stored in 5% buffered formalin. Fish larvae were also these embayments concurrently with P. auratus is separated and stored in 70% ethanol. Fish larvae were unknown. identified to the best taxonomic resolution possible using Ichthyoplankton sampling by the Western Australian a dissecting microscope and available literature (Moser et Department of Fisheries is routinely conducted in al. 1984; Neira et al. 1998; Leis & Carson-Ewart 2000). Cockburn Sound and surrounding areas during the peak Larval concentrations were expressed as the number of reproductive period of P. auratus (new moon in larvae per cubic metre of water and maps of egg and November) to facilitate spawning biomass estimates larval distribution were generated with Surfer 8.0 using the daily-egg-production method.' The fish larvae (Golden Software Inc.) using kriging interpolation. Body collected in these samples were available for this study. lengths (standard length, SL) of all P. auratus larvae and The primary aim was to identify the distributions and a subsample of 100 Engraulis australis larvae were abundances of fish larvae in Cockburn Sound and measured to the nearest 0.1 mm using an eyepiece adjacent waters to ascertain which species use these micrometer. embayments as spawning areas concurrently with P. Analysis of variance (ANOVA) using SPSS 17.0 auratus. It was hypothesised that the larval fish followed by Tukey's post-lmc tests were used to test for assemblages of Cockburn and Warnbro Sounds would differences in mean larval fish concentrations and differ from those of the surrounding exposed coastal numbers of families between locations (fixed factor). A waters due to their protected, semi-enclosed nature and logarithmic transformation was made to larval the retentive dynamics of their seasonal gyres. concentrations to condense the spread of the data and conform to the assumption of homogenous variances METHODS between locations, based on the gradient of the lineal relationship between the logarithms of the standard Ichthyoplankton was sampled during daylight from 10- deviation and mean abundances of species according to 12 November 2007 (coinciding with the new moon Clarke & Warwick (2001). Analyses of larval fish phase). Sampling was conducted at 88 stations arranged assemblages were carried out using the software in a 1 nautical mile grid in Cockburn and Warnbro PRIMER (Version 6.1.4: Clarke & Gorley 2006). Larval Sounds and a 1.5 nautical mile grid in Owen Anchorage fish concentrations were log (x+1) transformed and the and Five Fathom Bank (Figure 1). The sampling period Bray-Curtis distance measure was used to preserve the (relative to the peak temporal reproductive period of P. abundance structure of the dataset when constructing a auratus) and station locations were determined from a similarity matrix. Non-metric multidimensional scaling previous study (Wakefield 2010). The stations were (nMDS ordination) was used to explore the spatial sampled over three consecutive days starting with those variation in the larval fish assemblages. An analysis of in Cockburn Sound, followed by Owen Anchorage and similarities (ANOSIM) was used to test for significant northern Five Fathom Bank, and then Warnbro Sound differences, following which, pair-wise comparisons and southern Five Fathom Bank on the last day. were conducted. ANOSIM is based on rank similarity, Spawning periodicity of P. auratus in these embayments with the significance assessed using the R statistic, which is temporally and spatially specific and highly correlated is comparable among all pair-wise tests. R statistic values with environmental cycles, with diel spawning events range between -1 and 1, with values of 1 indicating that occurring over consecutive nights during the three hours larval fish assemblages at all stations within a location following the high tide (Wakefield 2006, 2010). Thus, are more similar to each other than any stations from eggs and larvae of this species were considered equally different locations. Significant ANOSIM results were available in each embayment during each of the three further explored using similarity percentages (SIMPER: days sampled. Clarke & Gorley 2006) to identify the larval fish taxa that Ichthyoplankton was collected using bongo nets with typified and distinguished locations. Typifying taxa had 60 cm diameter openings and 500 pm mesh. A 'General similarity to standard deviation ratios (Sim/SD) >1 and Oceanics' mechanical flowmeter was attached to the percentage contribution >5% to that location. mouth of one net to establish the volume filtered. Nets Distinguishing taxa were determined by dissimilarity to were towed obliquely for two minutes with the vessel standard deviation ratios (Diss/SD) >1 and percentage travelling at ~2 knots. The warp (length of rope) used for contributions >5% to the locations being compared. each tow was about 2.5 times the depth at each station, The biota and environment matching routine allowing the net to be towed through the water column (BIOENV: Clarke & Gorley 2006) was used to test the from just above the seafloor to the surface (Wakefield degree to which environmental variables influenced 2010). The two plankton nets were washed thoroughly larval fish assemblages. Benthic habitats were ascertained and their contents placed into a single sample jar and from available maps (Department of Environmental preserved with 5% buffered formalin. At each station, Protection 1996) and the dominant habitat type within a
105 Journal of the Royal Society of Western Australia, 95(2), July 2012
radius of 500 m (centred on each station) was assigned to (Appendix 1). Pegasus volitans larvae (Pegasidae) were each station. Draftsman's plots were used to examine specific to Cockburn and Warnbro Sounds (0.013 and whether any variables were correlated and whether any 0.015 larvae nr3, respectively) and Creedia haswelli larvae transformations were needed based on the spread of (Creediidae) were specific to the waters of Five Fathom data. Environmental data were normalised as variables Bank (0.006 larvae m'3: Appendix 1). were in different units and the Euclidean distance There was a significant difference between the measure was used. numbers of families recorded among locations (ANOVA, F = 33.7, p <0.001). Five Fathom Bank was significantly less diverse at the family level than all other locations (p RESULTS values <0.05). Warnbro Sound had significantly richer family level diversity than all other locations (p values Larval fish assemblages <0.001), particularly in the southern waters with up to 16 A total of 13 270 fish larvae (30 families and 53 taxa) was families recorded. Jervoise Bay, in northeastern Cockburn collected with an overall mean concentration of 1.91 m° Sound, also showed a high family richness. (± s.e. 0.28, range 0.01-12.34 nr3). Larval fish Overall, there was significant separation between concentrations were significantly different between larval fish assemblages among locations (ANOSIM global locations (ANOVA, F = 47.6, p <0.001: Figure 2). Larval R = 0.617, p <0.001: Figure 4). Tight clustering of concentrations in Cockburn Sound (3.69 nv3 ± s.e. 0.50) assemblage data from stations in Cockburn and Warnbro were significantly higher than in all other locations (all p Sounds in the nMDS plot indicated low variability values <0.001), and concentrations in Warnbro Sound between stations, whereas data from stations at the more (1.71 nr3 ± s.e. 0.44) were significantly higher than in exposed sites of Owen Anchorage and Five Fathom Bank Owen Anchorage (0.92 m3 ± s.e. 0.29) and Five Fathom showed greater separation and thus higher variability in Bank (0.16 nr3 ± s.e. 0.02) (p <0.01 for both). Much of the larval assemblages (Figure 4). Pair-wise comparisons variation between locations was attributed to markedly showed assemblages from all locations to differ higher abundances of larvae of the anchovy £. australis in significantly from one another (p values <0.01: Table 1). Cockburn Sound (3.05 nr3 ± s.e. 0.61) and, to a lesser Cockburn Sound and Five Fathom Bank had the most extent, Owen Anchorage (0.34 nr3 ± s.e. 0.21) than in dissimilar larval fish assemblages (R = 0.728) and Five other locations (Figure 2). Fathom Bank and Owen Anchorage the most similar (R = All locations displayed distinct larval fish 0.380). Typifying taxa were identified by SIMPER with compositions with a suite of families often characterising Cockburn Sound characterised by £. australis (yolk sac each (Figure 3). Owen Anchorage was dominated by and larvae) and C. goodladi, and Warnbro Sound larval fish of the families Engraulidae (36%) and characterised by C. goodladi, S. sagax, P. auratus and Pomacentridae (31%) and Five Fathom Bank was Pseudocaranx sp. (Table 2). With the exception of dominated by Percophidae (23%) and Monacanthidae Pseudocaranx sp., these typifying species essentially (19%). The Cockburn Sound assemblage was clearly distinguished the larval assemblages of both sounds dominated by Engraulidae (83%, including yolk sac stage from those of the other locations (Table 2). Owen larvae), whereas Warnbro Sound was dominated by Anchorage was distinguished from Five Fathom Bank Clupeidae (30%), Callionymidae (16%) and Sparidae and Warnbro Sound by the higher prevalence of (14%). Larvae of Parapercis haackei (Pinguipedidae) were Monacanthidae (Table 2). recorded in considerably higher concentrations in Cockburn Sound (0.049 m'3) than all other locations Abundance and distribution of taxa Eggs of P. auratus occurred in three highly concentrated 4.5 nodes in each of the deeper central waters of the adjacent 38 embayments of Warnbro Sound (40-44 nv'), Cockburn 4.0 Sound (12-16 nv3) and Owen Anchorage (4-8 nr3. Figure __ 3.5 5). Larvae of P. auratus occurred in relatively high CO concentrations in Warnbro Sound (0.70-0.77 nr3) and 3.0 moderate concentrations (0.07-0.21 nr3) in northern Cl) 2.5 CC Cockburn Sound, but were markedly less abundant in £ _cd 2.0 JZ 1.5 C/) Table 1 ANOSIM pairwise comparisons between Ll 1.0 locations showing R statistics and p values 0.5 Groups R statistic p value 0 Cockburn Sound vs Five Fathom Bank 0.728 0.001** Owen Five Fathom Cockburn Warnbro Anchorage Bank Sound Sound Cockburn Sound vs Warnbro Sound 0.683 o.ooi** Five Fathom Bank vs Warnbro Sound 0.548 0.001** Figure 2 Mean concentration of fish larvae (nr3 ± s.e.) for Cockburn Sound vs Owen Anchorage 0.510 0.003** each location off southwestern Australia. Engraulis Owen Anchorage vs Warnbro Sound 0.404 0.001** australis larvae are indicated as white bars and all other Five Fathom Bank vs Owen Anchorage 0.380 0.001** taxa by grey bars. Number of stations sampled per location shown above the bars. ** significant at 1% sig. level.
106 Breheny et air Ichthyoplankton and P. auratus spawning aggregations
Table 2 Larval fish taxa identified by SIMPER as those that typified larval assemblages within each location or distinguished larval assemblages between locations.
Owen Five Fathom Cockburn Warnbro Anchorage Bank Sound Sound
Owen Anchorage - - - - (OA)
Five Fathom Bank Monacanthidae spp. (OA) - - - (FF)
Cockburn Sound £. australis(CS) E. australis The location where distinguishing taxa from each pairwise comparison were most abundant is indicated by superscripts, ys, yolk sac larvae. Owen Anchorage (Figure 5). The majority of P. auratus Monacanthidae spp. The highest concentrations of C. larvae were in the preflexion stages of development goodladi larvae were recorded at Mangles Bay in southern (including some still with yolk sac), with a mean body Cockburn Sound (1.20-1.35 nr3), the deeper central length of 3.2 mm (± s.e. 0.06, n = 240: Figure 6). waters of Cockburn Sound (0.45-0.90 nr3) and southern Larvae of E. australis were almost exclusively collected Warnbro Sound (0.45-0.60 nr3: Figure 5). The highest in the deeper central waters of Cockburn Sound (10.5- concentrations of Monacanthidae larvae were recorded 12.0 nr3: Figure 5). The majority of £. australis larvae were in the east of Cockburn Sound at Jervoise Bay (0.10-0.45 in yolk sac or very early preflexion stages of development nr3), Mangles Bay (0.1-0.3 m'3) and in northwestern with a mean body length of 3.2 mm (± s.e. 0.09, n = 100: Owen Anchorage (0.1-0.3 nr3: Figure 5). Figure 6). Larvae of S. sagax were exclusive to Warnbro Larvae of other commercially and recreationally Sound, with relatively high concentrations (0.8-1.6 nr3) important taxa collected included those of Pseudocaranx occurring in the southeast of this embayment (Figure 5). sp. and Sillaginidae spp. (Figure 5). Larvae of The most ubiquitous larval fish taxa collected across Pseudocaranx sp. occurred in the three adjacent all four locations were Callionymus goodladi and embayments, with higher abundances (0.04-0.07 nr3) Owen Anchorage Five Fathom Bank ^Engraulidae Monacanthidai Callionymidae .Callionymidae Sillaginidae Other 'Clupeidae Creedidae Monacanthidae Percophidae Labridae Sparidae Pomacentridae Tripterygiidae Unidentified Cockburn Sound Warnbro Sound Sparidae Engraulidae (ys) Unidentified Figure 3 Percentage contributions (nr3) by dominant teleost families to the ichthyoplankton at each location, Clupeidae (ys, yolk sac stage larvae). Engraulidae 107 Journal of the Royal Society of Western Australia, 95(2), July 2012 2D Stress: 0.17 A ■a • • ■ ■ • As.® • ► • * o °o *• mO ■ » Figure 4 Non-metric • • o Region multidimensional scaling (nMDS) plot showing the distinction of ■ Owen Anchorage larval fish assemblages for each • Five Fathom Bank station from each of the four A Cockburn Sound locations sampled. O Warnbro Sound occurring in Warnbro Sound (Figure 5). Larvae of spawn at the same time as P. auratus. The larval fish Sillaginidae occurred in higher concentrations in assemblages recorded in the four areas of this study were southeastern Warnbro Sound (0.06-0.11 nr3) and in low markedly different within a relatively small geographic to moderate concentrations (0.02-0.08 nr3) along Five area (-500 km2), with concentrations generally higher in Fathom Bank and Cockburn Sound (Figure 5). the protected marine embayments of Cockburn and Larval fishes hatched from demersal versus pelagic Warnbro Sounds. The larval fish concentrations were eggs showed different spatial distributions across the greater than those recorded in previous studies from study area. Larvae originating from demersal eggs were southwestern Australia (Gaughan cl al. 1990; Kendrick most strongly clustered around stations in northwestern 1993; Young & Potter 2003; Muhling el al. 2008a) and in Owen Anchorage (<2.2 larvae nv3), whilst larvae similar marine embayments in southeastern Australia, originating from pelagic eggs occurred predominantly in e.g. Port Phillip Bay (Jenkins 1986; Neira & Sporcic 2002). the deeper central waters of Cockburn Sound (<12.5 The differences in both taxonomic diversity and larvae nr3). The spatial distribution of larvae hatched abundances of fish larvae between the four locations over from pelagic eggs was largely dominated by the single such a small geographic range could be attributed to dominant species, E. australis. Excluding this species, differences in shelter from wave energy, benthic habitat other larval fishes hatched from pelagic eggs showed and anthropogenic influences. However, the seasonally highest concentrations in southeastern Warnbro Sound occurring hydrodynamic gyres in Cockburn and (2.80-3.15 larvae nr3), and moderate concentrations in Warnbro Sounds, which are driven by local Mangles Bay (1.40-1.75 larvae m'3) and the deeper central geomorphology and the prevailing southwesterly winds waters of Cockburn Sound (0.70-1.05 larvae nr3). during the austral spring/summer (Steedman & Craig 1983; Gersbach 1993), could result in high levels of Environmental and physical parameters retention in Cockburn and Warnbro Sounds during this period. Retention of fish eggs and larvae by the gyres has Daily wind conditions during the 14 days leading up to been predicted from particle tracking in hydrodynamic sampling displayed a characteristic late spring pattern, models based on typical weather conditions (Apai 2001) with winds predominantly from the east to southeast (p and applied to seasonal patterns of P. auratus egg = 18.9 km/h) in the morning before shifting south to abundance (Doak 2004). The spatial distribution of P. southwest and increasing in velocity (p = 25.8 km/h) in auratus eggs and larvae from this study provides the afternoon. Depth of stations ranged from 3 to 28 m, circumstantial evidence of the retention of their early sea-surface temperatures averaged 20.0°C (± s.e. 0.09) pelagic life stages in Cockburn and Warnbro Sounds. and surface salinity averaged 35.6 %0 (± s.e. 0.02). BIOENV was used to relate environmental and physical Ichthyoplankton sampling was purposely conducted parameters to larval assemblages and revealed significant during the peak reproductive period for P, auratus and (p <0.001) but weak positive correlations with benthic the highly concentrated nodes of eggs in each of the habitat, depth, salinity and sea-surface temperature. The embayments in 2007 concurs with the findings of a most predictive environmental variable was benthic previous study (Wakefield 2006, 2010). However, despite habitat (p = 0.345). all three embayments having high concentrations of P. auratus eggs, similar concentration nodes of larvae were only apparent in Cockburn and Warnbro Sounds. The DISCUSSION hydrodynamic retention of P. auratus progeny in Cockburn and Warnbro Sounds is similar to that This study provides circumstantial evidence that, to reported for eggs and larvae of P. auratus from discrete varying degrees, many temperate Australian fish species locations in Shark Bay, Western Australia (Nahas et al. 108 Breheny et al: Ichthyoplankton and P. auratus spawning aggregations co CO CO CO in in in in o o CM CM CM CM CO CO CO CO LU o -'3' o LU LO CO ino LU o ■'3- oin UJ in CO o in LU o oin LU in co oin LU o o in LU Figure 5 Spatial distributions of Pagrus auratus eggs, and larvae (nr 3 ) those species that typified assemblages in each location. 109 Journal of the Royal Society of Western Australia, 95(2), July 2012 P. auratus australis appears to compete unsuccessfully with S. sagax for optimal upwelling zones in shelf waters (Ward et al. 2001; Dimmlich et al. 2004). Dimmlich et al. (2004) identified the northern waters of Spencer Gulf and Gulf St Vincent, where egg and larval retention is maximised, to be the centres of anchovy spawning activity in South Australia. Engraulis australis commonly spawns in protected nearshore areas across temperate Australia (Neira & Sporcic 2002), which appears to be a common trait among this genus globally (Yoklavich et al. 1992; Allain et al. 2003; Takasuka et al. 2003). Tire distributions of some of the larval fishes in this study appeared ubiquitous across all four locations with the most abundant of these being larvae of Callionymus goodladi, Monacanthidae and Sillaginidae. Larvae of C. goodladi have been recorded consistently in other larval fish studies at coastal and estuarine locations off southwestern Australia (Gaughan et al. 1990; Jonker 1993; Kendrick 1993). The broad spatial distribution of larvae of Monacanthidae probably reflects the diverse habitats occupied by the large number of species (>25) that occur Standard length (mm) in temperate Western Australian waters (Hutchins & Swainston 1999). Larvae of Sillaginidae were one of the Figure 6 Length-frequency distributions of Pagrus few taxa to occur in moderate concentrations at stations auratus and Engraulis australis larvae from the study area. along Five Fathom Bank and in Cockburn and Warnbro Sounds. King George whiting (SUlaginodes punctatus) is a species known to use Mangles Bay in southern Cockburn 2003). In South Australia and Victoria, such spawning Sound as a nursery area (Hyndes et al. 1998), but as this aggregation locations have been identified as important species spawns from winter to early spring in recruitment sources, responsible for large or entire southwestern Australia, the preflexion whiting larvae contributions of annual cohorts to adult stocks in coastal identified in this study were not likely to be S. punctatus. waters (Fowler et al. 2005; Hamer el al. 2005). It is likely Myomere counts indicated that they were probably from that Cockburn and Warnbro Sounds also represent the genus Sillago (Neira et al. 1998). Species within this important recruitment locations for adjacent adult stocks genus spawn from late spring to early autumn with only of P. auratus, considering the egg and larvae retention Sillago schomburgkii spawning in nearshore areas, indicated in this study and their significance as juvenile suggesting that the larvae found in this study were likely habitat areas (Lenanton 1974; Wakefield et al. 2011). to belong to this species (Hyndes & Potter 1997). The larvae of some other fish species (e.g. Pscudocaranx Northwestern Owen Anchorage had moderate larval sp. and Pnrapercis haackei) were also exclusive to the three fish concentrations with a high proportion of these larvae adjacent embayments. The wavy grubfish, P. haackei, originating from demersal eggs. The variety of benthic commonly occurs in sand and rubble habitats in habitats including seagrass meadows and subtidal reefs relatively protected waters (Hutchins & Swainston 1999) adjacent to Owen Anchorage may provide appropriate and the trevallies of the genus Pseudocaranx are usually substrates for attachment of their demersal eggs associated with sand and seagrass habitats in coastal (Department of Environmental Protection 1996). areas including estuaries (Hutchins & Swainston 1999). Demersal eggs would be expected to limit tire offshore dispersal of early life history stages and Muhling et al. Larval distributions of two pelagic fish species also (2008a) found Marmion lagoon (40 km north of Owen displayed affinities with the sheltered waters of the Anchorage) to be dominated by larvae of small, inshore marine embayments. Early stage larvae of £. australis reef- and seagrass-inhabiting fishes such as the exhibited the highest concentrations of all taxa collected Tripterygiidae, Clinidae, and Gobiidae and and were most prominent in Cockburn Sound and, to a Monacanthidae that produce demersal eggs or are live- lesser extent, Owen Anchorage. The reproductive period bearers. Larvae of Pomacentridae sp. 1, which were of this pelagic species peaks during the austral summer specific to Owen Anchorage were most likely to belong (January to March) throughout its temperate and to the genus Parma. Kendrick (1993) noted that larvae of subtropical distribution (Hoedt & Dimmlich 1995; Ward this taxon exhibited a discrete distribution pattern et al. 2003; Dimmlich et al. 2004). The spatial distribution around sites off Rottnest Island and Carnac Island of larvae of another clupeiform species, S. sagax, was (southwestern Owen Anchorage). Pomacentridae restricted to Warnbro Sound, and thus segregated from generally produce demersal eggs and commonly the £. australis larvae. Larvae of S. sagax are found year- male will exhibit brooding care for up to a week round in shelf waters off southwestern Australia and the (Gladstone 2007; Curley & Gillings 2009). influence of the Leeuwin Current with respect to growth rates and larval abundance of this species has been At the deeper stations, west of Five Fathom Bank, investigated (Muhling et al. 2008b). In South Australia, larvae of the slender sandburrower, Creedia haswelli were the spawning locations of these two clupeiform species found. Similarly, Muhling et al. (2008a) found C. hasivelli are discreet between shelf and gulf waters, where £. to only make significant contributions to larval 110 Broheny et ai: Ichthyoplankton and P. nurntus spawning aggregations assemblages in southwestern Australian shelf waters at Curley B G & Giluncs M R 2009. Population connectivity in the stations in 40 m and 100 m depths rather than their temperate damselfish Parma microlepis: analyses of genetic structure across multiple spatial scales. Marine Biology 156, shallower nearshore station (18 m). Larvae of the broad 381-393. duckbill, Enigmapercis reducta (Percophidae), also D’Adamo N 1992. Hydrodynamics and recommendations for occurred in greatest abundances at stations along the further studies in Cockburn Sound and adjacent waters. deeper western margin of Five Fathom Bank. Environmental Protection Authority Western Australia, Technical The inherent vulnerability of a species increases Series 41. markedly when a particular life history stage is restricted Department of Environmental Protection 1996. Southern to a specific area or limited habitat which may be affected Metropolitan Coastal Waters Study (1991-1994). Department of Environmental Protection Western Australia 17. by anthropogenic disturbance. In the marine environment, disturbance is commonly encountered Dimmuch W F, Breed W G, Geddes M & Ward T M 2004. Relative importance of gulf and shelf waters for spawning and within protected bays as their sheltered nature benefits recruitment of Australian anchovy, Engraulis australis, in many anthropogenic activities (e.g. ports, industry, South Australia. Fisheries Oceanography 13, 310-323. aquaculture, fishing, mining, naval defence operations). Doak C 2004. Hydrodynamic modeling of snapper (Pagrus In southwestern Australia, P. auratus forms spawning auratus) eggs and larvae in Cockburn Sound. BSc (Hons) aggregations in the few sheltered marine embayments thesis. University of Western Australia, Perth (unpubl.). along the essentially linear southwest coast during a Fowler A J, Gillanders B M & Hai l K C 2005. Relationships short annual spawning period (Wakefield 2006). The between elemental concentration and age from otoliths of importance of these embayments in contributing to adult snapper (Pagrus auratus, Sparidae): implications for cohorts of exploited adult stocks of P. auratus warrants movement and stock structure. Marine and Freshwater further investigation. Although the biology is poorly Research 56, 661-676. known for most of the other teleost species with larvae Fowler A J & Jennings P R 2003. Dynamics in 0+ recruitment recorded in this study, those with larval distributions and early life history for snapper (Pagrus auratus, Sparidae) in South Australia. Marine and Freshwater Research 54, 941- similar to P. auratus could also have a higher inherent 956. vulnerability, given the paucity of nearshore marine Gawarkiewicz G, Monismith S & Largier J 2007. Observing larval embayments along the coast of southwestern Australia transport processes affecting population connectivity. and increasing anthropogenic activities in these areas. Oceanography 20, 40-53. Gaughan D J, Neira F J, Beckley L E & Potter I C 1990. Composition, seasonality and distribution of ACKNOWLEDGEMENTS ichthyoplankton in the lower Swan estuary, south-western Australia. Australian journal of Marine and Freshwater Research We would like to thank the Western Australian 41, 529-543. Department of Fisheries, Fremantle Ports and Murdoch Gbrsbach G 1993. Physical oceanography of Warnbro Sound. University for funding the study. Further, thanks to the BSc (Hons) thesis, University of Western Australia, Perth Department of Fisheries for conducting the plankton (unpubl.). sampling and, specifically, jan Richards for separating Gladstone W 2007. Selection of a spawning aggregation site by the eggs and larvae from the plankton samples. Thanks Chromis hypsilepis (Pisces: Pomacentridae): habitat structure, transport potential, and food availability. Marine Ecology to Anthony Miskiewicz for his assistance in larval fish Progress Series 351, 235-247. identification, Jeffrey Leis for the verification of several Hamer P A, Jenkins G P & Gillanders B M 2005. Chemical tags identifications, and Fiona Valesini and James Tweedley in otoliths indicate the importance of local and distant for providing statistical assistance. We also thank the settlement areas to populations of a temperate sparid, Pagrus Journal reviewers for their comments. auratus. Canadian journal of Fisheries and Aquatic Sciences 62, 623-630. Hegge B, Eliot I & Hsu J 1996. Sheltered sandy beaches of southwestern Australia, journal of Coastal Research 12, 748- REFERENCES 760. Allain G, Petitgas P, Grellier P & Lazure P 2003. The selection Hoedt F E & Dimmuch W F 1995. Egg and larval abundance and process from larval to juvenile stages of anchovy (Engraulis spawning localities of the anchovy (Engraulis australis) and encrasicolus) in the Bay of Biscay investigated by Lagrangian pilchard (Sardinops neopilchardus) near Phillip Island, simulations and comparative otolith growth. Fisheries Victoria. Marine and Freshwater Research 46, 735-743. Oceanography 12, 407-418. Hutchins J B & Swainston R 1999. Sea fishes of southern Australia. Apai D 2001. Oil spill fate and trajectory modelling in Cockburn Swainston Publishing, Smithfield. Sound. BSc (Hons) thesis. University of Western Australia, Hyndes G A, Pi.atell M E, Potter 1 C & Lenanton R C J 1998. Perth (unpubl.). Age composition, growth, reproductive biology, and Clarke K R & Gokley R N 2006. PRIMER v6: User Manual/ recruitment of king george whiting, Sillaginodes punctata, in Tutorial. Natural Environment Research Council, Plymouth. coastal waters of southwestern Australia. Fishery Bulletin 96, Clarke K R & Warwick R M 2001. Change in marine communities: 258-270. an approach to statistical analysis and interpretation (2nd Hyndes G A & Potter I C 1997. Age, growth and reproduction edition). Natural Environment Research Council, Plymouth. of Sitlago schomburgkii in south-western Australian, nearshore Colin P L, Sadovy Y J & Domeier M L 2003. Manual for the waters and comparisons of life history styles of a suite of study and conservation of reef fish spawning aggregations. Sitlago species. Environmental Biology of Fishes 49, 435-447. Society for the Conservation of Reef Fish Aggregations Special Jackson G 2007. Fisheries biology and management of pink Publication 1. snapper Pagrus auratus, in the inner gulfs of Shark Bay, Coutin P, Cashmore S & Sivakumuran K P 2003. Assessment of the Western Australia. PhD thesis, Murdoch University, Perth snapper fishery in Victoria. FRDC (Project No. 971127). Marine (unpubl.). and Freshwater Resources Institute, Melbourne. Jackson G & Cheng Y W 2001. Parameter estimation with egg 111 Journal of the Royal Society of Western Australia, 95(2), July 2012 production surveys to estimate snapper, Pagrus auratus, Neira F J & Sporcic M A 2002. Use of ichthyoplankton ecology biomass in Shark Bay, Western Australia. Journal of to evaluate ecosystem changes: a case study in a large, semi- Agricultural, Biological, and Environmental Statistics 6, 243-257. enclosed Australian bay. Marine and Freshwater Research 53 Jenkins G P 1986. Composition, seasonality and distribution of 339-354. ichthyoplankton in Port Phillip Bay, Victoria. Australian Paulin C D 1990. Pagrus auratus, a new combination for the Journal of Marine and Freshwater Research 37, 507-520. species known as 'snapper' in Australasian waters (Pisces: Jonker L J 1993. Comparison of the larval fish assemblages in Sparidae). New Zealand Journal of Marine and Freshwater healthy and degraded seagrass meadows in Cockburn Research 24, 259-265. Sound, Western Australia. BSc (Hons) thesis, Murdoch Pineda J, Hare J A & Sponaugle S 2007. Larval transport and University, Perth (unpubl.). dispersal in the coastal ocean and consequences for Kendrick A J 1993, Assemblages of larval fishes in coastal population connectivity. Oceanography 20, 22-39. waters of south-western Australia. BSc (Hons) thesis, Sadovy Y & Cheung W L 2003. Near extinction of a highly Murdoch University, Perth (unpubl.). fecund fish: the one that nearly got away. Fish and Fisheries 4 Leis J M & Carson-Ewakt B M 2000. The larvae of Indo-Pacific 86-99. coastal fishes: an identification guide to marine fish larvae. Brill, Steedman R K & Craig P D 1983. Wind-driven circulation of Leiden. Cockburn Sound. Australian Journal of Marine and Freshwater Lemm A J, Hegge B J & Masselink G 1999. Offshore wave climate, Research 34, 187-212. Perth (Western Australia), 1994-96. Marine and Freshwater Takasuka A, Aoki I & Mitani I 2003. Evidence of growth- Research 50, 90-102. selective predation on larval Japanese anchovy Engraulis Lenanton R C J 1974. The abundance and size composition of japonicus in Sagami Bay. Marine Ecology Progress Series 252 trawled juvenile snapper Chrysophrys unicolor (Quoy and 223-238. Gaimard) from Cockburn Sound, Western Australia. Wakefield C B 2006. Latitudinal and temporal comparisons of Australian Journal of Marine and Freshwater Research 25, 281- the reproductive biology and growth of snapper, Pagrus 285. auratus (Sparidae), in Western Australia. PhD thesis, Lenanton R C J, St John J, Keay I, Wakefield C B, Jackson G, Murdoch University, Perth (unpubl.). Wise B & Galghan D J 2009. Spatial scales of exploitation Wakefield C B 2010. Annual, lunar and diel reproductive among populations of demersal scalefish: implications for periodicity of a spawning aggregation of snapper Pagrus management. Part 2: Stock structure and biology of two auratus (Sparidae) in a marine embayment on the lower west indicator species, West Australian dhufish (Glaucosoma coast of Australia. Journal of Fish Biology 77, 1359-1378. hebraicum) and pink snapper (Pagrus auratus), in the West Wakefield C B, Fairclouch D V, Lenanton R C J & Potter I C Coast Bioregion. Final FRDC Report - Project 2003/052. 2011. Spawning and nursery habitat partitioning and Department of Fisheries, Western Australia, Fisheries Research movement patterns of Pagrus auratus (Sparidae) on the lower Report 174. west coast of Australia. Fisheries Research 109, 243-251. Moser H G, Richards W J, Cohen D M, Fahay M P, Kendall A Ward T M, Hoedt F E, McLeay L, Dimmuch W F, Jackson G, W & Richardson S L 1984. Ontogeny and systematics of Rogers P J & Jones K 2001. Have recent mass mortalities of fishes. American Society of Ichthyologists and Herpetologists the sardine Sardinops sagax facilitated an expansion in the Special Publication 1. distribution and abundance of the anchovy Engraulis australis Muhling B A, Beckley L E, Gauchan D J, Jones C M, Miskiewicz in South Australia? Marine Ecology Progress Series 220, 241- A G & Hesp S A 2008b. Spawning, larval abundance and 251. growth rate of Sardinops sagax off south-western Australia: Ward T M, Staunton-Smith J, Hoyle S & Halliday I A 2003. influence of an anomalous eastern boundary current. Marine- Spawning patterns of four species of predominantly Ecology Progress Series 364,157-167. temperate pelagic fishes in the sub-tropical waters of Muhunc B A, Beckley L E, Koslow J A & Pearce A F 2008a. southern Queensland. Estuarine, Coastal and Shelf Science 56, Larval fish assemblages and water mass structure off the 1-16. oligotrophic south-western Australian coast. Fisheries Yoklavich M M, Stevenson M & Cailliet G M 1992. Seasonal Oceanography 17, 16—31. and spatial patterns of ichthyoplankton abundance in Nahas E L, Jackson G, Pattiaratchi C B & Ivey G N 2003. Elkhorn Slough, California. Estuarine, Coastal and Shelf Science Hydrodynamic modelling of snapper Pagrus auratus egg and 34, 109-126. larval dispersal in Shark Bay, Western Australia: Young G C & Potter I C 2003. Do the characteristics of the reproductive isolation at a fine spatial scale. Marine Ecology ichthyoplankton in an artificial and a natural entrance Progress Series 265, 213-226. channel of a large estuary differ? Estuarine, Coastal and Shelf Neira F J, Miskiewicz A G & Trnski T 1998. Larvae of temperate Science 56, 765-779. Australian fishes: laboratory guide for lanrnl fish identification. University of Western Australia Press, Perth. Received 26 April 2012; accepted 26 June 2012 112 Breheny et al.\ Ichthyoplankton and P. auratus spawning aggregations Appendix 1 Total counts and mean larval fish concentration (nr3) for each taxon at each location. Larval fish taxa Owen Anchorage Five Fathom Bank Cockburn Sound Warnbro Sound Overall Total Mean Total Mean Total Mean Total Mean Total Mean Clupeidae Sardinops sagax 1 0.0014 11 0.0041 0 0 245 0.4276 257 0.0405 Spratelloides robustus 9 0.0123 5 0.0018 15 0.0052 0 0 29 0.0043 Hyperlophus vittatus 0 0 1 0.0004 0 0 0 0 1 0.0001 Unidentified 0 0 0 0 0 0 44 0.0837 44 0.0076 Unidentified (ys) 0 0 0 0 0 0 20 0.0352 20 0.0032 Engraulidae Engraulis australis 240 0.3349 5 0.0021 6587 2.1594 1 0.0016 6833 0.9714 Engraulis australis (ys) 7 0.0096 6 0.0024 2631 0.8938 0 0 2644 0.388 Synodontidae Synodontid sp. 0 0 0 0 0 0 1 0.002 1 0.0002 Syngnathidae1’ Stimalopora nigra 5 0.0064 1 0.0004 3 0.0009 5 0.0057 14 0.0018 Syngnathid sp. 0 0 0 0 2 0.0008 0 0 2 0.0003 Pegasidae Pegasus volitans 1 0.001 1 0.0004 40 0.0131 8 0.0148 50 0.0073 Scorpaenidae Gymnapistes marmoratus 3 0.004 1 0.0004 1 0.0003 0 0 5 0.0007 Scorpaenid sp. 0 0 2 0.0008 0 0 0 0 2 0.0003 Triglidae Lepidotrigla papilio 1 0.0015 3 0.0011 8 0.0024 10 0.0176 22 0.0032 Lepidotrigla spp. 0 0 0 0 4 0.0013 2 0.0028 6 0.0008 Unidentified 0 0 2 0.0007 0 0 1 0.0019 3 0.0004 Platycephalidae Platycephalus speculator 0 0 1 0.0004 2 0.0006 1 0.002 4 0.0006 Platycephalid sp.l 0 0 0 0 1 0.0004 0 0 1 0.0002 Platycephalid sp.2 0 0 0 0 1 0.0004 0 0 1 0.0002 Unidentified 0 0 1 0.0004 0 0 2 0.0033 3 0.0004 Serranidae Epinephelinae sp. 0 0 1 0.0003 0 0 0 0 1 0.0001 Unidentified 1 0.0013 0 0 0 0 0 0 1 0.0001 Plesiopidae3 Paraplesicips sp. 0 0 0 0 1 0.0003 0 0 1 0.0001 Terapontidae Pelates Odolinealus 0 0 0 0 0 0 6 0.0103 6 0.0009 Pelates sexlineatus 0 0 1 0.0004 1 0.0004 8 0.0129 10 0.0015 Pelates sp. 0 0 0 0 1 0.0003 1 0.0019 2 0.0003 Unidentified 0 0 0 0 1 0.0004 3 0.0056 4 0.0007 Sillaginidae Sillago bassensis 0 0 5 0.002 0 0 0 0 5 0.0007 Sillago spp. 0 0 7 0.0028 33 0.0114 16 0.0286 56 0.0085 Unidentified 0 0 1 0.0004 0 0 1 0.002 2 0.0003 Carangidae Pseudocaranx sp. 2 0.0028 1 0.0004 13 0.0041 20 0.0324 36 0.0052 Sparidae Pagrus auratus 10 0.0141 1 0.0004 124 0.0405 105 0.1811 240 0.0357 Rhabdosargus sarba 0 0 0 0 17 0.0052 0 0 17 0.0023 Unidentified 7 0.0096 2 0.0008 37 0.0118 37 0.0632 83 0.0122 Pomacentridae3 Pomacentrid sp.l 202 0.2844 11 0.0048 0 0 1 0.0011 214 0.0342 Pomacentrid sp.2 0 0 0 0 1 0.0004 0 0 1 0.0002 Mugilidae Mugilid sp. 0 0 2 0.0007 0 0 0 0 2 0.0003 Labridae Labrid sp.l 3 0.0043 8 0.003 0 0 12 0.0168 23 0.0031 Labrid sp.2 0 0 7 0.0025 0 0 17 0.0244 24 0.0031 Labrid sp.3 0 0 0 0 0 0 2 0.0028 2 0.0003 113 Journal of the Royal Society of Western Australia, 95(2), July 2012 Appendix 1 (cont.) Larval fish taxa Owen Anchorage Five Fathom Bank Cockburn Sound Warnbro Sound Overall Total Mean Total Mean Total Mean Total Mean Total Mean Labrid sp.4 0 0 0 0 0 0 3 0.0054 3 0.0005 Unidentified 0 0 4 0.0015 2 0.0006 23 0.0368 29 0.0042 Odacidae Odax cyanomelas 2 0.0027 1 0.0004 0 0 4 0.0072 7 0.0011 Haletta semifasciata 0 0 0 0 1 0.0003 0 0 1 0.0001 Siphonognalhus argyrophanes 0 0 0 0 0 0 1 0.0014 1 0.0001 Odax sp. 0 0 1 0.0004 0 0 0 0 1 0.0001 Percophidae Enigmapercis reducta 12 0.0168 102 0.0376 14 0.0046 3 0.0041 131 0.0179 Pinguipedidae Parapercis haackei 2 0.0028 2 0.0008 152 0.049 1 0.0017 157 0.0219 Creediidae Creedia haswelli 0 0 18 0.0064 0 0 1 0.0011 19 0.0024 Tripterygiidaed Tripterygiid sp.l 3 0.0043 11 0.0041 5 0.0016 5 0.0083 24 0.0034 Tripterygiid sp.2 8 0.0109 9 0.0037 29 0.0102 1 0.0011 47 0.0071 Blenniidaed Parablennius postoculomaculatus 4 0.0054 0 0 3 0.0011 1 0.0011 8 0.0012 Gobiidaed Gobiid spp. 15 0.0207 9 0.0033 120 0.0404 7 0.0122 151 0.0221 Gobiesocidaed Gobiesocid sp.l 2 0.0028 3 0.0012 4 0.0014 45 0.0531 54 0.0062 Gobiesocid sp.2 2 0.0027 1 0.0004 0 0 3 0.0041 6 0.0008 Gobiesocid sp.3 0 0 0 0 0 0 1 0.0014 1 0.0001 Gobiesocid sp.4 0 0 1 0.0005 0 0 0 0 1 0.0002 Alabes sp. 0 0 1 0.0004 0 0 0 0 1 0.0001 Unidentified 4 0.0051 5 0.0021 0 0 1 0.002 10 0.0015 Callionymidae Callionytnus goodladi 58 0.074 37 0.0144 953 0.3064 160 0.2736 1208 0.1708 Bothidae Bothid sp. 0 0 0 0 1 0.0004 0 0 1 0.0002 Paralichthyidae Pseudorhombus jenynsii 4 0.0056 6 0.0022 19 0.0064 35 0.0597 64 0.0096 Monacan thidaed Monacanthid spp. 46 0.0641 80 0.0308 244 0.0803 29 0.0473 399 0.0575 Tetraodontidaed Tetraodontid spp. 0 0 0 0 2 0.0006 0 0 2 0.0003 Unidentified/Damaged 10 0.0121 46 0.0168 100 0.0327 120 0.2102 276 0.0407 The number of stations sampled at each location was: Owen Anchorage (10), Five Fathom Bank (32), Cockburn Sound (38) and Warnbro Sound (8). Families known to produce demersal eggs or brood are indicated by superscript d or b, respectively. ys, yolk sac larvae. 114 Journal of the Royal Society of Western Australia, 95:115-124, 2012 Importance of Lake MacLeod, northwestern Australia, to shorebirds: a review and update D BERTZELETOS *, R A DAVIS & P HORWITZ School of Natural Sciences, Edith Cowan University, 270 Joondalup Drive, Joondalup, WA 6027, Australia. ' Corresponding author [email protected] A number of surveys have been undertaken to determine Lake MacLeod's significance for shorebirds (also called waders). Given the recent global declines in shorebird populations, a current understanding of the significance of Lake MacLeod to shorebirds is timely. We review all survey data in the context of the following criteria: presence of significant national, international and staging populations in the East Asia-Australasian Flyway; species covered by international agreements; and species listed as being of conservation concern in the Action Plan for Australian Birds 2011. Lake MacLeod hosts significant populations of 10 shorebird species. Of these, three species occurred in internationally significant numbers (red knot Calidris canutus, red-necked stint Calidris ruficollis and curlew sandpiper Calidris ferruginca), five species occurred in numbers representing significant proportions of the Australian populations (common greenshank Tringa nebularia, black-winged stilt Himantopus leucocephalus, banded stilt Cladorhynchus leucocephalus, red¬ necked avocet Recurvirostra novaehollandiae and red-capped plover Charadrius mficapUhts) and two species (sharp-tailed sandpiper Calidris acuminata and greater sand plover Charadrius leschenaultii) had populations that met the significant population staging criterion. The most notable of these populations were those of the curlew sandpiper and banded stilt that accounted for up to 31% and 47% of their Flyway and Australian populations, respectively. Lake MacLeod is of great importance to shorebird conservation both in Western Australia and internationally. We recommend that future management strategies focus on maintaining and enhancing shorebird habitats at Lake MacLeod. KEYWORDS: East Asian-Australasian Flyway, Lake MacLeod, monitoring, shorebirds. INTRODUCTION plover (Charadrius ruficapillus) (Rio Tinto Minerals 2008; George 2009). In 2003, it was determined that, at a single Lake MacLeod is a large salt lake (2200km2) located point in time, up to 16% of the Fly way population of the between 23°30'S and 24°40’S and 113°30’E and 114°00'E, curlew sandpiper and up to 26% of the world population on the northwestern coast of Australia, north of of the banded stilt may be present at Lake MacLeod (Rio Carnarvon (Davis el al. 2001; Ellison & Simmonds 2003). Tinto Minerals 2008). Consequently, Lake MacLeod is The northern sections of the lake are fed seawater via a included in A Directory of Important Wetlands in Australia karst system that connects the lake bed to the Indian (Environment Australia 2001) and is eligible for listing as Ocean and results in 2000 km2 of the lake being a Wetland of International Importance under the Ramsar permanently inundated (Logan 1987; Davis et al. 2001; Convention on Wetlands (Rio Tinto Minerals 2008; Russel 2004). This results in a unique, permanent, inland George 2009). saline wetland that is of national and international importance to migrant and resident bird species as a Globally, 52% of shorebird populations are declining, stopover, wintering and drought refuge area (Rio Tinto while only 8% are increasing (Delany & Scott 2006). Minerals 2008; George 2009). Reclamation projects at stop-over sites such as Bohai Bay in China and Saemangeum in South Korea pose Currently, 37 species of shorebirds have been recorded significant threats to shorebirds as they restrict feeding at the lake (Jaensch & Vervest 1990; George 2009). Of habitat available to the birds during this critical period these, many are trans-equatorial migrants that use the and can cause significant population declines (Moores East Asian-Australasian Flyway to spend the austral 2006; Rogers et al. 2006, 2010; Pain et al. 2011; Yang et al. summer (their non-breeding season) in Australia, with 2011). In Australia, development, agricultural pressures the remaining nine species being nomadic residents that on water-tables and human disturbance have contributed move around the continent in response to rainfall to population declines and at some local sites, monitoring (Jaensch & Vervest 1990; Geering et al. 2007; George has detected decreases of those populations by up to 99% 2009). Lake MacLeod has held more than 1% of the for species such as the red-necked stint and curlew Flyway populations of three migrant species: red knot sandpiper (Gosbell & Clemens 2006; Nebel et al. 2008; (Calidris canutus), red-necked stint (Calidris ruficollis) and Wainwright & Christie 2008; Creed & Bailey 2009; Singor curlew sandpiper (Calidris ferruginca)-, and has held more 2009). than 1% of the national population of three endemic species: red-necked avocet (Recurvirostra novaehollandiae), As a result of these threats, 16 species of shorebird are banded stilt (Cladorynchus leucocephalus) and red-capped listed as threatened in Australia in the Action Plan for Australian Birds (Garnett et al. 2010). International agreements between the government of Australia and © Royal Society of Western Australia 2012 those of the People's Republic of China (CAMBA), Japan 115 Journal of the Royal Society of Western Australia, 95(2), July 2012 (JAMBA) and, most recently, the Republic of Korea Department of Agriculture Fisheries and Forestry, (ROKAMBA) have been signed for the protection of CSIRO, Museum of Western Australia, Birds Australia migratory birds (JAMBA 1974; CAMBA 1986; (now Birdlife Australia), Shorebirds 2020 and BirdLife ROKAMBA 2006). These listings and agreements bind International. The domestic databases and publications the Federal Government (and thereby other levels of of some of these organizations were also searched with government in Australia) to the conservation of the same keywords. Where appropriate, interviews with shorebirds and their habitats and encourage scientific key personnel were sought. research at the national and international level. All relevant records received were stored in Excel™ The broad objective of this paper is to evaluate the files. Data manipulation was focused on summarising importance of Lake McLeod to shorebirds. Specifically, total species counts in relation to survey periods, and we aim to: (i) identify instances where the lake supports assessing counts against the various criteria (e.g. 1% more than 1% of the global or national populations of a criterion) as previously outlined. species; (ii) determine if the lake is an important stopover site for a species (i.e. counts of 0.25% or more of the Evaluating Lake MacLeod's significance for shorebirds flyway population); (iii) evaluate the importance of the From the database search, 15 reports and journal lake for listed declining shorebird species; and (iv) publications were discovered that contained information establish the significance of the lake for those species on the shorebirds of Lake MacLeod (Appendix 1). Only protected under international conventions. 12 of these can be described as surveys, or contained enough information to be used for the purposes of this study. These surveys also identified areas within the lake METHODS that were of the most importance to shorebirds (Table 1). The Shorebirds 2020 database included 18 counts from Data collection and interrogation Lake MacLeod. Of these, 11 represent the counts from We were interested in all bird count data from Lake the above surveys whilst the remaining seven only cover MacLeod. The methodology for our data collection was part of the lake and due to uncertainty as to methods and adapted from Horwitz et al. (2010). Online database coverage, were excluded from this study. engines such as Google Scholar™, ISI Web of Shorebird counts for Lake MacLeod were Knowledge™, BioOne™ and others, were used for a consequently available from the following surveys: desktop literature search. The key words used were 'Lake Jaensch & Vervest (1990), Davis et nl. (2001), Davis (2002, MacLeod' and 'Carnarvon Basin' in conjunction with the 2003a, 2003b, 2004) and Hassell (2004, 2005, 2006). For following secondary key words: Fauna Survey of. Avian more information see Appendix 1. All counts from all Survey of. Fauna Report of. Avian Report of. surveys are available in George (2009). Birdwatching Event of, Shorebirds of. Waders of, Rio Tinto, Dampier Salt and others. Organisations included East Asian-Australasian Flyway population estimates in the search were Dampier Salt Ltd, Western Australian of trans-equatorial migrant species were acquired from Department of Environment and Conservation, Bamford et al. (2008). Australian population estimates for Commonwealth Department of Environment, Water, migrant species (that is the number of individuals of a Heritage and the Arts (now DSEWPaC), Commonwealth migrant species visiting Australia) and for species Table 1 Areas of importance to shorebirds in Lake MacLeod. Site Coordinates Habitat Important for: Use 1. Jack's Vent (and 23°57’40"S 113°38'25"E Mangrove-lined channels Bar-tailed godwit, great knot, common Feeding surrounding channels between the vents and greenshank, black-winged stilt, and vents) open-water areas banded stilt 2. 'Roost Site' 23°55'40"S 113°38'10"E Dry samphire cover Various shorebird species Roosting 3, 4. Godwit Beach to 23°57'07"S 113°38'48"E Extensive shallow mudflats, Red-necked stint, curlew sandpiper, Feeding Linda's Creek stands of stunted dead other trans-equatorial migrant mangroves shorebird species. 5. Stilt Lagoon 23°58'20"S 113°38'34"E Shallow expanse of water Red-necked avocet, banded stilt Feeding 6. Oystercatcher Flat 23°58'43"S 113°38'24"E Large mud-bank barrier, Banded stilt, red-necked avocet Feeding no mangrove background 7. Stilt Flat 23°58'59"S 113°38'24"E Inundated mud flat Banded stilt, black-winged stilt, other Feeding trans-equatorial migrant shorebird species. 8. Dogleg Channel 23°59'36"S 113°36'47"E Shallow mud channel Red-necked stint, curlew sandpiper, Feeding sharp-tailed sandpiper, other trans- equatorial migrant shorebird species. 9. Phalarope Flat 24°00'27"S 113°37'57"E Shallow water over sandy Red-necked stint, curlew sandpiper, Feeding, mudflats banded stilt roosting. Adapted from Davis et al. 2001; Davis 2002, 2003a, 2004; Hassell 2004, 2005, 2006. 116 Bertzeletos et al.: Importance of Lake MacLeod to shorebirds endemic to or resident in Australia were acquired from RESULTS Geering et al. (2007). To determine the importance of Lake MacLeod to Lake MacLeod's significance to shorebirds shorebirds, five criteria were used. Using the Ramsar criteria (Ramsar Convention Bureau 1. To examine if a locality was internationally 2000) and Watkins (1993), 10 shorebird species were significant, Criterion Six of the Ramsar Convention was found to have significant counts at Lake MacLeod. used. This criterion states: 'a wetland should be Internationally, the lake holds significant numbers of red considered internationally important if it regularly knot, red-necked stint and curlew sandpiper (Appendix supports 1% of the individuals in a population of one 2). Nationally, the counts for these three species are also species or sub-species of water bird' (Ramsar Convention significant. The area also holds nationally important Bureau 2000). A site is determined to 'regularly support' populations of an additional five species: common 1% of a population if the 1% threshold is achieved in two greenshank, black-winged stilt (Himantopus out of three seasons, or is met by the mean of at least five leucocephalns), banded stilt, red-necked avocet and red- maximum annual counts (Ramsar Convention Bureau capped plover (Table 2). 2000). For migrant taxa, the 0.25 % staging criterion was 2. To determine if a locality is important during reached for six species: the common greenshank, red migration as a stopover site, the guidelines present in the knot, red-necked stint, sharp-tailed sandpiper (Calidris Ramsar convention were used. These state that a locality acuminata), curlew sandpiper and greater sand plover is significant as a stopover site if counts are equal to, or (Ch. leschenaultii) (Table 2). greater than 0.25% of the fly way population, at least once The average numbers of these species present on the (Ramsar Convention Bureau 2000). Non-migratory lake, the frequency with which these numbers meet the species were not assessed under this criterion. various criteria and the percentages of the East Asian- 3. To investigate if Lake MacLeod is significant on an Australasian Flyway populations (Bamford et al. 2008) Australian-only level, the two criteria proposed in that these numbers represent are shown in Table 2. Watkins (1993) were examined. These criteria signify that Detailed counts for all 37 species are in Appendix 2. a site is of national importance for shorebirds if at least Lake MacLeod hosts 12 species that are listed in the 1% of the Australian population or 10 000 or more Near Threatened or more severe categories (Appendix 2). individuals have at any one time been recorded at the Of these, the Asian dowitcher (Limnodromus site (Watkins 1993). Furthermore under this plan, any semipalmatus), ruddy turnstone (Aremria interpres), grey¬ counts that are internationally important are tailed tattler (Heteroscelus brevipes) and black-tailed automatically considered to be of national importance godwit (Limosa limosa) are listed as Near Threatened, the (Watkins 1993). great knot (C. tenuirostris), red knot, curlew sandpiper, 4. To examine if Lake MacLeod was a significant eastern curlew (Numenius madagascariensis), bar-tailed locality for shorebirds under threat of extinction, the godwit (L. lapponica), greater sand plover and grey plover conservation status (IUCN rankings) of the shorebird (Pluvialis squatarola) as Vulnerable and the lesser sand species occurring at Lake MacLeod was acquired from plover (Ch. mongolus) as Endangered. All 28 migratory Garnett et al. (2010) which has the following rankings: species occurring on the lake are protected under the Least Concern (LC), Near Threatened (NT), Vulnerable three international agreements (Appendix 2). (VU), Endangered (EN), Critically Endangered (CE), Extinct in the Wild (EW) and Extinct (EX). 5. Lastly it was checked to see if these species were DISCUSSION also covered by the three international agreements Apart from George (2009), this is the first study to use all (CAMBA, JAMBA and ROKAMBA). Although this was the survey data collected from Lake MacLeod to examine also done for the Lake MacLeod Management Plan (Rio the lake's importance both on a national level and as a Tinto Minerals 2008), it was discovered that this plan did staging area during shorebird migration. Consequently, not list all the species present in these agreements and we were also able to provide an update on the lake's thus a repeat of the procedure was deemed necessary. international importance, against the latest East Asia- Since the three agreements largely cover all migrant Australasian Fly way population estimates (Bamford et al. shorebird species occurring at Lake MacLeod (JAMBA 2008). 1974; CAMBA 1986; ROKAMBA 2006), they can be considered collectively for this criterion. Ten species of shorebird were found to have significant numbers at Lake MacLeod. Of these, eight The above analysis was undertaken for all 37 species have had international and nationally significant of shorebird recorded at the lake. For species meeting the numbers, while for the other two species, the area may 1% criterion either globally or for Australia, the average be important as a staging area. number of birds was calculated as well. This was done to determine if the site holds significant numbers of these species on average. Tire number of times that surveys International and national significance for shorebirds met the 1% criterion was also noted, to determine how The internationally significant numbers of red knot, red¬ frequently the lake holds significant numbers of the necked stint and curlew sandpiper have been previously particular species, Finally maximum counts for these highlighted (Hassell 2006; Rio Tinto Minerals 2008; species were used to specify the maximum percentage of George 2009). The counts for these species are also the population, Flyway arid Australian, using Lake nationally significant. Lake MacLeod also held 1% of the MacLeod. Australian populations of an additional five species: 117 Journal of the Royal Society of Western Australia, 95(2), July 2012 Table 2 Shorebird species for which Lake MacLeod holds significant numbers. Species 1%F/Wa l%Aust.a 0.25%>’ Average' %F/WC %Aust.c MaxJ %F/W“ %Aust. Common greenshank 0 2 2 128 0.21 0.67 300 0.5 1.58 Red knot 1 1 3 452 0.21 0.34 2566 1.17 1.90 Red-necked slint 8 8 9 5418 1.67 2.01 25000 7.69 9.26 Sharp-tailed sandpiper 0 0 2 155 0.10 0.11 850 0.53 0.61 Curlew Sandpiper 9 9 10 21216 11.79 17.98 55000 30.56 46.61 Black-winged stilt N/A 1 N/A 645 N/A 0.22 3008 N/A 1.00 Banded still N/A 9 N/A 19691 N/A 9.56 100000 N/A 46.61 Red-necked avocet N/A 3 N/A 700 N/A 0.65 2401 N/A 3.30 Red-capped plover N/A 3 N/A 830 N/A 0.87 3125 N/A 3.30 Greater sand plover 0 0 1 61 0.05 0.08 515 0.46 0.69 a The 1% F/W and 1% Aust. represent the number of surveys (out of 12 total) for which Lake MacLeod was found to hold at least 1% of the East Asia-Australasian Flyway (EAAF) and Australian populations of these species in accordance with the Ramsar (Ramsar Convention Bureau 2000) and A National Plan for Shorebird Conservation in Australia (Watkins 1993) criteria. b For migrant taxa, the 0.25% column represents tire number of surveys for which the 0.25% staging criterion was met (Ramsar Convention Bureau 2000). c The average number of individuals recorded from the surveys together with the percentages these represent of the flyway and Australian populations. d The 'Max' column lists the highest counts of these species recorded at the lake and the proportions of the flyway and Australian populations these represent. N/A (Not Applicable) entries are for species that are resident and/or endemic to Australia and thus do not use stopover sites or the flyway. EAAF population estimates adapted from Bamford et al. 2008, Australian population estimates adapted from Geering et al. 2007.' common greenshank, black-winged stilt, banded stilt, inundation), meaning that high counts may coincide with red-necked avocet and red-capped plover. The site's conditions where other inland wetlands are dry. national significance for the last three species is already known (Hassell 2006; Rio Tinto Minerals 2008; George Staging area, species threat status and coverage by 2009), however, it is the first time that it has been realised international treaties for the common greenshank and black-winged stilt. As a staging area. Lake MacLeod may be important for Of these eight species occurring at Lake MacLeod at six migrant shorebird species as some counts for these nationally and internationally significant levels, only species equaled or exceeded the 0.25% of their estimated three (red-necked stint, curlew sandpiper and banded Flyway populations. These are the common greenshank, stilt) have had significant counts in most of the surveys, red knot, red-necked stint, sharp-tailed sandpiper, and the proportions of the populations that occur are also curlew sandpiper and greater sand plover. This is also high. For the curlew sandpiper and banded stilt, the supported by observations during the start of the maximum proportions of the East Asian-Australasian northbound migration of fat birds (curlew sandpipers Flyway and Australian populations recorded in the area and red-necked stints) in partial or complete breeding are double those presented in the Lake MacLeod plumage at Lake MacLeod (Davis 2003a). However, this Management Plan (Rio Tinto Minerals 2008); 16% then is complicated by the fact that Australia is a non-breeding versus 31% now of the fly way population for the curlew season destination making it difficult to distinguish sandpiper and 26% versus 47% for the endemic banded between non-breeding and stop-over birds (Geering et al. stilt. The high frequency with which these large numbers 2007; Bamford et al. 2008). In order to fully understand occur at Lake MacLeod suggests that for the curlew the use of the lake as a staging area one would have to sandpiper the area may be a non-breeding site of major calculate the 'turnover rate' of birds that are in transit importance. Likewise for the non-migratory banded stilt, using the lake (Colwell 2010). This would be best the area appears to be an important drought refuge area. achieved by colour-marking or radio-tagging individuals T he species has not been recorded breeding in the area (Iverson et al. 1996; Warnock and Bishop 1998; Battley & and departs when suitable conditions present themselves Rogers 2007; Colwell 2010). elsewhere (Hassell 2006). All 28 migratory species occurring at Lake MacLeod For the other five species, the number of significant are protected by international treaties (JAMBA 1974; counts and the populations using Lake MacLeod, are CAM BA 1986; ROKAMBA 2006) (Appendix 2). This much lower. These suggest that for these species the lake includes the 12 species at risk of extinction that occur in is of lesser significance or that significant counts are the area (Appendix 2) (Garnett et al. 2010). However, only unusual events, for example, when regional weather the red knot, curlew sandpiper and greater sand plover conditions make the lake attractive to these species. have been recorded as having internationally significant Conversely, regional weather conditions may result in populations there and, of these three, only the curlew Lake MacLeod being the only water source present away sandpiper has occurred in significant numbers for the from the coast (due to its karst link and permanent majority of the surveys. Based on this and the large 118 Bertzeletos et al.: Importance of Lake MacLeod to shorebirds numbers using the site, Lake MacLeod is considered to between DSL and ECU. We gratefully acknowledge the be of vital conservation importance to at least the curlew assistance of Journal reviewers M Bamford and M Craig sandpiper. in improving this manuscript. Lake MacLeod's importance for shorebirds compared with other sites in Australia REFERENCES Unlike wetland areas in the southeast and southwest of Bamford M, Watkins D, Bancroft W, Tischler G & Wahl J 2008. the continent where some shorebird populations have Migratory shorebirds of the East Asian-Australasian Flyway; declined, the numbers of shorebirds at Lake MacLeod population estimates and internationally important sites. appear to have remained high. This is especially true for Wetlands International - Oceania, Department of Sustainability, Environment, Water, Population and the curlew sandpiper where the number of birds utilising Communities Canberra. sites in Victoria, for example, has declined by 80% from Battley P F & Rogers D 12007. Migration. In: Geering A, Agnew 30 000 birds in the early 1980s to about 5000 between L & Harding S (eds) Shorebirds of Australia, pp. 35-49. CSIRO 2004 and 2006 (Gosbell & Clemens 2006). The species has Publishing, Melbourne. a negative trend of 3% per year and similar declines have CAMBA (Ouna-Australia Micratory Bird Agreement) 1986. been observed in bar-tailed godwit, red-necked stint, Agreement between the Government of Australia and the sharp-tailed sandpiper and the Australian populations of Goverment of the People's Republic of China for the other species in Victoria (Gosbell & Clemens 2006; Nebel Protection of Migratory Birds and their Environment. 119 Journal of the Royal Society of Western Australia, 95(2), July 2012 Horwitz P, McLure N & Hewitt P 2010. Lake MacLeod: Rogers D I, Moores N & Battley P F 2006. Northwards biophysical characteristics. Data audit and research migration of shorebirds through Saemangeum, the Geum framework for the Northern Ponds of Lake MacLeod, Estuary and Gomso Bay, South Korea in 2006. Stilt 50, 73-89. Western Australia. Centre of Ecosystem Management, Edith Rogers D I, Yang H Y, Hassell C J, Boyle A N, Rogers K G, Cowan University, Perth. Chen B, Zhang Z W & Piersma T 2010. Red Knots (Calidris Iverson G C, Warnock S E, Butler R W, Bishop M A & Warnock canutus piersmai and C. c. rogersi) depend on a small N 1996. Spring migration of Western Sandpipers along the threatened staging area in Bohai Bay, China. Emu 110, 307- Pacific coast of North America: a telemetry study. Condor 98, 315. 10-21. ROKAMBA (Republic of Korea-Australia Migratory Bird Jaensch R & Vervest R 1990. Waterbirds at remote wetlands in Agreement) 2006. Agreement between the Government of Western Australia 1986-88 Part 2: Lake MacLeod, Shark Bay, Australia and the Government of the Republic of Korea on Camballin Floodplain and Parry Floodplain. RAOU Report the protection of migratory birds and exchange of 69 notes. 120 Appendix 1 Survey effort and methodology at Lake MacLeod. H jij 45 ^ 'T' re 111 J1J u3 <3 re < ^ CD .2 ^ 'fa ^ o - 2 ’■£ • »HOJ in < 3 ■*- re 2 O re G-OJ > .fa 3 33 2 re E - s c 2 ■a 6 *5 cu £ O H o .5- h^< X> -J“±S ■3 •§aJ i 5JS^ "T3 £ Pi CD 43 3 CM 2U re xoj JX roo3 O) 2CD O 3 c o (5 O^ H3 45 CD S- £ OJ fa h 3 QJ Li u > CD (X) Of >, g ^ £ 3 S3 g cd ^ M CD § 8 &c >- C2 W Id T3 « O 1 c o c at re ^ «. ,-i s "? E -§ ? — re £ o — o jjj _CD 0J CC 3 0J O- C Q _ S. Q CD £ Q Cu (0 o £ C fa CD 53 a £ 3? TS Pi « ;* = «|c§2 iis < CDO> B.2 =<*£S “’ —bJQJ"O 44 35*-c 1 E£§B •d 2^ s c U. 3UJ _ racn J2 oj re o oi jv Li CD CD 3 2 £ 1 Si g >o u $ re cd»_, ^ =o.S 5; to"2"c J O' 00 IN Oj g*"* OQ-CD < CN oo ire vO ~ Op 33 Pi S "du 00 ■5 SS 13 siIS ^ -a - a. 13 c -J rfS ^-s re oj x* OJ CJ ■“ n r3cd Ic ■ ..x-hre Bertzeletos etal.\ImportanceofLakeMacLeodtoshorebirds OS 43 2 re H = ~ "o , re fa CD PiPh •2 3 3 T5 Q^Ui T3 'H- 're c Q 4P O < re>oU; 5* cSCre■ —1 01CD3O i> re^ ’g -cID ^3-3 =0 ^ ! o 2 - T3 D. 2i CD .2i CDO < o 4= 3 '5 ^ ■a § s sj4 *- QJ-3 "cj ■*-• & o re g-g Scg a> CD 3 Con 3 CTn O 3 32 - ^reOJ - ^-eom a; S-aS 3 22 o o 44 w ^ore CL .-3 O *Sr-"±2 o OX. 3 re o> •*-> ^ *C g 2 re on o» o re S CD V > 0> > d t- CTO\o 44 “CCD (DidwUIm X 273«JSa —; 0) K ■a cx ^ 2mfflOJ si SEcuX,CDLi oc > CD 3 UH 5, Xtn _> H.5 vn 2.3 ^ £« id *red h oT5 ^ C £30 C >23L-32 c OJ>C £ tScj re o<^ . >23 2 Si ° O5 • Qj01 Q 5 Ji^g o CD' <, o ^ CQ ■s % D I1 ^ re ub oj^3 ■£ -3- ■8S § 1 gs ■si 1 2 E 2 5c| O •*_L-. C -3 >-. re g;0J| O *-,-r1 L- CDCJ Oi 3O V T3 121 §o‘S u OI ■P 6SBi13 5 g<-2'g •5 S2§ CQ £,n m K < w x 2 < 2 2 < 2 < A C40ON ■°-l ^ 2 < — 3^-343 *1 40reo ■334= « oj ~3c'y o ^E2 a> recdc E bo ra -3A CD T2reO'-1 s CDOJra >x G c| eg c <“ O —3IN. 3 ^C35 r 3o\o 3 ‘*5jS “ §I CD C re ■*- CD Li a» CL o T3 a» m ■*•* CD£ re* .23re O rerH ju .52 8 £ 52 o >x OJ 'S TJ -u ■s. —. £ O 3 o ^ s -a Aon O O XCDCN Z Pi CD .S3 x>o -3 re 3 32 > reon 0044 ON *H QJ, O 3 CL re .fa3§ >44 o > o QJ A A 44 t>JD!Z! QJ O s 2 OJ _Q re ro c^ 40-3 re o 0 2 O Ol u g2 O *CD rx T2h s 3> o c C > ^ CD X2 CD O re 3 *5 l. rz > u aj j A OJ £ % re v- S!li QJ T3 •§ o Id 2 2 g aj re .re2 3 cu. c>o oj' 2 o X 32 >x _ o-1 'cd re OON r, rre ^ 43 I>x X) z o o > O 40 CD 44 £ O 44 u O OJ 3 c o c c 3 £ C 3 o January 2002 Survey” Unknown Unknown 18-21/1/2002 4 - - - Water levels low, access difficult. Southern basin not visited. Low numbers of birds recorded. I do not have access to the original report. Information from Davis 2003 Journal of the Royal Society of Western Australia, 95(2), July 2012 cn ^ 60 toO ^ > x B S u/ G cn 7311 ~ S 73 £ 73 In > o 73 too ro .G cn •— f O B 01 33 cn G3 oi ■“ too § .£ £ 3 OJ . 3 tn to ro O 33 s &-§ ci to £ 01 > 2 toCc •— i_ «- > '7- u JPS’O ,«2 ~ too J- Ol 3 (3 G -w 73 73 O (8 O u C £ 3 3 E<—i 8 O 2 ^ 8 ^o> O J to 3 <— cn S3 toC-^ > e g frje 33 u 2 £ O 23 o £ ^ oi Q £ 33 K & 2 S O U > 8 ,5 5 (3 23 73 73 -3 — re « = r'- 73 toO D >• o •83 cfl Ol 73 cd ~ - 8 3 * - £ 3 2-1 £ S O ~ 5 ■o "S cn qj CD f’S S p 73 to3 E ‘-3—_ 73u -3— 8cn 3 o c;.° g I« i to^; j£ 3 |g toO 8 r > jo •- C to C • =3 U 73 3 ns X rr- to " o > K vw G X ut O cn 8 *.S Cj l— 3 ^ 73 O ’ « « ^ ^ 73 c8“ G m •X 73 cn > y 3 § V 5> ro (8 X f8 *!— o •£ .£ ■§ 2 top s $ 5 | >- cn _ Sfi CU E CO cn 8 -O^ g G C — £ f8 s j| bx 8 ~ Sul, » a* ^ 3 o ^ — -g SO G fco^ 0-0 &- S ■73 too ^ o ^ SB I zj .2 c •-03^ 2U •G re on 01 CQ O' >- r- _cn *- o 8 2: u 3 u £ 2 OJ* > 3 cn ^ « cn 73 u >H CJ O Cn ~ — ° § >- y o (T- 73 ■^ £ to 6 p g y o G 07 5 £ > 8 S-gg ■3 G 22 o» g S"0 g 0 0^3 ■G *3 53 3 O 11 x-r >> o ill -J 73 .2 ■£ 12 o~S 2 H U a) Z u c J3U* U n 13 3 8 2 *5 Q 73 6 tn U J3 >. cn CD 5-^. G ° 33 C Pn (8 C <2 £ O. o O o O o a> < & fc Z 2 2 2 2 < a 'S 8 6 '3 2 2 r | S. " n> <: 2 •« 3 C 2 Q - CO o o (N CN £ S CN & — Cp = 8 O 2 CD ^ CN GO .s o 73 8 3 01 U u u U 2 cn cn cn cn and Autl Davi 2002 Davi Davi 2004 Davi 2003 2003 u s 0> ^ o QJ <8 D- o —cn c8Ih £ Q. 3 CN X X Ui oi SJ 3 0) CO hJ h O o LD £ o 3 ko £ 2 s O U- V B i Nw CD CD o 1 OJ S Cl* | O 0) Cii ^3 i a ^ ^ g E 73 CN U.) ^ g QJ J- ° *- 73 73 O ^ IN g ^ — 3 ni O O (~si °f 73 a O CN 3 « O u ^ jg 5 O ^ 8 3S O 73 O) 0J S •a ^ S J ^ 2 >,s J tN 2 . u o ra j •—1 o V O QJ ID QJ 3 hJ “? w -S £ ^ ci V ■ 10 cn > y cn 3 a -q o frgJ1?• O •a S u CO TJ # I S 73 £ -3 £ g 8 33° 8 CN & »L o £ H i- 23 8 >■ ra G •7 cn 1/1 « o 73 O g -o . J2 3 S y O CN £ 3 ^ c~ QJ 2 2 2 >n «s° >. m S J I ° 01 « -J ai j 03 K ga 01 > 0> O 3 01 rvj c 73 ^ o o (£ '5fc2 Cl •* £ > “A -cs £ to n O 8 V- |_. 73 ra o to 3 8 3 ■ O < Ch i—J CN >—J on 122 Bertzeletos et alImportance of Lake MacLeod to shorebirds o o o o m in co00 °in LT> < CN on m 10 2 < 2 .5 Oh LO CN S g. CO O O V05 s a CN 53 QJ o H CN CO O ^ QJ CO CN Oh 0) -G H-» z s 03G 0) u C 03 G bJD « o a» I G Oh O CO QJ O CO CN col T3 G 0 o o o o o in 03 C _o vo o o ^ 13 c '-M o • ° QJ Oh O oj o CO fN 03 X 4—* ’G O OJ H-I u 03 JO a O) os •5- I •S S Bo fQ G u I ~ -4 U « C 60 £ *5 .co .52 .JQ .52 « S »- ^ V. Sh. G &0 &0 §0 3 :s QJ S *Q is K ^ ^ ^ ^ rs u ? *QS "S "a ,s CO 5 Z £ £ u u u u u <3 U u u p .X ■s c o 03 Oh hG £ X Oh CO QJ "O c/) "Oh C QJ < .s < O- < < o £ G o £ ' < ’5- 60 T3 s a p f l" Oh CN s O Ou s c 15 ^ y a« T3 u u Oh Oh 03 > •J J a c T3 ~o X G 60 ■s s -G T3 QJ B 4 ■? 4 • pH a " G T3 £ 5 60 _ —H U T3 C C o "Z QJ P T3 0 C 3 o S S U — 03 Oh G o ra ^ „ ■a £ T3 t- § C/^ *-\ H-H —H s •o “Or^Li-fUrros0 r > r G*-To- 3 c'^ Oh ■ Si-H T3Cl c r o H P s o ' ° u QJ U § u .% H 3 C 03 •- ojU oU oj U r § o P g 5 o < U CO CO Z > u $ Jh JU JOZ W 2 O K CO hJ ^hJhJhJOhhJcO [Eu>cq ™0h 123 Haematopodidae Pied oystercatcher Haematopus 19 6222166682 24 NA NA NA NA 10000 LC longirostris C < 2 < 2 2 < 2 CO H J2 U V. E 3■ o PP hJ c to Journal oftheRoyalSocietyWesternAustralia,95(2),July2012 ~ <3 & bK a. =>>*o 2 < < 2 < 2 < 2 ft o o U *8 1 nl ** cn 13 CQ c ^ a §S*§ c 2 < I o Pi xj U u o > R :c •a 2 3 On o o o 5 £>,y P- CuUPJJUU^OJ V, 2.3-g co Ui.‘ooUO2-_v_x U l.PS Q J Instructions to Authors The Journal of the Royal Society of Western Australia publishes peer-reviewed articles dealing with research in all aspects of the natural and physical sciences of relevance to Western Australia. Research papers reporting original research should not normally exceed 5000 words; review papers should not exceed 10 000 words. Short papers not exceeding 2000 words will be given priority in publication. Discussions and Replies of recently published papers in the journal should not exceed 1000 words. Authors are encouraged to submit a list of potential reviewers in Australia or overseas, but choice of reviewers is at the discretion of the Editor. Submission of a paper is taken to mean that the results have not been published or are not being considered for publication elsewhere. Authors are solely responsible for the accuracy of all information in their papers, and for any opinions they express. Papers accepted for publication become copyright of the Royal Society of Western Australia and authors will be required to assent to the transfer of copyright. A pdf of the published paper is provided free to corresponding authors and their coauthors; hard copy offprints may be ordered at proof stage at a rate of $100 (incl. GST and postage) for 50 offprints. Manuscripts should be submitted as pdfs or Word (doc or docx) files to the Editor in the following format: Title page; Abstract; Key words; Text; References; Figure captions; Tables; Appendices (if any); and Figures. Headings: Authors may use up to four levels of headings, all of which are left justified: first-order headings are bold capitals; second-order headings are bold upper and lower case; third-order headings are capitals; and fourth- order headings are upper and lower case. All pages should be numbered consecutively, including those carrying tables and captions to figures. Illustrations: All illustrations, both line drawings and photographs, are to be numbered as figures (not plates) in sequence, and each must be referred to in the text in consecutive order. In composite figures made up of several photographs or diagrams, each part should be designated by a lower-case letter (e.g. Figure 2b). Authors are advised to use the most recent issues of the Journal as a guide to the general format and bibliographic style of the manuscript, and to conform as closely as possible to the style in preparing the manuscript. The metric system (SI units) must be used. Taxonomic papers must follow the appropriate International Code of Nomenclature, and geological papers must adhere to the International Stratigraphic Guide. Spelling should follow the Concise Oxford Dictionary. Following the review process, authors should revise their manuscript and send one printed copy together with a digital (doc or docx) file to the Editor for final technical editing. The Title should begin with a keyword. The Abstract should not be an expanded title, but should include the main substance of the paper in a condensed form. Six to ten Keywords should be provided. The Text should be organised under the following headings: Introduction, which should contain the reasons for doing the work and essential background material, but not the results or conclusions; Methods (if applicable), which should provide sufficient details of techniques to allow them to be repeated; Results, which should not include material more appropriate to the discussion; Discussion, which should emphasise the significance of the results and place them in the context of related work; Conclusions, which summarise the main findings of the paper; Acknowledgements, which should acknowledge any appropriate assistance from colleagues and also the journal reviewers. References must be set out as follows, with journal titles spelled out in full. Paper: Collins L B, Testa V, Zhao J & Qu D 2011. Holocene growth history and evolution of the Scott Reef carbonate platform and coral reef. Journal of the Royal Society of Western Australia 94, 239-250. Book: Harper J L 1977. Population biology of plants. Academic Press, London. Chapter in Book: Davie P J F & Short J W 1995. Decapod—Anomura, Brachyura. In: Wells F E, Hanley J R & Walker D I (eds) Marine biological survey of the southern Kimberley, Western Australia, pp. 118-126. Western Australian Museum, Perth. High-quality Illustrations are required, no larger than 300 mm x 400 mm, with sans serif lettering suitable for reduction to appropriate publication size (81, 115 or 170 mm wide). Scale must be indicated on illustrations. Photographs must be good quality prints, not exceeding 170 mm x 252 mm. Colour figures incur a charge of $300 (incl. GST) per colour page. Additional printing costs for folding maps and charts will be charged to authors: please consult the Editor before submitting these items. Figures should also be provided as high-resolution (300 dpi) tif, jpg oreps files to ensure high-quality reproduction. Extensive sets of data, such as large Tables or Appendices, may be classed as Supplementary Papers and not printed with the paper but lodged with the Society's Library (held in the Western Australian Museum, 49 Kew Street, Welshpool, WA 6106) and with the National Library of Australia (Manuscript Section, Parkes Place, Barton, ACT 2600). They should be supplied in printed and digital form (preferably as pdfs). Copies of a Supplementary Paper may be obtained from either institution upon payment of a fee. Contributions should be sent to: Editor-in-chief Journal of the Royal Society of Western Australia 104 Hensman Street South Perth WA 6151 or (preferably) L! [email protected] Also available from the Royal Society of Western Australia: (Prices are ex GST; discount prices for members are in parentheses) Regular issues of the Journal (price per part) $12 ($8) The Leeuwin Current: An Influence on the Coastal Climate and $20 ($15) Marine Life of Western Australia (Volume 74,1991) Plant Diseases in Ecosystems: Threats and Impacts $20 ($15) in South-Western Australia (Volume 77, Part 4,1994) De Laeter Symposium on Isotope Science $20 ($15) (Volume 79 Part 1,1996) Journal of the Royal Society Western Australia Volume 95 Part 2 July 2012 Symposium on the Design of Reserves for Nature Conservation $20 ($15) in South-western Australia (Volume 79, Part 4,1996) Mini-symposium on Remote Sensing and its Applications in $20 ($15) Western Australia (Volume 80, part 1,1997) Granite Outcrops Symposium (Volume 80, Part 3,1997) $20 ($15) Centennial Issue 1897-1997 (Volume 81, Part 1,1998) $20 ($15) Management of Granite Outcrops (Volume 83, Part 3, 2000) $20 ($15) Leschenault Inlet Estuary (Volume 83, Part 4, 2000) $50 ($35) Fire and Wetlands on the Swan Coastal Plain (Volume 88, Part 3, 2005) $20 ($15) Holocene sedimentation, stratigraphy, biostratigraphy, and history of the $25 ($20) Canning Coast, north-western Australia (Supplement to Volume 91 Part 1 2008) Leeuwin Current Symposium 2007 (Volume 92 Part 2, 2009) $65 ($45) The Habitats of the Western Australian soldier crab Mictyris occidentalis Unno 2008 $25 ($20) (Brachyura: Mictyridae) across its biogeographical range (Supplement to Volume 92 Part 3, 2009) Symposium on Evolutionary Biology (Volume 92 Part 4, 2009) $45 ($35) Symposium on Kimberley Marine and Coastal Science (Volume 94 Part 2 2011) $65 ($45) Please add postage and handling at the following rates: Australia Overseas Regular issue $5.00 $6.00 Thematic Issues $6.00 $10.00 (Surface Mail) The Leeuwin Current: An Influence on the Coastal Climate and Marine $12.00 $18.00 (Surface Mail) Life of Western Australia DeLaeter Symposium on Isotope Science $12.00 $18.00 (Surface Mail) Granite Outcrops Symposium $12.00 $18.00 (Surface Mail) Leschenault Inlet Estuary $15.00 $30.00 (Surface Mail) Leeuwin Current Symposium 2007 $12.00 $20.00 (Surface Mail) Kimberley Marine and Coastal Science Symposium 2011 $12.00 $32.00 (Surface Mail) Combined postage discounts apply, airmail POA Please send your order, including payment (all prices in Australian dollars), to: Honorary Treasurer, Royal Society of Western Australia, PO Box 7026, Karawara, Western Australia 6152. journal of the Royal Society of Western Australia© is published quarterly by: Royal Society of Western Australia, PO Box 7026, Karawara, Western Australia 6152. 0 [email protected] ISSN 0035-922X Print Post Approved PP642409/00008. Claims for non-receipt of the Journal should be submitted within 12 months of publication, to: RSWA Librarian, PO Box 7026, Karawara, WA 6152, 53 [email protected] Address changes should be submitted to Membership Secretary, PO Box 7026, Karawara, WA6152. © ROYAL SOCIETY OF WESTERN AUSTRALIA 2012 This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced by any process without written permission from the Editor-in-chief of the Royal Society of Western Australia, El [email protected]. A permissions line must be provided crediting author(s) and the Journal of the Royal Society of Western Australia as the source. Typeset and printed in Western Australia