sign in sign up
<<
Home , Baw Baw frog, Blue Mountains tree frog, Blue poison dart frog, Booroolong Frog, Broadwater National Park, Chytridiomycosis

Declines and Disappearances of frogsAUSTRALIAN

Edited by Alastair Campbell Biodiversity Group Environment GPO Box 787 Canberra ACT 2601 © Commonwealth of Australia 1999

Published by Environment Australia.

ISBN 0 642 54656 8

Published December 1999

This work is copyright. Information presented in this document may be reproduced in whole or in part for study or training purposes, subject to the inclusion of acknowledgment of the source and provided no commercial usage or sale of the material occurs. Reproduction for purposes other than those given requires written permission from Environment Australia. Requests for permission should be addressed to Assistant Secretary, Corporate Relations and Information Branch, Environment Australia, GPO Box 787, Canberra, ACT, 2601.

For copies of this publication, please contact Environment Australia’s Community Information Unit on freecall 1800 803 772.

The views expressed in this report are not necessarily those of the Commonwealth of Australia. The Commonwealth does not accept responsibility for any advice or information in relation to this material.

Front cover photo: rheocola, Creek Environment Australia Library

Photo by: Keith McDonald

Designed by: Di Walker Design, Canberra Contents

Foreword

Preface

The Gordian Knots of the International Declining Populations Task Force (DAPTF) Stan Orchard 9

A Review of Declining in Northern Keith McDonald and Ross Alford 14

Chytrid Fungi and Amphibian Declines: Overview, Implications and Future Directions Lee Berger, Rick Speare and Alex Hyatt 23

Measuring and Analysing Developmental Instability as a Tool for Monitoring Frog Populations Ross Alford, Kay Bradfield and Stephen Richards 34

An Assessment of Frog Declines in Wet Subtropical Australia Harry Hines, Michael Mahony and Keith McDonald 44

The Status of Rainforest Stream Frogs in North-Eastern NSW: Decline or Recovery? Ross Goldingay, David Newell and Mark Graham 64

Frogs in the Timber Production Forests of the Dorrigo Escarpment in Northern NSW:An Inventory of Present and the Conservation of Francis Lemckert and Rachael Morse 72

Review of the Declines and Disappearances within the Bell Frog Species Group (Litoria aurea species group) in Australia Michael Mahony 81

A Preliminary Assessment of the Status of the Green and Golden Bell Frog in North-Eastern NSW Ben Lewis and Ross Goldingay 94

Loss and Degradation of Red-Crowned Toadlet in the Region Karen Thumm and Michael Mahony 99

Status of Temperate Riverine Frogs in South-Eastern Australian Graeme Gillespie and Harry Hines 109

3 Potential Impacts of Introduced Fish and Fish Translocations on Australian Graeme Gillespie and Jean-Marc Hero 131

Population Declines and Range Contraction in Australian Alpine Frogs William Osborne, David Hunter and Greg Hollis 145

Implementation of a Population Augmentation Project for Remnant Populations of the Southern ( corroboree) David Hunter,Will Osborne, Gerry Marantelli and Ken Green 158

Husbandry; Science or Art? Are Captive Technologies Ready to Contribute to Recovery Processes for Australian Frogs? Gerry Marantelli 168

Conservation Status of Frogs in Western Australian Dale Roberts, Simon Conroy and Kim Williams 177

Toxicological Issues for Amphibians in Australia Reinier Mann and Joseph Bidwell 185

Declines and Disappearances of Frogs: Risk Assessment and Contingency Strategies Michael Mahony, John Clulow, Robert Browne, and Melissa Pomering 202

Community Involvement in threatened frog surveys, monitoring and recovery in Australia Harold Ehmann 212

Applications of Assisted Reproductive Technologies (ART) to Endangered Anuran Amphibians John Clulow, Michael Mahony, Robert Browne, Melissa Pomering and Andrew Clark 219

Appendices

Appendix 1: Checklist of Australian Frogs ‘The Action Plan For Australian Frogs’ 226 Appendix 2: Threatened Frogs: Endangered Species Protection Act 1992 229 Appendix 3: Threatened Frogs: Australian and Environment and Conservation Council list of threatened Australian fauna 230 Appendix 4: Recommendations from ‘The Action Plan For Australian Frogs’ 231 Appendix 5: The of Australian Frogs,‘The Action Plan for Australian Frogs’ 233

4 Foreword

Experts continue to reaffirm that a significant proportion of Australia formed a National Threatened Frogs Working the world’s frog fauna is in decline. Group in August 1997 and held a national workshop on the issue in November 1997. This collection of papers identifies While frogs have been great survivors for many millions of many of the issues confronting our amphibian biologists, years, habitat loss and degradation are imposing themselves managers and enthusiasts. I commend it to the reader. at increasing levels. More alarmingly frogs in pristine environments are being affected. Even in protected rainforest frogs have suffered a mysterious fate and despite extensive searching, some frog species can no longer be located in the wild.

The state of our frogs, who breathe and absorb water through their skin, and their relevance as potential early warning signals for water and air borne pollutants makes Stephen Hunter these signs especially important. Head of Biodiversity Group Enviroment Australia The mystery of our disappearing frogs remains to be solved. While a host of theories on the phenomenon have been elucidated and keenly debated, investigations continue on the many pressures and threats affecting frogs in Australia. Perhaps some day soon we will be in a position to reverse the negative trends we have witnessed in recent times.

5 Preface

With over 200 species, Australia has one of the most diverse The purpose of the Workshop was to: frog assemblages in the world. For many Australian frog 1. provide an overview of the current status of knowledge about species however the prognosis is grim. Dramatic population species which are declining; declines in some Australian frog species have been reported since the 1980s, some of the more serious crashes occurring 2. provide an overview of what’s known about the decline and in pristine habitats. Frustratingly the causal factors for many how it is being tackled; declines remain elusive. 3. identify gaps in knowledge;

In April 1997, Environment Australia published the Action 4. recommend priorities for national effort for research, Plan for Australian Frogs.This plan identified 27 Australian management and community attention; frog species at threat and a further 14 species that may be of 5. help set future priority actions for the Working Group; and concern but which were poorly understood. Recovery 6. establish on-going communication links between all outlines, identifying those research and management actions participants and the wider community. required, were presented for the 27 species believed to be at most threat. Following the Workshop the Working Group met and emphasised: the need to maintain a broad perspective and to In May 1997, the National Threatened Frog Working Group continue monitoring and research on a range of possible recognised a need to bring together specialists in research, causal factors; the need to support research into new areas management and policy making to discuss their understanding — particularly disease as a potentially major factor influencing of the continuing declines to our frog fauna and to prioritise global declines; the need to support strategic research and to future action for addressing the issue. As a result a two day develop protocols to address this issue; and the need to ‘National Threatened Frog Workshop’ was held at the support research on fluctuating asymmetry as a possible early University of Canberra in November 1997 sponsored by warning tool for predicting declines.They also requested that Environment Australia, the NSW National Parks and Wildlife the workshop proceedings be published; that a brochure on Service, the University of Canberra and the Worldwide Fund declining frogs be prepared; that a web site to provide up to for Nature.The Workshop brought together some 80 people date information about declining frog issues be established from throughout Australia and overseas involved in frog and that a national but restricted internet discussion group research, management and policy development and included on frogs should be established. representatives from government agencies, educational institutions and non-government organisations.

6 I am pleased to observe that the majority of these initiatives, I believe this symposium makes an extremely valuable along with the implementation of the recommendations of contribution to our knowledge of our threatened frog fauna. the Action Plan for Australian Frogs, are in hand and While the recovery of threatened frogs and the search for continue to be supported by a wide range of sponsor our missing frog species continues to be highly problematic it organisations. In particular, a number of important projects remains a work of the highest priority. are supported through the Commonwealth Government’s Natural Heritage Trust.

The agreement by Environment Australia to publish the Alastair Campbell proceedings of that workshop has led to the development of Environment Australia this set of 20 papers. It includes regional overviews of the status of threatened frog species, summaries of current research efforts, and much discussion on the technical tools and priorities for action.The papers presented here represent the dedicated work of some 34 authors, many of whom have observed drastic population crashes first hand. I would like to express my thanks for their efforts and patience.

I would also like to thank the following 33 referees for kindly agreeing to review the papers presented here — Will Osborne,Tim Halliday, Ray Nias, Marg Davies, Steve Richards, Arthur White, Grahame Pyke, Dale Roberts, Graeme Watson, Murray Littlejohn, Michael Mahony, Grant Wardell- Johnson, Andrew Burbidge, Gordon Grigg, Bruce Waldman, Mike Tyler, Roy Swain, Harry Hines, Aurel Moise, Bill Buttermeir, Bruce Male, Geoff Larmour, Chris Banks, Hal Cogger,Tony Robinson, Stan Orchard, Gerry Marantelli, Harald Ehmann, Don Driscoll, Keith McDonald, Ann Jelinek, Simon Conroy, and Michael Scroggie.

7 8 The Gordian Knots of the International Declining Amphibian Populations Task Force (DAPTF)

Stan A. Orchard*

ABSTRACT are many sound reasons for the logical sequence of having scientific investigation precede conservation I am not a member of the international Declining action, but the DAPTF must not lose sight of the Amphibian Populations Task Force (DAPTF), need for real success on the conservation side of its but I am nevertheless a very keen observer of its mission. Should conservation be neglected at the activities and progress. My interest stems from my expense of a relentless pursuit for more and more participation in Canada’s own well-established and data, the logic of the DAPTF’s very existence could well-regarded task force.Thus, my views are rightfully be brought into question. personal, and I do not speak on behalf of the international DAPTF. INTRODUCTION The international research effort to unravel the mysteries of Amphibian biologists owe a great debt to the declining amphibian populations has had considerable international DAPTF because, against substantial “success” since its formation in 1992 — but how do we measure its success? More scientists are involved in studying cultural resistance, important and historic strides the conservation biology of amphibians than ever before.The have been made in many positive, mostly scientific, communications links between amphibian researchers have directions — but what about the amphibians? Are never been better. Mountains of data are being gathered, analysed and the results published. Public awareness of the they receiving as many benefits as the amphibian issues is improving and public sympathy appears to be biologists?- well not necessarily at this stage.There growing. In spite of all this, there are still very few examples

* Canadian Amphibian and Reptile Conservation Network, 1745 Bank Street,, British Columbia, CANADA,V8R 4V7.

9 of how this newly acquired knowledge and public sentiment The problem of how to popularise the amphibian will be applied to successfully reverse declining population conservation cause and thence foment a general public trends.Thus, while those of us involved in these research and concern for their plight is far from being solved. Undaunted, monitoring efforts perceive success and growing momentum, Halliday and Heyer (1997) recently attempted to summarise from an amphibian survival perspective the future is not what, in their view, are the most compelling reasons why much brighter today than it was in 1992. One of the DAPTF’s people should care enough about amphibians to at least try espoused tenets is to promote means by which declines can to explain, arrest and reverse these horrendous population be halted and reversed, but we may be largely ignoring this declines. Ethical reasons topped their list. Amphibians, they aspect of our work in our rush to make scientific discoveries. say, have a right to exist. People have a moral responsibility to Ultimately, we should be striving to create a “frog-friendlier” help amphibian populations recover since environmental world, and not simply adding more statistically defensible vandalism by people is probably, more often than not, at the documents to the historical record of amphibian demise. root of the problem.Though not religious advocates, they At the heart of the declining amphibian populations nevertheless go on to invoke biblical references wherein it is conundrum is a complex of Gordian Knots that are ecological, said that “humans are a special creation of God and are given psychological, sociological, political, logistical, and philosophical, dominion over the rest of living organisms on Earth”.They and it may not be humanly possible to unravel them all. argue that this biblical passage should be interpreted by us to mean that our interrelations with other organisms should be “a It is now so well-established that the amphibian populations stewardship, not a slaughter.” However, this moral position is crisis is real that I will not explore its history in any detail certainly not confined to Judeo-Christianity, it is a widely held here.The phenomenon has been documented on several tenet of many of the world’s philosophies. Unfortunately, it is continents, most disturbingly in what appear to be remote, almost universally ignored, at least by the most environmentally pristine habitats, and the problem is being actively if not reckless and politically influential segments of each culture. intensively investigated by over 1200 scientists organised into 90 working groups worldwide (Halliday and Heyer 1997). Incidentally, what Halliday and Heyer failed to point out is that Though mobilised and motivated, the DAPTF will only if you accept that “humans are a special creation of God” succeed if it can secure the tools to carry out and complete then logically so are all other species.Thus, to destroy a its mission. Much hinges on its ability or inability to impress species is to profane a divine creation and to mock the upon a largely apathetic public and their legislative wisdom of God for having created it in the first place. representatives that declining amphibian populations research These species are after all, according to biblical text, some of is important, necessary and environmentally, culturally and the remnants of “Paradise” which God created before He economically beneficial. created people — more food for ecumenical thought.

Religion aside, Halliday and Heyer’s second reason to care WHY CARE ABOUT AMPHIBIANS? about amphibians is because they are interesting organisms. People are intrinsically self-indulgent, and thus most of us who They present the example of the poison dart frogs of South gravitate towards studying amphibians do so because we find and whose habits are complicated, unique, it to be personally rewarding. In other words, people who like and whose extinction should therefore be perceived as a and study amphibians are usually not conscripted but rather great loss to humanity. On the surface, this is a reasonable seem to generate spontaneously.There is, in fact, a very rationale for preventing extinctions but, on the other hand, substantial cross-cultural resistance to these eccentric people how many of the 4,500 or so species of living amphibians are with their curious predilections. It is probably fair to say that, as “interesting” as poison dart frogs? How will this help us historically, there has been little glory or professional save all those frogs and who are insufficiently advantage to specialising in amphibian biology. Fortunately, “interesting” to the general public? Interesting species have history also seems to prove that there will always be at least a the distinction of also being the most desired commodities in small group of dedicated, but marginalised, biologists who care the commercial amphibian pet trade. Ultimately, there may be about the technical details and conservation of amphibians, a real battle to convince some modern, urban youth that any but for reasons that may never be fully comprehensible real frog can be as interesting as a digital or virtual frog. except to others within their own subculture. Getting down to business, the third point that Halliday and I am sure that we all recognise that to the great mass of Heyer raise in defence of amphibians is the mercenary humanity amphibian survival has little meaning. In ecological interest we should all have in preserving nature. For example, terms, amphibian specialists properly view the current pharmaceutically important “amphibians may provide direct situation as being catastrophic. Amphibians are low on the benefit to humans.” As in the previous argument, however, food chain and, when abundant, are extremely important in problems arise when one extends this line of reasoning. It is energy flow, particularly through wetland, damp forest and not easy to see how such discoveries will ever translate into aquatic ecosystems.There are analogous situations, from the protection of wild populations. Once a useful drug is past, where dramatic population declines in other vertebrate discovered you may well be able to either synthesise it or to groups have been reversed through human intervention, but farm the species under laboratory conditions.Thus, in this the greatest successes are inevitably tied to those species that scenario, it becomes important to preserve amphibians only have broad and perennial public appeal, or are easily because they may have some pharmaceutical benefit, but if a convertible to cash on either the open or black markets. given species is thoroughly investigated and found to have no Amphibians are a hard sell, and may always be a hard sell. medicinal value — then what? It is nevertheless the DAPTF’s mission to untangle these psycho-socio-political knots — but how?

10 The last reason on Halliday and Heyer’s list is possibly their products.Virulent pathogens are being found that may or most compelling, or at least marketable. It is the proposition may not have been introduced with non-native game fish, and that due to their complicated semi-aquatic life-styles, assorted environmental stressors are well known to suppress amphibians may be especially vulnerable to the accumulation the immune system and thereby induce . Perhaps of environmental stresses.They say,“amphibians are important amphibians are losing a natural ability to ward off infections indicators of general environmental health” and thus they from common microorganisms that they would normally not “may provide an early warning about deteriorating be susceptible to.The ozone layer has been seriously environments”. After all, their environments are our compromised, especially over Australia, and while early environments — contaminated or not, we ingest the same experimental results on the harmful effects of UV-B on water and breathe the same air. Amphibians could, therefore, amphibian embryos have not yet been replicated, UV-B still be a natural early-warning system that alerts people to a remains an interesting candidate. A 20 year study in northern growing lethal contamination while its concentration is still Ontario, Canada has connected acidification of lakes to sub-lethal to humans. People may even be moved and increased UV-B exposure by showing that acidity can dissolve comforted by the notion that noble frogs are sacrificing free-floating organics that would otherwise effectively screen themselves to save human lives. Accordingly, it is worth UV-B radiation. Studies are underway into hormone mimics stressing the physiological similarities between amphibians and and hormone blockers that can disrupt the endocrine system other vertebrates, including people. seriously enough to cause sex-reversal.While it is generally agreed that global warming is a fact, the effects that this will One thing that Halliday and Heyer did not discuss explicitly ultimately have, or may already be having, on amphibians were the concepts of “shame” and “guilt”.These are the time- remains mostly speculative. Another fertile area of scientific honoured psychological forces that organised religions have inquiry is how to contend with burgeoning populations of used to such great effect in controlling our baser motives. I released, predatory exotics — such as the think that many people are developing a stronger personal (Rana catesbeiana) and the (Bufo marinus). Can environmental ethic because they have been shamed into it. these be eradicated, and, if so, should they be? Finally, will For example, they have been forced to know about and look wetlands and waterways ever be managed for amphibians with at the senseless brutality of whaling and it is now political the zeal that they are today for predatory fish and waterfowl? suicide, in most parts of the world, to advocate a commercial whale hunt. It is my firm impression that, at least in Canada, Quite recently, malformations in amphibian populations younger generations are beginning to grasp the magnitude of suddenly became a hot topic.This is particularly true in the environmental damage that has been wrought within the where a minor media sensation occurred past century.They find it to be shameful, even if they are not when grossly deformed leopard frogs were found in able to fully grasp the data that documents each crisis or the unprecedented numbers by school children in the State of subtleties of each scientific argument. Minnesota. Parasites and contaminants have been implicated but the basis of many amphibian malformations still remain All things considered, amphibians are harmless, defenceless, mysterious. It is critical to differentiate between traumatic oddly beautiful, yet ecologically, culturally and economically events and true developmental abnormalities, but there are significant. How can it be acceptable to ignore their plight? very few specialists qualified to make this assessment. Thus groups such as the DAPTF might be more successful in getting the public to support their initiatives if the case for This year (1997) the United States National Aeronautics and amphibians were presented more as a struggle between Space Administration (NASA) became involved in DAP civility and brutality. Civilised people should be as alarmed research when Ron Heyer (Smithsonian Institution) and about the extinction of species, even uninteresting Cynthia Carey (University of Colorado, Boulder) were amphibians, as they would be about a malicious assault on awarded a grant from the NASA Mission to Planet Earth the unique treasures of a museum, art gallery or one’s own Program.The purpose of the project is to determine whether community.We need metaphors and similes and even NASA has any climate change data, gathered by Earth- paintings, poems and songs to help people, with no biological monitoring satellites, that could be useful in trying to account training, to conceptualise the magnitude of the problem and for known, well-documented amphibian declines scattered personalise the loss — the required political support should across the globe. A team of researchers from various parts of follow. Hitherto, the DAPTF has failed to garner the vast the world collaborated on this project and a report was due moral, financial, and political forces that it needs to adequately late in 1998. confront this crisis. In the main, this subject is still treated as a novelty in the popular media, and policy makers and major Increasingly, amphibian researchers are acknowledging the funding agencies remain, for the most part, fitfully interested possibility that they themselves could be unwittingly but so far unmoved. transporting lethal agents and pathogenic organisms (e.g. viruses, viroids, bacteria, ) into amphibian habitats. While marking and measuring the they study, THE SEARCH FOR CAUSES traditionally, researchers and their research subjects are in the Generally speaking, it is not too difficult to find reasons why most intimate contact. For this reason, some researchers now amphibian populations are dwindling. For example, the go to great lengths to minimise the possibility of commercial trade in live frogs and frog meat is huge. Acid contaminating critical study sites.This area of concern has not rain has been implicated in the decline of some species.The received much public discussion and is not mentioned in commonly used fertiliser, ammonium nitrate, can be toxic to recent manuals on amphibian monitoring methodologies. amphibians, but it is only one of myriad batrachicidal However, in the summer of 1998 the DAPTF produced a contaminants including herbicides, , and petroleum fieldwork code of practice that is an important first attempt

11 to fill this gap (DAPTF et al. 1998). Historically, standardised BUREAUCRACY AND RED TAPE methods of field study have presumed that it is harmless to In our universe, red tape and bureaucracy are seemingly as amphibians for people to handle them directly — which may inevitable as entropy. I was recently reminded of how yet prove to be true. In the meantime however, especially commonly people can lose their perspective on a problem when studying vulnerable populations, it would be most when they become totally immersed in the process of prudent to err on the side of caution and adhere to the formulating committee decisions.Take for example the case code of practice as closely as possible. of Dr Jaime Péfaur, Universidad de Los Andes,, whose conservation efforts are being thwarted by, of all HALTING AND REVERSING things, the newly revised criteria for listing a species to the AMPHIBIAN DECLINES World Conservation Union’s (IUCN), Red Data Book of It is routinely argued that the results of research into the endangered species around the globe. Dr Péfaur knows that causes of declining amphibian populations will be applied to species of frogs are disappearing rapidly in his country and he the problem of halting and reversing declining population needs the resources to study this and try to arrest or reverse trends — though earnestly stated, this is seldom the case. the trend. However, the government of Venezuela will not act There are, sadly, very few instances where an intensive, to fund his research unless the species in question are remedial conservation program has been the logical declared “red-listed” by the IUCN, and the IUCN will not list consequence of research results from declining amphibian them until conclusive statistics on population densities are populations studies. Hitherto, the standard approach by available. Alas, Dr Péfaur finds himself trapped in a perfect scientists and wildlife managers in dealing with the declining paradox. He can not meet the IUCN’s demand for statistics amphibian populations phenomenon is to: because his government will not give him the resources and therefore the frogs will continue to disappear — unstudied a. attempt to identify the species and locations where and unprotected. populations are dwindling; b. attempt to identify the underlying cause, or causes, of The most valuable lesson to learn from Dr Péfaur’s population declines; experience is to never permit a doctrine or prevailing policy to take on more importance than the problematical c. establish extensive population monitoring programs; situations that it was designed to prevent. Unpredictable d. revise endangered species lists; problems will arise and it is best to say,“this is the problem e. produce problem analyses and management plans; — how do we fix it?”, rather than abandoning hope because, “this problem is not covered by the policy”. f. launch public education campaigns.

There is nothing inherently wrong with any of these actions, POLITICS AND SCIENCE except that they are often used, for political reasons, as stall If the prevailing political environment is not to our liking, how tactics and can consume a disproportionate amount of time can scientists go about changing it? In the past, we have and money.When faced with a crisis, at some point, the best tended to address our appeals to the established government information that is available must be applied and action must bureaucracies assigned to environmental matters. Usually, be taken or the species may be lost. Consequently, it is these agencies are staffed by fellow biologists who we can sensible to funnel the greatest proportion of available relate to and who tend to be reassuringly sympathetic. resources to the most vital need.This may perforce involve Unfortunately, we are often disappointed when little action is making fateful decisions and even leaving one’s office. taken, but we are content that everything that could have It is supremely important for amphibian survival and for the been done was. In many cases, our expectations and our future of the DAPTF to demonstrate that human assessments are naive, because government bureaucracies intervention can not only explain declining population trends exist to preserve the current political condition.Thus, we must but can successfully reverse them. Some risks may have to be get our issues on the political agenda if we want to effect real run in order to achieve this, and alas someone must shoulder change in public policies.To do this we need to demonstrate the responsibility because in the process honest mistakes may genuine support for our cause from a wide constituency. be made. Examples of conservation recommendations that are commonly ignored include habitat creation and HERPETOLOGICAL EDUCATION reclamation, the eradication of exotics, captive breeding, A novel development to spring from the Third World head-starting tadpoles, and re-introducing species back into Congress of Herpetology, held in August 1997 in Prague, localities within the historical range.The greatest care must be Czech Republic, was the formation of an International taken to act ethically and prudently. Nevertheless, as we now Committee on Environmental Education on Amphibians and know, amphibian populations can crash so suddenly that Reptiles.The round-table discussion was attended by about 40 there is no time for debate. A reasonably comprehensive plan people representing: Hungary, the United States, Italy, Sweden, (e.g.Tyler 1997) should be adopted and the logistical details Slovenia, Israel, Switzerland, Canada, the Netherlands, Croatia, of carrying out its provisions should be agreed upon well in the United Kingdom, New Zealand, Denmark, Poland and advance of a worst case scenario so that time and effort are Australia. It got me thinking about this long-neglected aspect not wasted when the need suddenly arises. of our work and how it relates to the success of the DAPTF.

12 One of the unique problems amphibians face in their struggle it “has legs”.Thus, the external support and media attention to coexist with people is the fact that many people express a that the DAPTF has received over the past six years very phobic reaction to the very sight of them. Oddly, since we likely has at least as much to do with its entertainment and have no statistical benchmarks, herpetologists have only the curiosity value as with altruistic public sentiment. vaguest of notions as to how innate, culturally pervasive, or extreme this attitude is today. Are attitudes changing? — or So what happens some day, maybe soon, when the mysteries can they be changed? As critical as this information is to all of the declining amphibian populations phenomenon are aspects of the DAPTF’s work, we are woefully ignorant about resolved and we are back to where we were prior to the how to convey our message to the general public for formation of the task force — out of the limelight and maximum effect. Consequently, it would be extremely tussling with the mundane problems of amphibian valuable to understand more fully the psychological and conservation. From a conservationist’s perspective, the issues cultural dimensions of herpetological education.The positive will be no less grave and melancholy, but from the public’s ramifications of this line of inquiry would be in: a) soliciting view the glinting aura of the inexplicable will have been public sympathy and political support; b) developing more peeled away. The DAPTF will have to induce its own effective methods in herpetological education in schools; c) metamorphosis, since it will have come to strongly personify improving the effectiveness of conservation actions and not a cause but a catastrophe. It will have to cultivate genuine associated public relations campaigns; d) providing a statistical goodwill and concern from the general public, and for this it benchmark to which future surveys may be compared in will need to demonstrate successes in reversing declining order to monitor shifts in the perceptions of the general population trends through the integration of science, remedial public towards amphibians. Is it possible to popularise a conservation actions and public education and participation. If universal eco-etiquette that includes amphibians and is lasting the DAPTF fails in these arenas, its DAP research discoveries and cross-cultural? If education is the way to achieve this, we may come to little, and the task force and its great potential had better act quickly and it had better work because, in an could tragically go the way of the . already crowded world, the United Nations’ highest current projection is that the human global population will grow from REFERENCES 5.9 billion in 1998 to 11 billion people by 2050 (National DAPTF, (1998) The DAPTF Fieldwork Code of Practice. Geographic Magazine, October 1998:5). Copies of this leaflet are available by contacting John Herpetological education can also inspire some people to try Wilkinson, Biology Department,The Open University, forging careers for themselves as amphibian specialists.This Walton Hall, Milton Keynes, United Kingdom, MK7 6AA has always been a dicey proposition — but never more so ([email protected]). than today. In North America, for example, the traditional Halliday,T.R. and Heyer,W.R., (1997) Are amphibian objective scientific and moral authority of museum curators populations disappearing?: a task force status report and university professors in directing conservation biology 1996-1997.The Boreal Dip Net 2(1):2-6. research is being appropriated by politically hamstrung and Joel, L.S., (1998) Population. National Geographic, October policy-driven government administrators.The DAPTF’s 1998:5. mission relies heavily upon authoritative scientific advocacy Tyler, M.J., (1997) The Action Plan for Australian Frogs.Wildlife and a desire to integrate and infuse more of the details of Australia, Canberra. amphibian conservation biology into the broader culture.This means that there is a growing need for full-time career positions for real amphibian specialists with extensive personal experience and it is in the DAPTF’s long-term interest to state this plainly at every opportunity. It is a strange and cruel irony that as environmental crises mount, the support systems and respect for independent opinion, objective scholarship and academic freedom, at least in some parts of the world, seem to be quietly slipping away.

THE FUTURE Amphibian researchers have been, to some degree, led out of the wilderness by the declining amphibian populations crisis. Their studies have never before received such wide attention, and as a group they have never been so fully roused, united, and justifiably alarmed. On the other hand, the spotlight that has shone upon the declining amphibian population issue may be difficult to sustain. At this moment, in public relations terms, the DAPTF’s greatest asset is the element of mystery that still surrounds its mission. People are always fascinated by the mysterious and always inspired by the challenge of solving a riddle. Factor into this the possibility of an ecological calamity and you have a combination of irresistible forces. In media terms, this story is still “sexy”- or at least, malformations aside,

13 A review of Declining Frogs in Northern Queensland

Keith McDonald1 and Ross Alford2

ABSTRACT University. Recovery Plans have been compiled but have not been signed off by Government. Future Eleven frog species have declined or disappeared efforts need to incorporate experimental ecology, since 1985 and 1989 in the Central Queensland management of water usage and wider community- Coast and Wet Tropics Biogeographic Regions based participation in monitoring and surveys. respectively of northern Queensland. Declining species are in upland areas and declines occurred INTRODUCTION suddenly for highly susceptible species with close There has been a plethora of scientific and popular articles associations with streams. Despite considerable on declines of frog populations in northern Queensland since effort the causal factor(s) have not been determined the 1980’s (e.g.Winter and McDonald 1986; McDonald 1990; although several hypotheses have been proposed. Tyler 1991; Richards et al. 1993; Ingram and McDonald 1993; Dennis and Mahony 1994; McNellie and Hero 1994;Trenerry The taxa of concern are phylogenetically diverse and et al. 1994; Martin and McDonald 1995,1996, Laurance et al. exhibit a variety of morphologies, reproduction, 1996; Laurance 1996a, b; Mahony 1996; Hero 1996; Alford behaviour and microhabitat preferences. and Richards 1997; Alford et al. 1998; Berger et al. 1998). Numerous meetings, discussions, workshops and debates also Species outside rainforests are under anthropogenic have occurred since 1990, in an effort to determine causes of these frog declines in apparently secure areas of National pressures especially in sugar cane expansion areas Park and State Forest, some of which are further protected of the coastal lowlands. in a World Heritage Area. A wealth of unpublished reports and recovery plans have been devoted to the search for a Current activities are focused on population solution to this problem. A benefit of these studies has been monitoring by community groups, the Department a vast increase in our biological knowledge of stream frogs in north Queensland. of Environment and Heritage and James Cook

1 Conservation Strategy Branch, Queensland Department of Environment, PO Box 834 Atherton 4883, Queensland. 2 Department of Zoology and Tropical Ecology and Cooperative Research Centre for Tropical Rainforest Ecology and Management, James Cook University,Townsville 4811, Queensland.

14 TABLE 1: Frogs numbers in the Wet Tropics and Central FIGURE 1: Wet Tropics and Central Queensland Coast Queensland Coast Biogeographic Regions including those shared. Biogeographic Regions showing rainforest areas.

Species Total species W T 31 50 CMC 10 29 Shared other bioregions 19 TOTAL 60

TABLE 2: Endemic, rare and threatened and declining frogs numbers in the Wet Tropics and Central Queensland Coast Biogeographic Regions.

WT CMC Endemic 23 (46%) 3 (10.3%) Endangered 72 Rare 13 (2) 2(1) Declined 8 (7 End, 1 Rare) 3 (2 End. 1 Com.)

Many hypotheses concerning possible causes of declines have been advanced, but little information has been presented or published yet that allows any of these hypotheses to be tested. Identifying the cause or causes of the declines seems to be little closer now than a decade ago. Identification of the causal factor or factors is necessary before they can be managed or mitigated. Much of the research and monitoring carried out to date has been entirely hypothesis-free and exploratory (e.g., Richards et al. 1993; Speare 1995), has attempted to generate hypotheses but not to test them A total of 60 frog species (50% of Queensland frogs) are (e.g., Laurance et al. 1996), or has attempted to test found in these biogeographic regions; 19 of these species also hypotheses using data that are too sparse to provide rigorous occur in other regions. Fifty species have been recorded from tests (e.g., Dennis and Mahony 1994; Laurance 1996 a,b). the WT and 29 from the CQC (Table 1).

We argue that what is most needed at present is research The biogeographic regions have endemic species reflecting carefully targeted at testing and potentially rejecting hypotheses. the influence of rainforest habitats (Table 2). Nearly 50% (23) The rigorous rejection of some hypotheses will allow further of the WT frogs are endemic.This is reflected in the research and monitoring efforts to be concentrated on refining conservation status of WT species with high numbers of rare, and testing those hypotheses that remain. geographically-localised species — a consequence of the interactive influences of the rainforest habitat and topography, This paper reviews population declines in northern climate and history. Many species are restricted to upland and Queensland and proposes future directions for research and mountain-top habitats. Only two of the WT species listed as monitoring. Much of the information presented is derived rare under the regulations to the Queensland Nature from a current research program of the Queensland Conservation Act 1992 are shared with other areas. All Department of Environment and Heritage, the Department endangered species which have had population declines are of Zoology and Tropical Ecology at James Cook University endemic to the WT and CQC biogeographic areas. and the Cooperative Research Centre for Tropical Rainforest Ecology and Management, and independent research funded The declining species are representatives of the two most by the Wet Tropics Management Authority. diverse families in Australia; and . No declines have been reported for the other native frog families (the and Ranidae) within the areas. STUDY AREAS AND SPECIES COMPOSITION The eleven species which have experienced declines are from five genera (Table 3). Some have declined in parts of their This review covers rainforest areas in the Central distribution, while others have not been seen for some time Queensland Coast (CQC),Wet Tropics (WT), and Cape York (Table 3). Several declining species have very localised Peninsula (CYP) biogeographic regions of northern distributions while others were widespread in suitable habitat Queensland (Figure 1). Population declines of rainforest throughout the rainforest blocks within the biogeographic stream-dwelling frogs have occurred in two of these regions (Table 3). biogeographic regions (CQC,WT).These areas are characterised by high rainfall, uplands of granite with steep topography, and moist closed forests or tall open forests and woodlands (Stanton and Morgan 1977).

15 TABLE 3: Declining species in the biogeographic regions, conservation status, distribution within rainforest blocks and decline period. NCA Regs. cons. stat. — conservation status in Queensland Nature Conservation Act (Wildlife Regulations) 1994.;Action Plan Cons. Stat. — conservation status in the Action Plan for Australian Frogs (Tyler 1997).

Genus Species Bio. Reg. Declines Action Plan NCA Regs. Distribution* Decline Cons.Status Cons.Status period Litoria lorica WT Y E E 1 1991 Litoria nannotis WT Y E E 3 1989–1993 Litoria nyakalensis WT Y E E 2 1990 Litoria rheocola WT Y E E 2 1989–1994 Litoria genimaculata WT+ Y R 3 1990 –1994 dayi WT Y E E 3 1989–1994 Taudactylus acutirostris WT Y E E 2 1989–1994 Taudactylus rheophilus WT Y E E 1 1989–1991 Adelotus brevis CMC + Y C 2 1985–86 Rheobatrachus vitellinus CMC Y E E 1 Jan 1985–Mar 85 Taudactylus eungellensis CMC Y E E 2 1985–86

* % of latitudinal spread in biogeographic rainforest block 1. < 20% 2. 20–60% 3. 60–100%

REGIONAL ASSESSMENTS OF Wet Tropics DECLINING FROGS Eight species in the WT have experienced population Declines occurred in the CQC during 1985-86 and in the declines (Table 3; Richards et al. 1993; Laurance et al. 1996; WT between early 1989 and early 1994. Martin and McDonald 1996; McDonald unpubl.).Three species (Litoria nannotis, L. rheocola and Nyctimystes dayi) Central Queensland Coast have stable populations at lower altitudes (below approximately 400 m). In the CQC, two species, Rheobatrachus vitellinus and Taudactylus eungellensis have declined precipitously (McDonald Known declines in the WT occurred between 1988 and late 1990). Another species, T. liemi, appears to be secure, although 1994 (Table 3; Richards et al. 1993; Laurance et al. 1996). other Taudactylus species have suffered declines. Taudactylus acutirostris, L. lorica and L. nyakalensis have not been observed in substantial numbers since late 1994 Declines of T. eungellensis commenced in 1985 and the (Ingram and McDonald 1993, McDonald, unpubl.) with species was last seen in monitored sites in 1986 (McDonald the most recent reports being of a single individual of 1990). An individual was observed in 1992 (Couper 1992) T. acutirostris in Slaty Creek in 1994 (Roberts pers. comm.), a and small populations have been located since November sub-adult at Big Tableland in January 1995 (McDonald 1993 in the area around Mt Dalrymple (McNellie and Hero unpubl.), and a report of the call of a single frog near Millaa 1994; Retallick pers. comm.). Populations have not been Millaa. Small populations in two creeks of T. rheophilus has located in the southern and northern areas of the known recently been relocated on the CarbineTableland and the distribution ( McNellie and Hero 1994; McDonald 1990; Bellenden Ker Range ( Marshall 1998). Retallick pers. comm.). In each case where population crashes affected monitored The small populations located since 1993 are not as large or populations, tadpoles survived and metamorphosed after the as extensive as those found in the 1970’s and 1980’s and only adult population had crashed, indicating that the causal factor one population is nearing the numbers found at that time did not affect this stage of the frog life cycle (Richards et al. (Retallick pers. comm.).The three largest populations are 1993; Dennis and Mahony 1994; Laurance 1996 a,b; Laurance found in one catchment. Other populations are smaller and et al. 1996; Martin and McDonald 1996).Tadpoles removed appear to be stable with most of the existing populations at from declining populations successfully metamorphosed, but lower altitudes. Studies conducted by Richard Retallick all individuals died before attaining adult size (Dennis and indicate that the sizes of monitored populations are slowly Mahony 1994). It is possible that post-metamorphic mortality increasing but it is not known if small populations in other was more rapid in individuals having close contact with catchments are increasing or if the ranges of populations stream water, but this conclusion is not strongly supported by are expanding. statistical analysis (Dennis and Mahony 1994). Rheobatrachus vitellinus has not been observed since At a monitored site at Big Tableland (15° 42’S 145° 16’ E, March 1985, despite periodic intensive and extensive 620 m) four species of frog declined suddenly in late 1993 searches (McDonald, unpubl.;Retallick, pers. comm.; McNellie (Figure 2).The population of Litoria genimaculata at the site and Hero 1994). decreased at the same time, but did not disappear entirely. Similar patterns have been reported for L. pearsoniana, Adelotus brevis, and L. lesueurii elsewhere in Queensland; there have been no total disappearances but populations have

16 FIGURE 2: Total frog population declines (Litoria nannotis, L. rheocola, Nyctimystes dayi and Taudactylus acutirostris) at Big Tableland, northern Wet Tropics Biogeographic Region.

decreased in size (Ingram and McDonald 1993).The species NATURAL HISTORY/ECOLOGY which did not totally disappear were those having the largest Declining frogs in north Queensland breed in rainforest number of pigmented eggs of all the declining species. streams in upland areas (Table 4). However they differ widely At O’Keefe Creek, Big Tableland (400 m) species declined in taxonomic affinity, morphology, microhabitat preferences, simultaneously with those at the primary monitoring site. breeding strategies and behaviour. These sites have been monitored every 4-6 weeks since mid 1992. Since June 1995 L. rheocola and N. dayi (but not Habitat T. acutirostris and L. nannotis) have occasionally reappeared The declining frog species have a range of microhabitat near the 400 m site, but have not established resident preferences. In earlier reports (e.g., Richards et al. 1993) we populations.This site is immediately above a steep referred to all species breeding in streams as stream-dwelling escarpment that forms a disjunction between upland and frogs. It is now clear that frogs which breed in streams differ lowland populations.The lowland populations still persist. in their fidelity to them.Those with the tightest association No colonisation of areas above the O’Keefe Creek site may be the most likely to disappear or decline (Dennis and (400 m) has occurred, which may indicate that the factor Mahony 1994;Table 4 ). Rheobatrachus vitellinus,Taudactylus which caused the declines is still operating. acutirostris,T. eungellensis and T. rheophilus and members of In December 1997 two N. dayi were heard calling below the the Litoria nannotis group spend a greater proportion of their primary monitoring site (620 m) indicating some individuals time in or immediately adjacent to streams than do may be moving upstream although tadpoles and eggs have not L. genimaculata and Adelotus brevis (Table 4). been located. N. dayi activity along streams is very seasonal (McDonald unpubl.) and appears to spend more time away Daily active period from streams than do the other species (McDonald unpubl.). Most declining species are nocturnal but several species are diurnal and there appears to be no link between activity Cape York Peninsula periods and declines. Taudactylus eungellensis and T. acutirostris No declines occurred among the Cape York rainforest stream are diurnal while T. rheophilus is diurnal but can be active frogs during 1993-1998 (McDonald unpubl.). Litoria longirostris, during the evening. All of the other species are nocturnal L. eucnemis and Rana daemeli populations monitored yearly in although during heavily overcast and rainy conditions some upland areas of the McIlwraith Range between 1993 and the individuals become active and call during the day. present have remained stable. L. eucnemis and R. daemeli have considerable latitudinal and altitudinal ranges and both species Morphology occur in New Guinea. L. longirostris is endemic to, and patchily The declining frog species exhibit a range of morphologies.The distributed in, rainforest streams above 400 m in the species vary in size: the snout-vent length (SVL) of the smallest, McIlwraith Range (McDonald and Davies unpubl.). T. acutirostris, ranges from 18.8–25.5 mm in males and 22.7-31.2 mm in females and the SVL of the largest, L. genimaculata, ranges in size from 32-54.4 mm (males) and 42.6–86.6 mm (females). All species display a degree of in

17 size, a characteristic which is most striking in L. genimaculata and ratio for males of 0.40:1 and for females of 0.53:1. No data N. dayi (Table 5). Adelotus brevis shows a reverse dimorphism, are available on weights for A. brevis, L. lorica, L. nyakalensis, with males larger than females (Moore 1961). and T. rheophilus. Weights reflect general appearance of the frogs. T. acutirostris Eggs is the lightest and the heaviest is R. vitellinus (Table 5). R. vitellinus is the most robust species with a weight to SVL Most declining species, except for L. genimaculata,have ratio (g:cm) for males of 3.47:1 and for females of 4.81:1. relatively small numbers of large eggs (Table 6). Eight species T. acutirostris is the most gracile species with a weight: SVL have unpigmented eggs while those of T. acutirostris, A. brevis and L. genimaculata are pigmented. Known egg deposition strategies include: foam nests in pools (A. brevis); clumped TABLE 4: Stream fidelity of declining and non-declining rainforest frogs. egg masses in clear jelly in pools (L. genimaculata); on or under rocks in streams for (T. acutirostris , L. rheocola, and Species Stream-assoc. Decline category T. eungellensis); under rocks in streams (N. dayi); under crevices on waterfalls (L. nannotis).The egg deposition site Adelotus brevis 3b 2 of R. vitellinus is unknown but the stomach certainly carries Litoria chloris 3b 1 L. eucnemis 2b 1 tadpoles (McDonald unpubl.). Egg deposition sites for L. genimaculata 2b 2 T. rheophilus, L. nyakalensis and L. lorica are unknown. L. lesueuri 2b 1 L. longirostris 1b 1 Tadpoles L. lorica 1b 3 L. nannotis 1b 3 Tadpoles of the declining species are highly variable (Table 6). L. nyakalensis 1b 3 Definition of tadpole forms follow Altig and Johnston 1989. L. rheocola 1b 3 L. xanthomera 3b 1 The tadpoles of R. vitellinus do not have keratinised mouth Mixophyes schevilli 3a 1 parts, are unpigmented and are carried in the stomach of the M. fasciolatus 3b 1 female. Species of the L. nannotis group (except L. lorica which Nyctimystes dayi 2b 3 is unknown) and N. dayi have lotic suctorial tadpoles with Rana daemeli 3a 1 large suctorial mouths for holding on to rocks in the swift Rheobatrachus vitellinus 1a 3 Taudactylus acutirostris 2a 3 flowing streams (Davies and Richards 1990; Richards 1992). T. eungellensis 1b 3 Taudactylus acutirostris,T. eungellensis, Adelotus brevis and T. liemi 1b 1 L. genimaculata have lotic-benthic form tadpoles with small T. rheophilus 1b 3 oral discs.The tadpoles of T. rheophilus are unknown.

Summary: At least some tadpoles of each species except Rheobatrachus degree of association with streams vitellinus and A. brevis overwinter in the stream. T. rheophilus obligate tight to loose to and L. lorica are suspected to overwinter, while overwintering moderate very loose has been documented in the other seven species (Richards No changes 2 2 5 et al. 1993; Richards 1992; Retallick pers comm.). Declines 0 1 1 Disappearances 7 2 0 NON-DECLINING FROGS Fisher's Exact test, 2-tailed P = 0.019 Within the rainforests and their margins of the CQC and WT Degree of association with streams: there is no indication of declines of populations of Litoria 1 a. obligate, aquatic.The species is found only in the water lesueurii, a species which has declined in the rainforests of b. obligate, non-aquatic.The species is reliably found within south-east Queensland (Ingram and McDonald 1993).The the stream banks over the entire year widespread WT species, Litoria xanthomera and Mixophyes 2 a. tight.The species is found within the stream banks over schevilli, have not declined at any site, although their most of the year, but regularly ventures away from the populations have not been intensively and systematically stream monitored. None of the 14 species of litter-dwelling b. moderate.The species is found with some reliability microhylids appear to have declined. However, relatively little within the stream banks over an extended season is known of this group because standardised monitoring has 3 a. loose. Larvae occur predominantly in streams, but adults visit them irregularly been conducted at only one site, Paluma near Townsville. b. very loose. Larvae do not occur predominantly in Presence/absence monitoring for microhylids at monitoring streams, adults may visit them irregularly sites for stream frogs also indicates no loss of species (Dennis 4 none. Neither larvae nor adults occur within stream and Mahony 1994; McDonald unpubl.). banks with any frequency In the CQC at Eungella, non-stream species such as L. chloris Decline categories and M. fasciolatus appear secure. T. liemi has experienced no 1 no declines declines (McDonald pers. obs.; Retallick pers. comm.). 2 some declines, no disappearances 3 at least some disappearances

18 TABLE 5: Snout-vent length and weights of declining frogs in the Wet Tropics and Central Queensland Coast Biogeographic region.

Max SVL Min SVL Species Sex No. (mm) (mm) Mean SVL Std Dev Source Litoria genimaculata Male 1021 54.4 32 42.93 3.04 McDonald and Martin unpublished Female 102 86.6 53 68.75 6.06 L. lorica Male 7 33.1 29.6 Davies and McDonald 1979 Female 3 37.3 32.9 L. rheocola Male 756 36.4 27 31.52 1.56 McDonald and Martin unpublished Female 68 41.2 28.4 36.27 2.54 L. nannotis Male 99 51.8 41.2 47.04 2.09 McDonald and Martin unpublished Female 41 59.1 48.8 53.63 2.27 L. nyakalensis Male 6 32.3 29.7 31.6 2.975 Liem 1974 Female Nyctimystes dayi Male 265 41.9 32.7 37.22 1.65 McDonald and Martin unpublished Female 4 58.8 54 57.2 2.2 Adelotus brevis Male 14 43.5 34.2 Moore 1961 Female 7 38 29.1 Rheobatrachus vitellinus Male 48 58.3 42.6 52.11 4.04 K. McDonald unpublished Female 59 72.2 47 59.78 6.31 Taudactylus acutirostris Male 99 25.5 18.8 20.62 1.48 McDonald and Martin unpublished Female 39 31.2 22.7 24.6 1.62 T. eungellensis Male 490 32 23.8 27.63 1.2 R. Retallick pers comm. Female 640 37.11 25.5 32.33 1.72 T. rheophilus Male 12 27.4 24.3 26.1 Liem and Hosmer 1973 Female 9 30.5 24.1 28.5

Max WT Min WT Species Sex No. (gms) (gms) Mean WT Std Dev Source Litoria genimaculata Male 1021 9.6 2 4.66 1.06 McDonald and Martin unpublished Female 102 38.4 7.5 22.64 6.45 L. rheocola Male 756 3.3 1.2 2 0.27 McDonald and Martin unpublished Female 68 4.55 1.8 3.09 0.66 L. nannotis Male 99 12 6 8.4 1.05 McDonald and Martin unpublished Female 41 17 8.2 11.89 2.17 Nyctimystes dayi Male 265 3.9 1.9 2.56 0.3 McDonald and Martin unpublished Female 4 11 8.4 10.23 1.24 Rheobatrachus vitellinus Male 48 26.5 11 18.13 3.8 K. McDonald unpublished Female 59 46.5 13 28.78 8.91 Taudactylus acutirostris Male 99 1.6 0.6 0.83 0.19 McDonald and Martin unpublished Female 39 2.5 0.9 1.3 0.27 T. eungellensis Male 489 3.7 0.8 2.32 0.39 R. Retallick pers comm. Female 640 5 1.9 3.48 0.57

POSSIBLE CAUSES OF DECLINES expected ranges with no periods of prolonged . No declines were associated with periods of low rainfall Although numerous possible causes have been proposed to (Laurance 1996 a,b). Attempts to correlate environmental explain declines of frog populations elsewhere, few of these changes with declines depend strongly on information about have relevance to north Queensland populations. when the environment affected populations; recent evidence Habitat has remained intact with no clearing or logging of (Alford et al. 1997) suggests that environmental effects preceded forests in the Wet Tropics World Heritage Area since 1988. some declines by 1–2 years, so many of these analyses may need Species were present even though logging and mining of the to be repeated. However, data from other populations rainforest had occurred in the past (McDonald 1992). (SVL/weight ratios; McDonald 1990) failed to indicate any decline However, this does not rule out more subtle, unmeasured in health of monitored populations before declines occurred. environmental changes affecting declining frogs. There has been repeated speculation that UV-B light has No obvious environmental changes were detected during stream caused declines. However this possibility appears to be monitoring at the times when declines occurred (Richards et al. unlikely in north Queensland. All declining species in this area 1993; McDonald unpubl.).Water characteristics including pH, are found in rainforest with dense canopy cover. Several are conductivity, water temperature, concentrations of metal ions and active only nocturnally, several lay eggs under rocks biocides, and dissolved oxygen measured at a large number of (exceptions are L. genimaculata and A. brevis), tadpoles are still sites have not significantly differed between locations which had present when adults disappear (most UV-B hypotheses declines and those that did not . Rainfall has remained within concern effects on aquatic stages, not adults; Blaustein et al. 1996). Declines have been rapid, occurring over 2–3 month

19 TABLE 6: Egg and tadpole characteristics of declining frogs in the Wet Tropics and Central Queensland Coast Biogeographic Regions.Terms for tadpole forms follow Altig and Johnston 1989.

Species Egg size Egg no. Pigmentation Tadpole Reference Taudactylus acutirostris 2.2–2.7 25–40 Lotic benthic Liem & Hosmer 1973 Yes McDonald & Richards unpublished T. rheophilus 1.8–2.4 35–50 Lotic benthic? Liem & Hosmer 1973 Litoria nannotis 2.7–3.4 No Lotic suctorial Liem 1974 1.98–2.93 136–216 Hero & Fickling 1996 L.lorica No Lotic suctorial? Davies & McDonald 1979 L. rheocola 1.4–1.8 No Lotic suctorial Liem 1974 2.4–2.6 46–63 Hero & Fickling 1996 L.nyakalensis No Lotic suctorial Richards 1992 L. genimaculata 2.1–2.56 843 Yes Lotic benthic Davies 1989 Nyctimystes dayi 2.3–2.6 107 No Lotic suctorial Davies & Richards 1990 Rheobatrachus. vitellinus 22 No In stomach McDonald & Tyler (1984) T. eungellensis 2.2–2.6 30–50 No Lotic benthic McDonald pers. obs., Retallick pers comm., Liem & Hosmer 1973. Adelotus brevis 1.7–1.8 270 Yes Lotic benthic/ Tyler 1994 lentic benthic L. eucnemis 1000+ Yes Lotic benthic McDonald unpublished L. longirostris 2.0–2.4 28–60 Yes Lotic benthic McDonald unpublished, McDonald & Storch 1994

periods (Figure 2). All of these factors make it unlikely that some cases (Alford unpubl.). Many coastal lowland habitats such as declines have resulted from direct effects of UV-B on swamps and forests are important for the maintenance populations. Most importantly it is well established that there of frog populations. Species such as Notaden melanoscaphus which have been no significant changes in stratospheric ozone in the has isolated populations in the coastal areas near Townsville are tropical areas which can be linked with UV-B increases as disappearing at rates that are alarming because of loss of habitat experienced in higher latitudes (Madronich and de Gruijl through real estate expansion. 1993; McPeters et al. 1996; Moise pers. comm.)

A recent hypothesis suggests that a virulent pathogen, possibly CURRENT ACTIONS a virus or Chytrid , has decimated frog populations Monitoring (Laurance et al. 1996; Berger et al. 1998). Circumstantial evidence from observations of rapid declines and location of Monitoring studies have been conducted in the Wet Tropics sick and dying animals provides a strong link to a disease being and Eungella by James Cook University and the Queensland the proximate cause of some declines (Dennis and Mahony Department of Environment and Heritage for medium to 1994; Laurance et al. 1996; Berger et al. 1998). However, there long term periods. are alternative interpretations (Alford and Richards 1997) for all lines of evidence suggesting that a disease might be the sole Wet Tropics cause of declines (Laurance et al. 1996), and recent evidence A transect at Birthday Creek, near Paluma has been (Alford et al. 1997) suggests that the health of adults began to monitored since 1987 and one at Big Tableland since 1992. decline long before populations crashed. Decreases in body Other sites have been visited less frequently at differing condition were not observed in Rheobatrachus vitellinus prior intervals throughout the WT over the last seven years that to declines (McDonald 1990).The disease hypothesis clearly requires further testing and possibly refinement. cover a wide range of latitudinal, altitudinal and historical occurrences of frogs.

ANTHROPOGENIC INFLUENCES Eungella

The research and documentation of declines have focused to date At Eungella monitoring by the Queensland Department of on rainforest species, which do not appear to have suffered direct Environment and Heritage occurred at selected sites during anthropogenic effects. Little systematic survey and monitoring 1980–89 (McDonald 1990); since 1993 these and additional effort has been devoted to species in other habitats. Monitoring sites have been monitored by James Cook University.The around Townsville and in the Wet Tropics open forests indicates no current monitoring is conducted by Richard Retallick with declines other than those related to obvious anthropogenic causes assistance from national park rangers. such as clearing of habitat.There is local loss of habitat around real estate development or in the coastal lowlands through the sugar Public participation in the monitoring program has been in the expansion program in the last five years.The construction of , form of assistance to researchers. Monitoring has also been quarries and road-side ditches may have enhanced breeding conducted by the Cape York Herpetological Society at two habitat for some species, although the effects of these structures locations in the WT for the last 18 months and it is expected on population dynamics are uncertain and may be negative in that this program will expand to cover additional sites.

20 Because the results of Alford et al. (1997) suggest that those in areas on the coastal lowlands which developmental stability analysis may be a useful tool for detecting are under pressure from sugar expansion programs. the effects of environmental changes before they lead to There has been a preoccupation with the rainforest species changes in population sizes, measurements of limb asymmetry but work is urgently required in the lowland open forests. have recently been incorporated into monitoring programs.The Baseline data on the distribution, habitat preferences, techniques used follow those outlined by Alford et al. (1998). breeding and population fluctuations of lowland species is The lengths of the hindlimbs, from heel to the outer end of the urgently needed.This will require surveys and monitoring of tibio-fibula, are measured three times on each side of the , frog populations. using callipers and alternating between sides.The person handling the callipers gives them to a second person, who reads 5. Intensive monitoring of sites designed to address quantitative the length and records it.This gives three independent replicate population assessment. Intensive studies are needed to measurements on each side of the animal, which can be address population changes and recruitment through mark analysed to examine asymmetry, following the methods recapture and monitoring for population stability.This will suggested by Palmer and Strobeck (1986). Preliminary results of enable sites less frequently visited to be placed in the this monitoring indicate that asymmetry levels in populations of context of known population fluctuations determined L. genimaculata are generally low.The technique is discussed in from intensive long-term studies. detail elsewhere in this volume (Alford et al. 1999). ACKNOWLEDGMENTS RECOVERY PLANS Thanks to P.Brooke, J. A. Covacevich, A. Campbell, A. Dennis, Two recovery plans for the Wet Tropics and Eungella have A.B Freeman, M. Hero, D.House, M. Mahony,W.E. Martin, been written (Martin and McDonald 1995, 1996) and A. Moise, S.J. Richards, R. Retallick and volunteers and approved by the Northern Threatened Frogs Recovery Team. students who have assisted sometimes in the most arduous The plans have not yet been adopted by the Federal or State of field conditions. agencies responsible for the environment. Recovery Plans Funding has been provided by our respective institutions, need constant updating as more and better information Environment Australia, the Australian Research Council, the comes to hand.There is a need to respond quickly to Wet Tropics Management Authority, and the CRC for Tropical developments, especially with projects which exploit an event Rainforest Ecology and Management. such as a species die off which occurs quickly (Figure 2). Organisational inertia can be a deterrent to implementing Steve Richards and Dr Marg Davies made constructive actions in these circumstances.This can be addressed by a criticisms of earlier manuscript drafts.This assistance is most yearly review and modification of action plans. A process gratefully acknowledged. should be in place to permit the rapid implementation by State and Federal environment agencies of such modifications. REFERENCES FUTURE DIRECTIONS Alford, R. A., Bradfield, K.S. and Richards, S.J., (1997) Predicting declines in rainforest frog populations. Abstract, CRC- There are many challenges facing frog research and TREM annual conference, Cairns, Queensland. monitoring in the Wet Tropics and Eungella biogeographic regions. Future directions include: Alford, R. A., Bradfield, K.S. and Richards, S.J., (1999) Measuring and analysing developmental instability as a tool for 1. Experimental work — experimentation and manipulation monitoring frog populations. Pp34–43 in Declines and to test the hypotheses which have been advanced Disappearances of Australian Frogs ed by A.Campbell. regarding possible causes of declines. Environment Australia: Canberra. 2. Management of water usage — such as dams and water Alford, R. A., and Richards, S.J., (1997) Lack of evidence for extraction, and examination of impact on remnant frog epidemic disease as an agent in the catastrophic decline of populations. Experimental research and monitoring of impacts Australian rain forest frogs. Conservation Biology of water management regimes is essential as many water 11:1026-1029. extraction processes are proposed in the high rainfall areas of Altig, R. and Johnston, G.F., (1989). Guilds of anuran larvae : Eungella and the Wet tropics, e.g., the proposed Finch Hatton relationships among developmental modes, morphologies, and water extraction in the Johnston River catchment. and habitats. Herpetological Monographs 3 : 81-109. 3. Increased public participation — there is a need to have greater Atkinson, R., (1996) Proc. Conf. on Health Consequences of public participation, especially amongst herpetological clubs Ozone Depletion. Cancer Forum Vol. 20 No. 3 Spec. and conservation groups, because more information can be Ed.168-173. obtained through this process than by a few active scientists. Management staff in Government departments need to Berger, L., Speare, R., Dasak, P., Green, D.E., Cunningham, A. A., become aware of the declines and actively participate in Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., monitoring programs particularly on national parks.Water McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. and resource managers should incorporate potential impacts of Parkes, H., (1998) Chytridiomycosis causes amphibian proposed water extraction and impoundment in any planning. mortality associated with population declines in the 4 Community monitoring of common species — the public can rainforests of Australia and Central America. Proc. Natl. assist in obtaining information on the distribution and current Acad. Sci. USA 95 : 9031-9036. status of species currently thought to be common, especially

21 Blaustein, A.R., Hoffman, P.D., Hokit, D.G., Kiesecker, J.M.,Walls, biogeographic region of northeast Queensland. (submitted S.C. and Hays, J.B., (1996) UV repair and resistance to to Environment Australia, Canberra). solar UV_B in amphibian eggs : A link to population Martin,W.E. and McDonald. K.R., (1996) Recovery plan for declines? Proc. Natl. Acad. Sci., USA 91 : 1791-1795. the stream-dwelling rainforest frogs of the Eungella area, Couper, P.J.,(1992) Hope for our missing frogs.Wildlife mid-east Queensland. (submitted to Environment Australia. 29(4): 11-12. Australia, Canberra). Covacevich, J.A. and McDonald, K.R., (1993) Distribution and McDonald, K.R., (1990) Rheobatrachus Liem and Taudactylus conservation of frogs and reptiles of Queensland Straughan and Lee (Anura: Leptodactylidae) in Eungella rainforests. Memoirs of the Queensland Museum National Park, Queensland: distribution and decline.Trans. 34(1):189-199. R. Soc. S. Aust. 114(4): 187-194. Davies, M. and McDonald, K.R., (1979) A new species of McDonald, K.R., (1992) Distribution patterns and stream-dwelling hylid frog from northern Queensland. conservation status of north Queensland rainforest frogs. Transactions of the Royal Society of Conservation Technical Report 1., Queensland 103(7): 169-176. Department of Environment and Heritage, Brisbane. Davies, M. and Richards, S.J. (1990) Developmental Biology of McDonald, K.R., Covacevich, J.A., Ingram, G.J. and Couper, P.J., the Australian Hylid Frog Nyctimystes dayi (Gunther). (1991) The status of frogs and reptiles. Pp 338-345 In Trans. Roy. Soc. Sth Aust. 114: 207-212. Ingram, G.J and Raven, R.J. (eds),‘An atlas of Queensland’s Dennis, A., (1982) A brief study of the Sharp-snouted Torrent frogs, reptiles, birds and mammals’. (Queensland Museum, Frogs Taudactylus acutirostris. North Queensland Naturalist Board of Trustees: Brisbane). 391pp. 50: 7-8. McDonald, K. R., and Tyler, M. J., (1984) Evidence of gastric Dennis, A. and M. Mahony, M., (1994) Experimental brooding in the Australian leptodactylid frog Rheobatrachus translocation of the endangered sharp-snouted day frog vitellinus.Trans. R. Soc. S. Aust. 108: 226. Taudactylus acutirostris and observations of the cause of McNellie, M. and Hero, J.M., (1994) Mission amphibian.The declines among montane riparian frogs. Unpublished report search for the missing rainforest frogs of Eungella.Wildlife prepared for Wet Tropics Management Authority, Cairns. Australia 31(4): 22-23. Hero M-J., (1996) Where are Queensland’s missing frogs? McPeters, R.D., Hollandsworth, S.M., Flynn, L.E., Herman,J.R., Wildlife Australia 33(3): 8-13. and Seftor. C.J. (1996) Long-term trends derived from the Ingram, G. and McDonald, K.R., (1993) An update on the 16 year combined Nimbus 7 / Meteor 3 TOMS version 7 decline of Queensland’s frogs. Pp 297-303 in Herpetology record. Geophys. Res. Let. 23: 3699-3702. in Australia: a diverse discipline eds D. Lunney and D. Ayers. Moore, J.A. (1961) The Frogs of Eastern . Transactions of the Royal Zoology Society of New South Bull Amer. Mus. Nat. Hist. 121: 149-386. Wales. Mosman. Palmer, A. R, and Strobeck, C., (1986) Fluctuating asymmetry: Laurance,W. F., (1996) Why are Queensland’s frogs croaking? measurement, analysis, patterns. Annual. Review of Ecology Australian Nature 25 (4) : 56-62. and Systematics. 17:391-421 Laurance,W.F., (1996) Catastrophic declines of Australian Richards, S.J., (1992) The tadpole of the Australian frog Litoria rainforest frogs: is unusual weather responsible? Biol. nyakalensis (Anura, Hylidae), and a key to the torrent Conser. 77 : 203-212. tadpoles of northern Queensland. Alytes 10(3): 99-103. Laurance,W.F., McDonald, K.R. and Speare, R., (1996) Richards, S.J., McDonald, K.R., Alford, R.A., (1993) Declines in Epidemic disease and the catastrophic decline of Australian populations of Australia’s endemic tropical rainforest frogs. rainforest frogs. Conserv. Biol. 10(2) : 406-413. Pacific Conservation Biology 1:66-77. Liem, D.S., (1974) A review of the Litoria nannotis species Speare, R., (1995) Preliminary study on diseases in Australian group and a description of a new species of Litoria from Wet Tropics amphibians. Deaths of rainforest frogs at north-east Queensland, Australia. Memoirs of the O’Keefe Creek, Big Tableland. Final Report to Department Queensland Museum. 17(1): 151-168. of Environment and Heritage. Unpublished Report QDEH. Liem, D.S. and Hosmer,W., (1973) Frogs of the Stanton, J.P.and Morgan, M.G., (1977) The rapid selection and Taudactylus with description of two new species (Anura: appraisal of key endangered sites.The Queensland case Leptodactylidae). Memoirs of the Queensland Museum study. Uni. of New England, School of Nat. Res. Report 16(3): 435-457. No. PR 4. Madronich, S. and de Gruijl, F.R., (1993) Skin cancer and UV Trenerry, M. P.,Laurance,W. F., and McDonald, K. R., (1994) radiation. Nature 366 : 23. Further evidence for the precipitous decline of endemic Mahony, M., (1996) The decline of the Green and Golden Bell rainforest frogs in tropical Australia. Pacific Conservation Frog Litoria aurea viewed in the context of declines and Biology 1: 150-153. disappearances of other Australian frogs. Aust. Zool. 30 (2) Tyler, M.J., (1991) Where have all the frogs gone? Aust. Nat. : 237-247. Hist. 23 (8) : 618-625. Marshall, C., (1998) The reappearance of Taudactylus Tyler, M.J., (1997) The Action Plan for Australian Frogs. (Anura:Myobatrachidae) in north Queensland streams. Environment Australia, Canberra. Pacific Cons. Biol. 4: 39-41. Winter, J. and McDonald, K.R., (1986) Eungella, the land of Martin,W.E. and McDonald. K.R., (1995) Recovery plan for cloud. Australian Natural History 22(1): 39-43. the stream-dwelling rainforest frogs of the Wet Tropics

22 Chytrid fungi and amphibian declines: Overview, implications and future directions

Lee Berger1, 2, Rick Speare2 and Alex Hyatt1

ABSTRACT exposure. Tadpoles appear to be unaffected by the fungus which infects their keratinised mouthparts. A recently described chytrid fungus, genus Batrachochytrium can probably also survive and grow Batrachochytrium, killed free-living and captive in the environment. Based on the epidemiology of amphibians in Australia, Central America and the the amphibian declines, chytridiomycosis appears to USA. There is epidemiological, pathological, and be an emerging disease causing mortality in many experimental evidence that some amphibian species of anurans and has caused the disappearance populations suddenly declined due to mass and presumed extinction of some species. These mortalities caused by chytridiomycosis. species may have been more vulnerable to These were notably high altitude, stream dwelling extinction due to a combination of characteristics of rainforest anurans in protected areas of Queensland their distribution and biology which suited and . Chytrid fungi caused a widespread Batrachochytrium, as well as rendering them less able of the skin resulting in hyperkeratosis, to recover from population declines. Here we sloughing and erosions of the epidermis, and present an overview of the published and occasional ulcerations.There was minimal unpublished data on the amphibian chytrid fungus, inflammation in the skin. Infection occurs through discuss the implications of these findings, and waterborne zoospores that invade the superficial suggest future directions that should be taken to layers of epidermis, and experimentally infected investigate and manage this problem. frogs became terminally ill 10–47 days after

1 CSIRO, Australian Animal Health Laboratory Ryrie St, Geelong,Vic 3220 2 School of Public Health and Tropical Medicine James Cook University,Townsville, Qld 4218

23 INTRODUCTION In this paper we collate the data on the amphibian chytrid and expand on previously presented hypotheses, with a focus Amphibian declines in some regions have been attributed to on Australian circumstances. habitat disturbance including pollution, cattle damage, fish introduction and such as logging and wetland degradation (Hayes and Jennings 1986,Tyler 1997). BIOLOGY OF BATRACHOCHYTRIUM However, habitat disturbance does not explain the rapid AND THE disappearance of high-altitude stream-dwelling rainforest The amphibian chytrid has been placed in a new genus, amphibians from many protected areas in Australia (Richards Batrachochytrium (Phylum Chytridiomycota, Class et al. 1993, Mahony 1996) and Central America (Lips 1998). Chytridiomycetes, Order Chytridiales) and an isolate from a There was no correlation between frog population declines captive blue (Dendrobates azureus) that died and changes in ground level solar UV-B radiation in at the National Zoological Park in Washington has been Queensland (Moise, unpubl. data). Several factors in the described as B. dendrobatidis (Longcore et al. 1999). The declines indicated that a waterborne infectious disease, of ultrastructural morphology, amphibian host and 18S rDNA high to adults of some species, had entered a sequence of Batrachochytrium show that it is distinctly population previously unexposed to it. These factors were: different from other chytrid fungi (Berger et al. 1998, 1) sudden, severe declines occurred over a few months; Longcore et al. 1999).

2) declines were asynchronous and spread as a front; Chytridiomycete fungi are a large and diverse group and have 3) adults died while tadpoles survived and metamorphs been found in almost every type of environment, including died when they subsequently emerged; rainforests, deserts and arctic tundra (Powell 1993). They are frequently found in soil and water where they digest 4) no environmental changes were detected; substrates such as chitin from insect cadavers, cellulose from 5) only stream dwelling frogs disappeared and, vegetable matter, keratin from hair and skin, or pollen. These 6) in two intensively monitored sites, mass mortalities were species function as important primary biodegraders and are observed at the time of significant population declines possibly vital to the ecosystem. Others are parasites of (Laurance et al. 1996, Lips 1999,Trenerry et al. 1994). insects, fungi, algae, and nematodes and a few of these cause significant disease (Barr 1990, Powell 1993). In these two montane rainforest locations — Big Tableland, Australia (1993) and Fortuna, Panama (1997) — sick and Powell (1993) discusses the significance and inherent value of dying anurans (including Taudactylus acutirostris, Litoria rheocola chytridiomycetes and reviews the ability of parasitic species to and L. nannotis) were collected for pathological examination cause disease. The onset of chytridiomycete parasitism of and found to be infected with chytrid fungi in the skin phytoplankton is often correlated with a rapid decline in host (Berger et al. 1998). This fungus has been placed in a new population and so has a major impact on the ecology of the genus, Batrachochytrium (Longcore et al. 1999). host. Synchytrium endobioticum causes black wart disease of potatoes in and Canada, and was introduced to the USA Along with the epidemiological evidence above, we present in the early 1900’s but has since been eradicated. Coelomomyces pathological and experimental data that demonstrate that has been considered for use in biological control of mosquitoes. chytrids are pathogenic to amphibians. The pattern of the Apart from species found among the normal rumen flora of population declines is consistent with being caused by ruminants, chytridiomycetes have not been found in vertebrates Batrachochytrium as it is waterborne, is virulent to adults, does other than amphibians (Barr 1990, Berger et al 1998). not kill tadpoles (Berger et al. 1999), prefers cooler temperatures (Longcore et al. 1999), and is not dependent Sparrow (1960) describes the evanescent nature of chytrid upon the highly susceptible species for its continued epidemics, with their sudden appearance, brief period of rapid existence. Similar waves of mass mortalities, described as the multiplication and then decline and disappearance.This pattern is post-metamorphic death syndrome, have been reported in related to their virulence, ability for rapid reproduction, and the various amphibian populations in western North America loss of optimal environmental conditions. Factors affecting the (Scott 1993). Although the cause(s) was not determined, epidemiology of chytrid blooms include seasonal temperature chytridiomycosis has recently been discovered in populations changes, water pH, light, nutrition and dissolved oxygen (Sparrow of endangered north American frogs, including leopard frogs 1968). These may be relevant considerations when attempting to (Rana yavapiensis and R. chiricahuensis) in Arizona (Nichols et isolate Batrachochytrium from the environment and when al. 1998, Morell, 1999) and chytrids were also seen as investigating the causes of outbreaks of chytridiomycosis. For incidental findings in six percent of a group of wild cricket example, epidemics in populations of Litoria caerulea in southern frogs (Acris crepitans) in Illinois (Pessier et al. 1999). Queensland and northern NSW occurred in the winters of 1996, 1997 and 1998, demonstrating seasonal regularity (Table 1). Although Batrachochytrium has a broad amphibian host range and is currently widespread, not all susceptible species have declined. Findings from studies of other aquatic zoosporic fungi may be The selectivity of the declines may be due to a combination of pertinent here. The abundance of Saprolegniaceae in environmental factors and host biology that provide the California was correlated with altitude (Sparrow 1968). necessary conditions for expression of disease, as well as As aquatic phycomycetes are probably very sensitive to rendering species less able to recover after population crashes. contaminants they are considered good biological indicators Declining species from high altitude rainforests have restricted of pollution (Sparrow 1968). ranges and smaller clutch sizes (Williams and Hero 1998).

24 Most chytrids (i.e. members of the order Chytridiales) occur Figure 1: Map of Australia with stars indicating where in aquatic habitats. They have motile flagellated zoospores Batrachochytrium has been found on amphibians. which develop within a stationary sporangium. Sporangia of This distribution correlates with searching intensity. some species form one or more discharge tubes through which the zoospores are released. Zoospores often display chemotaxis towards their particular substrate enabling them to reach hosts or nutrients in the vicinity which are not abundant, although water flow is probably the main method of dissemination (Sparrow 1968). Zoospores of Batrachochytrium are waterborne, can live for over 24 hours (Berger, unpubl.) and are infective to frogs and tadpoles. Zoospores of many fungi produce an adhesive as they encyst on their host (Bartnicki-Garcia and Sing 1986). Encysted AUSTRALIA zoospores of Batrachochytrium in culture take 4-5 days to grow into mature sporangia containing numerous zoospores (Longcore et al. 1999). Sporangia of Batrachochytrium grow in the keratinised epidermis of amphibians, but as they can be grown in culture and grew on boiled snake skin (keratin), they may also be able to exist and proliferate as saprobes in the environment (Longcore et al. 1999). Rhizoids supply the developing sporangia with nutrients, and are formed whether the sporangia are in the epidermis or in culture (Longcore et al. 1999). Batrachochytrium is inoperculate and develops either monocentrically or colonially (Longcore et al. 1999). northern NSW, montane and foothill forests of Victoria, and Some chytrids have a thick walled, resistant resting spore suburban/semi-rural populated areas of Brisbane, Adelaide stage which can survive for decades in extreme conditions and (Figure 1,Table 1). Captive axolotyls from (Powell 1993) but such a stage has not been observed in Townsville (Speare, unpubl.) and Perth (Aplin and Kirkpatrick, Batrachochytrium (Longcore et al. 1999), which may be a unpubl.) were also found with chytridiomycosis. relatively fragile species. Overseas, chytrids have been found in rainforest anurans in Culture media for the amphibian chytrid contained tryptone, montane Panama in 1997 (Berger et al. 1998), in captive frogs gelatin hydrolysate and lactose (Longcore et al. 1999). In in National Zoo,Washington DC and other American zoos culture B. dendrobatidis developed most rapidly at 23C and since 1988 (Nichols et al. 1998, Pessier et al. 1999), and in grew at 28C, but did not grow significantly at 29C (Longcore wild frogs in Arizona (Nichols et al. 1998) and Illinois (Pessier et al. 1999). Cultures grew well at 15C and survived for et al. 1999). Recently, chytridiomycosis was diagnosed in more than three months at 4C (Longcore, unpubl. data). anurana that died in high altitude regions of Equador, including Chytridiomycetes do not generally survive freezing well, adults of Telmatobius niger collected in1989 from the Azuay although some success with storage in liquid nitrogen has province, adults of Atelopus sp. (bomolochos complex) been achieved (Hohl and Iselin 1986). Species without resting collected in 1991 from the Riobamba province and captive spores are less able to be preserved in an inactive state metamorphs of Gastrotheca pseustes that died in 1999 (Hohl and Iselin 1986). (Merino-Viteri and Coloma, unpubl.). As investigations into frog disease become more common, the known distribution No significant ultrastructural morphological differences were of the amphibian chytrid is likely to greatly expand, and by observed between isolates from Australia, the USA and examining archived specimens the dates of first occurrence Central America (Longcore et al. 1999, Berger et al. 1998) will be pushed back. Investigating the origin and historical and DNA comparisons are needed to determine how many range of the amphibian chytrid may determine whether it has amphibian chytrid species exist. It is likely all isolates belong to been introduced or is endemic. a single species. The 18S rDNA sequence of chytrids from a wild caught Australian L. caerulea and a captive American Apparently healthy frogs can exist as carriers; clinically normal D. azureus had only five base pairs different out of about 1700 frogs (including L. genimaculata) collected from the wet bp sequenced, and four of these differences were deletions tropics in Queensland have been found with low levels of which may be due to error (James, Porter and Longcore, infection (Speare and Freeman, unpubl.). Although unpubl.). Preliminary sequencing of a more variable region, the Batrachochytrium has not yet been found in non-amphibian rDNA internal transcribed spacers (ITS), demonstrates that hosts, it is possible that other cold-blooded animals with similar strains (<2% sequence divergence) can infect a range keratinised surfaces could harbour infections and play a role of Australian frog species (Morgan, unpubl.). More studies are in the persistence and spread of the disease. As the chytrid is needed to define significant variations. susceptible to desiccation (Berger, unpubl.), transportation of the disease long distances is most likely to have occurred via DISTRIBUTION the movement of amphibians, either deliberately (for example in the pet trade), or accidentally (such as in fruit boxes). Both In Australia, Batrachochytrium has been found in frogs since these scenarios have been observed and epidemics in captive 1989 and has been observed in various regions — including collections have occurred after the arrival of a few infected rainforests of southern, central and northern Queensland and frogs (Marantelli, pers. comm.). Cane toads are also potential

25 TABLE 1: Species, locations and dates of amphibians in Australia found infected with Batrachochytrium.

Hylidae Litoria adelaidensis Perth,WA Aug 1998 # J Litoria caerulea Casino, NSW July1996 * A 96/692 Brisbane, Qld July–Sep1996 * A 96/961 Brisbane, Qld June–July 1997 * A 97/845 Rockhampton, Qld Aug 1997 A 97/845 Mackay, Qld Sep 1997 A 97/845 Bellingen, NSW July 1998* A 98/871 Lanitza, NSW Aug 1998 A 98/871 Litoria chloris Dorrigo Plateau, NSW Dec 1997–Feb 1998 A 97/1754 Litoria citropa Captive, Sydney May 1999 A 99/232 Litoria ewingi Woodville, SA Oct 1998 A 98/1331 Litoria genimaculata Kirrama State Forest, Qld Nov 1989 A 98/927 Windsor Tableland, Qld April 1993 J Topaz,Qld July 1998 A 98/1231 Wright’s Creek, Qld Oct 1998 J 98/398 O’Keefe Creek, Qld Dec 1998 J 99/28 Litoria infrafraenata Captive,Adelaide, SA Oct 1998 A 99/232 Litoria lesueuri Mary River, Qld May-June 1996 * A 96/570 Eungella NP,Qld June 1996 A 96/1136 Goomburra, Qld April, 1997 A 97/574 Buckwong Ck,Vic Feb 1998 # A 98/320 Litoria moorei Perth,WA Sept 1998 # A 98/927 Ferndale, Perth,WA Nov 1998 # J 99/12 Manjimup,WA Nov 1998 # J 99/12 Mt Barker,WA Nov 1998 # J 99/12 Mt Helena,WA Nov 1998 # J 99/12 Sawyers Valley,WA Nov 1998 # J 99/12 Witchcliffe,WA Nov 1998 # J 99/12 Litoria nannotis Big Tableland, Qld Oct 1993 * J 93/290 Litoria pearsoniana Mebbin State Forest, NSW Dec 1997 A 97/1747 Kroombit Tops,Qld May 1998 A 98/320 Litoria raniformis Captive,Adelaide, SA Sep 1998 A 98/1331 Litoria rheocola Big Tableland, Qld Oct 1993 * J 93/290 (tadpoles & adults) Westgid Ck, Qld May 1999 ® J (tadpoles & adults) Frenchman Ck, Qld July 1999 ® J Litoria spenceri Bogong Ck,Vic March 1996 A 96/370 Taponga River,Vic Feb–March 1998 * A 98/320 Nyctimystes dayi Tully, Qld Dec 1998 A 98/1810 Myobatrachidae Adelotus brevis Brisbane, Qld Nov 1995 A 99/927 Brisbane, Qld Aug 1998 A 99/927 Lamington NP,Qld Jan 1999 A 99/951 pseudinsignifera Sawyers Valley,WA Nov 1998 # J australiacus Springwood, NSW May 1998 A 98/871 Heleioporus eyrei Captive, Perth,WA Nov 1998 # J dorsalis Woodlands, Perth,WA May 1998 # J Goomburra, Qld, Dec 1996 * A 96/1429 Mt Compass, SA Oct 1998 A 98/1331 Captive Melbourne,Vic Nov 1998 A 98/1469 Valley View, SA Jan 1999 A 99/232 Limnodynastes tasmaniensis Adelaide, SA May–June 1996 * A 96/495 Captive,Adelaide Aug 1998 A 98/871 Lechriodus fletcheri Cunningham’s Gap, Qld Jan 1999 A 99/951 Neobatrachus kunapalari Captive Melbourne,Vic June–Aug 1996 A 96/1431 Mixophyes sp. (tadpoles) Goomburra Dec 1996 A 96/1428 Mixophyes fasciolatus Captive Melbourne,Vic Dec 1996–Feb 1997 * A 96/1431 Mt Glorious, Qld Aug 1998 A 98/1159 Mixophyes fleayi Cunningham’s Gap, Qld Aug–Oct 1996 * A 96/ 962 Pseudophryne pengilleyi Captive, Melbourne,Vic Nov 1997 A 97/1510 Taudactylus acutirostris Carbine Tableland, Qld Nov 1990 A 98/927 Big Tableland, Qld Nov 1993–Jan 1994 * J 93/290 Captive Townsville, Qld Dec 1993 * J 93/290 Taudactylus eungellensis Eungella National Park, Qld Oct 1995 * A 96/657 Bufonidae Bufo marinus Captive Geelong,Vic June–Aug 1996 * A 96/673 Ambystomatidae Ambystoma mexicanum Captive Townsville, Qld Feb 1999 J

Accession numbers are provided: "A" refers to specimens deposited at AAHL, "J" refers to specimens deposited at JCU. * denotes records from Berger et al. 1998. # denotes records from Aplin et al. 1999 “ denotes records from Retallick 1999

26 carriers as they continue to expand their distribution in FIGURE 2: Captive-bred metamorph of Mixophyes fasciolatus Australia, as evidence suggests that free-ranging toads are (AAHL accession no. 96/1431) from the Amphibian Research infected with chytrids1 (Parkes, pers. comm.). Centre with naturally acquired chytridiomycosis. This frog is in the terminal stages of the disease - note depressed attitude, half We hypothesise that Batrachochytrium was introduced to closed and accumulations of sloughed skin over the body (arrow head). Bar = 0.5 cm. Australia in the 1970’s around Brisbane (where the first precipitous declines occurred), although as yet there are no hard data to support this assumption. Batrachochytrium has since become established in many areas on the east coast, around Adelaide and in south west .

CHYTRIDIOMYCOSIS:THE DISEASE Typical clinical signs in Australian frogs with chytridiomycosis were lethargy, inappetence, skin discolouration, presence of excessive sloughed skin, and sitting unprotected during the day with hind legs held loosely to the body (Berger, Speare and Marantelli, unpubl.). Frogs became moribund in terminal stages with loss of righting reflex, and death usually occurred FIGURE 3: Unstained skin slough from a dead adult of Litoria a few days after the onset of clinical signs. Most diseased caerulea from Thagoona,Qld (AAHL accession no. 97/845) frogs were in reasonable body condition and 45 out of examined by light microscopy. Note refractile round and oval 69 were considered to have moderate or large fat bodies. chytrid fungi. Most are empty, but one contains developing zoospores (arrow). Occasionally empty sporangia are seen Gross pathology of internal organs was generally divided by thin septae. E = epidermal cell. Bar = 20 µm. unremarkable. Some notable species variation in the clinical signs was observed. The rainforest frogs from Big Tableland, northern Queensland (Speare 1994) predominantly showed neurological signs; most commonly, abnormal sitting posture with hind legs adducted, lethargy, and slow response to tactile stimuli.When handled, these frogs (Litoria rheocola, L. nannotis, Taudactylus acutirostris) became rigid and trembled with extension of the hindlimbs and flexion of the forelimbs. Less common signs included minute skin ulcers and haemorrhage in skin, muscle or . Many frogs appeared anaemic, with pale muscles and internal organs. In contrast, individuals of Litoria caerulea often became intensely red on the belly, legs and feet. These frogs had marked congestion and reddening of organs internally, and skin ulcers were occasionally seen (Berger and Hines, unpubl.) Captive metamorphs of Mixophyes fasciolatus became slow and sloughing skin FIGURE 4: Histological section of heavily infected skin from the accumulated over the body (Figure 2). The presence of toe of a wild-caught adult of Litoria caerulea (AAHL accession shedding skin on the body is a nonspecific sign which is seen no. 97/845). Batrachochytrium does not invade through the epidermis but occurs just under and within the superficial in other diseases where frogs become lethargic and fail to keratinised layer which becomes thickened. Solid, immature wipe themselves, but with chytridiomycosis, excessive sporangia are present (arrow), as well as mature sporangia shedding may occur in response to infection of the skin. containing numerous dark zoospores (Z). Many sporangia are empty after the flagellated zoospores have swum out through the discharge tube (below the D). Immature stages are present In frogs with chytridiomycosis, light microscopy of unstained in the superficial layer of viable epidermal cells and older, empty pieces of sloughing skin revealed masses of round to oval stages are commonly seen in the outer dead layers of keratin refractile fungi (Figure 3), enabling rapid diagnosis. On histologic which are sloughed. Batrachochytrium does not form hyphae. (Haematoxylin & eosin). E = epidermis, K= keratin. Bar = 50 µm. examination, chytrid fungi were observed inhabiting the outer keratinised layers of the epidermis (Figure 4) and were associated with thickening or erosions and occasional ulcerations of the skin (Berger et al. 1998, Pessier et al. 1999). Different frogs varied greatly in their intensity of infection, but skin of the ventral body, limbs and feet were most consistently infected. Bacteria often colonised the layers of sloughing keratin. There was negligible inflammatory response in the dermis. Scanning electron microscopy of diseased skin revealed fungal discharge tubes emerging through the surface of the skin and confirmed

1 Fourteen out of 63 wild-caught juvenile toads from Lennox heads, NSW which were brought into AAHL in june 1996 died between 1 and 44 days after arrival and chytridiomycosis was diagnosed in 5 which were examined histologically (Parkes, unpubl. data).

27 hyperkeratosis and extensive sloughing of the epidermis (Figures FIGURE 5: Scanning electron micrograph of skin from an infected adult of Litoria lesueuri from Goomburra, Qld 5 and 6). The internal structure of Batrachochytrium was (AAHL accession no. 97/5741), showing fungal discharge tubes revealed by transmission electron microscopy (Figure 7). protruding through the surface of epidermal cells. Bar = 10µm. The lack of inflammation in the skin could be due to a lack of stimulation of the host immune system – perhaps due to the superficial site of infection, insufficient epidermal damage, or the chytrid may have low inherent antigenicity.

There were few specific internal lesions in sick frogs, suggesting that the ultimate cause of death was metabolic or toxic. Focal necroses, vacuolation, or cloudy swelling were sometimes apparent in a range of internal organs (Speare and Berger, unpubl.).

Histologic examination of organs involved in immunity — i.e. spleen and bone marrow, revealed no evidence of immunosuppression, apart from in one frog. Tests (including viral and bacterial culture, electron microscopy and haematology) for infectious organisms did not detect any other significant pathogens. Of 147 wild and captive frogs with FIGURE 6: Scanning electron micrograph of skin from the toe of an infected adult of Litoria lesueuri (AAHL accession no. chytridiomycosis, concurrent disease was diagnosed in 16 97/5741) with extensive peeling and degeneration of the (11%). However, apart from one frog with immunosuppression, keratinised skin surface. The surface of healthy skin from control these diseases were not considered to be the primary cause of frogs appeared smooth. Bar = 100µm. death. The concurrent diseases included septicaemia (four frogs), microsporidial hepatitis (one frog) and hyphal mycotic dermatitis (two frogs), all of which may have occurred secondary to chytridiomycosis. The other diseases - biliary hyperplasia and/or fibrosis (six frogs), foreign body myositis (one frog) and mild, localised mucormycosis (one frog) — all appeared chronic and inactive and may not have contributed to the deaths. A variety of helminth and protozoan parasites were identified as incidental findings (Berger and Speare unpubl.).

During our survey of amphibian diseases in Australia, 272 wild and captive sick frogs were examined between 1989 and 1999, but the only disease that was found consistently was chytridiomycosis which occurred in 54% and accounted for almost all cases of unusual mortality in the wild. A range of diseases was diagnosed in individual sick frogs that did not have chytridiomycosis, including viral, bacterial, protozoal, fungal, tapeworm, neoplastic, traumatic and congenital diseases, FIGURE 7: Transmission electron micrograph of a mature and inappropriate ingestion (Berger and Speare, unpubl.). sporangium cultured from the skin of a sick Nyctimystes dayi from Tully, Qld (AAHL accession no. 98/18103). It is filled with developing zoospores, each with a single flagellum (arrow head). Batrachochytrium is not a ubiquitous parasite. In a histological The discharge tube is closed with a gelatinous plug which survey of toeclips from 348 apparently healthy frogs from deliquesces when the zoospores are ready to be released. Queensland and Victoria collected between 1989 and 1999 Bar=3µm. only 7 (2.0%) were infected (Speare, Berger and Kent, unpubl.).

In experimental infections in Australia using Mixophyes fasciolatus a terminal illness was reached in 6/6 frogs 10–18 days after exposure to infected skin scrapings at 24C (Berger et al. 1998). In the USA, 2/2 Dendrobates tinctorius died 23 and 31 days after being exposed to broth culture of B. dendrobatidis (Longcore et al. 1999). Small doses of the pathogen have now been shown to cause fatal chytridiomycosis in metamorphs of Mixophyes fasciolatus — 3/3 frogs each exposed to an estimated 1000 zoospores died or became terminally ill between 23 and 38 days post exposure, 3/3 frogs exposed to approximately 100 zoospores died between 35 and 47 days post exposure, however 3 frogs exposed to approximately 10 zoospores did not succumb to chytridiomycosis and 2 have remained healthy for over 3 months (the other frog died by misadventure) (Marantelli and Berger unpubl.). The metamorphs were

28 housed individually and were infected by bath exposure to Once a new pathogen has emerged, the rapid global transport zoospores (cultured from a captive Limnodynastes dumerilli) by air of people, goods and animals will assist in its in 10ml water for 24 hours. dissemination both within and between continents.

In epidemics in captive collections of M. fasciolatus and Bufo If long-term degradation of the environment was the key marinus, almost all tadpoles survived to metamorphose, but problem in declines of amphibian populations, we would suffered high mortality rates 2-3 weeks after metamorphosis expect reproductive and nutritional status to be affected (Berger et al. 1998). Tadpoles were found infected with before fatal immunosuppression occurs. However, moribund Batrachochytrium, but were never found dead or dying as a frogs were found which were gravid (Mahony 1996), and we consequence of infection (Berger et al. 1998). Tadpoles were have seen many with adequate fat reserves. Also, with severe infected in the mouthparts (see Berger et al. 1999), the only immunosuppression a range of opportunistic infections is keratinised area of tadpole skin. After metamorphosis, the skin likely to be involved, rather than a solitary chytrid fungus. The over the body becomes keratinised allowing the fungal infection experiments in captivity demonstrate that the chytrid fungus to spread. The resistance of tadpoles to chytrid fungi is consistent can cause 100% mortality in conditions where uninfected with the epidemiology of the amphibian declines where tadpoles animals remained healthy. There is no evidence that have been seen after adults have disappeared (McDonald and predisposing immunosuppression is necessary for epidemics Alford 1999, Lips 1999). During the declines in Panama in 1997, of chytridiomycosis to occur. many abnormal tadpoles were seen with partial or complete loss of keratinised mouthparts, although tests for fungal infection in Although Batrachochytrium has a wide distribution and broad these animals were not reported (Lips 1999). Tadpoles with host range, we suspect only some species had the necessary abnormal mouthparts have not been observed in Australia. combination of characteristics to render them vulnerable to decimation by disease. The declining species from high Two theories have been proposed to explain how a fungus altitude rainforests in eastern Australia have significantly smaller restricted to the superficial epidermis could kill frogs (Berger et al. clutch sizes, occupy restricted geographic ranges, have aquatic 1998, Pessier et al. 1999). The first hypothesis is that the chytrid larvae associated with streams, and many spend a large releases proteolytic enzymes or other active compounds that are proportion of their time in or adjacent to streams (Williams absorbed through the permeable skin of the frog. The second and Hero 1998, McDonald and Alford 1999). Populations of hypothesis is that damage to skin function results in disturbance these species are therefore less able to recover from declines of water or electrolyte balance resulting in death.We tend to due to any cause, and also inhabit environments that would favour the first explanation. Anurans have evolved to manage support Batrachochytrium i.e. cooler, riparian habitats. water stress by shifting water and electrolytes from the brain and cerebrospinal fluid and can withstand levels of dehydration that Other aspects of host biology, genetic factors (see Leberg would be fatal to other classes of vertebrates (Hillman 1988). and Vrijenhoek 1994), humidity and complex ecological The epidermal damage caused by chytridiomycosis does not factors affecting the lifecycle of the chytrid, could be appear severe enough to result in the major changes in water important in determining the selectivity of the declines. and electrolyte balance required to cause death. As Batrachochytrium can probably exist as a free-living organism in the environment (Longcore et al. 1999), can be DISCUSSION AND IMPLICATIONS carried by healthy tadpoles and has a broad host range The observations of high mortality rates resulting in the (Berger et al. 1998), this may explain an ability to persist and disappearance of populations suggests there has been some cause disease even when the density of adults from particular recent change in the balance between host and pathogen. species has been greatly reduced (Daszak et al. 1999). Three explanations are suggested: Although most of the frog species that have disappeared 1. Batrachochytrium has been introduced recently to these areas, from Queensland rainforest since 1979 (e.g. Rheobatrachus 2. it has suddenly become more pathogenic to the host, or vitellinus, R. silus,Taudactylus diurnus,T. rheophilus, Litoria lorica and L. nyakalensis (Tyler 1997)) were not found dying and 3. the frogs have a lowered resistance due to environmental tested for disease, the epidemiological evidence suggests that changes allowing the endemic parasite to cause mass chytridiomycosis caused mass mortalities of these species, as mortalities. was observed in declining frogs (including T. acutirostris) at the The first explanation seems the most plausible, due to the Big Tableland site in north Queensland in 1993 (Berger et al. epidemiological factors already discussed, and acute or chronic 1998). The possible role of chytridiomycosis in frog declines environmental problems in these protected rainforest areas in southern Australia is less clear. were not detected (Richards et al. 1993). The amphibian The episodes of population declines described above now chytrid fungus may have originated by the escape of a pathogen appear to have passed in many areas. Numbers of some from a local environment, transfer to new hosts, followed by species, decreased significantly when other species rapid progression through these non-essential hosts. Two disappeared, but are now increasing.These species include T. recent examples of emerging infectious diseases following this eungellensis (McDonald and Alford 1999), L. genimaculata on pattern are Ebola virus in and Vibrio cholerae 0139. The Big Tableland (McDonald unpubl.) and L. pearsoniana from first has had several episodes of escape, from an as yet southeast Queensland (Hero et al. 1998). As residual undetected host into the human population of Zaire populations are increasing after significant declines although (Dobson and Carper 1996), while the second, an environmental chytrids are still present (Berger et al. 1998), this suggests bacterium, arose from the Bay of Bengal in 1992 and has spread resistance is present in some members of the population. A rapidly through the human population of Asia (Lee et al. 1996). balance between Batrachochytrium and the frogs may be

29 developing in areas where we suspect it has been present for distribution of chytridiomycosis by examination of frogs at least five years, and although disease still occurs, it does not preserved in museum collections to determine whether devastate the populations. Batrachochytrium may now behave chytrids were introduced to Australia or to these habitats and as an endemic pathogen with outbreaks of disease occurring to determine from where and how chytrids have spread. when conditions are optimal. The long term prognosis may Information on the current distribution may also help in be good for species which have survived and are recovering, formulating management plans to prevent new outbreaks — as long as remaining habitats are protected to allow damaged for example, in deciding where exposed or uninfected frogs populations to reestablish. However, since we currently know should be released after captive breeding. This survey relies little about the interaction between the amphibian chytrid on a collaborative approach (see Appendix 1). As many and hosts, and many frog species are currently considered healthy tadpoles may carry chytrids for extended periods, threatened or their status is insufficiently known (Tyler 1997), sampling of tadpoles may provide a more sensitive means we should not be complacent. Also, if uninfected areas still of assessing a location. Examination of sick frogs is the exist in Australia then locally endemic species in those areas most sensitive way to detect chytrids, but these are may be at high risk. not always available.

Infectious disease is important in the population biology of wild Mycological studies on the chytrid are needed to learn more animals, as it is in humans and domestic animals (May 1988). A about its lifecycle, requirements and survival in the wild. review of infectious disease and animal populations concluded Knowledge about the ecology and hosts of the fungus is that disease is an important factor affecting survival, essential to understanding the spread, and therefore to the reproduction, dispersal, community structure and genetic management of the disease. Information about the particular diversity, and should therefore be considered by ecologists conditions that encourage growth may help in understanding examining host population-dynamics (Scott 1988). Disease the factors which precipitate disease epidemics. becomes a threatening force when environmental degradation puts pressure on populations, and international trade and Further transmission experiments are required to confirm the smuggling continually threaten to introduce new pathogens to pathogenicity in a range of species, and also to determine which the native fauna has no resistance (Scott 1988, Daszak et what environmental conditions e.g. temperature, are required al. 1999). Exotic diseases can have effects similar to those of feral for expression of the disease in frogs. Treatments for adults predators, with susceptible native species facing extinction while and tadpoles are also being tested which will aid captive the ecosystem readjusts, and are another example of how breeding of endangered species. By producing large numbers increased global homogeneity leads to reduced biodiversity. of frogs in captivity, it may be possible to help species to Although the initial >99% mortality rate of myxomatosis was survive and evolve immunity. not sufficient to exterminate rabbits from Australia (Fenner and Studies of the pathogenesis of chytridiomycosis are important Ratcliffe 1965), it is plausible that a similar mortality rate could to understanding this disease, and immunologists in the USA wipe out frog species with limited distributions and relatively will investigate the innate and acquired immune response of infrequent breeding. The introduction of avian malaria is frogs to Batrachochytrium. suspected to have caused the extinction of birds in Hawaii (Warner 1968). A protozoan parasite of (Toxoplasma gondii) Further work will be done using DNA analysis to compare was probably introduced to Australia during the European chytrids from various species from localities across Australia, invasion, and marsupials are among the most susceptible animals Central America and the USA. The number of species of (Reddacliff et al. 1993). Phytophthora cinnamomi is an example of Batrachochytrium can be determined by using this information a pathogenic, introduced zoosporic fungus which threatens many combined with morphological . Molecular biology as native Australian species, and quarantine measures are a tool for molecular epidemiology can also be used to recommended to prevent the invasion of the pathogen into provide clues about the origins and spread of the fungus. new areas. Some plant species are highly susceptible, whereas others only become diseased after periods of stress such as a Data are being collected to evaluate diagnostic tests. drought (Dawson and Weste 1985;Wills 1993). Infrastructure Histology and examination of skin scrapings are highly specific exists to prevent and manage exotic disease outbreaks in tests, but may not be very sensitive when used to detect domestic animals, but little concern is shown for wildlife where chytrids on healthy specimens. Production of antibodies has many diseases are yet to be discovered and understood, and commenced to enable more sensitive testing, and perhaps for monitoring the disease status of populations currently appears to use in detecting chytrids in the environment. be the responsibility of no one. Regulations regarding quarantine, testing, treatment and movement of amphibians need to be introduced to prevent FUTURE DIRECTIONS further spread of Batrachochytrium within Australia and Work on amphibian chytridiomycosis must continue — to internationally. Although our understanding of the role of confirm or reject the hypothesis that it is the primary cause of chytridiomycosis in amphibian declines is far from complete, we the declines, to determine how the epidemic began, to find ways believe it is crucial to take immediate preventative measures to manage the problem in areas where the fungus is established, rather than risk waiting for more scientific data to be accrued. and to prevent it occurring in new regions. Knowledge gained from these investigations will be useful in preventing similar ACKNOWLEDGEMENTS population crashes occurring in frogs again, and in developing Amphibian disease investigations would not have been possible management strategies for other wildlife species. without the tremendous support of herpetologists throughout We will continue to map the temporal and geographic Australia. A huge collaborative effort was involved in collecting

30 sick frogs from the field and urgently forwarding them to the Hayes, M. P.and Jennings, M. R., (1986) Decline of ranid frogs lab.We are indebted to many people who have been involved, species in Western North America: Are bullfrogs (Rana including those who provided sick or healthy specimens, catesbiana) responsible? J Herpetol, 20: 490-509. allowed us access to their collections or gave advice, especially Hero, J-M., Hines, H., Meyer, E., Morrison, C., Streatfeild, C. and Gerry Marantelli, Harry Hines, Keith McDonald, Mike Tyler, Ken Roberts, L., (1998) New records of “declining” frogs in Aplin, Alastair Freeman, Craig Williams, Rebecca Short, Graeme Queensland, Australia (-February 1998). Froglog 29. Gillespie, Nick Sheppard, David Page, David Byrnes, Steve Hillman, S. S., (1988) Dehydrational effects on brain and Williams, Craig Taylor, D.A. Stewart, Michael Mahony, Lance cerebrospinal fluid electrolytes in two amphibians. Physiol Tarvey, Michael Smith, Richard Retallick, Marc Hero, John Clarke, Zool, 61: 254-259. Michael Healey, Brent Dadds, Lothar Voigt, Adrian Wayne, Jeremy Morante, Danny Wotherspoon and members of the Hohl, H. R. and Iselin, K., (1986) Liquid nitrogen preservation public.We are very grateful to Lee Skerratt, Deborah Middleton of zoosporic fungi. Pp 143-145 in Zoosporic Fungi in and Murray Littlejohn for comments on the manuscript. Thanks Teaching and Research. ed by M. S. Fuller and A. Jaworski. to Frank Fillipi for photography, Julia Hammond for bacteriology, Dept of Botany, University of Georgia, Athens GA. and Megan Braun, Gail Russell,Terry Wise and Andrew Kent for Laurance,W. F., McDonald, K. R. and Speare, R., (1996) the many tissue sections prepared. Peter Hooper and Mark Epidemic disease and the catastrophic decline of Australian Williamson are thanked for their support and pathology rainforest frogs. Conserv Biol, 77: 203-212. expertise.We also thank people who allowed us to cite their Leberg, P.L. and Vrijenhoek, R. C., (1994) Variation among unpublished data. Lee Berger has been supported by a grant desert topminnows in their susceptibility to attack by from Environment Australia. exotic parasites. Conserv Biol, 8: 419-424. Lee, S. H., Lai, S. T., Lai, J.Y. and Leung, N. K., (1996) Resurgence REFERENCES of cholera in Hong Kong. Epidemiol Infect, 117: 43-49. Aplin, K., Speare, R., Berger, L. and Cowan, M. A., (1999) Lips, K. R., (1998) Decline of a tropical montane fauna. Outbreak of fungal disease in Western Australian Frogs. Conserv Biol, 12: 106-117. J Royal Soc WA (submitted). Lips, K. R., (1999) Mass mortality and population declines of Barr, D. J. S., (1990) Phylum Chytridiomycota. Pp. 454-466 in anurans at an upland site in western Panama. Conserv Biol, Handbook of Protoctista. ed by L. Margulis, J. O. Corliss, 13: 117-125. M. Melkonian and D. J. Chapman. Lubrecht and Kramer, Longcore, J. E., Pessier, A. P.and Nichols, D. K., (1999) Monticello, New York. Batrachochytrium dendrobatidis gen. et sp. nov., a chytrid Bartnicki-Garcia, S. and Sing,V. O. (1986) Adhesion of pathogenic to amphibians. Mycologia, 91: 219-227. zoospores of Phytophthora to solid surfaces. Pp. 279-283 Mahony, M., (1996) The decline of the green and golden bell in Zoosporic Fungi in Teaching and Research. ed by frog Litoria aurea viewed in the context of declines and M. S. Fuller and A. Jaworski. Dept of Botany, University of disappearances of other Australian frogs. Aust Zool, 30: Georgia, Athens GA. 237-247. Berger, L. Speare, R. Daszak, P., Green, D. E, Cunningham, A. A., Marantelli, G., (1999) Husbandry: science or art ? — Are Goggin, C. L., Slocombe, R. Ragan, M. A. Hyatt, A. D., McDonald, captive technologies ready to contribute to recovery K. R. Hines, H. B., Lips, K. R., Marantelli, G. and Parkes, H., (1998) processes for Australian frogs? Pp 168-176 in Declines and Chytridiomycosis causes amphibian mortality associated with Disappearances of Australian Frogs ed by A.Campbell. population declines in the rainforests of Australia and Central Environment Australia, Canberra. Proc Natl Acad Sci America. , 95: 9031-9036. May, R. M., (1988) Conservation and Disease. Conserv Biol, Berger, L., Speare, R. and Kent, A., (1999) Diagnosis of 2: 28-30. chytridiomycosis in amphibians by histologic examination. McDonald, K.R. and Alford, R.A., (1999) A review of declining Proc Frog Symposium: Frogs in the Community , Queensland frogs in northern Queensland. Pp 14-22 in Declines and Museum, Brisbane, February 1999 (submitted). Disappearances of Australian Frogs ed by A.Campbell. Daszak, P.1,Berger, L., Cunningham, A. A., Hyatt, A. D., Green, D. Environment Australia, Canberra. E. and Speare, R., (1999) Role of Emerging Infectious Morell,V. (1999) Are pathogens felling frogs? Science, 284: Diseases in Amphibian Population Declines and Global 728-731. Implications. Emerg Inf Dis, (in press). Nichols, D.K., Pessier, A.P.,and Longcore, J.E., (1998) Dawson, P.and Weste, G., (1985) Changes in the distribution Cutaneous chytridiomycosis: an emerging disease? Proc Am Phytophthora cinnamomi of in the Brisbane Ranges National Assoc Zoo Vet, 1998:269-271. Park between 1970 and 1980-81. Aust J Bot, 33: 309-315. Pessier, A. P.Nichols, D. K., Longcore, J. E. and Fuller, M. S., Dobson, A. P and Carper, E. R., (1996) Infectious disease and (1999) Cutaneous chytridiomycosis in poison dart frogs Bioscience human population history. , 46: 115-126. (Dendrobates spp.) and White’s tree frogs (Litoria caerulea). Fenner, F. and Ratcliffe, F. N., (1965) Myxomatosis, Cambridge J Vet Diag Invest, 11: 194-199. University Press, New York. Powell, M. J., (1993) Looking at mycology with a Janus face: Groff, J. M., Mughannam, A., McDowell,T. S.,Wong, A. Dykstra, A glimpse of chytridiomycetes active in the environment. M. J., Frye, F. L. and Hedrick, R. P., (1991) An epidemic of Mycologia, 85: 1-20. cutaneous zygomycosis in cultured dwarf African clawed Reddacliff, G.L., Hartley,W.J., Dubey, J.P.and Cooper, D.W., (Hymenochirus curtipes) Basidiobolus ranarum frogs due to .J (1993) Pathology of experimentally-induced, acute Med Vet Mycology , 29: 215-223. toxoplasmosis in macropods. Aust Vet J.70: 4-6.

31 Retallick, R., (1999) Translocations and experimental ecology APPENDIX 1 of declining frogs. Update for the North Queensland Frog Recovery Team,August 1999. Unpublished report. What herpetologists can do to assist Richards, S. J., McDonald, K R. and Alford, R. A., (1993) SURVEY OF SICK AND HEALTHY FROGS Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conserv Biol, 1: 66-77. We wish to examine any diseased frogs or cane toads that are found in order to determine the cause of death and to Scott, M. E., (1988) The impact of infection and disease on screen for the presence of chytrids. Frog tissues deteriorate animal populations: Implications for conservation biology. very rapidly after death, so if a sick frog is found that is likely Conserv Biol, 2: 40-56. to survive another 24 hours, it should be sent by courier to Scott, N. J., (1993) Post metamorphic death syndrome. the Australian Animal Health Laboratory or James Cook Froglog, 7: 1-2. University after contacting us. Frogs or tadpoles found dead Sparrow, F. K., (1960) Aquatic Phycomycetes. 2nd re. ed. should be fixed or frozen immediately to preserve the University of Michigan Press, Ann Arbor, Michigan. tissues. They should be fixed in 10% buffered neutral Sparrow, F. K., (1968) Ecology of Freshwater Fungi. pp 41-93 formalin, but 70% ethanol can also be used. It is important to in The Fungi. ed by G. C. Gainsworth and A. S. Sussman, slit open the belly, and to ensure the frog is well covered in Academic Press, New York. fixative so that tissues are preserved rapidly. Details of what Speare, R., (1994) Preliminary study on diseases in Australian to do with sick or dead frogs have been posted on the Wet Tropics amphibians. Deaths of rainforest frogs at O’ World Wide Web at http://www.jcu.edu.au/dept/school/phtm/ Keefe Creek, Big Tableland. Final report to Queensland PHTM/frogs/pmfrog.htm. Department of Environment and Heritage. Unpublished Please send collection data with any frog submitted, and we report, Queensland Department of Environment and will keep you informed about the results of the post mortem. Heritage. Pathology is required for diagnosis, as the clinical signs of Taylor, S. K.,Williams, E. S.,Thorne, E. T., Mills, K.W.,Withers, chytridiomycosis are not highly specific. D. I. and Pier, A. C., (1999) Causes of mortality of the Wyoming toad. J Wildl Dis, 35: 49-57. We have prepared a frog mortality questionnaire (http://www.jcu.edu.au/school/phtm/PHTM/frogs/pmques.htm) Trenerry, M. P.,Laurance,W. F. and McDonald, K. R., (1994) which details the type of data that are important to observe Further evidence for the precipitous decline of endemic and record if you encounter a mass mortality event. rainforest frogs in tropical Australia. Pacific Conserv Biol, 1: 150-153. For our examination of archived frogs for chytrids, we need Tyler, M. J., (1997) The Action Plan for Australian Frogs. Wildlife skin samples from amphibians from a wide range of localities Australia, Canberra. and dates.We want skin from the pelvic areas and toes from Viggers, K. L., Lindenmayer, D. B. and Spratt, D. M., (1993) any formalin-fixed or ethanol-fixed frog that has collection The importance of disease in reintroduction programmes. data.We especially require frogs from inland Australia, Wildl Res, 20: 687-98. Northern Territory and northern WA, as none have been examined from these regions. A protocol is available at Warner, R. E., (1968) The role of introduced diseases in the http://www.jcu.edu.au/school/phtm/frogs/pmskin.htm. extinction of the endemic Hawaiian avifauna. The Condor, 70: 101-120. People doing skeletochronology on histological sections of Williams S. E. and Hero J-M., (1998) Rainforest frogs of the toes could simultaneously check the skin for chytrids (see Australian Wet Tropics: guild classification and the Figure 4). In healthy frogs, the level of infection may be very ecological similarity of declining species. Proc Roy Soc Lond, low, with only occasional sporangia present along the skin B 265: 597-602. surface. For detailed diagnostic histological features, see Wills, R. T., (1993) The ecological impact of Phytophthora Berger et al. (1999). cinnamomi in the Stirling Range National Park,Western We are attempting to maintain a comprehensive list Australia. Aust J Ecol, 18: 145-159. of confirmed cases of chytridiomycosis (http://www.jcu.edu.au/school/phtm/PHTM/frogs/chyspec.htm) and hope that data will be submitted for inclusion. This list will enable management decisions to be made based on current knowledge.

32 MANAGEMENT OF CHYTRIDIOMYCOSIS IN CAPTIVITY PREVENTING SPREAD OF DISEASE IN THE WILD If any epidemics of chytridiomycosis occur in captive To prevent the spread of chytrids or other diseases when collections, various antifungal drugs could be administered, performing field work, disinfection of equipment should be and the results communicated. performed.We need more information on the resistance of this fungus to heat, desiccation and disinfectants; so at present Benzalkonium chloride is a disinfectant that has been used at are recommending measures for disinfection that have been 2 mg/l to successfully treat a similar superficial mycotic proven against highly resistant organisms. (see protocol — dermatitis in dwarf African clawed frogs (Hymenochirus http://www.jcu.edu.au/school/phtm/PHTM/frogs/prevent.htm). curtipes) reported to be caused by Basidiobolus ranarum (Groff et al. 1991). The regime used experimentally was Disease status should always be considered when 30 minutes of bath treatment, on three alternate days. translocating animals (Marantelli 1999,Viggers et al. 1993) and This was repeated in 8 days (i.e. 6 treatments in total). Oral attempts should be made to reduce the chance of has also been used to treat B. ranarum infections introducing disease to a naïve population.We recommend (Taylor et al. 1999). One micro bead from 100mg the screening of healthy frogs by histologic examination of itraconazole capsules was administered daily for 9 days to toe clips, or by sacrificing a few in the group for more Wyoming toads (Bufo baxteri) at the first signs of disease. extensive skin examination (Berger et al. 1999). Testing for Benzalkonium chloride (1mg/L), amphotericin and fluconazole chytrids on sick frogs is more sensitive, so any deaths in a are effective against Batrachochytrium in vitro (Berger, unpubl.). valuable group of animals should be submitted to a pathology laboratory for testing. To screen a group of healthy tadpoles, In captivity, routine quarantine procedures (Marantelli 1999) some need to be sacrificed so that their mouths can be have been adequate in restricting outbreaks to certain tanks, examined histologically (see Berger et al. 1999).We have little and no airborne transmission has been observed (Marantelli, information on the sensitivity of these tests, so it is impossible unpubl.). Each group of frogs should be kept completely to recommend a statistically significant number of animals. separate to ensure no water borne transmission of disease Also, until we have more data on the distribution of chytrids can occur. By changing and discarding gloves between every in amphibian populations around Australia, it will be difficult tank, avoiding splashing water between tanks, and disinfection to make decisions about the release of infected animals. of tanks and implements before reuse using 2% hypochlorite, many frogs have been housed in close proximity without New regulations are being proposed to control the transmission of disease. movement and trade of adult amphibians and tadpoles, but before these are introduced, quarantine measures should become routine.

33 Measuring and analysing developmental instability as a tool for monitoring frog populations

Ross Alford, Kay Bradfield and Stephen Richards*

ABSTRACT INTRODUCTION

Levels of developmental instability (DI), usually Developmental stability can be defined as the ability of a measured by examining the degree of asymmetry of genotype to repeatedly produce precisely the same phenotype when exposed to the same environment during bilateral structures, increase as stress increases or development (Zakharov 1992;Tracy et al. 1995; Moller 1996). health decreases in many species. Several studies Developmental stability is produced by the action of complex have shown that DI levels increase before or during feedback mechanisms which ensure that random population declines. Increases in population DI levels environmentally-induced deviations from the genetically may therefore indicate that populations are at risk determined phenotype are corrected as development progresses (Zakharov 1992; Graham et al. 1993a; Graham of decline. This suggests that DI measurements et al. 1994; Clarke 1995). These deviations are termed should be incorporated in population monitoring “developmental noise” (Palmer and Strobeck 1986; Graham programs. Because these techniques are relatively et al. 1993a). It is thought that developmental stability is highest new, they have not been commonly used, and may when well-integrated genomes are expressed in benign be inaccessible to many field workers.We discuss environments (Leary and Allendorf 1989; Graham 1992). Poorly integrated genomes or poor-quality environments can how to measure DI in frogs, using blind interfere with the ability of organisms to correct for the effects measurement techniques that reduce the possibility of developmental noise, leading to greater variability in the of observer bias and maximise measurement phenotypes produced. This increased variation is termed precision.We also provide examples of data analysis, developmental instability (DI) (Palmer 1996). interpretation and presentation. The approach we Because the levels of developmental instability in populations outline will allow field biologists to incorporate of animals should reflect their genetic makeup and the quality DI measurement into their research programs. of the environment they experience during growth and

* School of Tropical Biology and Cooperative Research Centre for Tropical Rainforest Ecology and Management, James Cook University,Townsville, Queensland 4811 Australia.

34 development, it has often been proposed (Pankakoski 1985; populations of two frog species (Litoria nannotis and Leary and Allendorf 1989; Parsons 1990a, b; Freeman et al. L. genimaculata) experienced large and prolonged increases in 1996; Clarke 1994) that measurements of the degree of DI DI commencing nearly two years before population declines existing in populations should provide a sensitive indicator of became apparent through standard monitoring techniques. their general health. Studies on a variety of taxa have shown Similar results were found for fruit flies by Tsubaki (1998). that declining populations (Tsubaki 1998) or populations in Alford et al. suggested that monitoring of DI levels should be stressful environments (Valentine et al. 1973; Moller 1996; incorporated into routine population monitoring for many Sarre 1996; Siimäki and Lammi 1998) exhibit increased levels species, because increases in DI may act as an early warning of DI. Measuring DI requires the investigator to determine signal of stresses that may later lead to population declines, how far the phenotypes of a series of individuals deviate giving time to investigate and ameliorate the causes of stress from those they would have had if their development had before they adversely affect the population biology of species. proceeded under ideal conditions. For most phenotypic attributes of most animals, the ideal state that would be The aim of this paper is to provide guidelines for measuring produced by any particular genotype is unknown, making and analysing DI using techniques which should be practical developmental instability difficult or impossible to measure. to incorporate into regular monitoring programs for many One class of features, however, have well-defined ideal states: species of frogs. Our technique for measuring DI is designed Bilateral features of species that are normally bilaterally to meet the following criteria: symmetrical should be perfectly symmetrical. Because of this, • It must produce highly precise measurements. Differences of developmental stability analysis usually involves the 1% or less in the sizes of structures can be important in DI examination of departures of traits from symmetry analysis and, since the legs of frogs are small structures, (Graham et al. 1993b). precision on the order of 0.1 mm is critical for useful results.

Most features of normally bilaterally symmetrical animals • It must be as immune as possible to measurer bias. exhibit some degree of asymmetry, which can be of three • It must be possible to carry out the selected types (Palmer and Strobeck 1986): directional asymmetry measurements on irreplaceable museum specimens and occurs when the mean of the left-right differences in sizes of on living animals, so the measuring technique must not be bilateral structures is not equal to zero. Directional invasive or destructive. asymmetry is an element of the genetic and developmental program of certain structures in many species. A common MEASURING METHODS example is the heart of mammals, in which the left side is normally larger than the right (Graham et al. 1993a). Structures measured and effects of body Characters that are directionally asymmetrical are not suitable size on measurements for developmental instability analysis because their ideal state cannot be certainly known. Antisymmetry occurs when a We recommend measuring the lengths of the most easily structure is less than perfectly symmetrical in the majority of accessible long bones on the forelimbs and hindlimbs, plus individuals, but in any one individual, either side is equally adjoining tissue: for the forelimbs, this is the radio-ulna, and likely to be the larger one. This leads to a bimodal for the hindlimbs, the tibio-fibula. distribution of left-right differences in structure size, with a The forelimb measurements are made from the ‘elbow’ to mean of zero. Antisymmetry is a part of the normal the palm of the flexed ‘hand’. The hindlimb measurements developmental program of some structures in some species. are easier to make on live animals, as they are from the knee An example is the chelae of some species of lobsters (Palmer to the ankle, and can be made on most frogs very and Strobeck 1986), in which it is normal for the claw on one conveniently while they are in a normal resting posture. side to be larger than the other, but which side has the larger It is also possible to measure the length of the upper tarsal claw is determined by random stimuli. Antisymmetry may also bones on the hindlimb, from the ankle to the base of the occur when normal developmental processes, such as toes. The precise starting and ending locations of these feedback systems controlling the allocation of resources measurements are not as important as is consistency in the between body parts during development, fail to function choice of starting and ending locations. properly (Graham et al. 1994; Emlen et al. 1994). The causes of this failure may be either genetic or environmental. One possible problem in measuring and interpreting amounts Environmentally-induced antisymmetry can be used to of asymmetry is that the measured values of left-right measure developmental instability, because the ideal differences will be correlated with the sizes of the parts phenotype, perfect bilateral symmetry, is known. The final measured. Comparing or combining asymmetry form of asymmetry is fluctuating asymmetry (FA), in which the measurements taken in absolute units such as millimeters population mode and mean is perfect symmetry of bilateral from animals over a wide range of body sizes can present structures, with left-right differences normally (or at least difficulties. It may be possible in some cases to correct for unimodally and symmetrically, Palmer and Strobeck 1986) size effects using standard morphometric techniques (see distributed with a mean of zero. It is generally acknowledged below and Palmer and Strobeck, 1986). A simpler approach that levels of FA provide a good index of levels of DI within is to take all measurements on animals of similar sizes. populations (Palmer and Strobeck 1986; 1992). Fortunately, this is easily accomplished for most frog species, Measurements of DI may provide a more sensitive indication as the most commonly encountered individuals are adult of the general health of animal populations than can be males, which tend to occur in a limited range of body sizes. obtained through routine monitoring of population sizes. If animals measured vary by more than +/- 10% or so in Data collected by Alford et al. (1997) demonstrated that body size, potential size effects should be checked using

35 scatterplots and correlation analyses comparing unsigned calipers and then hands them to the recorder, who lifts the asymmetry with an overall linear size measurement such as tape, reads the measurement, and records it, then resets the snout-vent length (SVL). If a correlation is found, it may be calipers to zero. This eliminates the possibility of bias, since possible to correct for it by performing subsequent analyses on the measurer cannot set the calipers to previous residuals of regressions of asymmetry against body size, or by measurements as he or she does not know their values. By converting individual measurements to percentages of mean measuring the limbs in sequence around the animal, each size for left and right structures (Palmer and Strobeck 1986). replicate measurement of each limb is taken as far as possible in the sequence from other replicate measurements of the The basic measurement protocol same limb. This technique provides a basis for calculating levels of measurement error that should be as close as is When measurements are carried out on whole animals practical to the technique outlined above. rather than bones, there are likely to be measurement errors, particularly when the animals are alive. Results might also be Another technique that works well in the laboratory but influenced by the measurers’ expectations. To minimise these which we have not tried in the field is the use of digital effects, we devised the protocol that follows. electronic calipers with data outputs. This has the advantage that a single person can do measurements without a helper. First, the SVL (snout-vent length) and the fore- and hind- A piece of opaque tape or other object is secured over the limbs on the left side of the body are measured to the readout on the calipers so that the measurer cannot see the nearest 0.05mm, using calipers. These measurements are values obtained. The data are fed directly into the recording recorded along with an identifying number for each individual computer by activating the data output of the calipers after (if measuring animals that are not marked and will not be setting them for each measurement, in a sequence similar to marked, house them in individually numbered containers such that outlined above. The calipers are reset to zero after each as click seal bags for the duration of measurements). This measurement is sent to the computer. Again, this technique step is repeated for each series of individuals being measured. provides good estimates of the level of measurement error Second, taking care to choose animals in random sequence, and eliminates the possibility of bias on the part of the and without data on left sides visible, the limbs on the right person performing the measurements. side of each animal’s body are measured and those data recorded. Ideally, the random sequence should be generated Variation among observers using random numbers. If this is not possible, care should be Data obtained by different observers may differ slightly. taken to ensure that the sequence is thoroughly mixed Measurements of asymmetry are relatively robust to small between repeat measurements, for example by physically differences in technique or in judgement regarding the shuffling animals or their containers. locations of landmarks for measurement. This is because the final item of interest is not the absolute sizes of structures, The two steps above are carried out a total of three times but the magnitude of differences between sides. Asymmetry for each animal. Each new set of measurements is taken measurements are sensitive to lack of care in measurement without reference to previously taken sets. Ideally, the three technique. All observers must therefore strive for precision in sets of measurements should be recorded on separate data their measurements. The use of replicate measurements of sheets, so that seeing the values of previous measurements each structure, as we recommend, will also increase precision cannot possibly influence the measurer. of the final mean value. Observers should be carefully trained This protocol is single-blind and, if carried out carefully by a and supervised, and the possibility of observer effects should conscientious operator, should ensure that precision is be taken into account when evaluating data sets collected by maximised by the repeated measurements, while measurer more than one observer. bias is virtually eliminated. DATA ANALYSIS Modifications to the basic protocol We use data analysis techniques similar to those outlined by When measuring living animals in the field, it is important to Palmer and Strobeck (1986), with some modifications. This minimise the amount of handling and the time that animals section presents the techniques used and sample outputs, spend in bags, and to release animals at the point of initial using an artificial example data set we constructed. The data capture after as short a time delay as possible. in that artificial data set are very similar, in levels of asymmetry and patterns of asymmetry, to data collected on One technique that preserves the “blind” nature of replicate real populations of rainforest frogs (Alford et al. 1997). measurements but allows animals to be measured and released immediately after capture, is to work in two-person Within-sample tests of significance and teams, which is also desirable for safety reasons any time field examination of variance components work is being performed in remote and potentially hazardous locations. One person acts as the measurer, and the other We suggest initially analysing data using the ANOVA plus reads the calipers and records the data. A piece of opaque extraction of variance components technique detailed by tape is used to cover the area of the calipers from which Palmer and Strobeck (1986, pages 402-408). This allows readings are taken. The measurer works around the animal separate examination of variance due to directional three times, measuring the left fore- and hindlimbs and then asymmetry, FA + antisymmetry, and measurement error. the right fore- and hindlimbs each time. For each The ANOVAS can be used to test for whether there are measurement, the measurer handles the animal and sets the significant levels of directional asymmetry, and for whether

36 the combination of FA + antisymmetry measured is These ANOVAs must extract effects due to side of the body, significantly above the level produced by measurement error. individual, the interaction of side with individual, replicate, the Using the techniques suggested by Palmer and Strobeck, the interaction of side with replicate, the interaction of individual mean squares for FA+antisymmetry and for error can be with replicate, and the three-way interaction of side, used to estimate a population index of FA. individual, and replicate. The ANOVAs are performed to construct the tables of sums of squares and mean squares To perform these analyses, the data must be laid out as needed to construct tests of hypotheses and estimates of illustrated in Table 1. Separate fully-specified analyses of FA+antisymmetry, not for testing hypotheses. variance are performed for each character measured (forelimb length and hindlimb length) in each sample. Appendix 1 presents the commands needed to analyse the data using the SAS (SAS Institute Inc., 1989) command language. Appendix 2 presents the equivalent syntax TABLE 1: Format of data for analysis using modification of the ANOVA approach of Palmer and Strobeck (1986). Sample refers commands for SPSS version 6.1. These files are also available to the date or place of sampling; in this shortened example, all from the authors , as is a full set of sample data for testing three animals are from the same sample. Individual number can be either sample-specific as in this table or can be any unique the command files or testing other statistical packages, and individual identifier, as from a toe-clipping or PIT tagging full outputs from both SAS and SPSS. program. Replicate indicates which of the three replicate measurements of each limb is presented, side =1 for left, 2 for The results of an analysis of forelimb length in sample 1 of our right (any coding could be used). example data set appear in Table 2. Most of the sums of squares and mean squares are not useful, although finding Sample Individual Replicate SVL Side Forelimb Hindlimb large effects of individual, for instance, might suggest that number length length variation among individuals in body size is such that one of the 1 1 1 47.7 1 11.7 15.8 techniques suggested by Palmer and Strobeck (1986) should 1 1 1 47.7 2 11.5 16.1 be used to correct for body size effects, and the analysis 1 1 2 47.7 1 11.7 15.8 should be redone on the corrected data. The important 1 1 2 47.7 2 11.4 16.2 variance components extracted are the effects of side and the 1 1 3 47.7 1 11.8 15.8 interaction of side with individual, which are related to the 1 1 3 47.7 2 11.5 16.1 magnitudes of directional asymmetry and FA+antisymmetry, 1 2 1 44.9 1 11.4 14.4 respectively, and the interaction of side, individual, and 1 2 1 44.9 2 11.4 14.3 replicate, which serves as an error term for tests of the 1 2 2 44.9 1 11.4 14.4 hypothesis that the FA+antisymmetry effect is not produced 1 2 2 44.9 2 11.4 14.3 1 2 3 44.9 1 11.3 14.3 by measurement error. The mean squares for these 1 2 3 44.9 2 11.3 14.3 components can be used to construct tests of significance as 1 3 1 48.1 1 12 16.6 presented in Table 3. The SAS program in Appendix 1 does 1 3 1 48.1 2 12.1 16.1 this automatically, while the SPSS code (Appendix 2) leaves it 1 3 2 48.1 1 12 16.6 to the analyst to construct the tables by hand. 1 3 2 48.1 2 12 16.3 1 3 3 48.1 1 12 16.6 Table 3 presents results of the tests of the hypotheses that 1 3 3 48.1 2 12.1 16.1 directional asymmetry and FA+antisymmetry of forelimb length in sample 1 of our example data set are significantly greater than zero. There is no evidence for directional TABLE 2: Analysis of variance table for data on forelimb asymmetry (P = 0.3903), while there is a substantial degree asymmetry of 30 individuals in sample 1 of the example data set.Abbreviations: IND_NUM = individual number, REP = of FA+antisymmetry (P < 0.0001). If the test for directional replicate measurement number. asymmetry was significant it would suggest that DI analysis may be difficult to carry out on this population. This is because, when directional asymmetry is present, the degree Factor DF Sum of squares Mean square of asymmetry cannot be used in any simple way as an index SIDE 1 0.0642 0.0642 of DI, as the ideal condition (the ideal degree of asymmetry) IND_NUM 29 60.9591 2.1020 cannot be known. Because the error term for the test of SIDE*IND_NUM 29 2.4491 0.0845 FA+antisymmetry is the interaction of the effects of side, REP 2 0.0034 0.0017 individual, and replicate (= measurement error), the significant SIDE*REP 2 0.0081 0.0041 IND_NUM*REP 58 0.1632 0.0028 test for FA+antisymmetry indicates that measurement error SIDE*IND_NUM*REP 58 0.1186 0.0020 is small relative to the levels of FA+antisymmetry expressed in our example data set. If this test was not significant, it

TABLE 3: Tests of the hypotheses of no directional asymmetry (consistent effect of side on limb length) and no FA+antisymmetry (significant interaction between the effects of individual and body side on limb length) for forelimb data from sample 1 of the example data set. Hypothesis tests were constructed using the mean squares from the ANOVA presented in Table 2.Abbreviations as in that table.

Source DF Mean Square Error term DF Mean Square F Pr > F SIDE (Directional asymmetry) 1 0.0642 SIDE*IND_NUM 29 0.0845 0.76 0.3903 SIDE*IND_NUM (FA + antisymmetry) 29 0.0845 SIDE*IND_NUM*REP 58 0.0020 41.32 0.0001

37 would suggest that the species being examined has levels of FA+antisymmetry so low as to be indistinguishable from TABLE 4: Complete analysis of variance table for forelimb measurements of individuals in both samples in the example measurement error in the sample of individuals being data set.Analysis was carried out to estimate the mean squares examined. This result would not, however, necessarily indicate for the interaction of sample with side and individual, which is related to how much FA+antisymmetry change between samples, that DI analysis should not be incorporated in monitoring for and the interaction of sample, side, individual, and replicate this species; it might simply indicate that, at the present time measurement, which serves as an error term for in the F-statistic in the population of animals measured, very little DI is being examining whether the change in FA+antisymmetry between samples is significant.Abbreviations as in Table 2. expressed. This might simply indicate that the population is healthy and experiencing little environmental stress. Factor DF Sum of Mean After performing significance tests, an index of squares square FA+antisymmetry can be estimated using the following equation (Palmer and Strobeck 1986; page 406): SAMPLE 1 9.4738 9.4738 SIDE 1 0.1440 0.1440 FA+antisymmetry = [(mean square for side X individual SAMPLE*SIDE 1 0.5444 0.5444 number interaction) — (mean square for side X IND_NUM 29 75.1879 2.5927 SAMPLE*IND_NUM 29 42.9712 1.4818 individual number X replicate interaction)] / number of SIDE*IND_NUM 29 4.0977 0.1413 replicate measurements. SAMPLE*SIDE*IND_NUM 29 3.9672 0.1368 REP 2 0.0029 0.0014 For the analysis presented in Table 2, this would be: SAMPLE*REP 2 0.0009 0.0004 FA+antisymmetry = (0.0845 — 0.0020) / 3 = 0.0275. SIDE*REP 2 0.0080 0.0040 SAMPLE*SIDE*REP 2 0.0016 0.0008 This indicates that the mean squared asymmetry of the IND_NUM*REP 58 0.1488 0.0026 forelimbs of animals in sample 1, with differences caused by SAMPLE*IND_NUM*REP 58 0.1441 0.0025 SIDE*IND_NUM*REP 58 0.1103 0.0019 measurement error extracted, is 0.0275 mm. Taking the SAMPLE*SIDE*IND_NUM*REP 58 0.1168 0.0020 square root of this, the mean unsigned asymmetry, corrected for measurement error, is estimated as 0.166 mm.

Among-sample tests for significant would conclude that there is strong evidence that the level of differences in levels of FA+antisymmetry FA+antisymmetry of the forelimbs differs between samples 1 and 2 of our example data set. One of the main reasons for incorporating DI analysis into the monitoring of frog populations is to allow detection of changes in levels of expressed DI over time within Graphs and statistics using means of populations, or comparisons of levels of expressed DI among replicate measurements populations. Such tests can be constructed using the same Once the ANOVA techniques outlined above and discussed data employed for the within-sample tests and estimates in greater detail by Palmer and Strobeck (1986) have been discussed above. Code for testing for differences among used to determine whether significant levels of samples (or calculating the MS necessary to construct tests FA+antisymmetry exist and to estimate mean levels for differences among samples) is incorporated in the SAS corrected for the effects of measurement error, it can be and SPSS programs presented in Appendices 1 and 2. very useful to examine the levels of asymmetry shown by the individuals in each sample. A complete table of the analysis of variance carried out to determine whether levels of FA+antisymmetry of forelimbs The first step in doing this is to take the mean of the three differ between samples in our example data set is presented replicate measurements of the length of each limb. These in Table 4. Although it can be interesting to look at the means can then be used to calculate signed (right length — magnitudes of the variance components in this table, they do left length) and unsigned (absolute value taken of signed) not mean a great deal, particularly if the data are unbalanced, asymmetry for the front and hind limbs of each individual. that is, if different numbers of individuals were measured in The unsigned asymmetry values for the front and hind limbs the different samples. The F-statistic to test for differences can be added together to get an estimate of the total limb between samples is calculated as: asymmetry for each individual. Means and derived measurements for the 60 individuals in both samples of our F = (mean square for the interaction of sample, side, x,y example data set appear in Table 5. and individual)/(mean square for the interaction of sample, side, individual, and replicate measurement). Signed asymmetry values can be used to directly examine the distribution of asymmetry in each sample by producing plots The x and y are the numerator and denominator degrees of of the magnitude and direction of expressed asymmetry freedom, respectively. For our example data set, this leads to: similar to Figure 1. This figure indicates that the asymmetry expressed in forelimb measurements in sample 1 of our F29,58 = 0.1368/0.0020 = 68.4. example data set is true FA, rather than antisymmetry, as the The probability of an F as great or greater than this can be distribution is unimodal (within the limits of resolution looked up in an F-table or can be calculated using a variety of possible with a relatively small data set), and is symmetrical available programs (including the FDIST(F,x,y) function of about a mean of zero. If the ANOVA test for significant levels Microsoft Excel). In this case, it is less than 0.0001, so we of FA+antisymmetry had not been significant, examining a

38 TABLE 5: Means of the three replicate measurements for each limb of each individual in each sample of the example data set, plus derived differences used to examine and plot levels of asymmetry.

Sample Individual Forelimb length Hindlimb length Signed difference (R-L) Unsigned difference Left Right Left Right Forelimb Hindlimb Forelimb Hindlimb Total

1 1 11.733 11.467 15.800 16.133 -0.267 0.333 0.267 0.333 0.600 1 2 11.367 11.367 14.367 14.300 0.000 -0.067 0.000 0.067 0.067 1 3 12.000 12.067 16.600 16.167 0.067 -0.433 0.067 0.433 0.500 1 4 12.067 12.100 16.167 15.900 0.033 -0.267 0.033 0.267 0.300 1 5 11.533 11.600 15.367 15.700 0.067 0.333 0.067 0.333 0.400 1 6 11.567 11.200 15.367 15.033 -0.367 -0.333 0.367 0.333 0.700 1 7 10.333 10.333 13.667 13.833 0.000 0.167 0.000 0.167 0.167 1 8 11.533 11.467 15.400 15.633 -0.067 0.233 0.067 0.233 0.300 1 9 11.000 10.833 14.600 14.667 -0.167 0.067 0.167 0.067 0.233 1 10 11.967 11.533 15.333 15.467 -0.433 0.133 0.433 0.133 0.567 1 11 11.433 11.300 14.700 15.033 -0.133 0.333 0.133 0.333 0.467 1 12 10.600 10.233 13.533 14.233 -0.367 0.700 0.367 0.700 1.067 1 13 10.233 10.067 13.567 13.800 -0.167 0.233 0.167 0.233 0.400 1 14 11.167 11.300 14.767 15.067 0.133 0.300 0.133 0.300 0.433 1 15 10.100 10.467 13.500 13.433 0.367 -0.067 0.367 0.067 0.433 1 16 9.833 9.700 12.900 12.533 -0.133 -0.367 0.133 0.367 0.500 1 17 10.867 11.167 14.367 14.767 0.300 0.400 0.300 0.400 0.700 1 18 11.567 11.367 15.500 15.100 -0.200 -0.400 0.200 0.400 0.600 1 19 10.867 11.367 14.600 14.767 0.500 0.167 0.500 0.167 0.667 1 20 11.500 11.600 15.133 15.567 0.100 0.433 0.100 0.433 0.533 1 21 10.600 10.700 14.100 15.133 0.100 1.033 0.100 1.033 1.133 1 22 12.200 11.633 15.867 15.300 -0.567 -0.567 0.567 0.567 1.133 1 23 10.767 10.800 14.600 14.100 0.033 -0.500 0.033 0.500 0.533 1 24 10.767 10.667 14.600 14.733 -0.100 0.133 0.100 0.133 0.233 1 25 10.167 10.333 13.833 13.933 0.167 0.100 0.167 0.100 0.267 1 26 11.067 10.833 14.400 14.733 -0.233 0.333 0.233 0.333 0.567 1 27 11.533 11.733 15.200 15.533 0.200 0.333 0.200 0.333 0.533 1 28 11.300 11.300 15.200 15.333 0.000 0.133 0.000 0.133 0.133 1 29 10.967 10.833 14.200 14.667 -0.133 0.467 0.133 0.467 0.600 1 30 11.200 11.333 14.667 15.067 0.133 0.400 0.133 0.400 0.533 2 1 11.067 11.067 13.867 14.967 0.000 1.100 0.000 1.100 1.100 2 2 11.000 10.700 15.400 13.867 -0.300 -1.533 0.300 1.533 1.833 2 3 12.200 12.233 16.600 16.867 0.033 0.267 0.033 0.267 0.300 2 4 12.433 12.467 15.667 16.133 0.033 0.467 0.033 0.467 0.500 2 5 11.300 11.200 14.667 14.300 -0.100 -0.367 0.100 0.367 0.467 2 6 11.600 11.500 13.833 14.867 -0.100 1.033 0.100 1.033 1.133 2 7 11.233 10.967 15.833 15.100 -0.267 -0.733 0.267 0.733 1.000 2 8 11.500 12.133 14.867 15.133 0.633 0.267 0.633 0.267 0.900 2 9 11.967 11.633 15.500 15.333 -0.333 -0.167 0.333 0.167 0.500 2 10 10.667 10.700 13.400 14.467 0.033 1.067 0.033 1.067 1.100 2 11 12.000 11.700 15.300 15.467 -0.300 0.167 0.300 0.167 0.467 2 12 12.200 12.000 16.900 16.167 -0.200 -0.733 0.200 0.733 0.933 2 13 12.133 12.333 16.600 16.700 0.200 0.100 0.200 0.100 0.300 2 14 10.733 10.833 14.000 14.367 0.100 0.367 0.100 0.367 0.467 2 15 10.867 10.833 15.033 15.733 -0.033 0.700 0.033 0.700 0.733 2 16 10.300 10.633 13.100 14.700 0.333 1.600 0.333 1.600 1.933 2 17 11.567 12.000 15.467 16.033 0.433 0.567 0.433 0.567 1.000 2 18 11.133 10.667 14.600 13.900 -0.467 -0.700 0.467 0.700 1.167 2 19 11.633 11.733 16.100 16.100 0.100 0.000 0.100 0.000 0.100 2 20 11.300 11.400 14.900 15.000 0.100 0.100 0.100 0.100 0.200 2 21 10.700 10.800 14.167 13.900 0.100 -0.267 0.100 0.267 0.367 2 22 10.700 11.733 14.400 14.400 1.033 0.000 1.033 0.000 1.033 2 23 11.367 11.933 16.433 16.333 0.567 -0.100 0.567 0.100 0.667 2 24 11.033 11.467 15.000 13.967 0.433 -1.033 0.433 1.033 1.467 2 25 10.167 10.567 14.167 14.233 0.400 0.067 0.400 0.067 0.467 2 26 11.500 11.100 14.500 15.067 -0.400 0.567 0.400 0.567 0.967 2 27 11.300 11.667 15.933 16.000 0.367 0.067 0.367 0.067 0.433 2 28 11.900 12.000 16.500 17.033 0.100 0.533 0.100 0.533 0.633 2 29 11.700 12.500 16.200 16.100 0.800 -0.100 0.800 0.100 0.900 2 30 12.033 12.267 16.167 16.633 0.233 0.467 0.233 0.467 0.700

39 plot like Figure 1 would allow determining the magnitude of significant differences between the samples: Sample 2 animals measurement error. A strongly bimodal distribution of signed are substantially less symmetrical than are sample 1 animals. If asymmetry values would indicate that the forelimbs of sample 2 had been taken later in time, from the same animals in sample 1 showed antisymmetry. population as sample 1, this would suggest that environmental stress levels experienced by the population were increasing, Unsigned asymmetry values can be added to give total and would give cause for increased monitoring and efforts to asymmetries for each individual, or can be examined determine the sources of the increased stress. separately for front and hind limbs. Signed asymmetry values should never be totalled, as positive and negative The mean values for individuals can be used to calculate a asymmetries within animals could cancel out, leading to totals variety of indices of asymmetry in addition to the signed and for highly asymmetrical animals that were similar to totals for unsigned asymmetry of each structure and the total unsigned highly symmetrical individuals. Figure 2 presents a plot of the asymmetry. Palmer and Strobeck (1986) discuss these indices means and standard deviations of unsigned asymmetry for and their application in detail. For ease of interpretation, the front and hind limbs of animals in samples 1 and 2 of our however, the simple signed or unsigned asymmetry is example data set, showing why the ANOVA detected excellent, and in most practical applications is probably nearly as powerful as any of the other indices they discuss.

FIGURE 1: Frequency distribution of signed asymmetry of INCORPORATING DI MEASUREMENTS IN forelimbs of animals in sample 1 of the example data set. FIELD MONITORING PROGRAMS We hope that the techniques presented above, plus the 8 computer programs in Appendices 1 and 2, are sufficient to allow amphibian biologists to begin incorporating DI analysis into their regular population monitoring programs. It appears

6 likely that DI analysis can provide a more complete, and probably more sensitive, indication of the “health” of populations than simple monitoring of numbers. This should be particularly true for animals such as frogs, whose numbers 4 at the breeding sites where censuses are usually carried out can fluctuate greatly over short periods, complicating the

Number in class determination of population sizes and status (Pechmann et al. 1991). 2 Using our measurement techniques it should be possible, with some practice, for field workers to take three measurements of each of the limbs of 10-20 animals in one 0 hour. After data on several samples of animals have -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 accumulated, performing an ANOVA followed by Minimum difference in class examination of variance components may reveal that levels of measurement error are very low, on the order of 0.1 mm or less at most. In that case, it is probably reasonable to take FIGURE 2: Mean unsigned asymmetry of the fore- and hindlimbs only a single measurement of each limb on each animal, of animals in both samples of the example data set, illustrating the increase in asymmetry levels between the first and second samples. which would greatly speed up data collection in the field. Measurements should always be taken “blind”, however, 0.65 following one of the protocols we have suggested or some modification of one of them.

One problem that is likely to be encountered in using DI analysis in regular monitoring programs is that of repeatedly measuring the same animals. To be certain of avoiding this, 0.45 animals need to be marked, PIT tagged, or photographed for identification. If individuals cannot be identified, measurements can still be taken, but will need to be interpreted cautiously, as it is not likely that levels of asymmetry change greatly once animals attain adult body sizes and growth slows or ceases. 0.25 If the same individuals are measured repeatedly, data collected on older, less asymmetrical animals might obscure

Mean unsigned asymmetry +/-1 S.E. increases in asymmetry apparent only in younger animals during periods of increasing environmental stress. This problem could be reduced by examining the distributions of signed 0.05 (Figure 1) or unsigned asymmetry for each sample taken. Forelimb Hindlimb Forelimb Hindlimb A tendency towards spreading the tails of the distribution Sample 1 Sample 2 might provide cause for concern even if mean levels have not Measurement Type and Sample changed greatly.

40 CONCLUSION M. A. Lewis, 136-58. Philadelphia, American society for testing and materials. The incorporation of measurements of limb symmetry into regular amphibian monitoring programs, and their analysis Graham, J. H., Freeman D. C., and Emlen J. M. (1994) using developmental stability analysis techniques, holds Antisymmetry, Directional Asymmetry, and Dynamic considerable promise for improving the ability of managers Morphogenesis. In Developmental Instability: Its Origins and biologists to detect declines in the health of populations and Evolutionary Implications. 123-39. Dordrecht, Kluwer before changes in population size become apparent. Academic Publishers. Techniques for the collection and analysis of data on DI, while Leary, R. F. and Allendorf, F.W. (1989) Fluctuating Asymmetry available in the literature, have not been presented in a form As an Indicator of Stress: Implications for Conservation that makes them readily accessible to many potential users. Biology. Trends in Ecology and Evolution, 4: 214-17. By presenting our techniques in detail we hope to make it Moller, A. P.(1996) Parasitism and Developmental Instability of possible for more field biologists and managers to use this Hosts: a Review. Oikos, 77: 189-96. technique, which should increase their ability to detect Palmer, A. R. (1996) Waltzing With Asymmetry. BioScience, 46: deterioration in the health of populations before drastic 518-32. declines in animal numbers occur. Palmer, A. R. and Strobeck, C. (1986) Fluctuating Asymmetry: Measurement, Analysis, Patterns. Annual Review of Ecology ACKNOWLEDGEMENTS and Systematics, 17: 391-421. Our research was supported by grants from the Australian Palmer, A. R. and Strobeck, C. (1992) Fluctuating Asymmetry Research Council and the Cooperative Research Centre for As a Measure of Developmental Stability: Implications of Tropical Rainforest Ecology and Management. Sascha Non-Normal Distributions and Power of Statistical Tests. Frydman provided many long, valuable discussions of the Acta Zoologica Fennica, 191: 57-72. concepts and techniques of DI analysis and interpretation. Pankakoski, E. (1985) Epigenetic Asymmetry As an Ecological Gary Fellers and Britta Grillitsch provided useful comments Indicator in Muskrats. Journal of Mammalogy, 66: 52-57. on much earlier drafts of this manuscript. Parsons, P.A. (1990a). Fluctuating Asymmetry and Stress Intensity. Trends in Ecology and Evolution 5: 97-98. REFERENCES Parsons, P.A. (1990b) The Metabolic Cost of Multiple Alford, R. A., Bradfield, K. S., and Richards S. J. (1997). Increasing Environmental Stresses: Implications for Climatic Change Fluctuating Asymmetry Precedes Frog Population Declines. and Conservation. Trends in Ecology and Evolution, 5: Abstract of talk presented in the Declining Amphibian 315-17. Symposium,Third World Congress of Herpetology, Prague. Pechmann, J.H.K., Scott D.E., Semlitsch R.D., Caldwell J.P.,Vitt L., Clarke, G. M. (1994) Developmental Stability Analysis: an and Gibbons J.W. (1991). Declining Amphibian Populations: Early-Warning System for Biological Monitoring of Water The Problem Of Separating Human Impacts From Natural Quality. Australian Biologist, 7: 94-104. Fluctuations. Science 253: 925-940. Clarke, G. M. (1995) Relationships Between Developmental Sarre, S. (1996) Habitat fragmentation promotes fluctuating Stability and Fitness: Application for Conservation Biology. asymmetry but not morphological divergence in two Conservation Biology, 9: 18-24. geckos. Researches in Population. Ecology. 38, 57-64. Emlen, J. M., Freeman D. C., and Graham J. H.. (1994) Siimäki, P,and Lammi, A. (1998) Fluctuating asymmetry in Nonlinear Growth Dynamics and the Origin of Fluctuating central and marginal populations of Lychnis viscaria in Asymmetry. In Developmental Instability: Its Origins and relation to genetic and environmental factors. Evolution Evolutionary Implications. 79-98. Dordrecht, Kluwer 52:1285-1292. Academic Publishers. Tracy, M., Freeman D. C., Emlen J. M., Graham J. H., and. Freeman, D. C., Emlen J. M., Graham J. H.,. Mara R. L,Tracy M., Hough R. A. (1995) Developmental Instability As a and Alados C. L.. (1996) Developmental Instability as a Biomonitor of Environmental Stress. In Biomonitors and Bioindicator of Ecosystem Health. In Proceedings: Biomarkers As Indicators of Environmental Change. Editor Shrubland Ecosystem Dynamics in a Changing F. A. Butterworth, L. D. Corkum, and J. Guzman-Rincon, Environment. Compilers J. R. Barrow, E. D. McArthur, R. E. 313-37. New York, Plenum Press. Sosebee, and R. J. Tausch, 170-177. Ogden, UT, U.S. Tsubaki,Y. (1998) Fluctuating asymmetry of the Oriental Fruit Department of Agriculture, Forest Service, Intermountain Fly (Dacus dorsalis) during the process of its extinction Research Station. from the Okinawa Islands. Conservation Biology 12:926- Graham, J. H. (1992) Genomic Coadaptation and 929. Developmental Stability in Hybrid Zones. Acta Zoologica Valentine, D.W., Soulé, M. E. and Samollow P., (1973) Fennica, 191: 121-31. Asymmetry analysis in fishes: a possible statistical indicator Graham, J. H., Emlen, J. M., and Freeman, D. C. (1993a) of environmental stress. Fisheries Bulletin 71: 357-370 Developmental Stability and Its Applications in Zakharov,V. M. (1992) Population Phenogenetics: Analysis of Ecotoxicology. Ecotoxicology, 2: 175-84. Developmental Stability in Natural Populations. Acta Graham, J. H., Freeman D. C., and Emlen J. M. (1993b) Zoologica Fennica, 191: 7-30. Developmental Stability: A Sensitive Indicator of Populations Under Stress. In Environmental Toxicology and Risk Assessment. Editors W. G. Landis, J. S. Hughes, and

41 APPENDIX 1 if ((_source_ eq ‘SIDE’) and (_type_ eq ‘SS1’)) then select; SAS code for full analysis of DI measurements taken on two when(_name_ = ‘FL’) do; flms0 = ss/df; fldf0=df; end; samples of animals and stored as in Table 1. For additional when(_name_ = ‘HL’) do; hlms0 = ss/df; hldf0=df; end; samples, some of the code would be duplicated. Readers end; familiar with the SAS command language should be able to if ((_source_ eq ‘SIDE*INDID’) and (_type_ eq ‘SS1’)) then do this easily. Note that the data within each sample must be select; balanced: each animal must have three measurements taken when(_name_ = ‘FL’) do; flms1 = ss/df; fldf1=df; end; of each limb. If the data are not balanced, the hypothesis tests when(_name_ = ‘HL’) do; hlms1 = ss/df; hldf1=df; end; and calculations of variance components done within each end; sample will be incorrect, since the Type I sums of squares will if ((_source_ eq ‘SIDE*INDID*REP’) and (_type_ eq ‘SS1’)) be order-dependent. The final section, which tests for then differences in levels of FA+antisymmetry between samples, select; should produce correct results even if there are not the when(_name_ = ‘FL’) do; flms2 = ss/df; fldf2=df; end; same numbers of animals in each sample. when(_name_ = ‘HL’) do; data all; hlms2 = ss/df; /*change to path to your data file*/ hldf2 = df; infile ‘d:\sas\asym\example\data.prn’ end=eof; flfdir = flms0/flms1; input; /*skips first line with variable names*/ flffa = flms1/flms2; do while(not eof); flfdirp = 1-probf(flfdir,fldf0,fldf1); input date ddmmyy8. sample indid rep svl side fl hl; flffap = 1-probf(flffa,fldf1,fldf2); output; hlfdir = hlms0/hlms1; end; hlffa = hlms1/hlms2; stop; hlfdirp = 1-probf(hlfdir,hldf0,hldf1); hlffap = 1-probf(hlffa,hldf1,hldf2); /*if you have data for more than one sample, this sorts them flasym = (flms1 — flms2) /3; so that SAS can do all the calsulations in one step*/ hlasym = (hlms1 — hlms2) /3; proc sort data=all; output; by sample; end; end; /*the following GLM does the main work, calculating oldsamp = sample; the SS and MS needed to estimate FA + antisymmetry retain flms0 hlms0 fldf0 hldf0 fldf1 hldf1 according to Palmer and Strobeck, and also testing fldf2 hldf2 flms1 flms2 hlms1 hlms2 for whether directional asymmetry and FA+antisymmetry oldsamp; are significantly greater than 0*/ run; proc glm data=all outstat=stats; by sample; proc print data=temp; var sample flasym hlasym; class side indid rep; title ‘Palmer and Strobeck (1986) estimates of model fl hl=side|indid|rep; FA+antisymmetry for each character’; test h=side e=side*indid; run; test h=side*indid e=side*indid*rep; run; proc print data=temp; /*following is optional—it calculates and prints var sample flfdir fldf0 fldf1 flfdirp flffa fldf1 fldf2 flffap; tables showing the FA+antisymmetry estimates title ‘tests for significance of directional asymmetry and for forelimb length and hindlimb length FA+antisym for FL’; and the tables of the tests of hypothesis run; for directional and FA+antisymmetry for each sample*/ proc print data=temp; data temp; var sample hlfdir hldf0 hldf1 hlfdirp hlffa hldf1 hldf2 hlffap; set stats; title ‘tests for significance of directional asymmetry and if (oldsamp ne sample) then do; FA+antisym for HL’; flms0 = .; run; hlms0 = .; fldf0 = .; /*the following GLM is not according to Palmer and hldf0 = .; Strobeck but is legitimate—it tests for whether the fldf1 = .; FA+antisymmetry component of variance differs between hldf1 = .; samples*/ fldf2 = .; proc glm data=all; hldf2 = .; class sample side indid rep; flms1 = .; model fl hl=sample|side|indid|rep; flms2 = .; test h=sample*side*indid e=sample*side*indid*rep; hlms1 = .; title ‘analyses for difference in FA + Antisym component hlms2 = .; between samples’; end; run;

42 APPENDIX 2 /ANALYSIS hll /DESIGN side ind_num ind_num*side ind_num*rep*side . SPSS code for full analysis of DI measurements taken on two FILTER OFF. samples of animals and stored as in Table 1. This code USE ALL. assumes that the data have already been read in or loaded. EXECUTE . For additional samples, some of the code would be MANOVA duplicated. Readers familiar with the SPSS command language fll BY side(1 2) ind_num(1 30) rep(1 3) sample(1 2) should be able to do this easily. Note that the data within /NOPRINT PARAM(ESTIM) each sample must be balanced: each animal must have three /METHOD=UNIQUE measurements taken of each limb. If the data are not /ERROR WITHIN+RESIDUAL balanced, the hypothesis tests and calculations of variance /ANALYSIS fll components done within each sample will be incorrect, since /DESIGN side ind_num sample side*ind_num the Type I sums of squares will be order-dependent. The final side*ind_num*rep sample*ind_num section, which tests for differences in levels of side*sample*ind_num side*sample*ind_num*rep . FA+antisymmetry between samples, should produce correct MANOVA results even if there are not the same numbers of animals in hll BY side(1 2) ind_num(1 30) rep(1 3) sample(1 2) each sample. Unlike the SAS code in Appendix 1, which /NOPRINT PARAM(ESTIM) carries out hypothesis tests and calculates estimates of /METHOD=UNIQUE corrected FA+antisymmetry values, all hypothesis tests and /ERROR WITHIN+RESIDUAL estimates would have to be constructed by hand using the /ANALYSIS hll MS reported by SPSS. See Table 3 and the text for examples. /DESIGN side ind_num sample side*ind_num This may simply reflect the authors’ relative ignorance of the side*ind_num*rep sample*ind_num SPSS syntax language. side*sample*ind_num side*sample*ind_num*rep . USE ALL. COMPUTE filter_$=(sample = 1). VARIABLE LABEL filter_$ ‘sample = 1 (FILTER)’. VALUE LABELS filter_$ 0 ‘Not Selected’ 1 ‘Selected’. FORMAT filter_$ (f1.0). FILTER BY filter_$. EXECUTE . MANOVA fll BY side(1 2) ind_num(1 30) rep(1 3) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL /ANALYSIS fll /DESIGN side ind_num ind_num*side ind_num*rep*side . MANOVA hll BY side(1 2) ind_num(1 30) rep(1 3) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL /ANALYSIS hll /DESIGN side ind_num ind_num*side ind_num*rep*side . USE ALL. COMPUTE filter_$=(sample = 2). VARIABLE LABEL filter_$ ‘sample = 2 (FILTER)’. VALUE LABELS filter_$ 0 ‘Not Selected’ 1 ‘Selected’. FORMAT filter_$ (f1.0). FILTER BY filter_$. EXECUTE . MANOVA fll BY side(1 2) ind_num(1 30) rep(1 3) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL /ANALYSIS fll /DESIGN side ind_num ind_num*side ind_num*rep*side . MANOVA hll BY side(1 2) ind_num(1 30) rep(1 3) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL

43 An assessment of frog declines in wet subtropical Australia

Harry Hines1, Michael Mahony2 and Keith McDonald3

ABSTRACT Although a lack of long term systematic baseline information on presence/absence or relative The decline of frog populations is a worldwide abundance has hampered our assessment, the phenomenon and a major conservation issue in species of concern fall into three groups, based on Australia. In south-eastern Australia the threats. Six species of stream breeding frogs have understanding of the frogs affected, their ecology, suffered declines from unknown causes. Two of patterns of decline and the causal agents is these, endemic to the wet subtropics, have not been generally poor. The wet subtropical region seen since the early 1980s. Eight species of mesic encompasses the coast and ranges from about forest frogs, that are not dependant upon streams Newcastle north to Rockhampton. Twenty-three for breeding, show no evidence of declines. They are species of conservation concern from this region of conservation concern because of their rarity. The are examined. The extent of declines and known or third group, frogs of the lowlands, are increasingly at likely threatening processes for each species, is risk due to loss or degradation of habitat. investigated by comparing the distribution of records pre-1990 and 1990 onwards, and Knowledge of the population dynamics and biology considering other relevant information. The state of the frogs in this review is generally poor. This of knowledge on the ecology of these species is needs to be greatly improved to ensure conservation assessed and the threats to them summarised, as a of these species, and in particular, for determining basis for identifying what additional information is the cause(s) of decline in stream breeding frogs. required for successful species recovery. There is an urgent need to establish recovery processes for the habitat of wallum or ‘acid’ frogs.

1 Queensland Parks and Wildlife Service, Conservation Resource Unit, PO Box 42 Kenmore Qld 4069. 2 Department of Applied Science and Technology,The University of Newcastle, Callaghan NSW 2308. 3 Queensland Parks and Wildlife Service,Wet Tropics District Office, PO Box 834 Atherton Qld 4883.

44 INTRODUCTION TABLE 1: Rare or threatened frogs of the wet subtropical Over the past decade there has been an increasing awareness of region of Australia massive declines in frog populations in many parts of the world (e.g. Tyler and Davies 1985b, Blaustein and Wake 1990; Richards Species Qld1 NSW2 Comm3 Action ANZECC5 et al. 1993; Drost and Fellars 1996; Lips 1998). Despite the Plan4 disappearance of several species and declines in populations of Assa darlingtoni RV many others in south-eastern Australia since the 1970’s (Tyler Crinia tinnula VV I and Davies 1985b; Osborne 1989; Czechura and Ingram 1990; Lechriodus fletcheri R Watson et al. 1991, Ingram and McDonald 1993), understanding #Mixophyes balbus VV of the problem is poor (Tyler 1997). For most species affected Mixophyes fleayi EV E there has been little or no synthesis of information across their Mixophyes iteratus EV E entire range, nor in a framework that allows easy comparison Kyarranus kundagungan RV among species in a region. Reviews of the status of frogs in Kyarranus loveridgei RV south-eastern Australia have either focused on small Kyarranus sphagnicolus VI geographical areas (e.g. — Mahony 1993), Rheobatrachus silus EEEE single species (e.g. Litoria spenceri — Gillespie and Hollis 1996; Taudactylus diurnus EEEE Pseudophryne corroboree — Osborne 1989) or groups of closely Taudactylus pleione VV + related species (e.g. L. aurea complex — Pyke and Osborne Litoria aurea EE # 1996). Others have been constrained by political or other Litoria booroolongensis EI Litoria brevipalmata RV I biologically artificial boundaries (e.g. Ingram and McDonald 1993; Litoria cooloolensis R Ehmann 1997a;White and Pyke 1996). In this and other papers Litoria freycineti VI in these proceedings we examine the available data on the Litoria olongburensis VV V distribution and declines of species throughout their range, based Litoria pearsoniana EI on broad ecological boundaries of the rainforests of mid- and Litoria revelata R north-eastern Queensland (Qld) (McDonald and Alford 1999), #Litoria subglandulosa VV I the wet subtropical region of eastern Australia (this review), streams in the temperate zone of south-eastern Australia Source: (Gillespie and Hines 1999) and the subalpine region (Osborne 1 Queensland Nature Conservation (Wildlife) Regulation 1994 et al. 1999) to overcome these problems. Only a 2 New South Wales Threatened Species Conservation Act 1995 few species of concern, such as Adelotus brevis and L. lesueuri, 3 Commonwealth Endangered Species Protection Act 1992 4 Tyler 1997 span two or more of these biological regions. 5 The ANZECC list of Threatened Vertebrate Fauna (1995)

We define the wet subtropical region of Australia as those Status abbreviation (refer to relevant source for definitions): parts of the summer rainfall, subtropical climate zone E = endangered (Bureau of Meteorology 1989, Figure 5), that receive in V = vulnerable excess of 800 mm median annual rainfall (Bureau of R = rare Meteorology 1989, Figure 2), but excluding the upland I = insufficiently known temperate areas of the NSW . The status of species listed in this table are reviewed here except This region encompasses the escarpment, foothills and coastal for those marked: lowlands south from Rockhampton Qld to about Newcastle + reviewed by Mahony (this volume) NSW. The region includes large areas of subtropical # reviewed by Gillespie and Hines (this volume) rainforests, Melaleuca swamps and wallum, three environments especially important for frogs in eastern Australia. The region experiences wet, warm to hot summers, with drier, mild winters (Bureau of Meteorology 1989). and monitoring i.e. A. brevis and L. lesueuri (Hines and Mahony unpubl. data); and/or c) anecdotal evidence of past declines Over 50 species of frogs are known from the wet subtropical i.e. A. brevis and L. lesueuri (Ingram and McDonald 1993), region, and about half of these are endemic.We have chosen and Mixophyes fasciolatus (Corben in McDonald 1991). to assess 23 species of conservation concern. These include Four additional rare or threatened species known from wet 17 species that are listed as Rare,Vulnerable or Endangered subtropical Australia (Table 1) are not assessed here because under state or federal legislation (Table 1), three undescribed the majority of their distribution is elsewhere. They are species which meet criteria for listing under one of these discussed in other papers in these proceedings (Gillespie and categories and three stream breeding species potentially in Hines 1999; Mahony 1999; Osborne et al. 1999). decline. Some of these species are listed because of their rarity and or restricted distributions, while others are Our aim is to provide an overview of the status of frogs of threatened by loss of habitat. However, a large proportion conservation concern in the wet subtropical region of are stream breeding species that have suffered population Australia.We examine available knowledge on the declines or have disappeared, without a clear understanding distribution, ecology and threats to these species, as a basis of the causal agent(s). Inclusion of three additional stream for identifying additional information needed for their breeding species in this review was based on: a) conservation. For each species we assess: demonstrable decline in other regions i.e. A. brevis (Gillespie and Hines 1999); and/or b) an apparent decline in distribution or abundance based on data from recent surveys

45 1. recorded distribution, subtropical Australia is poor. Despite increasing awareness of 2. current distribution with discussion of declines, declines during the 1980’s there was little survey or monitoring of frog populations. However, since 1989, large 3. threats, scale biodiversity inventories have been undertaken in the 4. knowledge of biology, and wet subtropical region such as regional forest assessments in 5. conservation status. south-eastern Qld and north-eastern NSW. During this period there were also targeted frog survey and monitoring METHODS programs (e.g. Ingram and McDonald 1993, Ehmann 1997a, Distributional data for each of the 23 species of concern Mahony 1996, Hines unpubl. data, McDonald unpubl. data) were collated from existing computer databases and and a number of site or species specific studies (e.g. Mahony augmented with records from unpublished reports, published and Knowles unpubl. data on Mixophyes and Kyarranus). literature and naturalists. Although not complete, this process Major sources of information on the current distribution enabled a large data set to be rapidly assembled. Digital data (up to December 1997) of frogs are presented in Table 2. were received from the Australian Museum (AM), Many sites where rare or threatened frogs were recorded Queensland Museum(QM) and South Australian Museum prior to 1990 were revisited during this time. (SAM), the National Museum of Victoria (NMV),The Australian National Wildlife Collection (ANWC), Atlas of There are a number of potential limitations in the use of NSW Wildlife (NSW Atlas), Atlas of Victorian Wildlife (Vic species records such as these including, biases in survey Atlas) and South East Queensland Regional Forest effort, undetected errors in georeferences or species Assessment (SEQRFA) databases.Vetting of data sets for identification and poor taxonomic resolution of some species obvious species mis-identification and coarse georeference groups.When necessary these issues are discussed in the errors was aided by examining maps of the records for each species accounts, where we also draw on additional species. Many records lacked georeferences so where an information (e.g. relative abundance measures) to aid in unambiguous locality description was provided a assessments of declines. georeference was obtained from maps or gazetteer. A brief account is provided for each species, summarising or To help assess regional population declines, maps for each assessing available information on historical and current species were produced showing the distribution of records distribution (1990-1997), threats, biology and conservation pre-1990 and 1990 onwards. Nineteen-ninety was selected status. Species have been grouped according to similarities in for two reasons. Apart from the documented disappearances their ecology. of Taudactylus diurnus and Rheobatrachus silus in the late 1970’s to early 1980’s (see species accounts below), knowledge of the temporal patterns of frog declines in

TABLE 2: Major sources of information on current distribution of frogs of the wet subtropical region of Australia

Source Type of data Target species or areas Museums specimens all species, throughout Queensland Parks and Wildlife survey, monitoring threatened frogs, especially stream-breeding species Service (QPWS) and research (Hines unpubl. data) and Fraser Island (McDonald unpubl. data, QPWS unpubl. data) Queensland Department of monitoring historical locations of R. silus and T. diurnus (Borsboom unpubl. data) Natural Resources (QDNR) QPWS/QDNR survey surveys in southeast Queensland for regional conservation planning QDNR/QPWS monitoring and research T. pleione Brisbane Frog Society monitoring T. diurnus at Mount Glorious NSW National Parks and survey biodiversity surveys, east coast NSW, particularly NE, for regional Wildlife Service conservation planning State Forests of NSW survey environmental impact assessments, east coast of NSW, particularly NE University of Newcastle survey, monitoring most species, especially L. pearsoniana sensu lato,A. darlingtoni, and research Kyarranus spp, and Mixophyes spp throughout study area, and the Watagan Mountains area University of Queensland monitoring and research L. pearsoniana at Mount Glorious Australian National University survey and research stream breeding species, particularly L. pearsoniana and M. fasciolatus (Parris unpubl. data) Southern Cross University survey, monitoring various species, mostly in far NE NSW and research Various biologists survey, research and all species, throughout incidental records

46 SPECIES ACCOUNTS Cooloola sedgefrog Litoria cooloolensis Coastal lowland species Distribution: Restricted to Qld, from Lake Woongeel, Fraser (“wallum” or “acid” frogs) Island (24° 53´S 153° 14´E — QPWS unpubl. data), through the sandmasses of Cooloola, with a disjunct population on Wallum or acid frogs (Ingram and Corben 1975a, Kikkawa et North Stradbroke Island (27° 32´S 153° 29´E — AM and al. 1979) inhabit low nutrient soils (mostly deep sands) of near QM specimens). The species has not been recorded on coastal lowlands and sand islands.Vegetation types typical of Moreton Island despite targeted surveys (McDonald and these environments include heathland, Melaleuca swamp, Stewart QPWS unpubl. data). Figure 2. sedgeland and Banksia woodland. Recent survey and monitoring results (Table 2) indicate that populations are Current distribution: As above, except for localised relatively stable in protected habitat. However these species extermination of populations due to loss or fragmentation are at risk from continuing loss of habitat through clearing for of habitat. agriculture, pine plantations, housing and infrastructure such as Biology and threats: Information on breeding biology and canal development, drainage projects and transport corridors. habitat usage of L. cooloolensis is provided by Liem (1974) and They occur in an area with the highest growth rate of human James (1996). Some physiological aspects of L. cooloolensis population in Australia. In the period 1974-1989, 33% of the larval development in acidic, ion-deficient breeding waters 1974 bushland cover of coastal south-eastern Qld mainland have been investigated by Meyer (1997). Threats from high was cleared, including 50% of Melaleuca forest and 34% of human visitation to the fresh water lake habitats important heathland (Catterall and Kingston 1993). Other threats include for L. cooloolensis include trampling of reed beds and pollution habitat degradation through changes to hydrological regimes, of water (e.g. increased nutrients, sunscreen residues) increased nutrients or sediments (e.g. storm water, sewage), (James 1996). weed invasion, inappropriate fire management, competition from invading frog species and predation from introduced fish James (1996) examined the genetic structure of this species (Gillespie and Hero 1999). Regular monitoring of these species and found that populations on North Stradbroke Island has occurred on Fraser Island, where biyearly surveys of sites diverged significantly. On Stradbroke Island it is restricted to have been undertaken since 1992 (QNPWS unpubl. data). reed beds in and around freshwater lakes and swamps.Water extraction for domestic purposes and sandmining has, and Wallum froglet Crinia tinnula will continue to, significantly impact on this population Distribution: Lake Woongeel, Fraser Island SEQ (24° 53´S 153° through loss of habitat or degradation resulting from changes 14´E — QPWS unpubl. data), south to Kurnell MENSW (34° to hydrological processes and water quality. 02´S 151° 13´E). Figure 1. The species is widespread on Fraser There is no information on population size, structure, or Island (QPWS and McDonald unpubl. data, QM specimens) dynamics, non-breeding habitat requirements or factors contrary to Ehmann’s (1997c) statement that it was apparently limiting distribution. absent from the island. There are records of C. tinnula on the mainland north of Woodgate, to Littabella National Park near Conservation status: The species is currently considered Bundaberg SEQ (24° 41´S 152° 05´E — QPWS unpubl. data Rare in Qld. Some populations (Stradbroke Island) of L. — indicated by a question mark in Figure 1). The identity of cooloolensis are under threat, so its legal status should be these populations requires confirmation. reviewed both in Qld and nationally (currently not listed). Litoria cooloolensis is not the subject of a recovery process. Current distribution and threats: As above, except for localised extermination of populations due to loss or Wallum rocketfrog Litoria freycineti fragmentation of habitat. The recent range extensions of this species (i.e. new records north and south of its previously Distribution: Northern end of Fraser Island SEQ (25° 6´S known range) are almost certainly due to increased survey 153° 11´E — QPWS unpubl. data), south to Jervis Bay effort using improved field identification characters and aids MENSW (35° 08´S 150° 43´E — ANWC, NMV, QM (field guides and audio recordings), rather than recent specimens). Figure 3. Several inland records (from far migration or deliberate or accidental introductions, as NENSW and MENSW have not been included in Figure 3 suggested by Ehmann (1997c). as they require confirmation).

Biology: Not well documented. Some information on Current distribution and threats: As above except for breeding biology and habitat usage is presented by Straughan localised extermination of populations due to loss or and Main (1966), Ingram and Corben (1975a) and Ehmann fragmentation of habitat. (1997c). Meyer (1997) investigated some physiological Biology: Poorly known. There is general information on aspects of tadpole development in the acidic ion-poor waters habitat use (Ingram and Corben 1975a) and breeding biology typical of this species’ habitat. There is no information on (Straughan 1966). Meyer (1997) investigated some population size, structure, dynamics or genetics, non-breeding physiological aspects of tadpole development in acidic ion- habitat requirements or factors limiting distribution. deficient waters typical of this species’ habitat. There is no Conservation status: The data presented here indicate that information on population size, structure, dynamics or the current legislative status of C. tinnula in Qld and NSW, as genetics, non-breeding habitat requirements or factors Vulnerable, is appropriate. Accordingly its status nationally limiting distribution. (currently not listed) should be reviewed. Crinia tinnula is not the subject of a recovery process.

47 Conservation status: The data presented here indicate that suitable conditions of the sites where McDonald (1974) first the current legislative status of L. freycineti in Qld as recorded this species in Queensland (the two most south- Vulnerable, is appropriate. Given the rate of habitat loss westerly records; Figure 5). Over the past five years additional throughout its range, its status in NSW and nationally should populations have been located as a result of increased survey be reviewed (currently not listed). Litoria freycineti is not the effort and better understanding of the factors affecting subject of a recovery process. detectability of this species. The impacts of habitat disturbance on L. brevipalmata are not known. The lack of records from Wallum sedgefrog Litoria olongburensis cleared environments and the decline at suggests that L. brevipalmata is vulnerable to major habitat Distribution: Lake Woongeel, Fraser Island (24° 53´S 153° perturbations. Given the rate of habitat loss in the lowlands (e.g. 14´E — QPWS unpubl. data), south to near Woolgoolga (30° Catteral and Kingston 1993), some localised populations must 08´S 153° 11´E — Liem and Ingram 1977; AM, QM have been exterminated before their existence was recorded. specimens) including Bribie, Moreton and North Stradbroke Other disturbances to its habitat include inappropriate fire Islands, SEQ (Figure 4). The records from Woolgoolga regimes, timber harvesting, grazing by domestic stock, weed (above), and Yamba (NSW NPWS invasion and changes to hydrological regimes and water quality. 1995, NSW Atlas, Hines unpubl. data) show that the It is not possible to assess the effects of these disturbances as distribution of this species extends much further south than knowledge on the biology of L. brevipalmata is poor. stated by Ehmann (1997e). Biology: Limited information. Anstis (1994) described larval Current distribution and threats: As above except for development and Czechura (1978), McDonald (1974), localised extermination of populations due to loss or McEvoy et al. (1979), Nattrass and Ingram (1993) and Aridis fragmentation of habitat. (1997) provide qualitative information on habitat at some Biology: Limited information on breeding biology and habitat sites. Ledlin (1997) provides information on larval morphology usage of L. olongburensis is provided by Liem and Ingram and ecology, and the habitat characteristics of the species in (1977), James (1996) and Ehmann (1997e). Meyer (1997) mid-eastern NSW. Genetic studies (allozymes and DNA) of investigated some of the physiological aspects of larval eleven populations in north-eastern NSW and one in south- development in acidic ion-deficient waters typical of this eastern Qld (north from Watagan State Forest to Border species’ habitat. Considerable genetic structuring occurs Ranges National Park) revealed no cryptic species (Donnellan within L. olongburensis reflecting its disjunct distribution and Foster 1997). Additional samples are required from Qld (James 1996). There is no information on population size, to fully assess genetic diversity. There is no information on structure or dynamics, non-breeding habitat requirements population size, structure, or dynamics, non-breeding habitat or factors limiting distribution. requirements or factors limiting distribution.

Conservation status: The data presented here indicate that Conservation status: The legislative status of L. brevipalmata the current legislative status of L. olongburensis in Qld and in Qld is Rare and in NSW it is Vulnerable. Litoria NSW, as Vulnerable, is appropriate. Accordingly its status brevipalmata is not the subject of a recovery process. nationally should be reviewed (currently not listed). Litoria olongburensis is not the subject of a recovery process. Mesic forest species Within subtropical Australia, mesic forest frogs are those that Sclerophyll forest species are predominantly dependent upon rainforest and wet Green-thighed frog Litoria brevipalmata sclerophyll forest communities. This vegetation is typically in the foothills and ranges of the Great Divide. In contrast to Distribution: Patchily recorded from low to mid altitudes the habitat of the preceding frogs, the mesic forests have from Cordalba State Forest SEQ (25° 10´S, 152° 10´E — generally been cleared at a slower rate in recent times. In the Hines unpubl. data) south to Darkes Forest, MENSW (34° past however, some large areas of rainforests on flat 14´S, 150° 55´E — AM specimens) (Figure 5). It is a difficult topography have been lost. Significant extensive disturbances species to survey as it breeds for only a short period, usually to remaining habitat include timber harvesting, hydrological after heavy spring and summer rains. changes, cattle grazing, altered fire regimes, feral animals and weed invasion. The effects of these disturbances are largely Current distribution and threats: The species has declined in unknown but from the knowledge of the frogs’ ecology they the Ourimbah Creek area since the time it was first detected in are likely to be highly significant in many situations. 1972 (Mahony 1996). This site is visited by many herpetologists from Sydney and the Central Coast and the large populations Within the frog fauna of the mesic forests there are three reported in the early 1970’s have not been observed in recent groups that differ in breeding biology; non-stream breeding, years. It is most likely that this decline is due to clearing for facultative stream breeding and obligate stream breeding. agricultural development (Mahony 1996). However, several other Species falling into each of these groups have undoubtedly species (e.g. Mixophyes balbus, M. iteratus, and L. aurea) previously declined due to localised loss or degradation of habitat. A known from this and nearby areas have declined dramatically number of stream breeding species have suffered additional without any apparent reason (Mahony 1993). Therefore it is not declines from unknown causes, that are apparently not possible to discount other, unknown threat(s) as the cause of related to habitat disturbances. Unlike temperate riverine decline of L. brevipalmata in the Ourimbah Creek area. frogs (Gillespie and Hines 1999), there is no evidence of introduced fish causing declines of frogs in subtropical Elsewhere there are no reports of declines or disappearances streams. Most streams either lack fish or support a small of this species.We are not aware of any recent surveys under

48 FIGURE 1: Distribution of wallum froglet Crinia tinnula. Closed FIGURE 2: Distribution of Cooloola sedgefrog Litoria cooloolensis circles with ‘?’ are records 1990 onwards with questionable taxonomic identity.

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

FIGURE 3: Distribution of wallum rocketfrog Litoria freycineti FIGURE 4: Distribution of wallum sedgefrog Litoria olongburensis

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

49 number of native species. However the impact of introduced Mountain frogs Kyarranus species fish is potentially very serious and we recommend strict Three Kyarranus species have been formally described; controls on the introduction or movement of any fish species K. kundagungan, K. loveridgei and K. sphagnicolus, all from in south-eastern Australia (see Gillespie and Hero 1999 for south-eastern Qld and north-eastern NSW. Recent studies of detailed discussion). A fourth group, includes a single species, allozyme variation (Knowles 1994; Mahony and Knowles Taudactylus pleione, because its breeding biology remains 1994) revealed that there are six allopatric species. All are unknown. rare with restricted distributions; at least two species are known from less than 10 sites each. Members of the genus Mesic forest species — non-stream breeding Kyarranus are sometimes placed in the genus with Pouched frog Assa darlingtoni P. frosti. For the purposes of this study we have combined all records of Kyarranus because: Distribution: Disjunct distribution from the Conondale and Blackall Ranges SEQ (26° 34´S 152° 52´E — NMV specimen a. taxonomic revision is pending; from near Kondalilla) south to the Dorrigo Plateau NENSW b. there has been considerable confusion over the identity of (30° 22´S 152° 44´E — AM specimen, NSW Atlas, NSW individuals or populations in the past and NPWS 1994a), with populations on the D’Aguilar, Main, c. each species occupies similar ecological niches (i.e. very Gibraltar and Border Ranges (Figure 6). similar habitat requirements and breeding biology). Current distribution: As above. The lack of recent records Distribution: Six allopatric species occurring in the from Mount Nothophagus on the Qld/NSW border probably headwaters and seepage zones in mid to high elevation mesic reflects a lack of survey effort. Under suitable conditions large forest from Mount Mistake SEQ (27° 53´S 152° 21´E — numbers can be heard calling in the Conondale and Eastern Ingram and Corben 1975b) south to the Comboyne Plateau Border Ranges (Hines unpubl. data, Ehmann 1997b). The in NENSW (31° 40´S 152° 18´E — AM specimens, NSW small, recently discovered population on D’Aguilar Range, Atlas), Figure 7. For a detailed description of the distribution west of Brisbane, is regularly detected during monitoring of of each species refer to the “current distribution” section of stream frogs undertaken by the University of Qld species profiles 7 — 12 in Ehmann (1997a).Wotherspoon’s (Cunningham pers. comm.). (1981) record from Barrington Tops National Park (151° 32’E Threats: In the past considerable clearing of this species’ 32° 10’S) is rejected because of probable mis-identification habitat occurred (e.g. Dorrigo and Maleny plateaux) and (Knowles and Mahony 1997d) and lack of corroborating much of its remaining habitat has been logged. However, the records from the area despite intensive survey work (e.g. majority of this habitat is now reserved or excluded from Knowles and Mahony 1997d; NSW NPWS 1994a). timber harvesting. There is some evidence that timber Current distribution: As above. There are no documented harvesting may have negative impacts on this species declines for the genus. Under suitable weather and seasonal (Lemckert 1999). conditions members of the genus have been readily located at Biology: The biology of A. darlingtoni is reasonably well sites where previously recorded (Hines unpubl. data; Knowles documented, in particular its unusual breeding (Straughan and and Mahony unpubl. data). Targeted surveys for this genus Main 1966;Tyler 1972; Ingram et al. 1975; Ehmann and Swan have located many new populations (e.g. Smith et al. 1989a,b; 1985; Mahony 1992). In north-eastern NSW its distribution NSW NPWS 1994a; Knowles and Mahony 1997a,b,c,d,e,f). was statistically modelled using a pool of 24 environmental Threats: In some areas considerable clearing of these species’ predictors (NSW NPWS 1994a). The model highlighted the habitat occurred (e.g. Acacia, Dorrigo and Comboyne species’ preference for landscapes dominated by mesic plateaux) and much of the remaining habitat has been logged. forests, in areas with mild temperatures and moderate to However, the majority of this habitat is now reserved or high rainfall. This, in conjunction with information on local excluded from timber harvesting. These species are likely to distribution and habitat association provided by Czechura be susceptible to upstream disturbances that affect (1991), Ingram et al. (1975), and Ehmann (1997b), dispels the hydrological regimes or water quality but there has been no misconception that A. darlingtoni is closely associated with research to assess this. Direct damage to breeding habitat by Antarctic Beech (Nothophagus moorei) forest at high altitude domestic stock has been observed at a number of sites (e.g. (Straughan and Main 1966; Cogger 1996). Knowles and Mahony 1997e; Hines pers. obs.). There is no information on population size, structure or Biology: The factors affecting distribution have been dynamics. A limited genetic study of the three major disjunct investigated at local (Knowles 1994,Webb 1989) and regional populations in NSW (Eastern Border Ranges, scales (NSW NPWS 1994a). The eggs and larval development and Dorrigo Plateau) found no evidence of species level are well documented (Anstis 1981; deBavay 1993; Ingram and differences among populations (Donnellan and Mahony, Corben 1975b). Diet in one population has been studied unpubl. data). (Webb 1989). Little is known about non-breeding habitat Conservation status: Listed as Rare in Qld and Vulnerable in requirements, population size, structure or dynamics. NSW but not considered threatened nationally.We have not Conservation status: Kyarranus are listed as Rare in Qld and identified any additional information to suggest that its status Vulnerable in NSW. The status of each species will need to be in either state or nationally requires review. Assa darlingtoni is reviewed after the taxonomy of the group is revised and the not the subject of a recovery process. distribution and threats to each are known. Recovery plans have not been prepared for any of the species of Kyarranus.

50 FIGURE 5: Distribution of green-thighed frog Litoria brevipalmata FIGURE 6: Distribution of pouched frog Assa darlingtoni

● 1990 onwards record ❍ pre-1990 record ■ capital cities

● 1990 onwards record ❍ pre-1990 record ■ capital cities

FIGURE 7: Distribution of mountain frogs Kyarranus spp FIGURE 8: Distribution of black-soled frog Lechriodus fletcheri

● 1990 onwards record ❍ pre-1990 record ■ capital cities

● 1990 onwards record ❍ pre-1990 record ■ capital cities

51 Black-soled frog Lechriodus fletcheri Plateau (SEQ), and on the Tooloom Range and Acacia Plateau (NENSW) (Hines unpubl. data). It has been detected at most Distribution: South from Mistake Mountains SEQ (27° 58´S other sites in the remainder of its range in recent years 152° 23´E — SEQRFA, Hines unpubl. data) to near (Mahony unpubl. data). MENSW (33° 25´S 151° 20´E — Covacevich and McDonald 1993). Figure 8. Moore (1961) referred to a specimen from Threats: Relative to the other mesic frogs considered in this Ravenshoe on the Atherton Tableland in north-eastern Qld. review, L. revelata is more frequently detected in heavily However McDonald and Miller (1982) argued that the disturbed sites (e.g. farm dams). Some of these sites continue to locality recorded for this specimen (AM 19947) was be disturbed through clearing, timber harvesting and associated probably erroneous. activities, and cattle grazing. There have been no studies on the impact (negative or positive) of these disturbances on L. revelata Current distribution: Corben (in McDonald 1991) felt that and the long-term viability of these populations is not known. L. fletcheri had declined in abundance but lacked any quantitative data to support the suggestion. Recent regional Biology: Poorly known. The description of the species surveys (e.g. Mahony 1996, NSW NPWS 1994a) and site- (Ingram et al. 1982) provides basic information on specific monitoring (e.g. Hines unpubl. data) have recorded distribution and habitat. Larvae from the Clarke Range this species frequently throughout its known range. population have been described by Hero et al. (1996). There is no information on population size, structure, genetics or Threats: Lechriodus fletcheri is commonly encountered in dynamics. Resolution of taxonomic confusion in the southern wet sclerophyll forest as well as rainforest. In the past population with other members of the L. ewingii complex is considerable clearing of this species’ habitat occurred needed to accurately determine the distribution and ecology (e.g. Comboyne, Dorrigo and Beechmont plateaux) and of L. revelata in that area. much of its remaining habitat has been logged. Although rainforest is now largely excluded from timber harvesting, Conservation status: Currently considered Rare in Qld, due logging of wet sclerophyll forest continues. The impact of to its restricted and disjunct distribution. A review of its this on L. fletcheri is unknown. status in both states and nationally may be necessary once a better understanding of the taxonomy of the L. ewingii Biology: The eggs and larval development of L. fletcheri are complex is gained. well documented (Moore 1961,Watson and Martin 1973, Martin 1967). The factors affecting this species’ distribution in Mesic forest species — facultative stream breeders north-eastern NSW (the vast majority of its range) were examined by modelling presence-absence data with a pool of Tusked frog Adelotus brevis 24 environmental predictors (NSW NPWS 1994a). The model showed that L. fletcheri had a marked preference for Distribution: Disjunct population in the Clarke Range MEQ mesic forests in areas of high rainfall and deeper soils, on mid (21° 02´S 148° 36´E — 21° 19´S 148° 14´E — QM and AM slopes or plateaux. These environments provide a greater specimens) then Shoalwater Bay MEQ (22° 40´S 150° 41´E abundance of ephemeral water bodies, an important — ANWC specimens) south along the coast and ranges to breeding requirement for this species (Moore 1961, Hines near Moss Vale MENSW (34° 37´S 150° 30´E — SAM unpubl. data). The role of larvae in the food web of water specimen), inland to Blackdown Tableland (23° 43´S 149° filled tree holes was investigated by Kitching and Callaghan 04´E — QM specimens) and Carnarvon Gorge (24° 55´S (1982). There is no information on population size, structure, 148° 05´E — QM specimen), Figure 10. dynamics, or genetic variation. Current distribution: In south-eastern Qld there has been little Conservation status: Lechriodus fletcheri is considered Rare in recent survey work west of the Great Divide so the status of Qld because of its narrowly restricted distribution in that inland populations is not known.Within the subtropical zone A. state. There is no evidence of declines or of major threats brevis has not been recorded along the across its range so we believe there is no need to review its from the NSW border to the Bunya Mountains despite recent legal status. There is also no need for it to be included in a surveys and intensive monitoring in these areas (Hines unpubl. recovery process. data). This area is climatically similar to the adjoining NSW northern tablelands where A. brevis appears to have suffered a Whirring Treefrog Litoria revelata serious decline (Gillespie and Hines 1999). Elsewhere in far south-eastern Qld the species is readily detectable in suitable Distribution: Three allopatric populations; Atherton Tableland, seasonal and weather conditions (Ingram and McDonald 1993, NEQ , Clarke Range MEQ, (Covacevich and McDonald Hines unpubl. data). In far north-eastern NSW, Mahony (1996) 1993), and in the south from Mount Tamborine SEQ (27° frequently located A. brevis in stream and pond situations but only 55’S 153° 10’E — QM specimens) to Ballina NENSW (28° below 400m. In that area there are few historical records above 52’S 153° 34’E — QM specimens; Ingram et al. 1982). Figure 400m making it difficult to assess whether declines have occurred. 9. Records of L. revelata further south (e.g. Covacevich and Adelotus brevis has apparently declined from elevated sites in the McDonald 1993) are possibly records of other members of Clarke Range mid-eastern Qld (Ingram and McDonald 1993; the L. ewingii complex (Mahony and Knowles 1994) and have McDonald and Alford 1999) and from the NSW northern not been included in this review. tablelands (Gillespie and Hines 1999). Localised declines have also been reported from the Blackall and Conondale Ranges south- Current distribution (southern population only): No eastern Qld (1978 to 1984 — Ingram and McDonald 1993) but reports of declines. Under suitable conditions it has been no declines were apparent in the Watagan Mountains in mid- found recently in abundance at several ponds on O’Reilly’s eastern NSW (Mahony 1993).

52 FIGURE 9: Distribution of whirring treefrog Litoria revelata in FIGURE 10: Distribution of tusked frog Adelotus brevis subtropical Australia

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

FIGURE 11: Distribution of great barred-frog Mixophyes fasciolatus FIGURE 12: Distribution of Fleay’s barred-frog Mixophyes fleayi

● 1990 onwards record ❍ pre-1990 record ■ capital cities

● 1990 onwards record ❍ pre-1990 record ■ capital cities

53 Threats: Ill and dead A. brevis have recently been found from often uses unconnected pools near streams and also urban Brisbane, the Lamington Plateau and D’Aguilar Range billabongs and dams well away from streams. Although (Hines unpubl. data) and near Lismore (Tarvey NSW NPWS M. fasciolatus is commonly found in disturbed environments it pers. comm.). Preliminary post-mortem examinations of does not tolerate habitat clearing. In the Dorrigo area specimens from Brisbane suggest that a chytrid fungus (NENSW), Lemckert (1999) found that M. fasciolatus was (Berger et al. 1998) is the cause of death (Berger unpubl. less abundant in more recently logged areas. The effect of data). A major threat to A. brevis is loss and degradation of other disturbances (e.g. fire, grazing) has not been examined. habitat, especially in the lowlands (see discussion on coastal Berger et al. (1998) identified a chytrid fungal infection as the lowland frogs), through agriculture and urban development. cause of death in a captive population and more recently in Although A. brevis persists in heavily disturbed sites the an animal from the wild (Mount Glorious SEQ — Berger viability of these populations is unknown. Introduced fish, unpubl. data). including Gambusia holbrooki, are also widespread in some lowland areas. Their impact on A. brevis is not known. Biology: Poorly known. There are qualitative descriptions of Within forest catchments timber harvesting, cattle grazing and broad habitat use (e.g. Cogger 1996; Barker et al. 1995; altered fire regimes may also affect this species. For example, Straughan 1966) and its eggs and larvae are inadequately Lemckert (1999) found that in the Dorrigo area (NENSW) described. Current studies by University of Newcastle and A. brevis was apparently dependent upon patches of QPWS are providing more detailed information on habitat undisturbed forest. requirements breeding biology and genetic population structuring. Captive husbandry techniques have been Biology: Poorly known. Katsikaros and Shine (1997) examined developed for this species at a number of institutions sexual dimorphism with respect to diet, habitat use and mating (Amphibian Research Centre, Melbourne Zoo, Lone Pine systems. Moore (1961),Watson and Martin (1973) and Daly Sanctuary). There is no information on population size, (1995) described the nests, eggs and larvae. There is no structure, dynamics or genetics, non-breeding habitat information on population size, structure, genetics or dynamics. requirements or factors limiting distribution.

Conservation status: This is the first time that possible Conservation status: Mixophyes fasciolatus in not listed as a declines of A. brevis over large areas have been identified. It threatened species in either state or nationally. It has not had previously been considered a secure species that was shown any significant reduction in range. Due to ecological widespread and common (e.g. Tyler 1997; Covacevich and similarities with stream dependant frogs that have declined or McDonald 1993; Ingram and McDonald 1993). It has therefore disappeared we recommend that this species be included in not been considered for listing as a threatened species. The monitoring programs, particularly to assess changes in relative information presented here and in the review by Gillespie and abundance. Mixophyes fasciolatus is and should continue to be Hines (1999) should be used as a basis for reassessment of its used as a model for research into threatening processes and status, particularly in NSW. Targeted surveys and inclusion of conservation management for the two species of Mixophyes this species in monitoring and research programs for other currently considered endangered. declining species are recommended. Mesic forest species — obligate stream breeders Great barred-frog Mixophyes fasciolatus Fleay’s barred-frog Mixophyes fleayi Distribution: Clarke Range MEQ (20° 50’S 148° 28’E — 21° 17’S 148° 30’E — AM, QM and SAM specimens); Kroombit Distribution: Narrowly and disjunctly distributed in wet Tops SEQ (24° 22´S 151° 01´E — e.g. QM specimens) to forests from the Conondale Range SEQ (26° 43´S 152° 35´E Gosford MENSW (33° 26’S 151° 20’E — AM, MV and SAM — Hines unpubl. data), south to Yabbra Scrub in NENSW specimens). Records from the Blue Mountains and further (28° 37´S 152° 29´E — Mahony unpubl. data), Figure 12. south (e.g. Cogger 1996) require confirmation. Figure 11. Current distribution: Corben (in McDonald 1991) reported that Current distribution: Frequently recorded, often in large M. fleayi declined in the Conondale Range in the late 1970’s. breeding congresses throughout its known range (e.g. Ingram and McDonald (1993) reported that it had not been Mahony 1996; Ingram and McDonald 1993; NSW NPWS seen in the Conondale Range since the summer of 1990-91. 1994a). Corben (in McDonald 1991) reported that Since Ingram and McDonald’s review, targeted surveys for M. M. fasciolatus declined markedly in the Conondale Range in fleayi have been undertaken. A population was found in the the mid 1970’s, but no other researchers at the time upper reaches of three neighbouring streams in the Conondale reported declines in this area. Regular surveys and monitoring Range (Hines unpubl. data) despite surveys of historical sites in this and other areas of south-eastern Qld (Hines unpubl. lower down in the streams which failed to locate the species. In data) support Ingram and McDonald’s (1993) statement that Qld other populations are currently known from Lamington there is no decline in this species. In the Conondale Range it plateau and the northern section of Main Range (Hines unpubl. is one of the few species that persisted at sites such as data), Mount Barney area (Hero unpubl. data) and Currumbin Booloumba Creek where four species of stream dependant (Dadds unpubl. data) and Tallebudgera Creeks (QM specimen, frogs disappeared. Hines unpubl. data) below Springbrook Plateau. There have been no records of M. fleayi from the extensively developed Mt Threats: The apparent resilience of M. fasciolatus to Tamborine area since 1976, despite targeted surveys (Hines disturbance and factors causing the declines in other stream unpubl. data, Hero unpubl. data). In NSW it is known from breeding species may be linked to its broader habitat Lever’s Plateau (Border Ranges — Cunningham pers. comm.), utilisation and differences in breeding biology. Mixophyes Yabbra and Tooloom Scrubs (Mahony unpubl. data, Dadds fasciolatus is not dependant upon streams for breeding, but unpubl. data), Mt Warning (Southern Cross University unpubl.

54 data),Terania Creek in and Sheepstation Creek 1994b, AM specimen). North of this there is currently a large in the Border Ranges (Mahony 1996, Mahony et al. 1997a). population in the Dorrigo — Coffs Harbour area, North In the past two summer seasons there have been no sightings Washpool State Forest and Bungawalbin State Forest of M. fleayi at Terania Creek despite intensive searches, although (Mahony unpubl. data, Southern Cross University unpubl. prior to this only very low numbers had been observed data). In far north-eastern NSW, M. iteratus is known from (Mahony unpubl. data). only three sites, despite intensive surveys.

Mixophyes fleayi has disappeared from some sites.Whether it In south-eastern Qld M. iteratus is currently known from has declined at others is difficult to assess due to a lack of scattered locations in the Mary River catchment downstream historical records of relative abundance. The very low to about Kenilworth, Upper Stanley River, Caboolture River numbers recorded from many well surveyed sites suggest and Coomera River (Hines unpubl. data, Marshall unpubl. that declines in abundance may have occurred. data). During the early 1980’s M. iteratus declined and disappeared from at least two streams in the Conondale Threats: The reasons for declines or disappearance of Range where it was well known (Corben in McDonald 1991, M. fleayi are not known. Large areas of this species’ habitat McDonald unpubl. data, Ingram unpubl. data). Recent surveys have and continue to be degraded by feral animals (e.g. feral and monitoring (see R. silus) in this area failed to locate it pigs in the Conondale Range), domestic stock (Main Range) despite a population being discovered in a nearby catchment and invasion of weeds. Upstream clearing, timber harvesting (Parris unpubl. data). The Bunya Mountains (Straughan 1966) and urban development (e.g. Mt Tamborine) are all likely to and Cunningham’s Gap (Straughan 1966, AM specimen, have affected flow regimes and water quality. A chytrid fungal Corben pers. comm.) previously supported M. iteratus but infection has been identified as the cause of illness and these and nearby sites have recently been the subject of death of M. fleayi on Main Range and Lamington plateau targeted survey or intensive monitoring (Hines unpubl. data), (Berger et al. 1998). without locating the species. Assessing the extent of the decline is difficult because of the lack of baseline data on its Biology: Poorly known. This species was first described in distribution and abundance. 1987 by Corben and Ingram, who provided very basic information of the species’ ecology. Current studies by Threats: Many sites where M. iteratus occurs are the lower Knowles (University of Newcastle) and QPWS are providing reaches of streams which have had major disturbances such detailed information on habitat requirements, breeding as clearing, timber harvesting and urban development in their biology, population structure and dynamics. Samples are being headwaters. In the Dorrigo area (NENSW), Lemckert (1999) collected to enable examination of genetic structure. found that M. iteratus was less abundant in recently logged areas and at sites where there was little undisturbed forest. Conservation status: Mixophyes fleayi is currently listed as The impacts of feral animals, domestic stock, weed invasion Endangered in both Qld and NSW and in the Action Plan and disturbance to riparian vegetation, all potential threats to for Australian Frogs (Tyler 1997), but it is not listed nationally. current populations, are unknown. Based on its rarity, restricted distribution, possible declines and current threats its status nationally should be reviewed. Biology: Poorly known with only basic information on its A draft recovery plan has now been prepared and ecology (Straughan 1968, Mahony et al. 1997b). Current implemented (Hines 1997). A captive husbandry project has studies by Knowles and Mahony (University of Newcastle) been initiated at Lone Pine Koala Sanctuary. are providing more detailed information on habitat requirements, breeding biology, population structure and Giant barred-frog Mixophyes iteratus dynamics. A study of genetic variation within the species has Distribution: Belli Creek near Eumundi SEQ (26° 31´S 152° been completed (Donnellan and Mahony, unpubl. data) and 49´E — Barden pers. comm.) south to Warrimoo MENSW no species level differences were found among the six NSW (33° 43´S 150° 36´E — AM specimen), Figure 13. Cogger populations compared. (1996) states that M. iteratus was distributed south “to about Conservation status: Mixophyes iteratus is currently listed as Narooma” (36° 13´S 150° 08´E), but there are no specimens Endangered in both Qld and NSW and in the Action Plan for nor other records this far south to substantiate the claim. Australian Frogs (Tyler 1997), but is not listed nationally. Current distribution: Mixophyes iteratus has suffered major Based on declines in distribution and the current threats to declines in the southern portion of its range. There are no M. iteratus, its status nationally should be reviewed. A draft recent records from the Blue Mountains (Mahony et al. recovery plan has now been prepared and implemented 1997b), although there were only a few historical records in (Hines 1997). that area. In the Watagan Mountains M. iteratus is currently known from a small recently rediscovered population Southern gastric Rheobatrachus silus (Mahony unpubl. data). Although not common there in the brooding frog past it was frequently recorded (Mahony 1993). Between the Distribution: Restricted to elevations of 400 — 800m in the Hunter River and Macleay catchment there is currently only Blackall and Conondale Ranges SEQ between Coonoon one known population, at Mount Seaview (Hobcroft pers. Gibber Creek (26° 33’S 152° 42’E — QM specimens) and comm.), but survey effort in this area has been relatively low. Kilcoy Creek (26° 47’S 152° 38’E — Covacevich and There were only two confirmed historical records in this area McDonald 1993, QM specimens), Figure 14. (AM specimens from the Upper and Middle Brother State Forest). A population was recently located in the southern catchment (NSW NPWS

55 Current distribution: Not sighted in the wild since 1981 Current distribution: Not sighted in the wild since 1979 despite continued efforts to relocate the species. Since despite continued efforts to relocate the species. Since Ingram and McDonald’s review (1993) the following surveys Ingram and McDonald’s review (1993) the following surveys and monitoring for this species have been undertaken: and monitoring for the species have been undertaken: a. Regular monitoring at Ingram’s (1983) study site — Beauty a. T. diurnus was present at most sites R. silus occurred, so Spot 100 on Booloumba Creek, Bundaroo, Peters and East surveys and monitoring for that species (see above) were Kilcoy Creeks in the Conondale Range (Borsboom unpubl. likely to detect T. diurnus. data, QPWS unpubl. data) and at Picnic Creek (the type b. Regular (near fortnightly) diurnal monitoring at the type locality near Kondalilla) on the Blackall Range (QPWS locality (Greene’s Falls) and nearby streams at Mount unpubl. data); Glorious by Brisbane Frog Society for a year (1995-1996 b. 1995 intensive “frog search” of Conondale Range — Holdway pers. comm.). (Cunningham pers. comm.); c. A study on L. pearsoniana at nearby Love Creek c. 1997 “frog search” of the headwaters of Kilcoy, North (Cunningham pers. comm.) since September 1995 failed Booloumba and Bundoomba Creeks, Conondale Range to detect T. diurnus despite some diurnal censuses and (Hines unpubl. data); regular tadpole surveys. d. since 1996 systematic surveys of many streams in the Threats: Like R. silus, the reason(s) for the disappearance for Conondale and Blackall Ranges. Some sections of streams T. diurnus remains unknown (Martin et al. 1997). This species’ were visited on many occasions over a range of weather habitat is currently threatened by feral pigs, invasion of weeds conditions. Poorly surveyed streams in the Upper Stanley (especially mist flower) and altered flow and water quality River were targeted (Hines unpubl. data, Dadds unpubl. due to upstream disturbances. data); and Biology: Czechura and Ingram (1990) reviewed knowledge of e. opportunistic surveys by various frog biologists. the biology of T. diurnus. There is a reasonable understanding The species declined rapidly in late 1979, with only a single of the factors affecting distribution of the species, its patterns specimen located after that, in 1981 (Richards et al. 1993). of behaviour, breeding biology, diet, and thermal relations and water balance. There is no information on population size, Threats: The reason(s) for the disappearance of this species structure, genetics or dynamics. remains unknown (Tyler and Davies 1985b). Populations of R. silus were present in logged catchments between 1972 and 1979. Although R. silus persisted in the streams during these activities, the effects of timber harvesting on this aquatic species were never investigated. Its habitat is currently FIGURE 13: Distribution of giant barred-frog Mixophyes iteratus threatened by feral pigs, invasion of weeds (especially mistflower Ageratina riparia), and altered flow and water quality due to upstream disturbances.

Biology: Relative to other frog species in subtropical Australia, R. silus was well studied before its disappearance. There was detailed examination of its highly unusual reproductive strategy (gastric brooding; e.g. Corben et al. 1974;Tyler 1985). Ingram (1983) studied the ecology of R. silus at Booloumba Creek and gathered some information on population size, structure and dynamics before its disappearance. There is no information on the genetic structuring of populations or on the factors limiting distribution.

Conservation status: It is listed as Endangered in Qld, nationally and in the Action Plan for Australian Frogs (Tyler 1997). Rheobatrachus silus is a species (IUCN 1996) that is possibly extinct. A draft recovery plan ● 1990 onwards record has been prepared (Martin et al. 1997) but recovery is ❍ pre-1990 record dependant upon locating an extant population. Monitoring of ■ capital cities historical locations and surveys of potential habitat will continue as part of ongoing frog conservation actions in south-eastern Qld.

Southern dayfrog Taudactylus diurnus Distribution: Occurred in disjunct populations in the Blackall, Conondale and D’Aguilar Ranges SEQ from Coonoon Gibber Creek in the north to Mount Glorious in the south (26° 33’S 152° 42’E — 27° 23´S 152° 47´E — QM specimens) Figure 15.

56 Conservation status: It is listed as Endangered in Qld, streams. During recent surveys of mesic forests in north- nationally and in the Action Plan for Australian Frogs (Tyler eastern NSW, L. lesueuri was infrequently recorded above 1997). Taudactylus diurnus is a critically endangered species 400m (5% of 36 sites surveyed — Mahony unpubl. data). (IUCN 1996) that is possibly extinct. A draft recovery plan This observation may be additional evidence for decline of has been prepared (Martin et al. 1997) but recovery is this species from montane rainforest streams, but our dependant upon locating an extant population. Monitoring of assessment is limited by a lack of historical data on historical locations and surveys of potential habitat will distribution and abundance in this area. There have been no continue as part of ongoing frog conservation actions in reports of declines from lower altitudes or from open forests. south-eastern Qld. In the temperate zone L. lesueuri is still frequently encountered at high altitude (Gillespie and Hines 1999). Stony creek frog Litoria lesueuri Threats and biology: Gillespie and Hines (1999) summarise Distribution: Endeavour Falls, Cooktown NEQ (15° 25’S known or likely threats to L. lesueuri as well as knowledge of 145° 05’E — Covacevich and McDonald 1993) to Lederderg its biology and taxonomy. Gorge Vic (37° 37’S 144° 25’E — Gillespie and Hines 1999). Figure 16. Conservation status: Litoria lesueuri is not listed as a threatened species in any of the states in which it occurs. Current distribution: Within the wet subtropical region of Given its past decline from some rainforest streams and its Australia, L. lesueuri remains one of the most widespread, ecological similarity with other declining frogs, it is abundant and frequently recorded stream breeding frogs. recommended that L. lesueuri be included in monitoring There are however reports of declines. Ingram and programs, to assess changes in relative abundance. McDonald (1993) reported declines from rainforest streams in the Blackall and Conondale Ranges of south-eastern Qld Cascade treefrog Litoria pearsoniana between the late 1970’s and early 1980’s, but the species The identity of various populations referred to this species remained common in wet sclerophyll forests. This period are currently under review. Recent allozyme and DNA coincides with the disappearance of T. diurnus and R. silus, and studies (Mahony et al. unpubl. data; McGuigan et al. 1998) declines of other species from the same area. Following their indicate that the population at Kroombit Tops (24° 24´S 151° survey in 1993 Ingram and McDonald considered that 01´E — Czechura 1986, QM specimens) may be an populations of L. lesueuri were recovering along these

FIGURE 14: Distribution of southern dayfrog Taudactylus diurnus FIGURE 15: Distribution of southern gastric brooding frog (after Martin, McDonald and Hines 1997). Shaded areas are: Rheobatrachus silus (after Martin, McDonald and Hines 1997). light grey -state forest or timber reserve, dark grey - national Shaded areas are: light grey -state forest or timber reserve, dark parks and conservation parks. Some towns and larger streams grey - national parks and conservation parks. Some towns and are also shown. larger streams are also shown.

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ towns ■ towns

57 undescribed taxon. These studies will also more clearly was abundant over several kilometres of three streams in the delineate the southern range limit of L. pearsoniana. Conondale Range; at Bundaroo Creek 61 individuals were counted on a 100m transect (Hines unpubl. data). However, Distribution: Kandanga State Forest SEQ (26° 26´S 152° at other sites where seemingly suitable habitat exists, 24´E — SEQRFA) south to Gibraltar Range NENSW (29° L. pearsoniana is currently at low densities. For example only 31´S 152° 25´E — Mahony et al. unpubl. data), Figure 17. one individual was recorded during regular monitoring of a 1 200 m transect along a rainforest stream at Cunningham’s Current distribution: Czechura (1991) and McDonald and Gap south-eastern Qld (33 censuses totalling 24 km — Davies (1990) recorded declines of L. pearsoniana in the late Hines unpubl. data). 1970’s to early 1980’s from the Conondale and Blackall Ranges south-eastern Qld. Corben (in McDonald 1991) Threats: The reasons for population declines are unknown. suggested that this species had not suffered a conspicuous In the Blackall and Conondale Ranges they coincide with the decline in the Conondale Range, but that it had disappeared period in which T. diurnus and R. silus disappeared. Large areas from some streams in Brisbane Forest Park. Ingram and of this species’ habitat have and continue to be degraded by McDonald (1993) found L. pearsoniana breeding in small introduced animals, e.g. feral pigs and domestic stock, invasion numbers in the Conondale, Border and Main Ranges. During of weeds and timber harvesting (see Parris and Norton 1997 their survey only two were heard during ideal weather for discussion). Upstream clearing and urban development conditions at East Kilcoy Creek (McDonald unpubl. data) have reduced habitat and are likely to have affected where it had previously occurred in hundreds during the downstream flow regimes and water quality in some localities study of McDonald and Davies (1990). Ingram and McDonald (for example Kondallila Falls). Infections of a chytrid fungus (1993) did not find it at Kondallila Falls in the Blackall Range, (Berger et al. 1998) have been found on dead individuals although it was common there in the 1970’s (McDonald from Main Range south-eastern Qld and from the closely unpubl. data). There are no reports of declines in NSW. related taxon at Kroombit Tops (Berger unpubl. data). Other ill and dead L. pearsoniana have been found in the Conondale More recent studies have found the species to be reasonably Range but have not yet been examined (Hines unpubl. data). widespread (Mahony 1996; NSW NPWS 1994a) or “appear to have recovered at some sites” (McGuigan et al. 1998). Biology: Relatively well documented. McDonald and Davies Surveys and monitoring carried out over the past two (1990) described adult behaviour and breeding biology. seasons by QPWS in south-eastern Qld largely support these Environmental and habitat preferences are provided by findings. It is one of the most frequently recorded species McDonald and Davies (1990), Covacevich and McDonald along streams in mesic forests. In May 1996 L. pearsoniana (1993), NSW NPWS (1994a) and Parris and Norton (1997).

FIGURE 16: Distribution of stony creek frog Litoria lesueuri in FIGURE 17: Distribution of cascade treefrog Litoria pearsoniana subtropical Australia

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

58 Genetic structuring within L. pearsoniana has been reported by T. pleione are included in the unpublished documents of McGuigan et al. (1998) who found significant divergence Cunningham and James (1994) and Borsboom et al. (1998). among populations from different wet forest isolates. There is The breeding biology is unknown and there is no information no published information on population dynamics or structure. on population size, structure or genetic variation.

Conservation status: Litoria pearsoniana is currently listed as Conservation status: Taudactylus pleione is currently Endangered in Qld but is not considered threatened in NSW considered Vulnerable in Qld and in the Action Plan for or nationally. Reassessment of its status, based on a more Australian Frogs (Tyler 1997) but is not listed nationally. Its thorough analysis of recent survey and monitoring data, is rarity, extremely restricted distribution, coupled with possible warranted. Resolution of taxonomic problems within the population declines and the identification of a number of group is needed, particularly for the population at Kroombit likely threats suggests that the status of this species should be Tops.A draft recovery plan for L. pearsoniana has been reviewed for listing as Endangered. A range of interim prepared and implemented (Hines 1997). management measures, i.e. monitoring, handling protocols, stock exclusion and fire management are being implemented Mesic forest species — breeding biology unknown pending the finalisation of the recovery plan for the species (Borsboom et al. 1998). Kroombit tinkerfrog Taudactylus pleione Distribution: Very restricted distribution. Confined to eight DISCUSSION small patches of rainforest above 600m at Kroombit Tops, Taudactylus diurnus and R. silus, species endemic to the wet south-west of Gladstone SEQ (24° 22´S 151° 01´E) subtropical region of Australia remain missing and the factors (Clarke et al. in press). causing their disappearance are unknown. Four other obligate Current distribution: Surveys in February 1997 greatly or facultative stream breeding frogs, A. brevis, M. fleayi, expanded the known distribution of T. pleione from three to M.iteratus and L.pearsoniana, have suffered significant local or eight sites with all additional sites in Kroombit Tops National regional declines. For L. pearsoniana, there is evidence of Park (Clarke et al. in press). The only monitored population, population recovery. In the case of A. brevis we (this review; in the head of Kroombit Creek, appears to have declined. Gillespie and Hines 1999) present data for the first time that The species was regularly encountered prior to 1997 but indicates a major decline in populations from the northern T. pleione was not heard or seen at this site during the tablelands of NSW north along the Great Divide into south- eastern Qld. The magnitude and extent of declines in these 1997/98 season despite systematic monitoring (Clarke unpubl. four stream breeding species is difficult to quantify due to a data). During 1997/98 little other survey and monitoring work lack of long term systematic baseline information. Determining was undertaken in the area but T. pleione was heard calling at the cause of these declines is made more difficult as many three of the recently discovered populations. potentially threatening processes have operated over the Threats: Clarke et al. (in press) and Borsboom et al. 1998 list six range of these species. In many cases the impacts of these main potential threats to T. pleione; timber harvesting, domestic threats are unknown, let alone likely synergistic effects. More and feral animals, visitor pressure, wild fire and the unknown detailed discussion of threats to stream-breeding frogs is agent(s) responsible for declines in other Qld frogs. Timber provided by Gillespie and Hines (1999). harvesting has ceased in the catchments above all known There is no evidence of regional declines in two facultative populations. A fence to exclude domestic stock has been stream breeding species, M. fasciolatus and L. lesueuri although constructed but stock continue to impact habitat at the head of both have declined from high altitude rainforest streams in Kroombit Creek. Feral pigs, which have the potential to prey some areas. Given the serious declines in other stream upon T. pleione or destroy habitat, have recently been found breeding frogs in eastern Australia (see also Gillespie and nearby.A recent wildfire burned into many rainforest patches Hines 1999, McDonald and Alford 1999) it is important that used by this species and may at least be partially responsible for monitoring and research projects incorporate these species decline of the species at the monitoring site. Modified fire due to their potential vulnerability. management procedures have now been put in place. Eight species of mesic forest frogs, that are not dependent on Taudactylus Four species of from similar habitat elsewhere in streams for breeding, show no signs of decline. They are of Qld declined dramatically or disappeared due to unknown conservation concern because of their rarity and or causes (e.g. Ingram and McDonald 1993). In May 1998, dead restricted distribution. L. pearsoniana (sensu lato see above) were found in Kroombit Tops (Hines and Clarke unpubl. data). Chytrid fungus, found Five lowland species are increasingly at risk due to loss or by Berger et al. (1998) associated with frog deaths and degradation of habitat. There is an urgent need to establish declines elsewhere in Australia and Central America, was recovery processes for the habitat for wallum or acid frogs isolated from these animals (Berger unpubl. data). This, in — many threatened vegetation communities, plants and conjunction with the probable decline of T. pleione at the other animals are dependent upon these environments. monitoring site has heightened concern for this species. Generally our knowledge about the distribution, population Biology: Knowledge of the biology of T. pleione is extremely size ,structure, natural fluctuations, and biology of the frogs poor. There are only two publications concerning this species reviewed is poor. This needs to be greatly improved to assist (Czechura 1986 a,b) which provide qualitative information of in our understanding of the potential causes of declines and habitat, calling and activity patterns, at the type locality. other threats and ultimately to ensure recovery of populations. Additional information on these aspects of the biology of

59 This review has concentrated on frog species currently ACKNOWLEDGMENTS considered rare or threatened, and additional stream Species locality data were provided by the Australian, breeding species for which there was evidence of declines. Queensland and South Australian Museums, Australian This approach is somewhat subjective and potentially National Wildlife Collection,The National Museum of Victoria, overlooks species which may have declined.We recommend Atlas of New South Wales Wildlife and the South-east that a thorough analysis of the patterns of change in Queensland Regional Forest Assessment Fauna Database. distribution of all frog species in south-eastern Australia be These data made up the vast majority of records used in the undertaken urgently. Such a review needs to address a production of species distribution maps.We thank the number of limitations in the available data sets. The best relevant staff and curators of these collections and databases historical information on species’ distributions lies in our for their generous assistance. A. Borsboom, M. Cunningham, B. museum collections. For each specimen the identity, locality Dadds, R. Goldingay, M. Hero, R. Knowles, C. Marshall, E. and georeference need to be checked. This process has to Meyer, C. Morrison and K. Thumm provided additional species records and opinions on species’ status. Staff of the be coordinated by both State and Federal conservation Parks and Wildlife Service at Moggill also assisted especially B. agencies. It will result in a much more robust data set for Dadds with data manipulation, R. Brown and H. Preece with comparison with ongoing data collection. In order to geographical information system support and J. Angus who complete this process taxonomic resolution of species typed sections of the manuscript.We thank N. Campbell,W. complexes is needed. In this review we identified Litoria Drake, S. Richards and A.White for their valuable comments. cooloolensis, L. lesueuri, L. pearsoniana and L. revelata and Kyarranus spp., as requiring further taxonomic studies. Financial support was provided by Queensland Parks and For many species more robust identification characters need Wildlife Service, NSW National Parks and Wildlife Service, to be determined. Environment Australia, the University of Newcastle and the Natural Heritage Trust. Whilst refinement of historic data sets is required this and other reviews in these proceedings have repeatedly identified REFERENCES major gaps in our understanding of species’ distribution, Anstis, M. (1981). Breeding biology and range extension for ecology and threatening processes. In eastern Australia the New South Wales frog Kyarranus sphagnicolus (Anura: priority in addressing these gaps must be given to stream Leptodactylidae). Australian Journal of Herpetology 1: 1-9. breeding forest species (see also Gillespie and Hines 1999; Anstis, M. (1994). The larval development of Litoria McDonald and Alford 1999), to assist in the identification and brevipalmata (Anura: Hylidae). Memoirs of the Queensland management of the causal agent(s) of the declines. Museum 37: 1-4. Ongoing investigations of ill and dead frogs from south- Aridis, J.J. (1997). A range extension for Litoria brevipalmata (Anura: Hylidae). Memoirs of the Queensland Museum eastern Australia and elsewhere have for the first time 42: 310. identified a pathogen (a chytrid fungus — Berger et al. 1998) apparently causing morbidity and associated with population Barker, J., Grigg, G.C., and Tyler, M.J. (1995). A Field Guide to Australian Frogs. Chipping Norton, NSW: Surrey Beatty & declines. Much work is needed to fully test the disease Sons. hypothesis. Although it is necessary to continue this research urgently, investigations of other potential agents (pollutants, Berger, L., Speare, R., Daszak, P., Green, D.E., Cunningham, A.A., Goggin, C.L., Slocombe, R., Ragan, M.A., Hyatt, A.D., climate change, UV-B and synergistic effects) must also be McDonald, K.R., Hines, H.B., Lips, K.R., Marantelli, G. and carried out simultaneously. Parkes, H. (1998). Chytridiomycosis causes amphibian The greatest threat to lowland frogs is the rapid rate of mortality associated with population declines in the rain habitat loss or degradation. Some ecosystems and many forests of Australia and Central America. Proceedings of the National Academy of Science USA 95: 9031-6. other species of plants and animals are dependent upon coastal environments. This is a major issue that conservation Blaustein, A.R. and Wake, D.B. (1990). Declining amphibian agencies must address. populations: A global phenomenon? Trends in Ecology and Evolution 5: 203-204. An important component in future research and Borsboom, A., Clarke, J. and Cunningham, M. (1998). Draft management of all species is population monitoring. By recovery plan for the Kroombit tinker frog Taudactylus carefully selecting target species, monitoring sites and pleione 1997–2001. Queensland Department of Natural methods, many of the gaps in our current knowledge will be Resources and Queensland Department of Environment filled. Monitoring will allow more rigorous quantitative (May 1998). assessments of declines, (see for example McDonald and Bureau of Meteorology (1989) Climate of Australia. Australian Alford 1999) as well as assisting in the investigations of Government Publishing Service, Canberra. threatening processes. In south-eastern Australia monitoring Catterall, C.P.and Kingston, M. (1993). Remnant Bushland of has largely been ad hoc and recent, despite awareness of South East Queensland in the 1990’s: Its Distribution, Loss, population declines for many years. State and Federal Ecological Consequences, and Future Prospects. Institute of conservation agencies need to work together to ensure the Applied Environmental Research, Griffith University and establishment and continuation of a network of systematic Brisbane City Council, Brisbane, Australia. monitoring throughout eastern Australia.

60 Clarke, J.M., Borsboom, A.C., Cunningham, M. and Hines, Ehmann, H. (1997b). 1. Marsupial Frog, Assa darlingtoni. H.B. (in press). The recovery process for the Kroombit pp. 42-47 in H. Ehmann (ed) Threatened Frogs of New tinkerfrog, Taudactylus pleione.WWF South-east South Wales: Habitats, Status and Conservation. Frog and Queensland Rainforest Recovery Conference Proceedings. Tadpole Study Group of NSW Inc, Sydney South, Australia. Cogger, H.G. (1996). Reptiles and Amphibians of Australia. Ehmann, H. (1997c). 2.Wallum Froglet, Crinia tinnula. Reed Books, Chatswood, NSW. pp. 48-55 in H. Ehmann (ed) Threatened Frogs of New Corben, C.J. and Ingram, G.J. (1987). A new barred river frog South Wales: Habitats, Status and Conservation. Frog and (Myobatrachidae: Mixophyes). Memoirs of the Queensland Tadpole Study Group of NSW Inc, Sydney South, Australia. Museum 25: 233-237. Ehmann, H. (1997d). 17. Green-thighed Frog, Litoria Corben, C.J., Ingram, G.J. and Tyler, M.J. (1974). Gastric brevipalmata. pp. 156-163 in H. Ehmann (ed) Threatened brooding: Unique form of parental care in an Australian Frogs of New South Wales: Habitats, Status and frog. Science 186: 946-947. Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Covacevich, J.A. and McDonald, K.R. (1993). Distribution and conservation of frogs and reptiles of Queensland Ehmann, H. (1997e). 21.Wallum Sedgefrog, Litoria rainforests. Memoirs of the Queensland Museum 34: olongburensis. pp. 182-187 in H. Ehmann (ed) Threatened 189-199. Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Cunningham, M. and James, C. (1994). Taudactylus pleione,a Sydney South, Australia. report on the conservation status and distribution of the Kroombit tinker frog. Department of Zoology and Centre Ehmann, H. and Swan, G. (1985). Reproduction and for Conservation Biology,The University of Queensland, development in the Marsupial Frog, Assa darlingtoni unpublished report to Queensland Forest Service and (Leptodactylidae, Anura). pp. 279-285 in G. Grigg, R. Shine Queensland Department of Environment and Heritage. and H. Ehmann (eds) Biology of Australasian Frogs and Reptiles. Surrey Beatty and Sons/Zoological Society of Czechura, G.V. (1978). A new locality for Litoria brevipalmata New South Wales, Chipping Norton. (Anura: ) from south east Queensland. Victorian Naturalist 95: 150-151. Gillespie, G.R. and Hero, J.M. (1999) Potential impacts of introduced fish and fish translocations on Australian Czechura, G.V. (1986a). A new species of Taudactylus amphibians. Pp 131-144 in Declines and Disappearances (Myobatrachidae) from southeastern Queensland, Australia. of Australian Frogs ed by A.Campbell. Environment Memoirs of the Queensland Museum 22: 299-307. Australia: Canberra. Czechura, G.V. (1986b). Kroombit Tops Torrent Frog Gillespie, G.R. and Hines, H.B. (1999) Status of temperate Taudactylus pleione with a key to the species of Taudactylus. riverine frogs in south-eastern Australia. Pp 109-130 in Queensland Naturalist 27: 68-71. Declines and Disappearances of Australian Frogs ed by Czechura, G.V. (1986c). Distant exiles: Frogs and reptiles A.Campbell. Environment Australia: Canberra. recorded from Kroombit Tops, southeastern Queensland. Gillespie, G.R. and Hollis, G.J. (1996). Distribution and habitat Queensland Naturalist 27: 61-67. of the Spotted Tree Frog, Litoria spenceri Dubois (Anura: Czechura, G.V. (1991). The Blackall-Conondale Ranges: Frogs, Hylidae), and an assessment of potential causes of reptiles and fauna conservation. Chapter 19. pp. 311-324 population declines.Wildlife Research 23: 49-75. in G.Werren and P.Kershaw (eds) The Rainforest Legacy: Hero, J.M., Fickling, S. and Retallick, R. (1996). The tadpole of Australian National Rainforests Study,Volume 2. Special Litoria revelata Ingram, Corben and Hosmer, 1982 (Anura: Australian Heritage Publication Series, 7, Commonwealth Hylidae). Transactions of the Royal Society of South Government Printer, Canberra. Australia 120: 71-73. Czechura, G.V. and Ingram, G.J. (1990). Taudactylus diurnus Hillis, D.M. and Moritz, C. (1996). Molecular Systematics, and the case of the disappearing frogs. Memoirs of the Sinauer Associates Inc., Sunderland. Queensland Museum 29: 361-365. Hines, H.B. (1997). Draft recovery plan for southern barred- Daly, G. (1995). Observations on the Tusked Frog Adelotus frog, Fleay’s barred-frog, giant barred-frog and cascade brevis (Anura: Myobatrachidae). Herpetofauna 25: 32-35. treefrog 1998-2002. Queensland Department of de Bavay, J. (1993). The developmental stages of the Environment (June 1997). frog, Kyarrunus sphagnicolus Moore (Anura: Ingram, G.J. (1983). Natural history. pp. 16-35 in M.J. Tyler Myobatrachidae). Australian Journal of Zoology 41: (ed) The Gastric Brooding Frog. Croom Helm, London. 151-201. Ingram, G.J., Anstis, M. and Corben, C.J. (1975). Observations Donnellan, S. C. and Foster, R. (1997) Genetic study of the on the Australian leptodactylid frog, Assa darlingtoni. Green-thighed Frog. Unpublished report for the Roads Herpetologica 31: 425-429. and Traffic Authority of NSW. Ingram, G.J. and Corben, C.J. (1975a). The frog fauna of Drost, C.A. and Fellars, G.M. (1996). Collapse of a regional North Stradbroke Island, with comments on the ‘acid’ frog fauna in the Yosemite Area of California Sierra frogs of the wallum. Proceedings of the Royal Society of Nevada, USA. Conservation Biology 10: 414-425. Queensland 86: 49-54. Ehmann, H.(ed) (1997a). Threatened Frogs of New South Ingram, G.J. and Corben, C.J. (1975b). A new species of Wales: Habitats, Status and Conservation. Frog and Kyarranus (Anura: Leptodactylidae) from Queensland, Tadpole Study Group of NSW Inc, Sydney South, Australia. Memoirs of the Queensland Museum Australia. 17: 335-339.

61 Ingram, G.J., Corben, C. and Hosmer,W. (1982). Litoria Knowles, R. and Mahony, M. (1997f). 12. A Mountain Frog, revelata: A new species of tree frog from eastern Australia. Philoria sp. nov. 3. pp. 118-123 in H. Ehmann (ed) Memoirs of the Queensland Museum 20: 635-637. Threatened Frogs of New South Wales: Habitats, Status Ingram, G.J. and McDonald, K.R. (1993). An update on the and Conservation. Frog and Tadpole Study Group of NSW decline of frogs. pp. 297-303 in D. Lunney Inc, Sydney South, Australia. and D. Ayers (eds) Herpetology in Australia. Surrey Beatty Ledlin, D. (1997). Ecology of the Green-thighed Frog (Litoria and Sons, Sydney. brevipalmata). B. Env. Sc (Honours) Thesis. The University IUCN (1996). International Union for the Conservation of of Newcastle. Nature (IUCN) Red List Categories. Species Survival Lemckert, F. (1999). Impacts of selective logging on frogs in a Commission, Gland, Switzerland. forested area of northern New South Wales. Biological James, C. (1996). Conservation genetics of island and Conservation 89: 321-328. mainland populations of the sedge frogs Litoria cooloolensis Liem, D.S. (1974). A new species of the Litoria bicolor species and Litoria olongburensis. Department of Zoology and group from southeast Queensland, Australia (Anura: Centre for Conservation Biology,The University of Hylidae). Memoirs of the Queensland Museum 17: 169-174. Queensland, unpublished final report to Queensland Liem, D.S. and Ingram, G.J. (1977). Two new species of frogs Department of Environment and Heritage. (Anura: Myobatrachidae, Pelodryadidae) from Queensland Katsikaros, K. and Shine, R. (1997). Sexual dimorphism in the and New South Wales.Victorian Naturalist 94: 255-262. tusked frog, Adelotus brevis (Anura, Myobatrachidae) — Lips, K.R. (1998). Decline of a tropical montane amphibian the roles of natural and sexual selection. Biological Journal fauna. Conservation Biology 12: 106-117. of the Linnean Society 60: 39-51. Mahony, M. (1992). A frog in my pocket.Wildlife Australia 24-25. Kikkawa, J., Ingram, G.J. and Dwyer, P.D. (1979). The vertebrate Mahony, M. (1993). The status of frogs in the Watagan fauna of Australian heathlands — an evolutionary mountains area the central coast of New South Wales. perspective. pp. 231- in R.L. Specht (ed.) Ecosystems of pp. 257-264 in D. Lunney and D. Ayers (eds) Herpetology the World 9A. Heathlands and Related Shrublands. in Australia. Surrey Beatty and Sons, Sydney. Chapter 8. Elsevier Scientific Publishing Company, Amsterdam. Mahony, M. (1996) Survey of the distribution and abundance of declining frog species in northern New South Wales. Kitching, R.L. and Callaghan, C. (1982). The fauna of water- Unpublished report to Australian Nature Conservation filled tree holes in box forest in southeast Queensland. Agency. Australian Entomological Magazine 8: 61-70. Mahony, M. (1999) Review of the declines and Knowles, R. (1994). The systematics of the endangered disappearances within the bell frog species group (Litoria montane rainforest frogs of the genus Philoria (Anura: aurea species group) in Australia. P 81-93in Declines and Myobatrachidae), and implications for conservation Disappearances of Australian Frogs ed by A.Campbell. management. Honours Thesis, Department of Biological Environment Australia: Canberra. Sciences, University of Newcastle. Mahony, M. and Knowles, R. (1994). A taxonomic review of Knowles, R. and Mahony, M. (1997a). 7. Red and Yellow selected frogs of north-east NSW forests. North East Mountain Frog, Philoria kundagungan. pp. 84-89 in Forests Biodiversity Study Report No. 3g. NSW National H. Ehmann (ed) Threatened Frogs of New South Wales: Parks and Wildlife Service, unpublished report. Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Mahony, M., Knowles, R. and Pattinson, L. (1997a). 5. Silverblue-eyed , Mixophyes fleayi. pp. 72-77 in Knowles, R. and Mahony, M. (1997b). 8. Masked Mountain H. Ehmann (ed) Threatened Frogs of New South Wales: Frog, Philoria loveridgei. pp. 90-95 in H. Ehmann (ed) Habitats, Status and Conservation. Frog and Tadpole Study Threatened Frogs of New South Wales: Habitats, Status Group of NSW Inc, Sydney South, Australia. and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Mahony, M., Knowles, R. and Pattinson, L. (1997b). 6. - eyed Barred Frog, Mixophyes iteratus. pp. 78-83 in H. Knowles, R. and Mahony, M. (1997c). 9. Sphagnum Mountain Ehmann (ed) Threatened Frogs of New South Wales: Frog, Philoria sphagnicola. pp. 96-105 in H. Ehmann (ed) Habitats, Status and Conservation. Frog and Tadpole Study Threatened Frogs of New South Wales: Habitats, Status Group of NSW Inc, Sydney South, Australia. and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Martin, A.A. (1967). Australian anuran life histories: Some evolutionary and ecological aspects. pp. 175-191 in A.H. Knowles, R. and Mahony, M. (1997d). 10. A Mountain Frog, Weatherby (ed.) Australian Inland Waters and Their Fauna. Philoria sp. nov. 1. pp. 106-111 in H. Ehmann (ed) ANU Press, Canberra, Australia. Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Martin,W.E., McDonald, K.R. and Hines, H.B. (1997). Recovery Inc, Sydney South, Australia. plan for southern platypus frog and southern dayfrog. Unpublished report to Environment Australia. Knowles, R. and Mahony, M. (1997e). 11. A Mountain Frog, Philoria sp. nov. 2. pp. 112-117 in H. Ehmann (ed) McDonald, K.R. (1974). Litoria brevipalmata, an addition to the Threatened Frogs of New South Wales: Habitats, Status Queensland amphibian list. Herpetofauna 7: 2-4. and Conservation. Frog and Tadpole Study Group of NSW McDonald, K.R. (1991). Report of a workshop on declining Inc, Sydney South, Australia. frog populations in Queensland. Queensland National Parks and Wildlife Service, unpublished report.

62 McDonald, K.R. (1992). Distribution patterns and Pyke, G.H. and Osborne,W.S. (1996). The green and golden conservation status of north Queensland rainforest frogs. bell frog Litoria aurea: biology and conservation. Australian Queensland Department of Environment and Heritage Zoologist 30: 132-258. Conservation Technical Report 1: 1-51. Richards, S.J., McDonald, K.R. and Alford, R.A. (1993). Declines McDonald, K.R. and Alford, R.A. (1999) A review of declining in populations of Australia’s endemic tropical rainforest frogs in northern Queensland. Pp 14-22 in Declines and frogs. Pacific Conservation Biology 1: 66-77. Disappearances of Australian Frogs ed by A.Campbell. Smith, A.P.,Hines, H.B., Pugh, D. and Webber, P.(1989a). Environment Australia: Canberra. Mammals, reptiles and amphibians of the rainforests of the McDonald, K.R. and Davies, M. (1990). Morphology and Focal Peak region. Department of Ecosystem Management, biology of the Australian tree frog Litoria pearsoniana University of New England, Armidale. (Copland) (Anura: Hylidae).Transactions of the Royal Smith, A.P.,Hines, H.B. and Webber, P.(1989b). Mammals, Society of South Australia 114: 145-156. reptiles and amphibians of the rainforests of the Mount McDonald, K.R. and Miller, J.D. (1982). On the status of Warning Caldera region. Department of Ecosystem Lechriodus fletcheri (Boulenger) (Anura: Leptodactylidae) in Management, University of New England, Armidale. northeast Queensland.Transactions of the Royal Society of Straughan, I.R. (1966). An analysis of species recognition and South Australia 106: 220. species isolation in certain Queensland frogs. PhD Thesis, McEvoy, J.S., McDonald, K.R. and Searle, A.K. (1979). Mammals, University of Queensland. birds, reptiles and amphibians of the Kilcoy Shire, Straughan, I.R. (1968). A taxonomic review of the genus Queensland. Queensland Journal of Agricultural and Mixophyes, (Anura, Leptodactylidae). Proceedings of the Animal Sciences 36: 167-180. Linnean Society of New South Wales 93: 52-59. McGuigan, K., McDonald, K., Parris, K. and Moritz, C. (1998). Straughan, I.R. and Main, A.R. (1966). Speciation and Mitochondrial DNA diversity and historical biogeography polymorphism in the genus Crinia Tschudi (Anura, of a wet forest-restricted frog (Litoria pearsoniana) from mid-east Australia. Molecular Ecology 7: 175-186. Leptodactylidae) in Queensland. Proceedings of the Royal Society of Queensland 78: 11-28. Meyer, E.A. (1997). Acid tolerance in larval Litoria fallax and L. cooloolensis. B.Sc (Hons) Thesis, University of Queensland. Tyler, M.J. (1972). A new genus for the Australian leptodactylid frog Crinia darlingtoni. Zoologische Mededelingen Moore, J.A. (1961). The frogs of eastern New South Wales. Rijksmuseum Van Naturolijke Historie Te Leiden Bulletin of the American Museum of Natural History 121: 47: 193-201. 149-386. Tyler, M.J. (1985). Gastric brooding: A phenomenon unique to Nattrass, A.E.O. and Ingram, G.J. (1993). New records of the Australian frogs. Search 16: 157-158. rare Green-thighed Frog. Memoirs of the Queensland Museum 33: 348. Tyler, M.J. (1997). The Action Plan for Australian Frogs. Wildlife Australia, Canberra, ACT. NSW NPWS (1994a). Fauna of north-east NSW forests. North East Forests Biodiversity Study Report No. 3. NSW Tyler, M.J. and Davies, M. (1985a). A new species of Litoria National Parks and Wildlife Service, unpublished report. (Anura: Hylidae) from New South Wales, Australia. Copeia 1: 145-149. NSW NPWS (1994b). Results of vertebrate fauna surveys of north-east NSW forests North East Forests Biodiversity Study Tyler, M.J. and Davies, M. (1985b). The Gastric Brooding Frog Report No. 3a,Vol. 1, Site and transect based methods. NSW Rheobatrachus silus. pp. 469-470 in G. Grigg, R. Shine and National Parks and Wildlife Service, unpublished report. H. Ehmann (eds) Biology of Australasian Frogs and NSW NPWS (1994c). Results of vertebrate fauna surveys of Reptiles. Surrey Beatty and Sons/Royal Zoological Society north-east NSW forests. North East Forests Biodiversity of New South Wales, Chipping Norton. Study Report No. 3a,Vol. 2, Incidental records. NSW Watson, G.F., Littlejohn, M.J., Hero, J.M. and Robertson P. National Parks and Wildlife Service, unpublished report. (1991). Conservation status, ecology and management of NSW NPWS (1995).Vertebrates of upper north-east NSW. the spotted tree frog (Litoria spenceri). Arthur Rylah NSW National Parks and Wildlife Service, unpublished Institute Technical Report Series No 116. Department of report to the Natural Resources Audit Council. Conservation and Environment,Victoria. Osborne,W.S. (1989). Distribution, relative abundance and Watson, G.F. and Martin, A.A. (1973). Life history, larval conservation status of Corroboree Frogs, Pseudophryne morphology and relationships of Australian leptodactylid corroboree Moore (Anura: Myobatrachidae). Australian frogs. Transactions of the Royal Society of South Australia Wildlife Research 16: 537-547. 97: 33-45. Osborne,W.S., Hunter, D.A. and Hollis, G.J. (1999). Polulation Webb, G.A. (1989). Notes on the biology and conservation declines and range contraction in Australian Alpine Frogs. of Philoria sphagnicolus (Moore 1958) (Anura: Pp 145-157 in Declines and Disappearances of Australian Myobatrachidae). Herpetofauna 19: 1-6. Frogs ed by A. Campbell, Environment Australia: Canberra. White, A.W. and Pyke, G.H. (1996). Distribution and Parris, K. and Norton,T. (1997). The significance of State conservation status of the green and golden bell frog Forests for conservation of Litoria pearsoniana (Copland) Litoria aurea in New South Wales. Australian Zoologist and associated amphibians. pp. 521-526 in P.Hale and 30: 177-189. D. Lamb (eds) Conservation Outside Nature Reserves. Wotherspoon, D. (1981). Note on the range of the Centre for Conservation Biology,The University of Sphagnum Frog Philoria sphagnicolus. Herpetofauna 2: 36. Queensland.

63 The status of Rainforest Stream Frogs in north-eastern New South Wales: decline or recovery?

Ross Goldingay, David Newell and Mark Graham*

ABSTRACT commonly encountered on forest tracks at night. Mixophyes fasciolatus was detected at 29 sites with Rainforest stream-dwelling frogs have declined an average abundance of 2.9 individuals per dramatically throughout eastern Australia.We transect. The distribution and abundance of these collected baseline information on the distribution species were not significantly influenced by and relative abundance of these frogs in north- elevation. Single surveys of sites in early 1999 eastern New South Wales.We established confirmed the distribution data. Mixophyes iteratus 100 m transects along rainforest streams at and Mixophyes fleayi were each found at nine sites 54 sites and traversed these transects at night, from 1997 to 1999. The former was confined to low counting all stream-dwelling frogs seen and heard calling. Repeat surveys were conducted at sites in elevation sites while the latter occurred almost 1997-98 to accommodate the influence of prevailing exclusively at high elevation. The rarity of these weather conditions on frog activity. species is of great concern despite moderately high numbers at 2-4 sites. Further monitoring is required Litoria pearsoniana was widespread in our study in order to resolve whether these frog populations area, occurring at 38 sites at an abundance of 7.1 are still in decline or have stabilised, and to identify individuals per transect. In contrast, L. lesueuri key locations where intensive management occurred at lower densities (4.2 individuals per programs should be initiated. transect, across 29 sites) but stream surveys may underestimate its abundance because it was

* School of Resource Science and Management, Southern Cross University, Lismore, NSW.

64 INTRODUCTION Many species of frog were recorded during this study including Adelotus brevis, Assa darlingtoni, Lechriodus fletcheri, There has been much recent attention devoted to assessing the status of amphibian populations in many parts of the world (e.g. Litoria spp., Mixophyes spp. and Philoria spp. However, we only Richards et al. 1993; Drost and Fellers 1996). This has stemmed report on the stream-breeding species Litoria pearsoniana, from the recognition that many species have experienced Litoria lesueueri and three species of Mixophyes.We present sudden and unexplained declines, and some species have the maximum counts for each species at each survey site for disappeared totally (Carey 1993; Pounds and Crump 1994; Lips the 1997–98 period. An additional count was made at each 1998). In Australia, as many as 9 frog species have disappeared creek site if a breeding aggregation was heard away from the in the last 15 years (Ingram and McDonald 1993; Laurance et al. fixed transect site. Searching was also conducted beyond the 1996; Mahony 1996). The seriousness of this situation is fixed transect at sites where there were historic records of reflected by the fact that no frog extinctions are known from M. fleayi in order to more fully assess its presence. the last 100 years in Australia. A set of handling and disinfection procedures was followed. Many frog species are still in decline and there is a need to Boots were disinfected between catchments by means of a assess the status of frog communities throughout Australia. The importance of monitoring programs has been dilute bleach mixture. Handling of frogs was kept to a demonstrated by studies in north Queensland (Richards et al. minimum and plastic bags (used only once) were everted 1993;Trenerry et al. 1994). Similar studies of rainforest stream over hands if frogs were handled. frogs are urgently needed at other locations because this group seems especially vulnerable to population declines. RESULTS

The aim of the present study was to locate suitable sites and Status of Tree Frogs establish a monitoring program for rainforest stream-dwelling frogs in north-eastern New South Wales (NSW). This area Litoria pearsoniana was detected at 39 sites in 1997–98 (Table 1) included three species, Litoria pearsoniana, Mixophyes fleayi and with an average abundance of 6.9 individuals per transect. Large Mixophyes iteratus, which have been reported to have suffered numbers (16–40) of individuals were recorded at several sites. serious declines (Ingram and McDonald 1993; Mahony 1996) Litoria pearsoniana was detected at 35 sites in 1999 (Table 1) and which are recognised as threatened or endangered with an average abundance of 4.4 individuals per transect. Litoria species (Lunney et al. 1996; Parris and Norton 1997). lesueuri was detected at30 sites in 1997–98 (Table 1) with an average abundance of 4.3 individuals per transect. It was METHODS detected at 23 sites in 1999 with an average abundance of 2.7 Study Area individuals per transect. The lower abundance in 1999 of both species was likely a result of only a single survey being The study area included forested areas in the Border Ranges conducted at each site which would have included periods of National Park (NP), Mt Warning NP,Limpinwood Nature lower activity. Litoria lesueuri was more widespread and abundant Reserve, Nightcap NP,Mebbin State Forest (SF), Nullum SF, than indicated by these data. It was regularly observed crossing Mooball SF,Whian Whian SF, Richmond Range NP and Toonumbar NP (Figure 1).Within all areas except Mebbin SF, roads in the Border Ranges NP,Mebbin SF and Whian Whian SF where some streams may be ephemeral, transects were when these areas were traversed at night, and the occurrence of established along permanent streams with a rainforest verge. large numbers of juveniles indicated that recent breeding had Between one and 12 transects measuring approximately 100 m been successful. Each of these species were found at sites that in length were established in each area (Table 1), depending on ranged in elevation from 70 m to 780 m. stream availability and access. All but eight of the stream sites were located within 500 m of an access road. Status of the Barred Frogs

Survey Methodology Mixophyes fasciolatus was detected at 31 sites in 1997–98 (Table 1) with an average abundance of 2.9 individuals per Site selection began in late 1996 when a small number of sites transect. It was detected at 24 sites in 1999 (Table 1) with an were selected and surveyed in Whian Whian SF and Border average abundance of 3.5 individuals per transect. Sites where Ranges NP.Most sites were selected in 1997 and surveys were this species was found ranged in elevation from 70 m to 780 m. conducted between January 1997 and April 1998. Because this study aimed to provide baseline data for long-term monitoring, Mixophyes iteratus was found at only seven sites in 1997–98, 2–3 surveys were conducted at night at each site under various all in Mebbin SF at low elevation (Table 1). It had an average weather conditions to ensure that we obtained appropriate data abundance of 4.2 individuals per transect. Juvenile frogs were for a range of species. This included at least one night when rain occurred. Two sites in Nightcap NP and seven sites in Whian detected at three sites, indicating that recent breeding had Whian SF were surveyed at least 14 times each between late been successful. At one site no adults were detected. In 1999, November 1997 and mid-March 1998. The majority of these M. iteratus was again found at seven sites in Mebbin SF but sites, in addition to a few extra sites, were surveyed on a single this included two sites where it was not detected the night in January-February 1999. A total of 54 transects were previous year. It was also found at two other locations which included in the study overall (Appendix 1); 47 were surveyed in were not included in the 1997-98 survey (Table 1). It had an 1997–98 while 49 were surveyed in 1999. Each transect was average abundance of 3.4 individuals per transect. searched using a 30 watt spotlight for a minimum of 20 person- minutes (number of observers x time).

65 FIGURE 1: Location of broad survey areas in north-eastern New South Wales.

Byron Bay

Key: 1. Whian Whian State Forest 8. Border Ranges National Park 2. 9. Mooball State Forest 3. Nullum State Forest 10. 4. Mount Jerusalem National Park 11. Richmond Range National Park 5. National Park 12. Bungdoozle Flora Reserve 6. Wollumbin State Forest (Richmond Range State Forest) 7. Mebbin State Forest

66 TABLE 1: Relative abundance of frogs at survey sites. A. 1997-98, B. 1999.Values are the maximum adult count of each species along 100 m transects (or maximum count of the nearest aggregation). - = none detected; j=presence indicated by 6 juveniles. NP = National Park, SF = State Forest.

Site (Elevation – m) L. pearsoniana L. lesueuri M. fasciolatus M. iteratus M. fleayi Mooball SF A. B. A. B. A. B. A. B. A. B. M1 (50) – – – 0 (2) 1 3 – – – – M2 (80) 12––54–– –– M3 (90) 11–5–2–– –– Mt Warning NP MW1 (430) 4 0 (2) – – – – – – 1 2 MW2 (220) 753–1––– –– Nullum SF N1 (270) 2 (6) 0 (4) – – 2 – – – – – N2 (130) 2 – 0 (1) 0 (1) 0 (1) – – – – – N3 (210) 8 0 (6) – – – 0 (2) – – – – N4 (230) 2 (6) 2–––––– –– Mebbin SF Mb1 (135) –54––25– –– Mb2 (145) –11––0 (1)–––– Mb3 (200) –31112 (3)––– Mb4 (170) 231––––0 (1)–– Mb5 (155) 183–1–11 –– Mb6 (170) –54–1–21 – Mb7 (165) 2 (14) 9 6 11 1 – – 2 – – Mb8 (165) – 7 (1) 2 0 (2) – 0 (1) j 4 (7) – – Mb9(155) 733–1–4– –– Mb10 (150) 40 6 0 (1) – – 2 (2) 0 (1) 1 (2) – – Mb11 (130) – 14 (8) – 3 – – 9 8 – – Whian Whian SF W1 (180) 22 1 (4) 4 1 8 – – – – – W2 (380) 1022–21–– –– W3 (540) 22––––– – W4 (360) 4–512––– –– W5 (330) 36414––– –– W6 (370) 5–3–5––– –– W7 (320) 2–7–2––– –– W8 (160) 522–4––– –– W9 (70) 4–424–– –– W10 (250) 4 16333––4 –– Nightcap NP Nc1 (200) 713233– –3 Nc2 (180) 12 6 13 1 3 – – – – – Nc3 (550) 3 2 – 0 (3) 3 4 (3) – – – – Nc4 (600) 2 – 2 0 (3) 2 0 (2) – – – – Nc5 (780) 2 nd 2 nd 3 nd – nd – nd Nc6 (520) 5–––23–– –– Nc7 (460) –2––38–– –15 Nc8 (470) nd 6 nd – nd 8 nd – nd 2 Border Ranges NP B1 (570) 11 6 3 0 (5) – – – – 2 – B2 (750) 5 1 (2) – – – – – – 1 0 (5) B3 (150) 4 nd – nd 3 nd – nd – nd B4 (730) 12 nd – nd – nd – nd – nd B5 (240) 7–5273–– –– B6 (675) nd – nd – nd – nd – nd 16 B7 (880) –––––––– –– Limpinwood NR L1 (560) 9 nd 4 nd 2 nd – nd – nd L2 (500) 101––22–– –– L3 (380) 8 – 6 0 (2) 5 2 – – – – Richmond Range NP RR1 (460) 18 2 (4) – 2 – 0 (7) – – – – RR2 (560) nd 7 nd 0 (1) nd 2 nd – nd – RR3 (300) –––3–8–6 –– Tooloom NP T1 (710) nd – nd – nd – nd – nd 8 T2 (690) nd – nd – nd 3 nd – nd 10 (10) Toonumbar NP T1 (210) 16 3 (4) 22 2 – 5 – – – –

67 Mixophyes fleayi was found at nine sites overall; three sites in and penetrated the rainforest areas of north-eastern NSW 1997-98 and eight sites in 1999, ranging from 200 m to 765 m by the early 1980’s and that it has been present here for well elevation (Table 1). This species was more abundant in 1999 over 10 years. Indeed, Mahony (1993) reported the with >5 individuals detected at four sites. These observations disappearance of several stream-dwelling frogs in the mid- represent new records for this species at Mt Warning NP,at 1980’s from the central coast of NSW. This may suggest that Tuntable Falls in Nightcap NP and at Kangaroo Ck in Tooloom the most virulent phase, when species decline rapidly, has NP. These results confirm that this species is rare in the study passed and that there may be hope to recover those species area and that populations are widely separated. that have declined but are still present. However, it would be unwise to be too optimistic because species that are reduced Influence of Elevation to very low levels of abundance may be demographically unstable. Surveys such as those employed here should be Elevation has been implicated as an influence on the decline continued to provide an on-going appraisal of the relative of stream-dwelling frogs in north Qld (Richards et al. 1993). stability of affected species. We investigated whether elevation influenced the distribution and abundance of our five target species by dividing up the In the present study, most sites were characterised by small sites into high (>300m) and low (<300 m) elevation sites. numbers of individuals. There were only three sites where We rely on the data collected in 1997–98 because multiple more than 20 individuals of any species were found. The visits were conducted to a site in order to reduce weather- absence of any studies reporting densities of these species induced variability in the detection of these frogs. prevent a resolution of how indicative these abundances are. These abundances may reflect a post-decline environment or Litoria pearsoniana was detected at 18 high and 21 low continued decline, assuming that a pathogen has been present. elevation sites but this distribution was independent of Sick or dead frogs have been reported from a small number elevation (G=1.80, P >0.1). Its abundance was not significantly of locations throughout our study area such as Mebbin SF different (t=0.44, P =0.65) at high (6.4 ± 1.1 individuals) (Berger pers. comm.). Studies in Qld have suggested that versus low (7.4 ± 2.0) elevation sites. Litoria lesueuri was species’ declines were more apparent at high elevation, but we detected at 10 high and 20 low elevation sites but this were unable to find an influence of elevation on distribution and distribution was independent of elevation (G=2.28, P >0.1). abundance for three species (L. pearsoniana, L. lesueuri, Its abundance was not significantly different (t=0.52, P =0.61) M. fasciolatus). Moreover, the two species of greatest at high (3.8 ± 0.6 individuals) versus low (4.7 ± 1.2) conservation concern (M. iteratus, M. fleayi) were found almost elevation sites. exclusively at either low or high elevation.We are now entering a period when active management of many populations is Mixophyes fasciolatus was detected at 13 high and 18 low needed and monitoring will be required to evaluate the success elevation sites but this distribution was independent of or otherwise of any management activities. elevation (G=0.01, P >0.9). Its abundance was not significantly different (t=0.05, P =0.96) at high (2.9 ± 0.3 individuals) versus Survey Methodology low (2.9 ± 0.5) elevation sites. Mixophyes iteratus only occurred at sites ≤300 m elevation. In contrast, all but one of the nine This study aimed to employ a survey method equivalent to sites where M. fleayi was found were >400 m elevation. that used by Richards et al. (1993). It was not possible to survey streams frequently, but rather it was intended to DISCUSSION obtain an index of frog abundance at each site. It became apparent that the activity levels of different species were not Frog Declines synchronised at a given site, largely due to different triggers for breeding. For example, the calling of male L. pearsoniana Frogs and other amphibians have undergone sudden and appears to peak during rainfall events while that of M. iteratus pronounced declines in protected rainforest areas in many and M. fleayi appears to precede rainfall (pers. obs.). To parts of the world (Heyer et al. 1988; McDonald 1990; accomodate these differences, we surveyed all sites during Ingram and McDonald 1993; Richards et al. 1993; Pounds and wet and dry periods in 1997-98. In early 1999, we were only Crump 1994; Lips 1998). In Australia, there has been able to survey a site once so these data should be speculation about the cause of the rainforest frog declines interpreted more cautiously. and its pattern of spread (see Laurance et al. 1996, 1997; Alford and Richards 1997; Hero and Gillespie 1997). The One short coming of our survey methodology was that focus of this debate has been whether a water-borne transects were located arbitrarily at a site. This may have led pathogen is the most probable cause of the decline, and the to the exclusion of important microhabitat components of a environmental conditions over which such a pathogen may particular species. Often suitable microhabitats (e.g. fringing be virulent. There appears to be some consensus that the vegetation) occurred just outside the transect. The practical frog declines have spread from the Conondale Ranges in way to account for this was to include an additional count of about 1979, based on the disappearance of the locally the nearest breeding aggregation detected outside the transect endemic Taudactylus diurnus and Rheobatrachus silus (Ingram if one was heard calling. This technique provided many useful and McDonald 1993; Laurance et al. 1996; Mahony 1996). data that produced a more accurate assessment of relative abundance and of the distribution across all survey sites. Most attention has been devoted to the northward progression of disappearing species from this area in south- east to north-east Queensland (Qld). This would suggest that if the agent of decline (e.g. a pathogen) can move as freely as it appears to have, that it must have moved south

68 Tree Frogs a further site (Byrrill Ck) where it was apparently detected by Manning (pers. comm.).We detected M. fleayi at the Border The most widespread and abundant species encountered Ranges sites in both years of the study but only in 1999 at during this survey were L. pearsoniana and L. lesueuri. Ingram Terania Creek, Nightcap NP. We were unable to locate it at and McDonald (1993) suggested that L. pearsoniana was a Byrrill Ck, Mebbin SF, despite >15 visits to this site.We detected species of some conservation concern. They found that it M. fleayi at two new locations (Mt Warning NP and Tooloom had disappeared from the Blackall Range and was present in NP). Each of these represent significant range extensions low numbers at other sites where it had previously been because these sites are well isolated from other locations. common. Laurance et al. (1996) suggested that it had declined by 90%. It is now listed as an endangered species in The abundance of M. fleayi was extremely low in 1997–98 Qld (Parris and Norton 1997). Litoria lesueuri has also with only 1–2 males detected at each site. However, larger declined in many areas of south-east Qld but appears to be numbers were detected in early 1999, which may have been recovering (Ingram and McDonald 1993). These observations due to the above-average rainfall experienced in that period suggest that these two species will be important for and also in late 1998. Mahony et al. (1997a) reported six evaluating the recovery of rainforest frog populations and individuals from Terania Creek in 1994 but none in 1996.We may provide insights that can be applied to other species. detected none there between late 1996 and March 1998, despite 17 visits under various weather conditions. However, We found that L. pearsoniana was widespread in the study three individuals were detected in 1999. Mahony et al. area and relatively abundant at a number of sites. This (1997a) reported no individuals at Brindle Ck in 1994-1996 situation offers good potential for assessing its population but we detected one individual in 1997-98 and five in 1999. stability and using it as an indicator of any changes in Moderate numbers (8–16) of M. fleayi were detected at four management. Litoria lesueuri was also widespread in the study sites in early 1999. This suggests that these may be more area but generally occurred at low abundance along the important sites and will require continued monitoring. transects. This may not be an accurate reflection of the abundance of this species because it was commonly Management Implications encountered when driving through the forest at night, under both wet and dry conditions. Therefore, the survey method There are several implications arising from this study: may not be suitable for assessing changes in the population 1. that surveys of these sites must be continued on an size of this species. Our surveys in 1997-98 suggest that the annual basis to provide an on-going assessment of the distribution and abundance of each of these species were not stability of the populations of three key species (M. fleayi, influenced by elevation. M. iteratus, and L. pearsoniana), Barred Frogs 2. that some assessment be made as to whether the factor that led to the decline of these species is still present and Mixophyes fasciolatus was the most common of the barred virulent, and frogs, being found at 42 of 53 sites over the period of this 3. that key locations be identified for detailed management study. Its distribution and abundance in 1997–98 was not intervention. influenced by elevation. This species is able to breed in ponds as well as streams (Mahony 1993) from which it may We assume that the rainforest stream-dwelling frogs of have derived some benefit which allows it to persist at many north-east NSW have been subject to some agent of decline sites. Mahony (1993) suggested that M. fasciolatus was not as for at least 10 years (see above). This suggests that these abundant on the central coast of NSW as it used to be. frogs have survived the most serious period of decline, unlike Whether its abundance has declined in north-east NSW is many species in Qld. Surveys must be continued to unknown but we have now established base-line data which determine whether these species are undergoing a more can be used for comparison by future studies. gradual decline or whether they may be recovering. In the case Ingram and McDonald (1993) reported that M. iteratus may of M. fleayi, it has only ever been known from a small number have disappeared from highland sites and that its status was of sites in NSW which makes it difficult to assess whether it uncertain. Mahony et al. (1997b) reported that this species has declined here. The discovery of new sites is the result of had disappeared from much of its southern range.We only more intensive survey, rather than an expansion in its range. detected this species at low elevation sites, at three broad The presence of M. fleayi at low and high density offers the locations. Mahony et al. (1997b) recorded 5 adults at Byrrill opportunity to investigate factors that may limit its abundance. Ck, in Mebbin SF.We recorded 9 adults there in 1998 and Detailed population studies at several sites should be initiated 8 adults in 1999.We found that M. iteratus was widespread in for M. fleayi and M. iteratus which are both listed as threatened Mebbin SF. This species could be readily detected during most in NSW. This would provide some calibration to the visits to this area, and the presence of juveniles at several sites population indices generated by occasional monitoring surveys. suggests that this population may currently be stable. Monitoring of populations is a first step to assessing the Mixophyes fleayi has a restricted distribution in north-east NSW status of populations and identifying focal species for and south-east Qld (Corben and Ingram 1987). Ingram and management. The second step is to determine whether an McDonald (1993) noted that M. fleayi was a species of agent of decline is still present in the environment of these conservation concern. Mahony et al. (1997a) noted the very species. Thus, although the factor(s) causing declines in this restricted distribution and extreme rarity of M. fleayi in north- frog assemblage is not resolved, it is still possible to ask eastern NSW. The current study included each of the sites whether the factor may still be operating. This could be where this species was detected by Mahony et al. (1997a) and assessed by conducting experiments using tadpoles of

69 affected species (e.g. L. pearsoniana). Tadpoles could be Lips, K.R., (1998) Decline of a tropical amphibian fauna. housed at a variety of sites to assess development and Cons Biol, 12: 106-17. survival. If sites are identified where survival is high, then Lunney, D., Curtin, A., Cogger, H. G. and Dickman, C. R., translocations could be considered to boost populations at (1996) An ecological approach to identifying the those sites or assistance could be provided to reduce endangered fauna of New South Wales. Pac Cons Biol, 2: mortality of tadpoles born in situ. Any attempt at 212-31. translocation should be mindful of mixing gene pools and Mahony, M.J., (1993) The status of frogs in the Watagan should limit translocation to relatively short distances Mountains area, the central coast of New South Wales. (e.g. 5–10 km). Pp. 257-64 in Herpetology in Australia: a diverse discipline. For species such as M. fleayi and M. iteratus, key sites should ed by D. Lunney and D. Ayers. Surrey Beatty & Sons, be identified where active management of their populations Chipping Norton. could be pursued. This may initially include trying to enhance Mahony, M.J., (1996) The decline of the Green and Golden tadpole survival in situ and ex situ, but may also involve the Bell Frog Litoria aurea viewed in the context of declines identification of sites for translocation studies. In the case of and disappearances of other Australian frogs. Aust Zool, M. fleayi, it is likely that captive breeding will be needed to 30: 237-47. prevent the loss of remaining individuals in NSW. This study Mahony, M., Knowles, R. and Pattinson, L., (1997a) 6. Gold- has shown that there are currently several sites available in eyed barred frog. Pp. 78-83 in Threatened frogs of New north-east NSW for further studies directed at resolving South Wales: habitats, status and conservation. ed by management issues for each of these species. H. Ehmann. Frog and Tadpole Study Group of NSW, Sydney South. ACKNOWLEDGEMENTS Mahony, M., Knowles, R. and Pattinson, L., (1997b) 5. We thank Lance Tarvey (NPWS) for sharing our enthusiasm Silverblue-eyed barred frog. Pp. 72-77 in Threatened frogs for this study. The NPWS Declining Frog Working Group of New South Wales: habitats, status and conservation. provided financial support that facilitated the field surveys in ed by H. Ehmann. Frog and Tadpole Study Group of NSW, early 1999.We thank Drs Steve Richards and Arthur White Sydney South. for many constructive comments on a draft of this paper. McDonald, K.R., (1990). Rheobatrachus Liem and Taudactylus Straughan and Lee (Anura: Leptodactylidae) in Eugella REFERENCES National Park, Queensland: distribution and decline. Trans Roy Soc South Aust, 114: 187-94. Alford, R.A. and Richards, S.J., (1997) Lack of evidence for Parris, K.M. and Norton,T.W., (1997) The significance of State epidemic disease as an agent in the catastrophic decline of Forests for conservation of Litoria pearsoniana (Copland) Australian rain forest frogs. Cons Biol, 11: 1026-29. and associated amphibians. Pp. 521-26 in Conservation Barker, J., Grigg, G.C. and Tyler, M.J., (1995) A Field Guide to Outside Nature Reserves. ed by P.Hale and D. Lamb. Australian Frogs. Surrey Beatty & Sons, Chipping Norton. Centre for Conservation Biology,The University of Carey, C., (1993) Hypothesis concerning the causes of the Queensland, Brisbane. disappearance of boreal toads from the mountains of Pounds, J.A. and Crump, M.L., (1994) Amphibian declines and Colorado. Cons Biol, 7: 355-62. climate disturbance: the case of the golden toad and the Corben, C.J. and Ingram, G.J., (1987) A new barred frog harlequin frog. Cons Biol, 8: 72-85. (Myobatrachidae: Mixophyes). Mem Qld Mus, 25: 233-37. Richards, S.J., McDonald, K.R. and Alford, R.A., (1993) Declines Drost, C.A. and Fellers, G.M., (1996) Collapse of a regional in populations of Australia’s endemic tropical rainforest frog fauna in the Yosemite Area of the California Sierra frogs. Pac Cons Biol, 1: 66-77. Nevada, USA. Cons Biol, 10: 414-25. Trenerry, M.P.,Laurance,W.F. and McDonald, K.R., (1994) Hero, J-M. and Gillespie, G.R., (1997) Epidemic disease and Further evidence for the precipitous decline of endemic amphibian declines in Australia. Cons Biol, 11: 1023-25. rainforest frogs in tropical Australia. Pac Cons Biol, Heyer,W.R., Rand, A.S., Goncalvez da Cruz, C.A. and Peixoto, 1: 150-53. O.L., (1988) Decimations, extinctions, and colonizations of frog populations in southeast Brazil and their evolutionary implications. Biotropica, 20: 230-35. Ingram, G.J. and McDonald, K.R., (1993) An update on the decline of Queensland’s frogs. Pp. 297-303 in Herpetology in Australia: a diverse discipline. ed by D. Lunney and D. Ayers. Surrey Beatty and Sons, Chipping Norton. Laurance,W.F., McDonald, K.R. and Speare, R., (1996) Epidemic disease and the catastrophic decline of Australian rain forest frogs. Cons Biol, 10: 406-13. Laurance,W.F., McDonald, K.R. and Speare, R., (1997) In defence of the epidemic disease hypothesis. Cons Biol, 11: 1030-34.

70 APPENDIX 1 Survey site details, including names and Australian Map Grid (AMG) references. Site names are from the creek name or nearest road name. Sites are grouped within broad locations. S = south, N = north, E = east, FR = Flora Reserve.

Site Number Site Name AMG Site Number Site Name AMG Mooball SF Whian Whian (cont.) M1 Palmvale Spur 54630 686100 W9 Quandong 53750 683300 M2 Barooka 1 54600 685930 W10 Rocky Ck 2 53460 683530 M3 Barooka 254600 685900 Nightcap NP Mt Warning NP Nc1 Protestors 53040 683930 MW1 Breakfast Creek 52780 685850 Nc2 Terania 53040 683970 MW2 Rest Area 52930 685880 Nc3 Bat Cave 53160 683950 Nullum SF Nc4 McNamaras 53190 684030 N1 S Chowan 1 53850 685030 Nc5 Geebung 53230 684160 N2 S Chowan 2 53720 685020 Nc6 Mulgum 52730 684180 N3 Sand Ridge 53480 684870 Nc7 Tuntable 1 52890 684120 N4 Scrub 53510 684810 Nc8 Tuntable 2 52910 684110 Mebbin SF Border Ranges NP Mb1 Swifts N 51620 685480 B1 Sheepstation FR 50340 685790 Mb2 Swifts S 51620 685460 B2 Brindle Ck 50670 686080 Mb3 Tank S 51630 685380 B3 Lynches 50250 685540 Mb4 Kolonga 51740 685400 B4 Gradys Ck 50730 686270 Mb5 O’Connors 51750 685380 B5 Sawpit Ck 48260 686170 Mb6 Bullocks Head 51700 685300 B6 Long Ck 48570 686770 Mb7 Plantation N1 51740 685200 B7 Collins Ck 51370 685420 Mb8 Plantation N251740 685190 Limpinwood NR Mb9 Plantation SE 51810 685170 L1 Hopping Dicks 51850 687400 Mb10 Lemon Tree 51860 685140 L2 Hidden Ck 51930 686720 Mb11 Byrill Ck 51870 685320 L3 Upper Oxley 51530 686800 Whian Whian SF Richmond Range NP W1 Big Scrub FR 53250 683170 RR1 Cambridge Plateau 47410 681230 W2 Duffs 53600 683130 RR2 Bungdoozle 47080 683530 W3 Tungun 53350 683950 RR3 Peacock Ck 47220 682950 W4 Rocky Ck 1 53530 683770 Tooloom NP W5 Boomerang Ck 53600 683380 T1 Kangaroo Ck 1 44410 684980 W6 Boggy Ck 53760 683520 T2 Kangaroo Ck 1 44090 684950 W7 Minyon 53780 683510 Toonumbar NP W8 Boomerang FR 53220 683620 Tr1 Iron Pot Ck 47530 684495

71 Frogs in the timber production forests of the Dorrigo escarpment in northern New South Wales: An inventory of species present and the conservation of threatened species

Francis Lemckert1 and Rachael Morse2

ABSTRACT at more than 30 sites, but the majority of frogs were recorded on less than 10 occasions. The five Surveys for frogs were performed at water bodies in the Dorrigo area of northern New South Wales. threatened species appeared no less numerous than The searches consisted of both aural and visual the majority of species considered to be more components and covered 182 breeding choruses. common. Protection of frogs within timber Additional records were obtained from opportunistic production forests has previously relied upon buffer sightings. Twenty nine species of frogs were strips of undisturbed vegetation around riparian located of which ten appeared to be restricted to zones. Recently, additional protection has been forested areas. provided for recognised threatened species through

Five are listed on the NSW Threatened Species the use of larger site specific buffer strips, the Conservation Act. Several species appeared to be provision of corridors connecting catchments and widespread and common in that they were located the recognition of soakage areas for protection.

1. Research Officer, Native Forest Management Systems, Forest Research and Development Division, State Forests of New South Wales. PO Box 100, Beecroft, NSW. 2119. Australia. 2. Technical Officer, Native Forest Management Systems, Forest Research and Development Division, State Forests of New South Wales. PO Box 100, Beecroft, NSW. 2119. Australia.

72 The retention of all rainforest, old growth forest STUDY AREA patches greater than 25 hectares and the A detailed description of the study area can be found in the maintenance of a minimum of 50% undisturbed State Forests of NSW Management Area report (Forestry Commission NSW 1980), but the general features are as habitat in harvested compartments has provided follows. The DMA covers 213 338 hectares of an escarpment further protection to probable foraging and shelter on the Great Dividing Range in northern NSW (see Figure 1). The land rises from 600m elevation in the south-east to over habitats. Specific research is required to determine 1000m in its north-western corner, with annual rainfall how forestry activities affect frogs and so determine decreasing along the same gradient from 2000 mm to 700mm if these protective measures will be as effective as (Forestry Commission NSW 1980). Temperatures are mild with a mean minimum of 14.5°C and maximum of 26°C. The planned through the long term. majority of land is privately owned, but 95 886 hectares are located within state forest and crown timber lands and a INTRODUCTION further 46 885 hectares are held in National Parks and Nature Reserves found wholly or partly within the DMA (State Frogs are a poorly known component of the New South Forests NSW 1995). The forest types in the area vary from Wales (NSW) vertebrate fauna with relatively few warm temperate rainforests through to dry open sclerophyll comprehensive and systematic surveys having been performed forests. Logging of these forests commenced in the 1900s and for this group. This poor state of our knowledge on both the increased in intensity until the 1930s and 1940s as a result of status of species and even their basic ecology is of concern increasing mechanisation. Currently, logging is of a selective given the recent noted serious declines in Australia (Mahony nature under a rotation cycle consisting of thinning operations 1993; Richards et al. 1993) and the rest of the world at 15 and 30 years and a cut of saw quality logs at 60 years. (Pechmann and Wilbur 1994). In 1992, the initiation of the Low intensity fires are prevalent in the dryer forests, both as a Threatened Species (Interim) Protection Act (TSIP Act) by the fire suppression mechanism and as a result of the activities of NSW government provided for the general protection of frogs graziers, but the moister forests are rarely subject to fire. and the targeted protection of threatened species which were listed under “Schedule 12” of the Act. Importantly, the process SURVEY METHODS of determining which species required listing as threatened Surveys for amphibians were performed over three separate made clear the very limited knowledge available on the biology time periods: and status of most frog species (Lunney et al. 1996). It also a. 26th of September to the 19th of December 1993 — provided a direction to determine the current status of when weather conditions were mostly warm and dry with threatened species, the processes threatening their survival and occasional light showers; the means to conserve each species into the future. b. 11th to the 17th of February 1994 — when moderate to In 1993, as part of a series of Management Area impact heavy rains fell before and during the survey period; and assessments, the Forestry Commission of New South Wales c. 21st to the 23rd of June 1994 — a period of dry and cool (now State Forests of New South Wales or SFNSW) conditions. commenced a major amphibian survey within the Dorrigo Management Area (DMA) of northern NSW. This work Survey techniques used followed those described in York et covered the requirements of the TSIP and Environmental al. (1991) and included both surveys of water bodies and the Protection and Assessment (EPA) Acts. The surveys in this traversing of roads to locate migrating frogs. The aim was to study covered all land tenures within the DMA and were search as many water bodies as possible (with an emphasis performed over an extended period of time in order to on forest areas) to provide as broad and intensive a coverage as possible in the time available.Water bodies were located cover varying seasons and weather conditions, thus both from maps and by travelling along roads to find maximising the opportunities to locate frogs. The information breeding sites visually or by listening for choruses of frogs. gained was used to assess the habitat requirements of The selection of sites was not systematic in terms of animals present within the DMA, and to assess the potential stratifying the study or selecting equal effort for the different effects of forestry operations on all species present, types of water bodies available, and so a measure of the particularly those considered to be threatened (see State relative abundance of each species can only be made with Forests NSW 1995). some caution. It was of greater importance to determine the distribution of species within the DMA and locate sites at This paper details the findings of these surveys. In particular, it which threatened species were present so that protective provides an assessment of the current status in the Dorrigo prescriptions could be applied for their conservation. Notably, area of species listed under the Threatened Species this approach almost certainly reduced the opportunities to Conservation Act (which replaced the TSIP Act in 1995) and locate terrestrial breeding species as their breeding sites were provides new information on their basic biology and habitat less obvious as to their location (a stream is obvious whereas requirements. Finally, the current conservation measures a small soakage is not). developed to protect frogs within areas of forest subject to forestry operations are discussed.

73 Each site was surveyed at least once in each of the first two hand Pseudophryne coriacea (70 sites) and Crinia signifera survey periods with a subset of sites being searched in the (66 sites) were never recorded on roads and all site records winter survey. Site surveys started with a three to five represent calling individuals, often more than ten at a site, minutes listening period to identify the species calling and making them the most abundant and widespread species their approximate numbers. This was followed by a spotlight located during the surveys. inspection of the water body and its surrounds to identify non-calling individuals which encompassed the entire bank The range of broad habitats from which each species were area of ponds (both temporary and permanent) or an recorded is also indicated in Table 1. Most species are notable average of 50 metres of both sides of streams. Longer lengths for utilising more than one habitat type, with a few species of streams were not generally traversed in order to increase appearing to favour either dry or moist forest types. Eight of the number of sites which could be covered on each night the species recorded on more than two occasions had 85% (the aim was to cover as many sites as possible). Additionally, or more of their records fall within areas substantially call playbacks were performed to elicit responses from non- covered by native forest (including plantations) and so are calling males and so further increase the chances of locating all of the species present at each breeding site. Road searches considered forest dependant. This includes four species of were also performed after dusk during periods of rainfall to conservation significance (Mixophyes iteratus, M. balbus, Assa identify frogs crossing roads. Searches eventually totalled a darlingtoni, Philoria sphagnicolus). minimum of 35 hours at pools, 45 hours at streams and 41 hours on road transects during this study. Endangered Species

On each occasion frogs were located, the position was The surveys located a total of five species listed as recorded to within the nearest 50 metres using speedometer threatened under the TSC Act with a greater concentration readings. The start and end time of each search was of these frogs being found on the eastern side of the DMA recorded along with the air temperature, humidity, cloud (Figure 3). Each of these frogs had previously been recorded cover, rainfall and wind strength. in the area, but the number of records was increased for each species except the sphagnum frog. Little information has RESULTS been published on the habits of these species and so general The Dorrigo Management Area proved to have a diverse accounts of the records obtained are provided as follows: frog fauna with 29 species located during these surveys. From the broad distributional maps of Cogger (1992) it was Hip-pocket Frog/Pouched Frog Assa darlingtoni considered that 40 species may potentially occur within the The six populations located are at the southern end of this DMA (see Table 1) and so 72.5% of those species potentially species’ range. All populations were within areas of wet present were located. However eight species (marked in sclerophyll forest or rainforest. Chorus sizes ranged from Table 1) are found in habitats not represented in the DMA more than 100 males to approximately 20 males with calling or are apparently extinct in this part of their range. Thus, being heard from September to November, although calling 29 of 32 of the species likely to be present (95%) were located. Notably, no “unexpected” species were located has been heard at other times of the year subsequent to this indicating that the predicted species list is relatively accurate. (Lemckert pers. obs.). Calling could be heard both during the day and at night, even during dry conditions. Calling sites Records of frogs consisted of 174 riparian and 8 terrestrial were under the leaf litter or, on rocky scree slopes, from breeding choruses and 176 incidental records of frogs on within spaces underneath or between rocks. The small size roads or in the forest (Figure 1). Thirty three records were (<30mm) and cryptic colouration of the hip-pocket frog obtained in rainforest, 90 in wet sclerophyll forest types, 136 meant that only aural surveys were effective for this species. in dry sclerophyll forest, 17 in eucalypt and native pine plantations and 82 within areas of cleared land. The hip-pocket frog appears to be adversely affected by logging of the forest in which they live. Of the six sites A species acquisition graph through time (Figure 2) demonstrates that half of the species located were recorded located during this study, only one was within forest that had within two days of the commencement of surveys and that been logged within at least the previous 30 years and at this the majority of species (20) were collected within six days. site logging had been performed within the previous six Additional species encountered after this time were the months. Several further populations have been located result of heavy rainfall which stimulated some species to call, subsequent to this study and all of these have also been in a change of survey season (for Pseudophryne bibronii — a late unlogged forest. Unfortunately a bush fire appeared to summer/autumn breeder) or specific searches for the destroy the population in the logged site so that its progress otherwise difficult to locate species Philoria sphagnicolus. could not be monitored, but the indications are that hip- The number of different sites from which a species was pocket frogs would have disappeared from the site. Back recorded during the frog surveys was highly variable, ranging burns reduced another population by 90% indicating that fire from 101 for Litoria lesueurii and 79 for Mixophyes fasciolatus may be a significant problem for this frog on any occasion it to just one for Litoria caerulea, Litoria jervisiensis, Litoria tyleri intrudes into the wet forests it inhabits (see discussion on and Pseudophryne bibronii. This is a misleading representation conservation protocols). however, in that the majority of records for the two common species were of single individuals on roads. On the other

74 FIGURE 1: Distribution of frog survey sites through Dorrigo area.

Prepared by Forest Research and Development Division, State Forests of NSW, February, 1999

75 TABLE 1: Species of frogs potentially occurring in the Dorrigo area and the number of records for each obtained during the surveys.

Number of record sites Myobatrachidae Rainforest Wet Scler Dry Scler Cleared Total Adelotus brevis * 227112 Assa darlingtoni *33 6 Crinia parinsignifera 314 Crinia signifera 212351766 Lechriodus fletcheri * 8146129 Limnodynastes dumerilii 112316 Limnodynastes ornatus + – Limnodynastes peronii 2 4 12 18 36 Limnodynastes tasmaniensis 639 Mixophyes balbus * 657 18 Mixophyes fasciolatus * 3 24 49 3 79 Mixophyes iteratus * 6 14 20 Philoria sphagnicolus 12 3 Pseudophryne bibronii 11 Pseudophryne coriacea * 2 20 46 2 70 Uperoleia fusca 312823 Uperoleia laevigata + – Uperoleia marmorata + –

Litoria aurea – Litoria booroolongensis – Litoria brevipalmata – Litoria caerulea 11 Litoria chloris 10 12 4 11 37 Litoria dentata 3 4 6 14 27 Litoria fallax 3 9 10 22 Litoria freycineti + – Litoria gracilenta + – Litoria jervisiensis 11 Litoria latopalmata *11112 Litoria lesueurii * 4 41 45 11 101 Litoria nasuta + – Litoria pearsoniana * 596323 Litoria peronii 5 6 26 11 48 Litoria phyllochroa 343616 Litoria piperata + – Litoria revelata 1416 Litoria rubella + – Litoria subglandulosa 36 312 Litoria tyleri 11 Litoria verreauxii 2 6 21 41 59

* Forest species + Species not likely to be present

New England Tree Frog Litoria subglandulosa to call playbacks. However, their habit of calling from raised positions on vegetation or from within logs made it difficult The 12 records for the were all the to locate visually. result of detecting chorusing males and were scattered across the DMA. This species was found in habitats ranging from Southern Barred/ Mixophyes balbus moist forest to generally cleared fields with strips of riparian Stuttering Frog vegetation (which held the largest recorded population). All populations were associated with flowing streams, varying in This frog was recorded at 18 sites (Figure 3), 11 of which width from between 30 cm to 3-4m, never still water sites. were located within moist vegetation, five were in dense Males were located calling from elevated sites amongst riparian vegetation in dry forest and two were from streams vegetation or logs located adjacent to or hanging over the with relatively open banks in dry forest. Calling males of this stream. This frog was only heard calling during September species were only found associated with flowing water, and October. This and later fieldwork indicates that the new although the flow was generally slow and the streams small england tree frog essentially finishes its chorusing by mid- (0.5m in width). On three occasions males were found calling November, which is very early for an Australian frog. This adjacent to trickles of water below stock dams where only species was easily detected by aural surveys, as it called very small pools were apparent and it appeared likely that regularly through the evening, and individuals responded well water flow was ephemeral, although the pools may be

76 FIGURE 2: Acquisition rate of frog species located during the Dorrigo frog surveys.

permanent. Calling populations consisted usually of only sclerophyll forest. It was evident during this study that 1–3 frogs, but minimums of 10 calling individuals were heard generalised survey techniques are inappropriate in locating at two different sites. Calling was heard from September to this species as they are often not associated with readily February. Males were generally found calling from under leaf identified riparian areas. For this species, very specific litter or within holes some 2–5 metres from the waters edge microhabitats needed to be targeted which could only be making this species difficult to locate visually. Several times reached by extensive foot travel in the DMA and so the during rainfall events this frog was found on roads at least status of this species could not be readily determined from 100m away from the nearest water body, and clearly some this study. individuals move widely through the forest when moist conditions prevail. DISCUSSION

Giant Barred Frog Mixophyes iteratus The Dorrigo Management Area, with 29 species, proved to be an area of relatively high frog diversity compared to other This was the most geographically confined of the three areas surveyed for forestry EISs. Surveys in the areas around Barred Frog species found in the DMA, with the Glen Innes (24 species — Smith et al. 1992) and 20 confirmed record sites occurring only at the eastern end (23 species — Schodde 1996) recorded the (see Figure 3). However, within this area it was found at every next greatest numbers of species with between 2 and 22 moderate sized (>2 metres wide), permanently flowing creek. species found in 13 other EISs (Lemckert 1996). This might Individuals of all sizes were easy to locate visually due to their indicate the DMA has a higher diversity of species than other large reflective eyes and habit of sitting above the leaf litter nearby areas, but general distribution maps (Cogger 1992; (as compared to M. balbus). Males were only heard calling Robinson 1993) would not suggest this. The majority of sporadically. All populations were found in wet sclerophyll species found in the DMA should have been present in these forest or rainforest. other areas. Rather, the difference is more a reflection of the Seven of these records came from timber plantations increased effort and flexibility directed at surveying frogs in indicating that this species is able to utilise disturbed habitats this study. The survey periods in the other EIS studies were to at least some degree. The DMA and the forests to the all constrained to a relatively narrow window of time (often east appear to now be the stronghold of this species as just one month; Lemckert 1996). In the DMA, surveys were populations are present throughout the drainage systems of performed over a nine month period allowing various this area, but have disappeared or severely declined to the seasons to be covered and more opportunity to survey north and south (Mahony 1993). The reasons for this decline under optimal conditions. The DMA was also surveyed far remain unclear as are the reasons for the continued more intensively, with the effort between two and five times abundance of this frog in the DMA. greater that applied to other MAs. This emphasises the fact that any survey attempting to understand the status and Sphagnum Frog Philoria sphagnicolus distribution of frogs within an area needs to take into account the difficulty in detecting this group of vertebrates. The sphagnum frog was recorded from only three sites, all of which were located at the far eastern end of the DMA. All three populations were located during diurnal searches and resulted from the daytime calling activity of this species. The sites inhabited were all soakages on steep slopes within wet

77 FIGURE 3: Location of record sites for threatened frog species.

Prepared by Forest Research and Development Division, State Forests of NSW, February, 1999

78 TABLE 2: Specific protective prescriptions for species listed under Schedules 1 and 2 of the Threatened Species Act.

Species Prescription Litoria aurea 5 hectare exclusion zone around known records Litoria brevipalmata 5 hectare exclusion zone around known records Mixophyes balbus 30m wide buffer 200m upstream and downstream of known sites Mixophyes fleayi 40m wide buffer 200m upstream and downstream of known sites Mixophyes iteratus 30m wide buffer 200m upstream and downstream of known sites Philoria kundagungan 50m radius exclusion zone around site Philoria loveridgei 50m radius exclusion zone around site Philoria sphagnicolus 50m radius exclusion zone around site Assa darlingtoni * Minimum 50m radius exclusion zone around southern populations

* Prescription applied voluntarily by SFNSW

The distributions and number of record sites for species The failure to locate Litoria brevipalmata was not a surprise as provided some unexpected results. Species such as Litoria this species is known to breed over only a few days of each tyleri and L. caerulea and to a lesser extent Limnodynastes year coinciding with heavy rainfalls (Barker et al. 1995). Even tasmaniensis and L. dumerilii were expected to be widespread regular checks of known breeding sites rarely result in this and common, but were not. The surveys were conducted species being found outside of breeding events (Lemckert during both wet and dry conditions which should have unpubl. data). The absence of Litoria aurea confirms the covered all species preferred breeding conditions and the findings of Mahony (1996). Although this frog has been reasons for the relative rarity of these species are unclear. recorded immediately to the east of the DMA, Mahony One possibility is that these frogs were at some form of (1996) found that Litoria aurea has disappeared from all cyclical low point in their abundance or detectability. Such upland areas within its historically known range. variations in amphibian abundances have been documented overseas (e.g. Meyer et al. 1998). Such a cycle has obvious Conservation of frogs in Timber implications for environmental management and a cautious Production Areas interpretation of survey results is required when determining Historically, the accepted method of protecting frogs in the status of a species.Whatever is the case, this result does forestry areas was the retention of “filter strips” of point out that surveys need to be intensive and cover several undisturbed vegetation along the banks of streams to seasons in order to have confidence in locating even some conserve breeding habitats. Recently however, the SFNSW apparently common species. Such time is rarely ever allowed and the NSW National Parks and Wildlife Service (NPWS) for these sorts of studies. have developed a set of agreed Conservation Protocols which provide multiple and broad ranging mechanisms to The habitats utilised by the species of conservation protect frogs within forests subject to logging operations significance were generally the same as the general habitat (SFNSW/NPWS 1997 and see Table 2). A core part of these requirements listed in standard reference texts, however protocols remains the retention of (most usually) a 20m wide there are a few notable differences. Barker et al. (1995) strip of undisturbed vegetation along the banks of streams records that Mixophyes balbus is found in rainforest whilst (=40m strip in total) to protect breeding habitat and Cogger (1992) generalises this to moist forests. In this study maintain water quality. This has however been extended to this species was also found within riparian areas in dry include both dams and soakage areas which protects non- sclerophyll forest. Only Robinson (1993) provides any stream breeding species. Additional protection has been assessment of the breeding season noting that for all granted to threatened species through the increase of the southern Mixophyes breeding appears to occur in late spring protective strips along streams from 20m to 30 or 40m for and early summer. In this study calling was heard from early 200m upstream and downstream of sites (see Table 2).Water spring to late summer for all three species of Mixophyes bodies utilised by Litoria aurea and L. brevipalmata are present in the DMA. Assa darlingtoni was recorded in both protected by the retention of 5 hectares of undisturbed rainforest and wet sclerophyll forest although both Cogger vegetation centred around the site whilst known sites for (1992) and Robinson (1993) note it to be restricted to species of Philoria are provided with a 50m buffer zone rainforest and antarctic beech forest. Most noticeably for (Table 2). Specific prescriptions for the protection of Assa Assa, all three texts indicate that the range of this frog darlingtoni were not set under the Conservation Protocols, extends approximately 50km into NSW, whereas Dorrigo is but SFNSW itself has applied a prescription whereby all areas approximately 200km south of the border. Litoria of forest inhabited by this species and a 50m buffer zone are subglandulosa is noted by both Barker et al. (1995) and reserved from forestry operations. Robinson (1993) as breeding from October to November. The calling season of this species did not appear to extend The need for dispersal routes has been considered in the beyond mid-December, but it was heard calling strongly in Protocols by the provision of both the retained strips of September at all of the sites in the DMA and should be stream-side vegetation and by the addition of cross considered an early spring breeder. catchment corridors. The latter provides for a minimum of either two 40m wide or one 80m wide corridor of vegetation to be retained every two kilometres of stream

79 length. Therefore, frogs have an opportunity to both disperse Barker, J., Grigg, G. C. and Tyler, M. J., (1995) A Field Guide to along streams and across ridges, although the effectiveness of Australian Frogs. Surry Beatty and Sons, Chipping such corridors as dispersal routes remains to be confirmed. Norton, NSW. Cogger, H. G., (1992) Reptiles and Amphibians of Australia The one aspect of the ecology of frogs that has been most (Revised Edition). Reed. Sydney, NSW. difficult to address has been their foraging needs. It is generally perceived that logging results in the clear-felling of FCNSW., (1980) Forestry Operations in the Dorrigo forest areas to leave a barren landscape. This may be true in Management Area. Forestry Commission of New South other countries or even some areas of Australia, however in Wales. Sydney, NSW. northern NSW logging has generally been of a selective Lemckert, F. L., (1996) Surveys for the Green and Golden Bell nature. The result is that trees are removed in a patchier Frog (Litoria aurea) by State Forests of NSW. Aust. Zool. manner, with the area logged being dependent on what 30: 208-213. percentage of an area had commercial value and the amount Lunney, D., Curtin, A., Ayers, D., Cogger, H. G. and Dickman, C. of riparian area that required protection. So even up to R., (1996) An ecological approach to identifying the recently, an average of 30% of any compartment remained endangered fauna of New South Wales. Pac. Cons. Biol. unlogged (Forestry Commission NSW 1980) and so 2: 212-231. significant areas of undisturbed vegetation were available for Mahony, M.J., (1993) The status of frogs in the Watagan foraging if a frog required such habitat. The Conservation Mountains area, Central Coast of New South Wales. Pp Protocols have provided additional protection and restrictions 257-264 in Herpetology in Australia: A Diverse Discipline. such that now no greater than 50% of the available logging ed by D.Lunney and D.Ayers. Surrey Beatty and Sons with area can be harvested, further increasing the area of the Royal Zoological Society of NSW, Sydney. undisturbed forest available for use. Mahony, M.J., (1996) The decline of the Green and Golden The newly increased protection available for frogs in general Bell Frog Litoria aurea viewed in the context of declines and threatened species in particular has been provided to and disappearances of other Australian frogs. Aust Zool, increase the likelihood that this group of vertebrates will 30: 237-247. persist in logging areas into the foreseeable future. However, Meyer, A. H., Schmidt, B. R. and Grossenbacher, K., (1998) the effectiveness of the Conservation Protocols remains Analysis of three amphibian populations with quarter unproved. Monitoring of populations is to be undertaken century long time-series. Proc. R. Soc. Lond. 265: 523-528. over the long-term, but if there are significant impacts over Pechmann, H. K. and Wilbur, H. M., (1994) Putting declining the long term declines may be detected only after they are amphibian populations in perspective: natural fluctuations difficult to reverse. There is a need for immediate, explicit and human impacts. Herpetologica 50: 65-84. studies to examine the effectiveness of these prescriptions. Such studies need to concentrate on each stage of the frog Richards, S. J., McDonald, K. R. and Alford, R. A., (1993) life-cycle to determine how and when impacts occur from Declines in populations of Australia’s endemic tropical forestry and so what changes can be made to management rainforest frogs. Pacific Conservation Biology 1, 66-77. practices to reduce the impacts of future operations. This Robinson, M., (1993) A Field Guide to Frogs of Australia: from program should most specifically target the species of Port Augusta to Fraser Island including . Reed. conservation significance which appear to be the species Sydney, NSW. most vulnerable to disturbance and/or have the most specific Schodde, R., (1996) Murwillumbah management area fauna habitat requirements. Effective prescriptions in combination survey. Report to State Forests of New South Wales. with the development of a comprehensive and adequate CSIRO. Canberra, ACT. reserve system are critical to ensure the continued survival of SFNSW., (1995) Environmental Impact Statement for the frog populations in the Dorrigo Management Area and in Dorrigo management area. Unpublished report by State other timber production forests. Forests of NSW. Sydney, NSW. SFNSW/NPWS (1997) Threatened species protocol: Survey ACKNOWLEDGEMENTS design and potential habitat. Joint unpublished report by Firstly, our thanks go to Garry Daly, Mark Fitzgerald, Liz Kelso, State Forests of New South Wales and the New South Allan Manning, Stefanie Pidcock and Jacqui Richards for their Wales National Parks and Wildlife Service. Sydney, NSW. help during the surveys. Thanks also go to the staff of the Smith, A.P.,Moore, D.M. and Andrews, S.P.,(1992) Proposed Dorrigo Office of SFNSW, and particularly Paul Roberts, in forestry operations in the Glen Innes Management Area providing help in various aspects of the surveys. Finally, we — Fauna Impact Statement. Report to State Forests of would like to thank Jacqui Recsei, Karen Thumm, Dr. Michael New South Wales. Austeco Pty Ltd, Armidale, NSW. Mahony and an anonymous reviewer for their critical reviews York, A., Binns, D. and Shields, J., (1991) Flora and Fauna of this document. Assessment in NSW State Forests. Survey Guidelines. Procedures for Sampling Flora and Fauna for REFERENCES Environmental Impact Statements.Version 1.1. Report Andrews, S. P., Gration, G. L., Quin, D. G. and Smith, A. P., (1994) for the Forestry Commission of NSW Sydney, NSW. Description and assessment of forestry impacts on fauna of the Urbenville Management Area. Report to State Forests of New South Wales. Austeco Pty. Ltd. Armidale, NSW.

80 Review of the declines and disappearances within the bell frog species group (Litoria aurea species group) in Australia

Michael Mahoney *

ABSTRACT Victoria nor have there been reports of widespread declines in the other three species (L. cyclorhynchus, Declines and disappearances of species and and L. moorei, in southern Western Australia, and L. populations of three species of bell-frogs have been dahlii in the north of the continent).The cause of reported, Litoria castanea has not been observed declines remains unknown; however, the introduced in the wild since the mid 1970’s, and L. aurea and fish Gambusia holbrooki may be responsible for, or L. raniformis have declined from parts of their play a role, in the declines in some regions.The bell former distribution.All populations of these three frog group has a number of features that make it a species formerly found on the Great Dividing Range suitable model to investigate the wider issue of in New South Wales have disappeared. declines among Australian frogs; captive breeding There has been a documented decline in and husbandry have been achieved, and large geographic range and abundance of L. aurea along numbers of offspring are available for the coastal plains of New South Wales.All extant experimental research. populations of L. aurea are below 150 m (ADH) and the majority of populations are found in coastal or near coastal habitats.Along the north coast the populations are apparently small and disjunct. There have been no declines in northeastern

* Department of Biological Sciences,The University of Newcastle, University Drive, Callaghan NSW 2308

81 INTRODUCTION with mottling of gold or bronze, all are relatively large, adult males reach to about 90 mm and adult females 100 mm The purpose of this paper is to review the decline and snout vent lengths. All occur in mesic environments and they disappearance of members of the bell frog species group are usually associated with water, their general distributions (Litoria aurea species group,Tyler and Davies 1978; Maxson are presented in Figure 1.The species group has one et al. 1982; King 1993). For each species there will be a brief representative in northern Australia, two representatives in review of the historical and present distribution, habitat the south-west of Western Australia, and three in requirements, biology, and the pattern of declines and southeastern Australia. No species penetrates the arid zone. disappearances where they have been documented, followed by a general discussion on the possible causes of declines and As might be expected for members of a species group they the research that has been conducted.The paper aims to also share similar habitat preferences and behaviour. Adults bring together the current state of understanding of the are usually associated with water, and they are active by day declines and disappearances within this group so that they basking in the sun. Individuals can usually be observed on may be compared with those in other amphibians. I have emergent vegetation above water or on the edge of water relied on published accounts and various technical reports bodies.When disturbed they drop into the water and remain (fauna surveys, environmental impact statements, submerged for a short period. Breeding generally occurs in still management plans, fauna impact statements and species water bodies ranging in size from ephemeral pools to large impact statements). Consequently, the review does not lakes, coastal floodplains and billabongs (Oxbow lakes) present the most recent findings from a number of research (Copland 1957; Main 1965; Humphries 1979;Tyler and Davies projects that have been in progress for some years and 1978; Barker et al. 1995). None of the species breeds in fast are ongoing. flowing streams, but they may use still pools in streams. A full account of the variation in breeding sites of Litoria aurea is SYSTEMATICS AND ECOLOGY provided by Pyke and White (1996), and details of the breeding habitat of the other species can be obtained from various Bell frogs as they are commonly known, are a distinctive sources (see Barker et al. 1995 for a general description). group within the Australo-Papuan hylid genus Litoria.The species group comprise six species L. aurea, L. castanea, L. These frogs are largely terrestrial, although termed “tree-frogs”; cyclorhynchus, L. dahlii, L. moorei, and L. raniformis (Tyler and they rarely climb trees, and have only small adhesive pads on Davies 1978; Barker and Grigg 1977; Barker et al. 1995). their fingers and toes.While they are generally closely associated All members of the group have rich green dorsal surfaces with water when observed by day, by night adults and juveniles

FIGURE 1: Distribution of members of the Litoria aurea species group.

82 may move widely in terrestrial situations when foraging range and apparent reduction in abundance, although the latter particularly during rainy periods. Individuals have often been has been more difficult to verify.The species can no longer be found considerable distances from the nearest water body found on the southern and , or on the (White 1995a), but evidence from a limited mark-recapture western slopes (see Figure 2). It has contracted from its former study of L. aurea indicated that some animals remain at or return northern extent along the New South Wales coast and extant to a breeding site when adult (Murphy 1996). However, the populations are apparently small and widely spaced (Clancy importance of movement (immigration, emigration and distances 1996). No populations have been detected on the eastern travelled when foraging) for different ages and sexes relative to hinterland of the Great Dividing Range in the past ten years edaphic features remains a matter of some debate. Seasonal (Lemckert 1996). Formerly it ranged in altitude from sea level activity varies among the species depending largely on the to about 700 m, but currently there are no populations known regional climate. In coastal areas of New South Wales L. aurea is above 150 m (Ravensworth in the Hunter Valley)(Interagency active at all times except in the midst of winter.At Homebush Advisory Group 1994;Wellington and Wells 1994; Resource Bay in Sydney and at Kooragang Island in the Hunter Valley Planning 1994; Mahony 1997). individuals overwinter in crevices in scree slopes and piles of rubble (Greer 1994, Hamer unpubl.). In the north L. dahlii may Most extant populations occur close to the coast. Since 1990 be observed at all times of the year basking on vegetation only 38 localities have been recorded and 19 of these are in the around wetlands, or in the dry months it hides in cracks in clay biogeographic region. Many of the populations are areas (Tyler and Davies 1986).The tableland species L. castanea small, estimates range from as few as 5 to 15 adults at some had a much shorter period of activity from October to April, isolated sites (Cogger 1993; Fanning and White 1994;White and adults required suitable sites for winter torpor (Humphries 1993, 1997;White and Pyke 1996).The largest known 1979; Courtice 1972; Courtice and Grigg 1975). Many of the populations are at Homebush Bay, the site of the year 2000 sites where the species was found experience frost during the Olympics development, and on the Kurnell Peninsula to the winter months, but reports of brumation sites are few. High southeast of Botany Bay.At Homebush the species occurs and altitude areas of the northern and southern tablelands have a breeds in semi-permanent pools in the uneven base of a large frost period that exceeds a median duration of 150 days per disused brick quarry (about 15 hectares and 20 m deep with year (Hobbs and Jackson 1977). Courtice and Grigg (1975) almost vertical sides (Greer 1994; Pyke 1995). Populations occur report that individuals of the New England tableland population in the land surrounding the quarry and have been recorded were found buried in soil beneath a tree stump, and under tin, breeding at several human-made permanent and ephemeral and Humphries (1979) reports that on the southern tablelands water bodies (Pyke 1995; Pyke and White 1996). Because of the groups were found under reeds at the edge of the breeding site. large developments proposed for the Homebush site, fauna In the south west of Western Australia L. cyclorhynchus and L. surveys were conducted and fauna impact statements prepared, moorei breed in spring and summer, and the species occur near including estimates of population sizes, habitat use and an swamps and permanent water (Main 1965;Tyler et al. 1984). I assessment of potential threatening processes (Greer 1994; have found L. moorei in late autumn in moist forest habitat Pyke 1995).With the aim of retaining an evolutionarily viable sheltering under sheet metal. Litoria raniformis breeds in spring population in the area inside and outside of the quarry, frog and summer and is active in the warmer months. Near Murray populations have subsequently been the subject of intensive Bridge and Mannum on the in South Australia monitoring, demographic studies, and investigation of animals have been found during winter in groups beneath thick microhabitat and habitat use by a postgraduate student Michelle beds of reeds on the edge of wetlands that occur adjacent to Christie from the University of Sydney, Drs Graham Pyke and the river (Mahony unpubl.). Arthur White from the Australian Museum Sydney and personnel from the Australian Museum Business Services. THE GREEN AND GOLDEN BELL-FROG Litoria aurea Wetlands on the Kurnell Peninsula provide significant habitat for the frog.The peninsula is primarily a series of pleistocene The biology and conservation of the green and golden bell- coastal sand dunes with formerly extensive holocene frog was the subject of a workshop held in Sydney at the freshwater wetlands (> 10 000 years of age)(Roy and Australian Museum and Taronga Zoo in 1994 and Crawford, 1978, 1979). Natural habitats in this area have contributions were published in a subsequent edition of the been greatly disturbed since first European settlement with at Australian Zoologist Volume 30(2) May 1996 edited by Pyke various times, cattle grazing, removal of vegetation, sand and Osborne.This publication has 17 papers that deal with mining and industrial and urban development. Large tracts of various aspects of the biology of the frog including the peninsula are currently covered by transgressive sand reproductive biology, habitat requirements, former and dunes destabilised by earlier activities (Pickard 1972; Skinner present distribution and abundance, cause of decline and 1973; Roy and Crawford 1979). Adduction sand mining has conservation management. resulted in mirror lakes, which are used by the frog for breeding. Some natural wetlands occur in swales between Distribution dunes on the northern end of the peninsula and in the The historical distribution of the green and golden bell frog has remnants of the Holocene freshwater wetland on the mid been reasonably well documented (Courtice and Grigg 1975). northern end of the peninsula (Mahony 1998).The continued White and Pyke (1996) compared the historical and current persistence of this population is threatened by an array of distribution prior to 1990 with the known distribution in 1996, developments that are currently in progress or proposed for and details of the former distribution and disappearance of the the peninsula, and there are few chances that interconnected southern tableland population were presented and reviewed natural habitat will be set aside (Gunninah 1997, 1998; by Osborne et al. (1996).There has been a contraction in Mahony 1997, 1998).

83 The remaining populations in the Sydney basin are all lands. At the same time he noted that the frog was not apparently small with estimates of less than 20 adults per site detected in adjacent agricultural lands where it would have (White and Pyke 1996). Outside this area larger populations been expected.This is at odds with the report of Gillespie are known to the south on the Illawarra coastline and to the (1996) who indicates that the species uses forest habitats in north in the Hunter Valley, however measures of the sizes of . However, Pyke and White (pers. comm.) these populations are based on counts taken at various times consider that the species is essentially a coastal and near- and not based on values for which estimation errors are coastal riparian wetland species in this region and is only available. Several large populations are known in coastal found in forest sites after moving from these breeding sites. swamp habitat south of Kiama and Nowra respectively They also note that the species does not use forest habitats (Daly 1995, 1996; Morgan and Buttemer 1996; Goldingay in New Zealand. A question remains whether the absence of 1996; Murphy 1996, NSW NPWS 1996a; Mahony 1997). the frog in forest habitats in New South Wales is a reflection In the Hunter Valley the largest population occurs on of its decline rather than its preference for other habitats. Kooragang Island near the mouth of the river. Much of the island is developed for industrial purposes but several large Habitat features at 74 sites, 18 where the species was extant natural and human constructed wetlands occur (Markwell with the breeding site identified, and 56 sites where the 1984; NSW NPWS 1996b). Disjunct populations also occur species no longer occurs and the breeding site was not on pastoral land 38, 98 and 124 kilometres up the valley. identified, were tabulated by Pyke and White (1996).The These populations are currently the most inland known in the result was a description of the habitat at the extant sites former distribution of the species. Populations on the north compared to historic sites, and therefore does not necessarily coast are widely spaced and the number in each population is identify the requirements of the species prior to declines. apparently small (Clancy 1996; Pyke and White 1996). They found that for a site to support a breeding population it In Victoria this species is restricted to east Gippsland (Brook should contain water bodies which are still, shallow, 1979a) where it occurs predominantly at low altitudes along ephemeral, unpolluted, unshaded, with aquatic plants and free the coast and hinterland. Gillespie (1996) compared its of Gambusia and other predatory fish, with terrestrial habitats current distribution with historical records, and found little that consist of grassy areas and vegetation no higher than evidence of a decline in the distribution or abundance of the woodlands, and a range of diurnal shelter sites. Breeding species in this region. He also observed that the landscape of occurred in a significantly higher proportion of sites with East Gippsland was relatively intact and possibly provides a ephemeral ponds rather than sites with fluctuating or greater continuity of habitat with a higher density of breeding permanent ponds, and where predatory fish were absent. or refuge areas than in regions where the species had According to their analysis both the level of water fluctuation declined further north in New South Wales. He reported and presence/absence of Gambusia have significant effects on that it was usually associated with stationary, mostly whether or not the frog breeds at a site.The coupling of permanent water bodies in both forested and cleared these two features may provide insight into the habitats that habitats. He also noted that apart from isolated occurrence in continue to support this species. Fish are purged from farm dams, Gambusia was absent from the region east of the ephemeral ponds during periods of extended dry and these in Victoria. ponds are often isolated, whereas once established in Populations occur on two offshore islands in New South permanent ponds or along drainage lines they remain and are Wales, Bowen Island in Jervis Bay (Osborne and McElhinney able to recolonise.The possible predatory role of Gambusia 1996) and Broughton Island north of Port Stephens holbrooki in the decline and continued pressure placed on (NSW NPWS Atlas 1998). It remains unknown whether populations of bell frogs will be discussed later. At this stage it these populations are relictual or the result of assisted is worthy of note that use of ephemeral breeding sites was translocation. Bowen Island is about 300 m and Broughton not a feature associated with members of the bell frogs Island about 2 km respectively from the mainland.The species group in earlier habitat descriptions (Courtice and Grigg has been successfully introduced to New Zealand (Robb 1975; Dankers 1977; Humphries 1979; Cogger 1992). 1980; Ford 1986,1989; Pickard and Towns 1988; Bell 1982), and (Cree 1984) and (Tyler 1979). Pattern of Decline Habitat Requirements The precise time at which populations began to decline and disappear is difficult to assess.White and Pyke (1996) provide The natural habitat requirements of the species have proved evidence that declines have occurred since 1990, but the difficult to define because it has been associated with almost pattern before that date is unclear. On the southern every type of water body except fast flowing streams (Pyke tablelands populations disappeared some time after 1977 and White 1996). Many of the sites where the species was formerly known have been disturbed, and the species still (Osborne et al. 1996). Assessment of the current distribution occurs at many sites that have experienced long periods of reveals that all remaining populations are at low altitude, the disturbance.This has led some to suggest that the frog is great majority are within a few kilometres of either the coast opportunistic or occupies a successional role (Pyke and or an estuary, and those few populations that are inland, such White 1996). Few historical records have come from within as those at Ravensworth and Bayswater in the Hunter Valley forest in New South Wales (Pyke and White 1996). Lemckert are in close proximity to a large lake.The significance of these (1996) detailed the results of numerous field surveys in distribution patterns relative to the broader phenomenon of forested and adjacent habitats in New South Wales and declining amphibians remains unknown, but they provide concluded that the species was rarely found within forested numerous avenues for further research into the causal agents.

84 Future Prospects NORTHERN TABLELANDS POPULATION Because of its occurrence in the large urban area of Sydney OF Litoria castanea and the fact that one of the largest remaining populations in Distribution the Sydney Basin occurs on the year 2000 Olympic development site, this species has received considerable The northern tableland population (L. flavipunctata of media attention.The plight of this species has therefore Courtice and Grigg 1975) was known from a relatively drawn public attention and awareness to the conservation restricted distribution centred around the town of Guyra on needs of frogs, and while it may not be as critically the New England Tableland (see Figures 1 and 2).This area is endangered as other members of the species group, it acts as on the higher section of the tableland and most historic sites a flagship species for all frogs.There are several ongoing and were above 1 000 m (ADH).There are 13 sites in this region long-term studies on this species that focus on its ecology, all of which have been verified by examination of museum causes of decline, and its potential for relocation and specimens or photographs (Mahony unpubl.).This population reintroduction. Captive husbandry has been reported by was allopatric from the southern tableland population several groups (Robinson 1993a) and two “A Class” zoos separated by a distance of about 500 km, it was not (Taronga Zoo, Mosman Sydney NSW; Australian Reptile Park, sympatric with any other member of the group. Somersby, NSW) have breeding programs dealing with There are no verified records of this population after 1975 specific populations. Hence we can expect that conservation (Ehmann and White, 1997), and the last specimen to be management of this species should continue to be enhanced placed in a museum was collected in 1973 (Australian and based on detailed field ecology. Museum register).The earliest record of this population on the northern tablelands, is a specimen from “Hillgrove” about THE SOUTHERN AND NORTHERN 25 km south-east of Armidale (see Figure 2) which until TABLELANDS BELL-FROG Litoria castanea recently was misidentified as a specimen of L. aurea. Details of Some uncertainty surrounds the taxonomic status of the the precise date of collection and the collector are unknown various populations of what is referred to here as Litoria but this specimen was registered with the Australian Museum castanea.Thomson et al. (1996) have provided a detailed prior to 1890. account of the issues involved, and in this account I have adopted their position and refer the northern and southern tableland populations to one species.

FIGURE 2: Declines and disappearance among members of the Litoria aurea species group.These maps are based on verified records in museum collections.The map for the southern tableland populations of L. aurea, L. castanea and L. raniformis is based in part on information from Osborne et al. (1996), and the map of L. aurea on the work of Pyke and White (1996).There have been reports of declines in L. raniformis in Victoria, however we not aware of any map that summarises the geographic pattern, for this reason its distribution is presented as a cross hatched area, based on the Atlas of frogs of Victoria by Brooks (1979).

85 Habitat Requirements Bibron’s toadlet (Pseudophryne bibronii) was encountered across the tablelands in many habitats, while recent surveys General features of the habitat and breeding biology were have failed to detect it at any site in the region except on the provided by Courtice and Grigg (1975). eastern fringes.These species were far more common, as far Pattern of Decline as can be ascertained by the number of records in the field register, and more widespread in distribution than the bell In the early 1970’s individuals of this population were frog. Sometime after 1973, along with the bell frogs, these apparently common in farm dams, near large tableland lakes species disappeared from across large areas of the northern and in still pools of streams. Professor Grigg (Department tablelands, to the point that they can now be regarded as of Zoology, University of Queensland) has a photo which regionally extinct. At the same time several other species shows eight individuals sitting on reeds basking in the manner such as Crinia signifera, C. parinsignifera, Limnodynastes dumerilli, typical of bell-frogs.The Australian Museum has a series of Lim. peroni, Lim. tasmaniensis, Litoria fallax, L. lesueuri, L. peroni 42 individuals collected in the vicinity of Little Llangothlin and L. verreauxi, that were also recorded in the register, Lagoon on 13/8/1971.There are also a number of anecdotal remain widespread and common in the region wherever reports, which indicate the species was once common. suitable habitat occurs.The available evidence indicates that Apparently staff from the Zoology Department at the the decline of the bell-frog was rapid, and that similar declines University of New England were able to collect reasonable affected two other species on the tablelands, however, these numbers to use for teaching purposes, and the landholder of species remain in populations at low altitudes. Habitats used a property (Millievale) south west of Guyra recounted that by these three species are markedly different; L. castanea students conducted field studies on the property and one breeds in still water situations, L. boorolongensis is closely feature was the bell-frogs around the farm house dam. associated with fast flowing rocky streams, and P. bibronii breeds in terrestrial soaks and swampy situations. Photographs in life show that members of this population were large and vividly marked and it may be expected that Gambusia is widespread on the New England Tablelands its disappearance would have been apparent. However, the however there is an absence of data on the time of extent of the demise of this population did not become introduction and rate of dispersion in the region (Harris 1995). apparent until the mid 1980’s when several herpetologists noted that it had not been observed since the mid 1970’s Future Prospects (Mahony 1996).There had been no systematic surveys of In the spring and summer of 1996-1997 a public awareness historic sites and other suitable habitat in the region in the program was conducted by the NSW National Parks and period between 1973 and the early 1990’s.The first detailed Wildlife Service in the hope that some remnant population systematic surveys were commenced in the summer of 1992 may be reported and to obtain historical information about (Mahony and Knowles 1993), and surveys continued for the species. A colour brochure with pictures of the frog, another three breeding seasons, and the region covered was typical habitat and maps of distribution was distributed, and a expanded. Unfortunately no population of the frog was travelling poster display was mounted at schools, libraries and found.The New South Wales Frog and Tadpole Study Group shire offices in the New England region. No population was also conducted a survey for this population in the summer of reported and no new information on the historic distribution 1993-1994 and 1994-1995 which focused on historical sites was obtained.The prognosis for this species is not good and potential habitat in the region, however no specimens unless an unknown population is discovered. were observed (Ehmann and White 1997).

To a great extent the knowledge that we have of this species SOUTHERN TABLELANDS POPULATION is based on the collections made by students, staff and visiting OF L. castanea researchers to the Department of Zoology at the University of New England. In the late 1960’s and early 1970’s there was Distribution a group of herpetologist working in the department with an Details of the decline and disappearance of the southern active field program in which series of animals were collected tablelands population of L. castanea have been presented and and placed into the department museum (Heatwole et al. reviewed by Osborne et al. (1996).The species apparently 1995).The collection and register is now held by the had a restricted distribution between Canberra and Bombala Australian Museum Sydney. Not only do we know of the on the southern tablelands at altitudes between 700 and 800 existence of this population and some details of its m. It was broadly sympatric with L. aurea in the north of its distribution because of their work, in addition the register range and with L. raniformis in the southwest of the region. provides invaluable information when trying to understand its pattern of decline and disappearance.Thus unlike many other Habitat Requirements regions in New South Wales there is a reasonable base-line upon which to base comparisons of distribution, abundance Details of its biology and ecology are available from a study and former community composition. of the dynamics of a breeding frog community conducted by Humphries (1979) at Oakdale (700 m alt.) near the During the same period of active collection in which the bell NSW/ACT border between 1973 and 1977.The study frogs were encountered, large series of the stream hylid provides some of the best evidence of the relative Litoria boorolongensis were registered.This species has now abundance of the southern population and its habitat also disappeared from the tablelands. Similarly, there are requirements. Litoria castanea was sympatric with L. aurea at numerous occasions in the register’s field notes where the the study site.The most significant finding was that both various species making up a frog community were recorded. L. aurea and L. castanea (dealt with by him under the name

86 L. raniformis, although he was aware that the taxon he was Pattern of Decline studying differed from L. raniformis sensu stricto), were In retrospect it is apparent that the disappearance of the residents at the permanent pond where they bred. By northern and southern tablelands population of L. castanea comparison with other species in the community he regarded occurred in a similar time frame, however, the pace at which the bell frogs as the only residents,“during the breeding the decline and disappearance occurred cannot be accurately season they are semi-aquatic, with males spending a major assessed, because no monitoring was conducted.The available fraction of their time in or near water. Few adults emigrated, evidence suggests that the decline was rapid, to the extent and if they did the movement was not related to that biologists were not aware that it was occurring. Osborne reproduction. Many juveniles left the pond after et al. (1996) concluded that the declines of members of this metamorphosis, but a considerable number remained, and group on the southern tablelands was rapid and occurred reached sexual maturity within the area bounded by the drift sometime between 1978 and 1981 and “did not involve a fence”. Of the six ponds in the study area the one most prolonged stage when the frogs were in low numbers”. intensely studied had been constructed as a stock water Surveys, discussions with experienced local herpetologists, pond in a drainage line, it was roughly circular in shape with a and comparison of historical field records confirmed that all diameter of about 20 m, and had extensive emergent and three species of bell frogs on the southern tablelands suffered floating vegetation. an extensive decline, with no confirmed records since 1980.

Bell frogs occurred in much lower densities than the other At the time the species disappeared no immediate concern “migratory” species, which used the pond for breeding, was raised when a population or species was not observed presumably because of their larger size (Humphries 1979). during field work.Variable climatic conditions, including The total number of adult L. aurea of both sexes at the study drought, coupled with the cryptic nature of some species pond was 21, 26 and 27 over successive seasons. were assumed to explain the failure to observe some species. Comparable figures for L. castanea cannot be stated precisely It was often assumed that a species was absent because field because an experimental introduction of 38 males was made work did not coincide with the preferred climatic conditions into this enclosed pond in November of 1975. Prior to this of the particular species. Often local anthropogenic impacts the number in the first year was 6 adults, and the number of were also considered responsible for the absence of a adult females did not rise above 3. Recapture rates within species and searches over wider areas were rarely initiated. In Australia there have been few close ecological studies of and between seasons were high for both species. frogs that have considered temporal patterns of reproduction At Oakdale the period of breeding was shorter than for with seasonal or climatic condition. Hence, when frogs first populations of bell frogs on the coast, starting a month or disappeared no alarm was raised and no wider or systematic two later, but adults were active until mid-April, when low surveys were conducted to determine the scale of declines. temperatures inhibited activity. After this time communal brumation occurred each year in the permanent pond, THE SOUTHERN BELL-FROG “Frogs of both sexes and species were found together, torpid Litoria raniformis beneath leaf litter at the base of large Carex appressa tussocks. Mild, wet weather occasionally permitted Distribution emergence, and some feeding.”(Humphries 1979). Fidelity of This species was distributed across a large area of south east the resident adults at the pond was facultative; the soil near Australia including Tasmania (Figure 1). In New South Wales two other ponds dried during autumn, and frogs of both and the Australian Capital Territory the range of the species species “emigrated from these ponds to overwinter”. It is not was centred on the Murray and valleys indicated to where these animals emigrated, and migration and their tributaries. It occurred throughout the Southern towards these ponds was observed during the breeding Tablelands.The species was also recorded on the central season. Newly-metamorphosed juveniles foraged later in the tablelands as far north as Bathurst (Ehmann and White year than adults, and used the sheltered belt of littoral 1997).The species was widespread across Victoria being only vegetation. At the well vegetated study pond many juveniles absent from the western desert regions and the eastern reached sexual maturity and were recruited into the breeding alpine regions (Littlejohn 1963,1982; Brook 1979a; Hero et al. population, while many others dispersed, and a few 1991). In South Australia the species is known to occur along individuals born elsewhere entered the pond. the lower Murray River Valley, the lower south-east to near Keith, and a small, apparently introduced population, in the Population age class structure indicated that males and Adelaide Hills (Tyler 1978). In Tasmania the species occurred females reached maturity in their third year and lived for at broadly across the north and east of the island, and on the least six breeding seasons. Furthermore, individuals, marked in Islands (Brook 1979b). October 1973, remained at the one breeding pond over six breeding seasons. Measures of fecundity were limited, two Habitat Requirements clutches of L. castanea had 1 885 and 3 893 eggs.These Habitat requirements are broadly similar to L. aurea, although values are considerably lower than reported for L. aurea at several authors mention that it was associated with low altitudes and maturity has been reached in the first year “permanent” water bodies, hence there may be some in this species under artificial conditions (White 1995a; question whether this species differs in habitat requirements Hamer and Mahony unpubl. data). from L. aurea which appears to use ephemeral pools in some cases (Tyler 1978; Hero et al. 1991; Cogger 1992; Robinson

87 1993b, Ehmann and White 1997). However, because of the the introduction of exotic fauna particularly predators considerable ambiguity associated with the word ephemeral (Bradford 1989; Blaustein and Wake 1990; Bronmark and it is difficult to compare habitats of bell frog species. It Edenham 1994;White and Pyke 1996). In this latter category, occurred across a greater altitude than all other members of it has been suggested that the introduced fish Gambusia may the species group, with populations ranging from sea level to have been responsible for the decline of members of the bell 1 300 m on the Southern Tablelands (Osborne et al. 1996). frog group in eastern Australia (Mahony 1993;Webb 1994; White and Pyke 1996; Morgan 1995; Morgan and Buttermer Pattern of Declines 1996;Webb and Joss 1997).

Declines have occurred in sections of the range of this Investigations of disappearances among bell frogs have focused species. Consistent with the pattern already reported in primarily on the declines in L. aurea and L. castanea.Two major other members of the species group, all tableland populations directions of research have been pursued, the role of increased have disappeared. In a review of the species distribution and ultraviolet radiation, and the impact of Gambusia. In several status, Ehmann and White (1997) noted that in New South cases the impact of Gambusia has also been linked with Wales the species has disappeared from sites in the central investigation of habitat requirements of the species. Attention and southern highlands.The disappearance of populations to these issues has been required because of the need to from the Southern Tablelands of NSW/ACT has been actively manage populations in certain areas of Sydney. reviewed by Osborne et al. (1996). It is currently known from the district near the Victorian border and is widespread throughout the Murray River valley but has THE ROLE OF GAMBUSIA disappeared from a number of sites along the Murrumbidgee Results of several studies are consistent with the hypothesis River. They noted that in Victoria the species “still has a that Gambusia contributes to the decline of frog populations. widespread distribution ....although some declines have been The first study to indicate the impacts of this introduced fish reported from central Victoria (Gillespie pers. comm.)”. on native frog communities was conducted by Dankers (1977) who found that the number of metamorphosing frogs To my knowledge there are no published accounts which was much higher in the part of the study pond from which document declines in populations of L. raniformis apart from the fish was excluded, compared with the remainder in which those in the upland areas on the southern tablelands of Gambusia activity was not reduced. Litoria aurea was present NSW (Ehmann and White 1997; Osborne et al. 1996).There in the pond community studied and it has subsequently are however numerous informal reports of declines of this disappeared. More recently several studies have shown that species in southern Victoria and in Tasmania which require Gambusia will attack and eat tadpoles including those of the validation by systematic survey and monitoring. green and golden bell frog (Harris 1995; Morgan and Buttermer 1996; Pyke and White 1996;Webb 1994;Webb Litoria cyclorhynchus, L. moorei and L. dahlii and Joss 1997). However, the importance of Gambusia as a To my knowledge there are no reports of declines of either predator relative to other factors in causing the decline of of the Western Australian or northern species in this bell frogs remains unclear.The role of introduced predators in species group. the decline of native species cannot be underestimated, along with habitat destruction they are particularly responsible for Reynolds (1995) examined the impact of Gambusia on the the extinction of island faunas (Clout 1996). Heyer (1973, eggs and larvae of six species of frog, including L. moorei,on 1976) demonstrated that predation and larval habitat the Swan Coastal Plain Western Australia. He found that preferences that evolved during speciation were significant in L. moorei coexisted with Gambusia in nine lakes in this region permitting coexistence. If fish predators were absent during and he suggested that the fish was responsible for reductions evolution, as may be the case with bell frogs, the frog would in frog population sizes in some instances. not be expected to have defensive adaptations, or it may have adaptations to cope with native fish but not with CAUSES OF DECLINES IN THE BELL introduced fish. FROG GROUP Studies on the impact of Gambusia have taken two lines, first Declines among Australian frogs are recognised as being part to show that eggs and tadpoles of bell frogs and other frog of a global pattern of declines among amphibians. A major species are palatable to the fish and that it is an active question is whether the cause of most declines is a single predator of tadpoles, and secondly, investigations of the widespread factor, which could explain the global extent of the distribution of the fish, the habitats it occupies, phenomenon, or a range of factors that are involved at and the relation between its distribution and that of frogs. regional and local levels? On the global scale a range of Gambusia was found to be a voracious predator on the factors have been suggested as possible causes, including tadpoles of green and golden bell frogs (Morgan and increased ultraviolet radiation related to ozone layer Buttermer 1996; Pyke and White 1996) and a number of depletion, pollutants dispersed in the atmosphere, climate other native frogs (Harris 1995; Morgan and Buttermer 1996; change and the action of a virulent pathogen (Blaustein and Webb and Joss 1997). Studies have dealt with various factors Wake 1990; Barinaga 1990; Pounds et al. 1997). On a including predator prey ratios, density of predators and the regional scale there have been examples of declines due to effect of available cover or substrate. In all cases the fish pollution (insecticides and herbicide use), reduced water rapidly killed tadpoles. quality (Warner et al. 1993), habitat destruction and isolation (Barinaga 1990), altered water flow regimes (Hero 1991), and

88 A criticism that can be leveled at the studies that have The relationship between the decline and disappearance of examined the impact of Gambusia is that rarely have they bell frogs and the role that Gambusia has played leads to a measured field densities of the fish and of tadpoles in the conundrum.There are sites where the frog has disappeared absence of fish and constructed experiments that deal with but where the fish is absent, and there are sites where the these natural densities. In general the laboratory densities of frog can be found and the fish are present. As an example of fish are probably higher than encountered in the field and the first case, I have compared two subcatchments of the therefore the results unrealistic. Hunter River near Ravensworth, one Bettys Creek (about 30 sq km) is free of any Gambusia, while the adjacent Gambusia are present in all states of Australia except catchment Mine Creek (about 40 sq km) has a large Tasmania (Arthington and Lloyd 1989), however, there is little population of the fish. Each catchment has much the same documented information on their spread in various regions. aspect, experiences the same climate, includes a variety of Harris (1995) examined a large number of waterbodies in waterbodies including permanent ponds, lotic stream habitats, the New England region around the town of Armidale and and ephemeral sites, and both have a long history of clearing the fish was found to be widespread in streams, dams and followed by pastoral use.The area is currently the site of lakes, but it did not occur in all waterbodies. Pyke and White several large open cut coal mines. Bell frogs have been (1996) reported the number of historic sites used by the known in the area for some time, the Museum of Victoria and green and the golden bell frog that had populations of fish. the Australian Museum have specimens collected at There is a great deal that remains to be understood about Ravensworth and nearby Lake Liddel in 1971.Two property the impact of Gambusia on individual species and on frog owners who have lived in the Mine Creek catchment for communities.These studies will have important implications over twenty years were able to describe the frog and for the management of habitats to maintain natural frog recognised that it was once common in dams on their community composition and the abundance of frogs.The properties. A small population is extant in one section of the variation found among water bodies makes extrapolation of Bettys Creek catchment (Wellington and Wells 1994; Mahony ecological conclusions difficult. Future studies need to focus 1995, 1997). No extant population of the frog has been on the relationship between the distribution and number of found in the Mine Creek catchment in the past three years ephemeral water bodies and distance from nearest despite intensive systematic searches, while a small population permanent water bodies. For species that are particularly has been detected in only two ponds during each summer susceptible to Gambusia, the occurrence of ephemeral season in the Bettys Creek catchment. Gambusia cannot be waterbodies surrounding permanent water bodies may the limiting factor to the distribution and abundance of the provide valuable refuge sites. Gambusia and frogs could frog in the Betty’s creek catchment, because the fish is absent coexist during periods of high rainfall due to the presence of from all water bodies.This argument has numerous caveats ephemeral waterbodies.These waterbodies provide a refuge because it may be that there are more than one limiting from the fish.While various studies have revealed that the factor to the distribution and abundance of the frog. eggs and tadpoles of bell-frogs are palatable to Gambusia Examples of the second case, where the fish and frog coexist, under laboratory conditions, to date there have been no are less common. I am aware of at least two sites, which have similar experiments under field conditions. It is unknown if in the latter characteristics, Coomaditchy Lagoon south of ponds with submerged weeds and cover whether some egg Wollongong (van der Mortel and Goldingay 1998), and on masses and tadpoles escape predation. Kooragang Island near the mouth of the Hunter River Could the introduction and dispersal of Gambusia explain the (Hamer and Mahony in prep.). In both cases there are a disappearance of the tableland populations of bell-frogs and variety of waterbodies available for reproduction and tadpole the decline of the low altitude populations of the green and development.Van de Mortel and Goldingay (1998) reported golden bell frog? that the lagoon contained high densities of the fish however tadpole survival and adult recruitment at this site appeared Several important pieces of information required to answer high. Unfortunately quantitative measures of fish and tadpole this question are absent.The dates of introduction of densities at sites have not been reported. A study of the Gambusia to many regions are not documented, and even if comparative recruitment success from different water bodies they were, the impact may have been gradual and the on Kooragang Island has not been completed. It may be that apparent rapid disappearance of the frogs the final phase of a in such situations the population of the frog is maintained by long process. Differences in declines between high and low recruitment from ephemeral pools and that these represent altitude populations may be due to differences in the length a situation where some sites are sinks and other sources in a of the activity period of frogs which is determined by climatic metapopulation sense. factors coupled to impacts such as predation. Finally, it should be borne in mind that Gambusia have not been implicated in Finally, while the importance of Gambusia relative to other the disappearance of any of the seven species of rainforest factors in causing the decline of the bell frogs remains unclear, stream frogs in eastern Queensland (see review by there seems little doubt that at sites where the frog is being McDonald and Alford 1999), and they are absent from managed the control of Gambusia is something for which numerous former breeding sites on the southern and strategies can be developed and implemented. Both the decline northern tablelands where various members of this group of the species in places where Gambusia was absent and the have disappeared (see Osborne et al. 1996, Mahony 1993). apparent ability of the green and golden bell frog to survive and breed in water bodies where Gambusia are present serve to highlight the likely importance of other factors.

89 INCREASED LEVELS OF ULTRAVIOLET CONCLUDING REMARKS RADIATION It is clear that many issues surrounding the decline and Experiments to test the hypothesis that increased ultraviolet- disappearances of bell frogs require in depth investigation. B (280–320 nm) radiation was adversely affecting L. aurea I have postulated elsewhere (Mahony 1996) that in one populations were conducted by van de Mortel and respect, unexplained rapid disappearances from upland Buttermer (1996).They conducted the experiment twice, populations, the bell frogs are similar to other species that firstly using three clutches of eggs (late December 1994) and have disappeared, while on the other hand they have marked secondly using only one clutch (mid February 1995). In the differences in reproductive biology and habitat. Several first experiment they found no significant effect on hatching aspects of the biology of the green and golden bell frog make success at two elevations (sea level and 600 m) between it an ideal candidate to study the cause of disappearance in sunlight filtered for UV-B, unfiltered sunlight and a filtered other Australian frogs. Study of this species, as an exception control (p=0.749). In the second experiment hatching success to the general situation, may increase our understanding of was significantly higher under the UV-B filter than under the broader pattern and as a result improve our abilities to unfiltered sunlight (p=0.017).The first experiment was conserve all endangered frogs.This frog has several conducted under cloudy conditions and the second under advantages as a research model. Successful captive husbandry clear skies.These experiments were in several ways is possible, something that has not been achieved for any preliminary. Impacts were investigated on hatching success other Australian species listed as critically endangered or and did not follow effects on further growth and endangered.When coupled with high levels of fecundity, bell development. Unfortunately, they were unable to measure frogs lay between 4 000 and 8 000 eggs, this enables large field UV-B levels, and they observed that it would be numbers of eggs and tadpoles to be available for desirable to construct the dose-response relationship of experimental studies. Hence, stock has been available to L. aurea eggs to UV-B radiation to assess their results.There is conduct experiments on UV-B tolerance and on the impact a vital need to further this line of investigation and build on of predators. Adults are large enough to be individually the results already obtained. marked using passive transponders enabling population demographic studies (Christie 1996), and a number of DISEASE populations of green and golden bell frog in the Sydney basin have been compared genetically and measures of genetic The occurrence of sick and moribund individuals has not diversity obtained (Colgan 1996).Two populations are known been associated with the declines and disappearances of from offshore islands and the species occurs on several members of this group. Absence of sick individuals does not pacific islands where it was introduced.This enables studies rule out the possibility of disease being a factor, it simply may on population demographic structure, dispersal and be that because there was no active research on these metapopulation structure. Several programs are currently in animals at the time disappearances occurred no sick progress to manage populations of the green and golden bell individuals were encountered.While this was the case in the frog, including habitat restoration and enhancement, period 1977 to 1985, it is less so in recent years during which reintroduction, and relocation.These provide invaluable there has been considerable attention given to low altitude opportunities to obtain information on the biology of this populations of the green and golden bell frog. Some sick and frog and should therefore provide insight into the declines dead individuals of L. aurea have been collected and and disappearances among other Australian frogs. deposited for pathology, but to date no pattern consistent with other cases where a pathogen has been implicated has ACKNOWLEDGMENTS emerged.We have proposed elsewhere (Mahony, 1995 ) that the bell frog group provides a useful model in which to test Special thanks to Steve Donnellan and Arthur White for hypotheses about the possible role of disease. critically reading the manuscript.

DROUGHT REFERENCES Attention has been drawn to the possible role of drought in Arthington, A.H. and Lloyd, L.N., (1989) Introduced Poeciliids the disappearance of bell frogs.The period between 1977 in Australia and New Zealand. In G.K. Meffe and F.F. and 1982, when the upland species and populations Snelson J.R. (eds) Ecology and Evolution in Livebearing disappeared, represented a period of below average rainfall Fishes (Poeciliidae). pp. 333-348. Prentice Hall, New Jersey. and drought conditions in the northern and southern Barinaga, M. (1990) Where have all the froggies gone? Science, tablelands (Osborne et al. 1996; Mahony unpubl. data). 247:1033-1034. However, examination of long-term records indicate that Barker, J., Grigg,G., and Tyler, M. J. (1977) A Field Guide to drought is a regular feature of these regions (Osborne et al. Australian Frogs. Rigby, Adelaide. 1996). On the northern tablelands several of the large mountain lagoons have contracted, but these sites have rarely Barker, J., Grigg,G., and Tyler, M. J. (1995) A Field Guide to completely dried out (Casanova and Brock 1996). Similarly, Australian Frogs. Surrey Beatty and Sons. Chipping on the southern tablelands several of the sites where Norton. members of the group occurred did not become dry during Bell, B. D. (1982) The amphibian fauna of New Zealand. the drought in the early 1980’s (Osborne et al. 1996). In: New Zealand Herpetology. NZ Wildlife Service Occasional Publication No. 2. (Ed: Newman, DG) N.Z. Wildlife Service, 27-89.

90 Blaustein, A.R., and Wake, D.B. (1990) Declining amphibian Environment Australia (1995).Thackway, R. and Cresswell populations: A global phenomenon. Trends Evolutionary I. D. eds. An interim biogeographic regionalisation for Ecology, 4(7): 203-204. Australia: a framework for establishing the national system Bradford, D. F. (1989) Allopathic distribution of native frogs of reserves,Version 4.0. Australian Nature Conservation and introduced fish in high Sierra Nevada lakes of Agency, Canberra. California: Implications of the negative effects of fish Fanning, F. D., and White, A.W. (1994) The Green and Golden introductions. Copeia 1989(3): 775-778. Bell Frog Litoria aurea. Management plan for the Rosebery Bronmark, I.C., and Edenham, P., (1994) Does the presence of population. Prepared in accordance with requirements of fish affect the distribution of tree frogs (Hyla arborea)? the Section 120 license for the site. Conserv. Biol. 8:841-845. Ford, M. A. (1986) Some observations of reproductive Brook, A.J., (1979a) Atlas of Frogs of Victoria. Department of behaviour in captive smooth-backed green frogs (Litoria Zoology, University of Melbourne. Publication No. 2. aurea). Tane 31:113-120. Brook, A.J., (1979b) Atlas of Frogs of Tasmania. Department of Ford, M. A. (1989):Vocalisations of the green frogs Litoria Zoology, University of Melbourne. Publication No. 3. aurea and Litoria raniformis, including species-specific male calls. N.Z. J. Zool. l. 16:7-23. Cassanova, M.T., and Brock, M. A. (1996) Wetland flooding regime and plant community establishment, Abstract. Gillespie, G.R. (1996) Distribution, habitat and conservation Wetlands for the future INTERCOL V Conference status of the Green and Golden Bell Frog Litoria aurea Abstract Book: p. 109. (Lesson 1829)(Anura: Hylidae) in Victoria. Aust. Zool. 30: 199-207. Clancy, G.P.(1996) The Green and Golden Bell Frog Litoria aurea in the Station Creek area of Yuraygir National Park. Goldingay, R.L. (1996) The Green and Golden Bell Frog Litoria Aust. Zool, 30: 214-217.Christy, M.T. (1996) The efficacy of aurea — from riches to ruins: conservation of a formerly using Passive Integrated Transponder (PIT) tags without common species. Aust. J. Zool, 30: 248-256. anaesthetic in free-living frogs. Aust. Zool, 30(2):139-142. Greer, A. E. (1994) Faunal Impact Statement for the proposed Clout, M. (1996) Introduced species:The greatest threat to development works at the Homebush Bay Brick Pit. global biodiversity? Species 34-35. Prepared for Property Services Group. Cogger, H. G. (1992) Reptiles and Amphibians of Australia. Gunninah Environmental Consultants. (1997) Assessment of 4th ed. Reed Books, Sydney. the flora and fauna issues for a proposed development at Cronulla (The North Cronulla integrated resort). Cogger, H. G. (1993) Faunal Impact Statement. Green and Golden Bell Frog (Litoria aurea) occurring on property Gunninah Environmental Consultants. (1998) Proposed sand at the corner of Dalmeny street and Kimberley Grove, overburden removal. Lot 8, DP586986 Captian Cook Rosebery. Drive Kurnell. Fauna and Flora Assessment. Colgan, D. (1996) Electrophoretic variation in the Green and Harris, K. (1995) Is there a negative relationship between Golden Bell Frog Litoria aurea. Aust. J. Zool, 30:170-176. Gambusia and tadpoles on the Northern Tablelands? BSc (Hons) Department of Ecosystem Management, Copland, S. J. (1957) Presidential address. Australian tree frogs University of New England, Armidale. of the genus Hyla. Proc. Linn. Soc. N.S.W. 82:9-108. Heatwole, H. H., de Bavay, J. M.,Webber, P.,and Webb, Courtice, G. (1972) A taxonomic revision of the Litoria aurea G. (1995) Faunal survey of New England. IV the frogs. complex (Anura: Hylidae). B.Sc. (Hons) Honours Thesis, Memoirs of the Qld. Mus, 38(1):229-249. University of Sydney, Sydney. 48 p. Hero, J-M. (1991) A froggy forecast. Wildlife Australia Autumn: Courtice, G. P.,and Grigg, G. C. (1975) A taxonomic revision 14-15. of the Litoria aurea complex (Anura: Hylidae) in south- eastern Australia. Aust. Zool, 18:149-163. Hero, J-M., Littlejohn, M., and Marantelli, G. (1991) Frogwatch Field Guide to Victorian Frogs.Victorian Department of Cree, A. (1984) Breeding biology, respiration and larval Conservation and Environment, Melbourne. development of two introduced frogs (Litoria raniformis and L. ewingi). New Zealand Journal of Zoology 11: 179-188. Heyer, R.W. (1973) Ecological interactions of frog larvae at a seasonal tropical location in Thailand. J. Herpetol. 7:337-361. Cree, A. (1988) Water balance responses of the hylid frog Litoria aurea. J. Exp. Zool, 247:119-125. Heyer, R.W. (1976) Studies in amphibian habitat partitioning. Smithsonian Contributions to Zoology, 242:1-27. Daly, G. (1995) Observations on the Green and Golden Bell Frog Litoria aurea (Anuran: Hylidae). Herpetofauna, 25:2-9. Hobbs, J.E., and Jackson, I.J. (1997) Climate. In An Atlas of New England. A.D.Lea, J.J. Pigram., and L. Greenwood (eds). Daly, G. (1996) Some problems in the management of the pp.75-93. Armidale, NSW Department of Geography. Green and Golden Bell Frog Litoria aurea (Anura: Hylidae) University of New England. at Coomonderry Swamp on the South Coast of New South Wales. Aust. J. Zool, 30: 233-236. Homebush Bay Corporation (1994) Homebush Bay Development Guidelines.Volume 1. Environmental Dankers, N. M. J. A. (1977) The ecology of an anuran Strategy. Draft 24/10/94. community. Ph.D., University of Sydney. Humphries, R. B. (1979) Dynamics of a breeding frog Ehmann, H., and White, A. (1997) Frog and Tadpole Study community. PhD dissertation, the Australian National Group of NSW. (1997).Threatened Frog Species of University, Canberra. New South Wales. Ed. By H.Ehmann. Frog and Tadpole Study Group of NSW: Sydney.

91 Interagency Advisory Group (1994) Plan of Management. Moore, J. A. (1961) Frogs of Eastern New South Wales. Bull. Revegetation & wildlife management in relation to the Amer. Mus. Nat. Hist, 121:149-386. proposed Mount Owen Mine. Prepared for Interagency Morgan, L. A. (1995) Analysis of the impact of predation by Advisory Group in relation to proposed expansion of Gambusia holbrooki on Litoria aurea tadpoles. Mount Owen Coalmine. B.Env.Sc.(Hons) Thesis. University of Wollongong: King, M. (1993) A cytotaxonomic analysis of Australian hylid Wollongong. frogs of the genus Litoria. In: Herpetology in Australia. A Morgan, L. A. and Buttemer,W. A. (1996) Predation by the Diverse Discipline. (Eds: Lunney, D., Ayers, D.) Royal non-native fish Gambusia holbrooki on small Litoria aurea Zoological Society of NSW, Sydney, 169-175. and L. dentata tadpoles. Aust. J. Zool, 30:143-149. Lemckert, F. (1996) Surveys for the green and golden bell Murphy, M. J. (1996) A capture/recapture study of the frog (Litoria aurea) by state forest of New South Wales. endangered hylid frog Litoria aurea. Herpetofauna Aust. Zool, 30:208-213. 25:19-21. Littlejohn, M.J. (1963) Frogs of the Melbourne area. Vict. Nat, Myers, G.S. (1965) Gambusia, the fish destroyer. Aust. Zool, 79:296-304. 13:102. Littlejohn, M. J, (1982) Amphibians of Victoria.Vict.Yearbook NSW National Parks and Wildlife Service (1996a) Seven Mile 85:1-11. Beach National Park and Comerong Island Nature Mahony, M. J. (1993) The status of frogs in the Watagan Reserve. Draft Plan of Management. mountains area of the Central Coast of New South NSW National Parks and Wildlife Service (1996b) Kooragang Wales. Pp. 257-64 in Herpetology in Australia: A Diverse and Hexham Swamp Nature Reserve. Discipline ed. by D. Loony & D. Ayers. Royal Zoological Draft Plan of Management. Society of New South Wales: Mosman, Australia. NSW National Parks and Wildlife Service (1998) Wildlife Atlas. Mahony, M. J. (1995) Experimental translocation of the Osborne,W.S. (1990) Declining frog populations and endangered sharp-snorted day frog Taudactylus acutirostris, extinction’s in the Canberra region. Bogong, 11: 4-7. and observations on the cause of declines among montane stream frogs.The Australian Society of Osborne,W.S. (1992) Declines and extinctions in populations Herpetologist Inc. Newsletter 37:113. of frogs in the A.C.T.: a discussion paper. Report 92/8, Australian Capital Territory Parks and Conservation Mahony, M. J. (1996) The decline of the Green and Golden Service: Canberra. Bell Frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust. Zool, Osborne,W. S., Littlejohn, M. J., and Thomson, S.A. (1996) 30:237-247. Former distribution and apparent disappearance of the Litoria aurea complex from the Southern Tablelands of Mahony, M. J. (1997) Land and Environment Court New South Wales and the Australian Capital Territory. Proceedings No. 10710 of 1997, Australian Properties-v- Aust. Zool, 30: 190-198. Sutherland Shire Council. Cronulla Integrated Resort: Fauna assessment, Species Impact for the endangered Osborne,W.S. and McElhinney, N.A. (1996) Status, habitat and green and golden bell frog (Litoria aurea). preliminary observations on calling of the Green and Golden Bell Frog Litoria aurea on Bowen Island, Jervis Bay Mahony, M. J. (1998) Land and Environment Court National Park. Aust. J. Zool, 30: 218-223. Proceedings No. 10349 of 1998, Kurnell Development Pty Ltd -v- Sutherland Shire Council. Fauna assessment, in Pechman, J. H. K., Scott, D. E., Semlitsch, R. D., Caldwell, J. P.,Vitt, particular issues dealing with the impact on the L. J. and Gibbons, J.W. (1991) Declining amphibian endangered green and golden bell frog (Litoria aurea). populations:The problem of separating human impacts from natural fluctuations. Science, 253: 892-895. Mahony, M. J., and Knowles, R. (1993) Research plan for the Yellow-spotted bell frog and Peppered tree-frog. Annual Pechmann, J. H. K, and Wilbur, H. M. (1994) Putting declining Report to Australian Nature Conservation Agency. amphibian populations into perspective: natural Unpublished Report. fluctuations and human impacts. Herpetologica, 50:64-84. Main, A. R. (1965) Frogs of Southern Western Australia. Pounds, J. A, Fogden M. P.L, Savage J. M, and Gorman G. C. (1997) Test of null models for amphibian declines on a Handbook No. 8.Western Australian Naturalists’ Club — tropical mountain. Conserv. Biol, 11:1307-22. Perth. Pickard , J. (1972) Rate of movement of transgressive sand Markwell, K.W. (1984) A biographical study of the frog fauna dunes at Cronulla, New South Wales. J. Geol. Soc. Aust, of Hexham Swamp. BA (Honours) Thesis, University of 19(2):213-216. Newcastle, Newcastle, NSW. Pickard, C. R. and Towns, D. R. (1988) Atlas of the Amphibians Maxson, L. R.,Tyler, M. J., and Maxson, R. D. (1982) and Reptiles of New Zealand. Conservation Sciences Phylogenetic relationships of Cyclorana and the Litoria Publication Number 1. Science and Research Directorate, aurea species-group (Anura: Hylidae): A molecular Dept of Conservation,Wellington. perspective. Aust. Zool, 30:643-651. Pyke, G. H. (1995) Fauna Impact Statement for proposed McDonald, K.R. and Alford, R.A. (1999) A review of declining development works at The Homebush Bay Development frogs in northern Queensland. Pp 14-22 in Declines and Area, excluding the Brickpit. Prepared for Olympic Co- Disappearances of Australian Frogs ed by A.Campbell. ordination Authority. Environment Australia: Canberra.

92 Pyke, G.H. and White, A.W. (1996) Habitat requirements for Tyler, M. J. and Davies, M. (1993) Family Hylidae. In: Fauna of the Green and Golden Bell Frog Litoria aurea (Anura: Australia. (Eds: Glasby, C. J., Ross, G. J. B., Beesley, P.L.) Hylidae). Aust. Zool, 30: 224-232. Australian Government Publishing Service, Canberra, Resource Planning Pty Ltd. (1994) Fauna Impact Statement 58-63. Proposed Extension of the Mt Owen Colliery near Tyler, M. J., Smith L. A., and Johnston, R.E. (1984) Frogs of Ravensworth, New South Wales. Prepared for Hunter Western Australia.Western Australian Museum. Valley Coal Corporation Pty Ltd. van de Mortel,T. F. and Buttemer,W. A. (1996) Are Litoria Reynolds, S.J. (1995).The impact of the introduced aurea eggs more sensitive to ultraviolet-B radiation than mosquitofish (Gambusia holbrooki) on mortality of eggs of sympatric L. peronii or L. dentata? Aust. J. Zool, premetamorphic anurans. Unpublished Honours thesis, 30: 150-157. Dept of Zoology, University of Western Australia. van de Mortel,T. F., and Goldingay, R. (1998) Population Robb, J. (1980) New Zealand Amphibians and Reptiles. assessment of the endangered Green and Golden Bell Collins, Auckland. Collins, Auckland. Frog Litoria aurea at Port Kembla, New South Wales. Aust. Robinson, M. (1993a) Raising the Green and Gold, or Smooth Zool. 30(4): 398-404. Swamp Frog Litoria aurea to maturity in captivity. Warner, S. C.,Travis, J., and Dunson,W. A. (1983) Effect of pH Herpetofauna 23:15-16. on interspecific competition between two species of Hylid Robinson, M. (1993b) A Field Guide to Frogs of Australia. tadpoles. Ecology, 74:183-194. From Port Augusta to Fraser Island including Tasmania. Watson, G. F. (1993) Diversity and distribution of Australian Australian Museum/Reed Books, Sydney. 112 pages. amphibians. In: Herpetology in Australia. A Diverse Roy, P.S., and Crawford, E. A. (1978) Holocene geological Discipline. (Eds: Lunney, D., Ayers, D) Royal Zoological evolution of the southern Botany Bay — Kurnell region, Society of NSW, Sydney, 3. Central New south Wales Coast. Engineering, Marine and Watson, G. F., and Littlejohn, M. J. (1985) Patterns of Environment Section, NSW Department of Minerals and distribution, speciation and vicariance biogeography of Energy, 100 pp. southeastern Australian amphibians. In:The Biology of Roy, P.S., and Crawford, E. A. (1979) Holocene geological Australasian Frogs and Reptiles. (Eds: Grigg,G. C; Shine, R; evolution of the southern Botany Bay — Kurnell region, Ehmann, H.) Surrey Beatty & Sons Pty Ltd, Sydney, 91-97. Central New south Wales Coast. Records of the Geological Webb, C. E. (1994) Does predation by holbrook Gambusia Survey of New South Wales 20(2):159-250. holbrooki (Atheriniformes: Poecilidae) contribute to Skinner, J. E. (1973) A barrier system, Kurnell peninsula, Botany declining frog populations? B.Sc. (Hons) Thesis. Macquarie Bay, N.S.W. B.Sc (Hons) school of Earth Sciences, University: Sydney. Macquarie University, 158 pp. Webb, C. E. and Joss, J. (1997) Does predation by the fish Thomson, S. A., Littlejohn, M. J., Robinson,W. A., and Osborne, Gambusia holbrooki (Atheriniformes: Poeciliidae) contribute W.S. (1996) Taxonomy of the Litoria aurea complex: a re- to declining frog populations? Aust. Zool, 30(3):316-324. evaluation of the Southern Tableland populations of the Wellington, R., and Wells, R. (1994) Herpetofaunal Survey of Australian Capital Territory and New South Wales. Aust. the Ravensworth State Forest area for the proposed Zool, 30: 158-169. extension of the Mount Owen Coal Project, Hebden, Tyler, M. J. (1976) Frogs. Collins, Sydney. New South Wales. Report prepared for Resource Planning Tyler, M. J., (1978) Amphibians of South Australia. South Pty Ltd. and included with FIS for proposed extension of Australian Museum, Adelaide. Mount Owen Colliery. Tyler, M. J., ed. (1979) The Status of Endangered Australasian Wells, R.W., and Wellington, C. R. (1985) A synopsis of the Wildlife. Proc. Centenary Symposium of the Royal Amphibia and Reptilia of New Zealand. Aust. J. Herpet. Zoological Society of South Australia, Adelaide, 21–23 Suppl. Series 1:62-64. Sept 1978. Published by Royal Zool. Soc. of S. Aust. White, A. (1993) Faunal Impact Statement. Proposed Tyler, M. J. (1997) The Action Plan for Australian Frogs. development of Telecom site at Roberts Road, Greenacre. Wildlife Australia, Canberra. 77 pp. White, A.W. (1995a) The Green and Golden Bell Frog (Litoria Tyler, M. J. and Davies, M. (1978) Species-groups within the aurea). Frog Facts 5:1-4. Australopapuan Hylid Frog Genus Litoria Tschudi. Aust. White, A.W. (1995b) Disappearing frogs. Aust. Zool, 30:1-9. Zool. Supplementary Series No. 63:1-47. White, A.W. (1997) M5 East Proposal. Strategy for Tyler, M. J. and Davies, M. (1978) Frogs of the Northern compensatory habitat, monitoring and long-term Territory. maintenance of the Marsh Street wetlands, Arnecliffe. Tyler, M.J. and Davies, M. (1986) Frogs of the Northern Report to the Roads and Traffic Authority. Territory. Conservation Commission of the Northern White, A.W. and Pyke, G.H. (1996) Distribution and Territory. conservation status of the Green and Golden Bell Frog Litoria aurea. Aust. Zool, 30: 177-189.

93 A Preliminary Assessment of the Status of the Green and Golden Bell Frog in north-eastern NSW

Ben Lewis and Ross Goldingay*

ABSTRACT A management plan is urgently needed for the bell frogs in Yuraygir National Park in order to address We conducted surveys for the green and golden bell disturbances to breeding sites and to establish frog (Litoria aurea) in north-eastern NSW, north of methods that assist bell frogs to breed at these sites. Coffs Harbour. Fifteen sites, including six where bell frogs occurred historically, were surveyed. Sites INTRODUCTION were visited during the day and at night between The green and golden bell frog (Litoria aurea) has declined so 1996 and 1998. Bell frogs were recorded at only two substantially in NSW during the last 20 years that it has been listed in NSW as a threatened species since 1992 (Lunney et of these sites, both within Yuraygir National Park al. 1996) and was recently (late 1997) listed by the where the total number of frogs detected was 10. Commonwealth as a (Commonwealth of No evidence of breeding was found. Australia 1997). The historic range of the species in NSW extended from the Victorian border to as far north as . It was known from 92 sites where it occurred in large We also conducted surveys for the introduced numbers (White and Pyke 1996). It has now disappeared plague minnow (Gambusia holbrooki) and found it at from about 50 sites and at only five sites are there known to be populations of more than 20 individuals (White and Pyke nine of the 15 sites. This fish was absent from the 1996; van de Mortel and Goldingay 1998). sites occupied by bell frogs. This study suggests A fundamental component of developing a conservation there may only be one population of bell frogs strategy for a species is to be able to conserve it throughout remaining in this part of northern NSW but we its geographic range (e.g. Murphy and Noon 1992). For recommend further surveys to verify this. widespread species, this is assisted by dividing up the species’

* School of Resource Science and Management, Southern Cross University Lismore NSW 2480.

94 TABLE 1: Number of surveys conducted in the study area. Survey sites (with Australian Map Grid references) are arranged from north to south. Values for surveys show the number of diurnal (D) and nocturnal (N) surveys. Bell frog present is the year of the most recent record prior to this study. BF = bell frogs detected in this study. Fish density is plague minnows per 30 litres.

Survey Sites Frog Surveys Bell Frog Fish Surveys Fish (D:N) Present (D:N) Density 1. Ocean Shores E 55220, N 684500 4:3 3:2 2.3 2. Tyagarah Nature Reserve E 55580, N 683290 6:4 1986 3:2 0.6 * 3. Byron Sewerage Works E 55630, N 683290 2:1 – P 4. Lake Ainsworth E 55760, N 681580 4:4 1985 3:2 0.2 * 5. Ballina Football Fields E 55550, N 680590 4:3 1972 3:2 12.0 6. Ballina Bicentennial Park E 55350, N 680850 5:2 3:2 2.8 * 7. Boundary Creek Road E 54670, N 679275 1:1 1:1 - 8. Evans Head Sewage Works E 54150, N 678140 3:2 2:1 1.1 * 9. Broadwater NP E 53890, N 678210 3:3 1:1 1.8 * 10. Wendoree Lagoon E 53340, N 676230 3:3 1:1 0.4 * 11. Primitive Area Lagoon E 53350, N 676080 2:1 1:1 – 12. Diggers Camp E 55280, N 670000 3:2 1993 1:1 – * 13. Station Creek E 52380, N 668640 5:4 1995 2:2 – 14. Station Creek Road E 52300, N 668630 4:3 BF 2:2 – * 15. Blue Lake E 52320, N 668480 6:4 1994, BF 2:2 – *

P = present; – = absent. * = the presence of firetail gudgeons. range into appropriate management units (Goldingay 1996). 1996;White and Pyke 1996). Another site (Ocean Shores) Conserving species in this way will preserve any genetically which we surveyed was mistakenly reported by White and distinct populations, reduces correlation of environmental Pyke (1996) as an historic location (White pers. comm.; factors and provides security against catastrophes. Tarvey pers. comm.). Bell frogs have never been detected at the dam site at Ocean Shores despite regular surveys during One of the key areas for conserving the green and golden the 1980’s (van Beurden pers. comm.). The other eight sites bell frog will be in the north-eastern portion of its range. contained suitable habitat (Pyke and White 1996) and Peripheral areas of species’ ranges may contain populations occurred within the historic range. One of these sites was that are genetically distinct from those in other portions of located approximately 1 km from Blue Lake where bell frogs the range (e.g. Lesica and Allendorf 1995).White and Pyke had been found most recently (Clancy 1996). (1996) provide evidence that just two distinct populations of green and golden bell frogs now occur in north-eastern Sites were surveyed for frogs at night and during the day, NSW, although several others were present historically. between March 1996 and March 1998. During each visit to a site at night, a hand-held spotlight was used to survey all The aim of this study was to further assess the status of the suitable habitat. A minimum of 30 minutes was spent green and golden bell frog on the far north coast of NSW. searching each site. This survey method was used extensively This will allow recovery planning for this species to begin to detect bell frogs at sites near Wollongong (van de Mortel in this region. and Goldingay 1998; Goldingay and Lewis unpublished data). All bell frogs seen and heard calling were counted; frogs were METHODS sexed based on the presence of dark nuptial pads in males >5 cm snout-vent length. A dip-net was used to search for We surveyed 15 sites (Figure1) between Red Rock (40 km bell frog tadpoles and involved continual random sweeping north of Coffs Harbour) and Ocean Shores (15 km north of whilst wading through the site. Different zones of the water Byron Bay) based on those described in the published column were surveyed to increase the surveys’ effectiveness. literature and the presence of suitable habitat. Historic records of bell frogs occurred at five of these sites (Clancy

95 FIGURE 1: Location of study area.

Survey sites are: 6. Ballina Bicentennial Park, 12. Diggers Camp 1. Ocean Shores 7. Boundary Creek Road, 13. Station Creek 2. Tyagarah Nature Reserve 8. Evans Head Sewerage Works, 14. Station Creek Road 3. Byron Sewerage Works, 9. 15. Blue Lake. 4. Lake Ainsworth, 10. Wendoree Lagoon 5. Ballina Football Fields, 11. Primitive Area Lagoon

96 The plague minnow (Gambusia holbrooki) (McDowell 1996) DISCUSSION has been implicated in the demise of the green and golden This study has provided further evidence of the grim situation bell frog in NSW (Morgan and Buttemer 1996; Pyke and for the green and golden bell frog in north-eastern NSW.We White 1996). Surveys were conducted at all but one site to were able to confirm its presence at just one of six sites in document the presence and density of this fish (Table 1). our study area where there was an older record of its Fish were surveyed using a dip-net (diameter 35 cm) to occurrence.We located one new site close to that historic scoop through approximately 30 litres of water in the water site. Our surveys were repeated several times (including day column. This was repeated three times in adjacent water and night) at most sites and spanned many months, so we columns (open water, ecotone between inshore and open believe that this is a true reflection of the bell frog’s status at water areas, and the inshore zone). Fish surveys were these sites. Many further sites must be surveyed across the conducted during the day and at night and averaged. Data study area before a conclusion can be reached that the are reported only for the inshore microhabitat at each site species’ distribution has contracted to Yuraygir NP. because other microhabitats gave an inadequate representation of fish density. We only detected a maximum of 10 bell frogs at Blue Lake in Yuraygir NP. This is a slight increase on the number recorded by RESULTS Clancy (1996). Similar surveys at sites near Wollongong have detected up to 89 adult bell frogs (Goldingay and Lewis unpubl. Despite multiple surveys across the 15 sites, we were able to data). Only adult frogs were detected at Blue Lake and there detect bell frogs at just two sites (Table 1). The sites with the were no signs of breeding, suggesting that breeding activity may older bell frog records were surveyed extensively without be suppressed. Our surveys in Wollongong have often detected success.We were able to confirm the historic presence of juvenile and subadult frogs and occasionally tadpoles. bell frogs at Lake Ainsworth, near Lennox Head. This observation was made in about 1985 by Dr E. van Beurden We were unable to detect bell frogs in the vicinity of Diggers who had continued his extensive surveys for cane toads in Camp. However, single bell frogs were found here in the area (see van Beurden and Grigg 1980). November 1993 at two sites about 2 km apart (Hines pers. comm.) and two bell frogs were found in October 1997 The two sites where bell frogs were detected are in Yuraygir (Hines pers. comm.) in the swamp we searched. None were National Park (NP) and about 1 km from each other. At Blue present there in September 1998. These observations Lake, two bell frogs were seen in September 1996 and five suggest a very small number of bell frogs has held on at this were seen at this site during warm conditions in August site which is located 15 km north of Blue Lake. 1997. These initial surveys focussed on the southern end of Blue Lake because of the ease of access and small area (ca. There was a high level of disturbance at most sites except 0.3 ha) compared to the whole lake (ca. 5 ha). In December those in Bundjalung NP (Wendoree and Primitive Area 1997, a canoe was used to survey the entire perimeter of Lagoons), although these lagoons are the legacy of sand- Blue Lake at night. This included regular stops to mimic the mining in the past. The wetlands in southern Yuraygir NP call of the frog and to illuminate the extensive beds of were subject to disturbance from wild horses which trampled cumbungi. Nine male bell frogs were detected. Two bell frogs the edges of the wetlands and fed on the aquatic plants. Such (at least one was a female) were detected during a similar disturbance has been present for many years (Clancy 1996). census in March 1998. Thus, we have direct evidence of 10 individuals at this site.We detected bell frogs during all visits The presence of exotic predators such as plague minnows, to Blue Lake. cane toads and foxes may have influenced the distribution of the bell frog. There is an apparent negative association At the Station Creek Road dam, no frogs were detected in between the plague minnow and the bell frog in our study 1996 but two males were detected by call in August 1997. area which is consistent with the findings of Pyke and White This new location is approximately 1 km from Blue Lake. (1996). However, this fish was absent from two sites without Diggers Camp was also surveyed during the 1997-98 season frogs. Bell frogs occur at one site near Wollongong where but no frogs were detected despite surveys being undertaken plague minnows are abundant but coexistence may be in both wet and dry periods. facilitated by submerged vegetation (van de Mortel and Goldingay 1998). Thus, it is difficult to fully implicate this We were unable to detect any evidence of breeding (i.e. exotic fish in the decline of the bell frog in northern NSW amplecting frogs, tadpoles, juveniles) by bell frogs at either of but it may have contributed to the decline. For example, the the two occupied sites. Searches for bell frog tadpoles were Ballina Football Fields were constructed on a large freshwater also conducted at the other sites but none were found. wetland and the remaining ponds now have high densities of The plague minnow was found at 9 of the 15 sites (Table 1). plague minnows. Management of sites with bell frogs should The densities observed were generally low except for one always endeavour to exclude this fish. historic site (Ballina Football Field) which had the greatest density of this exotic fish of any site. The number of sites with and without this fish and frogs can be assessed for an association by the Fisher exact test (Zar 1984). This revealed that bell frogs had a significant association (P=0.01) with sites without gambusia. A native fish, the firetail gudgeon, was present at 9 of the 15 sites, including those where bell frogs were present.

97 It is not known whether the cane toad has had an adverse REFERENCES effect on the bell frog but toads occur at high densities at Clancy, G.P.,(1996) The green and golden bell frog in the Lake Ainsworth where a small number (ca. 2) of bell frogs Station Creek area of Yuraygir National Park. Aust. Zool, 30: was recorded over several years up until about 1985 (van 214-17. Beurden pers. comm.). Toads have been present in this area since about 1968 (van Beurden and Grigg 1980). At present, Colgan, D., (1996) Electrophoretic variation in the Green and toads do not occur south of Evans Head in the study area. Golden Bell Frog Litoria aurea. Aust. Zool, 30: 170-76. Commonwealth of Australia, (1997) Declaration under The sites where bell frogs were detected in Yuraygir NP s18.(1) of the Endangered Species Protection Act 1992. require urgent management. Feral horses which have affected Goldingay, R.L., (1996) The Green and Golden Bell Frog the aquatic vegetation and trampled the edges of the water (Litoria aurea) — from riches to ruins: conservation of a bodies must be removed from the National Park. The formerly common species. Aust. Zool, 30: 248-56. microhabitats present on pond edges are required by bell frogs for shelter, foraging and breeding (van de Mortel and Lesica, P.and Allendorf, F.W., (1995).When are peripheral Goldingay 1998). populations valuable for conservation. Cons. Biol, 9: 753-60. There is no direct evidence of any breeding activity by bell Lunney, D., Curtin, A., Cogger, H. G., and Dickman, C. R., frogs in Yuraygir NP over the last 5 years (see also Clancy (1996) An ecological approach to identifying the 1996). Therefore, an attempt should be made to assist frogs endangered fauna of New South Wales. Pac. Cons. Biol,2: to breed. This could include either collecting of amplecting 212-31. pairs in order to collect spawn masses (see van de Mortel McDowell, R., (1996) Freshwater Fishes of South-eastern and Buttemer 1996) or providing artificial breeding ponds. Australia. Reed Books, Chatswood. The latter has been used as a technique for sampling the tadpoles of hylid frogs (e.g. Zimmerman and Simberloff 1996) Morgan, L.A. and Buttemer,W.A., (1996) Predation by the and should be trialed as a way of collecting bell frog spawn. In non-native fish Gambusia holbrooki on small Litoria aurea either case, tadpoles would be reared to a large size or to and L. dentata tadpoles. Aust. Zool, 30: 143-49. metamorphling stage before release. Small breeding ponds Murphy, D.D. and Noon, B.R., (1992) Integrating scientific may need to be constructed as a more permanent solution methods with habitat conservation planning: reserve to aid breeding at this site. The discovery of two frogs in a design for northern spotted . Ecol. Appl, 2: 3-17. dam approximately 1 km from Blue Lake suggests that bell Pyke, G.H. and White, A.W., (1996) Habitat requirements for frogs will readily colonise artificial breeding sites in this region the Green and golden Bell Frog Litoria aurea (Anura: as has been documented elsewhere (Pyke and White 1996). Hylidae). Aust. Zool, 30: 224-32. Further surveys are required throughout the study area over van Beurden, E.K. and Grigg, G.C., (1980) An isolated and the next two years to confirm the results of this study and to expanding population of the introduced toad Bufo marinus provide continued monitoring of the size of the bell frog in New South Wales. Aust.Wildl. Res, 7: 305-10. population in Yuraygir NP. The current absence of breeding van de Mortel,T.F. and Buttemer,W.A., (1996) Are Litoria may reflect several years of poor summer rain.Whatever the aurea eggs more sensitive to Ultraviolet-B radiation than case, the small size of this population shows that it will be very eggs of sympatric L. peroni or L. dentata? Aust. Zool, vulnerable to local extinction and a broader strategy should 30: 150-7. be considered to increase the number of occupied sites van de Mortel,T. and Goldingay, R., (1998) Population (Goldingay 1996) in order to conserve this most northern assessment of the endangered green and golden bell frog population. The population in Yuraygir NP now apparently Litoria aurea at Port Kembla, New South Wales. Aust. Zool, represents all that remains of the northern population of the 31: 48-54. green and golden bell frog. There is a major discontinuity in White, A.W. and Pyke, G.H., (1996) Distribution and range between this area and the nearest southern populations conservation status of the Green and Golden Bell Frog (see White and Pyke 1996). This population is therefore likely (Litoria aurea) in New South Wales. Aust. Zool, to be genetically distinct (e.g. Lesica and Allendorf 1995; 30: 177-89. Colgan 1996) from other bell frog populations and must be given a high priority for conservation. Zar, J.H., (1984) Biostatistical Analysis. Prentice-Hall, London. Zimmerman, B.L. and Simberloff, D., (1996) An historical ACKNOWLEDGMENTS interpretation of habitat use by frogs in a central Amazonian forest. J. Biogeog, 23: 27-46. We thank Dr Eric van Beurden and Harry Hines for sharing their unpublished bell frog observations with us. David Rohweder provided his canoe and assistance during some of the Blue Lake surveys.We thank two referees for constructive comments on a draft of this paper.

98 Loss and Degradation of Red-Crowned Toadlet habitat in the Sydney Region

Karen Thumm and Michael Mahony *

ABSTRACT six extant sites from across the Basin were studied and were generally within the top 40% of the slope The entire distribution of the red-crowned toadlet below the ridge-top. Breeding sites were in (Pseudophryne australis), a frog listed as vulnerable ephemeral situations.We postulate that the position under the New South Wales Threatened Species of the breeding sites is related to the geology and Conservation Act 1995, is within the geological region geomorphology associated with eroded Hawkesbury referred to as the Sydney Basin. Mapping of site Sandstone strata. Recognition of this relationship records revealed a relationship with the Hawkesbury resulted in a considerable reduction of the area Sandstone strata and the boundaries between this generally assumed to be suitable for this species. strata and the Wianamatta Shale above and Sydney, the largest metropolitan city in Australia, Narrabeen Group below. The Hawkesbury with a population of nearly 4 million people is Sandstones are exposed in about 27% of the Basin. centred within this area. Growth of the urban area is occurring largely on exposed Hawkesbury Sandstone Very few populations were found to occur within the strata leading to the incremental destruction and Narrabeen Group or Wianamatta Shale which are degradation of the habitat of the red-crowned exposed in about 42% and 10% of the Basin toadlet.When combined with a life history strategy respectively. Only three records were found on the adapted to an unpredictable ephemeral Quaternary Sands and Volcanics and one in the Coal environment it is evident that this species is Measures (less than 11% and 10% respectively). Fifty- particularly sensitive to urban pressures.

* Department of Biological Sciences,The University of Newcastle Callaghan NSW 2308.

99 INTRODUCTION preferences of the species at a regional scale is also presented to assist in gauging the impact of future The red-crowned toadlet, Pseudophryne australis, is a small development on this species. ground-dwelling frog (snout vent length 22–28 mm) with an orange v-shaped marking on its head and bold black and white ventral patterning (Cogger 1992). Reproduction and METHODS larval development are restricted to non-perennial Distribution watercourses that generally form below the ridge line. The egg mass is usually terrestrial but within the bed of a To determine the historic and current distribution of the red- non-perennial watercourse. Typically the eggs are placed in a crowned toadlet, location records of Pseudophryne australis moist, but not flooded situation, under some form of debris were collated from museums, from the New South Wales such as thick leaf litter or under rock. The clump of eggs has National Parks and Wildlife Service (NPWS) Atlas of NSW often been referred to as a nest (Harrison 1922;Woodruff Wildlife and from literature. Following collation of all records, 1976; Barker and Grigg 1977). Following rain, water flowing multiple records were removed. The geological strata within down the non-perennial watercourses passes through the the Sydney Basin were mapped and the areas covered by egg mass and releases the embryos from the egg capsules, each were calculated (1:500,000 Geological Sheet, NSW sweeping them into small ponds within this watercourse, Dept. of Mines 1969). The red-crowned toadlet records where the tadpole stage and metamorphosis occur (Parker were superimposed onto this map in order to calculate the 1940; Barker and Grigg 1977). These ponds are ephemeral relative number of records found in each geological and high embryonic and larval mortality due to desiccation formation. The extent of dense urban development was leads to a low recruitment rate (Thumm unpubl. data). estimated using a street directory (Sydway 1996).

Woodruff (1976, 1978) considered that the habitat Habitat requirements and reproductive biology of the red-crowned toadlet differed markedly from the other south-eastern To determine the habitats occupied by the red-crowned species of Pseudophryne (P. bibronii, P. dendyi, P. semimarmorata) toadlet at the landscape level the topographical features of which he reported to be seasonal breeders. He noted that in 56 extant red-crowned toadlet breeding sites including slope, these species the sites selected for breeding flooded and altitude (to the nearest 5 metres), distance below the ridge- retained water for extended periods. He also reported that top measured as a vertical distance, and position in relation this distinctive seasonal breeding pattern coincided with the to cliffs were recorded. The breeding area was defined by regular autumnal rainfall. surveying upslope and downslope from the areas where calls were first heard until no more calling males were located. Apart from the general descriptions of breeding sites provided by Harrison (1922) and Woodruff (1978) there is Rainfall limited data on the habitat features of the breeding sites of Daily rainfall statistics were obtained for the period between P. australis. Most descriptions of the distribution of the 1992 and 1996 for the Turramurra weather station (Bureau red-crowned toadlet simply note that it is restricted to the of Meteorology). This was the closest station with a Sydney Basin of New South Wales although some complete rainfall data set to the site where reproductive descriptions have considered underlying geological substrates biology was studied (approx. 10 km distance). (Woodruff 1976, Barker and Grigg 1977, Cogger 1992, Thumm and Mahony 1997). The Basin includes Sydney, the largest urban area in Australia with a population of almost Breeding Biology 4 million (Department of Urban affairs and Planning 1995). In order to assess the broad features of the reproductive It covers an area of about 26 478 km2 and the area of biology of the red-crowned toadlet a breeding site was studied dense urban development within this Basin is about over a six year period. The site chosen consisted of an 4 000 km2 (Sydway 1996) or 15%. This figure does not ephemeral watercourse that contained several breeding sites include large towns in coastal areas north and south of and ponds. Data was collected on the hydroperiod of breeding Sydney. A considerable portion of the urban areas to the sites, oviposition frequency and recruitment. The site in west of the central business district occurs on the Hornsby Heights (Australian Metric Grid (AMG) 56H 324000 Cumberland Plain which is primarily on Wianamatta shale 6274500), was in the core area of the distribution of this substrates, but to the north, south and west of this plain species (Figure 1). It contained two depressions, which filled urban development occurs mainly on Hawkesbury Sandstone after heavy rain to form ephemeral ponds. The lower pond substrates. It is in these areas where urban development has had three 0.9 m sides and was 0.13 m deep and the upper the potential to destroy the habitat of the toadlet and it is in pond was 0.3 m x 0.2 m x 0.05 m. There was also one semi- these areas where current and future urban development is permanent pond of approximately 0.30 m in diameter and concentrated. Continued destruction and degradation of this 0.12 m depth, which was located on a rock shelf. To assess habitat is inevitable as Sydney grows to accommodate an whether the hydroperiod of this site was representative of increasing population. red-crowned toadlet breeding sites, sixteen other ephemeral watercourses known to have populations of red-crowned This paper gives an overview of the characteristics of the toadlets were visited, and a total of 31 ponds measured. red-crowned toadlet’s habitat, its distribution and its breeding biology.We argue that the combination of these features The area used for the long-term study was visited 239 times make this species particularly sensitive to changes related to between September 1992 and August 1995 in order to urban development. An examination of the habitat collect data on oviposition frequency and recruitment.

100 FIGURE 1: Map of the distribution of Pseudophryne australis (red-crowned toadlet) in the Sydney Basin in relation to geological strata, based on records from Australian natural history museums, the National Parks and Wildlife Service Atlas of NSW Wildlife, and literature (6/12/95). This figure indicates that this species is most frequently found on Hawkesbury Sandstone. Records are indicated by a circle.

101 Monitoring was carried out at an average of 8.8 visits per Three specimens and one Atlas record (AM R9407 Jaffa month in the first year, 5 visits per month in the second year, Texas Ashord Downs, AM R70161 Lightning Ridge, Smith and 5.4 per month in the third year averaging out at 6.4 visits Lakes AM R78699 and the (NPWS) Border Ranges) were per month over the three years. Individual females were found to be misidentifications. identified by reference to their unique ventral pattern. The reliability of this method was tested using specimens in the One specimen from Point Lookout in the New England Australian Museum, by drawing the ventral markings of 20 region (AM R45704) had external features consistent with adult frogs, recording their registration number separately, and identification as a specimen of P. australis. The specimen was then asking Museum staff to use the drawing to identify faded but a distinct head mark typical of P. australis was individuals. To test the assumption that pattern does not apparent. This specimen was transferred to the Australian change with time, a series of 12 captive held individuals Museum from the University of New England Department of were compared over a period of up to nearly three years. Zoology collection. Comparison of the registers revealed that The frequency of visits to the main study site made it errors had occurred in transcription of the collectors’ identity possible to record any new clutches within a few days of and date. Further, there were three specimens listed in the being laid, to follow through the persistence of ephemeral University of New England register but only one specimen ponds and to record successful metamorphosis. was in the Australian Museum.When the collectors (de Bavay and Frazier) were contacted, they did not recall collecting the To prevent disturbance affecting the reproductive recruitment specimen in that area. There were similar problems with a at this study site, counts of egg mass size (numbers of specimen collected in “Armidale” (AM R45705). Firstly the eggs/clutch) were made at another nearby site in Berowra data was incorrectly transposed to the AM register, and Heights (AMG 56H 328400 6278650), approximately secondly, the collector (Frazier) did not recall this specimen. 4 km distant. A third specimen (AM R 45703) from the region “20 mls W of Armidale” cannot be verified as an accurate record Threats to Red-Crowned Toadlets from because none of the data in Pengilley’s (the collector listed in Urbanisation the NE and AM register) logbook is similar to the Museum register’s records. It is doubtful whether it will ever be To assess the impact of urban development fifty-six sites possible to discount these three records from the NE of were visited over 4 years (Figure 2). Disturbance indicators NSW completely, because it is not known who and where resulting from urbanisation were routinely recorded: siltation, the specimens were collected. However, considering the large soil pH, weed presence and species composition, presence number of inaccuracies relating to these specimens, it appears of stormwater outlets, quantities of remaining bushrock as as if these records should be given little weight. a percentage of the surface area, and fire hazard reduction activities. Three other specimens from outside the Sydney Basin could not be found in museums (Bowning AM R337, Gapstead RESULTS Railway Station MOV D6889 and Oberon AM R12269). One specimen from Tom Groggin (AM R12982) has been sent to Distribution the Philippines National Museum. It is assumed that this specimen was a P. dendyi, because in some parts of their A total of 431 records were obtained from museums (292), distribution, this species has a similar pattern to P. australis the NPWS Atlas of NSW Wildlife (130) and from the except that the crown is yellow. Specimens from Kiandra literature (9). Based on these records, there were 141 (R03350 — A/B, R03351) were sighted by Mark Hutchinson localities for which P. australis was recorded (Figure 1). and “are certainly not australis” (correspondence from The limits of the distribution were defined as between Hutchinson, curator of herpetology in the South Australian Mt.Victoria in the west to Pokolbin in the north and near Museum). There were two coding or transcription errors in the Barren Grounds in the south (Figure 1). NPWS Wildlife Atlas which were ascertained by the authors. The majority of records fell within the Sydney Basin (417 of There were historical records for red-crowned toadlets from the 431 records). Details relating to the fourteen records that areas of Sydney that are now most densely populated, e.g. fell outside this region were closely scrutinised. Those held by Chatswood. A few populations remained in natural bushland museums were inspected, the accession register checked and remnants within urban areas, e.g. Bradley’s Head, Manly Dam in a few cases the collector of the specimen was contacted. Reserve, Lindfield.

TABLE 1: Number of red-crowned toadlet locations within each of the geological formations of the Sydney Basin.

Geological Formation km2 % of total No. of localities % of total Sydney Basin 26 478 100% 141 100% Quaternary Sands and Volcanics 2 863 10.8% 2 1.4% Wianamatta Shale 2 843 10.7% 2 1.4% Hawkesbury Sandstone 7 045 26.6% 125 88.7% Narrabeen Group 11 144 42.1% 11 7.8% Coal Measures 2 583 9.8% 1 0.7%

102 Association with Geological Strata Breeding Biology

The relationship between distribution and geological strata is Low Recruitment Rate presented in Figure 1. Eighty-nine percent of locations were associated with the Hawkesbury Sandstone which is exposed At the long-term study site at Hornsby Heights, during the in 26.6% of the total area of the Sydney Basin (Table 1). period from Sept 1992–August 1995, there was water in the There were 7.8% of records within the areas dominated by ephemeral ponds on 18 occasions. Of 57 egg masses the Narrabeen Group of sandstones which comprise 42.1% of recorded from the site, 11 tadpoles metamorphosed the Basin (11 144km2). Only two records were located within successfully. A large proportion of the terrestrial egg masses the Wianamatta Shale formation of the Cumberland Plain, and dried up before the tadpoles hatched and were swept into both were based on relatively old (1889, 1923) museum ponds (Thumm unpubl. data). On two occasions the flow of specimens. It can be assumed that the locality data supplied water through the water course from heavy rain was so with these specimens only mentioned the closest settlement. strong that all tadpoles were swept out of the ponds to be One of the two records assigned to Quaternary Sands is an stranded in the leaf litter and die. older specimen (1894) and one lacks collection data. The one Clutch sizes averaged 22 eggs/clutch (average taken from record assigned to coal measures is on the scree slope below 55 clutches); hence the recruitment rate was less than the escarpment on the western rim of the Basin. Four localities 1% averaged across all clutches at this site. could not be assigned to a geological formation due to the vagueness of the locality data (e.g.“Sydney”). DISCUSSION There were records from the rim of the Basin, from Distribution Mt Victoria in the west, from near Barren Grounds in the south and from Pokolbin in the north (Figure 1). There were more Red-crowned toadlets showed a strong association with records in the north of the distribution area than in the south. Hawkesbury Sandstone substrates which covers about 27% Only 12% of all records were found in the southern half of the (7 045 km2) of the Sydney Basin (Table 1). The general distribution area (south of Botany Bay, 34°Lat S) although this perception (Barker and Grigg 1997) that red-crowned area makes up approx. 40% of the entire area of the Basin. toadlets are found on “Sydney sandstone” (including the Narrabeen Group) is misleading. Records on the Narrabeen Habitat Characteristics Group (including sandstones) made up only 7.8% of all records, although the Narrabeen Group comprises Geomorphological Relationship approximately 42% of the area of the Sydney Basin. Only two A diagrammatic representation of the relationship between red-crowned toadlets were found in the Wianamatta Shale red-crowned toadlet breeding sites and the geomorphology area of the Cumberland Plain (10.7% of the Sydney Basin). based on the examination of 56 extant sites is presented in Hawkesbury Sandstone, the geological formation on which Figure 2. Breeding sites were generally located just below the Red-crowned toadlets were most frequently found, is closer first escarpment (60% of sites studied) or in areas where in to the Central Business District than the Narrabeen Group there were large rock outcrops (80%) (Figure 3). All except of sandstones (Figure 1). For this reason it is more susceptible two were located within the first 100 m altitude from the to development. ridge-top (Figure 4). Sites were generally (87% of 56 sites) within the top 40% of the slope (Figure 5), and only four sites The observed difference in the abundance of sites between occurred on plateaus. Most breeding sites (67% of 56 sites) the north and the south of the Basin has not been previously were associated with a non-perennial natural drainage lines. recognised. The lower density of records in the south of No breeding sites were situated within swamps that Sydney may be attributable to a number of factors. This occasionally form on the plateaus in the Sydney Basin area region contains a number of water catchment areas that are (referred to as hanging swamps by Keith and Benson 1988 closed to access and accordingly there have been fewer and Keith and Myerscough 1993), although red-crowned collections made. Another possibility is that the topography toadlets were found in the small drainage lines that fed into of the Hawkesbury Sandstone in the south of Sydney is less or out of hanging swamps (e.g. Mt.Victoria). Ephemeral ponds rugged and perhaps provides less suitable habitat for the frog. had a mean size of 0.85 x 0.51 x 0.07 m. Lengths ranged This area also has a larger percentage of Wianamatta Shale from 0.22–3.50 m, widths from 0.11 m–2.0 m and depths outcropping than north of 34°S Lat. To determine whether from 0.05–0.25 m. the difference observed is real requires an unbiased sampling strategy. A similar approach would also be needed to Rainfall understand the link between geomorphology and habitats The rainfall pattern for Turramurra for three years is between these areas. Until this matter is resolved presented in Figure 6. Two broad features were apparent. consideration of the impact of habitat alteration in the south Firstly, there was no distinctive seasonal pattern, and secondly, of Sydney should take into account the lower density of total rainfall varies considerably between years with heavy localities in this part of the distribution area. rainfall or drought occurring in any month of the year. There are several sources of potential error in using historical records, databases and literature for mapping distribution of this species. Firstly, because these sources rely on records that were accumulated over time and were not the result of systematic surveys, the results are likely to be biased. Some

103 FIGURE 2: Map of the fifty-six extant red-crowned toadlet sites at which geomorphology was assessed (indicated as a triangle). Personal knowledge of these locations enabled more precise maping than was possible for Fig. 1, which included all available records, including historical records.This figure supports the conclusion that red-crowned toadlets are predominantly found on Hawkesbury Sandstone strata.

104 FIGURE 3: Schematic cross section of the Hawkesbury generalised locality data were presented with many museum Sandstone soil landscape illustrating that red-crowned toadlet specimens, making many records unsuitable for the purpose breeding sites are found below the first escarpment on the talus of precise mapping. Thirdly the scale of the geological maps slope. (Redrawn from Chapman, G.A. and Murphy, C.L. (1989) Soil Landscapes of the Sydney 1:100 000 Sheet. Soil leads to a level of inaccuracy when considering the intricate Conservation Service of N.S.W., Sydney). interweave of the different geological boundaries, leading to uncertainty as to the geological formation to which a record could be assigned. In spite of these constraints, the results add considerably to our knowledge of the distribution of P. australis which is vital when considering the management requirements of the species in the Sydney Basin.

Habitat There is a considerable amount of evidence that shows the red- crowned toadlet is a habitat specialist, relying on a combination of substrate and landform. All 56 extant breeding sites studied were ephemeral, with no breeding taking place in perennial creeks. Ephemeral ponds were small and were often found in depressions in rock shelves, worn by water pouring down the slope. Other ponds were below drops in the watercourse, with accumulations of leaf litter banked around the edges. No fish were found in these ponds, and there were very few larval Odonates or other macroinvertebrates, which are known to prey on tadpoles (Duellman and Trueb 1986; Richards and Bull 1990). Breeding areas were also on slopes, with the eggs being laid above the ponds, in the path of intermittent water flows. regions of the Basin have been subject to relatively close This pattern differed from the situation described by Woodruff scrutiny and collection while others have been poorly (1976) for the three species of Pseudophryne in south eastern studied. This was apparent in the number of records for the Australia where eggs are laid in a position where a rising frog from the north shore region of Sydney (area bounded groundwater table will inundate the eggs. Furthermore, the by 33°40’ — 33°50’ Lat, 150° 05’ — 150°15’ Long ) an area ponds in which the tadpoles develop were short-lived when which has a long history of urban development. The compared with these inundated areas used by other species distribution data show clearly that the species once occurred (Martin 1967;Woodruff 1976).We postulate that the abundantly in this region. By contrast there were few records reproductive strategy of red-crowned toadlets is adapted to take from the north west region of the Sydney Basin. This area advantage of ephemeral waterbodies that support few consists of deeply dissected plateaus that have few roads and predators. Perennial situations are not exploited in spite of a low access points. Another possible source of error was that only level of recruitment in ephemeral ponds.

FIGURE 4: Distance (in vertical metres) from the ridge-top of Red-Crowned Toadlet breeding sites, indicating a strong association of the breeeding sites with the area just below the ridge.

105 FIGURE 5: The position on the slope between the ridge and the valley of 56 red-crowned toadlet sites showing that most sites are found in the upper 40% of the slope.

Climate Threats to Red-Crowned Toadlets For a frog that relies on ephemeral water bodies for from Urbanisation breeding, rainfall patterns and quantities would be predicted The impacts of urban development include direct destruction to be of importance. Rainfall in Sydney is unpredictable, not of habitat and the adverse effects of housing and seasonal, and often falls as heavy spring/summer infrastructure which extend into areas beyond the urban thunderstorms or East Coast (winter) Cyclones (Bureau of fringe. Housing in Sydney’s suburbs frequently follows ridges. Meteorology 1991), with heavy falls or occurring in Development is in the headwater of catchments and is any month of the year. The Sydney Climatic Survey states accompanied by degradation of the bushland surrounding the that “an examination of the records reveals that the totals houses e.g. pollution, an increase in the amount of water from month to month and from year to year are quite reaching the soil, soil nutrients and weed propagules variable and the rain tends to fall in concentrated bursts”. (Clements 1983). There is “an average of 6 storms each year” (Bureau of Meteorology 1991). It appears that the opportunistic About 15% of the original distribution area available to red- reproductive strategy of the red-crowned toadlet in all crowned toadlets has been developed, and housing is seasons of the year has been shaped by the unpredictable continuing to expand, especially in the “outer ring” which rainfall in the Sydney Basin and the features of the habitats it includes many local government areas within predominantly occupies. Due to the lack of seasonality it is difficult to Hawkesbury Sandstone areas. The Department of Urban establish guidelines for habitat management. Affairs and Planning (DUAP) expects an increase in population to Sydney of 804 000 to 4.5 million by the year Breeding Biology 2021 (over a 30 year period), with 90% of new houses in the outer ring, of which 34% will be “in-fill” in established areas A combination of the ephemeral breeding habitat and the and 56% in “release areas” (DUAP Population Predictions uncertain nature of rainfall in the region were observed to be 1995). A recent ESD Study (Ecological Surveys and Planning responsible for the low levels of recruitment in this species. A 1998) pertaining to the proposed development of combination of observations made at the long-term study site “Landcom” housing in the Hornsby Shire identified 18 areas and the 56 sites examined for habitat features indicated that with a total of approximately 563 lots. All sites are on disturbance to breeding sites or their hydrology is likely to Hawkesbury Sandstone , and “most...are ridge-tops and spurs significantly affect recruitment and hinder the recovery or the upper slopes and benches” on the “fringes” of the potential of populations of this frog by disturbing the finely Hornsby plateau. (Ecological Surveys and Planning 1998). balanced breeding strategy which has evolved. Frequent Inevitably the need for increased housing will put pressure on disturbance associated with urbanisation may jeopardise red-crowned toadlet habitat as the edges of the city expand. reproductive success leading to the extirpation of populations. Red-crowned toadlets typically breed within the first 100 m below the ridge-top, in the talus slope below the first escarpment which forms the edges of the ridges. Historically these areas were not developed for housing because of the

106 FIGURE 6: Rainfall from the Turramurra Weather Station over a three year period plotted with the 123 year average rainfall from Sydney. This figure shows that there was no distinctive seasonal pattern, and secondly, that total rainfall varies considerably between years with heavy rainfall or drought occuring in any month of the year.

steep nature of the sites. However, new building technology is Housing developments also alter the quantities of water leading to a development of the talus slope below the first flowing through the ephemeral water courses used for steep escarpment. In established suburbs, land on the fringes breeding.When undisturbed these “upper laterals” (Harrison of the ridges which have previously been excluded from 1922) only flow after heavy rainfall, after which they are development due to the topography, are now being “in-filled”. observed to dry up, leaving a few ephemeral ponds. Housing developments typically seal the catchment and concentrate the Siltation is generally an “edge effect” of new housing flows into a few watercourses. Clements (1983) reported an developments, leading to an alteration of the characteristics increase in water entering the bushland in suburban areas of of the watercourse and to weed invasion (Buchanan 1989). 30–50%. It is suggested here that the change in flows created Siltation entering a nesting area has been seen to displace by the redirection of stormwater may affect levels of red-crowned toadlets from a preferred nest site (Thumm, recruitment. Even slight alterations to the hydrology of a unpubl. data). One site visited on eight occasions over a breeding site due to sealing of the catchment or diversion of period of two years, and where red-crowned toadlets were flows within a development, are likely to alter the balance of frequently observed to be present and to lay eggs under a conditions within which a terrestrial nest site will be successful. particular rock, was no longer used for breeding, after large quantities of silt entered the site and surrounded the rock. It There is the potential for conflict between the requirement has taken 4 years for this site to be used again, but there is to protect residents in Sydney from bushfire and the need to still no breeding under that particular rock (Thumm and conserve fire-sensitive flora and fauna. Fire hazard reduction Mahony unpubl. obs.). activities are generally carried out in the areas just below the escarpment, in the preferred breeding habitat of the red- It is generally accepted that many household chemicals, e.g. crowned toadlet. Frequent disturbance and degradation of insecticides, herbicides and fungicides, are deleterious to frogs the habitat of this species, clearing and fire hazard burning is (Tyler 1989). The increased nutrient load in the soils near likely to reduce the size of populations, due to the low suburbs (Clements 1983, Buchanan 1989) caused by recovery potential of this species. A colony with 15 calling chemicals such as fertilisers and detergents, results in weed males monitored prior to a wild fire in 1994 was revisited invasion down the water courses coming off the ridge-tops. after the fire. Only one male was calling on a day on which Stormwater from urban areas is typically directed into the there was a lot of activity at a nearby control study site. No non-perennial water courses used by red-crowned toadlets. leaf litter remained at all in the breeding area at the site of We have not observed this species breeding in polluted the fire. It was evident that there would be no nest sites conditions.Watercourses affected by stormwater adjacent to available, nowhere to refuge and no foraging areas. No calling pristine watercourses supporting red-crowned toadlet was heard at a second site three years after a fire in 1994. populations are not used by this species. It therefore appears likely that the decline in water quality associated with Osborne (1991) stated that fire was a “potential threat” to stormwater leads to the loss of this species. the corroboree frog, (P. corroboree), a congenor of the red-crowned toadlet, and suggested that fire may make them more “vulnerable to dehydration”. It has been observed that

107 red-crowned toadlets retreat to lower clay layers or into the Cogger, H. G., (1992) Reptiles and Amphibians of Australia. crevices of cliffs in dry times, but that they are found just 5th Edition. Reed Books, Sydney. below the leaf litter in wetter periods. It is not known which Cogger, H.G., Cameron, E.E., Sadlier R.A. and Eggler P., (1993) weather conditions should be chosen for fire hazard reduction Action Plan for Australian Reptiles. Australian Nature burns in order to create the type of fire which would have the Conservation Agency, Canberra. Endangered Species least effect on population numbers. This species does not Program Project No. 124, 156. have an “off-season” (hibernation or aestivation) making Department of Urban Affairs and Planning, (1995). Population forward planning difficult. Due to their low recovery rate, it is Predictions. New South Wales Government. considered that a minimum time span between prescriptions of all fire management activities within an individual red- Duellman W. E. and Trueb L., (1986) Biology of Amphibians. crowned toadlet site of about 10 years including burns, fire McGraw-Hill, New York. trail maintenance, turbo-mowing and clearing is kept in order Ecological Surveys and Planning (1998) Landcom ESD Study to avoid the cumulative impacts of each activity. 1. Ed. S. Douglas. Unpublished report for the Total Environment Centre Inc. Sydney. Another more subtle change associated with urban development Harrison, L.,(1922) On the breeding habits of some are the impacts on nearby adjacent natural habitats. Impacts of Australian frogs. Australian Zoologist, 3: 17-34. activities such as the removal of bush rock for gardens and the invasion of exotic plants are difficult to assess. Because red- Keith D.A. and Benson D.H., (1988) The natural vegetation of crowned toadlets lay eggs beneath rocks and seek refuge the Katoomba 1:100 000 map sheet Cunninghamia beneath them, reduction in the quantities of bushrock causes 2 (1): 107-144. degradation of their habitat. Typically bushrock is removed from Keith D. A. and Myerscough P.T., (1993) Australian Journal of ridge roads and tracks (Schlesinger and Shine 1994), as truck Ecology 18: 325-340. access is required to remove the rock. Areas adjacent to ridges Mahony, S., (1997) Efficacy of the “Threatening Processes’ are therefore primarily affected. Bushrock removal has been listed Provisions in the Threatened Species Conservation Act as a Key Threatening Process under the NSW Threatened Species 1995 (NSW): Bushrock removal and the endangered Conservation Act 1995, and red-crowned toadlets are listed as one broad-headed snake. Environmental and Planning Law of the species impacted by this process. Journal, 14 (1): 3-15. Martin, A. A., (1967) Australian Anuran Life Histories: ACKNOWLEDGMENTS Some Evolutionary and Ecological Aspects. Pp. 175-191 Thank you to the following museums and organisations for in Australian Inland Waters and the their Fauna. A.H. locality data: Australian Museum, South Australian Museum, Weatherley ANU Press. Canberra. Museum of Victoria, Museums and Art Galleries of the NSW Dept. of Mines, (1969) Sydney Basin 1:500 000 Northern Territory, Queensland Museum,Western Australian Geological Sheet. Museum, CSIRO and NPWS. Thanks also to all naturalists Osborne,W. S., (1991) The biology and management of the who generously provided us with locality data. Thank you to corroboree frog (Pseudophryne corroboree). In NSW NSW Ross Sadlier from the AM for checking the reliability of using National Parks and Wildlife Service Species Management ventral patterns for identification of individual red-crowned Report Number 8 NSW National Parks and Wildlife toadlets, to Jacquie Recsei and Marion Anstis for reviewing Service, Sydney. the manuscript, to John de Bavay for his help in researching Parker, H.W., (1940) The Australasian frogs of the family the New England specimens, and to Thomas Thumm for help Leptodactylidae. Novit. Zool. 42: 1-106. with mapping. The study of the geological preferences of red-crowned toadlets was supported by a Peter Rankin Richards, S. J. and Bull C.M., (1990) Size limited predation on Trust Scholarship. tadpoles in three Australian frogs. Copeia 4: 1041-1046. Schlesinger C.A. and Shine R., (1994) Choosing a rock: The manuscript was improved by the suggestions of two Perspectives of a bush-rock collector and a saxicolous anonymous referees. lizard. Biological Conservation 67: 49-56. Shine, R. and Fitzgerald, M., (1989) Conservation and REFERENCES reproduction of an endangered species: the broad-headed Barker J. and Grigg G., (1977) A Field Guide to Australian snake, Hoplocephalus bungaroides (Elapidae). Australian Frogs. Rigby, Adelaide. Zoologist, 25(3): 65-67. Buchanan, R. A., (1989) Bush Regeneration. Recovering Sydway (1996) Greater Sydney Street Directory 3rd Edition. Australian Landscapes. TAFE Student Learning Thumm, K and Mahony M., (1997) The red-crowned toadlet, Publications NSW. Pseudophryne australis: Pp. 143-156 in Threatened Frogs of Bureau of Meteorology, (1991) Sydney Climatic Survey, NSW Habitats, Status and Conservation Ed. Harald Ehmann, Australian Government Publishing Service, Canberra. published by the Frog and Tadpole Study Group of NSW Inc. Chapman, G.A. and Murphy, C.L., (1989) Soil Landscapes of Tyler, M. J., (1989) Australian Frogs.Viking O’Neil. Australia. the Sydney 1:100 000 Sheet. Soil Conservation Service Woodruff, D. S., (1976) Courtship, reproductive rates, and of NSW, Sydney. mating system in three Australian Pseudophryne (Amphibia, Clements, A., (1983) Suburban development and resultant Anura, Leptodactylidae) Journal of Herpetology. 10 (3):13-318. changes in the vegetation of the bushland of the northern Woodruff, D. S., (1978) Hybridization between two species of Sydney region. Australian Journal of Ecology 8: 307-319. Pseudophryne (Anura, Leptodactylidae) in the Sydney Basin Australia. Proc. Linn. Soc. of NSW, 102 (3):131-147.

108 Status of Temperate Riverine Frogs in South-eastern Australia

Graeme Gillespie1 and Harry Hines2

ABSTRACT factors limiting distribution and abundance, or identifying causative agents of declines.A range of Knowledge of the distribution and population potentially threatening processes exists throughout declines of temperate riverine frogs in south- the region, some of which are involved in some eastern Australia is reviewed. Patterns, nature and population declines. Systematic surveys are required potential causes of declines are examined, and level to fully ascertain the current status and distribution of current knowledge of demography and biology of of riverine species in the region.A strategic species for addressing declines is assessed.There are monitoring program is required to document on- nine obligate lotic species and three facultative going trends in populations of both declining and riverine species currently recognised in temperate non-declining species. Research is required on the south-eastern Australia. ecological requirements of most species, and to The systematics of some taxa are unresolved and examine impacts of specific threatening processes. this number is expected to increase. One species has disappeared, three species have suffered major declines, and a further five species are believed to have suffered minor population declines. For many species, current knowledge of distribution and abundance is inadequate for properly assessing magnitude and nature of declines. Information on ecology of most species is inadequate for assessing

1. Arthur Rylah Institute, Department of Natural Resources and Environment, PO Box 137, Heidelberg,Victoria 3084 Australia. Zoology Department, University of Melbourne, Parkville 3052,Victoria, Australia. 2. Conservation Resource Unit, Queensland Parks and Wildlife Service, PO Box 42, Kenmore, Queensland 4069.

109 INTRODUCTION DELIMITATION OF THE REGION Declines have been reported in the populations of numerous The region under consideration encompasses the mesic frog species in Australia (Osborne 1989, 1990; Czechura and south-eastern region of the continent, as defined by Littlejohn Ingram 1990; McDonald 1990;Tyler 1991;Watson et al. 1991; (1981), south of 29° S (Figure 1).This includes much of the coast and ranges of South Australia and Victoria, and the Ingram and McDonald 1993; Mahony 1993; Richards et al. coast and ranges of New South Wales as far north as the 1993;Trenerry et al. 1994; Hollis 1995; Gillespie and Hollis Northern Tablelands.This region generally falls within the 1996;White and Pyke 1996;Tyler 1997).A large proportion warm temperate climate type identified by Walter and Lieth of species reported to have declined in Australia consists of (1967 cited in Bridgewater 1987), which has a noticeable riverine species from the eastern sea-board region. Of the 41 winter, with year-round rainfall. Most of this region receives a species listed as Endangered,Vulnerable or Insufficiently median annual precipitation in excess of 500 mm, and the Known in Australia by Tyler (1997), 20 are lotic species.This annual average evaporation is less than 2400 mm (Bureau of number represents roughly 60% of the lotic frog species in Meteorology 1975, 1989).The region has a predominantly Australia. For many of these species the causes of decline are winter maximum rainfall regime in the south, and uniform rainfall in the east (Bureau of Meteorology 1975, 1989).The poorly known. north has tropical influences and receives a summer In the temperate regions of south-eastern Australia there are maximum rainfall regime. at least nine lotic species and several facultative stream- Species considered in this review are either restricted to, or breeders (Barker et al. 1995). Six of these are listed as have significant proportions of their distributions within this threatened by State or Federal agencies, either because of region.There is not a discrete latitudinal or altitudinal population declines and/or rarity (Table 1).This paper reviews boundary between temperate and sub-tropical regions, and the current status of lotic and facultative stream-breeding some species occur in both. One exception is the inclusion of species in temperate south-eastern Australia. Many biologists the Stream-bank Froglet Crinia riparia from the Flinders and organisations have contributed to the current level of Ranges, South Australia, which falls within the xeric Eyrean knowledge of species in this region.Whilst we acknowledge zoogeographical subregion (after Littlejohn et al. 1993).This species has been included to provide a complete review of this contribution, it is not an aim of this review to present a lotic species in south-eastern Australia. detailed synthesis of all knowledge of the distribution and biology of these species. Rather, this paper provides an overview of the current status, nature and extent of declines, and potential threatening processes, and identifies areas where further data are required to assist conservation agencies with the establishment of priorities for research and management.

TABLE 1: List of lotic and facultative stream-breeding frogs that occur in temperate south-eastern Australia, with their current status identified by State and National Authorities (NSW Threatened Species Conservation Act 1995;Tyler 1997; NRE 1999; Commonwealth Endangered Species Protection Act 1992).

Species Current Status listed by State and National Authorities

Obligate stream breeders Hylidae Litoria booroolongensis Endangered (NSW), Insufficiently known (Tyler 1997) Lesueur’s Frog L. lesueuri not listed L. citropa not listed Leaf-green Tree Frogs L. phyllochroa complex not listed Peppered Tree Frog L. piperata Vulnerable (NSW),Vulnerable (Tyler 1997) Spotted Tree Frog L. spenceri Endangered (Commonwealth); Endangered (Tyler 1997); Critically Endangered (Vic); Endangered (NSW) New England Tree Frog L. subglandulosa Vulnerable (NSW), Insufficiently Known (Tyler 1997)

Myobatrachidae Stuttering Frog Mixophyes balbus Vulnerable (Tyler 1997); Endangered (Vic);Vulnerable NSW) Stream-bank Froglet Crinia riparia not listed

Facultative stream breeders Myobatrachidae Tusked Frog Adelotus brevis not listed Giant Burrowing Frog Heleioporus australiacus Insufficiently known (Tyler 1997);Vulnerable (Vic); Vulnerable (NSW) Banjo Frog Limnodynastes dumerilii dumerilii not listed

110 FIGURE 1: The temperate region of south-eastern Australia, showing main geographic features

DATA COMPILATION AND ASSESSMENT during this time. Information for the period from 1990 OF POPULATION DECLINES onwards therefore provides a useful comparison of current distributions and status with historical information. Distributional data for each species were compiled from records held in the Australian Museum (AM), Museum Victoria Published and unpublished literature was reviewed for (MV), National Wildlife Collection (NC), South Australian assessments or reports of population declines, to assess Museum (SM), New South Wales National Parks and Wildlife current knowledge of distribution and population status and Service Wildlife Atlas (NSW Atlas), and the Department of level of knowledge of biology of each species; and to identify Natural Resources and Environment Atlas of Victorian Wildlife potential threatening processes.This review was augmented (Vic Atlas).These data were augmented with additional records where possible by discussions with other recognised authorities from the literature, unpublished reports and personal on particular species or those with local experience or observations/communications from herpetological authorities expertise in particular regions. where available. Data were vetted for obvious mistakes or doubtful records. From this information distribution maps were GUILD COMPOSITION AND compiled for each species, showing records pre-1990 and 1990 BIOGEOGRAPHY onwards. Although somewhat arbitrary, many declines were first reported during the 1980’s.The 1990’s roughly Lotic species are defined as those which invariably breed in incorporates the recent period of general heightened concern permanent-flowing streams or associated stream-channel about declining frogs and increased attention to conservation habitats. Numerous frog species breed in riparian habitats to status.There has been a relatively high level of surveying activity varying degrees in south-eastern Australia. In this review, for frogs throughout the region during this time.This has facultative stream-breeders are defined as those species included general fauna surveys conducted by NSW National which often breed in permanent-flowing stream habitats, but Parks and Wildlife Service Biodiversity Study, pre-logging and which also commonly reproduce in lentic habitats away from Regional Forest Assessment surveys in Victoria and NSW, and streams. Nine lotic species occur within this region (Table 1). targeted frog surveys (e.g., CNR 1993; Gillespie and Hollis Most of these are hylids and two are myobatrachids.The 1996; Mahony 1996 unpubl.; Ehmann 1997; Holloway and hylids comprise two species groups, which are distinctive on Osborne 1996 unpubl.; Hunter and Gillespie 1999). Many sites morphological and phylogenetic bases (see Tyler and Davies where species were historically known to occur were revisited 1978; Hutchinson and Maxson 1987; McDonald and Davies

111 1990).The systematics of some species complexes within these species mostly oviposit in lentic stream-side pools these groups are still being resolved (Donnellan, Mahony, (Gillespie pers. obs.) or slow-moving sections (Odendaal Knowles, Foster unpubl. data). et al. 1982).There are no reported declines in these species, and they are not considered in this review. The Litoria citropa group (Tyler and Davies 1978) includes the Blue Mountains Tree Frog (L. citropa), the Cascade Tree Frog SPECIES REVIEWS (L. pearsoniana), the Leaf-green Tree Frog (L. phyllochroa complex), the Peppered Tree Frog (L. piperata) and the New The nature and magnitude of population declines vary greatly England Tree Frog (L. subglandulosa).The Spotted Tree Frog (L. between temperate riverine species, as does the knowledge spenceri) arguably also resides in this group, based upon larval base for making informed assessments of current population and adult morphology and call structure (Watson et al. 1991; status or causes of decline. Available information on Hero and Gillespie 1993).The L. citropa group is restricted to demography and nature of population declines is described temperate south-eastern Australia, with the exception of L. for each species below: pearsoniana. Litoria pearsoniana also occurs in the temperate zone but is predominantly distributed within subtropical Booroolong Frog (Litoria booroolongensis) regions (McDonald and Davies 1990) and is discussed by Restricted to New South Wales, predominantly along the Hines et al. 1999. Species in the L. citropa group are western-flowing streams of the Great Divide from 200 to predominantly restricted to streams draining the eastern fall of above 1000 m above sea level (asl), from catchments draining the Great Dividing Range, with the exception of L. spenceri the Northern Tablelands to the in the southern (Watson et al. 1991). Highlands (Heatwole et al. 1995; Anstis et al. 1998; Hunter The L. lesueuri group in south-eastern Australia comprises and Gillespie 1999; NSW atlas records).This species was Lesueur’s Frog (L. lesueuri) and the Booroolong Frog formerly abundant along streams draining the Northern (L. booroolongensis) (see Barker et al. 1995; Anstis et al. 1998). Tablelands (Heatwole et al. 1995) (Figure 2). Further south, Litoria booroolongensis is predominantly restricted to streams there are relatively few historical records (AM records; NSW draining the western slopes of the Great Divide in New Atlas).There are problems with accurate identification of South Wales, while L. lesueuri is widespread along both sides L. booroolongensis which have confounded assessments of of the Great Divide, from Victoria up the east coast to north- current status, especially in the south of its range where the east Queensland (AM, NSW and Vic Atlas records). Its status species is superficially similar to L. lesueuri (Gillespie 1999). is also discussed by Hines et al. (1999) and McDonald and Some records, mostly from south-eastern New South Wales, Alford (1999). are not supported by specimens (e.g., NSW Atlas), and require confirmation. Only two obligate lotic myobatrachid species occur in temperate south-eastern Australia: the Stuttering Frog Litoria booroolongensis has not been recorded from the (Mixophyes balbus) and Crinia riparia (Barker et al. 1995). Northern Tablelands during the past 15 years despite extensive Mixophyes balbus is restricted to the eastern fall of the fauna surveys in recent years by the North-east Forest Great Dividing Range. Crinia riparia is restricted to streams Biodiversity Study (NSW NPWS 1994), Regional Forests draining the Flinders Ranges in South Australia (Odendaal Assessment Program and others (Harris, University of New and Bull 1983). England, pers. comm. in Anstis et al. 1998; Hines pers. obs.; Mahony, University of Newcastle pers. comm.).There are Three facultative stream breeders (all myobatrachids) occur several recent records of the species on the eastern slopes of in this region (Table 1). Adelotus brevis is a monotypic genus, the Great Divide, north of Newcastle and south of Wollongong occurring predominantly along the eastern fall of the Great (NSW Atlas), but specimens have not been lodged with any Divide, with a limited distribution in some western drainages museums. Intensive fauna surveys in north-eastern NSW over (AM, QM, and NSW Atlas records). It also occurs in the past eight years (e.g., NSW NPWS 1994) have failed to subtropical areas, and its status in those areas is discussed by locate this species, suggesting that its current status is extremely Hines et al. (1999) and McDonald and Alford (1999).The rare.The species may have also declined at former sites in the Giant Burrowing Frog (Heleioporus australiacus) is restricted Blue Mountains (Recsei pers. comm.). to the eastern fall of the Great Divide, and is the only east- coast member of this genus (Lee 1967; Gillespie 1990). Five The species is recently known from near Tamworth (Mahony subspecies of the Banjo Frog (Limnodynastes dumerilii) are pers. obs.); this is the only known extant population in recognised (Watson and Littlejohn 1985), only one of which, northern New South Wales. Further south the species persists L. dumerilii dumerilii, appears to commonly reproduce in in the and Winburndale Creek,Winburndale streams (Martin 1972; Gillespie, pers. obs.).This subspecies is Nature Reserve (Macartney, Ranger at Winburndale Dam, pers. widespread, predominantly on the tablelands and western comm.). Recent extensive stream surveys targeting lotic species slopes of the Great Divide, from South Australia to south- in the Southern Highlands have not located the species in the east Queensland (Martin 1972). Tumut River and Yarrangobilly Creek, where it historically occurred (NC records), and only located it in the Several other frog species occasionally breed along streams in , which may be close to the geographic south-eastern Australia, such as the Broad-palmed Frog limit of the species (Hunter and Gillespie 1999). (Litoria latopalmata), Peron’s Tree Frog (L. peronii), species of the L. ewingii group (Holloway 1997; Gillespie pers. obs.) and Overall, there have been very few records of the species in the Common Froglet (Crinia signifera) (Odendaal et al. 1982; the past 5 years, contrasting markedly with the 1980’s when Holloway 1997). However, these species rarely breed in the species was perceived to be abundant (Heatwole et al. flowing streams.Where they do breed along water courses, 1995).This situation may reflect, in part, a lack of survey

112 effort in many regions; however, this is not the case in north- Mountains Scheme structures affecting stream flow. Such eastern NSW, nor in the Northern Tablelands (NSW NPWS disturbances to natural flow regimes may displace L. lesueuri 1994).The species may have declined over its entire range. from streams.

Knowledge of the ecology of this species is limited. Broad Both dead and moribund specimens of L. lesueuri which habitat associations and reproductive biology are described had a chytrid fungal infection have recently been found along by Anstis et al. (1998).The species persists in a range of some streams in south-east Queensland and north-east habitat types and geographic regions. For instance, in the Victoria (Berger et al. 1998; Gillespie and Berger unpubl. Goobarragandra River, the species occurs in relatively data).The impact of this disease upon populations of this undisturbed reaches within Kosciuzsko National Park, as well species is unknown. as highly modified reaches in farmland downstream (Gillespie and Hunter unpubl. data). Several aspects of the biology of L. lesueuri have been Several potentially threatening processes have operated, examined. Information has been gathered on microhabitat use or are operating, in various parts of the range of of larvae, adult activity patterns, life history, and larval predator- L. booroolongensis. Introduced fish occur in many streams in prey relationships (Martin et al. 1966; Richards and Alford which the species has been recorded (Hunter and Gillespie 1992; Gillespie 1997a unpubl.; Holloway 1997). Litoria lesueuri 1999; Anstis, Berowra Heights, NSW, pers. comm.). Although possesses several ecological attributes that distinguish it from tadpoles of L. booroolongensis appear to be less palatable to most other temperate lotic species. It is frequently observed trout compared with some other temperate lotic species away from streams, sometimes on dry ridges several kilometres (Gillespie unpubl. data), introduced fish may still exert from water courses (Mahony 1996 unpubl.; Gillespie pers. significant predatory pressure upon populations of this obs.). Despite a predominantly lotic reproductive strategy, L. species (Gillespie and Hero 1999). Land clearance, forest lesueuri also often breeds in isolated streamside lentic pools grazing and timber harvesting have occurred adjacent to many (Anstis et al. 1998), and occasionally breeds in lentic habitats streams, or in the headwaters of catchments in which the away from streams, such as quarries and dams (Mahony 1996 species has been recorded (Gillespie and Hines pers. obs.). unpubl.; Anstis et al. 1998). In contrast to most other lotic Flow modification has occurred in many streams and weed species within the region, L. lesueuri persists along some invasion, particularly by willows, has grossly modified riparian habitats (Anstis et al. 1998; Hunter and Gillespie 1999). streams through cleared pastoral areas and urban fringes (AM, NSW and Vic atlas records).This may suggest greater tolerance Lesueur’s Frog (Litoria lesueuri) for habitat alteration or disturbance.Tadpoles of Litoria lesueuri are relatively unpalatable to introduced trout, which may Litoria lesueuri is widespread from Lerderderg Gorge in afford them more protection against predation from central Victoria, along the east coast as far as Cooktown, introduced fish compared with other sympatric lotic species north-eastern Queensland and along the western fall of the (Gillespie in review). Great Divide to south-east Queensland (Figure 3).The species is common, and occurs in a broad range of stream and forest habitats, from the coast to 1200 m asl (AM, NSW Blue Mountains Tree Frog (Litoria citropa) and VIC Atlas records). It is widely thought that this species Litoria citropa has a widespread distribution along streams east comprises a number of taxa, with two in temperate Australia of the Great Divide, from the Mitchell River in eastern Victoria (Barker et al. 1995; Mahony 1996 unpubl.; Donnellan and (VIC Atlas records), north to the Hunter River in NSW (Anstis Mahony unpubl. data). Litoria lesueuri (sensu stricto) occurs in and Littlejohn 1996) (Figure 4). It occurs up to 1000 m asl in the south, from Victoria and along the east coast and ranges the north of its range at Blackheath, NSW (Anstis and Littlejohn to Sydney, and along the western slopes of the Great Divide 1996), but is more widespread at lower altitudes (AM, NSW to the Australian Capital Territory (Donnellan and Mahony and VIC Atlas records). In southern NSW and Victoria it is unpubl. data). Hybrids of L. lesueuri and the taxon to the north may occur in the upper reaches of the Murrumbidgee restricted to below 500 m asl (NSW and Vic Atlas; AM River and tributaries around Canberra (Donnellan and records). Littlejohn et al. (1972) report the south-western limit Mahony unpubl. data).These taxa are treated in this review of this species at Aberfeldy,Victoria, based on an Australian as one species. Museum specimen with this locality. However, recent surveys found L. citropa to be abundant along the Mitchell River and Litoria lesueuri remains abundant at numerous localities, and streams further east in Gippsland, but absent in all streams west there is little evidence of any decline (Gillespie and Hollis of the Mitchell River, including the Valencia, Avon, Macallister, 1996; Mahony 1996 unpubl.). Notably, this species persists at Aberfeldy and Thompson Rivers (Gillespie and Hollis 1996; many sites, in some cases in abundance, where other species Gillespie unpubl. data).The Mitchell River catchment is also the have declined (Mahony 1993; Richards et al. 1993; Gillespie south-western biogeographical limit of a number of other warm and Hollis 1996). Litoria lesueuri is absent, or occurs in very low numbers, in some montane streams in north-eastern temperate elements, including several riparian and rainforest Victoria and southern NSW, which may indicate declines in plant species (Cameron, pers. comm. Flora Section, Department these regions prior to surveys, but this is unknown (Gillespie of Natural Resources and Environment,Victoria).The origin of and Hollis 1996; Hunter and Gillespie in 1999). the Aberfeldy specimen is therefore doubtful; it is more probable that the Mitchell River catchment is the south-western In the southern highlands, Hunter and Gillespie (1999) found limit of the range of this species. that L. lesueuri and other lotic frog species were absent from many streams below impoundments or other Snowy

113 Litoria citropa occurs along slow-flowing pool sections of Members of the complex are common and widespread in permanent or semi-permanent streams through dry low to mid-elevation forests. Litoria phyllochroa (sensu stricto) sclerophyll forest, temperate rainforest and coastal woodland occurs up to 1100 m asl in Barrington Tops. Litoria habitats (Tyler and Anstis 1975; Littlejohn 1981; Holloway barringtonensis (sensu stricto) also occurs in the Barrington 1997). It is restricted to streams with intact riparian and Tops area but the geographic distribution of this taxon in adjacent forest vegetation (Gillespie pers. obs.; Anstis pers relation to that of L. phyllochroa (sensu stricto) is currently comm.). Consequently, this species has been displaced from being resolved (Mahony pers. com.). Litoria nudidigitus occurs areas cleared for pasture or urbanisation in parts of its up to 1100 m asl in the upper Murray catchment (Gillespie historic range. Elsewhere, there is no evidence that this and Hollis 1996). species has declined. However, few surveys have been As with L. citropa, these species have been displaced from conducted through most of the range of this species, and areas cleared for pasture or urbanisation in parts of their only limited monitoring of populations has occurred. Mahony historic range. Elsewhere there is little evidence of any (1993) visited several sites in the Watagan Mountains at the widespread decline of these species (Mahony 1993; Figure 5). northern limit of L. citropa between 1977 and 1992 and Fauna surveys in eastern Victoria during the past 10 years reported that, although at low abundance, there was no have found L. nudidigitus to be common and widespread evidence of a decline in numbers of L. citropa. along many streams in lowland and coastal areas (VIC Atlas The species is well represented within existing reserve records). However, surveys have not been comprehensive systems of southern NSW and eastern Victoria (VIC and throughout the range of the species, and monitoring of NSW Atlas records). Many catchments in which this species populations has only occurred at a few sites. Mahony (1993) occurs are subject to intensive timber harvesting.This species found no evidence of declines of L. phyllochroa (sensu stricto) persists along some streams in catchments in which logging at several lowland sites monitored in the Watagan Mountains has occurred (Gillespie pers. obs.); however, the health or in NSW between 1977 and 1993. Some upland populations stability of these populations is unknown. Forest grazing and of L. barringtonensis (sensu stricto) may have declined in the land clearance up-stream also occur in some catchments. north of its range (Mahony pers. comm.). Otherwise, Again, the impact of these disturbances is unknown. L. phyllochroa (sensu lato) has been frequently recorded in recent surveys in north-eastern NSW (NSW NPWS 1994). Several aspects of the biology of L. citropa have been examined. Information has been gathered on microhabitat Above 500 m asl L. nudidigitus is relatively uncommon, with a very patchy occurrence (Gillespie and Hollis 1996; Hunter use of larvae, adult activity patterns, life history, and larval and Gillespie in press), and this may be due to population predator-prey relationships (Tyler and Anstis 1975; Anstis and declines prior to surveys in these areas.The tadpoles of Littlejohn 1996; Gillespie 1997a unpubl.; Holloway 1997; L. nudidigitus are palatable to introduced trout, which are Gillespie unpubl. data). common and widespread in upland streams of this region Leaf-green Tree Frog complex (Litoria (Gillespie 1997a, unpubl.).Trout are probably a major factor in determining the low density of this species in upland steams. phyllochroa (sensu stricto), L. nudidigitus and L. barringtonensis (sensu stricto)) Several populations of L. nudidigitus have been regularly monitored in north-eastern Victoria and the Southern The L. phyllochroa complex appears to comprise three species Highlands of NSW since 1994 (Gillespie unpubl. data). Most (Mahony, Knowles and Donnellan, unpubl. data). Litoria of these populations have remained relatively stable during this phyllochroa (sensu stricto) occurs from the Sydney region to time. However, in 1996 one population suddenly declined at the Coffs Harbour region. Litoria barringtonensis (sensu its upper limit of distribution (1100 m asl) in a trout-free stricto) (Copland 1957) occurs north from the Hunter River section of Bogong Creek, (Gillespie to near Gibraltar Range.The name L. nudidigitus was applied 1997a unpubl.; Gillespie 1998 unpubl.). Monthly surveys during to the southern taxon of Littlejohn (1967), after the the summers of 1996/97 and 1997/98 located very few frogs subspecies L. phyllochroa nudidigitus identified by Copland or tadpoles in this section of the stream. No frogs or tadpoles (1962). Litoria nudidigitus occurs from the Sydney region along were located in the summer of 1998/99. Further down- the eastern slopes of NSW to the Thompson River in stream, below 800 m asl, the population has remained eastern Victoria. It also has a limited distribution on the north relatively stable.The cause of this localised decline is unknown. and western sides of the Great Divide in tributaries of the in north-eastern Victoria (Gillespie and The biology of L. nudidigitus has been extensively studied. Hollis 1996;VIC Wildlife Atlas), and the Upper Murray and Information has been gathered on microhabitat use by adults Goodradigby Rivers in southern NSW, and the Cotter River and larvae, movement and activity patterns, population age- in the ACT (Gillespie and Osborne 1994; Hunter and structure and life history, larval competition and predator- Gillespie in press). prey relationships (Hero and Gillespie 1993; Gillespie 1997a, unpubl.; Holloway 1997; Gillespie in review).The biology of The specific taxonomic identities of specimens and records of L. phyllochroa (sensu stricto) and L. barringtonensis (sensu this complex held in Museums and wildlife atlas databases stricto) has not been examined specifically, but is likely to be have not yet been resolved, precluding the presentation of very similar to that of L. nudidigitus. accurate maps of the distributions of records of each taxa within the complex separately. All distribution records of the complex are presented in Figure 5.

114 FIGURE 2: Distribution of the Booroolong Frog Litoria FIGURE 3: Distribution of Lesueur’s Frog Litoria lesueuri. booroolongensis. Open circles indicate records collected pre- Open circles indicate records collected pre-1990; closed circles 1990; closed circles are those recorded 1990 onwards. are those recorded 1990 onwards.

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

FIGURE 4: Distribution of the Blue Mountains Tree Frog Litoria FIGURE 5: Distribution of the Leaf-green Tree Frogs of the Litoria citropa. Open circles indicate records collected pre-1990; phyllochroa complex. Open circles indicate records collected pre- closed circles are those recorded 1990 onwards. 1990; closed circles are those recorded 1990 onwards. Taxonomic boundaries between species within this complex have not yet been resolved. Australian Museum records are not included.

● 1990 onwards record ● 1990 onwards record ❍ pre-1990 record ❍ pre-1990 record ■ capital cities ■ capital cities

115 Peppered Tree Frog (Litoria piperata) similar to L. pearsoniana (Tyler and Davies 1985; McDonald and Davies 1990). It is possible that L. piperata represents Litoria piperata was formerly known from five streams morphologically distinct outlying populations of L. pearsoniana. draining the east of the Northern Tablelands, from 800 — Further genetic and morphometric studies are required to 1120 m asl, from Gibraltar Range to Armidale, northern resolve the systematics of these populations before the NSW (Tyler and Davies 1985) (Figure 6). Despite searches of conservation status of L. piperata can be properly assessed. the historic localities and other streams with similar habitat Further targeted surveys are also required of other streams in within the region (Mahony et al. 1997a; Hines unpubl. data.), the region to try and locate any other remaining populations. the species was last sighted in 1973.

Little is known about the biology of L. piperata.The Spotted Tree Frog (Litoria spenceri) morphological similarity of L. piperata to L. pearsoniana and Litoria spenceri is restricted predominantly to the western fall L. phyllochroa suggests that ecological similarities are also likely. of the Great Divide, from Lake Eildon to , Causes of the apparent disappearance of L. piperata are from 280–1100 m asl (Gillespie and Hollis 1996) (Figure 7). unknown; however, most of the historic sites and other Extensive systematic surveys have been conducted for this streams in the region have undergone substantial alteration species throughout eastern Victoria and southern NSW and suffered significant habitat disturbance through land (Watson et al. 1991; Gillespie and Hollis 1996; Hunter and clearance, grazing and timber harvesting (Hines pers. obs.). Gillespie 1999; Gillespie 1998 unpubl.). Litoria spenceri has Introduced predatory fish species (Eastern Gambusia only ever been found in 19 streams, and has always been Gambusia holbrooki and salmonids) also occur in these considered to be rare (Watson et al. 1991).The species is streams and may have displaced frog populations by now believed to be extinct in four of these streams, and has predation upon larvae. declined substantially in distribution and abundance along most others (Gillespie and Hollis 1996).These declines Populations of frogs which closely resemble L. piperata, were occurred in the 1970’s and early 1980’s (Watson et al. 1991); located on the Northern Tablelands north of the known but based upon the known demography of the species, it historic range of the species, in 1992 (NSW NPWS 1994) probably suffered population declines over a wider area (Figure 6).The advertisement calls of males in these earlier in this century, and possibly late in the last century populations were similar to that of L. pearsoniana (Mahony (Gillespie and Hollis 1996).The remaining streams comprise pers. comm. in Tyler 1997).The advertisement call of 12 discrete isolated populations (Gillespie and Hollis 1996; L. piperata is not known but the species is morphologically Hunter and Gillespie 1999). Based upon density estimates from mark-recapture studies, surveys, population monitoring, and habitat modelling, the largest of these populations is FIGURE 6: Distribution of the Peppered Tree Frog Litoria estimated to contain approximately 1000-1500 adults in the piperata. Open circles indicate records collected pre-1990; upper (Gillespie unpubl. data).The sizes of all closed circles with ‘?’ are records 1990 onwards with questionable other populations are estimated to be less than 1000 adults. taxonomic identity. The species is restricted to riffle and cascade stream sections with exposed rock banks, resulting in a highly patchy distribution along most streams (Gillespie and Hollis 1996). Individuals are highly sedentary, not venturing away from the stream (Gillespie 1997a unpubl.). Most adults appear to move less than 80 m over several years (Gillespie 1997a unpubl.). The patchy distribution of demes makes them highly prone to reduction because of local environmental and demographic stochastic extinction processes and unnatural disturbances.The long-term viability of remaining populations is therefore not secure.

The ecology and population dynamics of L. spenceri have been investigated in some detail (Gillespie 1997a unpubl.). Habitat associations of all life stages, movement and activity patterns, growth and population structure have been examined.The role of potentially threatening processes, such ● 1990 onwards record as introduced fish and stream disturbances, have also been ❍ pre-1990 record examined. A national Recovery Plan has been prepared for ■ capital cities the species (Robertson and Gillespie in review).

Human disturbances to streams, such as gold dredging, forest roads and recreational pressures, are correlated with the general pattern of decline of L. spenceri (Gillespie and Hollis 1996).Trout species occur throughout the distribution of L. spenceri, and are able to exert significant predation pressure on larvae (Gillespie 1997a unpubl.).Trout are believed to be a major cause of population declines of L. spenceri (Gillespie 1997a unpubl.; Robertson and Gillespie in review).

116 FIGURE 7: Distribution of the Spotted Tree Frog Litoria spenceri. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.

A program monitoring eight populations commenced in metamorphs emerged from this clutch. In 1997/98 only two 1994, and was recently expanded to include all populations of these juveniles were located; no other frogs were found (Robertson and Gillespie in review). Censuses have been (Gillespie 1998 unpubl.). In 1998/99 one of these juveniles conducted annually along 1 km transects on each stream. and one adult male, located in the summer of 1996/97, were Most adult populations have remained relatively stable during found (Gillespie unpubl. data). the five years of monitoring (Gillespie unpubl. data). However, the population of L. spenceri at Bogong Creek, Kosciuszko During the season prior to the population decline (1995/96), National Park, suffered a precipitous decline in 1996 several dead and one moribund frog were located.The (Gillespie 1997a unpubl.).This population is highly unusual. It moribund frog was found to be infected with a chytrid is mostly confined to a short (1.6 km) stream reach above a fungus, recently identified by Berger et al. (1998).This fungus waterfall and below a pondage and aqueduct for the Snowy has been associated with other declining frog populations Mountains Scheme, which excludes trout (Gillespie 1997a (Berger et al. 1998), but its role in these declines is not unpubl.). It is the highest elevation population, and was the known. An alternative cause of the decline may have been only high-density population known prior to this decline the record flood in Bogong Creek in October 1996 (Snowy (Gillespie and Hollis 1996). During the three years prior to Mountains Authority records). However this event the decline, the population was intensively studied and occurred prior to any breeding in that season, so no eggs or estimated to comprise 700–1000 adults and several thousand tadpoles were present in the stream.The species is adapted immature frogs (Gillespie 1997a unpubl.). In the summer of to living in mountain streams with high energy spring snow 1996/97 only six adults, three juveniles, and one clutch of melt (Hughes and James 1989) and the adult population is tadpoles were located (Gillespie 1998 unpubl.). Over 130 expected to be able to cope with these events. Numbers of

117 L. nudidigitus also declined from this section of stream, but the Several potentially threatening processes operate in population further down stream remained stable (Gillespie catchments containing populations of L. subglandulosa. 1998 unpubl.).This suggests that the causative agent of The species persists along streams in catchments in which decline was restricted to the upper reaches. timber harvesting has occurred (Anstis 1997); however the health and stability of these populations are unknown. Forest A chytrid fungus was detected in several dead and moribund grazing and aerial spraying has occurred at some sites in the adults of L. spenceri from three other populations in Victoria vicinity of the type locality (Anstis 1997).Trout have been during the summer of 1997/98 (Gillespie and Berger unpubl. introduced into many streams, and may be preying on data). Subsequent monitoring has not detected any declines tadpoles (see Gillespie and Hero 1999). of these populations (Gillespie unpubl. data).The role of the chytrid fungus in the population dynamics of this and other Stuttering Frog (Mixophyes balbus) frogs remains to be resolved. Mixophyes balbus is restricted to the eastern slopes of the New England Tree Frog Great Divide, from the catchment in far East (Litoria subglandulosa) Gippsland,Victoria; to tributaries of the Timbarra River near Drake, NSW (Mahony unpubl. data, AM records; NSW atlas; Litoria subglandulosa is restricted to the eastern fall of the VIC atlas records) (Figure 9).The species occurs over an Great Divide from Fal Brook in the Barrington Tops National altitude range of 20 to over 1400 m asl; from low to high Park, NSW, to near Stanthorpe, just north of the Queensland altitudes from south to north (Straughan 1968; AM records; border, at altitudes from 500 to 1400 m asl (Tyler and Anstis NSW atlas;VIC atlas records). It is typically found in 1975; Heatwole et al. 1995; Anstis and Littlejohn 1996; NSW association with permanent streams through temperate and Atlas; AM and QM records) (Figure 8). Recent taxonomic sub-tropical rainforest and wet sclerophyll forest (Mahony et studies suggest that there may be two sibling species within L. al. 1997b), and also in moist gullies in dry forest (Gillespie subglandulosa, separated north and south near the latitude pers. obs.; Hines unpubl. data). 30o 30’ S (Mahony, Knowles, Foster and Donnellan unpubl. data) but in this review it is treated as a single species. Mixophyes balbus was formerly more frequently encountered in the northern parts of its range (AM records; NSW atlas; Knowledge of the historical distribution of L. subglandulosa VIC atlas records) than south of Sydney (Figure 9). It may is limited. Prior to 1975, this species was known from only three have been uncommon in the south, or this may reflect limited localities (Tyler and Anstis 1975), and few other localities were historic searches in that region.Webb (1991) reported the reported until the 1990’s (Anstis and Littlejohn 1996). Most species as ‘common’ in streams at Bondi State Forest, in records of this species are from surveys since 1990 (Figure 8) particular, below 800 m asl.The species has only been found (see Anstis 1997).These surveys have not been comprehensive in Victoria on three occasions (Tennyson Creek, Cann River and more populations may exist. Consequently, there is a limited and Jones Creek) (VIC Atlas records), and has always been historical base for assessment of population declines.The species considered rare in that State. may have disappeared or suffered a drastic decline in three streams in the vicinity of the type locality, near Point Lookout Tyler (1997) reports that the species has declined during the (Anstis and Littlejohn 1996; Anstis 1997).These authors also past ten years in many areas where it was historically known. report that other lotic species historically common in these In the northern part of its range Mahony (1993) reported streams, L. booroolongensis, L. pearsoniana, and M. balbus, have also that M. balbus was common in the Watagan Mountains area declined (Anstis and Littlejohn 1996; Anstis 1997). Anstis (1997) in the early 1980’s, but had declined by the 1990’s.The species also reports that the type locality has undergone significant has declined at several sites in the vicinity of the type locality alteration since the 1970’s, when L. subglandulosa was common near Point Lookout, NSW, where it was formerly common in there. Much of the fringing riparian vegetation has gone, the 1960’s and 1970’s (Anstis and Littlejohn 1996; Anstis presumably from cattle grazing (Anstis and Littlejohn 1996; Anstis 1997). Elsewhere the species has only been found in low 1997). Introduced trout species have been released into these numbers (Mahony 1996 unpubl.; Mahony et al. 1997b). streams (Anstis 1997). Litoria subglandulosa persists at other localities documented in the 1970’s and 1980’s (Anstis and Surveys in south-eastern NSW since 1990 by State Forests Littlejohn 1996; Anstis 1997). However, surveys and monitoring NSW and NSW NPWS have located individuals of M. balbus are currently inadequate to assess population trends. at only a few sites (Lemckert et al. 1997; Daly 1998; Lemckert unpubl. data). Intensive fauna surveys in East Litoria subglandulosa occurs along slow-flowing pool sections Gippsland conducted in the past 15 years have not located of permanent streams, through dry and wet sclerophyll forest, this species (see CNR 1993). Extensive searches at historical rainforest, montane forest and heathland (Anstis and Littlejohn sites and streams in close vicinity over the past seven years 1996; Anstis 1997).The species has also been recorded along have not located the species (Holloway and Osborne 1996 streams through semi-cleared grazing land (Anstis 1997). unpubl.; Gillespie, Osborne and Holloway and unpubl. data). Knowledge of the biology of L. subglandulosa is limited.The reproductive biology and some aspects of life history have Mixophyes balbus may now be extinct in Victoria and some been documented (Tyler and Anstis 1975; Anstis and other parts of its range in NSW. However, the species is Littlejohn 1996). Apart from broad habitat associations, the highly cryptic and small populations may remain undetected. ecological requirements of adults and larvae are unknown.The Elsewhere the species is now mostly extremely rare.The species has a distinctive larval morphology, unique amongst collective observations of M. balbus (Mahony et al. unpubl. Australian hylids (Tyler and Anstis 1975), which may reflect data) suggest that the species tends to have a patchy specialised diet and microhabitat requirements. distribution along water courses.While actual estimates of population size are not available, where populations have

118 been recorded recently, the species appears to be in very Giant Burrowing Frog (Heleioporus low numbers (Mahony and Knowles pers. comm.). Further australiacus) surveys are required in the region to properly ascertain the status of the species. The species is confined to eastern slopes of the Great Dividing Range and coastal regions from the southern end of Knowledge of the biology of M. balbus is limited. It has more the Olney State Forest north of Sydney, NSW (AM records), ecological similarities with its sub-tropical and tropical to Walhalla, in the Central Highlands of eastern Victoria congeners, rather than other temperate zone lotic frog (Littlejohn and Martin 1967;VIC Atlas) (Figure 10).The species. Its reproductive biology is very similar to M. fleayi,a species has been found near sea level on the coast, and congeneric species in north-eastern NSW and south-eastern almost 100 km inland, along the escarpment of the Great Queensland (Knowles et al. 1998 unpubl.). Eggs are deposited Dividing Range, up to 1000 m asl. (Gillespie 1990, AM in very shallow, gently running water either in a shallow Records;Webb, pers. comm. in Daly 1996; Recsei 1997). Most excavation in the stream bed or pasted directly onto bed rock records are concentrated at the northern end of the range, (Knowles et al. 1998 unpubl.).Tadpoles are free-swimming, in the Sydney region on the Hawkesbury sandstone developing in pools and runs (Knowles et al. 1998 unpubl.). formation. Most other records are from the southern part of the range, in eastern Victoria and the south-east corner of The specific ecological requirements of adults and larvae are NSW (Gillespie 1990). However, these constitute only 40 poorly known.The species has been found away from sparsely-distributed documented records (Gillespie 1990;VIC riparian habitats in the north of its range (State Forests NSW Atlas; NSW Atlas).There is a notable disjunction in the 1995), and adults may routinely disperse into surrounding distribution of records between Jervis Bay and the Eden forests outside of the breeding season. No information is District (Figure 10). One individual has been reported from available on population structure and dynamics. In north- Narooma State Forests, NSW (Wellington and Wells 1994, eastern NSW, statistical modelling was used to investigate the unpubl.).The paucity of records in this region may be due to relationship of M. balbus with 24 environmental predictors the rarity of the species, although it may also reflect the (NSW NPWS 1994).The species showed a preference for limited survey effort for frogs in south-eastern NSW. the interiors of large forest tracts in areas with relatively cool mean annual temperatures. Available information indicates that H. australiacus is rare (Webb 1987; Gillespie 1990; Recsei 1997; NSW Atlas). Several potentially threatening processes have operated at However, it is widely regarded as a highly cryptic species, sites where M. balbus has been found, or up-stream in usually only detected at night after heavy rains (Gillespie catchments. Logging and associated forest management 1990; Daly 1996).This characteristic has hampered the practices have been carried out in some catchments where detection of the species during fauna surveys and M. balbus historically occurred, or currently occurs (Mahony assessments of relative abundance. It also makes any rigorous et al. 1997b).The health or stability of extant populations in assessment of changes in local abundance or population these disturbed catchments is unknown. Forest grazing and trends impossible. land clearance for pasture up-stream have also occurred in some catchments (Mahony et al. 1997b). Mahony et al. Historically, H. australiacus appears to have been locally (1997b) report that the species is not known from any common in the Sydney region (Barker and Grigg 1977). localities with disturbed riparian vegetation or significant Choruses of the species have been reported as “common” human impacts up-stream. However, populations of this after summer rain storms, particularly in and around Royal species have also disappeared in catchments with seemingly National Park, NSW, and other fringes of suburban Sydney in minimal human disturbance (Mahony et al. 1997b).This may the 1960’s and 1970’s (Lee pers. comm. in Gillespie 1990; indicate that the species is highly sensitive to perturbations in Grigg, Dept. Zoology, University of Queensland, pers. comm.). its environment. There are relatively numerous records of the species from the Sydney region in recent years (Figure 10), but this may Tadpoles have been found in sympatry with native fish reflect an increase in reports in response to heightened (Knowles, Mahony and Hines unpubl. data), and probably have interest in frog-population declines.With few exceptions, survival strategies to avoid predation from them (see these recent records have been of single or few individuals. Gillespie and Hero 1999). However, the impact of introduced Concerns have been raised that H. australiacus has declined fish, such as Eastern Gambusia (Gambusia holbrooki), carp throughout many parts of the Sydney — Hawkesbury (Cyprinus spp.) and salmonids, is unknown. Introduced fish sandstone region (Recsei 1997; Mahony pers. comm.). Local (salmonids) have been recorded at sites where M. balbus has declines have mostly been in areas which have suffered declined (Anstis 1997). Mahony et al. (1997b) did not habitat fragmentation, along with direct and indirect impacts observe introduced fish at any sites where they found from urbanisation, which is placing increasing pressure upon M. balbus. Other introduced fish, such as G. holbrooki and many remaining populations in the region (Recsei 1997). Cyprinus spp., may also occur in some streams within the range of this species (see Gillespie and Hero 1999). However, Elsewhere, particularly in the southern part of the range of M. balbus has also disappeared from many streams which do the species, only one or a few adults have been located at not contain introduced fish species (Knowles and Mahony any one time (Gillespie 1990; Daly 1996; NSW Atlas). In unpubl. data; Hines and Gillespie pers. obs.). Victoria, the most individuals ever located were four over several kilometres on a wet road (Gillespie pers. obs.).Webb A National Recovery Plan is in preparation for this species. (1987) has also reported numbers of H. australiacus crossing a road in south-eastern NSW. Intensive fauna surveys in eastern Victoria over the past 15 years (see CNR 1993),

119 along with a general increase in opportunistic searching FIGURE 8: Distribution of the New England Tree Frog Litoria during this time, have generated most of the records in the subglandulosa. Open circles indicate records collected pre-1990; state. However, the species has been detected infrequently. closed circles are those recorded 1990 onwards. Other regional fauna surveys in eastern Victoria prior to this period failed to detect H. australiacus at all (LCC 1985; Norris et al. 1983). A recent frog survey in East Gippsland, targeting areas where H. australiacus had been found historically, only detected one individual (Osborne and Holloway, unpubl data). Many sites in Victoria at which the species has been recorded have been revisited, in some cases on numerous occasions, without relocating the species (Gillespie pers obs.; Holloway and Osborne 1996 unpubl.).

Information on the species in south-eastern NSW is similar to that for Victoria. Survey effort has been much more ● limited in this region.The species was detected during surveys 1990 onwards record ❍ pre-1990 record conducted by Lunney and Barker (1986) near Bega, and by ■ capital cities Webb (1991) in Bondi State Forest. In recent years, survey effort has been increased by the NSW State Forest Service. The low detection rate of this species suggests that it is very rare; however, it may also reflect inadequate survey effort or un-targeted survey techniques.

The paucity of records and absence of any systematic population monitoring imposes limits upon assessment of population trends. However, in the context of the level of survey effort in eastern Victoria, the frequency of detection of H. australiacus suggests that the species was rare in the south before Europeans began collecting information on frog fauna. The increase in reports of the species over the past 10 years most likely reflects the increased survey effort and general awareness within the region rather than population trends. If population declines have occurred in the south in more recent times, they are unlikely to be detected. FIGURE 9: Distribution of the Stuttering Frog Mixophyes balbus. Open circles indicate records collected pre-1990; closed circles A range of threatening processes operate across the range of are those recorded 1990 onwards. H. australiacus.These include timber harvesting, cattle grazing, fuel reduction burning, introduced terrestrial and aquatic predators, and various disturbances resulting from urbanisation (Gillespie 1990; Gillespie 1997b unpubl.; Recsei 1997).The potential impacts of these processes have not been examined.

Information is lacking on the demography of the species and on the size of populations.The ecological requirements of H. australiacus are poorly known. Basic information has been collated on broad habitat associations and breeding biology ● 1990 onwards record ❍ (Gillespie 1990; Daly 1996; Recsei 1997; Gillespie 1997b unpubl.). pre-1990 record ■ capital cities Tusked Frog (Adelotus brevis) Adelotus brevis was previously widespread along the coast and ranges, predominantly on the eastern slopes of the Great Divide, from near Nowra, south of Sydney (AM records), to Eungella National Park in mid-eastern Queensland (Covacevich and McDonald 1993).The species occurs in a wide range of habitats, including rainforest, wet sclerophyll forest, dry forest communities, open swamps, cleared pasture and urban areas (Gillespie and Hines pers. obs.). Adelotus brevis breeds in temporary and permanent ponds, and flowing streams (Gillespie and Hines pers. obs.).The distribution of A. brevis is within the temperate and adjoining subtropical zone (Figure 11).

120 Within the temperate zone A. brevis appears to have declined FIGURE 10: Distribution of the Giant Burrowing Frog Heleioporus australiacus. Open circles indicate records collected from the Northern Tablelands of NSW.There are no recent pre-1990; closed circles are those recorded 1990 onwards. records of A. brevis in this region (Figure 11).The eastern section of the Northern Tablelands has been relatively well surveyed in recent years by the North-east Forests Biodiversity Study (NSW NPWS 1994), the Regional Forests Assessment Program and others (Hines pers. obs.; Mahony, ● 1990 onwards record pers. comm.).Targeted surveys for this species throughout its ❍ pre-1990 record known range in the temperate zone are required to more ■ capital cities thoroughly assess its status. Apparent declines of A. brevis have also been documented for high-elevation sites at Eungella in mid-eastern Queensland (McDonald and Alford 1999; Ingram and McDonald 1993), and from streams along the Great Dividing Range in south-eastern Queensland (Hines et al. 1999).

Several threatening process operate in the Northern Tablelands, which may have contributed to the decline of this species. Much of the region has been cleared or grossly modified for pastoralism. Introduced fish (cyprinids, salmonids and G. holbrooki) are also widespread, but their impact on this species is not known (but see Gillespie and Hero 1999).The species persists in heavily disturbed areas at lower elevations (e.g., in urban areas of Brisbane), but the viability of these populations is unknown.

Within the temperate zone the biology of A. brevis is poorly known. Elsewhere basic information has been collated on broad habitat associations and breeding biology (Daly 1995; Katsikaros and Shine 1997). Specific ecological requirements of A. brevis, and factors limiting distribution and abundance are unknown. Population structure and dynamics are also not known.

FIGURE 11: Distribution of the Tusked Frog Adelotus brevis. Open circles indicate records collected pre-1990; closed circles Eastern Banjo Frog (Limnodynastes are those recorded 1990 onwards. dumerilii dumerilii) Limnodynastes dumerilii dumerilii is common and widespread across south-eastern South Australia, from the northern plains of Victoria, along the western slopes of the Great Divide, on the Northern Tablelands of NSW, and along the Great Divide in the far south of Queensland (Martin 1972; SM, AM, NSW and VIC records).This is a cosmopolitan sub- species, occurring in virtually all types of habitats within its range, including grassland, woodland, dry and wet sclerophyll forests, montane and sub-alpine communities (Martin 1972; Barker et al. 1995; Gillespie and Hines pers. obs.). Limnodynastes dumerilii dumerilii also occurs in highly modified or disturbed habitats, such as cleared farmland and urban ● 1990 onwards record environments (NSW and VIC Atlas records).The sub-species ❍ pre-1990 record breeds in both ephemeral and permanent ponds, both ■ capital cities natural and artificial, along with permanent and ephemeral streams (Martin 1972).

A map of distribution of records has not been provided for this sub-species in this review, due to the difficulty of reliably differentiating subspecies within the complex in wildlife atlas records and museum specimens. It has been frequently recorded throughout its range in recent years, and there have been no reports of population declines. However, no detailed assessments have been made of its current status in many areas. An extensive survey of the northern plains of Victoria in 1993-1994 found this species to be one of the most abundant frogs in the region (Brown unpubl. data;VIC Atlas).

121 Odendaal and Bull (1982). It is therefore possible to assign FIGURE 12: Distribution of the Stream-bank Froglet Crinia riparia. Open circles indicate records collected pre-1990; closed most records to a species based on their geographic location circles are those recorded 1990 onwards. (Figure 12). Further surveys are required to confirm these records, but they suggest that C. riparia remains widespread throughout its formerly-known range.

There is no information on specific threats to C. riparia, although a number of potential threats operate across its ● 1990 onwards record range. Given its specialised lotic larval stage it is likely to be ❍ pre-1990 record susceptible to the types of threats causing declines in other ■ capital cities riverine frog species.

The ecology of C. riparia has been examined in some detail. Aspects of habitat use, life history, larval ecology and ecological interactions with other species, and factors limiting distribution have been investigated (Odendaal and Bull 1980, 1982, 1983; Odendaal et al. 1982, 1984, 1986). However, knowledge of non-breeding habitat requirements and population structure and dynamics is lacking.

ADEQUACY OF CURRENT KNOWLEDGE FOR ASSESSMENT OF NATURE OF POPULATION DECLINES In summary, one species of temperate riverine frog has disappeared, and three have clearly suffered substantial population declines across their ranges (Table 2). A further five species have suffered declines of some populations. Current knowledge of distribution and abundance is inadequate for proper assessments of the status of seven of the species reported to have declined, and declines may have been more extensive. In addition, the taxonomic status of at least four species groups/complexes is yet to be fully resolved.

The sub-species was found to be widespread in north-east Knowledge of Species Distributions NSW (NSW NPWS unpubl. data), and remains common on Over 90 % of records for temperate riverine species have the Northern Tablelands (Hines pers. obs.). A limited been collected since 1960 (all museum and wildlife atlas herpetofauna survey of the western slopes of the Great records combined). Prior to 1960 the distributions of most Divide in southern NSW in 1994-1995 found the sub-species species were very poorly known.The assessment of historical to be relatively common (Lemckert 1998). Populations are distribution is therefore based primarily on records collected currently known from the Main Range in south-eastern between 1960 and 1990. For most species the general Queensland (Hines unpubl. data). distribution and broad habitat associations have been identified. Generally the data are inadequate for modelling specific habitat Stream-bank Froglet (Crinia riparia) associations or patterns of decline. Some species have been This species is restricted to streams draining the Flinders surveyed extensively across their ranges; but for most, detailed Ranges in South Australia, from 400 to over 1000 m asl surveys have been conducted in only parts of their ranges. (Figure 12).The species typically occurs along swift-flowing With the exception of L. spenceri, detailed information on the rocky streams in this region (Odendaal and Bull 1982, 1983; number and sizes of remaining populations of each declining Odendaal et al. 1982). It is the only species of Crinia with a species are not known. Species such as L. citropa, L. phyllochroa, specialised lotic reproductive mode (Odendaal et al. 1982; L. subglandulosa and H. australiacus lack adequate base-line Odendaal and Bull 1983). survey data across their ranges to determine whether or not general population declines have occurred. Most surveys have The distribution of C. riparia was surveyed extensively from been non-systematic or at too low an intensity to provide 1977 to 1979 by Odendaal and Bull (1982).The species was more than presence-only data. Surveys of presence or absence found to be common along numerous streams within the of a species, or its relative abundance have been undertaken region.There have been no formal surveys conducted for this for only a few species and usually at only a few sites.This species in recent years. However, during water-sampling information deficit limits assessments of magnitude of declines, surveys conducted in the Flinders Ranges over the past four or analyses of patterns of population declines. years, the South Australian Environmental Protection Authority documented frog species (SA EPA unpubl. data). Comprehensive surveys ascertaining distribution and relative These records do not distinguish between C. riparia and abundance have only been completed for one species, C. signifera. However, these species are largely allopatric with a L. spenceri (Gillespie and Hollis 1996; Hunter and Gillespie in narrow zone of sympatry, which was surveyed in detail by press; Gillespie 1998 unpubl.).This has enabled analyses of

122 TABLE 2: Summary of population declines of temperate riverine frog species. See text for further explanations.

Species Disappeared Major decline Major regional Some local Some Local No evidence across most of decline; otherwise population population of declines species’ range persistent declines; declines, throughout status otherwise remainder of otherwise persistent and range indeterminate widespread Litoria piperata *

L. booroolongensis * L. spenceri * Mixophyes balbus * Adelotus brevis * L. subglandulosa * Heleioporus australiacus * L. lesueuri * L. phyllochroa complex * L. citropa * Limnodynastes dumerilii * Crinia riparia *

patterns of distribution, which has shed light on causes of population change have been identified, and the on-going status decline. In addition, this has provided information on of each population is known.This has also enabled quick population size and distribution, which has then been used to detection of new population declines (Gillespie 1997a unpubl.). target management actions for protection of the species (Robertson and Gillespie in review). Monitoring programs have also recently commenced in NSW and south-eastern Queensland on some other declining Population Monitoring species (e.g., Mixophyes species), but these programs need to be expanded to incorporate multiple sites across the Population monitoring is essential to examining patterns of distribution of each species so that regional patterns and change in populations, and when ascertaining whether variation in population trends can be observed. Other declining changes are natural fluctuations or unusual declines and “non-declining” riverine species also need to be (Pechmann et al. 1991; Pechmann and Wilbur 1994).The systematically monitored. In view of the number of species value of monitoring programs has been demonstrated in within the guild which have already suffered declines, it cannot several studies (Osborne 1989; Richards et al. 1993; Gillespie be assumed that those species currently considered secure will 1997a unpubl.; Lips 1998; Osborne et al. 1999). Monitoring not suffer declines in the future. Further to this, comparisons of may provide information not only on magnitude and time of the population dynamics and demography of declining and declines, but also on environmental factors influencing non-declining species may be informative of causes of decline. population changes. Repeated sampling also provides information on the biology of species, such as movement and Adequacy of Knowledge of Ecology of seasonal activity patterns.This information is also necessary Temperate Riverine Species for survey design, interpretation of survey results and other field observations. For most temperate lotic species, no The basic biology of most riverine species in south-eastern systematic population monitoring has occurred, or has only Australia, such as life history, general reproductive phenology recently commenced, despite awareness of declines of a and broad habitat associations of larvae and adults, is known. number of species for many years (see Tyler 1997). Only a However, information on the detailed ecology and few populations of some species have been visited repeatedly demography of most species is inadequate for interpreting or and over broad time scales (10–20 years) (Mahony 1993; determining causes of declines (Table 3). For some species, Anstis and Littlejohn 1996).These studies provide broad such as L. booroolongensis, L. subglandulosa and H. australiacus, comparisons of historic and current status at sites, but this information is relatively limited, or even non-existent in sampling has generally been too infrequent to enable detailed the case of L. piperata. Factors which limit distribution and assessment of the causal factors. abundance are unknown for most species.

A population-monitoring program for L. spenceri commenced in Habitat associations and larval ecology of C. riparia have been 1990. Intensive monitoring with mark-recapture studies have examined in some detail (see Odendaal and Bull 1980, 1982, been conducted on two populations since 1992. Annual low 1983; Odendaal et al. 1982, 1984, 1986). Studies of habitat intensity broad-scale monitoring has been conducted across use and larval ecology of M. balbus are under way (Mahony most populations since 1994. In 1998 monitoring was expanded and Knowles unpubl. data). Only one declining species, to include all extant populations. Populations of L. lesueuri and L. L. spenceri, has been extensively studied. Some information phyllochroa are also monitored at some of these sites.This on life history, habitat use and larval ecology of program has provided detailed information on the behaviour L. booroolongensis, L. lesueuri, L. citropa, and L. phyllochroa has and dynamics of individual populations, and on general patterns also been collected through this work. Knowledge of of population behaviour. Extraneous factors influencing L. spenceri should provide a useful model for aspects of the

123 ecology and population dynamics other species within the Most of the species that have not suffered major declines and L. citropa group, and a basis for developing some hypotheses remain common are either facultative stream breeders, or are to address population declines. known to venture considerable distances from riparian habitats.These species tend to have larger geographic ranges DIFFERENCES BETWEEN DECLINING and all occur at low altitudes down to sea level.With the AND NON-DECLINING SPECIES exception of declines of tableland populations of A. brevis, there is little evidence of decline of those species which Reported declines in temperate riverine species are not appear to tolerate gross habitat modification. associated with any particular phylogenetic assemblage. Species from two families and different species groups have POTENTIAL CAUSES OF DECLINES declined. Similarly, no clear geographic pattern is evident in declines. Populations of various species have declined on both The distributions of most species within south-eastern sides of the Great Dividing Range, from the Central Australia have undoubtedly been drastically affected by Highlands of Victoria to the Queensland Border. extensive clearing of the lowlands and tablelands since European settlement. For many species which are mostly Most population declines have been at altitudes above 500 m restricted to forested environments, these changes asl.The only species to have disappeared, L. piperata,is undoubtedly removed considerable areas of suitable habitat. restricted to this altitude range. Some species have also However, declines have occurred in catchments with minimal declined at low altitude sites, e.g., M. balbus (Mahony 1993; gross disturbance or alteration.The causes of these declines Holloway and Osborne 1996 unpubl.).The decline of are more subtle and difficult to resolve. A range of potential L. spenceri was found not to be associated with altitude threatening processes has been identified (Table 4). (Gillespie and Hollis 1996). Introduced Fish Declines are not restricted to one broad ecological guild of species, although obligate lotic species appear to be affected Several species of introduced fish occur in streams within the more than facultative lotic species.The following conclusions region (see Gillespie and Hero 1999). Species of trout are can be made about the four species which have clearly able to exert significant predation pressure upon larvae of suffered major declines (Table 2). L. spenceri, L. citropa and L. nudidigitus, and have been strongly 1. They are all obligate stream-breeders, although implicated in the decline of L. spenceri (Gillespie 1997a L. booroolongensis is known to reproduce in isolated unpubl.; Gillespie in review).Trout are also present in several streamside pools as well as in the flowing stream upland streams where L. booroolongensis, L. piperata, (Gillespie pers. obs.). L. subglandulosa and M. balbus, have declined or disappeared. In view of the vulnerability to trout predation of other 2. They are species which are mostly restricted to riparian members of the L. citropa group, these fish are likely to have habitats, rather than dispersing widely into surrounding had a significant impact on L. piperata and L. subglandulosa, terrestrial habitats. and may also have affected populations of other species. 3. The species are mostly restricted to forested or naturally vegetated areas, and have been displaced from cleared or grossly disturbed areas. A notable exception to this is L. booroolongensis, which persists along streams through cleared land. However, the viability of these populations is unknown.

TABLE 3: Summary of current knowledge base for assessing population declines of temperate riverine frog species.

Species Systematic assessment Systematic monitoring Knowledge of ecology of distribution and abundance of populations and assessment of potential threatening processes

Completed Limited None All Some None Advanced Limited None Litoria piperata P** L. booroolongensis A** L. spenceri AA* L. subglandulosa A** L. lesueuri AA* L. phyllochroa complex AP* L. citropa AP* Mixophyes balbus PA* Heleioporus australiacus P** Limnodynastes dumerilii P** Crinia riparia ** Adelotus brevis P**

A data available on population size or relative abundance P presence/absence data only • existent

124 TABLE 4: Summary of threatening processes likely (L) or potentially (P) contributing to population declines of temperate riverine frogs.

Species Habitat Forestry Stock Introduced Introduced Pesticides Increased Disease* Destruction Practices Grazing Fish Mammals pollutants UV*

L. piperata LLLLPP L. booroolongensis LPPL LPP L. spenceri LLPL PP M. balbus LLPPPPPP L. subglandulosa LPLL LPP H. australiacus LL PPPPP L. lesueuri LPP L. phyllochroa complex LPP L. citropa LPPP Limnodynastes dumerilii P PPP Crinia riparia PPP Adelotus brevis L PP PPP

* Due to the ‘non-localised’ nature of ultra-violet radiation or disease, no species can be considered secure from these potential threats at this stage.

Introduced Mammals by limiting sheltering sites and food availability. Experiments conducted with tadpoles of L. spenceri found that deposited Foxes and cats are common and widespread throughout fine sediment can retard development rates of tadpoles, thus south-eastern Australia and are potentially a major reducing their chances of survival (Gillespie 1997a unpubl.). threatening process to terrestrial frog species such as H. australiacus and M. balbus. Frog bones have been detected Agriculture and Grazing in fox scats (Triggs, Dead Finish,Victoria, pers. com.). Specimens of adult Limnodynastes dumerilii have been found Land clearance for agriculture has adversely impacted upon in the stomachs of foxes (Marks, Department of Natural the headwaters of some catchments in which declines have Resources and Environment,Victoria, pers. com.).The impact occurred, especially on the tablelands. Apart from direct of foxes and cats on frog populations is unknown and effects of habitat removal, other down-stream impacts are warrants investigation. likely, such as and fertiliser run-off, increased nutrient loads from grazing stock, and increased sediment loads from Forestry Activities erosion. Forest grazing is also likely to have significant effects upon some populations through destruction or modification Timber harvesting and associated forest management of riparian vegetation, increased nutrient levels, trampling of practices have occurred with varying intensities in many stream edges, and increased erosion damaging oviposition catchments throughout the ranges of both declining and non- sites and larval habitats (Parris and Norton 1997; Knowles et declining species in south-eastern Australia.Timber harvesting, al. 1998 unpubl.). associated road construction, and burning practices can significantly affect stream water temperature, sediment loads, Hydrological Changes turbidity, nutrient levels and flow regimes (Boughton 1970; Langford and O’Shaughnessy 1980; Flinn et al. 1983; Clinnick Flow regimes of some catchments have been modified for 1985; Cornish and Binns 1987; Campbell and Doeg 1989). irrigation or hydro-electric power generation. In the southern Growth and development of tadpoles are directly affected by highlands, Hunter and Gillespie (1999) found that L. lesueuri these factors (reviewed by Duellman and Trueb 1994). In and other lotic frog species were encountered less frequently contrast to species which are able to opportunistically breed in streams below impoundments or aqueducts affecting in standing temporary water bodies, in which fluctuations in stream flow. Increased stream flows during the warmer temperature, nutrient and oxygen levels may be extreme, the months are likely to have severe adverse impacts upon larvae of species that rely on streams are likely to be less temperate riverine frog populations. Significant rises in water tolerant of such changes because the normal stream level and velocity during this period are likely to flush eggs environment is more stable. and larvae downstream. Reduced temperatures of sub- surface releases of water from dams during the summer Recent investigations by O’Shaughnessy (1995 unpubl.) found months (SMEC 1997) are likely to inhibit larval growth and that forest roads were the main sources of sediment input development.These reduced temperatures may also favour into streams in eastern Victoria. Increased sediment loads may introduced trout (which may result in increased predation reduce the availability or quality of oviposition sites through pressure). Sub-surface waters may also be anoxic and have filling of interstitial spaces in the stream bed and blanketing different pH and higher concentrations of activated heavy substrates, resulting in increased mortality of eggs from metals (Doeg 1987; Ligon et al. 1995; Erskine 1996). Such predation, desiccation or flooding.This may be particularly alterations to water chemistry may be detrimental to important for species, such as L. booroolongensis and tadpoles. Reduced peak flows resulting from dams may allow L. spenceri, which oviposit in rock crevices (Gillespie pers. build up of sediments and colonisation of the stream channel obs.), or M. balbus which oviposit in the stream bed (Knowles by vegetation. Increased entrained sediments will reduce et al. 1998 unpubl.). Larval survival may similarly be reduced availability of oviposition sites and refugia for larvae by

125 blanketing the stream bed and in-filling of crevices between FUTURE DIRECTIONS rocks. Encroachment of vegetation will reduce basking sites A more strategic approach is required for assessing the for adult frogs, and shading lower stream temperatures, which nature and causes of population declines within this guild may reduce larval growth. throughout the region, rather than examining single species Climate Change separately.The current approach to recovery planning is likely to result in duplication of research, adoption of inconsistent Climatic change and increased levels of ultra-violet radiation methods, and lack of coordination between separate have been suggested as causative agents in the declines of programs which otherwise have similar objectives. frog populations (Blaustein et al. 1995; Broomhall 1997). It is possible that upland populations of some species may be Systematic surveys are required to ascertain the magnitude more susceptible to UV than lowland populations, which may and nature of observed declines of most species. Surveys explain some of the upland population declines in south- need appropriate sampling intensity, methods, scale and eastern Australian temperate riverine species.The effects of stratification. In conjunction with these surveys, information increased UV might be expected to be most pronounced in should be gathered on abundance of introduced fish, levels of upland southern populations of species which bask, such as habitat disturbance, both to streams and catchments, and L. spenceri and L. lesueuri (Gillespie and Hollis 1996). However, other potentially threatening processes.This data should be the current distributions and abundances of these species do quantifiable and suitable for appropriate multivariate analyses not support this. of distribution and abundance. Patterns of current distribution should also be examined at a landscape scale. Disease In conjunction with these surveys, appropriate monitoring sites Trenerry et al. (1994) and Laurance et al. (1996) suggested should be established to assess population dynamics and to that a pathogen may be responsible for some frog declines in detect future changes in abundance. Strategic monitoring Australia. Criticism of this hypothesis was largely based on should include ‘non-declining’ species, as there is no guarantee their inability to isolate and identify the pathogen (Hero and that these species are secure in the medium term. Monitoring Gillespie 1997; Alford and Richards 1997). Ongoing studies of sites should be stratified throughout the region to incorporate ill and dead frogs, including some collected during dramatic environmental variation across the ranges of all species. declines in north Queensland and Central America, now provide strong evidence that a chytrid fungus is the proximal Studies are required to examine comparative ecology and cause of death (Berger et al. 1998, Berger et al. 1999). habitat requirements of species and to ascertain factors Several species of temperate riverine frog have been found which limit distribution and influence abundance. Information infected with the chytrid fungus (Berger et al. 1998, Berger et is required on aquatic and terrestrial habitat use, population al. 1999). Some populations of frogs infected with this fungus structure and demography, and factors which influence have declined while others have not. It is not known if the recruitment and mortality. Both declining and non-declining chytrid fungus is responsible for these declines, or is an species should be examined as differences between these indication of other environmental stress.The fungus may be a taxa may be informative on causes of decline. normally innocuous opportunistic natural pathogen, which is Processes potentially responsible for observed declines able to kill frogs under certain environmental conditions. require specific examination, such as the impact of introduced Alternatively, this fungus may be a novel pathogen in aquatic and terrestrial predators, impacts of various stream populations of Australian frogs. disturbances, disease and climate change. Clearly there is a range of potentially threatening processes operating with varying impact throughout streams and across ACKNOWLEDGMENTS catchments of temperate south-eastern Australia.The We thank the Arthur Rylah Institute, Department of Natural interactions of these changes and disturbances within streams Resources and Environment,Victoria, the Queensland Parks and catchments are likely to be complex. Assessment of these and Wildlife Service, NSW National Parks and Wildlife impacts upon frog populations is likely to be further Service, and Environment Australia for financial support. complicated by interactions with other stochastic Museum records and other distribution data were gladly environmental processes.The persistence of populations in supplied by the Australian Museum, the Queensland Museum, streams affected by timber harvesting or some other the Museum of Victoria, the South Australian Museum, the disturbance process does not mean that they have not been Australian National Wildlife Collection, the South Australian significantly adversely impacted upon. Many of these processes Environmental Protection Agency, the NSW Wildlife Atlas, may serve only to reduce the ability of populations to cope and the Victorian Wildlife Atlas. R. Brown and H. Preece, with other catastrophes or environmental loads.The medium- Queensland Parks and Wildlife Service, provided geographical term and long-term influences of past and current information system support.This manuscript was greatly management practices on the viability of riverine frog improved by discussions with M. Mahony, M. Anstis, F. populations in south-eastern Australia cannot be dismissed. An Lemckert, D. Ayers and J. Recsei. R. Loyn provided comments applied research program is required, targeting a range of on the manuscript.The manuscript was greatly improved by threatening processes, so that their importance can be judged. referee reports from M. Littlejohn and W. Osborne, and comments from N. Clemann.

126 REFERENCES Clinnick, P.F. (1985) Buffer strip management in forested areas: a review. Aust. For., 48: 34-45. Alford, R. A. and Richards, S. J. (1997) Lack of evidence for epidemic disease as an agent in the catastrophic decline of CNR (1993) Statement of Resources, Uses and Values, East Australian rainforest frogs. Cons. Biol., 11: 1026-29. Gippsland Forest Management Area. Department of Conservation and Natural Resources, Orbost,Victoria. Anstis, M. (1997) The glandular frog Litoria subglandulosa. Chapter 25 In H. Ehmann (Ed.).Threatened Frogs of New Copland, S. J. (1957) Presidential address. Australian tree frogs South Wales: Habitats, Status and Conservation. Frog and of the genus Hyla. Proc. Linn. Soc. N.S.W., 82: 9-108. Tadpole Study Group of NSW, Sydney. pp. 213-221. Copland, S. J. (1962) Hyla phyllochrous Gunther (Amphibia) as Anstis, M. and Littlejohn, M. J. (1996) The breeding biology of an addition to the fauna of Victoria, with the description of Litoria subglandulosa and L. citropa (Anura: Hylidae), and a a new race and a note on the name of the genus. Proc. re-evaluation of their geographic distribution.Trans. Roy. Linn. Soc. N.S.W., 87: 137-40. Soc. S. A., 120: 83-99. Cornish, P.M. and Binns, D. (1987) Streamwater quality Anstis, M., Alford, R. A. and Gillespie, G. R. (1998) Breeding following logging and wildfire in a dry sclerophyll forest in biology of Litoria booroolongensis Moore and L. lesueuri south eastern Australia. For. Ecol. Manage., 22, 1-28. Dumeril and Bibron (Anura: Hylidae).Trans Roy. Soc. S. A., Covacevich, J.A. and McDonald, K.R. (1993) Distribution and 122:33-43. conservation of frogs and reptiles of Queensland Barker, J. and Grigg, G. (1977) A Field Guide to Australian rainforests. Mem. Qld. Mus., 34: 189-199. Frogs. Surrey Beatty, Chipping Norton, NSW. Czechura, G.V. and Ingram, G. S. (1990) Taudactylus diurnus Barker, J., Grigg, G. and Tyler, M. J. (1995) A Field Guide to and the case of the disappearing frogs. Mem. Qld. Mus., Australian Frogs (2nd Ed.). Surrey Beatty, Chipping 29: 361-365. Norton, NSW. Daly, G. (1995) Observations on the Tusked Frog Adelotus Berger, L., Speare, R., Daszak, P., Earl Green, D., Cunningham, brevis (Anura: Myobatrachidae). Herpetofauna, 25: 32-35. A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., Daly, G. (1996) Observations on the eastern frog McDonald, K. R., Hines, H. B., Lips, K. R., Marrantelli, G. and Heleioporus australiacus (Anura: Myobatrachidae) in Parkes, H. (1998) Chytridiomycosis causes amphibian southeastern New South Wales. Herpetofauna, 26: 33-42. mortality associated with population declines in the Daly, G. (1998). Review of the status and assessment of the rainforests of Australia and Central America. Proc. Nat. habitat of the stuttering frog Mixophyes balbus (Anura: Acad. Sci., 95: 9031-9036. Myobatrachidae) on the south coast of New South Wales. Berger, L., Speare, R., and Hyatt, A.D. (1999) Chytrid fungi and Herpetofauna, 28: 2-11. amphibian declines: overview, implications and future Doeg,T. J. (1987) Response of the macroinvertebrate fauna of directions. Pp 23-33 in Declines and Disappearances of the Mitta Mitta River,Victoria, to the construction and Australian Frogs ed by A.Campbell. Environment Australia: operation of the Dartmouth Dam. 2. Irrigation release. Canberra. Occ. Pap. Mus.Vic., 1: 101-128. Blaustein, A. R., Edmond, B., Kiesecker, J. M., Beattie, J. J. and Duellman and Trueb (1994) Biology of Amphibians. (2nd Ed.). Hokit, D. G. (1995) Ambient ultraviolet radiation causes John Hopkins University Press: Maryland, USA mortality in eggs. Ecol. Appl., 5: 740-743. Ehmann, H., (1997) (ed.) Threatened Frogs of New South Boughton,W. C. (1970) Effects of Land Management on Wales: Habitats, Status and Conservation. Frog and Quantity and Quality of Available Water. A Review.Water Tadpole Study Group of NSW Inc., Sydney South. Resources Laboratories Report No. 120, University of NSW, Sydney. Erskine,W. D. (1996) Downstream hydrogeomorphic impacts of the Eildon Reservior on the mid-Goulburn River, Bridgewater, P.B. (1987) The present Australian Environment Victoria. Proc. Roy. Soc.Vic., 108: 1-15. — Terrestrial and Freshwater. In D.W.Walton and G. R. Dyne (Eds.). Volume 1A General Flinn, D.W., Farrell, P.W., Stuart, H.T. L., Leitch, C. J. and Articles. Burea of Flora and Fauna, Australian Government, Hopmans, P.(1983) The effects of fire in eucalypt forest on pp. 69-100. soils, nutrient cycling, tree growth and catchment hydrology: a review with particular reference to fuel Broomhall, S. D. (1997) Comparative effects of ultra violet -B reduction burning. In Proceedings of the Symposium: (UV-B) radiation on two sympatric species of Australian Fighting Fire with Fire. Graduate School of Environmental anurans, Litoria verreauxii alpina and Crinia signifera. Studies, Monash University, Melbourne, pp. 146-85. Unpublished B. Sc. Honours thesis. Applied Ecology Research Group, University of Canberra. Gillespie, G. R. (1990) Distribution, habitat and conservation status of the giant burrowing frog, Heleioporus australiacus Bureau of Meteorology (1975) Climatic Atlas of Australia. (Anura: Myobatrachidae), in Victoria.Vict. Nat., 107: Map Set 5. Rainfall. Australian Government Publishing 144-153. Service, Canberra. Gillespie, G. R. (1997a, unpublished) The Biology of the Bureau of Meteorology (1989) Climate of Australia. Australian Spotted Tree Frog (Litoria spenceri) and Examination of Government Publishing Service, Canberra. Factors Responsible for Population Declines. Arthur Rylah Campbell, I. C. and Doeg, I. J. (1989) The impact of timber Institute, Department of Natural Resources and production and harvesting on streams: a review. Aust. J. Environment, Heidelberg,Victoria. Unpublished Report to Mar. Freshw. Res., 40: 519-539. the Biodiversity Group, Environment Australia, Canberra.

127 Gillespie, G. R. (1997b, unpublished) Survey Design and Hutchinson, M. N. and Maxson, L. R. (1987) Phylogenetic Management Prescriptions for the Giant Burrowing Frog relationships among Australian tree frogs (Anura: Hylidae: (Heleioporus australiacus) and the Stuttering Frog ): an immunological approach. Aust. J. Zool., (Mixophyes balbus). Arthur Rylah Institute, Department of 35: 61-74. Natural Resources and Environment, Heidelberg,Victoria. Ingram, G. J and McDonald, K. R. (1993) An update on the Unpublished report to the New South Wales National decline of Queensland’s frogs, In D. Lunney and D. Ayers Parks and Wildlife Service, Queanbeyan. (Eds.) Herpetofauna in Australia a Diverse Discipline.Trans. Gillespie, G. R. (1998, unpublished) Spotted Tree Frog Roy. Zool. Soc. NSW, Mosman, pp. 297-303. Recovery Program — Annual Report to April 1998. Katsikaros, K. and Shine, R. (1997) Sexual dimorphism in the Unpublished report to the Biodiversity Group, tusked frog, Adelotus brevis (Anura, Myobatrachidae) — Environment Australia, Canberra. the roles of natural and sexual selection. Biol. J. Linn. Soc., Gillespie, G. R. (1999) The booroolong frog Litoria 60: 39-51. booroolongensis Moore (Anura: Hylidae): an addition to the Knowles, R., Hines, H.B.,Thumm, K. Mahony, M., and frog fauna of Victoria.Vict. Nat., 116: 112-114. Cunningham, M. (1998 unpublished) Oviposition of the Gillespie, G. R. (in review) The role of the introduced trout in barred-frogs (Mixophyes species) in southeastern Australia the decline of the spotted tree frog Litoria spenceri. with implications for management. Unpublished abstract of Cons. Biol. a talk presented to the Australian Society of Gillespie, G.R. and Hero, J.-M. (1999) Potential Impacts of Herpetologists meeting, February 1998. Introduced Fish and Fish Translocations on Australian Langford, K. J. and O’Shaughnessy, P.J. (Eds.) (1980) Water Amphibians. Pp 131-144in Declines and Disappearances of Supply Catchment Research. Second Progress Report, Australian Frogs ed by A.Campbell. Environment Australia: Coranderrk. M.M.B.W. Report W0010. Melbourne and Canberra. Metropolitan Board of Works, Melbourne. Gillespie, G. R. and Hollis, G. J. (1996) Distribution and habitat Laurance,W. F. K., McDonald, K. R. and Speare, R. (1996) of the spotted tree frog Litoria spenceri Dubois (Anura: Epidemic disease and the catastrophic decline of Australian Hylidae), and an assessment of potential causes of rain forest frogs. Cons. Biol., 10: 406-13. population declines.Wildl. Res., 23: 49 — 75. LCC (1985) Report on the East Gippsland Area Review. Land Gillespie, G. R. and Osborne,W. S. (1994) Update on the Conservation Council, Melbourne. status of the spotted tree frog (Litoria spenceri) in the Lee, A. K. (1967) Taxonomy, ecology and evolution of the Australian Capital Territory.Vict. Nat., 111: 182-183. genus Heleioporus (Gray) (Leptodactylidae). Aust. J. Zool., Heatwole, H., De Bavay, J.,Webber, P.and Webb, G. (1995) 15: 367-439. Faunal survey of New England. IV.The frogs. Mem. Qld. Lemckert, F. (1998) A survey for threatened herpetofauna of Mus., 38: 229-49. the south-west slopes of New South Wales. Aust. Zool., Hero, J.-M. and Gillespie, G. R. (1993) The tadpole of Litoria 30: 492-500. phyllochroa (Anura, Hylidae). Proc. Roy. Soc.Vic., 105: 31-38. Lemckert, F., Potter, M., Smith, B. and Brues,T. (1997) Recent Hero, J.-M. and Gillespie, G. R. (1997) Epidemic disease and records of the southern barred frog (Mixophyes balbus) amphibian declines in Australia. Cons. Biol., 11: 1023-25. from the south coast of NSW. Herpetofauna, 27: 43-45. Hines, H.B., Mahony, M.J., and McDonald, K.R. (1999) An Ligon, F. K., Dietrich,W. E. and Trush,W. J. (1995) Downstream assessment of frog declines in wet subtropical Australia. Pp ecological effects of dams. Bioscience, 45: 183-192. 44-63 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Littlejohn, M. J. (1967) Patterns of zoogeography and speciation in south-eastern Australia Amphibia. Ch. 6 In A. Hollis G. J. (1995) Reassessment of the distribution, H.Weatherley (Ed.) Australian Inland Waters and their abundance and habitat of the baw baw frog Philoria frosti Fauna. Australian National; University Press, Canberra, pp. Spencer: Preliminary Findings.Vict. Nat., 112: 190-201. 150-174. Holloway, S. E. (1997) Survey protocols for the stream- Littlejohn, M. J. (1981) The amphibia of mesic southern breeding frogs of far East Gippsland: the implication of habitat modelling and an assessment of techniques. Australia: a zoological perspective. In A. Keast (Ed.) Unpublished M. Sc.Thesis, University of Canberra, Ecological Biogeography of Australia. Dr W. Junk,The Canberra. Hague, pp. 1304-1330. Holloway, S. E. and Osborne,W. S. (1996 unpublished) The Littlejohn, M. J., Loftus-Hills, J. J., Martin, A. A. and Watson, G. F. Stream-breeding Frog Fauna Associated with Montane (1972) Amphibian fauna of Victoria. Confirmation of Forests in Far East Gippsland. Unpublished Report to the records of Litoria (= Hyla) citropa (Tschudi) in Gippsland. Department of Natural Resources and Environment, Vict. Nat., 89: 51-54. Orbost,Victoria. Littlejohn, M. J. and Martin, A. A. (1967) The rediscovery of Hughes, J. M. R. and James, B. (1989) A hydrological Heleioporus australiacus (Shaw) (Anura: Leptodactylidae) in regionalization of streams in Victoria, Australia, with eastern Victoria. Proc. Roy. Soc.Vic., 80: 31-36. implications for stream ecology. Aust. J. Mar. Freshwater Littlejohn, M. J., Davies, M., Roberts, J. D. and Watson,W. F. Res., 40: 303-326. (1993) The Myobatrachidae. In C. J. Glasby, G. J. B. Ross and Hunter, D. and Gillespie, G. R. (1999) The distribution, P.L. Beesley (Eds.). Fauna of Australia.Vol. 2A Amphibia & abundance and conservation status of riverine frogs in Reptilia. Australian Government Publishing Service, Kosciuszko National Park. Aust. Zool., 31: 198-209 Canberra, pp. 41-57.

128 Lips, K. R. (1998) Decline of a tropical montane amphibian Odendaal, F. J. and Bull, C. M. (1982) A parapatric boundary fauna. Cons. Biol., 12: 106-116. between Ranidella signifera and R. riparia (Anura: Lunney, D. and Barker, J. (1986) Survey of the reptiles and Leptodactylidae) in South Australia. Aust. J. Zool., 30: 49-57. amphibians of the coastal forests near Bega, New South Odendaal, F. J. and Bull, C. M. (1983) Water movements, Wales. Aust. Zool., 22: 1-9. tadpole competition and limits to the distribution of Ranidella riparia R. signifera Mahony, M. J. (1993) The status of frogs in the Watagan the frogs and . Oecologia, 57: 361-67. Mountains area, the central coast of New South Wales. In D. Lunney and D. Ayers (Eds.) Herpetology in Australia a Odendaal, F. J., Bull, C. M. and Nias, R. C. (1982) Habitat Diverse Discipline.Trans. Roy. Soc. NSW, Mossman, pp. selection in tadpoles of Ranidella signifera and R. riparia 257-264. (Anura: Leptodactylidae). Oecologia, 52: 411-14. Mahony, M. J. (1996, unpublished) Survey of the Distribution Odendaal, F. J., Bull, C. M. and Richards, S. J. (1984) Interactions and Abundance of Declining Frog Species in Northern during feeding between tadpoles of Ranidella signifera and New South Wales. Unpublished report for the New South R. riparia. J. Herp., 18: 489-92. Wales National Parks and Wildlife Service, Hurstville. Odendaal, F. J., Bull, C. M. and Telford, S. R. (1986) Influence of acoustic environment on the distribution of the frog Mahony, M., Knowles, R. and Pattinson, L. (1997a) Peppered Ranidella riparia. Anim. Behav., 34: 1836-43. tree frog Litoria piperata. Chapter 22 In H. Ehmann (Ed.) Threatened Frogs of New South Wales: Habitats, Status Osborne,W. S. (1989) Distribution, relative abundance and and Conservation. Frog and Tadpole Study Group of conservation status of corroboree frogs, Psuedophryne NSW, Sydney, pp. 189-194. corroboree Moore (Anura: Myobatrachidae). Aust.Wildl. Res, 16: 537-47. Mahony, M., Knowles, R. and Pattinson, L. (1997b) Stuttering frog Mixophyes balbus. Chapter 4 In H. Ehmann (Ed.) Osborne,W. S. (1990) Declining frog populations and extinctions in the Canberra region. Bogong, 11: 4-7. Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of Osborne,W.S., Hunter, D.A. and Hollis, G.J. (1999) Population NSW, Sydney, pp. 65-71. declines and range contraction in Australian alpine frogs. Pp 145-157 in Declines and Disappearances of Australian Martin, A. A. (1972) Studies in Australian amphibia III.The Frogs ed by A.Campbell. Environment Australia: Canberra. complex (Anura: Leptodactylidae). Aust. J. Zool., 20: 165-211. O’Shaughnessy, P.and Associates (1995, unpublished) Water Quality Protection Measures for the Conservation of the Martin, A. A., Littlejohn, M. J. and Rawlinson, P.A. (1966) A key Spotted Tree Frog. Unpublished report to the Forest to the anuran eggs of the Melbourne area, and an addition Service,Victorian Department of Natural Resources, to the anuran fauna.Vict. Nat., 83: 312-15. Melbourne. McDonald, K. R. (1990) Rheobatrachus Liem and Taudactylus Parris, K. and Norton,T. (1997) The significance of State Straughan and Lee (Anura: Leptodactylidae) in Eungella Forests for conservation of Litoria pearsoniana (Copland) National Park, Queensland: distribution and decline.Trans. and associated amphibians. In P.Hale and D. Lamb (Eds.) Roy. Soc. S. A., 114: 187-194. Conservation Outside Nature Reserves. Centre for McDonald, K.R. and Alford, R.A. (1999) A review of declining Conservation Biology,The University of Queensland, St frogs in northern Queensland. Pp 14-22in Declines and Lucia, pp. 521-526. Disappearances of Australian Frogs ed by A.Campbell. Pechmann, J. H., Scott, D. E., Semlitsch, R. D., Caldwell, J. P.,Vitt, Environment Australia: Canberra. L. J. and Whifield Gibbons, J. (1991) Declining amphibian McDonald, K. R. and Davies, M. (1990) Morphology and populations: the problem of separating human impacts biology of the Australian tree frog Litoria pearsoniana from natural fluctuations. Science, 253: 892-95. (Copland) (Anura: Hylidae).Trans. Roy. Soc. S. A., 114: Pechmann, H. K. and Wilbur, H. M. (1994) Putting declining 145-56. amphibian populations in perspective: natural fluctuations Norris, K. C., Mansergh, C. A., Ahern, L. D., Belcher, C. A., and human impacts. Herpetologica, 50: 65-84. Temby, I. D. and Walsh, N. G. (1983) Vertebrate Fauna of Recsei, J. (1997) The giant burrowing frog Heleioporus the Catchment Victoria. Occasional Paper australiacus. Chapter 3 In H. Ehmann (Ed.) Threatened Series No. 1, Fisheries and Wildlife Division,Victoria, Frogs of New South Wales: Habitats, Status and Ministry for Conservation, Melbourne. Conservation. Frog and Tadpole Study Group of NSW, Sydney, pp. 55-64. NRE (1999) Threatened Vertebrate Fauna in Victoria — 1999. Department of Natural Resources and Environment, Richards, S. J. and Alford, R. A. (1992) Nest construction by an Victoria. Australian rainforest frog of the Litoria lesueuri complex (Anura: Hylidae). Copeia, 1992: 1120-23. NSW NPWS (1994) Fauna of North-east NSW Forests. North-east Forests Biodiversity Study Report No. 3, Richards, S. J., McDonald, K. R. and Alford, R. A. (1993) Declines in population of Australia’s endemic tropical Unpublished Report, NSW National Parks and Wildlife rainforest frogs. Pac. Cons. Biol., 1: 66-77. Service, Hurstville. Odendaal, F. J. and Bull, C. M. (1980) Influence of water speed on tadpoles of Ranidella signifera and R. riparia (Anura: Leptodactylidae). Aust. J. Zool., 28: 79-82.

129 Robertson, P.and Gillespie, G. R. (in review) Draft Spotted Walter, H. and Lieth, H. (1967) Klimadiagramm — Weltatlas. Tree Frog Recovery Plan. Prepared for the Biodiversity Jena, Fisher (Maps only). Group, Environment Australia, by Department of Natural Watson, G. F. and Littlejohn, M. J. (1985) Patterns of Resources and Environment, Melbourne,Victoria. distribution, speciation and vicariance biogeography of SFNSW (1995). Queanbeyan and Badja Management Area southeastern Australian amphibians. In G. Grigg, R. Shine Environmental Impact Statement. State Forests of New and H. Ehmann (Eds) Biology of Australasian Frogs and South Wales. Reptiles. Royal Zoological Society of New South Wales, SMEC (1997) Data Compilation. Prepared Sydney, pp. 91-7. by SMEC Australia Ltd for the Department of Land and Watson, G. F., Littlejohn, M. J., Hero, J-M. and Robertson, P. Water Conservation, Sydney. (1991) Conservation Status, Ecology and Management of Straughan, I.R. (1968). A taxonomic review of the genus the Spotted Tree Frog (Litoria spenceri). Arthur Rylah Mixophyes, (Anura, Leptodactylidae), 1968. Proc. Linn. Soc. Institute for Environmental Research Technical Report NSW. 93:52-59 Series No. 116. Department of Conservation and Environment, Melbourne. Trenerry, M. P.,Laurance,W. F. and McDonald, K. R. (1994) Further evidence for the precipitous decline of endemic Webb, G. A. (1987) A note on the distribution and diet of the rainforest frogs in tropical Australia. Pac. Cons. Biol., 1, giant burrowing frog, Heleioporus australiacus (Shaw and 150-53. Nodder 1795)(Anura: Myobatrachidae). Herpetofauna, 17(2): 20-22. Tyler, M. J. (1991) Declining amphibians — a global phenomenon? An Australian perspective. Alytes, 9, 43-45. Webb, G. A. (1991) A survey of the reptiles and amphibians of Bondi State Forests and surrounding areas, near Tyler, M. J. (1997) The Action Plan For Australian Frogs. Bombala, New South Wales. Aust. Zool., 27: 14-19. Biodiversity Group, Environment Australia, Canberra. Wellington, R. and Wells, R. (1994). Fauna Survey of the Tyler, M. J. and Anstis, M. (1975) Taxonomy and biology of Morisset District Central Coast, New South Wales, frogs of the Litoria citropa complex (Anura: Hylidae). Rec. Reptiles and Amphibians. Unpublished Report to State S. A. Mus., 17: 41-50. Forests of New South Wales. Tyler, M. J. and Davies, M. (1978) Species groups within the White, A.W. and Pyke, G. H. (1996) Distribution and Australopapuan hylid genus Litoria Tschudi. Aust. J. Zool. conservation status of the green and golden bell frog Suppl. Ser., 63: 1-79. Litoria aurea in New South Wales. Aust. Zool. 30: 177-89. Tyler, M. J. and Davies, M. (1985) A new species of Litoria (Anura: Hylidae) from New South Wales, Australia. Copeia, 1985:145-49.

130 Potential Impacts of Introduced Fish and Fish Translocations on Australian Amphibians

Graeme Gillespie1 and Jean-Marc Hero2

ABSTRACT subject to thorough research. Many Australian amphibian assemblages, including several threatened This review examines the potential impact of species, are potentially threatened by a variety of introduced fish on amphibians, with particular introduced fish species. Future research priorities and emphasis on Australian freshwater systems. Firstly, the ecological relationships between fish predators guidelines for examining the impact of introduced fish and their amphibian prey are examined, and how on Australian amphibians are outlined. Key management they can be altered when non-native fish are objectives for conservation agencies are identified. introduced into aquatic systems.The current knowledge and research on the impacts of INTRODUCTION introduced fish on amphibians both overseas and The reported declines of many amphibian populations both in Australia and around the world are now recognised as a within Australia is then reviewed. Evidence in the very real phenomenon.These declines pose a serious threat literature strongly suggests that introduction of to global amphibian diversity, and may result from recent exotic fish or translocation of native species could global environmental change associated with human activities. have enormous impacts on the amphibian The cause(s) of many species declines, particularly in apparently pristine tropical forests of Central America and assemblages of Australian freshwater systems. Australia, remain obscure (but see Lips 1998). However, in many cases one or other anthropogenic impacts, commonly Introduced fish have been implicated in the decline of identified as key threatening processes in the decline or several anuran species, though few cases have been extinction of other vertebrates (Meffe and Carroll 1994;

1. Arthur Rylah Institute, Department of Natural Resources and Environment, PO Box 137, Heidelberg,Victoria 3084 Australia. 2. School of Applied Science, Griffith University Gold Coast, PMB 50 Gold Coast MC, Queensland 4127 Australia.

131 Leakey and Lewin 1995), are associated with observed Fish may directly impact amphibian species by predation on declines.These include habitat destruction (Laan and larvae (Macan 1966; Heyer et al. 1975; Sih et al. 1988) or Verboom 1990; Johnson 1992;Wardell-Johnson and Roberts eggs (Grubb 1972). Consequently fish predators are capable 1991; Gillespie and Hollis 1996; Dubuis 1997;Waldick 1997), of eliminating larval amphibian species from some habitats pollution (Bishop 1992; Bidwell and Gorrie 1995; Bertram (Tyler 1963, Macan 1974, Petranka 1983). In a comprehensive and Berrill 1997), over-exploitation (Jennings and Hayes study by Petranka (1983) the larvae of the salamander 1985) and introduction of exotic predators (Orchard 1992; Ambystoma texanum were found to be restricted to the Bradford et al. 1993; Lannoo et al. 1994). In Australia, fish-free, upper portions of breeding streams and this was anthropogenic impacts on amphibian populations are only attributed to predation on larvae by endemic species of fish. now beginning to be appreciated.The relative significance of Kats et al. (1988) identified fish predation as a primary factor various potentially threatening processes to the maintenance influencing marked differences between larval amphibian of many amphibian communities is poorly understood.This is assemblages in ephemeral and permanent aquatic habitats. a reflection in part of the inherent difficulties associated with Similarly, Fickling (1995) found that Litoria nannotis and studying amphibian population dynamics, the high diversity of L. rheocola were restricted to streams without predatory fish amphibian assemblages present in Australia, and the small in the Tully gorge, northern Australia. number of ecologists and amount of resources available for studying them. Decisions of research and management TADPOLE SURVIVAL STRATEGIES priorities and allocation of resources must therefore be based upon careful assessments of current knowledge about the Amphibian larvae can physically evade fish predators through importance of all potentially threatening processes, from both spatial or temporal avoidance (Petranka 1983, Bradford 1989, within Australia and overseas. Holomuzki 1995, Hecnar and M’Closkey 1997, Hero et al. 1998). Many species of amphibian breed only in temporary After over-exploitation and habitat destruction, introduced water bodies in which fish are absent. Recent studies have predators have been identified as the main cause of mammal shown that females of some species of amphibian choose and bird species extinctions in modern times, particularly in oviposition sites in waterbodies without fish (Resetarits and (Meffe and Carroll 1994; Leakey and Lewin 1995). Wilbur 1989; Kats and Sih 1992; Bronmark and Edenhamn It is likely that introduced predators may also play a significant 1994; Hopey and Petranka 1994; Holomuzki 1995). role in the decline of some amphibian species.The following Alternatively, amphibians can avoid fish predators by review examines the potential impact of introduced fish on reproducing in a waterbody at times when fish predators are amphibians, with particular emphasis on Australian freshwater absent (e.g. streamside ponds that are isolated from the systems. Firstly, we examine the ecological relationships stream at some times of the year). between fish predators and their amphibian prey, and how these may be altered when non-native fish are introduced.We The larvae of many amphibian species occur in habitats then review current knowledge and research on the impacts of containing predatory fish, such as permanent lakes and introduced fish on amphibians both overseas and within streams. Survival or anti-predator strategies allow these Australia. Finally, we outline future research priorities and species to coexist with fish predators.These strategies include guidelines for examining the impact of introduced fish on cryptic colouration (Wasserug 1971), behavioural responses Australian amphibians, and identify key management objectives. such as use of refugia (Sih et al. 1988), schooling (Waldman 1982; Kruse and Stone 1984), protean flight (Taylor 1983), and chemical defences (Liem 1961;Wasserug 1971; Brodie et SIGNIFICANCE OF FISH PREDATORS IN al. 1978; Kruse and Stone 1984; Kats et al. 1988;Werner and AMPHIBIAN COMMUNITIES McPeek 1994). In contrast to species which typically occur in Predation is considered to be a major factor regulating the fish-free habitats, larvae of species which coexist with fish distribution of amphibian larvae (e.g. Calef 1973, Heyer et al. predators may possess one or a combination of these 1975, Duellman 1978, Scott and Limerick 1983, Smith 1983, survival traits (Kats et al. 1988). Many amphibian larvae which Woodward 1983,Wilbur 1984, Hayes and Jennings 1986, Kats coexist with predacious fish are unpalatable or noxious (Liem et al. 1988). Heyer et al. (1975) suggested that predation by 1961;Voris and Bacon 1966;Wasserug 1971; Brodie et al. aquatic predators, primarily fish, was the most important 1978;Walters 1975; Kruse and Stone 1984; Kats et al. 1988; biotic factor influencing the temporal and spatial composition Hero 1991;Werner and McPeek 1994). Amphibian larvae of tadpole communities.The combined direct and indirect which do not respond behaviourally to predatory fish are effects of fish predators on the local distribution of individual typically toxic or unpalatable to fish (Voris and Bacon 1966; species of tadpole consequently influence local and regional Kruse and Fransis 1977; Kruse and Stone 1984; Kats et al. amphibian assemblage structure. Recent studies have 1988). Antipredator strategies used by tadpole species against demonstrated the importance of fish predators in invertebrate predators, such as immobility (Azevedo-Ramos determining tadpole species-composition (species present) et al. 1992; Chovanec 1992;Werner and McPeek 1994) are and tadpole species richness (number of species) in not usually effective against fish predators that use visual cues temperate (Hecnar and M’Closkey 1997) and tropical (Hero 1991;Werner and McPeek 1994).Therefore the systems (Fickling 1995; Hero et al. 1988). In this section, we distribution of each larval amphibian species is related to the examine the ecological relationships between fish predators survival strategies it possesses and is strongly influenced by and their amphibian prey and how these may be altered the distribution of predatory fish. when non-native fish are introduced.

132 PREDICTED IMPACTS OF INTRODUCED OVERSEAS EVIDENCE FOR IMPACTS OF PREDATORS INTRODUCED FISH ON AMPHIBIANS Predator-prey relationships are maintained by a constant The consequences of introducing fish into breeding habitats evolution of both the predator to capture and the prey to for amphibians have been well documented overseas. A avoid capture, commonly described as the “evolutionary arms number of studies in Europe, North and have race” (Dawkins and Krebs 1979). Survival strategies tend to implicated or demonstrated that introductions of predatory be predator specific and are unlikely to be effective against all fish are responsible for the decline or extinction of some predators. For example, female amphibians may not be able amphibian species. to recognise the chemical cues produced by introduced fish Brönmark and Edenhamn (1994) suggest that, in Europe, the species and may inadvertently oviposit in a water body with widespread introduction of various fish species into farm exotic fish predators, resulting in levels of predation that dams and ponds has contributed to the decline of Hyla preclude survival of the species. Palatability of a species of arborea.They found that H. arborea in Sweden was tadpole can differ among different species of fish predator predominantly restricted to ponds in which fish had not been (Hero 1991; Holomuzki 1995). Hence, a species of tadpole introduced. No reproduction was recorded during a three may be unpalatable to the native fish predators with which it year period in ponds containing either pike (Esox lucius), coexists, but may not be unpalatable to a novel fish predator. perch (Perca fluviatilis), roach (Rutilus rutilus), Crucian carp Prey species may not identify introduced fish as predators (Carassius carassius), rudd (Scardinius erythrophthalmus) or and hence fail to use the appropriate survival strategies tench (Tinca tinca). Pike (Esox lucius), perch, and Crucian carp (temporal or spatial isolation), or the species of tadpole may have been shown in laboratory studies to readily feed on not have the necessary antipredator defences that allow H. arborea tadpoles and metamorphs (Brönmark unpublished, them to coexist with introduced fish species.The introduction in Brönmark and Edenhamn 1994). of an exotic predator is therefore likely to disrupt the arms race in favour of the predator. Macan (1966) reported a dramatic decrease in numbers of bufonid and ranid tadpole species following the introduction Fish predators can also influence tadpole assemblages of brown trout (Salmo trutta) into a British tarn. Braña et al. indirectly by consuming invertebrate tadpole-predators, such (1996) found that amphibian species’ numbers and amphibian as dragonfly naiads and predacious diving beetles (Wilbur and abundance were significantly lower in lakes of northern Spain et al. Fauth 1990;Werner and McPeek 1994; Hero 1998). containing introduced fish: brown trout, rainbow trout Along an environmental gradient Werner and McPeek (1994) (Oncorhynchus mykiss), tench, roach and European minnow found that only Rana catesbiana tadpoles were found in (Phoxinus phoxinus).They concluded that the presence of waterbodies with fish predators while R. clamitans was found these introduced species was responsible for the almost primarily in fishless ponds with high densities of invertebrate complete disablement of large permanent waterbodies for predators. Furthermore, the presence of fish predators can amphibian reproduction and subsequent decline of amphibian reduce densities of some species of tadpole, and this may species in the region. release other species from competition, thus enhancing their survival (Morin 1986;Werner and McPeek 1994).This In North America the introduction of salmonids into predator mediated release from competition may result in a previously fishless habitats has impacted upon numerous shift in species composition from species that are amphibian species. Burger (1950) reported the wide scale competitively dominant to species that are competitively elimination of tiger salamander (Ambystoma tigrinum) larvae inferior but have the survival strategies that allow them to from ponds in Colorado following stocking with trout. Fish coexist with fish. introductions, primarily trout, have been suggested as an important factor contributing to the decline of ranid frog The general pattern observed in natural systems is that species (Hayes and Jennings 1986; Liss and Larson 1991; species of tadpole that are vulnerable to predation by Hecnar and M’Closky 1996). Several species of introduced invertebrate predators survive in waterbodies with fish salmonids have profoundly affected the distribution of the (where the density of invertebrate predators is low due to Mountain Yellow-legged Frog (Rana mucosa) within the past predation by fish), and species of tadpole that are vulnerable century by eliminating the species from nearly all waters to predation by fish survive in waterbodies where predacious where fish have been introduced (Grinnell and Storer 1924; fish are absent (Hecnar and M’Closkey 1997; Hero et al. Bradford 1989; Bradford et al. 1993). Hayes and Jennings 1988).The introduction of predacious fish species will (1986) noted that the abundance of endemic Rana species in potentially result in the elimination of some tadpole species California was inversely correlated with densities of and a shift in the species composition to those species which introduced fish species, primarily trout.Tyler et al. (1998) have the survival-strategies that allow them to coexist with demonstrated that larval salamanders (Ambystoma the introduced predator. macrodactylum) were found in much higher densities in alpine lakes without fish than in lakes that contained introduced Theory therefore predicts that the introduction of exotic fish trout populations. to aquatic systems may lead to the elimination of some species of tadpole, resulting in changes in the species In Canada Liss and Larson (1991) reported the decline of composition of natural tadpole assemblages.These changes amphibian species in naturally fishless lakes after stocking with may be extremely detrimental to the long term survival of trout. Hecnar and M’Closkey (1996) concluded that the some species, undermine amphibian communities and disrupt presence of introduced predatory fish was responsible for the natural aquatic systems. decline of amphibian species in south-western Ontario.They

133 found that amphibian species richness was significantly lower amphibians in southern Chile (Formas 1995). Introduced in ponds containing introduced predatory fish. However, those salmonids are also thought to be responsible for the amphibian species with either large larval body size or large extinction of several Atelopus species in Costa Rica clutch size were less adversely affected than others, and (Pough et al. 1998). occurred more frequently with predatory fish. In most of the above cases, fish introductions have occurred The introduction or translocation of other species has also for recreational purposes. Hence, the frequent reports been implicated in the decline of some amphibian species in involving trout species, which have been widely introduced in North America. Introduced mosquitofish (Gambusia spp.) lakes and streams throughout both hemispheres due to their have been identified as the most likely cause of localised popularity with anglers. It should be emphasized that declines of Californian newts ( torosa) in southern translocation of native fish species into aquatic systems that California (Gamradt and Kats 1996). Petranka (1983) have not previously contained the species could have similar documented decimation of small-mouthed salamander impacts on amphibian fauna. (Ambystoma texanum) larvae in local pools in streams following colonisation by Green Sunfish (Lepomis cyanellus), INTRODUCED FISH IN AUSTRALIA and Sexton and Phillips (1986) noted a dramatic reduction in species richness after the introduction of this species. The list of fish introduced into Australia is extensive (Table 1). Semlitsch (1983) reported almost complete mortality of At least 24 exotic species have established self-sustaining Rana esculenta tadpoles following the addition of Pike to populations in Australian freshwater systems to date. In experimental ponds. addition, several native species have been translocated into aquatic systems in which they did not naturally occur.These Declines of some amphibians in South America have also include Murray cod (Maccullochella peelii), golden perch been attributed to introduced fish.The introduction of (Macquaria ambigua), Macquarie perch (M. australasica), bass various fish species: salmonids, European carp (Cyprinus (M. novemaculeata), barramundi (Lates calcarifer), catfish carpio), Odonthestes bonariensis and catfish (Ictalurus spp.), is (Tandanus tandanus) and rainbow fish (Melanotaenia spp.) thought to be a principal factor leading to the decline of (Raadik, Arthur Rylah Institute (ARI),Victoria, pers. comm.).

TABLE 1: List of exotic fish which have established populations in Australian waters.

Species Origin Occurrence in Australia

Salmonidae Rainbow Trout Oncorhynchus mykiss Nth.America NSW;Vic;Tas; s.w.WA; s.e. SA Chinook Salmon O. tshawytscha Nth.America, N.E.Asia s.w.Vic Brook Trout Salmo fontinalis Nth.America s.e. NSW;Tas Atlantic Salmon S. salar Europe, Nth America s.e. NSW;Vic;Tas Brown Trout S. trutta Europe,West Asia NSW;Vic;Tas; s.w.WA; s.e. SA Cyprinidae Goldfish Carassius auratus East Asia s.e. Qld; NSW;Vic; SA; s.w.WA European Carp Cyprinus carpio Europe s.e. Qld; NSW;Vic; SA; s.w.WA Rosy Barb Puntius conchonius * Asia s.e.Qld Roach Rutilus rutilus Europe Vic Tench Tinca tinca Europe, Central Asia sth. NSW;Vic;Tas Percidae Redfin Perch Perca fluviatilis Europe, Nth.Asia NSW;Vic; east SA;WA;Tas Poecilidae gambusia (Mosquito Fish) Gambusia holbrooki Nth.America Qld; NSW;Vic; SA;WA;Tas One-spot Livebearer Phalloceros caudimaculatus Sth.America s.w.WA Sailfin Molly Poecilia latipinna Cent.America s.e. Qld Guppy P. reticulata Cent.America, s.e. Qld Swordtail Xiphophorus helleri Cent.America s.e. Qld Platy X. maculatus Cent.America s.e. Qld Cyprinodontidae American Flag Fish Jordanella floridae Nth.America n.e. Qld Cobitidae Oriental Weatherloach Misgurnus anguillicaudatus East Asia s.e. Qld; s.e. NSW;Vic Cichlidae Blue Acara Aequidens pulcher Cent.America e. Qld Convict Cichlid Heros nigrofasciata Cent.America Vic Mozambique Mouthbrooder Oreochromis mossambicus East Africa e. Qld, s.w.WA Black Mangrove Cichlid Tilapia mariae West Africa Vic Zilles Cichlid T. zillii * Africa s.w.WA

* Species which established populations that either died out or were successfully removed. (Sources:Allen 1982; Cadwallader and Backhouse 1982; McKay 1984;Allen 1989; Faragher and Harris 1993;Arthington and Blühdorn 1995; Ryan 1995; McDowall 1996).

134 Introductions have occurred primarily either for recreational Gambusia are widespread throughout south-eastern Australia, fishing purposes, or through releases of species from the including the Murray-Darling system, and extend up the east aquarium trade.The one exception is the eastern gambusia coast as far as Townsville. It also occurs in south-western WA or mosquito fish (Gambusia holbrooki), which was misguidedly and some water courses near Alice Springs (Allen 1982, introduced to control mosquitoes (Myers 1965). 1989; McDowall 1996).The species is exceptionally hardy and is able to tolerate an extremely broad range of Salmonids have been widely introduced into streams and environmental conditions (McKay 1984). It is a highly invasive lakes for recreational fishing.The brown trout and rainbow species inhabiting marshes, lakes and dams, slow-flowing trout are the most successful and widespread species.These streams and associated billabongs and aqueducts. It is most occur in mainland streams along the Dividing Range from abundant in modified habitats and areas near human Victoria up to northern NSW, in the Adelaide Hills and in settlement (Allen 1989; McDowall 1996). south-western Western Australia (Allen 1982, 1989; The remaining species have all originated from the aquarium McDowall 1996). In south eastern Australia, brown trout are trade (McKay 1984; Allen 1989; Ryan 1995; McDowall 1996). abundant in all upland streams and are only excluded from a Only three of these have so far established extensive few small tributaries where waterfalls have blocked their up- distributions in the wild.The oriental weatherloach (Misgurnus stream passage (Cadwallader and Backhouse 1982; Faragher anguillicaudatus) occurs in streams along the east coast of NSW and Harris 1993; McDowall 1996). Rainbow trout have a and several south-flowing catchments in Victoria, such as the more patchy distribution but are in high abundance in many Yarra and Latrobe Rivers. It also occurs inland in south-eastern small upland water courses (McDowall 1996;Victorian Fish Australia, in the Murrumbidgee, Ovens and Murray Rivers Database, ARI,Victoria). Both species also occur in many lakes (McDowell 1996;Victorian Fish database, ARI,Victoria).The and in these regions. Stocking of lakes and streams Mozambique mouthbrooder (Oreochromis mossambicus) has occurs extensively in NSW, and several lakes are stocked in established populations in the lowland reaches of several coastal Victoria by Fisheries authorities.These species are also rivers in Queensland between Brisbane and Cairns, and has stocked in farm dams. Brook trout have established in been reported in several rivers in south-western Western Tasmania in lakes of the Tyndall Ranges (McDowall 1996), the Australia (McKay 1984; Allen 1989; McDowall 1996).The guppy Clarence Lagoon on the Central Plateau (Swain, University of (Poecilia reticulata) is widespread from Brisbane to north-east Tasmania, pers. comm.), and are currently restricted to one Queensland (Ryan 1995).The one-spot livebearer (Phalloceros stream on the mainland in Kosciuszko National Park, NSW caudimaculatus) has been recorded in ponds and drains around (Harris, NSW Fisheries, pers. comm.). Chinook salmon are Perth (Allen 1989).The sailfin molly (Poecilia latipinna), swordtail currently restricted to several lakes in south-western Victoria, (Xiphophorus helleri) and platy (X. maculatus), are restricted to a maintained by stocking (McDowall 1996). Atlantic salmon are few streams around Brisbane (McDowall 1996).The rosy barb stocked in lakes and reservoirs in south-eastern NSW, central (Puntius conchronius) also established itself in one stream in the Victoria and Tasmania (McDowall 1996).The species has Brisbane area but has apparently died out (Brumley 1991; McDowall 1996).The American flag fish (Jordanella floridae) has escaped from hatcheries into the Rubicon and Latrobe Rivers been recorded near Cairns (Allen 1989).The convict and black in Victoria (McDowall 1996). mangrove cichlid (Tilapia mariae) are restricted to the Goldfish and European carp were originally introduced as Hasellwood Pondage in Morwell Victoria, which contains warm ornamental fish and have spread throughout the Murray- water outflow from the power station. Zilles cichlid (T. zillii) was Darling system and other inland and coastal waterways in recorded in tributaries of the Swan River estuary in Western south-eastern Australia (Cadwallader and Backhouse 1982; Australia. in 1975, but is believed to have been successfully Faragher and Harris 1993; McDowall 1996).They also occur eradicated (Allen 1989).The blue acara (Aequidens pulcher) has been recorded in one stream in Brisbane (Ryan 1995). in south-western Western Australia (Allen 1982, 1989) and Tasmania (Swain pers. comm.).Tench and roach were Many of Australia’s inland waters, particularly in south-eastern introduced in the late 1800’s into lakes and rivers in Victoria regions, contain one or more introduced fish species. In some for fishing and have both spread. Movements of roach have cases these species have completely displaced native fish and also been recorded up rivers feeding lakes, such as the substantially modified aquatic ecosystems. In addition to those Howqua and Big Rivers in the catchment of Lake Eildon, species already established, there is continual interest from Victoria (Victorian Fish Database, ARI,Victoria). Introductions the recreational and commercial fishing industry to establish of roach still occur illegally for caurse fishing.Tench have also hatcheries or introduce more species. Several hundred been stocked for fishing in lakes and reservoirs in southern species of exotic ornamental fish have been imported into NSW and Tasmania. Australia for the aquarium trade (McKay 1984).The potential for more of these species to become established in natural Redfin have been widely introduced throughout the Murray- waters is high, particularly in tropical and subtropical regions system, lakes and farm dams in south-eastern (McKay 1984; McDowall 1996). mainland Australia,Tasmania and south-western Western Australia as a popular angling species (Allen 1982, 1989; With the possible exception of the Oriental Weatherloach, all Rowland 1989; McDowall 1996).This species has also of the introduced species have the potential to prey upon penetrated up major tributaries of some lakes, in some cases amphibian eggs and larvae, and many species may also prey upon adults. As indicated in the literature reviewed earlier, considerable distances, such as Eildon and Glenmaggie in this has already been demonstrated for most of the more Victoria (Victorian Fish database, ARI,Victoria). widely introduced species, such as the salmonids, cyprinids, redfin perch and gambusia, on other continents.

135 REVIEW OF IMPACTS OF INTRODUCED sources for trout, and refuge microhabitats for tadpoles, trout FISH ON AMPHIBIANS IN AUSTRALIA were able to impose a significant predation pressure on L. spenceri and L. phyllochroa. Brown and rainbow trout are now Few studies have been conducted to investigate the considered to be the primary cause of decline of L. spenceri relationships between introduced fish and amphibians in (Robertson et al. 1998, unpublished; Robertson and Gillespie, Australia.To date, the impacts of only three introduced 1998, unpubl.).These findings suggest that upland populations species, brown and rainbow trout, and gambusia, have been of the other species within the L. citropa complex, i.e. L. investigated. Collectively these studies have assessed impacts subglandulosa and L. pearsoniana, may also be highly vulnerable on only 16 species of frog to any degree (Table 2). Some to predation from trout. Although L. lesueuri complex species information is presented on redfin perch and carp; however, were less palatable, L. booroolongensis has also declined appropriate research is required to further examine the (Gillespie and Hines 1999; NSW NPWS Scientific Committee impacts of these species. Determination Advice No. 97/27). Other factors such as habitat degradation may be involved but it remains unclear Impact of Fish Predation on Adult Frogs what impact trout may have on these species. For example, an It is common knowledge among the fishing fraternity that egg mass of L. lesueuri was found in the stomach of a brown frogs make good bait for trout and redfin perch (Baxter, trout (Rardik, ARI,Victoria, pers. comm.).This fish was able to Victorian Fisheries, pers. comm.; Harris, NSW Fisheries, pers. take out most of the annual reproductive investment of a comm.; Lake, Department of Biological Sciences, Monash single female frog (several hundred eggs) in one sitting. It is University, pers. comm.).This suggests that trout and redfin expected that predation pressure by trout on these species perch may readily attack frogs in the wild. Collection of frogs will be high when alternative food resources are limited. for bait may place excessive pressure on some frog populations (Watson et al. 1991).The use of frogs for bait is Impact of Predation by Gambusia now banned in some States. However, it is likely that fish are Gambusia has so far received the most scrutiny regarding its able to exert their greatest impact on frog populations by potential impact upon Australian frog populations.The broad preying upon larval stages. distribution and wide range of habitats occupied by gambusia means that it may potentially impact upon many lentic and Impact of Predation by Trout Species lotic frog populations across a large area of Australia. Species most at risk from predation by trout are those which Only one study has examined predation by gambusia upon breed exclusively in streams in south-eastern Australia.There anuran eggs; Reynolds (1995) found that eggs of Crinia are eight such species, several of which have declined in insignifera and C. glauerti were unpalatable. Preliminary trials recent years and are considered endangered or vulnerable also suggested that eggs of Litoria adelaidensis, L. moorei and (Gillespie and Hines 1999).The spotted tree frog (Litoria Crinia georgiana may also be unpalatable (Reynolds 1995). spenceri) has always been considered rare (Ahern 1982); However, several studies have shown experimentally that however, declines were observed in most of the few known gambusia are capable of preying on small larvae of a number populations in the 1970s and 80s (Watson et al. 1991), and of Australian anuran species: Limnodynastes tasmaniensis, the species is now listed as endangered (Tyler 1997).Watson Litoria lesueuri and L. dentata (Harris 1995); Crinia insignifera et al. suggested that introduced trout may be contributing to and C. glauerti (Reynolds 1995); Litoria aurea and L. dentata this decline.Trout are present in all the streams in which the (Morgan and Buttemer 1996); Limnodynastes peronii and species is known to have occurred (Victorian Fish database, Crinia signifera (Webb and Joss 1997). ARI,Victoria). Surveys of the distribution and relative abundance of L. spenceri and other upland riverine species A number of studies have identified negative associations have found that L. spenceri only occurred in abundance in one between the presence of gambusia and frog species. Dankers reach of stream which was above a waterfall which trout (1977) found that tadpole numbers of several species were could not negotiate (Gillespie and Hollis 1996; Hunter and drastically reduced in ponds containing gambusia after early Gillespie 1999). Only a few high density upland populations December, coinciding with a seasonal increase in fish biomass. of the leaf-green tree frog (L. phyllochroa) have been located, McGilp (1994) found a negative correlation between the most of which have also been above waterfalls in trout-free occurrence of Brown Tree Frog (Litoria ewingii) and that of streams (Gillespie pers. obs.). In contrast, lesueur’s frog gambusia in waterbodies along the in Melbourne. (L. lesueuri) remains widespread and is abundant along many streams where trout are present (Gillespie and Hollis 1996; Blyth (1994) compared survival and recruitment of three Hunter and Gillespie 1999; Gillespie pers. obs.). species of Western Australian anuran larvae, Crinia glauerti, C. insignifera and Heleioporus eyrei, in the presence or absence of Gillespie (1997, unpubl.) examined the relative palatabilities of gambusia in experimental field enclosures.Tadpole survival of five riverine frog larvae, L. booroolongensis, all three species was significantly lower in the presence of L. citropa, L. lesueuri, L. phyllochroa and L. spenceri, to two gambusia at the end of the experimental period. However, sympatric native fish, mountain galaxias (Galaxias olidus) and the design of the enclosures allowed access for oviposition by two-spined blackfish (Gadopsis bispinosis), and introduced local frog populations, as evidenced by increases in numbers brown trout. All fish readily consumed tadpoles of of experimental animals in some enclosures. Other potential Limnodynastes peronii which occur in lentic habitats without fish; predators of premetamorphic stages also had access, such as however, only trout ate a significant proportion of tadpoles of invertebrates and birds. Furthermore, each species/fish any riverine tadpole species. Further in-stream experiments treatment was not replicated.These factors limit demonstrated that despite the availability of alternative food interpretation of the results of this study.

136 TABLE 2: List of introduced fish and native frog species-interactions that have been examined in Australia.

Fish species Frog species Source

Brown Trout Booroolong Frog Litoria booroolongensis Gillespie (1997), unpublished (Salmo trutta) Blue Mountains Tree Frog L. citropa Gillespie (1997), unpublished Lesueur’s Frog L. lesueuri Gillespie (1997), unpublished Leaf-green Tree Frog L. phyllochroa Gillespie (1997), unpublished Spotted Tree Frog L. spenceri Gillespie (1997), unpublished Limnodynastes peroni Gillespie (1997), unpublished Rainbow Trout Leaf-green Tree Frog L. phyllochroa Gillespie (1997), unpublished (Oncorhynchus mykiss) Spotted Tree Frog L. spenceri Gillespie (1997), unpublished Eastern gambusia or Litoria adelaidensis Reynolds (1995) Mosquito Fish Green and Golden Bell Frog L. aurea Morgan and Buttemer (1996); (Gambusia holbrooki) Pyke and White (1996) Kerferstein’s Tree Frog L. dentata Harris (1995); Morgan and Buttemer (1996) Moor’s Frog L. moorei Reynolds (1995) Lesueur’s Frog L. lesueuri Harris (1995) Tschudi’s Froglet Crinia georgiana Reynolds (1995) Glauert’s Froglet C. glauerti Blyth (1994); Reynolds (1995) Sign-bearing Frog C. insignifera Blyth (1994); Reynolds (1995) Common Froglet C. signifera Webb and Joss (1997) Moaning Frog Heleioporous eyrei Blyth (1994); Reynolds (1995) Striped Marsh Frog Limnodynastes peroni Webb and Joss (1997) Spotted Marsh Frog L. tasmaniensis Harris (1995) Goldfish Spotted Marsh Frog L. tasmaniensis M. Healey (unpublished data) (Carassius auratus)

Webb and Joss (1997) examined amphibian species richness presence of gambusia. However, gambusia showed a strong and abundance in relation to gambusia density and cover of preference for invertebrate prey (Daphnia sp. or mosquito emergent aquatic vegetation in ten ponds near Sydney.They larvae). Both groups were consistently consumed completely found a significant negative relationship between fish density before tadpoles in all trials. In a field enclosure experiment, in and frog abundance but no relationship for species richness. which tadpoles were also exposed to invertebrate predators, The descriptions provided for each waterbody indicate a high Reynolds (1995) found no significant difference in survival in degree of variability in habitat among pond sites. Unfortunately the presence or absence of gambusia.These results, in additional factors such as pool size and native vegetation cover, conjunction with his field survey data, suggest that the impact which may strongly affect frog abundance, were not considered of gambusia upon populations of these frog species is in their analyses.Tadpole density is easier to sample influenced by several factors, and under natural conditions systematically than adult frog density in pond habitats (Heyer et may be limited. al. 1994). Given that tadpoles are one of the life stages on which gambusia potentially preys upon, a measure of their Gambusia cannot consume large prey as these small fish are relative abundance, rather than that of adult frogs, will provide gape-limited predators.Webb and Joss (1997) conducted a more reliable indicator of the impact of gambusia. predation experiments examining the impact upon survival of different size classes of C. signifera and Limnodynastes peroni Reynolds (1995) examined the occurrence of six anuran tadpoles by hungry and pre-fed gambusia.They found species with gambusia in water bodies near Perth,Western significant differences between predation rates due to tadpole Australia. In contrast to the above studies, he found no size class and hunger status of fish.Tadpole species which are relationship between the presence/absence of fish and able to rapidly attain moderate to large size may therefore individual anuran species, with one exception, Crinia insignifera, which was found infrequently with gambusia. However, he minimize the impact of predation (Caldwell et al. 1980; observed that most of the sites used by C. insignifera were Crump 1984). ephemeral and unsuitable for gambusia. Species richness was Several studies have reported damage to the fins of larger generally lower at sites occupied by gambusia, but many of tadpoles from gambusia attack (Dankers 1977; Blyth 1994; these sites were also degraded, contributing to their Harris 1995).This could result in reduced survival of larger unsuitability as frog breeding habitats. tadpoles due to reduced mobility and feeding, inability to In addition Reynolds (1995) experimentally examined escape other predators, or reduced metamorphic fitness. predation by gambusia on several tadpole species in Western However, some tadpole species have been found to survive Australia.Trials with tadpoles indicated that gambusia were tail loss (Harris 1995).Wilbur and Semlitsch (1990) reported able to attack and kill tadpoles of L. adelaidensis, C. georgiana tail regeneration by tadpoles of Rana catesbeiana even after and H. eyrei. Controlled palatability experiments showed that considerable loss, and suggest that this may be a general survival of L. moorei tadpoles was significantly reduced in the mechanism to reduce the impact of predation.

137 Concerns for the role of gambusia in the decline of (Reynolds 1995). Further research is required to ascertain the amphibian species, particularly members of the L. aurea role of gambusia in the decline of amphibian species assemblages complex, have been expressed by several authors (Mahony with respect to other threatening processes. For instance, as 1993; Daly 1995; Morgan and Buttemer 1996;White and gambusia occur in areas which are mostly disturbed or modified Pyke 1996;White and Ehmann 1997). However, evidence in other ways, the relative impacts of these habitat changes upon linking gambusia to declines of frog populations in the L. aurea amphibian populations need to be differentiated from those complex is limited, due in part to conflicting findings and wrought by the fish. methodological limitations of some studies. Impacts of Predation by Redfin Perch and For example, Morgan and Buttemer (1996) conducted Carp Species controlled predation experiments examining the impact upon survival of tadpoles of L. aurea and L. dentata by gambusia. Research overseas suggests that redfin perch and carp species The influence of macrophytes on the predatory impact of may be major predators of some tadpole species. No gambusia was also examined.They found that in the absence published studies have addressed the impact of carp species of macrophytes gambusia were able to significantly reduce or redfin perch upon frogs in Australia. Leslie (1995) has tadpole survival of both species within 24 hours. In the attributed the decline of frogs in some wetlands in the presence of macrophytes, the effect was substantially reduced Murray-Darling Basin in part to predation on and no significant impact of gambusia could be detected on premetamorphic stages by carp species. Healey et al. (1997) L. aurea after three days. However, survival of L. dentata was observed no evidence of frogs breeding in four billabongs on still significantly reduced after two days.These findings indicate the Murrumbidgee floodplain and suggested that this may that presence of gambusia may significantly influence the have been explained by the presence of carp species. survival of tadpoles, but that this is likely to be strongly However, no observations were made at sites where carp were absent and the absence of tadpoles could be explained influenced by habitat structure and tadpole behaviour. Litoria by a number of alternative abiotic and biotic hypotheses. aurea larvae have also been found in sympatry with native Laboratory predation experiments have shown that tadpoles predatory fish (pers. obs.). In the absence of comparative of Limnodynastes tasmaniensis are palatable to goldfish and data on the impact of these natural predators upon larval redfin perch (Healey, Charles Sturt University,Wogga Wogga, survival, it is difficult to assess the relative ecological pers. comm.), indicating the potential for these species to significance of gambusia predation. consume tadpoles. However, it is unknown whether they prey Pyke and White (1996) surveyed waterbodies in the Sydney on tadpoles in the wild when alternative food is available. region for L. aurea, and examined associations between Carp are able to significantly modify the physical habitat of evidence of breeding, occurrence of introduced fish, and aquatic systems, by uprooting aquatic vegetation and habitat.They found that breeding was most strongly associated increasing turbidity (Roberts et al. 1995).These changes may with ephemeral rather than permanent or ‘fluctuating’ ponds, have indirect impacts on tadpoles through loss of food followed by the absence of introduced fish, primarily gambusia, resources, cover for protection from other predators, and and speculated that this fish was a major cause of decline of L. loss of oviposition sites. aurea (Pyke and White 1996). However, examination of their data reveals that pond permanency and occurrence of Broader Implications of Introduced Fish gambusia are highly correlated and so the results could also for Australian Amphibian Species be explained in terms of unmeasured features of pond The evidence presented here strongly suggests that permanency, or abundance of other predators. introduction of exotic fish or translocation of native species could have an enormous impact on the amphibian White and Ehmann (1997) suggest that gambusia is also assemblages of Australian freshwater systems. However, in implicated in the decline of L. flavipunctata, a closely related many cases the impacts have not been investigated. For species to L. aurea. However, Osborne et al. (1996) point out example, the impact of carp species on Australian amphibian that many of the sites from which this species has assemblages has not been examined, despite their disappeared do not contain gambusia. Furthermore, both widespread distribution and frequently-raised concerns about L. aurea and L. raniformis, an ecologically similar species which their adverse effects upon freshwater systems. hybridises with L. aurea (Watson and Littlejohn, 1985), have been recorded in abundance at some sites containing Introduced fish within mainland Australia are currently gambusia (van de Mortel and Goldingay 1998; Gillespie pers. generally restricted to the eastern sea board, Murray-Darling obs.; Pyke, Australian Museum, pers. comm.). system and south-west Western Australia.This distribution also overlaps with regions of high amphibian species richness The role of gambusia in the decline of L. aurea is unclear. Other (see Barker et al. 1996). Consequently a large proportion of factors require careful consideration, such as pond duration, Australian anurans are potentially affected by one or more habitat quality, presence of other aquatic predators and introduced fish species. availability of refugia. Evidence of gambusia having a major impact on the abundance of other Australian amphibians is also unclear. Species most likely to be affected are those which breed in However, many of the studies to date have demonstrated that permanent aquatic habitats, such as streams and wetlands. gambusia are capable of killing a variety of tadpole species and However, many which breed in more ephemeral habitats, eggs. Considering the wide distribution of gambusia, it probably such as billabongs and temporary pools along flood plains of does have significant impacts on some native amphibian species, rivers, may also be affected as these habitats are seasonally particularly in the eastern states where seasonal peak fish colonised by introduced fish when water courses swell. abundance coincides with the larval stages of many species Changes to the rural landscape within these regions have

138 resulted in removal of many natural ephemeral aquatic can be drained to effect 100% removal. However, this option habitats and the expansion of more permanent habitats by is usually not available. Most introduced fish in Australia occur way of stock dams.These are the only breeding habitats in in streams or larger waterbodies which cannot easily be some areas for species which would otherwise breed in drained.There have been numerous attempts in the United ephemeral water bodies. Farm dams are often stocked with States of America to eradicate unwanted fish populations introduced and native angling species which are likely to from streams and lakes, using a variety of techniques, such as impact these amphibian assemblages.These habitats may have electrofishing, netting and poisoning.The only demonstrated become ecological sinks for some species. successful approach for complete removal of fish from these systems is with a toxicant.This approach has become a A large proportion of Australia’s threatened amphibian standard management technique throughout the USA species breed in habitats currently occupied by, or within the (Eschmeyer 1975), mainly as a fishery technique to improve range of, introduced or translocated fish. Lotic species populations of recreational over non-recreational species assemblages are particularly vulnerable.The range of (Ryan 1977). More recently this has expanded to aquatic introduced trout species includes part or all of the conservation to protect threatened fauna from introduced fish distributions of ten south-eastern Australian lotic species, five species. However, examples of treatments designed to of which have declined (Tyler 1997; NSW NPWS Scientific accomplish a complete kill, as required for long-term Committee Determination Advice No. 97/27; Gillespie and exclusion, are few and, of these, only few have been successful Hines 1999). Some populations of these species are probably (Rinne et al. 1981; Gresswell 1991; Stefferud et al. 1992). exposed to redfin perch and gambusia as well. There has been only one successful eradication of salmonids The three species currently recognised within the L. aurea from any Australian waters.This was conducted in several small complex have all declined. Gambusia occur throughout much mountain streams in eastern Victoria as part of the of the range of these species. Redfin perch and carp species implementation of the barred galaxias Recovery Plan (Raadik occur throughout most of the range of L. flavipunctata, 1993). Artificial trout barriers were established across the L. raniformis and in part of the range of L. aurea. Gambusia streams and all fish above the barriers and below remaining have already been implicated in the decline of this species native fish populations were killed with rotenone, allowing the group; redfin perch and carp species may also be contributing. native species to recolonise the rehabilitated zones (Raadic, ARI, unpublished data). Expanding this approach to larger water- Other regions of Australia which contain significant amphibian courses is problematic. It is more difficult to effect a complete assemblages, but are currently free of introduced fish, such as eradication due to an exponential increase in stream length and the Wet Tropics, may be at risk in the future if further complexity with increased catchment size, barrier construction introductions of other exotic fish species occur. becomes increasingly more difficult and expensive on larger Potential for Introducing Exotic Pathogens streams, and the risk of re-introduction also increases. Saddlier and Gillespie (1997) assessed the feasibility of excluding trout Recent studies have suggested that an introduced pathogen from streams to protect populations of L. spenceri. Of the 13 may be responsible for amphibian declines in Australia and streams examined, exclusion was considered feasible only on Central America (Blaustein et al. 1994; Laurance et al. 1996; reaches of three streams because of the above constraints. If Lips 1998; Berger et al. 1998).The potential for the successful this would afford protection to approximately 7 % of introduction of disease into Australian freshwater systems via the current range of the species. the importation of fish for the aquarium trade has been clearly identified (Mckay 1984; Laurance et al. 1996). Laurance The environmental and socio-economic costs of eradicating fish et al. (1996) has suggested that a pathogen introduced in this must also be measured against the longer-term benefits to way might be responsible for frog declines in north-east conservation. Several problems arise with eradication programs. Queensland, but at this time there is no evidence to support 1. Native fish species and some invertebrate groups are also this (Hero and Gillespie 1997; Alford and Richards 1997). affected by rotenone, which disables gill function. However, disease risk imposed by the continual importation 2. Most large waterbodies such as lakes and streams are also of live freshwater fish into the country cannot be ignored. used to supply water for human consumption and recreational purposes, including fishing; hence poisoning MANAGEMENT SOLUTIONS may risk human health. Furthermore opposition by the The importation of exotic fish for the aquarium trade should recreational fishing community is likely to influence the only be acceptable following rigid quarantine protocols that political decision-making process. eliminate the possibility of introducing pathogens either with 3. For waterbodies which are used for angling, there is a high the fish or the water they are transported in.The aquarium risk of reintroduction of popular angling species by trade should be advised of the potential impact of introduced members of the public.These factors further restrict the pathogens and fish species and a shift towards the use of range of circumstances in which eradication of introduced native fish species for the pet-trade encouraged. Similarly, fish is feasible. gambusia should not be introduced into new systems for mosquito control; alternatively native fish species local to the In the future it may be possible to develop biological agents area may be more suitable. to control introduced fish populations. However, this would be extremely costly and take many years to develop. Clearly, Once fish have been introduced into an aquatic system and in many instances removal of introduced fish for maintenance established self-sustaining populations, they are extremely of amphibian populations is not feasible at this time. difficult to remove. Small, confined water bodies, such as dams, Management should focus therefore on identifying and acting

139 on those habitats where the feasibility of removal of Information is required for most fish species which have introduced fish is high, and restricting the spread of existing established self-sustaining populations and a range of exotic exotic species and further introductions of more species. and native species which are readily introduced in frog habitats. Information is urgently required on the impacts of There is a strong fishing culture in Australia which has a large redfin perch and carp species, which are widespread and focus on introduced species. Many introduced species, such as potentially affect numerous frog species, particularly in the salmonids and redfin perch, continue to be considered desirable Murray Darling system and New England Tablelands region of alien species by State agencies. However, servicing this culture NSW where several frog species have disappeared. More continues to erode the biotic integrity of Australian freshwater information is required to ascertain the impact of gambusia in systems.There needs to be a shift in emphasis by State fisheries Eastern Australia, particularly on the L. aurea complex.The managers from introduced fish to native species.The ongoing impact of trout upon all upland temperate lotic anuran species commercial stocking programs for some introduced species in in south-eastern Australia also requires further investigation. some States, such as salmonids, pose a major threat to several significant amphibian assemblages. Discontinuation of stocking Broad-scale surveys are required to determine relationships programs, especially in regions where the fish populations are of occurrence of frog species in relation to the distribution of not self-sustaining, will greatly benefit the conservation of some introduced and native fish, and a range of other biotic and amphibian species.The needs of recreational fishing must be abiotic variables.The ability of fish species to impact frog balanced with the benefits derived from maintaining natural fish communities should be tested experimentally.The relative assemblages. Control of exotic fish stocks may enhance effectiveness of tadpole survival strategies amongst species in remaining native fish stocks which are also suited to recreational the community to native sympatric predators and introduced fishing pursuits, while maintaining natural assemblages of native species should be compared. Predation experiments should invertebrates and amphibians. It is important to emphasise that include adequate replication and use of known palatable and native species should not be released into systems in which unpalatable (where available) tadpole species as controls. Fish they did not occur naturally. density, fish size and tadpole sizes that replicate field observations should also be factored into experimental CONCLUSIONS AND FUTURE DIRECTIONS designs.The ability of fish to prey upon eggs should also be examined if possible as this may be a more vulnerable life In summary, fish are a major influence on amphibian assemblage stage for some species. structure. Hence, they play a major role in determining the distribution and abundance of amphibian species.The Ascertain the relative importance of the role of introduction of exotic fish to aquatic systems has the potential introduced fish in frog population declines with relation to to eliminate amphibian species. Additionally there is potential to other biotic or abiotic factors. introduce disease or pathogens into freshwater systems.These changes may be extremely detrimental to the long-term Relative impact of introduced fish must be examined in survival of some species, undermine amphibian communities conjunction with other factors potentially limiting survival, and disrupt natural aquatic systems. In view of the large number such as habitat degradation. Other biotic or abiotic factors of introduced fish species and extensive distribution of some of may either exacerbate or ameliorate the impacts of introduce these within Australia, many amphibian communities are fish on frog populations. Surveys to determine relationships currently vulnerable to impacts from exotic fish. between occurrence of fish and frogs should incorporate collection and analysis of confounding biotic and abiotic Limited research has been carried out in Australia on the habitat variables (e.g. hydroperiod, water quality, aquatic impact of introduced fish upon amphibian assemblages. vegetation, adjacent adult habitat, native predator abundance, However, there is strong evidence from both overseas and etc.).These should be designed where possible with adequate within Australia that those fish species which have been power in sample size (i.e. adequate number of waterbodies) introduced pose a serious threat to a range of anuran to assess the relative contributions of introduced fish and species, a number of which have already declined. In other factors which significantly influence occurrence of frog particular, trout have been shown to be responsible for the species. decline of at least one threatened species (L. spenceri), and gambusia has been suggested in the decline of others.The For some species which are rare or have limited distributions, impact and management of introduced fish therefore surveys of this kind are likely to have inadequate power.The warrants serious consideration in the development and relative impact of introduced fish can be examined in field implementation of recovery plans for declining frog species. experiments conducted in stream or pond enclosures which closely mimic conditions experienced in natural breeding Further information is required to assess the impact of bodies, incorporating natural levels of cover, sympatric introduced fish upon amphibian assemblages throughout the predators and hydroperiod. One or more variables can be range of habitats and regions in which they have spread. This manipulated, in conjunction with predator levels to assess is essential to gain a proper understanding of the role of their relative contributions to larval mortality. introduced fish in frog declines, and identify management objectives. Priority should be given to the following areas of The threats imposed by introduced fish to Australian investigation: amphibian assemblages require immediate and on-going attention by conservation and fisheries managers.The Determine which introduced or translocated fish species possibility of eradication of introduced species should be are impacting upon frog communities, and which frog assessed on a case by case basis; however, this is currently species and communities are most at risk. expected to have limited feasibility in many instances.There is

140 a strong need for development of improved effective Bertram, S. and Berrill, M. (1997) Fluctuations in a northern eradication techniques. However, the immediate priority for population of gray treefrogs, Hyla versicolor. In D. M. Green managers should be prevention of further translocations and (ed.) Amphibians in Decline Canadian Studies of a Global introductions of fish species. For those species under serious Problem. Chapture 6, Herpetological Conservation No. 1, threat from introduced fish, all extant populations currently Canadian Association of Herpetologists, Montréal, free of introduced predators should be identified and pp. 57-63. appropriate steps taken to ensure that fish are not introduced. Bidwell, J. R. and Gorrie, J. R. (1995) Acute Toxicity of a Herbicide to Selected Frog Species: Final Report.Western Current policies and management of introduced fisheries and Australian Department of Environmental Protection, Perth. the aquarium-trade require review and need to take into consideration the potential impact upon amphibian Bishop, C. A. (1992) The effects of pesticides on amphibians assemblages. Enhancement of native fisheries, rather than and implications for determining causes of declines in those based upon exotic species should be encouraged. amphibian populations. In C. A. Bishop and K. E. Pettit Stocking programs for introduced fish should be discontinued (eds.) Declines in Canadian Amphibian Populations: in aquatic systems known to support vulnerable amphibian Designing an National Monitoring Strategy. Occasional species.Tighter control of the importation and maintenance Paper No. 76, Canadian Wildlife Service, Ottawa, of ornamental species is required, particularly of those pp. 67-70. species with potential to establish self-sustaining wild Blaustein, A. R., Hokit. D. G., O’Hara, R. K. and Holt, R. A. populations and impact upon native biota.The public also (1994) Pathogenic fungus contributes to amphibian losses need to be educated about adverse effects of releasing or in the Pacific Northwest. Biological conservation, 67: translocating fish on amphibians and other biota. 251-254. Blyth, B. (1994) Predation by Gambusia holbrooki on Anuran ACKNOWLEDGMENTS Larvae at the RGC Wetlands Centre, Capel Western Australia. RGC Wetlands Centre Technical Report No. 22, We thank the Arthur Rylah Institute, Department of Natural Capel,W.A. Resources and Environment,Victoria, and Griffith University for financial support. H. Hines, B. Magnusson, R. Loyn,T. Raadik, R. Bradford, D. F. (1989) Allopatric distribution of native frogs Swain and W. Osborne provided comments on the manuscript. and introduced fishes in the high Sierra Nevada lakes of California: implication of the negative effect of fish introductions. Copeia, 1989: 775-8. REFERENCES Bradford, D. E.,Tabatabai, F. and Graber, D.M. (1993) Isolation Alford, R. A. and Richards, S. J. (1997). Lack of evidence for of remaining populations of the native frog, Rana mucosa, epidemic disease as an agent in the catastrophic decline by introduced fishes in Sequoia and Kings Kanyon National of Australian rainforest frogs. Conservation Biology, Parks, California. Conservation Biology, 7: 882-8. 11: 1026-29. Braña, F., Frechilla, L. and Orizaola, G. (1996) Effect of Ahern, L. D. (1982) Threatened Wildlife in Victoria and Issues introduced fish on amphibian assemblages in mountain Related to its Conservation. Ministry for Conservation, lakes of northern Spain. Herpetological Journal, 6: 145-148 Fisheries and Wildlife Service Paper 27. Brodie Jr, E. D., Formanowicz, D. R. and Brodie III, E. D. (1978). Allen, G. R. (1982) A Field Guide to Inland Fishes of Western The development of noxiousness of Bufo americanus Australia.Western Australian Museum, Perth, Australia. tadpoles to aquatic insect predators. Herpetologica, Allen, G. R. (1989) Freshwater Fishes of Australia.T. F. H. 34: 302-306. Publications, New Jersey. Brönmark, C. and Edenhamn, P.(1994) Does the presence of Arthington, A. H. and Blühdorn, D. R. (1995) Improved fish affect the distribution of tree frogs (Hyla arborea)? Management of Exotic Aquatic Fauna: Research and Conservation Biology, 8: 841-5. Development for Australian Rivers. Land and Waters Brumley, A. R. (1991) Cyprinids in Australia, In: I. J.Winfield Resources Research and Development Corporation. and J. S. Nelson (eds) Cyprinid Fishes: Systematics, Biology Azevedo-Ramos, C.,Van Sluys, M., Hero, J.-M. and Magnusson and Exploitation, Chapman and Hall, London. pp. 265-283. W. E. (1992) Influence of tadpole movement on predation Burger,W. L. (1950) Novel aspects of the life history of two by odonate naiads. Journal of Herpetology, 26: 335-38. Ambystomas. Journal of the Tennessy Academy of Science, Barker, J., Grigg, G. and Tyler, M. J. (1995). A Field Guide to 25: 252-257. Australian Frogs. 2nd Edition. Surrey Beattie, Chipping Cadwallader, P.L. and Backhouse, G. N. (1982) A Guide to the Norton, NSW. Freshwater Fish of Victoria.Victorian Government Printing Berger, L., Spear, R., Daszak, P., Green, D. E., Cunningham, Office, Melbourne, Australia. A. A., Goggin, C. L., Hines, H. B., Lips, K. R., Marantelli, G. Caldwell, J. P.,Thorp, J. H., and Jervey,T. 0. (1980) Predator-prey and Parkes, H. (1998) Chytridiomycosis causes amphibian relationships among larval dragonflies, salamanders, and mortality associated with population declines in the frogs. Oecologia, 46: 285-9. rainforests of Australia and Central America. Proceedings Calef, G.W. (1973) Natural mortality of tadpoles in a of the National Academy of Sciences, 95: 9031-9036. population of Rana aurora. Ecology, 54: 741-58. Chovanec, A. (1992) The influence of tadpole swimming behaviour on predation by dragonfly nymphs. Amphibia- Reptilia, 13: 341-9.

141 Crump, M. L. (1984) Ontogenetic changes in vulnerability to Grubb, J. C. (1972) Differential predation by Gambusia affinis predation in tadpoles of Hyla pseudopumma. on the eggs of seven species of anuran amphibians. Herpetologica, 40: 265-71. The American Midland Naturalist, 88: 102-108. Daly, G. (1995) Observations on the green and golden Harris, K. (1995) Is there a negative relationship between bell frog (Litoria aurea) (Anura: Hylidae) in southern gambusia and tadpoles on the Northern Tablelands? B.Sc. New South Wales. Herpetofauna, 25: 2-9. Honours thesis, University of of New England, Armadale, New South Wales, Unpublished. Dankers, N. M. J. A. (1977) The ecology of an anuran community. Ph D. thesis, University of Sydney. Unpublished. Hayes, M. P.and Jennings, M. R. (1986) Decline of Ranid frog species in Western North America. Are bullfrogs (Rana Dawkins, R. and Krebs, J. R. (1979) Arms races between and catesbeiana) responsible? Journal of Herpetology, 20: within species. Proceedings of the Royal Society of 490-509. London, 205: 489-511. Healey, M.,Thompson, D. and Robertson, A. (1997) Dubuis, L. A. (1997) Effects of logging on terrestrial Amphibian communities associated with billabong habitats amphibians of coastal British Columbia. Chapture 20 In on the Murrumbidgee floodplain, Australia. Australian D. M. Green (ed) Amphibians in Decline Canadian Studies Journal of Ecology, 22: 270-78. of a Global Problem. Herpetological Conservation No. 1., Hecnar, S. J. and M’Closkey, R.T. (1997) The effects of Society for the Study of Amphibians and Reptiles, Saint predatory fish on amphibian species richness and Louis, Missouri, pp. 185-90. distribution. Biological Conservation, 79: 123-31. Duellman,W. E. (1978) The biology of an equatorial Hero, J.-M. (1991) Predation, palatability and the distribution herpetofauna in Amazonian Ecuador. University Kansas of tadpoles in the central Amazon rainforest. Unpublished Museum of Natural History Miscellaneous Publication, PhD thesis, Griffith University, Brisbane, Australia, 234 pp. 65: 1-352. Hero, J.-M., Gascon, C. and Magnusson,W. E. (1998) Direct Eschmeyer, P.H. (1975) (ed.) Rehabilitation of Fish Populations and indirect effects of predation on tadpole community with Toxicants: A symposium. North Central Division structure in the Amazon rainforest. Australian Journal of Special Publication 4. American Fisheries Society, Bethesda, Ecology, 23: 474-482. Maryland, USA. Hero, J.-M. and Gillespie, G. R. (1997). Epidemic disease and Faragher, R. A. and Harris, J. H. (1993) The historical and amphibian declines in Australia. Conservation Biology, current status of freshwater fish in New South Wales. 11: 1023-25. Australian Zoologist, 29(3-4): 166-76. Heyer,W. R., Donnelly, M. A., McDairmid, R.W., Hayek, L.-A. C. Fickling, S. (1995) The influence of fish predation on the local and Foster, M. S. (eds.) (1994) Measuring and Monitoring distribution of frogs and tadpoles in rainforest streams of Biological Diversity. Standard Methods for Amphibians. the Tully River. Unpublished Honours Thesis, James Cook Smithsonian Institution Press,Washington. University,Townsville, Australia. Heyer,W. R., McDiarmid, R.W. and Weigmann, D. L. (1975) Formas, J. R. (1995) Amphibians. Chapt. 44, In J. A. Simonetti, Tadpoles, predation and pond habitats in the tropics. M.T. K. Arroyo, A. E. Spotorno and E. Lozada (eds) Biotropica, 7: 100-11. Diversidad Biologica de Chile, pp.314-25. Holomuzki, J. R. (1995) Oviposition sites and fish-deterrent Gamradt, S. C. and Kats, L. B. (1996) Effect of introduced mechanisms of two stream anurans. Copeia, 1995: crayfish and mosquitofish on Californian newts. 607-613. Conservation Biology, 10: 1155-62. Hopey, M. E. and Petranka, J.W. (1994) Shorter contributions: Gillespie, G. R. (1997) The Biology of the Spotted Tree Frog ichthyology. Copeia, 1994(4): 1023-1025. (Litoria spenceri) and Examination of Factors Responsible Hunter, D. and Gillespie, G. R. (1999) The distribution, for Population Declines. Unpublished Report to the abundance and conservation status of riverine frogs in Biodiversity Group, Environment Australia, Canberra. Kosciuszko National Park. Australian Zoologist, 31: 198-209. Arthur Rylah Institute, Department of Natural Resources Jennings, M. R. and Hayes, M. P.(1985) Pre-1900 overharvest and Environment, Melbourne,Victoria. of the Californian red-legged frog (Rana aurora draytonii): Gillespie, G. R. and Hines, H.B. (1999) Status of temperate the inducement for bullfrog (Rana catesbeiana) riverine frogs in south-eastern Australia. Pp 109-130 in introduction. Herpetologica, 41: 94-103. Declines and Disappearances of Australian Frogs ed by Johnson, B. (1992) Habitat loss and declining amphibian A.Campbell. Environment Australia: Canberra. populations. In C. A. Bishop and K. E. Pettit (eds) Declines Gillespie, G. R. and Hollis, G. J. (1996) Distribution and habitat in Canadian Amphibian Populations: Designing an National of the spotted tree frog Litoria spenceri Dubois (Anura: Monitoring Strategy. Occasional Paper No. 76, Canadian Hylidae), and an assessment of potential causes of Wildlife Service, Ottawa, pp. 71-75. population declines.Wildlife Research, 23: 49-75. Kats, L. B., Petranka, J.W. and Sih, A. (1988) Antipredator Gresswell, R. E. (1991) Use of antimycin for removal of brook defences and the persistence of amphibian larvae with trout from a tributary of Yellowstone Lake. North fishes. Ecology, 69: 1865-1870. American Journal of Fisheries Management, 11: 83-90. Kats, L. B. and Sih, A. (1992) Oviposition site selection and Grinnell, J. and Storer,T. I. (1924) Animal life in the Yosemite. avoidance of fish by streamside salamanders. Copeia, University of California Press, Berkeley, California. 1992: 468-473.

142 Kruse, K. C. and Francis, M.G. (1977) A predation deterrent in Myers, G. S. (1965) Gambusia, the fish destroyer.Tropical Fish larvae of the bullfrog, Rana catesbeiana.Transactions of the Hobbyist, 13: 31-32. American Fisheries Society, 106: 248-252. Orchard, S. A. (1992) Amphibian population declines in British Kruse, K. C. and Stone, B.M. (1984) Largemouth bass Columbia. In C. A. Bishop and K. E. Pettit (eds) Declines in (Micropterus salmoides) learn to avoid feeding on toad Canadian Amphibian Populations: Designing an National (Bufo) tadpoles. Animal Behaviour, 32: 1035-1039. Monitoring Strategy. Occasional Paper No. 76, Canadian Laan, R. and Verboom, B. (1990) Effects of pool size and Wildlife Service, Ottawa, pp. 10-13. isolation on amphibian communities. Biological Osborne W. S, Littlejohn, M. J. and Thomson, S. A. (1996) Conservation, 54: 251-262. Former distribution of the Litoria aurea complex from the Lannoo, M. J., Lang, K.,Waltz,T. and Phillips, G. S. (1994) An southern tablelands of New South Wales and the altered amphibian assemblage: Dickinson county, Iowa, 70 Australian Capital Territory. In G. H. Pyke and W. S. years after Frank Blanchard’s survey. American Midland Osborne (eds) The Green and Golden Bell Frog (Litoria Naturalist, 131: 311-319. aurea): Biology and Conservation. Royal Zoological Society of NSW, pp. 190-98. Laurance,W. F., McDonald, K. R. and Speare, R. (1996) Epidemic disease and the catastrophic decline of Australian Petranka, J.W. (1983) Fish predation: a factor affecting the rain forest frogs. Conservation Biology, 10: 406-413. spatial distribution of a stream-breeding salamander. Copeia, 1983: 624-28. Leakey, R. and Lewin, R. (1995) The Sixth Extinction. Doubleday, New York. Pough F. H., Andrews, R. M., Cadle, J. E., Crump, M. L., Savitzky, A. H. and Wells, K. D. (1998) Herpetology. Prentice-Hall, Leslie, D. J. (1995) Moira Lake: a case study of deterioration New Jersey. of a River Murray natural resource. Unpublished M. Forestry Sc. thesis, University of Melbourne. Pyke, G. H. and White, A.W. (1996) Habitat requirements of the green and golden bell frog Litoria aurea (Anura: Liem, K. F. (1961) On the taxonomic status and the granular Hylidae) In G. H. Pyke and W. S. Osborne (eds) The Green patches of the Javanese frog Rana chalconota Schlegel. and Golden Bell Frog (Litoria aurea): Biology and Herpetologica, 17: 69-71. Conservation. Royal Zoological Society of NSW, pp.224-32. Lips, K. R. (1998) Decline of a tropical montane amphibian Raadik,T.A. (1993) A Research Recovery Plan for the Barred fauna. Conservation Biology, 12: 106-116. Galaxias, Galaxias fuscus Mack 1936, in South-eastern Liss,W. J. and Larson, G. L. (1991) Ecological effects of stocked Australia. Unpublished report to the Australian National trout on North Cascades naturally fishless lakes. Park Parks and Wildlife Service; Department of Conservation Science, 11: 22-23. and Natural Resources,Victoria. Macan,T.T. (1966).The influence of predation on the fauna of Resetarits,W. J. Jr and Wilbur, H. M. (1989) Choice of a moorland fish pond. Archiv Für Hydrobiologie, 61: 432-452. oviposition site by Hyla chrysoscelis: role of predators and Macan,T.T. (1974) Freshwater Ecology. Longmans, London. competitors. Ecology, 70: 220-28. Mahony, M. J. (1993) The status of frogs in the Watagon Reynolds, S. J. (1995) The impact of introduced mosquitofish Mountains area, the central coast of New South Wales. In (Gambusia holbrooki) on the mortality on premetamorphic D. Lunney and D. Ayers (eds.) Herpetology in Australia a anurans. Unpublished B. Sc. Honours thesis, University of diverse discipline.Transactions of the Royal Society of Western Australia, Perth,Western Australia. New South Wales, Mossman, NSW, pp. 257-264. Rinne, J. N., Minkley,W. L., and Hanson, J. N. (1981) Chemical McDowall, R. M. (ed) (1996) Freshwater Fishes of South- treatment of Ord Creek, Apache County, Arizona, to re- eastern Australia. 2nd ed., Reed, Sydney. establish Arizona Trout. Journal of Arizona-Navada McGilp, E. (1994) Distribution of anuran amphibians in the Academy of Science, 16: 74-78. lower Yarra River valley. Unpublished B. Sc. Honours thesis. Roberts, J., Chick, A., Oswald, L. and Thompson, P.(1995) University of Melbourne, Parkville,Victoria. Effect of carp, Cyprinus carpio L., an exotic benthivorous McKay, R. J. (1984) Introductions of exotic fishes in Australia. fish, on aquatic plants and water quality in experimental In:W. R. Courtenay Jr. and J. R. Stauffer Jr. (eds.) ponds. Marine and Freshwater Research, 46: 1171-80. Distributions, Biology and Management of Exotic Fishes. Robertson, P.and Gillespie, G. R. (1998 unpublished report) John Hopkins University Press, Baltimore & London pp. Draft Spotted Tree Frog Recovery Plan. Prepared for the 177-199. Biodiversity Group, Environment Australia, by Department Meffe, G. K. and Carroll, C. R. (1994) Principles of of Natural Resources and Environment, Melbourne,Victoria. Conservation Biology. Sinauer Associated, Sunderland, M. A. Robertson, P., Lowe, K. and Gillespie, G. R. (1998 unpublished Morgan, L. A. and Buttemer,W. A. (1996) Predation by the report) Flora and Fauna Guarantee Action Statement for non-native fish Gambusia holbrooki on small Litoria area and the Spotted Tree Frog, Litoria spenceri. Flora and Fauna L. dentata tadpoles. In G. H. Pyke and W. S. Osborne (eds) Branch, Department of Natural Resources and The Green and Golden Bell Frog (Litoria aurea): Biology Environment, Melbourne,Victoria. and Conservation. Royal Zoological Society of NSW, Rowland, S. J. (1989) Aspects of the history and fishery of the pp. 143-49. murray cod Maccullochella peelii (Mitchell) (Percichthyidae) Morin, P.J. (1986) Interactions between intraspecific Proceedings of the Linneaen Society of New South Wales, competition and predation in an amphibian predator-prey 111 (3): 201-13. system. Ecology, 67: 713-20.

143 Ryan, J. H. (1977) Non-game fish control at Frenchman Waldman, B. (1982) Sibling association among schooling toad Reservoir, Plumas County, California. Department of tadpoles: field evidence and implications. Animal Behaviour, Fisheries and Game, Inland Fisheries Administration 30: 700-713. Report 78-1, Sacramento, California, USA. Walters, B. (1975) Studies of interspecific predation within an Ryan, M. (Ed.). (1995) Wildlife of Greater Brisbane. amphibian community. Journal of Herpetology, 9: 267-79. Queensland Museum, Brisbane Wardell-Johnson, G. and Roberts, J. D. (1991) The survival Saddlier, S. R. and Gillespie, G. R. (1997) Feasibility Study of status of the rosea (Anura: Myobatrachidae) Trout Eradication for Spotted Tree Frog Conservation. complex in riparian corridors: biogeographical implications. Unpublished Report for New South Wales National Parks In D. A. Saunders, and R. J. Hobbs (eds) Nature and Wildlife Service, Jindabyne, Arthur Rylah Institute, Conservation 2:The Role of Corridors, Surrey Beatty & Department of Natural Resources and Environment, Sons, Chipping Norton, NSW, pp. 167-75. Victoria. Wasserug, R. (1971) On the comparative palatability of some Scott, N. J. and Limerick, S. (1983) Reptiles and amphibians. In dry-season tadpoles from Costa Rica. American Midland D. H. Janzen (ed.) Costa Rican Natural History. University Naturalist, 86: 101-9. of Chicago Press, Chicago, pp. 351-416. Watson, G. F.and Littlejohn, M. J. (1985) Patterns of Semlitsch, R. D. (1983) Effects of different predators on the distribution, speciation and vicariance biogeography of survival and development of tadpoles from the southeastern Australian amphibians. In G. Grigg, R. Shine hybridogenetic Rana esculenta complex. Oikos, 67: 40-46. and H. Ehmann (eds) Biology of Australasian Frogs and Reptiles. Royal Zoological Society of New South Wales, Sexton, O. J. and Phillips, C. (1986) A quantitative study of Sydney, pp. 91-7. fish-amphibian interactions in three ponds.Transactions of the Missouri Academy of Sciences, 20: 25-35. Watson, G. F., Littlejohn, M. J., Hero, J.-M. and Robertson, P. (1991) Conservation Status, Ecology and Management of Sih, A., J. Petranka,W. and Kats, L. B. (1988) The dynamics of the Spotted Tree Frog (Litoria spenceri) Arthur Rylah prey refuge use: a model and tests with sunfish and Institute for Environmental Research Technical Report salamander larvae. American Naturalist, 132: 463-83. Series No. 116. Department of Conservation and Smith, D. C. (1983) Factors controlling tadpole populations of Environment, Melbourne. the chorus frog (Pseudacris triseriata) on Isle Royale. Webb, C. and Joss, J. (1997) Does predation by the fish Ecology, 64: 501-10. Gambusia holbrooki (Atheriniformes: Poeciliidae) Strefferud, J. A., Propst, D. L. and Burton, G. L. (1992) Use of contribute to declining frog populations? Australian antimycin to remove rainbow trout from White Creek, Zoologist, 30(3): 316-24. New Mexico. In Hendrickson, D. A. (ed.). Proceedings of the Werner, E. E. and McPeek, M. A. (1994) Direct and indirect Desert Fishes Council.Volumes 22 and 23, 1990 and 1991, effects of predators on two anuran species along an Annual Symposia and Index for Volumes 16 through 23. environmental gradient. Ecology, 75: 1368-82. Desert Fishes Council, Bishop, California, USA, pp. 55-66. White, A. and Ehmann, H. (1997) Southern Highlands Bell Tayler, J. (1983) Orientation and flight behaviour of a neotenic Frog. In H. Ehmann (ed) Threatened Frogs of New South salamander (Ambystoma gracile) in Oregon.The American Wales. Frog and Tadpole Study Group of New South Midland Naturalist, 109: 40-49. Wales, Sydney, pp. 171-76. Tyler, M. J. (1963) A taxonomic study of amphibians and White, A.W. and Pyke, G. H. (1996) Distribution and reptiles of the central highlands of New Guinea. conservation status of the green and golden bell frog Transactions of the Royal Society of South Australia, 86: Litoria aurea in New South Wales. In G. H. Pyke and W. S. 105-30. Osborne (eds) The Green and Golden Bell Frog (Litoria Tyler, M. J. (1997) Action Plan for Australian Frogs.Wildlife aurea): Biology and Conservation. Royal Zoological Society Australia Endangered Species Program, Canberra. of NSW, pp. 177-189. Tyler,T., Liss,W. J., Ganio, L. M., Larson, G. L., Hoffman, R., Wilbur, H. M. (1984) Complex life cycles and community Deimling, E. and Lomnicky, G. (1998) Interaction between organisation in amphibians. In P.W. Price, C. N. introduced trout and larval salamanders (Ambystoma Slobodchikoff and W. S. Gaud (eds.) A New Ecology: macrodactylum) in high-elevation lakes. Conservation Novel Approaches to Interactive Systems. John Wiley, New Biology, 12: 94-105. York. pp. 195-224. van de Mortel,T. and Goldingay, R. (1998). Population Wilbur, H. M. and Fauth, J. E. (1990) Experimental aquatic food assessment of the endangered green and golden bell frog webs: interactions between two predators and two prey. Litoria aurea at Port Kembla, New South Wales. Australian American Naturalist, 135: 176-204. Zoologist, 30: 398-404. Wilbur, H. M. and Semlitsch, R. D. (1990) Ecological Voris, H. K. and Bacon, J. P.(1966) Differential predation on consequences of tail injury in Rana tadpoles. Copeia, tadpoles. Copeia, 1966: 594-98. 1990: 18-24. Waldick, R. (1997) Effects of forestry practices on amphibian Woodward, B. D. (1983) Predator-prey interactions and populations in eastern North America. Chapter 21 In D. breeding-pond use of temporary-pond species in a desert M. Green (ed) Amphibians in Decline Canadian Studies of anuran community. Ecology, 64: 1549-55. a Global Problem. Herpetological Conservation No. 1. Society for the Study of Amphibians and Reptiles, Saint Louis, Missouri. pp. 191-205.

144 Population declines and range contraction in Australian alpine frogs

William Osborne1, David Hunter1 and Greg Hollis2

ABSTRACT Most sites where L. v. alpina were still present were at the lower altitudinal limit of the species and almost all Over a period of six years (1993-1998) we undertook an extensive, broad-scale field assessment of the were associated with artificial water bodies such as status of frogs in the .The surveys dams and gravel excavation pits.The most extensive were targeted towards threatened species, and remaining populations were found on the Dargo High excluded species considered to be common. Plains in Victoria.If the observed trends in abundance continue, it is likely that the most restricted species, P. Specifically, extensive surveys were conducted for the corroboree and P.frosti, will become extinct in the near following species: Litoria verreauxii alpina (Alpine Tree future. In addition, the subspecies L. v. alpina is in Frog), Pseudophryne corroboree (Southern Corroboree immediate danger of extinction in New South Wales, Frog), P.pengilleyi (Northern Corroboree Frog) and and is highly threatened in Victoria.Programs to Philoria frosti (Baw Baw Frog). Our surveys indicated reduce current threats and, if possible, to identify the that these endemic, high-altitude species have all causes of decline are required urgently for these frogs. suffered extensive population declines, at least in part Proactive experimental management is likely to be of their range. Litoria v. alpina, P.corroboree and P.frosti important in re-establishing populations of these frogs. were found to have experienced the most dramatic declines, having disappeared from a large proportion of the sites at which they formerly were recorded.

1. Applied Ecology Research Group and CRC for Freshwater Ecology, University of Canberra. 2. Department of Natural Resources and Environment,Warragul,Victoria.

145 INTRODUCTION FIGURE 1: Location of main areas with subalpine and alpine Recent dramatic declines and extinctions of amphibians in environments in mainland south-eastern Australia.We undertook recent surveys for frogs (1995-1998) in the following locations: many countries have raised international concern about the Bimberi Range, Snowy Mountains, Davies Plain, Buffalo Plateau, likely causes of the declines (see reviews by Stebbins and Bogong High Plains, Dargo High Plains, and Baw Baw Plateau. The mountainous region to the east of Mt Buller has not been Cohen 1995; Phillips 1994).The reported disappearances and subject to recent surveys. Some symbols overlap several sites. declines of twelve species of frogs from the tropical montane rainforests of Queensland (McDonald 1990; Richards et al. 1993;Trenerry et al. 1994; Mahony 1996) and seven species from the Southern Highlands of south-eastern Australia (Gillespie et al. 1995; Osborne et al. 1996a) supports the view that at least a proportion of the declines cannot be attributed to localised anthropomorphic effects. In fact, the apparently undisturbed nature of many locations where declines have been reported has been suggested as cause for concern about possible global influences on frog populations (Blaustein and Wake 1990; Blaustein et al. 1994a;Trenerry et al. 1994).

Some processes in ecological systems can be detected only from the results of long-term studies (Spellerberg 1991; Cody and Smallwood 1996). Such studies are particularly important for developing an understanding of population dynamics (Semlitsch et al. 1996) and for modelling the demographic and environmental processes influencing populations of endangered species (Gilpin and Soule 1986; Congdon and Dunham 1994; Caughley and Dunn 1996).The concern about declining amphibian populations has raised debate about whether or not population declines are simply part of long-term population cycles.Whilst declines and extinctions have undoubtedly occurred in relatively undisturbed regions, the very remoteness and inaccessibility of these areas, combined often with a lack of a suitable unbiased survey design (Heyer et al. 1994) means that the results of surveys and monitoring are often, at face value, unreliable. In fact it has been suggested that the declining species may not have become extinct, but may survive in remote unsurveyed refugia (e.g. Crump et al. 1992; Hollis 1995; Gillespie and Hollis 1996), and that longer-term monitoring may be required to detect population cycles, particularly in frogs (Pechmann et al. 1991; Pechmann and Wilbur 1994). Unfortunately there have been few published long-term studies of frogs (Blaustein et al. 1994a).

In a recent review of the status of amphibians in the Australian Alps, Gillespie et al. (1995) concluded that nine of the 27 taxa that occur in this region are of particular conservation concern. However, the assessment of conservation status for most high-altitude species was based largely on anecdotal field observations and a decrease METHODS in wildlife atlas records for some species in recent years. In this paper, we report on our extensive field surveys that Description of the study area assessed the relative abundance and distribution of four The study was undertaken in the Australian Alps; the threatened subalpine and alpine species: Litoria verreauxii mountainous region stretching from immediately west of alpina (Alpine Tree Frog), Pseudophryne corroboree Canberra in the ACT to Mt Baw Baw in Victoria (Figure 1). (Southern Corroboree Frog), P. pengilleyi (Northern Within this region our surveys were restricted to areas above Corroboree Frog) and Philoria frosti (Baw Baw Frog). 1000 m altitude, with most surveys being conducted in the We also report on the results of long-term monitoring subalpine zone (between about 1400-1800 m) and alpine of P. corroboree, P. pengilleyi and P. frosti. zone (treeless areas above about 1800 m). Much of this region is subject to a winter covering of snow (Green and Osborne 1994; Green 1998). Precipitation is generally high,

146 particularly at higher altitudes, and frequently falls as snow were conducted by Hunter and Osborne.We surveyed most during the cooler months (Costin 1954; Land Conservation known historic sites (locations obtained from museum and Council 1977; Brown and Millner 1989).The physical and literature records, and from field notes of reliable observers) biological features of the Alps have been described in detail as well as hundreds of new sites. At each site, details of the elsewhere (Costin 1954, 1957; Land Conservation Council location and the number of pools, seepages or creek-lines 1977; Costin et al. 1979; McDougal 1982; Barlow 1986; surveyed were recorded. At sites where frogs were found the Good 1989, 1992; Green and Osborne 1994; Green 1998) number of frogs and the number of pools containing frogs or and will not be repeated here. tadpoles were recorded together with descriptive details of the habitat. Determining former distribution Potential breeding sites were visited during the day Lists of specimens held in the Australian Museum, Museum of (Pseudophryne spp., Philoria) and early evening (all species). No Victoria and the Australian National Wildlife Collection, surveys were conducted during cold or very wet weather. Canberra, were obtained and the records noted if locality Pseudophryne corroboree and P. pengilleyi were surveyed details appeared to be reliable.The date of collection of each specifically by their calling response to a loud human shout specimen was also recorded, if available, and the altitude of (Osborne and Hunter unpublished observations; see also the collection site determined from topographic maps. Osborne 1989, 1991).This technique was also used with P. Information obtained from the Victorian Wildlife Atlas frosti to confirm the calls of individuals calling sporadically. (Department of Natural Resources and Environment, Surveys for P. frosti involved listening for their advertisement Victoria) was also examined (this included many of the calls (Hollis 1995). Litoria v. alpina was surveyed at night by detailed records of L. v. alpina and P. frosti obtained from the listening for its characteristic call (Smith 1998). Calls of field notes of Littlejohn,Watson, Coventry and Malone). individuals at every location where calling was heard were tape-recorded and compared to calls of known The distribution of P. corroboree and P. pengilleyi was mapped L. v. alpina (Smith pers. comm.). At some sites the males calling previously by Osborne (1989 and unpubl. data), who could be counted individually, but where abundance was high, recorded the distribution of both species in surveys an estimate was made of the number calling. During later conducted in 1986 and 1987. Detailed surveys of the analysis of the data the following groupings of numbers calling distribution of P. frosti on the Baw Baw Plateau were also were used: 1-5, 6-10, 11-25, 26-50, 51-100, greater than 100. made in 1983 and 1984 by Malone (1985). Malone prepared detailed maps indicating the extent of the populations Tadpoles (L. v. alpina) were sampled with a dip-net or by surveyed. Although there has been no previous effort to direct visual count. Each pool in a potential breeding site was assess the distribution of L. v. alpina, Osborne previously made searched visually and by dip-netting in daylight, or was extensive field notes recording the distribution of L. v. alpina in checked at night using a small spotlight. If no tadpoles were the Snowy Mountains region from 1978 to 1987. In addition, observed, ten sweeps were made with a dip-net. Each sweep there are many museum records defining the occurrence of was made along the bottom of the pool, and through any many historic locations for this sub-species on the Bogong aquatic vegetation, to collect hidden tadpoles. Identification High Plains in Victoria and in the Kosciusko region. was made in the field or, where identification was uncertain, several individuals were collected and raised to Surveys of current distribution and relative metamorphosis. abundance Surveys for L. v. alpina and both species of Pseudophryne Following the recommendation of Gillespie et al. (1995), our were conducted throughout the Snowy Mountains, Fiery surveys concentrated on four species of frogs believed to be Range and in New South Wales. In the of concern for conservation at higher altitudes in this region: ACT and adjacent areas of NSW several locations were Pseudophryne corroboree, P. pengilleyi, Philoria frosti and Litoria surveyed along the Bimberi Range and Brindabella Range verreauxii alpina.These species have reasonably well-defined (specifically Ginini Flats, Snowy Flats, Blackfellows Gap, Rolling breeding seasons (Pengilley 1971; Green and Osborne 1994; Ground Gap, Leura Gap, Mt Bimberi, and Brumby Flats). In Hero et al. 1991; Hollis 1995) that allowed us to undertake Victoria, surveys were conducted at Davies Plain, Mount field work at times appropriate for detecting each species. Buffalo and on the Bogong High Plains. Two approaches to survey were adopted: (1) auditory censuses (call counts) (Osborne 1989; Zimmerman 1994), Long-term monitoring undertaken during the known calling season of each species; and (2) tadpole sampling (Shaffer et al. 1994), searches of Three species were subject to long-term monitoring: potential breeding pools undertaken by a combination of P. corroboree and P. pengilleyi have been monitored at varying visual search, spotlighting and dip-netting. Detailed intervals since 1986 by Osborne and Hunter (1986-1995); descriptions of survey techniques are given elsewhere: Philoria frosti has been monitored annually for the last five Osborne (1989, 1991) (P. corroboree and P. pengilleyi), years (1993-1997) by Hollis, and a decade earlier for two Hollis (1995) (P. frosti) and Hunter et al. (1998) (L. v. alpina). years (1983-1984) by Malone (1985).There was no monitoring of L. v. alpina. In most cases monitoring sites were We conducted an extensive, broad-scale survey throughout established subjectively at strategic locations throughout the the known distribution of each species.The surveys were known range of each species. Because of the rugged and conducted over five summers between September 1994 and remote nature of much of the study area, it was necessary to March 1998. Surveys at Mt Baw Baw were conducted by locate many monitoring sites within walking distance of Hollis and Hunter and surveys elsewhere in Victoria and NSW nearby roads and walking trails.

147 still present at most of these sites (Figure 3) (he was unable FIGURE 2: Former and current distribution of Pseudophryne to find the frogs in the vicinity of Hume Sawmill at the corroboree. Open circles, sites where P.corroboree were recorded by Osborne (1989) during the period 1986-1987; closed circles, northern extremity of its range). Osborne (1989) also found sites where the species was recorded by Osborne (1989) and still the species to be widely distributed and common throughout occurs; closed square, additional breeding sites at which the the Fiery Range and Bogong Mountains (Figure 3). species was detected during recent surveys (1995-1998). Some symbols overlap several sites. Philoria frosti Philoria frosti was discovered on the Baw Baw Plateau 100 years ago (1898) and, despite numerous searches in the intervening years, the species has never been found elsewhere (Malone 1985). In 1983 and 1984, Malone (1985) undertook a comprehensive assessment of the distribution of the species. He found the frogs to be widely-distributed and abundant in the western (Mt Baw Baw), central (Mt St Phillack) and north-western (Mt Whitelaw) parts of the plateau (Figure 4).There were fewer records from the eastern region (Mt St Gwinear), and the species was conspicuously absent from the south-eastern region (Figure 4). Malone estimated the population on the plateau to be over ten thousand individuals.

Litoria v. alpina There are numerous historical records that allowed us to assess the previous distribution of L. v. alpina.We obtained over 100 records of L. v. alpina from the Museum of Victoria; these included specifically the Bogong High Plains (6 locations), Dargo High Plains (34), Davies Plain (4), Lake Mountain (3), Baw Baw Plateau (14) (prior to the mid 1980’s, calling males and tadpoles of this subspecies were regularly encountered while surveying for P. frosti (Watson, Coventry, Fifteen sites were monitored for P. corroboree, eight sites for Robertson and Malone pers. comm.), Mt Buller region (3), P. pengilleyi and 24 for P. frosti. For convenience, sites are Mt Cobberas (1), Mt Hotham (3) and Mt Wellington (1).The grouped into the following sub-populations: P. corroboree — Australian Museum provided records from the Kosciusko northern Snowy Mountains (7 breeding sites), and southern region (Charlottes Pass, , Etheridge Range, Rams Snowy Mountains (8 sites); P. pengilleyi — Fiery Range Head Range, Perisher and Smiggin Holes).We also obtained (2 sites), Brindabella Range (3 sites) and Bimberi Range field note records of the subspecies from 41 locations in the (3 sites); P. frosti — plateau (21 sites), escarpment edge (4 sites).

RESULTS FIGURE 3: Former and current distribution of Pseudophryne pengilleyi. Open circles, sites where P.pengilleyi were recorded by Former distribution Osborne (1989) during the period 1986-1987; closed circles, sites where the species was recorded by Osborne (1989) and still occurs. Some symbols overlap several sites. Pseudophryne corroboree Prior to the detailed survey undertaken by Osborne (1989) there were museum records of P. corroboree from only seven 148E locations, all in the Snowy Mountains (Guthega, Smiggin Holes, Happy Jacks Plain, Round Mountain, Alpine Hut, Pretty Plain and Tooma Swamp — based on museum records). Osborne (1989) subsequently recorded the species at most of these sites (he was unable to find the species near Guthega and Alpine Hut) and in a survey that included 257 potential breeding sites recorded the species at 63 locations (Figure 2).

Pseudophryne pengilleyi Examination of museum records indicated that P. pengilleyi (recorded at that time as the northern form of P. corroboree; see Pengilley 1966; Osborne et al. 1996) was most frequently collected in the Brindabella and Bimberi Ranges near Canberra.We examined specimens from Snowy Flats, Ginini Flats, Bulls Head, Lees Spring, Coree Flats, California Flats and Hume Sawmill. Osborne (1989) found that the species was

148 Snowy Mountains (Osborne, field notes) (Figure 5).There are Current distribution and abundance also field records of L. v. alpina from four high altitude locations in the ACT (Rolling Ground Gap, Leura Gap and Pseudophryne corroboree Ginini Flats, and Snowy Flats) (Osborne, field notes). Based on A total of 170 potentially suitable breeding sites were these records it is obvious that L. v. alpina was widespread surveyed for P. corroboree. During the four summers of survey and abundant throughout much of the high country of work P. corroboree was detected at 63 different sites.These south-eastern Australia. sites were widely-spread across the known historic range of the species. Only a single individual was found in the southern-most extent of the former range, south of the Snowy River (Figure 2) — subsequent monitoring (Table 1) indicates that the species is now probably extinct in this region. Few extant populations were found along the entire FIGURE 4: Former and current distribution of Philoria frosti eastern edge of the former distribution (Figure 2) and only a breeding aggregations on the Baw Baw Plateau. Closed circles, single specimen was found at low-altitude sites near Tooma breeding sites where P.frosti still occurs, open circles, sites Dam in the northern Snowy Mountains.The central portion previously found by Malone (985) to have P.frosti present, but which no longer support the frogs (for details see Hollis 1995, of the former range, in the region just north of Mt Jagungal 1998). Some symbols overlap several sites. (encompassing the northern slopes of Mt Jagungal,Toolong Range, Round Mountain and the plains at upper Hell Hole Creek), were thoroughly surveyed each year. In this whole central region of the former distribution of the species the frogs were found at only 21 sites with the number of frogs at each site being critically low (only four sites had greater than one calling males per site; only one site had greater than ten calling males) (Figure 2).This represents an extensive collapse of the population in this region that previously was believed to be the core of the species distribution (Osborne 1988, 1989).

During the three years of repeated survey a total of 298 male frogs were recorded (based on the largest count obtained for each site within any one season). About 50 % of these frogs were recorded from only four sites (Figure 6). The remaining frogs were spread across the other sites in very low numbers, with generally between one and five calling males per site (Figure 6).

Pseudophryne pengilleyi Between 1994 and 1998, we undertook restricted surveys (mainly along vehicle trails) throughout the known range of P. pengilleyi in the Fiery Range and Bogong Mountains. More extensive surveys were conducted in the Brindabella Range and Bimberi Range (Osborne and Hunter unpubl. data).The species was still relatively abundant and widespread in the Fiery Range (Figure 3), however, we did not find it in the Yarrangobilly — Peppercorn Hill area where it was previously recorded by Pengilley (1966) and Osborne (1989). The frogs were found at breeding sites (often remote from each other) throughout suitable parts of the Brindabella and Bimberi Range, both in the ACT and contiguous areas of NSW.The numbers present at breeding sites in this region were considerably lower than was recorded by Osborne (1989 and unpubl. data) (Figure 6).

Philoria frosti During a series of extensive annual surveys carried out since 1993, we observed that there had been a considerable reduction in the abundance of P. frosti (Figure 4;Table 2; see Hollis 1995, 1997 for details). Malone (1985) recorded calling males in 73% (64 of 88) of frost hollows surveyed in 1983 and 1984, compared to 46% (22 of 48) recorded by Hollis (1995) in 1993. In a subset of 35 frost hollows surveyed in both 1983 and 1993, Malone (1985) recorded 3,694 males compared with 83 by Hollis (1995). Similarly, in a subset of

149 19 frost hollows surveyed in both 1984 and 1993, 885 males FIGURE 6: Changes in the relative abundance of Pseudophryne. were recorded in 1984 compared with 19 in 1993.We (Hollis corroboree and P.pengilleyi over a 12 year period (19986/1987, 1995) only recorded 2.2% and 2.1% of the number of calling open columns, and 1996/1998, closed columns). (a) P.corroboree; males recorded by Malone in 1983 and 1984 respectively. (b) P.pengilleyi, Brindabella – Bimberi Range population; (c) P.pengilleyi, Fiery Range population. During the subsequent four years (1994-1997) surveys were continued (Figure 4).These surveys indicated that the species is currently restricted to the western half of the Baw Baw plateau, with a contraction of the distribution from eastern and central areas. No calling males were recorded from the eastern and south-eastern regions of the plateau. In 1996 and 1997 the survey was extended to include gullies in adjacent montane forest on the southern (1966) and northern (1996/97) escarpment of the Baw Baw plateau, mostly at elevations between 1000-1300 m. In 1996, 79 calling males were recorded in 12 of the 27 gullies surveyed on the southern escarpment, and were recorded as low as 1080 m elevation. In 1997, 225 calling males were recorded in 18 of the 32 gullies surveyed on the southern escarpment, and were recorded as low as 990 m. No calling males were recorded in 34 gullies surveyed on the northern escarpment of the Baw Baw Plateau (Hollis unpubl. data). One calling male was recorded on the plateau in (a) montane forest in a catchment running north (Figure 4).

Litoria v. alpina Searches for L. v. alpina were made at 49 locations in Victoria, 92 locations in NSW (each location generally included a number of water-bodies that provided potential breeding sites), and nine in the ACT (Osborne and Hunter unpubl. data). All locations surveyed were within the Alpine National Park (Victoria), Kosciusko National Park (NSW), Bimberi Nature Reserve (NSW) and Namadgi National Park (ACT).

FIGURE 5: Former and current distribution of alpine tree frogs (Litoria verreuxii alpina) in the Snowy Mountains. Closed circles, sites where L. v. alpina were recorded during recent surveys (post 1995); open circles, sites where L. v. alpina were recorded by Osborne (unpublished field notes) between 1978 and 1987. Extensive searches for the frogs were made at each of these sites and throughout the entire central Snowy Mountains area between Kiandra and Mt Kosciusko. For further details see Hunter et al. (1997). Some symbols overlap several sites. (b)

(c)

150 Frogs identified as the sub-species Litoria v. alpina were Pseudophryne pengilleyi recorded at only seven locations in NSW.The subspecies was Long-term monitoring of P. pengilleyi was only undertaken in the only detected at three high altitude locations, all in Kosciusko Brindabella Range. Only one population, Ginini Flats — a National Park — the highest sites were at Charlotte Pass subalpine site (1600 m) in the ACT was subject to annual Village, 1780 m, and near Jacky’s Lookout, 1760 m. All other monitoring. Less-regular monitoring was undertaken at Coree sites with the sub-species were between 1200m and 1500m Flats (980 m) in NSW. Numbers present at Ginini Flats declined altitude. Despite surveying a number of historic locations on substantially during the first few years of monitoring and have the Kosciusko Main Range (Figure 5) no frogs were located in remained low ever since (Table 1). By contrast, the Coree Flats the alpine zone. population has supported a larger number of calling males (at Of the seven locations where extant populations of least during the years surveys were carried out,Table 1). L. v. alpina were located in Kosciusko National Park, four were However, monitoring at Coree Flats commenced after the major associated with artificial water bodies (see Hunter et al. 1998 drop in numbers had occurred at other sites (Table 1). Earlier for details).These ranged from small dams to reservoirs. A collecting and observations by Pengilley (1966 and pers. comm.) similar absence of tadpoles of L. v. alpina from small pools at this site indicated that the adult frog population was very large was also noted during extensive surveys in the Snowy (perhaps well over 500 individuals).The low numbers detected Mountains that we conducted for P. corroboree during January in 1998 are likely to be a direct response to the extreme and February 1997. drought conditions prevailing during the breeding season.

No L. v. alpina were found during our surveys at Baw Baw Philoria frosti Plateau, Davies Plain, and Bogong High Plains (Hunter and Twenty-four sites surveyed on the Baw Baw Plateau originally Osborne unpubl. data; Figure 7) in Victoria.We did, however, by Malone (1985) were re-surveyed annually for the locate several small populations to the south-east of Mt numbers of calling males present (Table 2). At all sites there Hotham near Dinner Plain (altitudinal range 1300 to 1600 has been a very large reduction in the numbers recorded. m), and found a more extensive population on the Dargo The mean number of calling males recorded at 24 sites High Plains (1400 to 1600 m) (Figure 7). Full details of this survey are not yet available (Hunter and Osborne in prep.).

In the Bimberi Range in the ACT searches were conducted at FIGURE 7: Location of sites surveyed for Litoria verreuxii alpina on the Bogong High Plains in 1996 and 1997. Closed circles, sites Ginini Flats and Snowy Flats during spring 1996 and no frogs where L. v alpina were found; open circles, potential breeding were recorded (a single L. verreauxii was heard calling at habitat surveyed but no frogs of this subspecies found; open Snowy Flats in November 1995, however the specimen could squares, sites of historic occurrence determined from examination of museum specimens, but not surveyed during this not be found to determine sub-specific status). During study. Some symbols overlap several sites. summer, pools were searched for tadpoles at these sites, and at the following additional sites in the Bimberi Range: Summit of Mt Bimberi (ACT), Cheyenne Flats (ACT), Leura Gap (NSW), Brumby Flats (NSW), Rolling Ground Gap (NSW), Blackfellows Gap (NSW). No tadpoles of L. verreauxii were observed at any of these sites. Further north, in the Brindabella Range, the widespread (L. v. verreauxii) was heard calling near Coree Flats, and tadpoles were subsequently found at this site.

Long-term monitoring

Pseudophryne corroboree Monitoring of 15 sites in the Snowy Mountains since 1986 indicates that there has been a substantial decline in the abundance of P. corroboree, with declines occurring in all regions. Numbers of frogs appear to have initially dropped rapidly up to 1987, with most populations then either going extinct, or remaining at very low numbers (Table 1). In the southern Snowy Mountains (south of the Snowy River; Figure 2), where all known P. corroboree sites have been monitored annually since 1996, the number of sites at which the species was detected declined from eight to one by 1991, and in 1997 and 1998 no frogs were recorded at any site (Table 1). In the northern Snowy Mountains, numbers at some sites have declined from very large choruses of over 100 individuals to less than five individuals (Table 1). Numbers of frogs recorded at breeding sites at higher altitudes in the northern Snowy Mountains remained reasonably high until 1993, some six years later than declines occurred in the southern Snowy Mountains (Figure 1).

151 TABLE 1: Relative abundance of calling male Pseudophryne corroboree and P.pengilleyi subject to occasional monitoring during the period 1986-1996. A larger selection of sites have been surveyed annually during the last few years but the results are not shown here (Osborne, Hunter, Green and Rauhala unpubl. data).

Site name 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Pseudophryne corroboree Maragle Range 15 20–50 –––––––10–15 14 10 19 Round Mountain 1 0 0 12 –801580000 Round Mtn 2 7–––000000000 Ogilives Ck 1 15 14 35 – 13 3 2 10–20 00001 Ogilives Ck 2 50–100 5––127–10100 Dargals >100 ––––––––30443295 Jagumba Fire Trail 5–10 ––––––––37913 Toolong Range 1 50–100 5 12 –––6––1255 Toolong Range 2 100–500 10–20 ––––20–50 20–50 6–10 8700 Mt Jagungal 1 50–100 –––––0––0000 Mt Jagungal 2 –––––––20–50 20–50 20–50 0 0 0 Mt Jagungal 3 20–50 ––––––––2000 Happy Jacks Plain 100–––0000000 Guthega 5032033000000 Blue Cow 4000000000000 Link Road 7505000000000 Link Road 2002000000000 Pipers Ck 1 11 422000000000 Pipers Ck 2 6010000000000 Pipers Ck 3 ––700241233100 Pipers Ck 4 1000000000000 Pipers Ck 5 2000000000000 Smiggin Holes 1000000000000 Pseudophryne pengilleyi Ginini Flats (1600m) 500+ 100–500 140 65 – 25 10 54 28 32 24 15 Coree Flats (980m) – – – 50–100 – – 100–500100–500 – – 50–100 50–100 5–10

declined from 124 individuals in 1983 to between 1.5 and 3.3 Only P. corroboree and P. pengilleyi are closely related individuals over the five years 1993 to 1997 (Table 2). In (Osborne and Norman 1991; Osborne et al. 1996b). 1983-1984 the maximum population size recorded at a monitoring site was 667 individuals, by contrast between A feature of all declining taxa in the high-country is that they 1993 and 1997 the largest population recorded was 41 breed in lentic situations such as in shallow pools or seepages. individuals. In 1997 frogs were recorded at only six of the 25 This is in contrast to other declining frog species in Australia, monitoring sites (Table 2). Monitoring of newly-discovered most of which are riverine species (e.g. see Gillespie and sites in montane forest has only been recently commenced Hines 1999).The range of pool types varies considerably: P. and is not reported here in detail. frosti breeds in very small seepage-fed depressions hidden amongst boulders and dense vegetation (Hollis 1995); P. corroboree uses shallow pools in sphagnum bogs (Osborne DISCUSSION 1990); P. pengilleyi breeds in bog pools at higher altitudes and On a broad scale, our surveys provide compelling evidence in shallow seepage pools in gullies at lower altitudes (Osborne that at least three taxa of high-altitude frogs (Litoria verreauxii 1990); L. v. alpina breeds in deeper pools, which include fens, alpina, Pseudophryne corroboree, and Philoria frosti) have stream cut-offs, lakes and reservoirs (Hunter et al. 1998). suffered serious population declines. A fourth species, Similarly, the range of life-history strategies employed varies Pseudophryne pengilleyi, has declined at higher altitudes across these species.Three of the species produce relatively (above 1400 m) but remains common at montane altitudes small clutches of eggs (P. corroboree, P. pengilleyi, P. frosti), a in the Fiery Range.These results confirm previous concern demographic feature that may be important in their response about the conservation status of these endemic taxa to changing environmental conditions. All have a tadpole stage; (e.g. Gillespie et al. 1995;Tyler 1997). the tadpole of P. frosti is non-feeding, depending instead on a large supply of yolk, tadpoles of both species of Pseudophryne As with frog declines in other parts of Australia, there does develop within the egg and do not hatch until the not appear to be any phylogenetic relationship between the terrestrial oviposition site floods in winter or the following species that have declined and those that have not declined spring, L. v. alpina lays its eggs in pools and the free-swimming (e.g. Richards et al. 1993; Mahony 1996).The frogs in decline larvae hatch within a few days of laying and the tadpoles in the highlands of south-eastern Australia do not comprise a complete development in the pools.Thus, no single aspect of natural taxonomic group; they include one hylid (L. v. alpina) the field biology of these frogs stands out as a feature in and three myobatrachids (P. corroboree, P. pengilleyi, P. frosti). common that may help to explain the declines.

152 TABLE 2: Relative abundance of calling male Philoria frosti recorded within different breeding units during surveys conducted in 1983 and 1984 (Malone 1985) and in 1993-1997 on the Baw Baw Plateau (Hollis 1995; Hollis unpubl. data). See Hollis (1997) for further details. Note that there are more sites in montane forest that are not listed here that have only been monitored for the last two years.

Breeding unit 1983 1984 1993 1994 1995 1996 1997 Plateau (P) Montane (M)

1. Access Road 1 43 21 23120M 2. Access Road 2 30 26 00000M 3. Access Road 3 0000000M 4. Chairlift 6800400M 5. Village Flat 183 149 30000P 6. Neulyne Plain 24 –22100P 7. La Trobe Plain 206 –32043P 8. Macallister Plain 82 –52010P 9. Pudding Basin 101 –21100P 10. Moondarra Flat 225 –00000P 11. Baragwanath Flat 167 245 11 4401P 12. Currawong Flat 536 –84520P 13. Currawong Flat 174 231 23000P 14. Creek Corner 52 49 00240P 15. Creek Corner A 9 12 ––501P 16. Tanjil Plain 120 –93341P 17. East Tanjil 71 –10020P 18. McMillians Flat 64 –00000P 19. The Morass 667 – 30 26 10 23 41 P 20. Freeman’s Flat 41 104 00000P 21. Wombat Flat – 18 1–––23P 22. 1 14900000P 23. Tyers River 2 7300000P 24. Mustering Flat 57 000000P 25. Gwinear Flat 93 200000P Total count for the 13 sites monitored annually 867 847 18 10 11 6 1

Do the declines represent natural montane valleys) that would prevent re-establishment of populations fluctuations or real declines? populations by means of normal dispersal.These observations strongly suggest that the observed declines in these alpine There is debate as to whether the population declines frogs are not typical of what would be expected from normal observed in amphibians in recent years represent temporary population fluctuations such as would result from decreased population fluctuations in response to variation in seasonal recruitment or local extinctions (e.g. Semlitsch et al. 1996). conditions, particularly precipitation, or are real catastrophic Moreover, the declines are similar to frog declines elsewhere declines (Blaustein 1994; Pechmann and Wilbur 1994). that have been inferred from comparisons of historic and Determining whether an observed decline conforms to one current distributions (Gillespie and Hollis 1996; Osborne et or the other of these situations ideally requires long-term al. 1996a; Fisher and Shaffer 1996). population monitoring and an understanding of the species demography (Pechmann and Wilbur 1994). In the last few The concurrence of population declines in areas as widely years we have undertaken detailed ecological studies of each separated as the Baw Baw Plateau and Snowy Mountains of the declining species that allow us to now consider this indicates that the causes are not likely to be localised.We question. Population declines in alpine and subalpine frogs in hypothesise that the factors causing the declines are due to south-eastern Australia have continued over a long enough factors operating on a broad rather than a local scale. period (greater than a decade) that we can speculate that Possible causes are discussed below. they are no longer related to normal cycles in weather patterns linked to the Southern Oscillation.The populations Is there a global influence that affects have shown no sign of recovery despite there being a lengthy high-altitude populations? period of more favourable weather conditions (1988-1996) (Osborne and Davis 1997). Existing hypotheses that have been suggested as contributing to high-altitude population declines include unusual weather On a broader scale, it is obvious that the declines have also patterns (Corn and Fogleman 1984; Pounds and Crump been extensive in a biogeographical sense.Three species 1994), acid precipitation (Dunson et al. 1992), increased UV-B (L.v. alpina, P. corroboree and P. frosti) have undergone partial radiation resulting from ozone depletion (Blaustein et al. contractions of their geographic ranges. In some situations 1994b), deposition of pesticides (Colborn and Clement the frogs have completely disappeared from some areas 1992) and virulent disease (Laurance 1996; Berger et al. separated by major biogeographic barriers (e.g. deep

153 1999).There is no evidence to date that acid precipitation is phases, while indirect effects include prevention of breeding a problem in most parts of the Australian Alps, however, activity and the lowering of the immune system making the frog there is no information available on the levels of possible more susceptible to pathogens (Pounds and Crump 1994). contaminants in precipitation and dust in this region. Despite the regions high precipitation, marked annual The catastrophic disappearance of organisms frequently fluctuations are a feature of the long-term rainfall pattern in indicates the action of a particularly virulent pathogen. the southern Australian Alps, and moderately severe droughts Laurance et al. (1996) proposed that a water-borne disease have occurred in this area previously. Records of weather may have caused declines and extinctions in populations of measured since late last century in the Snowy Mountains stream-breeding montane frogs in the wet tropics of indicate that annual precipitation has fluctuated between Australia.They further postulated that the pathogen was a periods marked with both drier and wetter conditions virus that occurred at a range of elevations but became (Osborne and Davis 1997).These oscillations generally relate highly virulent only in cool upland habitats. More recently, a to variability in the Southern Oscillation Index (McBride and fungal pathogen has gained some favour as a possible cause Nicholls 1983) and overall there has been no long-term of ongoing declines in some tropical and temperate trend of declining precipitation.The last extended dry period Australian frogs (see Berger et al. 1999). Uncertainty in (1979-1987) coincides with the period when the frogs identifying the likely causal agent emphasises the need for declined. Osborne and Davis (1997) noted that features of further research on populations of frogs in Australia. the 1979-1987 dry period that distinguished it from earlier extended dry periods identified at Kiandra included that it At least some declines have been attributed to changes in was two years longer and the mean annual precipitation long-term weather patterns (Osborne 1989; Pounds and during this period was slightly lower (less than 10 %). Annual Crump 1994; Stewart 1995; Osborne et al. 1996a), however maximum temperatures were above average for six of the other studies have ruled-out the possibility of changes in nine years in this period (based only on the 34-year record at climate being a possible cause of catastrophic declines Cabramurra in NSW).These trends are also evident in data (Laurance 1996). High-elevation regions, particularly at high obtained for the Victorian highlands and Erica near the Baw latitudes, have experienced increasing solar UV-B radiation Baw Plateau (Smith et al. in prep.). Despite these (Broomhall 1998) caused by depletion of the ozone layer. observations, we consider that it is unlikely that such slight Because potentially harmful intensities of UV-B can penetrate changes in the long-term weather patterns could result in several metres in clear freshwater (Schindler et al. 1996), well frog declines on the scale observed; especially because the below the depths used for egg laying and tadpole frogs were observed to be in much higher population levels development, there is growing concern that declines in alpine after earlier droughts.The possible interaction of lower species may relate to increased UV-B radiation (Blaustein et precipitation with increased atmospheric temperatures is al. 1994b), particularly given the increases in ultraviolet considered below. radiation experienced in alpine areas (Bluthaler and Ambach 1990). A recent study conducted in the Snowy Mountains Global warming has been considered a threat to high altitude (Broomhall et al. in press) showed that developing embryos species (e.g., Busby 1988; Bennett et al. 1991; Brereton et al. of L. v. alpina are significantly more sensitive to ambient UV-B 1995), however in south-eastern Australia there has been radiation than a non-declining species, Crinia signifera. only a very slight increase in annual maximum temperatures In artificial pools at high altitudes L. v. alpina experienced since 1951 (less than 0.1°C). Osborne and Davis (1997) very high mortality rates unless shielded from UV-B. found no appreciable trend in the annual temperature at Cabramurra in the Snowy Mountains (the longest In the discussion below we consider in more detail the temperature record available). By contrast, Davis (1998) possibility of prolonged dry weather preventing breeding or notes that there has been a very slight increase in mean dehydrating frogs over a period long enough to have caused winter maximum and minimum temperatures (May–Oct) a complete disappearance from most of the region. since the 1970’s, however this increase represents about Consideration of the hypothesis that alpine frogs have 0.5°C (calculated from data in Davis 1998). Particularly warm suffered adversely from increased levels of ultraviolet years occurred both prior to, and during, the period of the radiation is discussed elsewhere in this volume (Broomhall present declines. et al. in press). The species in the group that would be expected to be most Is there a link between long-term weather sensitive to a deterioration in long-term weather patterns patterns and frog declines? would be those with a prolonged larval development period Given the reliance of amphibians on moist environments for that might be affected by early pond-drying, or lowered both physiological maintenance and reproduction (Duellman water tables. In both P. corroboree and P. pengilleyi, the pre- and Trueb 1986) and the increasing concerns about global metamorphic stage usually lasts for ten to twelve months. climate change, it is not surprising that climate has been Philoria frosti may also be sensitive to reduced moisture implicated as a possible factor driving amphibian population availability because of its habit of breeding in very small declines in a number of species through out the world (see depressions in seepages. Possible response to changing above references). Several long term studies have correlated weather patterns is considered further below. population fluctuations in amphibians with climatic variables, If alpine frogs were responding to drier and warmer particularly annual variation in rainfall patterns (Stewart 1995; conditions it is likely that they would have disappeared Pechmann et al. 1991).The direct effect of climate on earlier from the lower-altitude, drier margins of their range. amphibians may include desiccation of both the larval and adult This feature is in fact partially evident for two species,

154 P. corroboree and P. frosti, but the reverse is true for the various components of this study. Maxine Davis assisted P. pengilleyi and L. v. alpina. Osborne and Davis (1997) noted with analysis of long-term weather records and Michael Smith that the decline in P. corroboree appears to have been confirmed all call records obtained for L. v. alpina.We thank progressive in its effect, occurring first in areas with lowest Craig Smith, Ken Green, Graeme Enders, Marjo Rauhala, Brett predicted rainfall (such as on the far eastern edge of the MacNamara, Graeme Gillespie and Gerry Marantelli for their distribution) and most recently in the wetter parts of the assistance and logistical support. John Coventry, Graeme species range (Mt Jagungal,Toolong Range and Round Gillespie, Murray Littlejohn, Brian Malone, Peter Robertson, Mountain). Interestingly, the largest remaining population Graeme Watson and John Wombey have provided much (in the Dargals Range) occurs in an area predicted to be encouragement and advice over the years.We also thank the wettest site. All low-altitude populations (below about members of the Baw Baw Frog Recovery Team and the 1400 m) have disappeared, but so too have most of the Corroboree Frog Recovery Team. highest altitude populations (above 1700 m) (such as at the many former sites near Mt Jagungal) (Osborne and Davis REFERENCES 1997; Hunter and Osborne unpubl. data).The situation with P. frosti is somewhat different. Although the species has Barlow, B.A. (ed), (1986) Flora and Fauna of Alpine Australasia. declined and disappeared from the drier eastern end of its Ages and Origins. CSIRO, Melbourne. former distribution (Hollis 1995; Figure 4), on the western Bennett, S., Brereton, R., Mansergh, I., Berwick, S., Sandford, side of the plateau, it has also declined at higher altitudes with K. and Wellington, C., (1991) The potential effect of the the largest remaining breeding aggregations occurring in enhanced greenhouse climate change on selected dense montane forest at the lower limit of the species Victorian fauna. Arthur Rylah Institute for Environmental distribution. Both P. pengilleyi and L. v. alpina have also shown Research,Victoria. greatest decline at the highest altitudes, sites that are also Berger, L., Spear, R. and Hyatt, A.D. (1999) Chytrid fungi and exposed to significantly higher precipitation. It is not at all amphibian declines: overview, implications and future clear whether there is a relationship between the extent of directions. Pp 23-33 in Declines and Disappearances of population declines in alpine frogs and long-term trends in Australian Frogs ed by A.Campbell. Environment Australia: precipitation.The potential influence of climatic change Canberra. requires further analysis, and underscores the need for Blaustein, A.R., (1994). Chicken Little or Nero’s fiddle? continued monitoring of these highly threatened species. A perspective on declining amphibian populations. Herpetologica, 50: 85-97. CONCLUSION Blaustein, A. R., and Wake, D. B., (1990) Declining amphibian Our surveys indicate that in the highlands of south-eastern populations: a global phenomenon? Trends in Ecology and Australia four species of frogs have experienced pronounced Evolution, 5:203-204. population declines.These frogs are Litoria verreauxii alpina Blaustein, A.R.,Wake, D.B. and Sousa,W.P., (1994a) Amphibian (Alpine Tree Frog), Pseudophryne corroboree (Southern declines: judging stability, persistence, and susceptibility of Corroboree Frog), P. pengilleyi (Northern Corroboree Frog) populations to local and global extinctions. Conservation and Philoria frosti (Baw Baw Frog). Pseudophryne pengilleyi is Biology, 8: 60-71. still widespread and abundant at lower altitudes, but there are Blaustein, A.R., Hoffman. P.D., Hokit, D.G., Kiesecker, J.M.Walls, few remaining substantial populations of the other three S.C. and Hays, J.B., (1994b) UV repair and resistance to species, which are faced with the likelihood of extinction in solar UV-B in amphibian eggs: a link to population declines? the short term if the current trends continue. Proceedings of the National Academy of Sciences USA, 91: 1791-1795. There appears to be an altitudinal influence on the extent of the declines, with the most serious declines all occurring at Bluthaler, M. and Ambach,W., (1990) Indication of increasing higher altitudes, particularly in the subalpine and alpine zones. solar ultraviolet-B radiation flux in alpine regions. Science, This apparent relationship between altitude and the extent of 248: 206-208. the population decline has been observed in other frog Bradford, D. F., (1991) Mass mortality and extinction in a species in Australia (Richards et al. 1993; Osborne et al. 1996; high-elevation population of Rana muscosa. Journal of Mahony 1996) and in other countries (Corn and Fogleman Herpetology, 25(2): 174-177. 1984; Bradford 1991; Carey 1993; Fellars and Drost 1993). Broomhall, S., (1998) The implications of ozone depletion for Tyler (1997) recognised the significance of this altitudinal the Australian Alps: a review. Pp 224-247 In K. Green (ed.) relationship, and recommended that further research be Snow. A natural history; an uncertain future. Australian Alps undertaken into the factors causing it.We support this view Liaison Committee, Canberra. and recommend increased multi-disciplinary research effort Brown, J.A.H. and Millner, F.C., (1989) Aspects of the and urgent experimental management aimed at stemming meteorology and hydrology of the Australian Alps. the decline in these endemic alpine amphibians. Pp 297-332 in Good, R. (ed.) The Scientific Significance of the Australian Alps.The Australian Alps National Parks ACKNOWLEDGEMENTS Liaison Committee, Canberra. We are particularly grateful to Environment Australia, the Brereton, R., Bennett, S. and Mansergh, J., (1995) Enhanced NSW National Parks and Wildlife Service, the Victorian greenhouse climate change and its potential effect on Department of Natural Resources and Environment, the selected fauna of south-eastern Australia: a trend analysis. Australian Alps National Parks Liaison Committee and the Biological Conservation, 72: 339-354. University of Canberra for providing the funding to support

155 Broomhall, S., (1998) The implications of ozone depletion for Gillespie, G.R. and Hines, H.B. (1999) Status of temperate the Australian Alps: a review. Pp 224-247 In K. Green (ed.) riverine frogs in south-eastern Australia. Pp. 109-130 in Snow a Natural History; an uncertain future. Australian Alps Declines and Disappearances of Australian Frogs ed by Liaison Committee, Canberra. A.Campbell. Environment Australia: Canberra. Busby, J.R., (1988) Potential impacts of climate change on Gilpin, M.E. and Soule, M.E., (1986) Minimum viable Australia’s flora and fauna. Pp 387-398 In G.I. Pearman populations: Processes of species extinction. Pp 19-34 In (ed.) Greenhouse: Planning for Climate Change. CSIRO, M.E. Soule (ed.) Conservation Biology:The science of scarcity Melbourne. and diversity. Sinauer Assoc., Sunderland, Massachusetts. Carey, C., (1993) Hypothesis concerning the causes of the Good, R. (ed.), (1989) The Scientific Significance of the disappearance of Boreal toads from the mountains of Australian Alps. Australian Alps National Parks Liaison Colorado. Conservation Biology, 7(2): 355-362. Committee, Canberra. Caughley, G. and Dunn, A., (1996) Conservation Biology in Good, R.B., (1992) Kosciusko Heritage.The conservation Theory and Practice. Blackwell, Cambridge, Massachusetts. significance of Kosciusko National Park. National Parks and Cody, M.L and Smallwood, J.A. (eds), (1996) Long-term Studies Wildlife Service of New South Wales, Hurstville, Sydney. of Vertebrate Communities. Academic Press, San Diego. Green, K. (ed), (1998) Snow. A natural history; an uncertain Colborn,T. and Clement, C. (eds), (1992) Chemically-induced future. Australian Alps Liaison Committee, Canberra. alterations in sexual and functional development:The Green, K. and Osborne,W.S., (1994) Wildlife of the Australian wildlife/human connection. Princeton Scientific Publishing Snow-country: A comprehensive guide to alpine fauna. Reed Co., Princeton, N.J. Books, Chatswood, N.S.W. Congdon, J.D. and Dunham, A.E., (1994) Contributions of Hero, J-M., Littlejohn, M. and Marantelli, G., (1991) Frogwatch long-term life history studies to conservation biology. Field Guide to Victorian Frogs. Department of Conservation Pp 181-194 In G.K. Meffe, C.R. Carroll (eds.) Principles of and Environment,Victoria. Conservation Biology. Sinauer Assoc., Sunderland, Heyer,W.R., Donnelly, M.A., McDiarmid, R.W., Hayek, LC. and Massachusetts. Foster, M.S. (eds), (1994) Measuring and Monitoring Corn, P.S.and Fogleman, J.C., (1984) Extinction of montane Biological Diversity. Standard Methods for Amphibians. populations of the (Rana pipiens) Smithsonian Institution Press,Washington. in Colorado. Journal of Herpetology, 18: 147-152. Hollis, G.J., (1995) Reassessment of the distribution, Costin, A.B., (1954) A Study of the Ecosystems of the Monaro abundance and habitat of the Baw Baw Frog Philoria frosti Region of New South Wales. Government Printer, Sydney. Spencer: preliminary findings. The Victorian Naturalist, Costin, A.B., (1957) The high mountain vegetation of Australia. 112: 190-201. Australian Journal of Botany, 5: 173-189. Hollis, G.J., (1997) Recovery Plan for the Baw Baw Frog Costin, A.B., Gray, M.,Totterdall, C.J., and Wimbush, D.J., (1979) (Philoria frosti). Unpublished report to the Endangered Kosciusko Alpine Flora. CSIRO/Collins, Melbourne. Species Program, Environment Australia, Canberra. Crump, M.L., Hensley, F.R., and Clark, K.L., (1992) Apparent Hunter, D.A., Osborne,W.S. and Smith, M.J., (1998) decline of the Golden Toad: underground or extinct? Distribution and abundance of the alpine tree frog (Litoria Copeia, 1992(2): 413-420. verreauxii alpina) in the Australian Alps National Parks. Report on the first seasons survey (1996-97). Unpublished Davis, C.J., (1998) Meteorological aspects of snow. Pp 3-34 In Report to the NSW National Parks and Wildlife Service. Green, K. (ed.) Snow. A natural history; an uncertain future. Applied Ecology Research Group, University of Canberra, Australian Alps Liaison Committee, Canberra. Canberra. Duellman,W.E. and Trueb, L., (1986) Biology of Amphibians. Land Conservation Council, (1977) Report on the Alpine Study Johns Hopkins University Press, U.S.A. Area. Land Conservation Council,Victoria. Dunson,W.A.,Wyman, R.L. and Corbett, E.S., (1992) A Laurance,W.F., (1996) Catastrophic declines of Australian symposium on amphibian declines and habitat acidification. rainforest frogs: is unusual weather responsible? Biological Journal of Herpetology, 26: 349-352. Conservation, 77: 203-212. Fellers, G.M. and Drost, C.A., (1993) Disappearance of the Laurance,W.F., McDonald, K.R., and Speare, R., (1996) Cascades Frog Rana cascadae at the southern end of its Epidemic disease and the catastrophic decline of Australian range, California, USA. Biological Conservation, 65: 177-181. rainforest frogs. Conservation Biology, 10: 406-413. Fisher, R.N. and Shaffer, H.B., (1996) The decline of amphibians Mahony, M., (1996) The decline of the Green and Golden Bell in California’s Great Central Valley. Conservation Biology, Frog Litoria aurea viewed in the context of declines in 10(5): 1387-1397. other Australian frogs. Australian Zoologist, 30: 237-247. Gillespie, G.R., Osborne,W.S. and McElhinney, N.A., (1995) Malone, B., (1985) Status, distribution and ecology of the Baw The Conservation Status of Frogs in the Australian Alps: Baw Frog (Philoria frosti). Arthur Rylah Institute Technical A Review. Unpublished Report to the Australian Alps Report No. 36. Department of Conservation, Forests and Liaison Committee, Canberra. Lands,Victoria. Gillespie, G.R. and Hollis, G.J., (1996) Distribution and habitat McBride, J.L. and Nicholls, N., (1983) Seasonal relationships of the spotted tree frog, Litoria spenceri Dubois (Anura: between Australian rainfall and the Southern Oscillation. Hylidae), and an assessment of the potential causes of Monthly Weather Review 111, 1998-2004. population declines. Wildlife Research, 23:49-75.

156 McDonald, K.R., (1990) Rheobatrachus Liem and Taudactylus Pengilley, R.K., (1971) Calling and associated behaviour of Straughan and Lee (Anura: Leptodactylidae) in Eungella some species of Pseudophryne (Anura: Leptodactylidae). National Park, Queensland: distribution and decline. Journal of Zoology, London, 163: 93-103. Transactions of the Royal Society of South Australia, Pengilley, R.K., (1966) The biology of the genus Pseudophryne 114: 187-194. (Anura: Leptodactylidae). Master of Science thesis, McDougall, K.L., (1982) The alpine vegetation of the Bogong Australian National University, Canberra. High Plains. Environmental Studies Publication No. 357, Phillips, K., (1994) Tracking the Vanishing Frogs. St Martin’s Ministry for Conservation,Victoria. Press, New York. Osborne,W.S., (1988) A survey of the distribution and Pounds, J.A., and Crump, M.L., (1994) Amphibian declines and habitats of Corroboree Frogs, Pseudophryne corroboree in climate disturbance: the case of the Golden Toad and the Kosciusko National Park: with a reference to ski resort Harlequin Frog. Conservation Biology, 8(1): 72-85. development. Report prepared for NSW National Parks Richards, S.J., McDonald, K.R. and Alford, R.A., (1993) Declines and Wildlife Service, Snowy Mountains Region, Jindabyne. in populations of Australia’s endemic tropical rainforest Osborne,W.S., (1989) Distribution, relative abundance and frogs. Pacific Conservation Biology: 1: 66-77. conservation status of the corroboree frogs, Pseudophryne Semlitsch, R.D., Scott, D.E., Pechmann, J.H.K. and Gibbons, J.W., corroboree Moore (Anura: Myobatrachidae): Australian Wildlife Research, 16: 537-547. (1996) Structure and dynamics of an amphibian community. Evidence from a 16-year study of a natural Osborne W.S., (1990) The conservation biology of pond. Pp 217-248 In Cody, M.L and Smallwood, J.A. (eds) Pseudophryne corroboree Moore (Anura: Myobatrachidae): Long-term Studies of Vertebrate Communities. Academic A study of insular populations. Ph.D. thesis, Australian Press, San Diego. National University, Canberra. Schindler, D.W., Curtis, P.J.,Parker, B.R., and Stainton, M.P., Osborne W.S., (1991) The biology and management of the (1996) Consequences of climate warming and lake Corroboree Frog (Pseudophryne corroboree) in NSW. acidification for UV-B penetration in North American Species Management Report Number 8, NSW National boreal lakes. Nature, 379: 705-708. Parks and Wildlife Service, Sydney. Shaffer, H.B., Alford, R.A.,Woodward, B.D., Richards, S.J., Altig, Osborne,W.S. and Davis, M.S., (1997) Long-term variability in R.G. and Gascon, C., (1994) Quantitative sampling of temperature, precipitation, and snow cover in the Snowy amphibian larvae. Pp130-141 In Heyer,W.R., Donnelly, Mountains: is there a link with the decline of the southern M.A., McDiarmid, R.W., Hayek, LC. and Foster, M.S. (eds) corroboree frog (Pseudophryne corroboree)? Report to Measuring and Monitoring Biological Diversity. Standard NSW National Parks and Wildlife Service, Snowy Methods for Amphibians. Smithsonian Institution Press, Mountains Region, Jindabyne. Washington. Osborne,W.S. and Norman, J.A. (1991) Conservation Smith, M.J., (1998) Intraspecific variation in the male genetics of Corroboree Frogs, Pseudophryne corroboree: advertisement call and morphology of Litoria verreauxii. population subdivision and genetic divergence. Australian Bachelor of Applied Science, Honours Thesis.Applied Journal of Zoology 39:285-297. Ecology Research Group, University of Canberra. Osborne,W.S., (1998) Draft Recovery Plan for the Southern Spellerberg, I. F., (1991) Monitoring Ecological Change. Corroboree frog (Pseudophryne corroboree). NSW Cambridge University Press, Cambridge. National Parks and Wildlife Service, Sydney. Stebbins, R.C. and Cohen, N.W., (1995) A Natural History of Osborne,W. S., Littlejohn, M. J. and Thomson, S. A., (1996a) Amphibians. Princeton University Press, Princeton, Former distribution and apparent disappearance of the New Jersey. Litoria aurea complex from the Southern Tablelands of New South Wales and the Australian Capital Territory. Stewart, M.M., (1995) Climate driven population fluctuations Australian Zoologist, 30: 190-198. in rainforest frogs. Journal of Herpetology, 29:437-446. Osborne W. S., Zentelis R. A. and Lau, M., (1996b) Trenerry, M.P,Laurance,W.F. and McDonald, K.R., (1994) Geographical variation in corroboree frogs, Pseudophryne Further evidence for the precipitous decline of endemic corroboree Moore (Anura: Myobatrachidae): A reappraisal rainforest frogs in tropical Australia. Pacific Conservation supports recognition of P. pengilleyi Wells and Wellington. Biology, 1: 150-153. Australian Journal of Zoology, 44: 569-587. Tyler, M.J., (1997) The Action Plan for Australian Frogs. Pechmann, J. F. K., Scott, D. E., Semlitsch, R. D., Caldwell, J. P., Environment Australia, Endangered Species Program, Vitt, L. J. and Gibbons, J.W., (1991) Declining amphibian Canberra. populations: the problems of separating human impacts Zimmerman, B.L. (1994) Audio strip transects. Pp 92-97, from natural fluctuations. Science, 253: 892-895. In Heyer,W.R., Donnelly, M.A., McDiarmid, R.K., Hayek, Pechmann, J.H.K., and Wilbur, H.M., (1994) Putting declining L.C. and Foster, M.S. (eds) Measuring and Monitoring amphibian populations in perspective: natural fluctuations Biological Diversity — Standard Methods for Amphibians. and human impacts. Herpetologica, 50: 65-84. Smithsonian Institution Press,Washington.

157 Implementation of a population augmentation project for remnant populations of the Southern Corroboree Frog (Pseudophyrne corroboree)

David Hunter1,William Osborne1, Gerry Marantelli2 and Ken Green3

ABSTRACT survivorship to metamorphosis of captive-reared and field embryos and tadpoles. Differences in the Over the past 18 years the Southern Corroboree Frog (Pseudophryne corroboree) has undergone a fitness of captive-reared versus field tadpoles are dramatic decline.While the results of extensive being measured by comparing tadpole size, surveys and monitoring suggest that P.corroboree developmental rate and date at metamorphosis. may become extinct in the near future, the factors The results of the first year’s attempt at reducing causing the decline of this species remain unknown. mortality below field levels is presented, along with A population augmentation project aimed at a discussion on the limitations and relevance of this developing techniques to reduce the rate at which project to the recovery process. small remnant populations of P.corroboree are becoming extinct was commenced in 1997. INTRODUCTION It involves increasing the size of several small The Southern Corroboree Frog (Pseudophryne corroboree) is breeding populations by means of experimental a strikingly marked species restricted to high montane and field and captive management intending to reduce sub-alpine bog habitats in the Snowy Mountains at altitudes egg and tadpole mortality. between 1300-1760 metres (Osborne 1989, Osborne et al. 1996). Like most of its congeners, P. corroboree lays its eggs in a terrestrial nest, which later floods, allowing the eggs to Initially, an assessment of our ability to reduce hatch and the tadpoles to move into an aquatic environment mortality during these early life-history stages is (Barker et al. 1995). In the early to mid 1980’s, P. corroboree being undertaken by comparing the level of underwent a dramatic decline in abundance, which resulted in

1 Applied Ecology Research Group, University of Canberra, ACT 2601, Australia. 2 Amphibian Research Centre, P.O. Box 424, Brunswick Victoria 3056, Australia. 3 New South Wales National Parks and Wildlife Service, Snowy Mountains Region, P.O. Box 2228, Jindabyne NSW 2627, Australia.

158 local population declines and extinctions throughout the substantial success at re-establishing populations at extinct range of the species (Osborne 1989, 1991). As documented sites, and enhancing small populations, was only achieved after by a long-term monitoring program, P. corroboree has failed to knowledge of the breeding site requirements and the factors show any signs of recovery since the initial population crash, which caused the decline in this species were identified and with many of the monitored populations continuing to mitigated (Denton et al. 1997). Further criticisms have also decline to extinction (Osborne 1998; Osborne et al. 1999). targeted poor experimental design and lack of adequate In 1996 an endangered species recovery team was follow-up monitoring to assess the success of repatriation established to implement the first phase of the recovery projects (Hein 1997). process as outlined in a draft Recovery Plan for this species (Osborne 1996), and in 1997 P. corroboree was assessed as Given the difficulty in determining the causal factors of recent being critically endangered using the 1994 IUCN criteria frog declines (refer to papers in this symposium), it is unlikely (Tyler 1997). that the reasons behind the decline in P. corroboree populations will be identified prior to further local extinctions (see A primary aim of endangered species recovery programs is Osborne et al. 1999 for a better discussion of hypotheses to identify the processes threatening the particular species so relating to declines in alpine frogs).There is a very real that threat abatement can then be addressed in the possibility of the species becoming extinct in the near future. development and implementation of recovery actions This situation was the impetus behind the Recovery Team for (Dickman 1996). As such, the uncertainty behind the factors P. corroboree supporting a population augmentation project. causing the decline in P. corroboree has proven problematic for Furthermore, information obtained on the demography of the recovery process. Extensive surveys conducted over the P. corroboree should aid both management and further efforts past few years have failed to locate P. corroboree at 85% of to determine the causal factors of decline in this species. In known historic localities (Osborne et al. 1999). Furthermore, this paper we present the methods employed in the project, 77% of extant populations consist of fewer than five calling discuss its limitations and the relevance of the project to the males while only three populations consist of more than 15 recovery process for P. corroboree, and present the results of calling males (Hunter and Osborne unpubl. data).These the first year of this study. results are particularly concerning given that the long term monitoring program has demonstrated a high propensity for METHODS populations consisting of fewer than five calling males to become extinct within a few years (Osborne 1998). Apart Choice of sites from the likelihood that the factors which initiated the At the outset of the study in March 1997 we were faced decline in P. corroboree populations are still operating, the with considerable logistical constraints that limited the detrimental effects associated with small population size number of populations available as experimental populations. may also be an important process working against the Only three populations were deemed suitable for population recovery of this species. augmentation during the first year of this project.These In an attempt to aid the recovery of small populations of included one population in the Dargal Range which had P. corroboree, a population augmentation project (direct 32 calling males (this site will be referred to as Site A), one manipulation of recruitment through to the terrestrial frog population from the Jugumba Range which had 13 calling stage to increase the adult population size) was commenced males (Site B) and one population near Round Mountain in 1997.There are two aims to this: which had two calling males (Site C).The exact location of these sites will not be disclosed in this paper.The number of 1. to assess whether it is possible to increase recruitment calling males refers to the number of males detected during through to the metamorphic stage via a combination of the 1997 breeding season.While Site A would not be captive-rearing and active field management to prevent considered a small population in comparison with other tadpole mortality in the field; and extant populations of P. corroboree, it was included so as to 2. to determine whether an increase in the level of increase the sample size of clutches for assessing whether we recruitment through to metamorphosis will result in an are capable of artificially reducing embryonic and tadpole increase in the breeding adult population size in small mortality. After this process, which is expected to take two to remnant populations of P. corroboree. three years, Site A will probably be removed from the experiment. As such, it was not deemed necessary to While population repatriation programs involving captive manipulate every clutch at this site and so a limit of ten breeding or rearing have been undertaken for a number of clutches was allocated for egg collection. Control populations bird and mammal species (for a review see Griffith et al. were chosen by randomly selecting sites from a list of other 1989), there have been considerably fewer projects involving known extant remnant populations. amphibians (for a review see Dodd and Seigel 1991). Such projects are often received favourably by the public. However, Field procedures their limited success has led a number of authors to question the value of repatriation projects as a conservation tool In January 1997 at each breeding location the calling sites of (Dodd and Seigal 1991; Snyder et al. 1996).This limited males were marked with flagging tape.These calling sites success has been largely attributed to a failure to remove the were then inspected in early March (after breeding had processes that caused the initial decline and a lack of finished) in order to collect half of each clutch for captive knowledge of the biological requirements of the species. rearing at the Amphibian Research Centre (ARC) in This was particularly well demonstrated with the Natterjack Melbourne. All eggs found were removed and placed in a Toad (Bufo calimita) recovery program in England where clean polypropylene container (Genfac Plastics) moistened

159 with water, and the numbers of fertile eggs, dead eggs and no more than 50 in each 180 litre tank. A drip irrigation empty capsules were counted. Half of the viable eggs were system provided for water exchange at a rate of replaced in the oviposition site and the nest carefully approximately 20% per day. returned to its original condition. Post-winter tadpole monitoring In order to monitor the survival of tadpoles from the half clutches left in the field, plastic enclosures (1.0m x 0.5m x After snow had melted in mid-spring the captive-reared 0.4 m) were constructed to link each oviposition site with an tadpoles were returned to the field where they were enclosed portion of the adjacent experimental pool.The released into field enclosures (as described above) within bottom and top were open to allow access to the pond their natal pools.To allow for a direct comparison of substrate, and to allow light and precipitation to enter the survivorship within clutches, all tadpoles released into enclosed part of the pool.The base of each enclosure was enclosures represented individuals from the same clutch. As a embedded deeply in the silt and moss in the pool, and precaution against the possibility of high mortality occurring extended into the moss bank beyond the position of the during the release process, the captive-reared tadpoles were hidden eggs.Two walls of the enclosure had large, rubberised released in two batches. Each batch represented half the fibreglass mesh-covered openings to allow for water tadpoles collected from each experimental clutch.The two movement between the enclosure and the rest of the pool. batches were released one month apart so that the first batch All materials used were inert and resistant to ultra-violet could be monitored for signs of failing to adjust to the field radiation.To investigate the effect of the field enclosure on environment. Field tadpoles, early release tadpoles and late the water temperature, Hobo data loggers (Onset Computer release tadpoles were housed in separate enclosures. For Corporation) and Tinytag data loggers (Gemini Data Loggers, assessing survivorship differences between the captive-reared Chichester, England) were placed inside and outside and field-reared tadpoles after snow melt, a census of tadpole enclosures in a number of experimental pools. numbers in each enclosure was conducted once a fortnight until metamorphosis in mid summer.Tadpole counts were Prior to snow fall in June 1997 the enclosures were checked conducted at night by torch light.The number of tadpoles in regularly to determine whether the eggs left in the field had an enclosure was assessed by dip-net removal until five hatched. A sub-sample of experimental nest sites was also minutes had elapsed without further tadpoles being found. checked for signs of excessive embryonic mortality. During each census, Gosner stage (Gosner 1960) and snout- vent length were recorded for five randomly chosen tadpoles Captive husbandry from each enclosure, or fewer if fewer tadpoles were present.

The eggs removed from the field were placed in small plastic Chi-squared contingency tests were used to test for containers packed with sphagnum moss and immediately significant differences between mortality levels both between transported by car to the ARC (an approximately six hour the treatments within sites, and between sites within drive). Upon arriving at the ARC, the eggs were removed treatments. A two-way ANOVA was used to compare from containers, further divided and placed into either empty differences in the size of tadpoles at Gosner stage 41 for clean containers or containers containing natural nest both between-treatments and between-sites.To meet the material.These containers were each floated inside a slightly assumption of homogeneity of variance, the measurements larger container containing sterilised water which in turn was were log transformed and three outliers were removed.Two placed in a water-bath.The eggs were maintained in a analyses were undertaken.The first was to test for size moistened terrestrial state, with the contact of each container differences between treatments and involved removing Site B within the water bath providing thermal profiles similar to from the analysis as this site lacked the field treatment due to that which might be expected in natural nesting sites. 100% winter mortality.The second was conducted to test for The entire system was enclosed within a controlled differences in tadpole size between sites and involved temperature room (a modified shipping container) with removing the field treatments of Site A and C from the ambient temperature set to approximate recorded field analysis, again due to the lack of a field treatment at Site B. temperatures based on data collected from Site A using two Hobo data loggers in winter 1996.The 1996 recordings Disease prevention protocol did not start until the beginning of June so that temperatures in the container up until June 1997 were based on spot A strict set of protocols was maintained for the purpose of temperature records from the field and updates from the minimising the potential spread of pathogens both from the field data loggers when they could be accessed. captive environment to the field, and between experimental sites. All equipment used at the experimental sites was either On hatching the tadpoles were moved to glass aquaria sterilised prior to use or was a new item that had never been (1 000 X 300 X 750mm) containers within the constant used in the field or with captive frogs. Prior to undertaking temperature room.Tadpoles from each pool were housed fieldwork in the area of the experimental sites footwear was together and each clutch was separated by screens made sterilised in bleach. Further precautions were taken at Site A from anodised aluminium and rubberised fibreglass mesh. where, prior to entering the sites, footwear were changed for To mimic the field environment, individual tadpole enclosures gumboots which were housed adjacent to the site. contained only material obtained from the natal pool of the respective tadpoles. All water used was filtered by reverse osmosis then buffered with aquarium salt to approximate mineral content of natural alpine bogs.Tadpoles were housed at below recorded field densities (Osborne pers. obs.) with

160 To prevent contamination of ex-situ work, this project was RESULTS maintained in isolation from the rest of the frog collection at the ARC. It had an independent water supply, a set of Comparison of survivorship between field dedicated equipment and was housed at a separate site. Only and captive treatments new, sterilised or field sourced materials were taken inside Collection of embryos from the field took place on 15 and the room, and a routine of showering, changing into clean 16 March, 1997, approximately six weeks from the end of the clothes and dedicated boots was practised before entry. breeding season. Of the 25 male nests investigated at the As a further precaution before release, 10 tadpoles (one three sites, 16 contained eggs (11 from Site A, 4 from Site B from each of the 10 natal pools represented in the study) and 1 from Site C). From these 16 nests, 374 eggs were were killed, preserved in 10% neutral buffered formalin and collected for captive rearing while 324 eggs were left in the sent to the CSIRO Australian Animal Health Laboratory for male nests for the field comparison. examination.This material was sectioned, stained and examined for signs of disease. Only seven of the tadpoles could be fully The greatest level of field mortality at all three sites occurred examined.These and the available sections of the other three during the over-winter stage, with Site B experiencing total showed no evidence of disease (Berger pers. comm.). mortality (Figure 1). On the other hand, the greatest level of mortality for the captive-reared tadpoles occurred during the Maintenance of pool water levels post-winter tadpole stage (Figure 1). After adjusting for the number of individuals removed from the captive stock for The first year of this project coincided with the drought pathogen screening and further captive husbandry research, during the spring/summer of 1997/98.To avoid the drying of the total number of animals which survived from collection pools and subsequent loss of tadpoles, a number of techniques to metamorphosis was significantly higher for the captive- were used to maintain water in pools at the experimental sites. reared tadpoles than those left in the field at Site B (X2 = At Site A, both a solar and petrol water pump were used to 43.8, P < 0.01, d.f. = 1) and Site C (X2 = 15.86, P < 0.01, d.f. pump water from a nearby stream into a large polyethylene = 1) whereas there was no significant difference between the tank. From this tank the water was then fed into the various captive-reared and field survivorship for Site A (X2 = 0.25, pools that contained experimental enclosures by gravity.The P > 0.1, d.f. = 1). As this analysis used the total number of breeding pools at Site B are slow flowing seepage lines which tadpoles which survived within a site, the lack of a significant allowed adequate water levels to be maintained by damming difference between the captive-reared and field tadpoles for the out-flow end of the pool. At Site C, gravity-fed water from Site A was largely due to one large clutch attaining a high a nearby stream maintained an adequate water level in the level of survivorship in the field.The mean clutch survivorship single breeding pool with enclosures. at Site A was higher for the captive-reared tadpoles (33.3%) than the field tadpoles (15.2%). Of the eggs removed from Site A, 38% survived through to metamorphosis compared with 31% survivorship to metamorphosis in the field animals. At Site B there was no survivorship through to metamorphosis for the field-reared animals whereas 53% of the captive-reared animals survived through to metamorphosis. At Site C 70%

FIGURE 1: Mortality levels between captive reared and field embryos and tadpoles at the three experimental sites. Shaded bars represent embryonic mortality at the time of egg collection, open bars represent the level of mortality between egg collection and spring snow melt and closed bars represent the level of post-winter tadpole mortality.

161 FIGURE 2: Post-winter mortality curves for the field, early and late release tadpoles at the three experimental sites. Diamonds represent field tadpoles, squares represent early release tadpoles while triangles represent late release tadpoles.

of the captive reared tadpoles reached metamorphosis three sites was usually about two weeks apart between the compared to 13% survivorship to metamorphosis in the field, early and late-release tadpoles (Figure 2).The exception field-reared animals. to this was the similar date at metamorphosis for the early and late-release tadpoles at Site C. The field tadpoles displayed a higher level of survivorship than the captive-reared tadpoles during the post-winter stage The water temperatures recorded inside and outside (Figure 1).The greatest level of mortality for the captive- enclosures from four pools are given in Table 1.There were reared tadpoles was during the first two weeks after their only minor differences between the temperature inside and release (Figure 2). Furthermore, there was a greater level of outside the enclosures with the greatest differences being survivorship for the early-release tadpoles than the late- observed between enclosures. release tadpoles (Figure 2). There was a significant difference in tadpole size (measured The post-winter developmental rate of tadpoles showed a at Gosner stage 41) between the three experimental sites similar pattern across the three experimental sites (Figure 3), (Figure 4) with tadpoles at Site C being the largest and Site B with the late-release tadpoles appearing to develop at a faster being the smallest (Table 2).There were no significant rate than the early-release tadpoles, which in turn developed difference observed between the field, early release or late faster than the field tadpoles. Date of metamorphosis at all release tadpoles (Table 2).

162 FIGURE 3: Developmental rates for the field, early and late release tadpoles at the three experimental sites. Diamonds represent field tadpoles, squares represent early release tadpoles while triangles represent late release tadpoles.

DISCUSSION levels of mortality between over-wintering eggs compared to over-wintering tadpoles are currently being undertaken. Comparison between field-reared and captive reared tadpoles The low level of recruitment through to the post-winter tadpole stage observed in this study may have been a feature The monitoring results from the first year of this project have of other P. corroboree populations as a survey conducted in confirmed our ability to increase recruitment significantly spring at seven other remnant populations failed to locate through to the terrestrial frog stage via captive rearing.The tadpoles at five of these (Hunter and Osborne unpubl. data). difference in survivorship between the captive-reared and Research conducted during the mid 1960’s on both field animals was largely due to the higher level of field P. corroboree and the Northern Corroboree Frog P. pengilleyi mortality experienced at all three sites during the winter (then considered to be P. corroboree) documented high annual (Figure 1).Whether winter mortality in the field occurred in variability in early life history mortality as a result of varying the egg or tadpole stage is unknown. Given that the low climatic conditions between seasons (Pengilley 1992). Given autumn precipitation resulted in very few nest sites being that the winter field mortality observed during this study may flooded prior to snowfall, and considering that remains of have been a product of the poor climatic conditions dead eggs were found in a number of nest sites the following experienced during the 1997 winter (ie. low autumn rainfall spring, much of this mortality may have occurred prior to the and late snow cover), there is a need to establish our ability eggs hatching. Further experiments investigating different to reduce mortality artificially during climatically good years.

163 While over winter mortality was greater in the field than in The relationship between date and size at metamorphosis in captivity, the opposite occurred during the post-winter stage amphibians is considered to be of importance to both where a greater level of mortality was exhibited by the juvenile survivorship and adult fitness (Semlitsch et al. 1988). captive-reared tadpoles (Figure 1). A number of factors may As such, any differences between captive-reared and field- have contributed to this result, including possible negative reared tadpoles for these features may result in differences in effects of the captive rearing process.The greater mortality the fitness of individuals between the two treatments.While exhibited by the late-release tadpoles than the early-release there was no significant difference between the size of tadpoles (Figure 2) indicates that a greater level of captive-reared and field tadpoles within sites there was a survivorship may be attained by releasing tadpoles earlier. significant difference in the size of tadpoles between the Considerable variation in mortality levels was also observed three sites (Figure 3).Tadpoles often exhibit a high level of between sites in both the field and captive-reared treatments phenotypic plasticity for a number of traits, including size (Figure 1).While this variation between sites may have been (Duellman and Trueb 1994) and as such it would be due to both the low sample size of clutches being monitored expected that size differences may be observed between and local environmental and genetic effects, it also indicates sites as a results of local environmental effects. the potential for high levels of early life-history mortality within these small remnant populations. Because lower temperatures may strongly correlate with slower developmental rates in tadpoles (Duellman and Trueb 1994), temperature may have contributed to the differences in the developmental rates observed between the field, early release and late release tadpoles (Figure 3).This possibility is supported by the observation of an increase in both tadpole TABLE 1: Comparison between temperatures recorded inside developmental rate (Figure 3) and temperature (Table 1) as and outside of enclosures. All figures are average temperatures with minimum and maximum in parenthesis. the season progressed further into summer. Even though the captive-reared tadpoles developed at a faster rate, the field tadpoles still metamorphosed at least two weeks earlier Oct/Nov Dec (Figure 3).This difference was due to the captive-reared Round Mt. Enclosure 8.4 (4.7/22.5) 9.2 (5.0/32.9) tadpoles being slightly less developed than the field tadpoles Round Mt. Free 8.7 (-0.4/23.9) 9.1 (5.2/34.3) at the time of release. It is difficult to speculate how this Snakey A Enclosure 13.6 (3.9/30.5) 15.5 (3.9/34.2) difference in date at metamorphosis may affect the fitness of Snakey A Free 13.8 (2.9/33.8) 14.7 (-2.7/35.3) the captive-reared animals. In any case, date at Snakey 9 Enclosure 13.1 (0.9/44.1) 15.1 (5.8/46.2) metamorphosis for the captive-reared tadpoles occurred Snakey 9 Free 13.3 (5.5/33.8) 15.8 (9.9/34.8) within the period tadpoles were observed metamorphosing Snakey B Enclosure 11.4 (5.0/20.5) 13.9 (7.5/22.4) at other non-experimental sites (Hunter pers obs.). Snakey B Free 13.5 (-1.7/32.7) 14.5 (2.6/35.2)

FIGURE 4: Comparison of snout-vent-length at Gosner stage 41 between the field, early and late-release tadpoles at the three experimental sites. Diamonds represent field tadpoles, squares represent early release tadpoles while triangles represent late release tadpoles. Error bars represent 95% confidence intervals.

164 TABLE 2: Results of the two-way ANOVA undertaken to investigate differences in the size of tadpoles at Gosner stage 41 both between sites and between treatments. Analysis (a) was undertaken to investigate size differences between treatments and involved removing Site B from the analysis, while analysis (b) was undertaken to investigate size differences between sites and involved removing the field treatments from the analysis.

Sum of Mean Source DF Squares Square F Value P>F

Analysis (a) Among cells 5 0.0114 0.0023 4.17 0.0026 Treatment 2 0.0012 0.0006 1.09 0.3430 Site 1 0.0079 0.0079 14.41 0.0003 Treatment*Site 2 0.0005 0.0003 0.60 0.6052 Within 59 0.0322 0.0005 Total 64 0.0437 Analysis (b) Among cells 5 0.0296 0.0060 14.62 0.0001 Treatment 1 0.0013 0.0013 3.23 0.0771 Site 2 0.0260 0.0132 32.68 0.0001 Treatment*Site 2 0.0002 0.0001 0.27 0.7637 Within 63 0.02552 0.0004 Total 68 0.05512

Assumptions and limitations of this project Furthermore, both this study, and other research conducted prior to the decline of P. corroboree (Pengilley 1992), have The rationale behind the approach taken in our research has indicated the potential for high levels of within-season been based on generalisations about amphibian population embryonic mortality, with some populations experiencing dynamics. Assuming that the process of captive-rearing does total embryonic mortality.These findings suggest that early not have a negative impact on the fitness of released life-history mortality may significantly contribute to population individuals, the results will be strongly influenced by the regulation and, as such, support our attempt to reduce extent to which mortality at the egg and tadpole stage mortality during this stage as a means of increasing the size of contributes to the current regulation of population size. adult populations. However, even if egg and tadpole mortality is a key factor in determining population size, an increase in recruitment through In comparing survival between the captive-reared and field to the terrestrial frog stage in the experimental populations is tadpoles within the same sites we have compromised our greatly restricted by the low number of breeding adults within capacity to relate future changes in the size of experimental these sites and the relatively small clutch size for this species. populations to the efforts of captive rearing.This is because As such, both the control and experimental populations may we will not be able to differentiate between individuals that remain equally susceptible to stochastic levels of juvenile and were captive-reared and individuals that were left in the field. sub-adult mortality, despite reduced egg and tadpole mortality Ideally, marking cohorts would be undertaken so that upon in experimental populations. reaching maturity they could be identified as either captive- reared or field individuals, however, this would be very difficult Our reluctance to extend the use of captive rearing to bridge because P. corroboree tadpoles become extremely cryptic mortality levels during the juvenile and sub-adult stages is a upon approaching metamorphosis. A more appropriate response to a lack of knowledge of the husbandry approach may have been to compare captive versus field requirements of the post-metamorphic stages. Moreover, survivorship between sites rather than within sites.This, returning individuals to the field as early stage tadpoles however, would have required the use of a greater number of reduces the possibility of other complications resulting from sites than is currently available. Also the comparison of the captive environment, such as artificial selection or the tadpole survivorship between treatments would not control need to imprint on the natal pool (Bloxam and Tonge 1995, for the variation in survivorship between sites (Figure 1). Dodd and Seigel 1991). At this stage, this limitation is considered an inherent part of While information on age-specific schedules of mortality for the initial phase of this project as priority has been given to P. corroboree is lacking, demographic information obtained refining the methodology and obtaining information to aid since the commencement of this project does lend some other aspects of the recovery process. support to the appropriateness of the current approach. Preliminary data on the age structure of extant populations of P. corroboree has indicated the presence of both young and old frogs at breeding sites (Hunter unpubl. data).This suggests that at least some level of recruitment into the breeding population has been occurring in recent years, along with reproductively mature adults surviving for several years.

165 Relevance of this project to the As a result of locating several additional populations of P. recovery process corroboree in suitable areas during the 1998 breeding season, an additional four experimental populations and four control The need to identify and mitigate the processes causing the populations were incorporated into the project during the decline in a species before attempting repatriation projects second year.This will greatly enhance our ability to relate has been strongly emphasised in the literature (Dodd and changes in the size of experimental populations to the process Seigel 1991; Griffith et al. 1989).While this approach is of reducing early life-history mortality. Our ability to obtain sensible, it may not always be an appropriate prerequisite as it meaningful results may also be influenced by the duration of this does not recognise the potential contribution of repatriation project.This is due to the possibility that P. corroboree may live experiments to other research and recovery actions for up to six years as a breeding adult (Hunter unpubl. data), (Armstrong et al. 1994, Soderquist 1994).With respect to the and, as such, successive years of population augmentation may recovery process for P. corroboree, hypotheses relating to the have an accumulative effect on the breeding population size. decline of this species have been formulated in the absence of The process of increasing survivorship through to any information on which life history stage(s) has suffered an metamorphosis needs to continue for a further three years. In increase in mortality.This greatly limits our ability to allocate addition, the development of a long term monitoring program priorities of research into the causal factors of decline. Even for P. corroboree by the New South Wales National Parks and so, those hypotheses which have been proposed are either Wildlife Service should provide the commitment and continuity virtually impossible to test using controlled scientific necessary for assessing the outcomes of this project. experiments (ie. those examining habitat or climate change) or require relatively high numbers of individuals for experimentation which are currently not available (ie. those ACKNOWLEDGMENTS examining disease or increased levels of UV-B radiation). This project has been funded by the Endangered Species Program of Environment Australia, the New South Wales Given the constraints currently imposed on the recovery National Parks and Wildlife Service, the University of process for P. corroboree, the potential contribution from the Canberra and the Amphibian Research Centre. A special population augmentation project will be much broader than thanks to members of the Victorian Frog Group, particularly just attempting to increase the size of small remnant Rebecca Hirst, for undertaking further fundraising to assist populations.The development of husbandry techniques and this project. For assistance with either fieldwork and or the reduction of mortality will allow more scope for conducting captive husbandry we thank Joanne Doherty, Graeme Enders, experiments and establishing a captive breeding colony Narelle Freestone, Jason Kirby, Fiona Morrow, Dave Reznick, without having to exert undue levels of harvesting pressure Griff Rose, Craig Smith, Mike Smith and Quinton Smith. on extant populations. Data obtained on the current levels of Thanks also to David Judge for statistical advice.We are embryonic and tadpole mortality will also assist in indebted to Dave Lawrence and his staff at the NSW determining whether early life history mortality is significantly National Parks and Wildlife Service office in Khancoban for contributing to the continued decline in P. corroboree.This their tireless efforts in maintaining water in breeding pools. information is particularly important for guiding research and Helpful comments on this manuscript were provided by management actions, because the current factors regulating Mani Berghout, Keith McDonald and Alastair Campbell. the population size of P. corroboree are not necessarily those This project is the result of continued support and guidance that caused the initial decline in this species. by the Southern Corroboree Frog Recovery Team.

FUTURE DIRECTIONS REFERENCES To account for annual variation in field mortality rates, and Armstrong, D. P., Soderquist,T., and Southgate, R. (1994). hence our ability to decrease embryonic and tadpole Designing experimental reintroductions as experiments. mortality artificially below field levels, further comparisons of Pp. 27-29 in Reintroduction Biology of Australian and New field versus captive survivorship will be undertaken during the Zealand Fauna, ed by M. Serrena. Surrey Beatty & Son, second field season. In addition to this comparison, Chipping Norton. techniques aimed at reducing early life-history mortality without having to remove eggs from their natal pools will be Barker, J., Grigg, G.C. and Tyler, M.J. (1995) A Field Guide to investigated.The successful development of such a technique Australian Frogs. Surrey Beattie & Sons. will hopefully reduce the need to remove eggs into captivity Bloxam, Q. M. C. and Tonge, S. J. (1995). Amphibians: suitable and allow greater allocation of resources in this area to candidates for breeding-release programs. Biodiversity and developing a breeding program for P. corroboree.The Conservation, 4: 636-644. production of high numbers of tadpoles, through captive Denton, J. S., Hitchings, S. P., Beebee,T. J. C. and Gent, A. breeding, for release into experimental populations may (1997). A recovery program for the Natterjack Toad (Bufo reduce the possibility that stochastic processes will override calamita) in Britain. Conservation Biology, 11: 1329-1338. our efforts of increasing survivorship. Successful breeding of P. Dickman, C. R. (1996). Incorporating science into recovery corroboree may also provide greater potential for conducting planning for threatened species. Pp. 63-73. in Back from experiments into the causal factors of decline and attempting the Brink: Refining the threatened species recovery to establish P. corroboree at former breeding locations. process. Ed. S. Stephens and S. Maxwell. Surrey Beatty & Son, Chipping Norton.

166 Dodd, C. K. and Seigel, R.A. (1991). Relocation, repatriation, Osborne,W. S. and Norman, J. E. (1991). Conservation and translocation of amphibians and reptiles: are they genetics of corroboree frogs, Pseudophryne corroboree: conservation strategies that work? Herpetologica, population subdivision and genetic divergence. Australian 47: 336-350. Journal of Zoology, 39: 285-297. Duellman,W.E. and Trueb, L. (1994) Biology of Amphibians. Osborne W. S., Zentelis R. A. and Lau, M., (1996) (2nd Ed.). John Hopkins University Press: Maryland, USA. Geographical variation in corroboree frogs, Pseudophryne Gosner, K. L. (1960). A simplified table for staging anuran corroboree Moore (Anura: Myobatrachidae): A reappraisal embryos and larvae with notes on identification. supports recognition of P. pengilleyi Wells and Wellington. Herpetologica, 16: 183-190. Australian Journal of Zoology, 44: 569-587. Griffith, B., Scott, J. M., Carpenter, J.W. and Reed, C. (1989). Pengilley, R. K. (1992). Natural history of Pseudophryne spp. Translocation as a species conservation tool: status and (Anura: Myobatrachidae) in the Southern Highlands of strategy. Science, 245: 477-480. NSW, Australia. Sydney Basin Naturalist, 1: 9-29. Hein, E.W. (1997) Improving translocation programs. Rienert, H.K. (1991).Translocation as a conservation strategy Conservation Biology, 11: 1270-1274. for amphibians and reptiles: some comments, concerns, and observations. Herpetologica, 47: 357-363. Osborne,W.S. (1989). Distribution, relative abundance and conservation status of Corroboree Frogs, Pseudophryne Semlitsch, R. D., Scott, D. E. and Pechmann, H. K. (1988).Time corroboree Moore (Anura: Myobatrachidae). Australian and size at metamorphosis related to adult fitness in Wildlife Research 16: 537-547. Ambystoma Talpoideum. Ecology, 69: 184-192. Osborne W.S., (1991) The biology and management of the Snyder, N.F.R., Derrickson, S.R., Beissinger, S.R.Wiley, J.W., Corroboree Frog (Pseudophryne corroboree) in NSW. Smith,T.B.,Toone,W.D. and Miller, B. (1996). Limitations of Species Management Report Number 8, NSW National captive breeding in endangered species recovery. Parks and Wildlife Service, Sydney. Conservation Biology, 10: 338-348. Osborne,W. S. (1996). Draft Recovery Plan for the Southern Soderquist,T. R. (1994).The importance of hypothesis testing Corroboree Frog. (Pseudophryne corroboree). Unpublished in reintroduction biology: examples from the report to the NSW National Parks and Wildlife Service. reintroduction of the carnivorous marsupial Phascogale tapoatafa. Pp. 159-164 in Reintroduction biology of Osborne,W. S. (1998). Recovery Plan for the Southern Australian and New Zealand fauna, ed by M. Serrena. Corroboree Frog (Pseudophryne corroboree) 1998-2001. Surrey Beatty & Son, Chipping Norton. NSW National Parks and Wildlife Service, Sydney. Tyler, M. J. (1997). Action Plan for Australian Frogs.Wildlife Osborne,W. S., Hunter, D.A. and Hollis, G.J. (1999) Population Australia, Endangered Species Program: Canberra. declines and range contraction in Australian alpine frogs. Pp 145-157 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra.

167 Husbandry: science or art? — Are captive technologies ready to contribute to recovery processes for Australian frogs?

Gerry Marantelli *

ABSTRACT have already outstripped our supply of information. It is suggested that efforts be made to rigorously The history of uses for frog husbandry in Australia investigate husbandry of as many species as possible, is extremely varied. Frogs are held in captivity for a and priority groups are outlined. Finally it is wide range of reasons. Such holding leads to both determined that care must be taken to ensure risks a need for, and the development of, husbandry are mitigated and needs are real before embarking technologies.This history, and its combined captive on captive conservation programs. achievements are reviewed.The risks and benefits of captive technologies as tools in conservation are INTRODUCTION summarised and priorities and protocols for the use The need to utilise all available resources in the effort to stem of husbandry in conservation are proposed. amphibian declines has seen a significant amount of attention focused on husbandry and the use of captivity as a tool in the The potential for captive husbandry to contribute to conservation process.The usefulness of husbandry, or more the conservation of Australia’s frogs is discussed, specifically captive breeding, as a conservation tool has been the subject of much debate, with some authors suggesting that many along with the roles of stakeholders, and suggestions complex issues are simply overlooked (see Gippoliti and for maximising their cooperative contributions with Carpaneto 1997; Snyder et al. 1996; Snyder et al. 1997). A illustrations drawn from case studies. It is concluded number of projects employing captive breeding or rearing to re- establish or bolster wild populations of frogs have been that we should accelerate our pace in this area of attempted with varying levels of success (Banks 1996; Denton et research as the needs for husbandry technologies al. 1997; Hunter et al. 1999;Tonge and Bloxam 1989). By their

* Amphibian Research Centre, PO Box 424, Brunswick,Victoria 3056.

168 nature, conservation projects incorporating captive care achieve II. Conducting investigations to assist in ecological among the highest profiles of all conservation efforts.They place research or recovery plans. conservation actions within population centres and provide the III. Incidental exposure to observations not readily public and media with easy access to otherwise ‘invisible’ species. accessible to field biologists. Such access and exposure brings with it the opportunity to IV. Producing technologies to reduce effects related to educate and influence; as well as to secure funds and resources confinement that may otherwise confound research. not readily available to other less popularised or more remotely 4. Generating funding for conservation and research by: located conservation efforts (Kleiman and Mallinson 1998; I. Sales of frogs, tadpoles and expertise to end users Snyder et al. 1996). If not carefully handled, this exposure may of frogs. also lead to an unrealistically optimistic estimation of the true II. Attracting sponsorship and public support for high value of husbandry in such conservation projects (Snyder et al. profile conservation programs. 1996). Potential stakeholders in captive conservation represent a III. Indirectly generating funds for all frog research by multitude of different interests and disciplines.The effective increasing community awareness and exposure. networking and correct utilisation of all such parties remains the 5. Captive breeding or raising of declining frog species to most effective way to ensure beneficial outcomes from the use produce stock for: of husbandry in the conservation of Australia’s frogs. I. Decline hypothesis testing. THE ROLE OF HUSBANDRY IN II. Translocation experiments. III. Re-release to the wild. CONSERVATION AND RESEARCH The Action Plan for Australian Frogs (Tyler 1997) estimates a RISKS ASSOCIATED WITH HUSBANDRY total of A.$5 million dollars should be committed over five RELATED CONSERVATION ACTIONS years to aid the recovery of Australia’s threatened frogs. Any manipulation of natural systems is associated with risk. 20% – $1 million, has been earmarked for husbandry-related Simply collecting specimens could put pressure on wild actions.There are also additional connections between populations in some cases. Introduction of additional animals husbandry and other research-based actions listed, including and manipulation of mortality schedules, which are often the toxicology and disease research. aim of captive programs, pose yet more risks (for a review The need for husbandry technology in conservation extends see Snyder et al. 1996). As populations decline, the possible well beyond its role in breeding of threatened species for damage done by each of the above processes escalates. release. In addition to increasing numbers of wild frogs by Unfortunately, it is also when populations decline that the release, captive supply can also minimise harvesting for other above interventions are given higher priority. Careful uses. Good husbandry can reduce abnormal behaviours attention must be given to all potential risks when any associated with confinement.This can lead to less variables in conservation measure involves manipulation of field situations. other research, a better understanding of, and ultimately an To follow the medical profession’s caveat,‘above all do no increased capacity to conserve, frogs. Captive frogs are usually harm’, would be a wise philosophy (Myers 1993). the most accessible to the public and the media, providing opportunities to educate as well as promote conservation COLLECTION AND GENETIC programs and generate financial assistance.The ease of REPRESENTATION monitoring captive frogs has also often led to observations Captive conservation requires founder stock. Procurement of not readily accessed in the field. stock should not exceed a population’s capacity to be harvested Husbandry can be broken down into three functional (Stevens and Goodson 1993).To reduce genetic impact, it is components: Holding (temporary care of animals), Rearing or desirable to proceed with collection of individuals as soon as it is single generational maintenance (for conservation purposes this recognised that collection is required and well before the is usually across a mortality barrier, with the intention of achieving population is at risk. Animals collected should be considered as greater than field survival e.g. recruitment enhancement projects) no longer available to the wild population; too many instances of and Breeding (production of animals from captive-held stock). failed reintroductions exist ( Wolf et al. 1996) to place reliance Captive breeding, rearing and holding can contribute to on individuals being used to develop husbandry protocols ever conservation and conservation research by the following means: contributing to wild populations.Wherever possible, collection should concentrate on animals of low ecological value to the 1. Reducing pressures on wild populations by provision of population and higher value in captivity. In most cases this would captive-bred frogs for end uses: constitute the earliest life history stages available, as larval and I. Pets. juvenile mortality are usually high in nature and significantly II. Medical, biological and conservation research. reduced in captivity. Collection of early life history stages also III. Education and display. necessitates forward planning, as adult captive animals will take IV. Food, or other animal products. longer to achieve.Where protocols for long-term holding or 2. Providing for and contributing to conservation education by: establishment of breeding colonies are being developed, a staggered approach is recommended. A small number of adult I. Increasing public exposure to frogs, thereby drawing frogs can be used to develop techniques for achieving attention to conservation issues relating to frogs. reproductive success, while larger numbers of juveniles are raised 3. Contributing to our knowledge of frogs by: to form a colony. By the time a colony is achieved, techniques for I. Performing specific investigations into the raising as well as breeding should have been developed and any husbandry of frogs. difficulties overcome.

169 Adequate representation of a population or species should effects of captivity could create lasting implications for the be considered in cases where a colony is formed to secure a populations into which captive raised animals are released. All species against extinction and in cases where captive stock release programs should be backed by sound monitoring and, will be used to reestablish or bolster wild populations.What wherever possible, by comparative analysis of fitness constitutes adequate representation must be determined for (measured by success in development and reproductive each situation and in extreme cases such representation may output) of released versus wild individuals. not even be feasible.The loss of genetic representation in captive populations (Briscoe et al. 1992), along with REMOVAL OF CAUSAL AGENTS extinctions or gene loss attributed to low founder numbers or inbreeding (Frankham 1995; Robinchaux et al. 1997), The IUCN guidelines recommend that release should not are well documented. proceed until the causal agents of decline are mitigated (IUCN 1998).While such guidelines may be sound for most other vertebrate groups, they assume that causal agents can RELEASE be identified.Two considerations may give cause to A number of issues and some suggested solutions have been reconsider these guidelines with respect to frogs: most frogs raised with respect to the release of animals which have are considerably cheaper, both financially and ecologically, to spent time in captivity. Genetic considerations rate highly produce than are the majority of other terrestrial (Backus et al. 1995; Ryman et al. 1995), while the potential to vertebrates, and; limited, planned experimental reintroduction release disease must be carefully assessed before release aimed at hypothesis testing may in many cases be a cost (Viggers et al. 1993).The relative fitness of released animals effective and viable technique for identifying causal agents. should also be assessed to ensure efforts are not being wasted (Crayford and Percival 1992; Hunter et al. 1999).This Once the necessary precautions have been taken and plans is best done by monitoring animals post-release. developed that maximise the chance of success, while providing opportunities to identifying any failings, programs The above precautions have not always been taken and aimed at using husbandry to produce frogs for wild release sometimes the consequence of such omissions have led to can be considered. At this point it is valuable to look at the very expensive failures and an inability to identify the reasons, history of frog husbandry in Australia and what our current or establish techniques to improve later attempts. Recently a knowledge has to offer. number of releases of captive-bred Green and Golden Bell Frogs, Litoria aurea failed completely. Although numerous AN HISTORICAL PERSPECTIVE explanations including: human disturbance, under-developed habitat, introduction of predatory fish, and predation by As with most areas of natural history discovery, the early waterbirds and feral carnivores are offered (Hobcroft 1998); period of frog biology was dominated by taxonomists. During the absence of adequate post-release monitoring has left this time taxonomists had cause to temporarily house live frogs these explanations without adequate support or in research collections, providing some of our earliest insights quantification (Meikle pers. comm.). Given the financial and into frog husbandry.The need to describe life histories saw ecological costs of captive projects, the greatest mistake we considerably more attention focused on techniques for can make is not failing to be successful, but failing to design a achieving frog spawn and raising tadpoles. Soon other uses of project that gives us the greatest possibility of identifying the captive frogs; for research, as pets and as display animals began reasons for our failure. to add to our knowledge. Each had their limitations: taxonomists rarely raised frogs past metamorphosis and the The recent discovery of the amphibian chytridiomycete supply of frogs for research, pets and display was typically fungus that may be the proximal cause of numerous frog replenished from wild sources. Although exhibit-based declines (Berger et al. 1998), highlights the need to consider collections sometimes kept individual specimens for very long the risk of disease being introduced with any introduction of times, there was seldom the need to breed replacement stock. frogs. Not only known disease but all diseases should be Amateur collectors often saw breeding as a yardstick by which controlled. Screening, where available, may be able to to measure their achievements and were among the first to eliminate some possible pathogens from release. Adequate breed Australian frog species (pers. obs.). Unfortunately, many quarantine, however, not only provides preventative control such firsts were achieved outside Australia, where the difficulty but protects against unknown or undetectable pathogens. in obtaining more specimens must have contributed to the efforts expended on achieving breeding success. More popular Fitness has always proven difficult to quantify. Failing to literature on husbandry of common Australian frogs has been include such assessments in any release, however, could lead published in North America than in Australia (e.g. de Vosjoli et to erroneous predictions about, or measures of, the success al. 1996). In Australia it was rare for there to be any need to of release programs. It has been demonstrated that crowding culture frogs for long periods of time, to raise young frogs to can cause developmental delays, smaller adult size and lower sexual maturity, or to house the larger numbers of individuals fecundity and fertility. Rarely are captive animals raised at field necessary to give robust results. Frog husbandry was, and in densities and so the question begs, are we really producing many cases still is, an art rather than a science -a means to an stock suitable for release? Smooth Froglets, Geocrinia laevis, end, be it for taxonomy, research, pet keeping or display. raised in crowded captive conditions took much longer to mature, reached sexual maturity at a significantly smaller size Declines in numerous species of Australian frogs coupled with and produced smaller clutches (Marantelli unpubl. data) than the apparent success of captive conservation programs for the greatest extremes measured for this species in nature other vertebrate animals (for a review see Wolf et al. 1996) (Scroggie, pers. comm.).These and other as yet unknown has seen a focus placed on captive programs for a number of

170 Australia’s frogs. Unfortunately the development of husbandry The commercial need to constantly improve techniques, for amphibians has not been subject to the same level of coupled with the high volumes of individuals produced, has interest or need as it has for other vertebrate groups. Large led to significant advances not otherwise able to be achieved commercial interests in mammals, birds, fish and to a lesser by holding small numbers of specimens. extent reptiles have driven the development of husbandry technologies; frogs have been largely overlooked. Dietary CAPTIVE ACHIEVEMENTS FOR information, medications, reproductive technologies and other AUSTRALIAN FROGS husbandry techniques developed for commercially important vertebrates and humans, are routinely applied with success to During 1996 and 1997 a survey of 14 institutions and captive care of non-commercial vertebrate species. 12 private collectors known to have kept significant numbers Significantly less commercial value has been placed on frogs. of frogs was conducted by The Amphibian Research Centre. The resulting comparative reduction in our warehouse of The survey asked respondents to indicate success or lack knowledge no doubt contributed to the poor success rate of thereof for a number of possible achievements in husbandry, some early attempts at captive care of declining amphibians. for each species of Australian frog which they had held in While Australian zoos had on the whole been successful with captivity. Items included: the introduction of various life captive conservation of other vertebrate groups, they history stages into captivity, maintenance of various life experienced considerably more difficulty with frogs.With few history stages for defined periods of time or across defined exceptions frogs had only quite recently been kept in developmental events, achievement of captive spawning Australia’s zoos and were little more than a minor inclusion in under various sets of circumstances and the number of predominantly reptile driven herpetofauna displays.This lack of generations achieved in captivity. Most respondents were experience coupled with the absence of information from visited and interviewed. In some cases survey sheets were commercial research placed us in a very precarious position distributed. In other cases data derived from communications, when attempts were first made to employ captivity in efforts publications or personal observations were used. to salvage declining frog populations. Over 100 of the now presumed extinct Taudactylus acutirostris died in captive Although not exhaustive, this survey gives a good impression collections during attempts to secure that species in captivity of the relative success Australia’s herpetologists and (pers. obs.); while over 50 tadpoles and metamorphs of the zookeepers have had in keeping Australian frogs. At least critically endangered Litoria spenceri were taken into captivity 150 species of Australia’s 205+ species of frogs have been at Melbourne Zoo, only one remained alive when the project held in captivity although at least 25 of these were never held to develop husbandry protocols for that species was moved for more than two months.The number of species and elsewhere (Gillespie pers. comm.). Despite the fact that these genera for which significant benchmarks have been achieved projects did include some successes and did increase our is included in Table 1.The number of species which are listed knowledge, and that other projects have been successful, they in the Australian Frog Action Plan that have reached each are stark evidence that we were not prepared to deal with benchmark in captivity are also included (Table 1.) last ditch efforts when it came to amphibians. Australia’s track Much of the information collected by our survey was record numbers at least four species Rheobatrachus silus, anecdotal. Few respondents were able to quote figures for R. vitellinus,Taudactylus diurnus and T. acutirostris that, while not all exact periods of time that specimens were kept, number of held specifically for conservation, possibly the last individuals of specimens kept or raised, or even in some cases the number their species died in the hands of their human keepers. of times breeding had been recorded. It should also be noted It has only been recently that any attention has been focused that much data was contributed by biologists with no specific on more disciplined attempts to establish protocols for interest in long-term husbandry, or who only kept species Australian frog husbandry. Links between exhibit based long enough to achieve spawn for life history descriptions. collections and universities have recently become more That species bred were seldom raised beyond F1 is more common, promoting increased scientific rigour in this area of due to lack of need than to failure.Those studying life husbandry research.The establishment of the Amphibian histories simply did not need to attempt to proceed any Research Centre (ARC) in 1994 (the first facility in Australia further. Despite these limitations, the results indicate that very driven purely by frog husbandry) has enabled the commercial few species of Australian frogs have been studied extensively uses of frogs to fund the development of husbandry in captivity.The typical spawning event (excluding those protocols for a number of common native frog species. species spawned prior to one month in captivity) was

TABLE 1: Captive achievements for Australian frogs.

Achievement Number of species Number of genera Number of listed species. (Frog Action Plan)

Adults kept > 2 months 125 20 13 Adults kept > 12 months 90 16 5 Spawned in captivity 41* 15 3 F1 progeny raised to at least metamorphosis 32 10 4 F2 or beyond achieved 14 5 1

* 11 of these 41 species have only been spawned after less than one month in captivity.

171 achieved after keeping a colony of frogs for several years. PUBLIC This must raise the question of what we are doing wrong The public can be both consumers of frogs and providers of which can cause a group of adult frogs to largely fail in their financial and political support for recovery actions.There is primary biological function — reproduction! demand for pet frogs and tadpoles, so much so that the child with a tadpole in a bucket is almost an Australian icon.The THE PLAYERS Australian public grew up with frogs, their sounds and the In all fields of endeavour there are usually a number of experience of raising a tadpole.Their desire to ensure this individuals and organisations who contribute to any successful right for their children can be counted amongst the most operation.The nature and popularity of conservation ensures powerful of all the recovery tools at our disposal. Husbandry that a multitude of players are involved in most actions and and captive breeding can ensure the public continued access the level of success of such actions usually depends on the to and contact with frogs and tadpoles.The Australian people precise use and coordination of these contributors (Cannon are not only potential direct financial supporters of recovery 1996; Kleiman and Mallinson 1998). Husbandry and its highly actions, but the voice of frogs to others, including corporate popularised subordinate — captive breeding, attracts perhaps sponsors and government. the widest diversity of participants and therefore involves considerable challenges in terms of effective coordination of COMMUNITY ORGANISATIONS contributions.The primary focus is considered here to be Community organisations galvanise the sentiments of the husbandry, irrespective of its objective being commercial, public and allow those involved in recovery processes direct research or conservation.The players and their connections access to a sympathetic audience. Frog groups around the are illustrated and discussed below and in Figure 1. country not only have the networks necessary to raise funding and political pressure, but act to ensure that those ENVIRONMENT AUSTRALIA who take an interest remain interested.The provision of Environment Australia (EA) administers Australian frog information, in the form of talks and newsletter articles from recovery plans, the National Threatened Frog Working Group conservation scientists, is an integral part of ensuring that and provides funding for conservation actions. these groups remain effective in their popularising of frog conservation issues. Such groups may also provide assistance RECOVERY TEAMS with some actions and are in a unique position to provide remuneration to sponsors in the form of direct contact with Recovery teams administer conservation decisions and an interested consumer base. It should also be noted that in actions for specific species.They are valuable sources for some cases members of such groups have contributed and communication of information pertinent to each species’ continue to contribute significant amounts of information to conservation. our knowledge of husbandry.

UNIVERSITIES/RESEARCH FACILITIES MEDIA If frog husbandry in Australia is to secure credibility as a science, Frogs held in captivity provide the media with easy access to it will be necessary for links to be firmly established between interesting stories.The media responds to the public, who in scientists and those who are investigating husbandry. In some turn rejuvenate their interest with each new story.The media cases these links already exist and have seen rigorous is in the unique position of being the main method available experimental investigations into husbandry (e.g. Hunter et al. to conservation biologists for rewarding the public and 1999). Universities and research collections are also end users sponsors for their efforts, while generating more interest of frogs for non-conservation research and could be provided from both. with stock, use of facilities and husbandry expertise from those investigating husbandry.As a newly developing area of research, EXHIBIT BASED COLLECTIONS husbandry offers a multitude of possibilities for student projects and since facilities to conduct such research are expensive, their House, display and in some instances breed amphibians.They efficient use depends on maximising their output. are in some cases active participants in husbandry research. Husbandry research ideally should be able to provide exhibit- CORPORATE SPONSORS based collections with expertise and “ecologically dead” stock for display.The placement of small numbers of specimens in Captive programs place threatened species in population collections specialising in public education offers us a further centres and provide access to the public and the media.This avenue for recognising contributors and promoting our access, coupled with the public’s perception of the importance conservation objectives to the public. of such programs and their desire to see action being taken, places such programs in a unique position with respect to DEVELOPMENT OF HUSBANDRY generating funds not otherwise available to less visible or less PROTOCOLS popularised forms of conservation. Sponsorship of such projects can provide resources beyond the scope of normal recovery The need to ensure a scientific approach to the development plan budgets, while providing valuable marketing opportunities of husbandry techniques will require the production of a set for the sponsor.The effective sale of recovery actions to of guidelines. Only a rigid experimental approach is likely to corporate sponsors almost invariably requires visibility of the yield the results that are needed in the available time. Lack of species concerned and, as a consequence, husbandry. communication and co-ordination will continue to lead to

172 FIGURE 1: Relationships between stakeholders in the use of frog husbandry for conservation.

Recovery Environment Universities/res plans/teams Australia earch facilities

Provided with access Provided with high Provided with use of to husbandry as a profile and effective stock, facilities and recovery and recovery actions. expertise. experimental tool. Provide assistance Provide projects Provide funds and in maintaining links with student labor direction for work. and funding and collaborative through grant seeking. recovery plans.

Exhibit based Community collections organisations Provided with Provided with expertise and regular information Husbandry ecologically dead on projects of stock for display. interest to members. Experimental, recovery Provide public Provide increased and/or commercial access to frogs, awareness and frog husbandry information about acknowledgment of frog conservation sponsors through projects and publications & acknowledgment newsletters etc. to sponsors.

Public Corporate Media Provided with access sponsors to pet/schoolroom Provided with easy stock as well as Provided with high access to good education of issues. profile wide ranging stories. Provide increased recognition of Provide public profile and contributions. access and incentive for Provide financial recognition of sponsors as well as support for sponsors. direct financial projects. input into projects.

173 repetition and wasted resources. In husbandry, two processes S.P.F. quarantine procedures are well documented in should be seen as vital to the development of protocols for husbandry literature, and should be practiced. Functional species reintroduction: provision of appropriate ecological quarantine for frog facilities has been achieved by wearing parameters and quarantine. Administratively, the processes of and disposing of latex gloves, avoiding contact of other communication and cooperation are critical. clothing or implements with enclosure contents, use of separate implements and water supply, splash guards and PROVISION OF ECOLOGICAL disinfection of enclosures and implements before reuse with PARAMETERS 1% sodium hypochlorite (followed by rinsing). Failure to provide an environment comparable to that Snyder et al. (1996) propose single species facilities separate naturally experienced by each species can lead to a multitude from multispecies collections. Individuals used in captive of problems.While many species can be maintained, bred and conservation measures which may ultimately include release, raised in quite sterile environments, the absence of natural should be sourced directly from the wild. Materials sourced conditions can lead to selection for captivity (Snyder et al. from their natal environment should be used to provide for 1996) and production of animals which are ecologically (Lyles their ecological requirements.These precautions should ensure and May 1987; Page et al. 1989) or immunologically (Viggers that exposure to pathogens exotic to the species range will et al. 1993) naive. Efforts to develop protocols should begin not occur, at the same time as facilitating the possibly with attempts to provide for all environmental parameters necessary exposure to natural pathogens. As an additional before minimisation of care requirements is attempted. precaution, samples should be taken for full pathological investigation prior to release of any stock (Viggers et al. 1993). Some considerations that should be included in the Pathology should not be seen as a suitable precaution in its establishment of captive facilities for each species are: stocking own right and extreme caution should be exercised if releases densities; spatial and temporal contact between individuals; of animals, which have not been adequately quarantined, are environmental variants including microenvironmental variants being considered. No release of animals held without and their daily, annual or other cycles (e.g. temperature, adequate quarantine should occur into extant populations or photoperiod, humidity, precipitation, water levels and flow sites that may facilitate exposure of extant populations. rates, barometric pressure and food availability); and availability of specific microenvironments used by each species COMMUNICATION AND COOPERATION (e.g. sites used for basking, calling, oviposition, sheltering, hunting, feeding and thermoregulation). It is recommended that a network be established representing the stakeholder groups (Figure 1) to co-ordinate Common omissions in captive efforts with frogs include failure husbandry research and its application to conservation of to provide non-breeding environments (most captive frogs are Australia’s frogs.The establishment of such a network remains housed permanently in environments which replicate the the most likely method of ensuring the required measure of habitat in which they reproduce), housing females with males success to secure husbandry as a contributor to (exposure of females to reproductively-primed males does conservation, and to prevent failures due to lack of co- not usually occur in nature until the female is ripe), holding ordination or design resulting in the demise of this potentially too few males to elicit chorusing, and housing animals at useful conservation tool. densities which modify normal reproductive behaviour. PRIORITIES FOR HUSBANDRY QUARANTINE RESEARCH IN AUSTRALIAN FROGS Although disease has been flagged as a general risk in wildlife Our survey indicated that husbandry knowledge for a conservation (May 1986, Simonetti 1995, Cunningham 1996), number of ecological and phylogenetic groups of Australian the recent publication of evidence supporting the possibility frogs is still lacking. Despite some 70% of Australia’s frogs of disease as a causal agent in Australian frog declines (Berger being kept at one time in captivity less than 20% have been et al. 1998 and Berger et al. 1999) places special emphasis on bred.Those successfully bred represent primarily lowland disease control in captive conservation efforts for frogs. Only species with lentic tadpoles and few if any unusual life history one method exists to adequately control the effect of both characteristics. Species for which significantly less husbandry known and unknown disease — quarantine. Disease and success have been recorded belong to groups which include parasites play an important role in natural systems.The stream-dwelling species, high altitude species, arid-adapted inclusion or exclusion of certain disease organisms in captive species and species exhibiting specialised life history programs can only be achieved by combining quarantine, characteristics (e.g. direct development or parental care). pathology and exposure to natural materials. Most of Australia’s threatened frog species belong to one or more of these ecological groups. Limited experience or Effective quarantine involves isolation of captive animals from success has been reported for declining taxa or close allies. exposure to materials, other individuals or species which have either not come from the same location, or have not As the costs of using captive technologies can be high and themselves been subjected to the same level of isolation their value has not yet been adequately assessed, it is (Snyder et al. 1996). Practically this involves the minimisation recommended that efforts to develop protocols for of contact and includes use of sterilised water, separate threatened taxa be viewed as a priority, but efforts to secure implements and separate housing at enclosure, room or even sustainable populations of threatened species in captive facility level. Animals should not be exposed to any biological collections should not be seen as a routine step in species material not sourced from their point of origin. recovery. Efforts to develop husbandry protocols should focus

174 on ecological and phylogenetic analogues of threatened species means not readily accessible to other conservation actions. and be tested and modified to suit threatened species only Such non-transferable funding has been the backbone of this once demonstrated to be successful on such analogue species. project and allowed for other financial resources to be directed elsewhere within the P. corroboree recovery plan. Captive breeding for reintroduction should not necessarily be Similar programs are in progress and should be achievable by seen as the ultimate goal of husbandry research. As most projects involving threatened frogs in captivity knowledge improves it should be applied to the conservation measures already discussed, very few of which actually include CONCLUSIONS captive production of frogs for release into natural habitat. Techniques such as recruitment enhancement should be seen With rising interests in husbandry and a number of conservation as useful tools which focus on husbandry skills already actions involving husbandry already underway, there is clearly a developed for Australian amphibians and which also mitigate need to utilise husbandry as a tool in conservation processes. many of the risks associated with longer term captive The need to move quickly to cater to the numerous actions programs.The role of husbandry in education and generation listed in recovery plans which require husbandry is real, as is the of both public interest and funds remains a significant reason need to recognise the limitations of husbandry as a developing for holding threatened species in captivity. science.While legitimate questions must still be raised about genetics, fitness and disease risks associated with the release of A CASE IN POINT, Pseudophryne corroboree animals exposed to captivity, the need for solutions in what AT THE AMPHIBIAN RESEARCH CENTRE seem to be otherwise unwinnable situations must be respected. Where possible it would be prudent to direct the focus to A recruitment enhancement program for the Southern those aspects of husbandry that have demonstrated repeated Corroboree Frog Pseudophryne corroboree has been success, to ensure we do not overextend ourselves — the operating since early 1997 (Hunter et al. 1999).The captive consequences of which are dire. Programs aimed at tadpole component of this project has been carried out at the ARC. rearing and manipulating mortality schedules across single The project has been housed in a dedicated facility on a site generations could be considered as other techniques are being separated from other ARC frog colonies. Husbandry perfected. Efforts to develop techniques for longer term protocols, provision of ecological requirements, isolation and management should be commenced immediately with analogue quarantine procedures, as well as routine pathological species, focusing on those phylogenetic and ecological groupings screening of specimens prior to each release, have been for which we have the least knowledge. followed as prescribed in this paper. Post-introduction monitoring to assess relative fitness of captive-reared animals Maximum use must be made of all the potential contributors is described and preliminary results reported in Hunter et al. to husbandry research. Skilful coordination of multidisciplinary (1999). Methodology for captive care was based on three groups within recovery teams remains our best hope of years of successful commercial breeding of P. semimarmorata. effecting viable use of husbandry as a tool.We must recognise the public perception of captive programs and This project has been supported by a wide range of make use of the support they can give, while taking care not participants and is presented here as an appropriate model for to promote the idea of captivity as a panacea.The effective the involvement of relevant stakeholders in captive use of media and the public must be explored to provide conservation.The project was supported by funds provided to alternative sources to traditional funding. the New South Wales National Parks and Wildlife Service (NPWS) by Environment Australia (EA). Over two years, Above all there is a need for pragmatism.While programs $21 000 has been contributed to the captive component of the relying on husbandry may not always be the cheapest or best P. corroboree project by the recovery team. Cooperation conservation actions, we cannot deny the support afforded between the ARC and the University of Canberra has ensured such visible conservation measures and the level of public the development of rigid experimental design following scientific awareness and sense of involvement they can bring. Such principles. Further cooperation between the ARC and Victoria programs should be able to generate much of their own University of Technology has provided student labour to funding, reducing competition for other conservation dollars. conduct two small research projects within the larger Sustainable commercial use of frogs must be encouraged as a husbandry program.The involvement of the Victorian Frog free mechanism for driving research.The situation that exists; Group (VFG), by regular inclusion of information about the end users with capital while husbandry research is severely project in their newsletter, led to the involvement of Australian under-resourced, begs the consummation of a very Geographic (AG) as a corporate sponsor leading to over $5 000 convenient marriage. Realistic attitudes to commercialisation being contributed to the project. Media coverage and support of wildlife conservation may provide the only solution to a generated by this publicity has led to a further $2 000+ in public rapidly drying government funding well. donations to the project.The volunteer resource drawn to the ARC by the presence of this and other projects has contributed ACKNOWLEDGMENTS greatly to the ARC’s development and continued success. As a consequence of this and its commercial operations, the ARC has Thanks to all those who have by their collective efforts contributed some $15 000+ in material resources and $30 000+ contributed to the knowledge base for Australian frog in labour and infrastructure resources to the project. husbandry as described in this paper, especially: Margaret Davies, staff at Taronga Zoo and Currumbin Sanctuary, Lothar In just under two years a project valued at over $70 000 has Voigt, Harvey Vaux, Karen Thumm, Jacqie Recsie,Will been carried out with less than a 30% contribution from Osborne, Mark Cowen, Ross Alford and Shane Gow. government. Over $50 000 has been contributed largely by

175 Staff at the ARC who assisted in conducting the frog husbandry (Pseudophryne corroboree). Pp 158-167 in Declines and survey, or who assisted with maintenance of the collections Disappearances of Australian Frogs ed by A.Campbell. while I was absent visiting interstate collections or preparing this Environment Australia: Canberra. paper: Natalie White, Mason Hill, Joanne Doherty, Fiona Hobcroft, D., (1998) Conservation program for the Green Morrow, Michelle Love, Anne Gaskett and Lindley Makay. and Golden Bell Frog Litoria aurea at Taronga Zoo: a cooperative recovery plan. Proceedings of 1997 To all those who donated their time to assist with the ARAZPA/ASZK Conference. Auckland Zoo, Auckland NZ. P. corroboree project: Fiona Morrow, Kristy Penrose, Lee Berger, Mason Hill, Joanne Doherty, David Reznick, Raelene IUCN (1998) Guidelines for re-introductions. IUCN/SSC Hobbs, Rebekah Hirst, Australian Geographic and the Re-introduction Specialist Group. IUCN, Gland, Switzerland members of the VFG. and Cambridge, UK. Kleiman, D. G. and Mallinson, J. C., (1998) Recovery and And to my wife Sally, whose tolerance of my work habits has management committees for Lion Tamarins: partnerships in been pivotal in all of the above. conservation planning and implementation. Conservation Biology, 12: 27-38. REFERENCES Lyles, A. M. and May, R. M., (1987) Problems in leaving the ark. NATURE, 326: 245-246. Backus,V. L., Bryant, E. H., Hughes, C. R. and Meffert, L. M., (1995) Effect of migration or inbreeding followed by May. R., (1986) The precautionary tale of the Black-footed selection on low-founder-number populations: implications Ferret. Nature, 320: 13-14. for captive breeding programs. Conservation Biology, Myers, N., (1993) Biodiversity and the precautionary principle. 9: 1216-1224. Ambio, 22: 74-79. Banks, C.B., (1996) A Conservation Program for the Page,G., Quinn, P., and Warriner, J., (1989) Comparison of the Threatened Romer’s Tree Frog (Philautus romeri). Breeding of Hand- and Wild-Reared Snowy Plovers. Advances in Herpetoculture 1: 5. Conservation Biology, 3: 198-201 Berger, L., Speare, R., Daszak, P., Green, D. E., Cunningham, A. Robinchaux, R. H., Frier, E. A. and Mount, D.W., (1997) A., Goggin, C.L., Slocombe, R., Ragan, M. A., Hyatt, A. D., Molecular genetic consequences of a population McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. and bottleneck associated with reintroduction of the Mauna Parkes, H., (1998) Chytridiomycosis causes amphibian Kea Silversword (Argyroxiphium sandwicense ssp. mortality associated with population declines in the sandwicense [Asteraceae]). Conservation Biology, rainforests of Australia and Central America. Proc. Natl. 11: 1140-1146. Acad. Sci., 95: 9031-9036. Ryman, N. and Laikre, L., (1991) Effects of supportive Berger, L., Speare, R. and Hyatt, A.D. (1999) Chytrid fungi and breeding on the genetically effective population size. amphibian declines: overview, implications and future Conservation Biology, 5: 325-329. directions. Pp 23-33 in Declines and Disappearances of Ryman, N., Jorde, P.E. and Laikre, L., (1995) Supportive Australian Frogs ed by A.Campbell. Environment Australia: breeding and variance effective population size. Canberra. Conservation Biology, 9: 1619-1628. Briscoe, D. A., Malpica, J. M., Robertson, A., Smith, G. J., Simonetti, J. A., (1995) Wildlife conservation outside parks is a Frankham, R., Banks, R. G. and Barker, J. S. F., (1992) Rapid disease-mediated task. Conservation Biology, 9: 454-456. loss of genetic variation in large captive populations of Snyder, N. F. R., Derrickson, S. R., Beissinger, S. R.,Wiley, J.W., Drosophila flies: implications for the genetic management Smith,T. B.,Toone,W. D. and Miller, B., (1996) Limitations of of captive populations. Conservation Biology, 6: 416-425. captive breeding in endangered species recovery. Cannon, J.,R. (1996) Whooping Crane recovery: a case study Conservation Biology, 10: 338-348. in public and private cooperation in the conservation of Snyder, N. F. R., Derrickson, S. R., Beissinger, S. R.,Wiley, J.W., endangered species. Conservation Biology 10: 813-821. Smith,T. B.,Toone,W. D. and Miller, B., (1997) Limitations of Crayford, J. and Percival, S., (1992) Born captive, die free. captive breeding: reply to Gippoliti and Carpaneto. New Scientist, 8: 21-25. Conservation Biology, 11: 808-810. Cunningham, A. A., (1996) Disease risks of wildlife Stevens, D. R. and Goodsen, N. J., (1993) Assessing effects of translocations. Conservation Biology, 10: 349-353. removals for transplanting on a high-elevation Bighorn Denton, J., Hitchings, S., Beebee,T. and Gent, A. (1997) Sheep population. Conservation Biology, 7 :908-915. A recovery program for the Natterjack Toad (Bufo Tonge, S. J. and Bloxam, Q. (1989) Breeding the Mallorcan calamita) in Britain. Consevation Biology, 11: 1329-1338. Midwife Toad Alytes muletensis in captivity. Int. Zoo Yb de Vosjoli, P., Mailloux, R., and Ready, D., (1996) Care and 28:45-53. Breeding of Popular Tree Frogs. Advanced Vivarium Tyler, M., J. (1997) The Action Plan for Australian Frogs. Wildlife Systems, inc., Santee, California. Australia, Canberra. Frankham, R., (1995) Inbreeding and extinction: a threshold Viggers, K. L., Lindenmayer, D. B. and Spratt, D. M., (1993) effect. Conservation Biology, 9: 792-799. The importance of disease in reintroduction programs. Gippoliti, S. and Carpaneto, G. M., (1997) Captive breeding, Wildl. Res., 20: 687-698. zoos and good sense. Conservation Biology, 11: 806-807. Wolf, C. M., Griffith, B., Reed, C. and Temple, S. A., (1996) Hunter, D., Osborne,W., Marantelli, G. and Green, K., (1999) Avian and mammalian translocations: update and reanalysis Implementation of a population augmentation project for of 1987 survey data. Conservation Biology, 10: 1142-1154. remnant populations of the Southern Corroboree Frog

176 Conservation status of frogs in Western Australia

Dale Roberts1, Simon Conroy1 and Kim Williams2

ABSTRACT STATUS OF FROGS IN WESTERN AUSTRALIA Seventy eight frog species are known from Seventy eight frog species are known from Western Australia Western Australia. Only three have been the from fifteen genera (Tyler et al. 1994; Roberts et al. 1997). subject of recent systematic survey to assess status. The Kimberley region (regions sensu Tyler et al. 1994) with For eleven others there are some objective data summer rainfall has 28 species, the arid zone with irregular but tending to summer rainfall maxima has 17 species and (post 1988) suggesting they are still widespread and the southwest with predominantly winter rainfall has 30 abundant.Three frog species are of conservation species. Several species overlap two zones. concern in south-western Australia: Geocrinia alba, Although there has been sporadic work on some species in G. vitellina and Spicospina flammocaerulea.All three the Kimberley and the arid zone (e.g. Roberts 1996;Watson have small extents of occurrence (< 200 km2), small and Gerhardt 1997) neither area has been the subject of any systematic survey to determine the status of anurans. Further, areas of occupancy and fragmented distributions. before the period of major frog declines starting in Australia Twenty three sub- populations of G. alba have gone in the mid seventies (Mahony 1996), many species in these extinct since 1983 and there are continuing threats areas were either undescribed, uncollected or their distributions were poorly known so shifts in distribution, local from grazing, vegetation clearing and fire. Geocrinia extinctions or declines are unlikely to have been detected. vitellina has a range of 6 km2 but there are no For example,Tyler et al. (1994) reported eight species of major threats except fire. Spicospina flammocaerulea Uperoleia from the Kimberley and arid zone of Western Australia — six were described after 1980 and there has is known from 13 populations over a range of been little published work since then. 194 km2, but has a highly fragmented range and In the south-west, winter rainfall zone, there has been more population sizes are uncertain. Its true status intensive work which collectively suggests most widespread is unclear. species still occur over all of their known range and are still

1 Department of Zoology, University of Western Australia, Nedlands,WA 6907 2 Department of Conservation and Land Management, North Boyanup Road, Bunbury,WA 6230

177 locally abundant.This view is supported by the number of known population sites) had been cleared since European frogs recorded and the number of sites where species were settlement with likely consequent loss of populations. heard during recent studies on geographic variation in male Geocrinia alba had a small natural range which has been advertisement call. One of us (JDR) worked on eight species radically reduced and is now severely fragmented (Figure 1). from 1988-1992 and Littlejohn and Wright (1997) reported on a ninth.These data are summarised in Table 2.These data Fifty six discrete populations of G. alba have been located in are further supported by population or population genetics the period 1983 to 1997.These populations can be further studies on Heleioporus psammophilus (Berry 1996), subdivided into 80 sub-populations based on variation in H. albopunctatus, Neobatrachus kunapalari, N. pelobatoides and adjacent land use (e.g. cleared and grazed, natural forest or Pseudophryne occidentalis (Davis 1997) and Crinia georgiana tree plantations), or division by physical features such as (McDonald 1998) which all reported abundant populations major roads crossing creek systems (Wardell-Johnson et al. of the species studied. 1995 define populations and sub-populations). Most populations occur on privately owned land (Wardell-Johnson Although there is some evidence that land clearing, increasing et al. 1995; Figure 1). Based on IUCN (1994) criteria, G. alba salinity and fragmentation of vegetation may have affected is critically endangered (criterion B, area of occupancy less population size and local occurrence of some widespread than 2.5 km2; B1, severely fragmented, and continuing decline species (e.g. Main 1990; Davis 1997), there is no evidence of observed in, B2 (b) area of occupancy and B2 (c) number of general, recent declines or extinctions in anuran populations sub-populations; Figure 1,Tables 2 and 3;Wardell-Johnson in Western Australia comparable with those in other parts of et al. 1995). Australia (Mahony 1996;Tyler 1997). However, it is also 2 important to note that, with three exceptions listed below, Geocrinia vitellina has a much smaller range at about 6 km there has been no systematic survey of the status of any frog with only six known populations but all of these are in State species in Western Australia since 1992. Forest or are proposed as conservation reserves and under no immediate threat from clearing or logging activity Three species, Geocrinia alba and G. vitellina (Wardell-Johnson (Appendix 2 of Wardell-Johnson et al. 1995; Figure 1). and Roberts 1989; Roberts et al. 1990) and Spicospina Geocrinia vitellina is vulnerable based on IUCN (1994) flammocaerulea (Roberts et al. 1997) are currently declared criterion D2. as “threatened” pursuant to section 14 (2) (ba) of the West Australian Wildlife Conservation Act 1950. Species listed as Spicospina flammocaerulea was discovered in 1994 and “threatened” generally fall into the IUCN (1994) categories described in 1997 (Roberts et al. 1997). Field work in 1997 of extinct, critically endangered, endangered or vulnerable. raised the number of known populations to 13.This species 2 In the most recent rankings of species according to IUCN has a limited extent of occurrence (around 194 km ), small 2 (1994) Red List criteria (released in May 1998), the West area of occupancy (0.78 km calculated by assuming an area 2 Australian Threatened Species Scientific Committee (TSSC) of occupancy of 0.06 km (probable area at Mountain Road listed G. vitellina as vulnerable (criterion D2), G. alba as one of the largest sites) at each site and multiplying by 13)) endangered (criterion C2a) and S. flammocaerulea as and a fragmented range (Figure 2). Based on these data it is vulnerable (criterion D2)1. Concern about the status of these clearly vulnerable (criterion D2). However, this evaluation species came from initial observations of small range, low takes no account of possible variation in population size. In population number or low population density, and Table 4, we report data on number of calling males at several subsequently from detailed studies of range, fragmentation sites from 1994 to 1997. At Mountain Road, numbers of and population size. calling males have dropped from high levels of 120 observed in 1994 to lows of 2 in 1997. If this represents a true decline or variation in population size then the status of this species THREATENED FROG SPECIES — may be worse. Field work in 1998 and 1999 is designed to WESTERN AUSTRALIA assess population size more directly using mark-release- Geocrinia alba and G. vitellina occur north and west of the recapture techniques and survey of tadpole populations. Blackwood River between Margaret River and Augusta Spicospina flammocaerulea is found in peat based swamps (Figure 1).They are members of a complex of four closely near the West Australian south coast, east and north-east of related species (G. alba, G. lutea, G. rosea and G. vitellina) that Walpole in an area of moderate relief with granite outcrops all have direct developing eggs, breed in spring and into early and associated ranges of hills rising to 300 — 400 m (e.g. summer and have population structures consistent with very Mt Frankland 411 m, Mt Roe 340m; Mt Lindesay 440 m; low levels of dispersal among even adjacent local populations Roberts et al. 1997). Five populations are on private property (Driscoll 1997, 1998a, b). north-west of Bow Bridge, with the remainder in the Mount Geocrinia alba has an extent of occurrence (range) of Frankland National Park or on land designated to form part approximately 130 km2 (Figure 1, minimum convex polygon of the Mount Roe-Mount Lindesay National Park but not yet method, IUCN 1994).Wardell-Johnson and Roberts (1993) declared (Figure 2). estimated that 70% of creek systems suitable for breeding (where suitability was defined by geomorphology shared with POPULATION SIZE: Geocrinia Driscoll (1996) reported counts of calling males and from 1 data reported in this paper were not available to the TSSC in late mark-release-recapture studies counted total numbers of 1997 when decisions on status were last considered. Status males calling over a season in breeding populations for assessments discussed below represent new evaluations based on G. alba and G. vitellina.These data indicate that for G. alba, data reported here.

178 FIGURE 1: Distribution of Geocrinia alba and G. vitellina in relation to land tenure.All populations ever known are included.White areas are publicly owned land. light shading, State Forest, darker shading, National Park.Arrow on inset indicates approximate location of area covered by main map.

the maximum number of calling males counted on any one night in a season represents an average of 89% of all males FIGURE 2: Distribution of Spicospina flammocaerulea in relation to land tenure. Shaded areas are privately owned land. found in the population over the whole season. For Remainder is managed as state forest, or national park (see G. vitellina this figure is slightly lower at 83%.This means text).Arrow on inset indicates approximate location of area counts of calling males are a reasonably accurate estimator of covered by main map. the population size for adult calling males. Driscoll (1996) reported counts of egg masses at two sites where he also counted males.The number of egg masses almost equalled the total number of males. If we assume one egg mass per female per season (Driscoll 1996) this would mean there is a 1:1 adult sex ratio making counts of calling males a good index of total adult population size.

Population estimates (October counts of calling males) are available for three populations of G. alba and two populations of G. vitellina for the period 1992–1997. Some earlier data from transect and quadrat counts were reported by Wardell- Johnson and Roberts (1991). For G. vitellina, populations at Spearwood North rose steadily until 1996 but declined radically in 1997 before this area was burnt. At Spearwood South, the population increased rapidly from 1992–1993, was stable, then rose again in 1997 (Figure 3).There was no obvious cause of decline or increase at either site.These sites are approximately 1 km apart precluding any difference due to major variations in local weather conditions.

179 Spearwood Creek was last deliberately burnt in a spring, fuel TABLE 1: Number of frogs recorded (total over all sites) and number of locations where species heard, for eight south-west reduction burn in 1983. As part of the management plan for frog species with ranges largely or wholly in the intermediate this species, Spearwood creek south of Denny Road and rainfall zone (the wheatbelt) of Western Australia. Data collected Geo Creek (the second and third most easterly records on 1988 - 1992 by Roberts. The complete range of all species was covered but not all species were sampled with equal intensity. Figure 1) form a 137 ha fire exclusion zone with the remainder of the range subject to normal fuel reduction burns at 7–8 year intervals. After population counts were Species No. recorded No. locations made at Spearwood North in October 1997 (Figure 3), a heard wild fire (an escape from a fuel reduction burn in an adjacent Neobatrachus block) burnt 50% of the total area of occupancy for this kunapalari 72 102 species, including 85% of the fire exclusion zone.Within the pelobatoides 27 9 fire exclusion zone, 23.6% of breeding habitat burnt at mild albipes 36 98 to moderate intensities (i.e. scorched foliage retained on Pseudophryne Leptospermum and Agonis species; cf.Wardell-Johnson and occidentalis 264 90 Roberts 1993) and a further 25.6% was subject to an intense guentheri 333 192 fire (total defoliation of the same plant species and almost Crinia complete removal of surface litter). Other areas were either pseudinsignifera 535 230 not burnt or burnt at lower intensities. subinsignifera 83 12 Heleioporus albopunctatus 106 77

FIGURE 3: Population estimates for two localities for Geocrinia vitellina. Counts are maximum number of calling males (generally from October)

FIGURE 4.: Population estimates for three localities for Geocrinia alba. Counts are maximum number of calling males (generally from October)

180 TABLE 2: Extinction rates for sub-populations of G. alba on Table 4: Maximum counts, by year, of calling males for all privately versus publicly owned land, 1983 to 1997. 75 sub- known populations of Spicospina. Mountain Road (site 1) and populations known 1983–1996. There was no significant Boronia Road 1 (site 2) are split into northern and southern difference in extinction rates on privately and publicly owned populations because of differences in fire management history. 2 land ( x = 0.05, d. f. = 1, p > 0.05) Asterisk indicates an estimated number; dash means site was not visited (or was unknown in that year). Sites 1–6 and 8 are on the Kent River drainage; site 7, Frankland River drainage; sites Extinct 1997 Extant 1997 8 –13 on Bow River drainage. Fire data supplied from CALM records (summer of 1953–54 to 1995–96), by Hagan, CALM, Public 6 15 Manjimup from CALM,Walpole records. Exact location of Collis Private 17 37 Road 2 is uncertain so no fire history entered. Fire histories were not available for sites on private property.

TABLE 3: Extinction rates for sub-populations of G. alba on Site Site 1994 1996 1997 Last privately owned land related to adjacent vegetation. Cleared No. Fire covers total clearing, partial clearing and sites cleared and developed for tree plantations. Intact includes all sites where 1a. Mountain Road north 60 11 2 1994 there has been no or minor modification of adjacent upland vegetation. No data available for two new sites discovered in 1b. Mountain Road south 60* 10* 0 1994 1997.There was a significantly lower extinction rates at sites 2a. Boronia Road 1 north 2 3 0 1994 with intact, upland vegetation ( x2 = 6.32, d. f. = 1, p <0.05). 2b. Boronia Road 1 south 0 0 2 1993 3. Middle Road 2 0 2 1993 4. Collis Road 1 - 10* 8 1993 Extinct 1997 Extant 1997 5. Nornalup Road - - 3 1994 6. Boronia Road 2 - - 3 1994 Cleared 14 16 7. Collis Road 2 - - 5 no data Intact 3 19 8. Boronia Road 3 - - 2 1987 9. Trent Road 1 - - 150* no data 10. Trent Road 2 - - 3 no data In 1994 the maximum, total number of adult G. vitellina was 11. Rose Road - - 4 no data estimated at 2 230 frogs (Wardell-Johnson et al. 1995). Based 12. Hazelvale Road 1 - - 8* no data on 1994 counts, 64% of this total population was burnt in the 13. Hazelvale Road 2 - - 10* no data 1997 fire.The fire burnt about half of Spearwood Creek (south of Denny Road), all of Geo Creek and most of Hutt Pool. Spearwood North and South (Driscoll 1997, 1998a). On Geo Creek, a tributary of Spearwood Creek, sixty three males were calling when the population was counted twelve Populations of G. alba have varied in size but the dramatic days before the fire. Counts twelve days after the fire reduction at Forest Grove in 1995 was also associated with a indicated 13 of 63 males had stopped calling.There are two low intensity, fuel reduction burn (Figure 4).The population possible causes for this decline.The fire occurred late in the declined to about 60% of pre-fire levels.This is consistent with breeding season when males are normally dropping out of the average effect reported by Driscoll and Roberts (1997) the chorus therefore some loss is expected. However, males for the related species G. lutea.There are some indications of a may also have been affected by the fire itself, either killed, or recovery of population size in 1997 (Figure 4). they stopped calling because conditions were no longer suitable. Driscoll and Roberts (1997) reported one male EXTINCTIONS AND NEW POPULATIONS G. lutea killed by fire. One of us (SC) found three, burnt female G. vitellina active after the 1997 fire. Fire may affect Of 75 sub-populations of G. alba known from 1983 to 1996 adults of both sexes but it is possible females are more (5 new subpopulations were discovered in 1997), 23 have susceptible to direct damage if they are not protected in gone extinct (includes 15 populations not split into sub- breeding burrows. populations). Extinction was defined as absence of calling frogs at any site in 1994 or 1997. For fourteen sites scored as Based on Driscoll and Roberts (1997)(who studied Geocrinia extinct in 1994 there was no evidence of recovery at any site lutea) it is likely that low intensity, spring fires will cause in 1997 suggesting these are permanent extinctions rather significant drops in the number of calling males in Geocrinia than an artefact of poor survey technique. populations in the two years following fire though the longer term pattern of recovery is unknown.The study sites used by In some cases there were obvious causes, e.g. one population Driscoll and Roberts (1997) all had a 30 year (or more) (GA4a — Brooks Road) was inundated by a dam history of fuel reduction burning implying that populations constructed for irrigating vineyards, but in most there was can recover from low intensity fires.The impacts of higher not. Extinctions were equally common on both privately and intensity fires have not been experimentally studied. publicly managed lands (Table 2). On privately owned land, extinctions were more common at sites with adjacent Although large populations at Spearwood South were not cleared land: sites more likely to have adjacent stock grazing burnt, the chances of recolonisation or augmentation of (Table 3) but where there may also be other confounding depleted populations at Spearwood North by natural impacts such as changed hydrology or fertiliser run-off . migration is low. Movements of adults males are extremely Extinctions may also be an inevitable outcome of small limited and genetic structuring of populations is consistent with population size related to stochastic loss or inbreeding little or no migration amongst populations for any life stage or depression (Lande 1988; Frankham and Ralls 1998; Saccheri sex, even at the very local scales of only 1 km separating et al. 1998 — see ‘Prognosis’ below for further discussion).

181 No populations of G. vitellina are known to have gone extinct. Geocrinia vitellina The status of populations where habitat burnt in 1997 A fire exclusion zone was in place for part of the range of cannot be reliably assessed until the spring of 1998 or 1999 G. vitellina covering Spearwood and Geo creeks but this (cf. Driscoll and Roberts 1997 who reported maximum failed in 1997.The fire exclusion zone covered about half the declines two years post fire). geographic range of this species but this area contained around 80% of the adult frog population counted in 1994 Spicospina flammocaerulea (see above).The remainder of the range is subject to fuel No reliable data on population size are available.Table 4 reduction burning, again restricted to spring and on an eight contains counts of calling males made in the period 1994 — year cycle as for G. alba.The range of G. vitellina is subject to 1997. At site 3 (Table 4), a high intensity fire in October 1994 high visitation rates by recreational users of the Blackwood River system which may increase the risk of further fire burnt all ground level vegetation back to bare earth with a incidents. Pig control has been conducted in the range of scorch height on the tree stratum of more than 5 metres. G. vitellina by volunteer shooters whose expenses are partly Many frogs were observed calling and breeding after this fire paid by the Department of Conservation and Land but numbers of calling males at this site have not been Management.There has been no evidence of pig damage at matched since (Table 4). It is possible that peak numbers of any known populations of G. vitellina. calling males are only achieved after fire because the reduced evapotranspiration generates elevated water tables or very Spicospina flammocaerulea open swamps allowing elevated water temperatures. Comparable conditions may only occur in unburnt sites on No specific management actions have been taken for this species. CALM volunteers are currently monitoring known very hot days in wet years leading to explosive breeding in populations and adjacent control sites for evidence of pig most years (cf. Neobatrachus and Myobatrachus species, damage. Over the summer of 1997–1998 there was little Roberts 1983, 1997). However, explosive breeding seems indication of pig damage at ten monitoring sites (five control unlikely. At site 9, approximate numbers were first observed and five known population sites). Sites were monitored by on October 30 and maintained over ten nights in the period walking five randomly placed 100 m transects and noting November 3 to 13, 1997. At site 10, numbers were damage in 1 m sections. apparently maintained over the period November 3 to December 21 but with only 3 nights of observation. Neither This species will also require careful fire management. Studies site was burnt in 1997. by Horwitz, Pemberton and Ryder (in press) on peat swamps in this area showed a very high potential for peat loss due to No data are available to calibrate counts of calling males. fire. Peat ignition could lead to total loss of swamp systems. These numbers represent minimum estimates of population The apparent decline in frog numbers at Mountain Road size but in the extreme, could also represent maxima. following wild fires in 1994 also suggests a possible risk from fire. However, the frequency of major fires varies amongst MANAGEMENT ACTIONS sites on CALM managed lands (Table 5) and all sites, except Collis Road 1, have experienced wild fires in the last 50 years. Geocrinia alba This suggests some capacity to recover post-fire but the time and conditions for full recovery which could set an optimal Fences funded by the Federal Endangered Species Program fire interval and intensity regime are unknown. (through the Geocrinia Recovery Team) have been constructed around 15 populations on nine, privately owned properties in an attempt to reduce damage from grazing by stock. Only one of the fenced populations contained more than 50 frogs in 1994 — most are therefore small populations.We do not know whether fencing will maintain TABLE 5: Fire histories for fire management units that contain populations in the long term but this procedure is designed populations of Spicospina. Site numbers are from Table 4. "Intense" means a single fire covered at least two adjacent blocks. to eliminate any threat from grazing by stock and may allow These are extensive wild fires or fires resulting from escapes from us to disentangle the impacts of grazing from other impacts fuel reduction burns. In our experience these have all been of clearing.The fenced populations are being monitored. intense fires (see text for comments on 1994 fire at site 1). Data sources — see Table 4. The Geocrinia Recovery Team has agreed that fuel reduction burning will be maintained on an eight year cycle, covering two Site All Mean Intense Mean generations, with fires constrained to early spring to minimise No. fires Interval fires Interval intensity.West Australian Department of Conservation and 1a. 7 6.3 4 11 Land Management (CALM) fire prescriptions are to be 1b. 8 5.5 5 8.8 reviewed to formalise this decision. Equivalent control of fire 2a. 7 6.3 5 8.8 on private property is not possible though advice to 2b. 6 7.3 1 44 landholders from the Geocrinia Recovery Team has been not to 3. 6 7.3 2 22 burn swamps at all (Geocrinia Recovery Kit — fire management 4. 5 8.8 0 – advice. Distributed to landholders in the range of G. alba). 5. 6 7.3 1 44 6. 7 6.3 5 8.8 8. 7 6.3 4 11

182 PROGNOSIS CONCLUSIONS

Geocrinia vitellina Two of three frog species from southwestern Australia of conservation concern, S. flammocaerulea and G. vitellina,have There are no identified threats to this species other than that status because of restricted range or inadequate inappropriate fire management and feral pigs. An historical knowledge. However, in both species the majority of analysis of the timing and intensity of fire events may give populations are in areas managed at least in part for nature some better insights into the importance of fire as a conservation with manageable threats. By contrast, Geocrinia management tool for this species. Despite fire in 1997, alba is of major concern as populations of this species are the long term prognosis for this species is good. disappearing at an alarming rate with most on privately owned land.The likely causes of decline are isolation of Geocrinia alba populations due to land clearing and associated small populations sizes exacerbated by low natural dispersal. Nearly 31% of all known sub-populations of this species have Long term survival will require active expansion of available apparently gone extinct in the period 1983–1997: a very high habitats on private land, careful management of fire regimes rate of loss. Driscoll (1998a) argued that in the long term, on public lands and reintroductions to mimic natural populations of G. alba have expanded and contracted and he migration patterns no longer attainable in the modified presented evidence for historical contraction into a number landscape. It is unlikely that declines have been caused by of discrete refugia, though these were not based on any introductions of novel diseases (cf. Berger et al. 1998) as obvious features such as major drainage systems. However, most subpopulations that have gone extinct have rarely been Driscoll (1998a) was unable to estimate a time scale for visited. Sites with high exposure to accidental disease these events.The low levels of movement both within and introduction, e.g. the major monitoring sites (Figures 3 and 4) between populations (Driscoll 1997, 1998a) mitigate against have not shown inexplicable declines that might not also be rapid recolonisation of sites where extinctions have occurred caused by other factors (eg, vegetation senescence at though this might happen over longer time scales (perhaps Spearwood Creek with a 14 year fire exclusion history). hundreds rather than tens of years) if fragmentation due to So far,Western Australia appears to have escaped the clearing does not inhibit dispersal patterns. inexplicable patterns of frog decline seen along much of the Most populations of G. alba are small, with less than fifty adult eastern coast. However, with little deliberate survey activity, male frogs (Wardell-Johnson et al. 1995).There must be a except for species known to be at risk, and increasing high risk of stochastic loss with such small population sizes salinity in areas extensively cleared and used for wheat production (e.g. Main 1990) frogs in Western Australia face (Lande 1988) particularly as most adults may only breed an uncertain future. once and die at two to four years old (Driscoll 1996). Small populations may also be subject to inbreeding depression which could cause extinction (cf. Frankham and Ralls 1998; ACKNOWLEDGEMENTS Saccheri et al. 1998). All subpopulations that went extinct in Research on Geocrinia was supported by funds from the the period 1994 to 1997 contained fifteen or fewer calling Geocrinia Recovery Team from the Endangered Species males in 1994 (average of 5.4 frogs, range 1–15).With nearly Program. Research on Spicospina supported by a grant to JDR all populations in this size range (Wardell-Johnson et al. 1995) from the Environment Forest Taskforce contracted through the there are likely to be many more local extinctions in the Department of Conservation and Land Management,Western immediate future if these are the result of small population Australia. Data on frog occurrences in the wheatbelt size. Population modelling may give some clues to future supported by ARC grants A18715137 and A19030931 to JDR. management of populations and directions for allocation of Thanks to Greg Voigt for assistance with surveys of Geocrinia conservation effort on all land tenure types but the prognosis populations in 1997 and the Geocrinia Recovery Team for its for G. alba is comparatively poor. support. JDR acknowledges help from Oliver Berry, Don Driscoll, Kelley Whitaker, Jason How and Phil Byrne for field Spicospina flammocaerulea work on Spicospina and Greg Freebury and John Tillman, Although this species has a current status of vulnerable this is CALM,Walpole for assistance and information. Don Driscoll based on inadequate knowledge in two areas. First, there are made initial counts for populations of G. vitellina and G. alba no reliable data on true population size. Second, major survey (1992–1994) reported in Figures 4 and 5. effort in 1997 located 8 new populations — with five of those in areas previously never surveyed.The potential for discovery of new populations is still high.

There are no major threats to populations on publicly managed lands that cannot be controlled by appropriate intervention (e.g. fire) but there has been no analysis of threats to populations found on private property.The prognosis for this species is fair which could upgrade to good with better data on distribution and abundance of populations.

183 REFERENCES Main, B.Y., (1990) Restoration of biological scenarios: the role of museum collections. Proc. Ecol. Soc. Aust. 16: 397-409 Berger, L., Speare, R., Daszak, P., Green, D. E., Cunningham, A. A., Goggins, C. L., Slocombe, R., Ragan, M. A., Hyatt, McDonald, L., (1998) The legal protection of fauna habitat in A. D., McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, Western Australia and the consequences for the long- G. and Parkes, H., (1998) Chytridiomycosis causes term survival of fauna. Unpublished Honours thesis, amphibian mortality associated with population declines in Department of Zoology, University of Western Australia. the rain forests of Australia and central America. Proc. Roberts J. D., (1983) Terrestrial breeding in the Australian Natl. Acad. Sci. U. S. A. 95: 9031-9036. leptodactylid frog Myobatrachus gouldii (Gray) Aust.Wildl. Berry, O., (1996) The effect of a naturally fragmented Res. 8: 451-62. breeding habitat on the population genetic structure of Roberts, J. D., (1996) Geographic variation in male call and the sand frog, Heleioporus psammophilus (Anura: species boundaries in tetraploid frogs of the Australian Myobatrachidae) in south-western Australia. genus Neobatrachus (Myobatrachidae). Aust. J. Zool. Unpublished Honours thesis, Dept of Zoology, University 45: 95-112. of Western Australia. Roberts, J. D., (1997) Call evolution in Neobatrachus (Anura: Davis, R., (1997) The effects of habitat fragmentation on frog Myobatrachidae): speculations on the origins of tetraploids. populations in the wheatbelt of Western Australia. Copeia 1997: 791-801. Unpublished Honours thesis, Murdoch University. Roberts, J. D., Horwitz, P.,Wardell-Johnson, G., Maxson, Driscoll D. A., (1996) Understanding the metapopulation L. R. and Mahony, M. J., (1997).Taxonomy, relationships and structure of frogs in the Geocrinia rosea complex through conservation of a new genus and species of myobatrachid population genetics and population biology: implications frog from the high rainfall region of southwestern for conservation and evolution. Unpublished Ph D thesis, Australia. Copeia 1997: 373-381. Department of Zoology, University of Western Australia. Roberts, J. D. and Wardell-Johnson, G., (1995) Call structure Driscoll, D. A., (1997) Mobility and metapopulation structure differences between isolated populations of Geocrinia of Geocrinia alba and Geocrinia vitellina, two endangered (Anura: Myobatrachidae) from south-western Australia. frog species from southwestern Australia. Aust. J. Ecol. Copeia 1995: 899-906. 22: 185-195. Roberts, J. D.,Wardell-Johnson, G. and Barendse,W., (1990) Driscoll, D. A., (1998a) Genetic structure, metapopulation Extended descriptions of two new species of Geocrinia processes and evolution influence the conservation (Anura: Myobatrachidae) from south-western Australia, strategies for two endangered frog species. Biol. Cons. with comments on the status of G. lutea. Rec.West. Aust. 83: 43-54. Mus. 14: 427-437. Driscoll, D. A., (1998b) Genetic structure of the frogs Saccheri, I., Kuussaari, M., Kankare, M.,Vikman, P., Fortelius,W. Geocrinia lutea and Geocrinia rosea reflects extreme and Hanski, I., (1998). Inbreeding and extinction in a population divergence and range changes, not dispersal butterfly metapopulation. Nature 392: 491-494. barriers. Evolution 52: 1147-57. Tyler, M. J., (1997) The Action Plan for Australian Frogs. Driscoll, D. A. and Roberts, J. D., (1997) Impact of fuel Wildlife Australia, Endangered Species Program, Canberra. reduction burning on the frog Geocrinia lutea in south- Tyler, M.J., Smith, L.A. and Johnstone, R.E., (1994) Frogs of west Western Australia. Aust. J. Ecol. 22: 334-339. Western Australia.West Australian Museum: Perth. Frankham, R. and Ralls, K., (1998) Inbreeding leads to Wardell-Johnson, G. and Roberts, J. D., (1989) Endangered! extinction. Nature 392: 441-442. Forest frogs. Landscope 5 (1): 17. Horwitz, P., Pemberton, M. and Ryder, D., in press. Wardell-Johnson, G. and Roberts, J. D., (1991) The survival Catastrophic loss of organic carbon from a management status of the Geocrinia rosea (Anura: Myobatrachidae) fire in a peatland in south-western Australia. In,Wetlands complex in riparian corridors: biogeographical implications. for the future, A. J. McComb & J. A. Davis (eds). Gleneagles pp. 167-175, in, D. A. Saunders and R. J. Hobbs (eds) Press: Adelaide. Nature Conservation 2:The Role of Corridors, Surrey IUCN, (1994) IUCN red list categories. IUCN Species Beatty and Sons, Sydney. Survival Commission, Gland, Switzerland. Wardell-Johnson, G. and Roberts, J. D., (1993) Biogeographic Lande, R., (1988) Genetics and demography in biological barriers in a subdued landscape: the distribution of the conservation. Science 241: 1455-1460. Geocrinia rosea (Anura: Myobatrachidae) complex in Littlejohn, M. J. and Wright, J. R., (1997) Structure of the acoustic south-western Australia. J. Biogeog. 20: 1-15. signals of Crinia glauerti (Anura: Myobatrachidae) from south- Wardell-Johnson, G., Roberts, J. D. , Driscoll, D. and Williams, western Australia, and comparison with those of C. signifera K., (1995) Orange-bellied and white-bellied frogs recovery from South Australia.Trans. R. Soc. S. Aust. 121: 103-117. plan.Wildlife management program No. 10, Department Mahony. M., (1996) The decline of the green and golden bell of Conservation and Land Management, Perth. frog Litoria aurea viewed in the context of declines and Watson, G. F. and Gerhardt, H. C., (1997) The breeding disappearances of other Australian frogs. Australian biology and advertisement call of Litoria splendida Tyler, Zoologist 30: 237-247. Davies and Martin.Trans. R. Soc. S. Aust. 121: 119-124.

184 Toxicological issues for amphibians in Australia

Reinier Mann1 and Joseph Bidwell2

ABSTRACT are phylogenetically distinct. Interspecific variation in chemical sensitivity has been demonstrated As a consequence of agricultural, urban and amongst northern hemisphere phylogenetic industrial development, the chemical profiles of groupings. Australian soils and waterways are being altered.

One could expect that this kind of habitat The discipline of ecotoxicology has much to alteration may have deleterious effects on native contribute to a better understanding of these risks, fauna such as frogs.At present however, it is unclear with several methodological approaches available, whether environmental contaminants pose a threat although standardised protocols still need to Australian amphibian populations.To date there development. have been very few toxicological studies examining the effects of environmental chemicals such as INTRODUCTION pesticides, fertilisers and metals, or the effects of In recent years amphibians have been proffered as good changes in soil chemistry on Australian fauna. indicators of environmental contamination because of their unique physiology. Amphibians are the highest vertebrate For some of these contaminants, there is a group to retain an essentially “naked” egg, and the only vertebrate group which has an aquatic larval stage and a reasonable volume of literature for northern terrestrial adult phase. Furthermore, the skin of an adult hemisphere species, although these studies are amphibian is a permeable organ used for respiration and lacking in uniformity, preventing useful water-balance whereas the tadpole stage relies predominantly on gills for respiration.This dual life cycle implies that comparisons. Furthermore, these studies may not amphibians may have more opportunities for exposure and reflect the risks posed to Australian species which

1 School of Environmental Biology, Curtin University of Technology, PO Box U1987, Perth,Western Australia, 6845. 2 Aquatic Toxicology Unit, School of Pharmacy and Medical Sciences, University of South Australia, North Terrace, Adelaide, South Australia, 5000.

185 more modes of exposure to environmental contaminants toxicology references accumulated by the Canadian Wildlife than other vertebrates, although the notion that amphibians Service (http://www.cciw.ca/green-lane/herptox/reference- exhibit greater sensitivity to environmental contaminants than list.html), Power et al. (1989), Ferraro and Burgin (1993) and other phylogenetic groups has yet to be verified (see below). Tyler (1994).This review will cover three main questions:

Amphibians do however, provide an extremely visible 1. Which chemicals pose a threat to Australian frogs? indicator of pollution.The recent discovery by a school group 2. What tools are available to study toxic effects in frogs? of several deformed frogs in Minnesota (Schmidt 1997) 3. What are the problems associated with amphibian brought public focus to a water contamination issue. Indeed, toxicology? the incidence of frog limb abnormalities has been proposed as a useful indicator of environmental contamination (Tyler WHICH CHEMICALS POSE A THREAT TO 1994; Ouellet et al. 1997; Read 1997) and has been used to AUSTRALIAN FROGS ? evaluate the hazards to wildlife posed by mining operations at Olympic Dam in South Australia (Read and Tyler 1990; Pesticides 1994; Read 1997) and Jabiru in the Northern Territory during the 1970s (Tyler 1994). The largest single group of potential chemical pollutants that Australian frogs might encounter are the various pesticides Amphibians as a group have only recently been included in employed in agriculture and pest management in Australia. routine toxicological assessment of environmental chemicals. Much of the recent work examining the effects of pesticides More than 45% of the peer reviewed literature which focuses on amphibians has concentrated on the newer generations of on amphibians in toxicology oriented studies has been pesticides such as pyrethroids, carbamates, and published in the last ten years. Presumably this reflects the organophosphates (see below), although there has been a increased scientific and community interest in amphibians resurgence of interest in the older organochlorine insecticides following much publicity regarding their apparent global (i.e. DDT) because of their environmental persistence and decline, as well as a greater appreciation of the importance of possible links to amphibian decline. For example, studies in the these animals as an integral part of the food chain in many of Sierra Nevada and Cascade Mountains of the USA indicate the world’s ecosystems. Amphibians may actually constitute that wind-born topsoil bearing pesticide residues from the the largest fraction of vertebrate biomass in some central Californian agricultural valley, may be responsible for ecosystems, making them an important source of food for declines of several species of amphibians (Cory et al. 1970; higher vertebrates such as fish, birds, reptiles and mammals, as Bradford et al. 1994; Fellers 1997). At this stage a direct causal well as important herbivores (tadpoles) and carnivores in relationship has not been well established. Dramatic declines these ecosystems (Blaustein et al. 1994). of the high altitude species Rana muscosa in Kings Canyon National Park in the Sierra Nevada mountains in California Australia has one of the most diverse amphibian assemblages were preceded by an outbreak of redleg disease caused by in the world with more than 210 species of frogs the pathogen Aeromonas hydrophila (Bradford 1991).The representing 29 genera and four families (Cogger 1992; reason for the outbreak is unknown, however, given that Barker et al. 1995;Tyler, 1997).While it is surprising that more immunosuppression has been reported in other vertebrate attention has not been paid to the effects of environmental species following exposure to DDT and other pesticides pollutants upon these animals, environmental toxicology is (Barnett and Rodgers 1994), one can speculate about the itself a relatively new field in Australia, with few toxicological immunosuppressive effects of pesticide exposure in Rana studies carried out using native fauna of any kind. muscosa. One possible contributing factor in recently Consequently, regulatory authorities such as the National documented declines in high elevation species along the Registration Authority for Veterinary and Agricultural eastern slopes of the Great Dividing Range of eastern Chemicals (NRA) and the Environmental Protection Agency Australia (Richards et al. 1993) is that of a fungal pathogen (EPA) rely on studies conducted with North American, South (Berger et al. 1999) and this also may be related to the African or European species. More than 80% of the immunosupressive effects of environmental contaminants. amphibian toxicological literature describes studies using representatives of only three genera — Rana spp, Bufo spp or Insecticides — organochlorines Xenopus laevis. In Australia, only one species belongs to the genus Rana (a relatively recent colonist from New Guinea, Organochlorines were the first synthetic chemical pest restricted to Cape York Peninsula) and the only Bufo species control agents, with DDT being the archetypal is represented by the introduced pest species Bufo marinus organochlorine insecticide. DDT was used extensively in (cane toad). developed nations until it was banned from use in the USA in 1972 and in Australia in the mid 1980s. DDT is still widely This review is intended to highlight some of the potential used as an insecticide in third world countries (Lambert 1993; risks to Australian frogs from exposure to chemical 1997).While there are a few early studies on the effects of contaminants and environmental imbalances. Moreover, we DDT on amphibians (Ellis et al. 1944; Herald 1949; Langford hope to place the current body of literature dedicated to 1949; Logier 1949; Speirs 1949;Tarzwell 1950;Vinson et al. toxicology in Australian species of frogs within the context of 1963; Isaacson 1968) research only commenced in earnest in the general amphibian toxicological literature. In doing so, we the 1970s. One of the main proponents of these studies was also intend to provide an introduction to the amphibian A.S. Cooke of the Monks Wood Experimental Station in toxicological literature. It is not however, an exhaustive Huntington England, who published much of the work on treatment of the available literature. Other valuable sources DDT toxicity in amphibians (Cooke 1970; 1971; 1972; 1973a; of information include the list of amphibian and reptile 1973b; 1974; 1979; Osborn et al. 1981).

186 DDT and most other organochlorines are characterised by 1990; Swann et al. 1996). Interestingly, adult amphibians high environmental persistence.Their persistence and appear to be tolerant of severe acetylcholinesterase inhibition subsequent biomagnification through the food chain (Meeks (Balasundaram and Selvarajan 1990;Wang and Murphy 1982) 1968; Licht 1976; Niethammer et al. 1984; Russell et al. 1995) and the developmental toxicity of organophosphates (Elliot- are among the factors which have led to their reduction in Feeley and Armstrong 1982; Snawder and Chambers 1990; use. However, more than 20 years after it was banned from 1993; Alvarez et al. 1995) may be more ecologically use in North America, DDT is still being detected in important. Also, there are a few studies which have indicated amphibians from those areas in which the pesticide was that these chemicals can bioaccumulate (Hall and Kolbe applied (Russell et al. 1995). In Australia, DDT and other 1980; Fleming et al. 1982; Powell et al. 1982; Hall 1990). organochlorines were also widely used, as indicated by a survey carried out in Western Australia which detected, In Australia, organophosphate insecticides are used amongst other organochlorines, DDT and dieldrin in 39.6% extensively in agriculture and within urban areas (e.g. golf and 39.0% of 11 248 soil samples respectively (EPA WA courses, turf clubs etc.).There are no studies which have 1989). Despite this widespread contamination, we are aware examined their potential risks to Australian frogs. of only one study that has measured organochlorine residues in a native Australian amphibian (Birks and Olsen 1987). Insecticides — carbamates Like organophosphates, the low persistence of carbamate Organochlorines are still being used to some extent in insecticides has led to their widespread acceptance as a Australia. Reports of fish kills in September 1997 in the Ord replacement for the more traditionally used organochlorines. River agricultural zone of the Kimberly region of north- Again however, there has been very little research to evaluate western Australia, were attributed to the application of the toxicity of these chemicals to amphibians, although there endosulfan to control cotton pests.We found only six studies are some indications that field concentrations of carbamates examining the effects of endosulfan on amphibians (Cockbill following application of these insecticides may be detrimental 1979; Gopal et al. 1981; Hashimoto and Nishiuchi 1981; to amphibians (Tucker and Crabtree 1969; Flickinger et al. Vardia et al. 1984; Abbasi and Soni 1991; Berrill et al. 1998). 1980; Marian et al. 1983; Bridges 1997). Furthermore, One of these studies (Gopal et al. 1981) found endosulfan to carbaryl has been demonstrated to penetrate amphibian skin be an order of magnitude more acutely toxic to Rana tigrina more rapidly than organochlorines (dieldrin and DDT), tadpoles than to the catfish Clarias batrachus and damselfly organophosphates (parathion) or pyrethroids (permethrin) nymphs (Enallagma spp.). Mulla et al. (1963), also noted that (Shah et al. 1983). Carbamate based insecticides have also endosulfan/toxaphene application was effective in producing been found to produce developmental malformations in an “almost complete kill” of “public nuisance” bullfrogs. skeletal tissue (Alvarez et al. 1995) and musculature (Rzehak Furthermore, in their review of the amphibian toxicological et al. 1977; Cooke 1981).The effects on Australian species literature, Power et al. (1989) ranked endosulfan as the have yet to be investigated. second most acutely toxic chemical thus far tested on amphibians, being surpassed only by the heavy metal mercury. Insecticides — pyrethroids To date there has been no research into the possible consequences of endosulfan exposure to frog species of Pyrethroids have gained a reputation as “safe” insecticides and Australia’s Kimberly region or other agricultural areas where are widely used in agricultural, aquatic and household this chemical is still being applied. products (Elliot et al. 1978; Smith and Stratton 1986).There is some indication however, that field application of these Organochlorines are known to produce developmental chemicals may be deleterious to amphibians (Jolly et al. 1978; abnormalities in amphibians (Cooke 1970; 1972; 1973b; Thybaud 1990; Berrill et al. 1993; Materna et al. 1995). Brooks 1981; Marchal-Ségault and Remande 1981; Osborn Pyrethroids appear to affect voltage-dependent et al. 1981; Gavilan et al. 1988; van der Bercken et al. 1989). neuromuscular sodium channels producing tremors, One of these references (Brooks 1981) examined the hyperexcitation and convulsions (van den Bercken 1977; teratogenic effects of dieldrin on the Australian frog Vijverberg et al. 1982; Ruigt and van den Bercken 1986). Limnodynastes tasmaniensis, however this is the only study of this kind for an Australian species. Although pyrethroids are used extensively in Australia, there are no published studies on the effects of these chemicals on Insecticides — organophosphates Australian frogs. As far as we are aware, a B.Sc. honours thesis by Millen (1995) presents the only information available Organophosphates replaced organochlorines by virtue of for an Australian species that we are aware of. It indicated their lower environmental persistence.The first notable studies temporary increases in acetylcholinesterase, growth inhibition to examine the effects of organophosphates on amphibians and behavioural effects following exposure to cypermethrin. were published in the early sixties (Edery and Schatzberg- Porath 1960; Mulla 1962; Mulla et al. 1963). More recently, Herbicides and Fungicides several studies indicated that standard field application rates of organophosphate insecticides may have a deleterious effect on The first publication to investigate the potential hazards of a amphibian populations (Anguiano et al. 1994; Berrill et al. herbicide to amphibians emerged in 1970 (Hazelwood 1970). 1994; 1995; Schuytema et al. 1995; Sparling et al. 1997). Since then approximately 70 articles dealing with the toxicology of herbicides and fungicides have been published. The established mechanism of organophosphate toxicity is The scope of chemical species which these references cover through non-reversible acetylcholinesterase inhibition (Dekins amounts to more than 40 different compounds.There are et al. 1978; Llamas et al. 1985; Balasundaram and Selvarajan however, more than 100 chemical compounds registered for

187 use as herbicides in Australia and more than 60 registered for Fertilisers use as fungicides (Department of Agriculture 1994). It is The potential hazards associated with agricultural fertilisers reasonable to suggest that no particular herbicide or have only recently been proposed as a potential threat to fungicide has been adequately studied. amphibians (Berger 1989). Berger (1989) noted a correlation Two of the more wide-ranging studies by Johnson (1976) and between amphibian declines and environmental increases in Sanders (1970), have examined the acute toxicity of nitrates and ammonia. Subsequent laboratory studies have numerous herbicides to tadpoles or adult frogs and provide a concentrated on the effects of nitrate on amphibian survival useful overview of herbicide toxicity.The only compounds and growth (Baker and Waights 1993, 1994; Hecnar 1995; that have received a notable level of attention are the: Watt and Oldham 1995; Hecnar and McLoskey 1996; de Wijer et al. 1997; Oldham et al. 1997; Xu and Oldham 1997) • phenoxyacid herbicides (i.e. 2,4-D and MCPA) (Sanders but the results are somewhat equivocal. Hecnar (1995) 1970; Cooke 1972; Zaffaroni et al. 1986a; 1986b; Zavanella reported 96 hour LC s of 13.6-39.3 mg/L for Bufo et al. 1988; Arias et al. 1989; Leone et al. 1994;Vismara et 50 americanus when exposed to ammonium nitrate. Similarly, in al. 1995; Bernardini et al. 1996;Vismara et al. 1996;Vismara two studies, one of which is the only study to have examined and Garavaglia 1997); an Australian species, Baker and Waights (1993, 1994) • dithiocarbamate fungicides (Prahlad et al. 1974; Zaffaroni reported reduced feeding and weight loss in Bufo bufo and et al. 1978; Arias and Zavanella 1979; Zavanella et al. 1979, Litoria caerula tadpoles at 40 and 100 mg/L sodium nitrate 1984; Seugé et al. 1983; Birch and Prahlad 1986a, 1986b); respectively. In contrast, Xu and Oldham (1997) reported a 96 2- and hour LC50 of over 1000 mg/L N03 (as ammonium nitrate) • paraquat/diquat herbicides (Sanders 1970; Anderson and for Bufo bufo tadpoles and increased growth at 2- Prahlad 1976; Johnson 1976; Cooke 1977; Paulov 1977b; 50 mg/L N03 , while de Wijer et al. (1997) reported a Bimber and Mitchell 1978; Hashimoto and Nishiuchi 1981; disparity in the toxicity of ammonium nitrate and calcium Dial and Bauer 1984; Dial and Bauer-Dial 1987; Lindquist nitrate. Until these discrepancies are clarified the role of et al. 1988; Linder et al. 1990; Dial and Dial 1995; agricultural fertilisers in amphibian decline remains contentious. Lajmanovich et al. 1998). It is worth noting that the Australian frog fauna has evolved Many of these studies have indicated that these three classes in an environment that is comparatively depauperate in of chemicals have teratogenic effects. nitrate and phosphate (Lamont 1994) which may accord Australian frogs a greater sensitivity to these chemicals. Three published studies have incorporated Australian species. Johnson (1976) used four species of Australian frogs: Impurities in fertilisers include cadmium, lead and mercury Adelotus brevis, Limnodynastes peronii, Limnodynastes (State of the Environment Advisory Council 1996) and will tasmaniensis and Litoria ewingii.This particular study be addressed in the following section on metals. represents the only comprehensive toxicological study for Australian species with regard to any environmental Metals contaminant. More recently, Bidwell and Gorrie (1995) and Metal contamination of agricultural land/waterways occurs as Mann and Bidwell (1998), presented data on the acute a consequence of fertiliser impurities such as cadmium, lead toxicity of glyphosate formulations to tadpoles or adult frogs and mercury (State of the Environment Advisory Council of the south-western Australian species, Litoria moorei, Litoria 1996). Similarly, coastal wetlands are contaminated by various adelaidensis, Crinia insignifera, Heleioporus eyrei, and metals through urban and industrial runoff.The effect that Limnodynastes dorsalis. these metals may have on Australian frogs has yet to be investigated.Wetland bioremediation systems are slowly being Pesticide mixtures incorporated as integral parts of urban and industrial waste Pesticides are rarely applied in isolation. Usually a combination management. Such wetlands could provide a valuable of pesticides is mixed together and applied as a single laboratory for the examination of metal exposure effects on application cocktail. Furthermore, commercial preparations Australian frogs. are often a combination of two or more pesticides, or they The literature which deals with the effects of metal incorporate various solvents, carriers or surfactants.These contaminants is relatively expansive and has been treated various combinations may have additive, synergistic or comprehensively by Power et al. (1989). Recent studies of note antagonistic toxicological effects (Landis and Yu 1995). A few include several which have examined the teratogenic effects of studies have examined the toxicology of pesticide mixtures various metals including heavy metals, particularly cadmium and on amphibians (Anderson and Prahlad 1976; Berrill et al. lead (Pérez Coll and Herkovits 1990; Nebeker et al. 1994; 1993, 1994; Howe et al. 1998) while others have noted Plowman et al. 1994; Sunderman et al. 1995; Herkovits et al. differences in toxicity between technical grade pesticides and 1997; Rowe et al. 1998) and divalent metals such as zinc, formulated products (Bidwell and Gorrie 1995; Schuytema copper, cobalt and nickel (Hopfer et al. 1991; Plowman et al. et al. 1995; Swann et al. 1996; Mann and Bidwell 1998). 1991; Luo et al. 1993, 1994; Sunderman et al. 1995).

188 Endocrine disrupting chemicals (EDCs) Salinity Many environmental contaminants are now known to behave Dry land salinity and irrigation-induced salinity are possibly as hormone mimics, and there is much concern that wildlife is the most important environmental problems facing Australia. being affected (Raloff 1994). One class of chemicals of By 1994 approximately nine percent of land cleared for concern is the alkylphenolic surfactants.World-wide, agriculture in south-west Western Australia (1.6 million approximately 500 000 tonnes of alkylphenol based hectares) was affected by dryland salinity, with an average surfactants are produced annually for use in detergents, increase of 0.07% per year. In 1992, 200 000 hectares of the paints, pesticides, textile and petroleum recovery chemicals, Murray-Darling Basin were similarly affected. In South metal working lubricants and personal care products (Renner Australia and Victoria, 1993 estimates stood at 400 000 and 1997).While not directly concerned with amphibians, one of 150 000 hectares respectively, with smaller estimates the most notable studies involving alkylphenols is that by beginning to come from other states. Shallow water-tables Jobling et al. (1996), which reported induction of with accompanying salinity problems were estimated to affect vitellogenesis (a process normally dependent on endogenous 360 000 hectares of the Murray-Darling Basin and 199 000 oestrogens) and concomitant inhibition of testicular growth in hectares of the Wakool, Deniliquin and Murrumbidgee male rainbow trout following exposure to alkylphenolic irrigation areas (State of the Environment Advisory compounds at the 30 ppb level. Council 1996).

Very little work has been published on the potential effects of The severity of salinity induced land degradation in Australia EDCs in amphibians and only one has appeared in the refereed is perhaps analogous to that of acid rain in the northern literature (Palmer and Palmer 1995). A few conference abstracts hemisphere. Certainly the widespread nature of the two however, indicate that amphibian development may be affected phenomena is comparable and the consequential changes in by oestrogen mimics such as DDT (Palmer and Palmer 1995; water chemistry can be expected to exert comparable Hayes and Noriega 1997), polychlorinated hydrocarbons physiological stresses on amphibians. A considerable volume (Glennemeier 1997) and alkylphenolic compounds (Palmer et of literature (more than 140 journal articles) is dedicated to al. 1996; Ramsdell et al. 1996). the examination of acid tolerance/sensitivity amongst amphibians (for reviews see Pierce 1985; Freda 1986; Ferraro and Burgin 1993). In contrast to this impressive body of literature there are fewer than 30 studies which have examined salt tolerance/sensitivity in amphibians.

TABLE 1: Studies which have investigated the toxic effects of environmental contaminants in Australian species. OC organochlorine, P pyrethroid.

Toxin Species Study type Reference

Organochlorines Limnodynastes tasmaniensis Fat residues Birks and Olsen 1987 Dieldrin (OC) Limnodynastes tasmaniensis Sublethal – development Brooks 1981 Cypermethrin (P) Litoria ewingi Sublethal – growth, behaviour Millen 1995 and acetylcholinesterase levels Herbicides Adelotus brevis; Limnodynastes peronii; Acute toxicity and Sublethal – Johnson 1976 L. tasmaniensis; Litoria ewingi thermal tolerance Glyphosate-based herbicide Crinia insignifera; Litoria adelaidensis; Acute toxicity Bidwell and Gorrie 1995 L. moorei Glyphosate-based herbicides Crinia insignifera; Heleioporus eyrei; Acute toxicity Mann and Bidwell 1998 Limnodynastes dorsalis; Litoria moorei Sodium fluoroacetate (1080) Limnodynastes tasmaniensis Intraperitoneal administration McIlroy et al. 1985 Sodium nitrate Litoria caerulea Sublethal- growth and behaviour Baker and Waights 1994 Sodium chloride Crinia pseudinsignifera; Heleioporus Acute toxicity and Sublethal Baumgarten 1991 albopunctatus; Pseudophryne guentheri – avoidance Field correlations Sodium chloride Limnodynastes peronii; Uperoleia laevigata Actute toxicity and Sublethal Ferraro 1992 – growth Sodium chloride Limnodynastes tasmaniensis Sublethal – growth Quincy 1991 Sodium chloride Multiple species Sublethal – righting reflex Tyler 1972 Radiation Limnodynastes tasmaniensis Sublethal – development Panter 1986 Radiation Limnodynastes tasmaniensis Sublethal – oxygen consumption Panter et al. 1987 Metals Neobatrachus centralis Limb abnormalities – monitoring Read 1997 Metals Neobatrachus centralis Limb abnormalities – monitoring Read and Tyler 1990 Metals Neobatrachus centralis Limb abnormalities – monitoring Read and Tyler 1994 Copper Limnodynastes dorsalis; Litoria raniformis Liver residues Beck 1956

189 A wide variation in salt tolerance in amphibians has been restricted to a suite of representative species which could be reported (see Liggins and Grigg 1985; Ferraro and Burgin reared in adequate numbers in captivity. Interspecific variation 1993). In general however, amphibians must maintain in acute responses is unlikely to be great enough to invalidate hyperosmoticity to their environment and salinities greater data collected in common or cultured species, provided there than 25% sea-water present an osmotic challenge.This is strict adherence to established protocols. challenge is occasionally met by increasing plasma concentrations of chloride and urea, thereby increasing About 30% of the amphibian toxicology studies reviewed plasma osmotic pressure (Liggins and Grigg 1985). Such a present acute toxicity data in one form or another. Since the response indicates a degree of adaptability in those species majority of this work has remained in the realm of examined, including Bufo marinus (Liggins and Grigg 1985), “herpetological studies” rather than “toxicological studies” Bufo viridis (Katz 1973), Bufo bufo (Ferreira and Jesus 1973), there has been virtually no adherence to standard Rana temporaria (Ackrill et al. 1969), Rana cancrivora (Gordon methodology. Sanders (1970) presented a reasonably detailed et al. 1961) and Xenopus laevis (Romspert 1976). It is not account of a methodology used to generate LC50 data for known whether Australian species show a similar level of several different pesticides on two species of anurans.This plasticity. In an examination on the likelihood of frogs crossing protocol was subsequently adopted by Johnson (1976) in his the Torres Strait,Tyler (1972) examined salt tolerance in studies with Australian animals and his data are somewhat several Australian and New Guinean species by exposing comparable to later more rigorous studies (e.g.Wohlgemuth them to seawater and observing the time required for loss of 1977; Jolly et al. 1978; Hall and Swineford 1980; Gopal et al. righting-reflex, but the results were somewhat equivocal. 1981;Thurston et al. 1985; Holcombe et al. 1987; Materna et Apart from Tyler (1972) we are aware of a further three al. 1995; Schuytema et al. 1995; Sparling et al. 1997; Xu and unpublished studies that have examined salt tolerance in Oldham 1997; Howe et al. 1998; Mann and Bidwell 1998). Australian species (Baumgarten 1991; Quincy 1991; Ferraro 1992 — see Table 1). Sub-chronic, chronic and non-lethal exposures SUMMARY OF TOXICOLOGICAL STUDIES The amphibian toxicological literature includes a number of INVOLVING AUSTRALIAN SPECIES sub-chronic and chronic exposure studies in which either long term survival or sub-lethal effects are documented. A selection A summary of Australian studies is presented in Table 1. of representative studies is presented in Table 1. Unfortunately There are eighteen studies listed, four of which are the absence of methodological uniformity precludes much unpublished theses. It is likely that there are other useful comparison between studies, although the tabulation unpublished theses residing within Australian universities.This system employed by Power et al. (1989) provides a useful table has been included to highlight the paucity of basis for comparing much of the earlier work.Various life information available for Australian species. stages have been used in these studies with most focusing on various aspects of embryo and tadpole growth, development WHAT TOOLS ARE AVAILABLE TO STUDY or survival. Studies with adults have been uncommon and will TOXICOLOGY IN FROGS? be discussed further in a later section.

Acute toxicity tests Frog Embryo Teratogenesis Assay-Xenopus One of the most widespread protocols for assessing toxicity (FETAX) of a chemical to an aquatic species is the acute toxicity test. The African clawed toad (Xenopus laevis) is one of the most Standard procedures for conducting these tests can be found widely used laboratory animals in the world. It is both extremely in a range of sources (U.S.EPA 1991; ASTM 1993a; OECD easy to maintain and breed in captivity, and its embryological 1993). ASTM (1993a) includes a specific protocol for testing development has been described in detail (Nieuwkoop and Faber tadpoles, fish and macroinvertebrates (ASTM 1993c). Acute 1975). FETAX is a recently developed technique (Dumont et al. Toxicity tests will usually run for multiples of 24 h, with 48 1983) which uses Xenopus laevis embryos to ascertain the and 96 h tests being most common.The results of such tests teratogenic potential of environmental chemicals and has been are usually expressed as point estimates such as the LC50- a adopted as a standard bioassay for teratogenicity (ASTM 1993b). It statistically or graphically estimated concentration that is has not however, been developed with frog conservation as the expected to be lethal to 50% of a group of organisms under primary motivation. Of approximately 70 studies published by the specified conditions. Other endpoints include the LD50 which end of 1997, a number focus on validation of the test as a bioassay is a lethality endpoint in which the toxicant is administered for teratogenicity, and only a few have examined chemicals which orally.The use of a 50% effect on the test population is for may pose a threat to amphibians in the field (Dawson et al. 1985; statistical reasons only, not an indication that only 50% of the Birch and Prahlad 1986b, 1988; Snawder and Chambers 1989, population should be protected. 1990; Hopfer et al. 1991; Sunderman et al. 1991; Hauptman et al. Acute toxicity tests provide an inexpensive, rapid and simple 1993; Luo et al. 1993; Presutti et al. 1994; Bernardini et al. 1996; method for accumulating base-line data on the toxicity of a Morgan et al. 1996;Vismara et al. 1996;Winchester et al. 1996; chemical to a specific organism. Such data can be used to Dumont and Bantle 1997; Schrock et al. 1997). Furthermore, compare the relative toxicity of different chemicals or the Xenopus laevis is a somewhat atypical amphibian belonging to the relative sensitivities of different species.These tests demand family , and may not provide information relevant to other the availability of large numbers of test animals.This species.The technique does however, provide a template for a requirement precludes the use of rare or endangered stringent and highly reproducible methodology for evaluating species. Indeed, it may be preferable if this approach were embryotoxic potentials, and could be adapted to other species.

190 TABLE 2: A representative sample of sub-chronic and chronic studies or studies in which sub-lethal endpoints have been observed. OC organochlorine, OP organophosphate, C carbamate, H herbicide, F fungicide, DOC dissolved organic carbon.

Endpoint Toxicant Exposure stage and duration Reference

Hatching Success pH Embryos – hatching Freda and Dunson 1985 Petroleum oil Embryos – hatching Mahaney 1994 pH/aluminium Embryos – hatching Tyler Jones et al. 1989 pH/temperature Embryos- hatching Griffiths and De Wijer 1994 Growth retardation pH/aluminium Newly hatched – metamorphosis Cummins 1986 and survival Triphenyltin/pH Newly hatched – metamorphosis Fioramonti et al. 1997 pH/metals/DOC Newly hatched – metamorphosis Horne and Dunson 1995b pH/aluminium Newly hatched – 96h Jung and Jagoe 1995 2,3,7,8-tetrachlorodibenzo Embryos-metamorphosis Jung and Walker 1997 -p-dioxin (TCDD) Petroleum oil Larvae/Embryos – metamorphosis Lefcort et al. 1997; Mahaney 1994 Lindane (OC) Embryos – metamorphosis Marchal-Ségault and Remande 1981 Organophospates Embryos – metamorphosis Mohanty-Hejmadi and Dutta 1981 N-methyl-N’-methyl urea (H) 2 days post hatch – 80 days post hatch Paulov 1977a Pyrasophos (F) 2 days post hatch – 80 days post hatch Paulov 1981 pH Larvae – metamorphosis Rowe et al. 1992 pH/aluminium Embryos – metamorphosis Tyler Jones et al. 1989 Ammonium nitrate Larvae – metamorphosis Watt and Oldham 1995 Morphological Herbicides; fungicides Embryos for up to 5 days Anderson and Prahlad 1976 effects and Dithiocarbamate fungicides Embryos for up to 10 days Bancroft and Prahlad 1973 deformities Dieldrin (OC) Embryos for up to 35 days Brooks 1981 (see also section on DDT; Dieldrin (OC); 2,4D (H) Various stages for various time spans Cooke 1970; 1972; 1973b FETAX) Methyl mercury Embryos for up to 5 days Dial 1975 Paraquat (H) Various stages for various time spans Dial and Bauer 1984; Dial and Bauer- Dial 1987; Lajmanovich et al. 1998 Organophosphates Embryos Fulton and Chambers 1985 DDT (OC) Embryos Gavilan et al. 1988 Nickel Embryos Hauptman et al. 1993 Corticosterone Larvae (Gosner 39-40) for 9 days Hayes et al. 1997 Cadmium Various stages for 72 hours Herkovits et al. 1997 Primacarb (C) Embryos for up to 9 weeks Honrubia et al. 1993 Lindane (OC) Embryos-metamorphosis Marchal-Ségault and Remande 1981 Ethanol Embryos Nakatsuji 1983 DDT (OC) Larvae for 2 days Osborn et al. 1981 Malathion; fenitrothion (OP); Embryos for up to 96 hours Pawar et al. 1983; Pawar and benzene hexachloride (OC); Katdare 1984 carbofuran (C) Lead Various stage embryos for 20 hours Pérez Coll and Herkovits 1990 Cadmium Embryos Pérez Coll et al. 1986 Dithiocarbamate fungicide Embryos for 7 days Prahlad et al. 1974 Thiosemicarbazide Various stages for various time spans Riley and Weil 1987 Coal-ash polluted water Embryos – 80 days post hatch Rowe et al. 1998 Carbaryl (C) Early stage larvae for various time spans Rzehak et al. 1977 Malathion (OP) Embryos for up to 96 hours Snawder and Chambers 1993 Postmetamorphic Parathion-methyl (OP); Embryos – postmetamorphic Alvarez et al. 1995 persistence of Pirimicarb (C) juveniles (14 wks) deformities Nickel; cadmium; cobalt Embryos – postmetamorphic juveniles Plowman et al. 1994 (14 wks) Enzyme activity Parathion (OP) Embryos or 22 day old larvae for 120 hrs Anguiano et al. 1994 Organophosphates Tadpoles for 96 hours Hall and Kolbe 1980 DDT (OC) Embryos for 31 days Juarez and Guzman 1986 Pyrasophos (F) 2 days post hatch-22 days post hatch Paulov 1981 Temephos (OP) Early stage larvae for 96 hours Sparling et al. 1997 Malathion (OP) Early stage larvae for up to 144 hours Venturino et al. 1992 Behavioural effects Carbaryl (C) Larvae (Gosner 25) for up to 48 hours Bridges 1997 DDT; Dieldrin (OC); 2,4D (H) Various stages for various time spans Cooke 1970; 1972; 1973a Napthalene Three week old larvae for 96 hours Edmisten and Bantle 1982 pH Late stage larvae (hind paddles) for 5-8 days Griffiths 1993 Distillery effluent Larvae Haniffa and Augustin 1989 pH/aluminium Newly hatched larvae for 96 hours Jung and Jagoe 1995 pH Early stage larvae for less than 24 hours Kutka 1994 Triphenyltin 20 day old larvae for 48 hours Semlitsch et al. 1995 Lead Larvae for 6 days Steele et al. 1989

191 Mesocosms, Microcosm and Artificial Ponds entry, oral intoxication by consumption of contaminated food must also be considered. Only a few studies have examined The inadequacies of single species, laboratory based assays toxicity through oral administration (Rosato and Ferguson for predicting the ecosystem consequences of anthropogenic 1968;Tucker and Crabtree 1969; Hall and Swineford 1979; Dial pollutants has been voiced by leading ecotoxicologists and and Dial 1995) with most concentrating on skin exposure. amphibian ecologists (Kimball and Levin 1985; Rowe and Dunson 1994; Cairns et al. 1996). Microcosms and The latter studies can be split into those which have used mesocosms are more complex systems which provide more isolated skin preparations (Yorio and Bentley 1973; Celentano realistic exposure while allowing some level of experimental et al. 1979;Webb et al. 1979; Fromm 1981; Ferreira and Hill control.The categorical terminology denotes the scale of the 1982; Salibian 1983; Ardizzone et al. 1990; Lippe et al. 1992; system. Microcosms are generally small systems which can be Natochin and Jones 1992) and whole animal studies. Among set up on a laboratory bench, while mesocosms or artificial the whole animal studies, acute toxicity protocols are rare ponds are large tanks or permanent outdoor systems. Some (i.e. Kaplan and Yoh 1961; Kaplan and Glaczenski 1965; Kaplan may even take the form of in situ enclosures in streams or et al. 1967; Zaffaroni et al. 1986a; Mudgall and Patil 1987; ponds.The relative benefits of the various systems are still Mann and Bidwell 1998) and most studies have examined an being assessed, with the inherent problem being that the extremely wide range of sub-lethal effects such as greater complexity often makes it more difficult to develop behavioural effects (i.e. Cooke 1974; Hall and Swineford causal relationships between the presence of a contaminant 1980; Roudebush 1988; Haniffa and Augustin 1989; Antony and response. and Ramalingam 1990;), or physiological parameters including thermal tolerance (Johnson and Prine 1976), enzyme activity The use of mesocosms for amphibian studies was pioneered (Guzman and Guardia 1978; Deshmukh and Keshavan 1987; by Morin (1981) by employing 1000 litre cattle watering Joseph and Rao 1990; 1991; Mendiola and De Costa 1991), tanks. Such tanks have been used extensively since then for tissue metallothionein levels (Suzuki et al. 1986;Vogiatzis and the study of amphibian community dynamics (for review see Lombourdis 1998), limb regeneration (Manson and Rowe and Dunson 1994). More recently, similar tanks have O’Flaherty 1978; Zavanella et al. 1984; Pfeiffer et al. 1985; been used to examine the effects of acidity (Clark and Hall Arias et al. 1989; Nebeker et al. 1994), and metabolism 1985;Warner et al. 1991, 1993; Rowe et al. 1992; Sadinski (Mudgall and Patil 1987). and Dunson 1992; Horne and Dunson 1995a; 1995b), pyrethroid contamination (Materna et al. 1995), hydrocarbon Finally, approximately 50 studies were located which have contamination (Mahaney 1994; Lefcort et al. 1997) and utilised administration by hypodermic injection to examine a fertilisers (de Wijer et al. 1997) on multiple species systems. similarly wide variety of sub-lethal effects (e.g. Nagel and Urich 1981;Woodall and Maclean 1992; Scadding 1996). WHAT ARE THE PROBLEMS ASSOCIATED One study examined the acute responses to intraperitoneal WITH AMPHIBIAN TOXICOLOGY? injection of sodium fluoroacetate (1080) in the Australian species, Limnodynastes tasmaniensis (McIlroy et al. 1985), Adult/Larvae (Terrestrial /Aquatic) although the ecological significance of such a study is dichotomy questionable. Indeed the existence of numerous nephrostome like structures associated with frog kidneys One of the reasons that amphibians are considered good allowing rapid egress of water from the body cavity bioindicators of environmental contaminants is that the (Tyler pers. comm.) suggests that protocols that utilise permeable skin of the adult terrestrial phase of their life cycle intraperitoneal injection of a toxic agent may severely confers greater sensitivity than other vertebrates. One would underestimate the toxicological threat posed by that toxin. expect therefore, that equal weight would be given in the literature to studies that examine toxicity in adults. In actuality, One notable similarity in all of those studies which have it has been rare for researchers to test the toxicity of utilised adult subjects is the absence of any uniform pollutants on adults.There are two reasons for this. Firstly, methodology. Clearly a range of physiological and behavioural various studies have indicated that the larval or tadpole biomarkers have to be developed as standard test stages are more sensitive to pollutants than eggs (Cooke parameters for adult animals. 1972; Pritchard-Landé and Guttman 1973; Dial 1975; Greenhouse 1976; Bimber and Mitchell 1978; Saber and Species variability Dunson 1978; Birge et al. 1979; Hall and Swineford 1980; Davis et al. 1981; Mohanty-Hejmadi and Dutta 1981; Dial and Toxicological testing has for a long time relied on a small suite Bauer 1984; Herkovits and Jatimliansky 1986; Dial and Bauer- of aquatic test species. Prominent amongst these are salmon, Dial 1987; Anguiano et al. 1994; Berrill et al. 1994) while trout, sunfish, daphnids and amphipods. Several studies have others have shown that post-metamorphic adults are less attempted to evaluate the relative sensitivity of amphibian susceptible than larval stages (Hall and Swineford 1980; larvae when compared to such species (Jolly et al. 1978; Schultz et al. 1983; Bidwell and Gorrie 1995; Mann and Thurston et al. 1985; Holcombe et al. 1987;Thybaud 1990; Bidwell 1998). Consequently, most work has concentrated on Herkovits et al. 1995; Deyoung et al. 1996; McCrary and the larval tadpole stages. Heagler 1997) but the issue remains contentious, since the results of such studies are often dependent upon the toxin to Another factor which has led to a reluctance on the part of which test species are exposed (Thurston et al. 1985; researchers to examine toxicity in adult amphibians is deciding Holcombe et al. 1987). Also, only a limited range of amphibian on the most likely mode of exposure. Although the phylogenetic groups has been assessed in this manner. permeability of amphibian skin provides the obvious point of

192 Species variation amongst amphibians has received even less must be considered as a potential factor if animals are to be attention, although a few studies that have examined several removed from an already declining population base. species in parallel tests have reported species differences. Commercial breeding of native species needs to be Berrill et al. (1993, 1994, 1995) reported differences in the encouraged if amphibian susceptibility to environmental sensitivity of Bufo sp, Rana sp, and Ambystoma maculatum to contaminants is to be pursued as a line of research in Australia. the organophosphate pesticide fenitrothion and the pyrethroid pesticide permethrin. Hall and Swineford (1981) ACKNOWLEDGEMENTS reported differences in sensitivity of up to one order of magnitude amongst Bufo sp, Rana sp, Ambystoma sp and Acris The authors wish to acknowledge The World Wide Fund for crepitans following exposure to the halogenated pesticides Nature (WWF) whose financial support permitted Reinier endrin and toxaphene. Hoppe and Mottl (1997) noted Mann to present this paper at the National Threatened Frog species-specific differences in the occurrence, type and Workshop in Canberra in 1997.We also acknowledge severity of malformations amongst field collected animals in Dr Michael Tyler, Dr Roy Swain and Dr Linda Broadhurst for Minnesota.Wyman (1988) correlated distribution differences their constructive comments on the manuscript. for several species of amphibians with soil pH. Interspecific variation in acid tolerance has also been reviewed in Freda REFERENCES (1986) and Pierce (1985). Abbasi, S. A. and Soni, R., (1991) Evaluation of water quality Apart from a recent study examining variation in sensitivity criteria for four common between representatives of four temperate Australian frog Alvarez, R., Honrubia, M. P.and Herraez, M. P., (1995) Skeletal genera (Mann and Bidwell 1998), the degree of species malformations induced by the insecticides ZZ-Aphox and variation amongst Australian frogs, or between Australian Folidol during larval development of Rana perezi. Arch species and standard test species like Xenopus laevis or Rana Environ Contam Toxicol, 28: 349-356. sp. is unknown. A comparative study by Mann and Bidwell Anderson, R. J. and Prahlad, K.V., (1976) The deleterious (unpubl. data) indicated little variation in sensitivity to effects of fungicides and herbicides on Xenopus laevis agricultural surfactants between the tadpoles of four embryos. Arch Environ Contam Toxicol, 4: 312-323. Australian species and the two exotic species Bufo marinus Anguiano, O. L., Montagna, C. M., Chifflet De Llamas, M., and Xenopus laevis under standard test conditions. Such Gauna, L. and Pechen De D’Angelo, A. M., (1994) studies however, fail to consider the influence of high Comparative toxicity of parathion in early embryos and temperature, low oxygen environments inhabited by many larvae of the toad, Bufo arenarum Hensel. Bull Environ Australian species.The extreme conditions experienced by Contam Toxicol, 52: 649-655. Australian species need to be considered in the development Antony, A. J. and Ramalingam, K., (1990) Phosalone toxicity on of appropriate toxicity tests using Australian species. the movements and feeding behaviour of Rana tigrina It is worth emphasising here that the sensitivity of an Daudin. Uttar Pradesh J Zool, 10: 180-181. organism is stage dependent. Not only do embryos and Ardizzone, C., Krokos, H. and Lippe, C., (1990) Actions Of adults differ from tadpoles in their relative sensitivities (see Carbaryl on the ionic transport across the isolated skin of previous section), but tadpoles at different stages of Rana esculenta. Comp Biochem Physiol C, 97: 49-52. development will also display differences in sensitivities Arias, E., Pacces Zeffaroni, N. and Zavanella,T., (1989) (Sanders 1970; Cooke 1972; Johnson 1976; Jordan et al. Teratological evaluation of the phenoxyacid herbicide 1977;Wohlgemuth 1977; Mohanty-Hejmadi and Dutta 1981; MCPA in the regenerating forelimb of the adult newt. Rao and Madhyastha 1987; Howe et al. 1998). Ecotoxicol Environ Safety, 17: 30-37. Arias, E. and Zavanella,T., (1979) Teratogenic effects of Animal supply manganese ethylenebisdithiocarbamate (Maneb) on Many commonly used aquatic test species are either cultured forelimb regeneration in the adult newt, Triturus cristatus in the laboratory, or, as in the northern hemisphere, are carnifex. Bull Environ Contam Toxicol, 22: 297-304. supplied by commercial companies.The advantage of such an ASTM, (1993a) ASTM Standards on Aquatic Toxicology and approach is a constant supply of test organisms with a uniform Hazard Evaluation. American Society for Testing and age structure and known history with respect to nutrition and Materials, Philadelphia, PA. disease. In Australia there are no commercial supply houses ASTM, (1993b) Standard guide for conducting the Frog able to provide native or exotic frogs in the large numbers Embryo Teratogenesis Assay-Xenopus (Fetax). Pp. 457- required for toxicology (although the Amphibian Research 467. In ASTM Standards on Aquatic Toxicology and Hazard Centre in Victoria may have that potential in the future) and, Evaluation. American Society for Testing and Materials, consequently, research on amphibian toxicology is largely Philadelphia, PA. dependent on field collected animals.This presents a number ASTM, (1993c) Standard practice for conducting acute tests of problems. Firstly, animal collection in some states and with fishes, macroinvertebrates and amphibians. Pp. 102- territories is illegal or regulated by legislative authorities. 121. In ASTM standards on aquatic toxicology and hazard Second, field collection is time consuming and can be relatively evaluation. American Society for Testing and Materials, expensive.Third, there is always uncertainty about the health of Philadelphia, PA. field collected animals and their field exposure histories. Finally, Baker, J. M. R. and Waights,V., (1993) The effect of sodium it is becoming increasingly difficult to justify the removal of nitrate on the growth and survival of toad tadpoles (Bufo animals from natural populations.While field collection has not bufo) in the laboratory. Herpetological J, 3: 147-148. been singled out as an important factor in amphibian decline, it

193 Baker, J. M. R. and Waights,V., (1994) The effects of nitrate on Birch,W. X. and Prahlad, K.V., (1986b) Effects of Nabam on tadpoles of the tree frog (Litoria caerulea). Herpetological J, developing Xenopus laevis embryos, minimum 4: 106-108. concentration, biological stability and degradative products. Balasundaram, K. and Selvarajan,V. R., (1990) Inhibition of Cytobios, 48: 175-184. acetylcholinesterase in the central nervous system of Rana Birge,W. J., Black, J. A. and Westerman, A. G., (1979) Evaluation tigrina by an organophosphate. J Biochem Toxicol, 5: 65-66. of Aquatic pollutants using fish and amphibian eggs as Bancroft, R. and Prahlad, K.V., (1973) Effect of bioassay organisms. Animals as monitors of environmental ethylenebis[dithiocarbamic acid] disodium salt (Nabam) pollutants.Washington, D.C., Abstracts, Pp. 108-118. and ethylenebis[dithiocarbamate] manganese (Maneb) on Birks, P.R. and Olsen, A. M., (1987) Pesticide concentrations in Xenopus laevis development. Teratology, 7: 143-150. some South Australian birds and other fauna. Trans R Soc Barker, J., Grigg, G. C. and Tyler, M. J., (1995) A field guide to South Aust, 111: 67-78. Australian frogs. Surrey Beatty & Sons, Norton, NSW. Blaustein, A. R.,Wake, D. B. and Sousa,W. P., (1994) Amphibian Barnett, J. B. and Rodgers, K. E., (1994) Pesticides. Pp. 191-212. declines: Judging stability, persistence, and susceptibility of In Immunotoxicology and Immunopharmacology. ed by populations to local and global extinctions. Conserv Biol, J. H. Dean, M. I. Luster, A. E. Munster, and I. Kimber. Raven 8: 60-71. Press, New York. Bradford, D. F., (1991) Mass mortality and extinction in a Baumgarten, D. S., (1991) Salinity tolerance and the effect of high-elevation population of Rana muscosa. J Herpetol, land salinisation on three common species of frog in the 25: 174-177. wheat-belt of Western Australia. BAppSci(Hon), Zoology, Bradford, D. F., Gordon, M. S., Johnson, D. F., Andrews, R. D. University of Western Australia, Perth. and Jennings,W. B., (1994) Acidic deposition as an unlikely Beck, A. B. (1956) The copper content of the liver and blood cause for amphibian population declines in the Sierra of some vertebrates. Aust J Zool, 4: 1-18. Nevada, California. Biol Conserv, 69: 155-161. Berger, L., (1989) Disappearance of amphibian larvae in the Bridges, C. M., (1997) Tadpole swimming performance and agricultural landscape. Ecol Internat Bull, 17: 65-73. activity affected by acute exposure to sublethal levels of carbaryl. Environ Toxicol Chem, 16: 1935-1939. Berger, L., Spear, R. and Hyatt, A.D. (1999) Chytrid fungi and amphibian declines: overview, implications and future Brooks, J. A., (1981) Otolith abnormalities in Limnodynastes directions. Pp 23-33 in Declines and Disappearances of tasmaniensis tadpoles after embryonic exposure to the Australian Frogs ed by A.Campbell. Environment Australia: pesticide dieldrin. Environ pollut Ser A, 25: 19-25. Canberra. Cairns, J. J., Bidwell, J. R. and Arnegard, M., (1996) Toxicty Bernardini, G., Spinelli, O., Presutti, C.,Vismara, C., Bolzacchini, testing with communities: microcosms, mesocosms, and E., Orlandi, M. and Settimi, R., (1996) Evaluation of the whole-system manipulations. Rev Environ Contam Toxicol, developmental toxicity of the pesticide MCPA and its 147: 45-69. contaminants phenol and chlorocresol. Environ Toxicol Celentano, F., Monticelli, G. and Orsenigo, M. N., (1979) Chem, 15: 754-760. Membrane water permeability alterations induced by Berrill, M., Bertram, S., McGillivray, L., Kolohon, M. and Pauli, B., pollutants. J Environ Health C, 13: 301-304. (1994) Effects of low concentrations of forest-use Clark, K. L. and Hall, R. J., (1985) Effects of elevated hydrogen pesticides on frog embryos and tadpoles. Environ Toxicol ion and aluminum concentrations on the survival of Chem, 13: 657-664. amphibian embryos and larvae. Can J Zool, 63: 116-123. Berrill, M., Bertram, S., Pauli, B., Coulson, D., Kolohon, M. and Cockbill, G. F., (1979) The effect of ultra-low-volume aerial Ostrander, D., (1995) Comparative sensitivity of amphibian applications of endosulfan applied against Glossina tadpoles to single and pulsed exposures of the forest-use (Diptera: Glossinidae) on populations of non-target insecticide fenitrothion. Environ Toxicol Chem, 14: 1011-1018. organisms in savanna woodland in Zimbabwe-Rhodesia. Berrill, M., Bertram, S.,Wilson, A., Louis, S., Brigham, D. and Bull Ent Res, 69: 645-655. Stromberg, C., (1993) Lethal and sublethal impacts of Cogger, H. G., (1992) Reptiles and Amphibians of Australia. pyrethroid insecticides on amphibian embryos and A.H & A.W. Reed, Sydney. tadpoles. Environ Toxicol Chem, 12: 525-539. Cooke, A. S., (1970) The effect of pp’-DDT on tadpoles of Berrill, M., Coulson, D., McGillivray, L. and Pauli, B. (1998) the common frog (Rana temporaria). Environ Pollut, Toxicity of endosulfan to aquatic stages of anuran 1: 57-71. amphibians. Environ Toxicol Chem, 17: 1738-1744. Cooke, A. S., (1971) Selective predation by newts on frog Bidwell, J. R. and Gorrie, J. R., (1995) Acute toxicity of a tadpoles treated with DDT. Nature, 229: 275-276. herbicide to selected frog species. Department of Cooke, A. S., (1972) The effects of DDT, dieldrin and 2,4-D Environmental Protection,Technical Series 79. Perth. on amphibian spawn and tadpoles. Environ Pollut, 3: 51-68. Bimber, D. L. and Mitchell, R. A., (1978) Effects of diquat on Cooke, A. S., (1973a) The effects of DDT, when used as a amphibian embryo development. Ohio J Sci, 78: 50-51. mosquito larvicide, on tadpoles of the frog Rana Birch,W. X. and Prahlad, K.V., (1986a) Effects of minute doses temporaria. Environ Pollut, 5: 259-273. of ethylenebisdithiocarbamate disodium salt Nabam and Cooke, A. S., (1973b) Response of Rana temporaria tadpoles its degradative products on connective tissue envelopes of to chronic doses of pp’-DDT. Copeia, 1973: 647-652. the notochord in Xenopus, an ultrastructural study. Arch Cooke, A. S., (1974) The effects of pp1-DDT on adult frogs. Environ Contam Toxicol, 15: 637-646. Brit J Herpe, 5: 390-396.

194 Cooke, A. S., (1977) Effects of field application of the Dumont, J. and Bantle, J., (1997) Methoprene exposed to herbicides diquat and dichlobenil on amphibians. Environ UV light causes developmental toxicity in Xenopus laevis. Pollut, 12: 43-50. SETAC 18th Annual Meeting. San Francisco, CA, Abstracts, Cooke, A. S., (1979) The influence of rearing density on the Pp. 91-92. subsequent response to DDT dosing for tadpoles of the Dumont, J. N., Schultz,T.W., Buchanan, M.V. and Kao, G. L., frog Rana temporaria. Bull Environ Contam Toxicol, (1983) Frog embryo teratogenesis assay: (FETAX)- a short 21: 837-841. term assay applicable to complex environmental mixtures. Cooke, A. S., (1981) Tadpoles as indicators of harmful levels of Pp. 393-405. In Short-term Bioassays in the Analysis of pollution in the field. Environ pollut Ser A, 25: 123-133. Complex Environmental Mixtures. ed by M. D.Waters, S. S. Sandhu, J. Lewtas, L. claxton, N. Chernoff, and S. Nesnow. Cory, L., Fjeld, P.and Serat,W., (1970) Distribution patterns of Plenum Press, New York. DDT residues in the Sierra Nevada Mountains. Pestic Monit J, 3: 204-211. Edery, H. and Schatzberg-Porath, G., (1960) Studies of the effects of organophosphorus insecticides on amphibians. Cummins, C. P., (1986) Effects of aluminium and low pH on Arch Int Pharmacodyn, 124: 212-224. growth and development in Rana temporaria tadpoles. Oecologia, 69: 248-252. Edmisten, G. E. and Bantle, J. A., (1982) Use of Xenopus laevis larvae in 96-hour, flow-through toxicity tests with Davis, K. R., Schultz,T.W. and Dumont, J. N., (1981) Toxic and napthalene. Bull Environ Contam Toxicol, 29: 392-399. teratogenic effects of selected aromatic amines on the Elliot, M., Janes, N. F. and Potter, C., (1978) The future of embryos of the amphibian Xenopus laevis. Arch Environ pyrethroids in insect control. Ann Rev Entomol, 23: 443-469. Contam Toxicol, 10: 371-391. Elliot-Feeley, E. and Armstrong, J. B., (1982) Effects of Dawson, D. A., McCormick, C. A. and Bantle, J. A., (1985) fenitrothion and carbaryl on Xenopus laevis development. Detection of teratogenic substances in acidic mine water Toxicology, 22: 319-335. samples using frog embryo teratogenesis assay — Xenopus (FETAX). J Appl Toxicol, 5: 234-244. Ellis, M. M.,Westfall, B. A. and Ellis, M. D., (1944) Toxicity of dichloro-diphenyl-trichlorethane (DDT) to goldfish and Dawson, D. A., Stebler, E. F., Burks, S. L. and Bantle, J. A., (1988) frogs. Science, 477. Evaluation of the developmental toxicity of metal- contaminated sediments using short-term fathead minnow EPA WA, (1989) Monitoring pesticides-a review.Western and frog Embryo-larval assays. Environ Toxicol Chem, Australian Environmental Protection Authority, Bulletin 7: 27-34. 407. Perth. de Wijer, P.,Watt, P.J. and Oldham, R. S., (1997) Amphibian Fellers, G. M., (1997) Status of declining amphibians in decline and aquatic pollution: effects of nitrogenous California, USA.Third World Congress of Herpetology. fertiliser.Third World Congress of Herpetology. Prague, Prague, Czech Republic, Abstracts, p. 67. Czech Republic, Abstracts, p. 53. Ferraro,T. J., (1992) Factors influencing the decline of frogs Dekins, M. S., Guy, H. R. and Edwards, C., (1978) Effects of (Amphibia: Anura) in western Sydney. Unpublished organophosphate insecticides on the cholinergic receptors B.App.Sci. (Hons), University of Western Sydney, of frog skeletal muscle. J Pharmacol Exp Ther, 205: 319-325. Hawkesbury, Richmond. Department of Agriculture, (1994) Pesticides Registered in Ferraro,T. J. and Burgin, S., (1993) Review of environmental Western Australia. Department of Agriculture,Western factors influencing the decline of Australian frogs. Pp. 205- Australia, Bulletin 4275. Perth. 218. In Herpetology in Australia: a diverse discipline. ed by Deshmukh, P.B. and Keshhavan, R., (1987) Activities of D. Lunney, and D. Ayers. Royal Zoological Society of NSW, amylase and lipase in the frog exposed to DDT and Sevin. Mosman, NSW. J Adv Zool, 8: 118-122. Ferreira, H. G. and Jesus, C. H., (1973) Salt adaptation in Bufo Deyoung, D. J., Bantle, J. A., Hull, M. A. and Burks, S. L., (1996) bufo. J Physiol, 228: 583-600. Differences in sensitivity to developmental toxicants as Ferreira, K.T. G. and Hill, B. S., (1982) The effect of low external seen in Xenopus and Pimephales embryos. Bull Environ pH on properties of the paracellular pathway and the Contam Toxicol, 56: 143-150. junctional structure in isolated frog skin. J Physiol, 332: 59-67. Dial, C. A. B. and Dial, N. A., (1995) Lethal effects of the Fioramonti, E., Semlitsch, R. D., Reyer, H.-U. and Fent, K., consumption of field levels of Paraquat-contaminated (1997) Effects of triphenyltin and pH on the growth and plants on frog tadpoles. Bull Environ Contam Toxicol, development of Rana lessonae and Rana esculenta 55: 870-877. tadpoles. Environ Toxicol Chem, 16: 1940-1947. Dial, N., (1975) Methylmercury:Teratogenic and lethal effects Fleming,W. J., de Chacin, H., Pattee, O. H. and Lamont,T. G., in frog embryos. Teratology, 13: 327-334. (1982) Parathion accumulation in cricket frogs and its Dial, N. A. and Bauer, C. A., (1984) Teratogenic and lethal effect on American kestrels. J Toxicol Environ Health, effects of paraquat on developing frog embryos (Rana 10: 921-927. pipiens). Bull Environ Contam Toxicol, 33: 592-597. Flickinger, E. L., King, K. A., Stout,W. F. and Mohn, M. M., (1980) Dial, N. A. and Bauer-Dial, C. A., (1987) Lethal effects of Wildlife hazards from Furadan 3G application to rice in diquat and paraquat on developing frog embryos and Texas. J Wildl Manage, 44: 190-197. 15-day-old tadpoles, Rana pipiens. Bull Environ Contam Freda, J., (1986) The influence of acidic pond water on Toxicol, 38: 1006-1011. amphibians: A review. Water Air and Soil Pollut, 30: 439-450.

195 Freda, J. and Dunson,W. A., (1985) The influence of external Hashimoto,Y. and Nishiuchi,Y., (1981) Establishment of cation concentration on the hatching of amphibian bioassay methods for the evaluation of acute toxicity of embryos in water of low pH. Can J Zool, 63: 2649-2656. pesticides to aquatic organisms. Jap J Pesticide Sci, Fromm, P.O., (1981) Effect of acid stress on sodium transport 6: 257-264. by isolated skins and on osmotic permeability of intact Hauptman, O., Albert, D. M., Plowman, M. C., Hopfer, S. M. and frogs. Bull Environ Contam Toxicol, 27: 160-166. Sunderman, F.W., Jr., (1993) Ocular malformations of Fulton, M. H. and Chambers, J. E., (1985) The toxic and Xenopus laevis exposed to nickel during embryogenesis. teratogenic effects of selected organophosphorus Ann Clin & Lab Sci, 23: 397-406. compounds on the embryos of three species of Hayes,T. B. and Noriega, N., (1997) Endocrine disrupting amphibians. Toxicol Lett, 26: 175-180. environmental contaminants: Mechanisms of transfer and Gavilan, E. J. F., Hermosilla, B. I., Alay, H. F. and Venegas, S.W., potential population level effects.Third World Congress of (1988) Teratogenic effect of DDT on the development of Herpetology. Prague, Czech Republic, Abstracts, p. 94. Caudiverbera caudiverbera Linne 1758 Anura Hayes,T. B.,Wu,T. H. and Gill,T. N., (1997) DDT-like effects as Leptodactylidae. Boletin de la Sociedad de Biologia de a result of corticosterone treatment in an anuran Concepcion, 59: 47-56. amphibian: is DDT a corticoid mimic or a stressor. Environ Glennemeier, K., (1997) Effects of endocrine-disrupting Toxicol Chem, 16: 1948-1953. pollutants on the development and behavior in the Hazelwood, E., (1970) Frog pond contaminated. Brit J Herpe, northern leopard frog, Rana pipiens.Third World Congress 4: 177-184. of Herpetology. Prague, Czech Republic, Abstracts, Pp. 81. Hecnar, S. and McLoskey, R., (1996) Amphibian species Gopal, K., Khanna, R. N., M, A. and Gupta, G. S. D., (1981) richness and distribution in relation to pond water The acute toxicity of endosulfan to freshwater organisms. chemistry in south-western Ontario, Canada. Freshwater Toxicol Lett, 7: 453-456. Biol, 36: 7-15. Gordon, M. S., Schmidt-Nielsen, K. and Kelly, H. M., (1961) Hecnar, S. J., (1995) Acute and chronic toxicity of ammonium Osmotic regulation in the crab-eating frog (Rana nitrate fertilizer to amphibians from southern Ontario. cancrivora). J Exp Biol, 38: 659-678. Environ Toxicol Chem, 14: 2131-2137. Greenhouse, G., (1976) The evaluation of toxic effects of Herald, E. S., (1949) Effect of DDT-oil solutions upon chemicals in freshwater by using frog embryos and larvae. amphibians and reptiles. Herpetologica, 5: 117-120. Environ Pollut, 11: 303-315. Herkovits, J., Cardellini, P., Pavanati, C. and Perez Coll, C. S., Griffiths, R. A., (1993) The effect of pH on feeding behavior in (1997) Susceptibility of early life stages of Xenopus laevis newt larvae (Triturus: Amphibia). Journal of Zoology, to cadmium. Environmental Toxicology & Chemistry, 231: 285-290. 16: 312-316. Griffiths, R. A. and De Wijer, P.,(1994) Differential effects of Herkovits, J., Herkovits, F. D. and Perez-Coll, C. S., (1995) pH and temperature on embryonic development in the Comparison of the susceptibility of amphibian (Bufo British newts (Triturus). J Zool (Lond), 234: 613-622. arenarum) embryos and fish (Cnesterodon decemmaculatus) Guzman, J. A. and Guardia,T., (1978) Effects of an to aluminium. SETAC 16th Annual Meeting.Vancouver, BC, organophosphorous insecticide on the cholinesteratic Abstracts, p. 142. activities of Bufo arenarum (H). Bull Environ Contam Toxicol, Herkovits, J. and Jatimliansky, J., (1986) Efectos del cadmio 20: 52-58. sobre el consumo de oxígeno en embriones de Bufo Hall, R. J., (1990) Accumulation, metabolism and toxicity of arenarum. Medicina, 46: 603. parathion in tadpoles. Bull Environ Contam Toxicol, Holcombe, G.W., Phipps, G. L., Sulaiman, A. H. and Hoffman, 44: 629-635. A. D., (1987) Simultaneous multiple species testing acute Hall, R. J. and Kolbe, E., (1980) Bioconcentration of toxicity of 13 chemicals to 12 diverse freshwater organophosporus pesticides to hazardous levels by amphibian, fish, and invertebrate families. Arch Environ amphibians. J Toxicol Environ Health, 6: 853-860. Contam Toxicol, 16: 697-710. Hall, R. J. and Swineford, D., (1979) Uptake of Methoxychlor Honrubia, M. P., Herraez, M. P.and Alvarez, R., (1993) The from food and water by the American Toad (Bufo carbamate insecticide ZZ-Aphox induced structural americanus). Bull Environ Contam Toxicol, 23: 335-337. changes of gills, liver, gallbladder, heart, and notochord of Rana perezi tadpoles. Arch Environ Contam Toxicol, Hall, R. J. and Swineford, D., (1980) Toxic effects of endrin and 25: 184-191. toxaphene on the southern leopard frog Rana sphenocephala. Environ pollut Ser A, 23: 53-65. Hopfer, S. M., Plowman, M. C., Sweeney, K. R., Bantle, J. A. and Sunderman, F.W., Jr., (1991) Teratogenicity of nickel in Hall, R. J. and Swineford, D., (1981) Acute toxicity of Xenopus laevis assayed by the Fetax procedure. Biol Trace toxaphene and endrin to larvae of seven species of Elem Res, 29: 203-216. amphibians. Toxicol Lett, 8: 331-336. Hoppe, D. and Mottl, E., (1997) Anuran species differences in Haniffa, M. A. and Augustin, S. M., (1989) Oxygen malformation frequency and severity in Minnesota. SETAC consumption, surfacing frequency and distance of travel 18th Annual Meeting. San Francisco, CA, Abstracts, p. 90. in Rana malabarica tadpoles exposed to distillery effluent. J Environ Biol, 10: 139-148. Horne, M.T. and Dunson,W. A., (1995a) Effects of low pH, metals, and water hardness on larval amphibians. Arch Environ Contam Toxicol, 29: 500-505.

196 Horne, M.T. and Dunson,W. A., (1995b) The interactive Katz, U., (1973) Studies on the adaptation of the toad Bufo effects of low pH, toxic metals, and DOC on a simulated viridis to high salinities: oxygen consumption, plasma temporary pond community. Environ Pollut, 89: 155-161. concentrations and water content of the tissues. J Exp Biol, Howe, G. E., Gillis, R. and Mowbray, R. C., (1998) Effect of 50: 785-796. chemical synergy and larval stage on the toxicity of Kimball, K. D. and Levin, S. A., (1985) Limitations of laboratory atrazine and alachlor to amphibian larvae. Environ Toxicol bioassays: the need for ecosystem-level testing. BioSci, Chem, 17: 519-525. 35: 165-171. Isaacson, P.A., (1968) Action of DDT on the peripheral Kutka, F. J., (1994) Low pH effects of swimming activity of nervous system of the grass frog Rana pipiens. Curr Sci, Ambystoma salamander larvae. Environ Toxicol Chem, 37: 530-531. 13: 1821-1824. Jobling, S., Sheahan, D., Osborne, J. A., Matthiessen, P.and Lajmanovich, R. C., Izaguirre, M. F. and Casco,V. H. (1998) Sumpter, J. P.,(1996) Inhibition of testicular growth in Paraquat tolerance and alteration of internal gill structure rainbow trout (Oncorhynchus mykiss) exposed to of Scinax nasica tadpoles (Anura: Hylidae). Arch Environ estrogenic alkylphenolic chemicals. Environ Toxicol Chem, Contam Toxicol, 34: 364-369. 15: 194-202. Lambert, M. R. K., (1993) Effects of DDT ground-spraying Johnson, C. R., (1976) Herbicide toxicities in some Australian against tsetse flies on lizards in NW Zimbabwe. Environ anurans and the effect of subacute dosages on Pollut, 82: 231-237. temperature tolerance. Zool J Linn Soc, 59: 79-83. Lambert, M. R. K., (1997) Environmental effects of heavy Johnson, C. R. and Prine, J. E., (1976) The effects of sublethal spillage from a destroyed pesticide store near Hargeisa concentrations of organophosphorus insecticides and an (Somaliland) assessed during the dry season, using reptiles insect growth regulator on temperature tolerance in and amphibians as bioindicators. Arch Environ Contam hydrated and dehydrated juvenile Western Toads, Bufo Toxicol, 32: 80-93. boreas. Comp Biochem Physiol A, 53A: 147-149. Lamont, B. B., (1994) Mineral nutrient relations in Jolly, A. L., Jr, Avault, J.W., Jr, Koonce, K. L. and Graves, J. B., mediterranean regions of California, Chile and Australia. (1978) Acute toxicity of Permethrin to several aquatic Pp. 211-235. In Ecology and Biogeography of animals. Trans Am Fish Soc, 107: 825-827. Mediterranean Ecosystems in Chile, California and Jordan, M., Rzehak, K. and Maryanska, A., (1977) The effect of Australia. ed by M.T. Kalin Arroyo, P. H. Zedler, and M. D. two pesticides, Miedzian 50 and Gesagard 50, on the Fox. Springer-Verlag, New York. development of tadpoles of Rana temporaria. Bull Environ Landis,W. G. and Yu, M-H., (1995) Introduction to Contam Toxicol, 17: 349-354. Environmental Toxicology. Impacts upon ecological systems. Joseph, K.V. and Rao, K. J., (1990) Aldrin toxicity on amphibian Lewis, Boca Raton. neuronal hepatic and muscular tissue oxidative enzymes. Langford, R. R., (1949) The effect of DDT on the natural Biochem Internat, 22: 173-178. fauna of the forest exclusive of insect pests. Pp. 13-18. In Joseph, K.V. and Rao, K. J., (1991) Protein degradation and Forest Spraying and Some Effects of DDT. Ontario Dept related enzyme profiles on a sublethal toxicity of aldrin in of Land and Forests. the tissues of Rana hexadactyla. Ecotoxicol Environ Safety, Lefcort, H., Hancock, K. A., Maur, K. M. and Rostal, D. C. (1997) 22: 32-35. The effects of used motor oil, silt and the water mold Juarez, A. and Guzman, J. A., (1986) Beta-glucuronidase activity Saprolegnia parasitica on the growth and survival of mole and thyroid hormones of Bufo arenarum (H) tadpole salamanders (Genus Ambystoma). Arch Environ Contam during anuran metamorphosis induced by DDT. Comun Toxicol, 32: 383-388. Biol, 4: 251-257. Leone,V. G., Bordone, L., Cristiani, S., Sporchia, L. and Vismara, Jung, R. E. and Jagoe, C. H., (1995) Effects of low pH and C., (1994) The FETAX (Frog Embryo Teratogenesis Assay- aluminum on body size, swimming performance, and Xenopus) test: Evaluation of teratogenic effects induced by susceptibility to predation of green tree frog (Hyla cinerea) xenobiotics on the embryogenesis of Xenopus laevis: tadpoles. Can J Zool, 73: 2171-2183. Modification and new applications. Ist Lomb Accad Sci Lett Rend Biol Med B, 128: 3-12. Jung, R. E. and Walker, M. K., (1997) Effects of 2,3,7,8- tetrachlorobenzo-p-dioxin (TCDD) on development of Licht, L. E., (1976) Postmetamorphic retention of C14 by anuran amphibians. Environmental Toxicology & Chemistry, wood frog tadpoles. Comp Biochem Physiol C, 55: 119-121. 16: 230-240. Liggins, G.W. and Grigg, G., (1985) Osmoregulation of the Kaplan, H. M., Arnholt,T. J. and Payne, J. E., (1967) Toxicity of cane toad, Bufo marinus, in salt water. Comp Biochem lead nitrate solutions for frogs (Rana pipiens). Lab Animal Physiol A, 82: 613-619. Care, 17: 240-246. Linder, G., Barbitta, J. and Kwaiser,T., (1990) Short-term Kaplan, H. M. and Glaczenski, S. S., (1965) Hematological amphibian toxicity tests and paraquat toxicity assessment. effects of organophosphate insecticides in the frog (Rana Pp. 189-198. In Aquatic Toxicology and Risk Assessment. ed pipiens). Life Sci, 4: 1213-1219. by W. G. Landis, and W. H. van der Schalie. American Society for Testing and Materials, Philadelphia, PA. Kaplan, H. M. and Yoh, L., (1961) Toxicity of copper for frogs. Herpetologica, 17: 131-135. Lindquist, N. G., Larsson, B. S. and Lyden Sokolowski, A., (1988) Autoradiography of carbon-14 paraquat or carbon- 14 diquat in frogs and mice accumulation in neuromelanin. Neurosci Lett, 93: 1-6.

197 Lippe, C., Bellantuono,V.,Valenti, G. and Ardizzone, C., (1992) Morgan, M. K., Scheuerman, P.R., Bishop, C. S. and Pyles, R. A., The nature of chloride secretion induced by carbaryl (1996) Teratogenic potential of atrazine and 2,4-D using across the isolated skin of Rana esculenta. Comp Biochem fetax. J Toxicol Environ Health, 48: 151-168. Physiol C, 103: 387-389. Morin, P., (1981) Predatory salamanders reverse the outcome Llamas, M. C. de, de Castro, A. C. and Pechen de Dangelo, A. of competition among three species of anuran tadpoles. M., (1985) Cholinesterase activities in developing Science, 212: 1284-1286. amphibian embryos following exposure to the insecticides Mudgall, C. F. and Patil, H. S., (1987) Toxic effects of dieldrin and malathion. Arch Environ Contam Toxicol, dimethoate and methyl parathion on glycogen reserves 14: 161-166. on male and female Rana cyanophlyctis. J Environ Biol, Logier, E. B. S., (1949) Effect of DDT on amphibians and 8: 237-244. reptiles. Pp. 49-52. In Forest Spraying and Some Effects of Mulla, M. S., (1962) Frog and toad control with insecticides! DDT. Ontario Dept of Land and Forests. Pest Control, 30: 64. Luo, S.W., Plowman, M. C., Hopfer, S. M. and Sunderman, F.W., Mulla, M. S., Isaak, L.W. and Axelrod, H., (1963) Field studies Jr., (1993) Embryotoxicity and teratogenicity of copper and on the effects of insecticides on some aquatic wildlife zinc for Xenopus laevis assayed by the Fetax procedure. species. J Econ Ent, 59: 1085-1090. Ann Clin & Lab Sci, 23: 111-120. Nagel, R. and Urich, K., (1981) Elimination and distribution of Mahaney, P.A., (1994) Effects of freshwater petroleum different substituted phenols by frog (Rana temporaria) contamination on amphibian hatching and metamorphosis. and crayfish (Astacus leptodactylus). Bull Environ Contam Environ Toxicol Chem, 13: 259-265. Toxicol, 26: 289-294. Mann, R. M. and Bidwell, J. R., (1998) The toxicity of glyphosate Nakatsuji, N., (1983) Craniofacial malformation in Xenopus and several glyphosate formulations to four species of laevis tadpoles caused by the exposure of early embryos southwestern Australian frogs. Arch Environ Contam Toxicol, to ethanol. Teratology, 28: 299-305. In Press. Natochin,Y.V. and Jones, M. M., (1992) Chelating agent Manson, J. M. and O’Flaherty, E. J., (1978) Effects of cadmium reversal of cadmium effects on ionic transport in the on salamander survival and limb regeneration. Environ Res, isolated frog skin Rana temporaria. Comp Biochem Physiol C, 16: 62-69. 101: 531-535. Marchal-Ségault, D. and Remande, F., (1981) The effects of Nebeker, A.V., Schuytema, G. S. and Ott, S. L., (1994) Effects of lindane, an insecticide, on hatching and postembryonic cadmium on limb regeneration in the northwestern development of Xenopus laevis (Daudin) anuran salamander Ambystoma gracile. Arch Environ Contam Toxicol, amphibian. Environ Res, 24: 250-258. 27: 318-322. Marian, M. P., Arul,V. and Pandian,T. J., (1983) Acute and Niethammer, K. R.,White, D. H., Basket,T. S. and Sayre, M.W., chronic effects of carbaryl on survival, growth and (1984) Presence and biomagnification of organochlorine metamorphosis in the bullfrog (Rana tigrina). Arch Environ chemical residues in oxbow lakes of northeastern Contam Toxicol, 12: 271-275. louisiana. Arch Environ Contam Toxicol, 13: 63-74. Materna, E. J., Rabeni, C. F. and La Point,T.W., (1995) Effects Nieuwkoop, P.D. and Faber, J., (1975) Normal Tables of of the synthetic pyrethroid insecticide, esfenvalerate, on Xenopus laevis (Daudin). North Holland, Amsterdam. larval leopard frogs (Rana spp.). Environ Toxicol Chem, OECD, (1993) OECD guidelines for testing of chemicals. 14: 613-622. Organisation for Economic Co-operation and McCrary, J. E. and Heagler, M. G., (1997) The use of a Development, Paris. simultaneous multiple species acute toxicity test to Oldham, R. S., Latham, D. M., Hilton Brown, D.,Towns, M., compare the relative sensitivities of aquatic organisms to Cooke, A. S. and Burn, A., (1997) The effect of ammonium mercury. J Environ Science & Health A, 32: 73-81. nitrate fertiliser on frog (Rana temporaria) survival. Agric McIlroy, J. C., King, D. R. and Oliver, A. J., (1985) The sensitivity Ecosyst Environ, 61: 69-74. of Australian animals to 1080 fluoroacetic-acid sodium salt Osborn, D., Cooke, A. S. and Freestone, S., (1981) Histology poison VIII. amphibians and Reptiles. Aust Wildl Res, of a teratogenic effect of DDT on Rana temporaria 12: 113-118. tadpoles. Environ pollut Ser A, 25: 305-319. Meeks, R. L., (1968) The accumulation of 36Cl ring-labeled Ouellet, M., Bonin, J., Rodrigue, J., Desgranges, J. L. and Lair, S., DDT in a freshwater marsh. J Wildl Manage, 32: 376-398. (1997) Hindlimb deformities (Ectromelia, Ectrodactyly) in Mendiola, P.and De Costa, J., (1991) The effects of free-living anurans from agricultural habitats. Journal of temperature and pH on the kinetic properties of heart Wildlife Diseases, 33: 95-104. muscle lactate dehydrogenase from anuran amphibians. Palmer, B. D., Massie, K. and Zepeda, F., (1996) Vitellogenin as Comp Biochem Physiol B, 98: 529-534. a biomarker for xenobiotic estrogens in amphibian and Millen, S. A., (1995) Sublethal effects of cypermethrin on reptillian model systems. SETAC 17th Annual Meeting. Litoria ewingi tadpoles. BSc (Hons), Department of Washington, DC, Abstracts, p. 1. Zoology, University of Tasmania, Hobart. Palmer, B. D. and Palmer, S. K., (1995) Vitellogenin induction by Mohanty-Hejmadi, P.and Dutta, S. K., (1981) Effects of some xenobiotic estrogens in the red-eared turtle and African pesticides on the development of the indian bull frog Rana clawed frog. Environ Health Perspect, 103: 19-25. tigerina. Environ pollut Ser A, 24: 145-161. Panter, H. C. (1986) Variations in radiosensitivity during development of the frog Limnodynastes tasmaniensis. J Exp Zool, 238: 193-200.

198 Panter, H. C., Chapman, J. E. and Williams, A. R. (1987) Effect Pritchard-Landé, S. and Guttman, S., (1973) The effects of of radiation and trophic state on oxygen consumption of copper sulfate on the growth and mortality rate of Rana tadpoles of the frog Limnodynastes tasmaniensis. Comp pipiens tadpoles. Herpetologica, 29: 22-27. Biochem Physiol A, 88: 373-376. Quincy, L. M., (1991) The effect of high salinities on growth Paulov, S., (1977a) The effects of N-Phenyl-N’-Methyl Urea and development of the spawn and larvae of the spotted (defenuron) on the development of amphibia (Rana grass frog — Limnodynastes tasmaniensis. BSc (Hons) temporaria L). Biologia (Bratislava), 32: 607-611. Department of Zoology, University of Adelaide, Adelaide. Paulov, S., (1977b) Ucinky herbicidu Gramoxone S (paraquat Raloff, J., (1994) The gender benders: Are environmental dichlorid) na vyvoj a bielkoviny obojzivelnikov (Rana “hormones” emasculating wildlife? Sci News, 145: 24-27. temporaria L.). Biologia (Bratislava), 32: 127-131. Ramsdell, H. S., Blandin, D. A. and Schmechel,T. R., (1996) Paulov, S., (1981) Hazards of the fungicide Afugan Developmental effects and biochemical markers of (pyrasophos) for the development of amphibia (Rana alkylphenol exposure in frog larvae. SETAC 17th Annual temporaria L.). Biologia (Bratislava), 36: 135-140. Meeting.Washington, DC, Abstracts, Pp. 140. Pawar, K. R., Ghate, H.V. and Katdare, M., (1983) Effect of Rao, I. J. and Madhyastha, M. N., (1987) Toxicities of some malathion on embryonic development of the frog heavy metals to the tadpoles of frog Microhyla ornata Microhyla ornata (Dumeril and Bibron). Bull Environ Contam Dumeril and Bibron. Toxicol Lett, 36: 205-208. Toxicol, 31: 170-176. Read, J., (1997) Comparative abnormality rates of the trilling Pawar, K. R. and Katdare, M., (1984) Toxic and teratogenic frog at Olympic Dam. Herpetofauna, 27: 23-27. effects of fenitrothion, BHC and carbofuran on embryonic Read, L. J. and Tyler, M. J., (1990) The nature and incidence of develpment of the frog Microhyla ornata.Toxicol Lett, post-axial, skeletal abnormalities in the frog Neobatrachus 22: 7-13. centralis (Parker) at Olympic Dam, South Australia.Trans Pérez Coll, C. S. and Herkovits, J., (1990) Stage dependent R Soc South Aust, 114: 213-217. susceptibility to lead in Bufo arenarum embryos. Environ Read, J. L. and Tyler, M. J., (1994) Natural levels of Pollut, 63: 239-246. abnormalities in the trilling frog (Neobatrachus centralis) Pérez Coll, C. S., Herkovits, J. and Salibian, A., (1986) at the Olympic Dam mine. Bull Environ Contam Toxicol, Teratogenic effects of cadmium on Bufo arenarum during 53: 25-31. gastrulation. Experientia, 42: 1174-1176. Renner, R., (1997) European bans on surfactant trigger Pfeiffer, C. J., Nagai,T., Fujimura, M. and Tobe,T., (1985) transatlantic debate. Environ Sci Technol, 31: 316A-320A. Teratogenic effects of carcinogenic agents on limb Richards, S. J., McDonald, K. R. and Alford, R. A., (1993) regeneration in the Japanese newt Cynops pyrrhogaster. Declines in populations of Australia’s endemic tropical Teratog Carcinog Mutagen, 5: 137-148. rainforest frogs. Pacific Conservation Biology, 1: 66-77. Pierce, B. A., (1985) Acid tolerance in amphibians. BioSci, Riley, E. E. and Weil, M. R., (1987) The effects of 35: 239-243. thiosemicarbazide on development in the wood frog Rana Plowman, M. C., Grbac-Ivankovic, S., Martin, J., Hopfer, S. M. sylvatica II: critical exposure length and age sensitivity. and Sunderman, F.W., Jr., (1994) Malformations persist Ecotoxicol Environ Safety, 13: 202-207. after metamorphosis of Xenopus laevis tadpoles exposed Romspert, A. P., (1976) Osmoregulation of the African clawed to Ni-2+, Co-2+, or Cd-2+ in FETAX assays. Teratog frog, Xenopus laevis, in hypersaline media. Comp Biochem Carcinog Mutagen, 14: 135-144. Physiol A, 54: 207-210. Plowman, M. C., Peracha, H., Hopfer, S. M. and Sunderman, Rosato, P.and Ferguson, D. E., (1968) The toxicity of endrin- F.W., Jr., (1991) Teratogenicity of cobalt chloride in resistant mosquito fish to eleven species of vertebrates. Xenopus laevis assayed by the fetax procedure. Teratog BioSci, 18: 783-784. Carcinog Mutagen, 11: 83-92. Roudebush, R. E., (1988) A behavioral assay for acid sensitivity Powell, G.V. N., DeWeese, L. R. and Lamont,T. G. (1982) A in two desmognathine species of salamanders. field evaluation of frogs as a potential source of secondary Herpetologica, 44: 392-395. organophosporus insecticide poisoning. Can J Zool, Rowe, C. L. and Dunson,W. A., (1994) The value of simulated 60: 2233-2235. pond communities in mesocosms for studies of amphibian Power,T., Clark, K. L., Hernfenist, A. and Peakall, D. B., (1989) A ecology and ecotoxicology. J Herpetol, 28: 346-356. review and evaluation of the amphibian toxicological Rowe, C. L., Kinney, O. M. and Congdon, J. D., (1998) Oral literature. Canadian Wildlife Service, Headquarters, deformities in tadpoles of the bullfrog (Rana catesbeiana) Technical Report 61. Ottawa, Canada. caused by conditions in a polluted habitat. Copeia, 1998: Prahlad, K.V., Bancroft, R. and Hanzely, L., (1974) 244-246. Ultrastructural changes induced by the fungicide Rowe, C. L., Sadinski,W. J. and Dunson,W. A., (1992) Effects of ethylenebis [dithiocarbamic acid] disodium salt (Nabam) acute and chronic acidification on three larval amphibians in Xenopus laevis tissue during development. Cytobios, that breed in temporary ponds. Arch Environ Contam 9: 121-130. Toxicol, 23: 339-350. Presutti, C.,Vismara, C., Camatini, M. and Bernardini, G., Ruigt, G. S. F. and van den Bercken, J., (1986) Action of (1994) Ecotoxicological effects of a nonionic detergent pyrethroids on a nerve muscle preparation of the clawed (Triton DF-16) assayed by ModFETAX. Bull Environ Contam frog, Xenopus laevis. Pestic Biochem Physiol, 25: 176-187. Toxicol, 53: 405-411.

199 Russell, R.W., Hecnar, S. J. and Haffner, G. D., (1995) Snawder, J. E. and Chambers, J. E., (1993) Osteolathyrogenic Organochlorine pesticide residues in southern Ontario effects of malathion in Xenopus embryos. Toxicol Appl spring peepers. Environ Toxicol Chem, 14: 815-817. Pharmacol, 121: 210-216. Rzehak, K., Maryanska-Nadachowska, A. and Jordan, M., (1977) Sparling, D.W., Lowe,T. P.and Pinkney, A. E., (1997) Toxicity of The effect of Karbatox 75, a carbaryl insecticide, upon the Abate to green frog tadpoles. Bull Environ Contam Toxicol, development of tadpoles of Rana temporaria and Xenopus 58: 475-481. laevis. Folia Biol (Kraków), 25: 391-399. Speirs, J. M., (1949) The relation of DDT spraying to the Saber, P.A. and Dunson,W. A., (1978) Toxicity of bog water vertebrate life of the forest. Pp. 141-159. In Forest to embryonic and larval anuran amphibians. J Exp Zool, Spraying and Some Effects of DDT. Ontario Dept of Land 204: 33-42. and Forests. Sadinski,W. J. and Dunson,W. A., (1992) A multilevel study of State of the Environment Advisory Council, (1996) Australia, effects of low pH on amphibians of temporary ponds. State of the Environment 1996. CSIRO Publishing, J Herpetol, 26: 413-422. Collingwood,Vic. Salibian, A., (1983) Studies on the effect of the pyrethroid Steele, C.W., Strickler Shaw, S. and Taylor, D. H., (1989) insecticide decamethrin on ionic transport through the in Behavior of tadpoles of the bullfrog Rana catesbeiana in vitro skin of Rana esculenta. Comp Biochem Physiol C, response to sublethal lead exposure. Aquat Toxicol, 76: 157-162. 14: 331-344. Sanders, H. O., (1970) Pesticide toxicities to tadpoles of the Sunderman, F.W., Jr., Plowman, M. C., Kroftova, O. S., Grbac western chorus frog Pseudacris triseriata and Fowler’s toad Ivankovic, S., Foglia, L. and Crivello, J. F., (1995) Effects of Bufo woodhousii fowleri. Copeia, 1970: 246-251. teratogenic exposures to Zn-2+, Cd-2+, Ni-2+, Co-2+, Scadding, S. R., (1996) Treatment of axolotls with retinoids for and Cu-2+ on metallothionein and metallothionein- limb regeneration studies. Int J Dev Biol, 40: 909-910. mRNA contents of Xenopus embryos. Pharmacol Toxicol, 76: 178-184. Schmidt, C.W., (1997) Amphibian deformities continue to puzzle researchers. Environ Sci Technol, 31: 324A-326A. Sunderman, F.W. J., Plowman, M. C. and Hopfer, S. M., (1991) Embryotoxicity and teratogenicity of cadmium chloride in Schrock, E., Bantle, J. and Santelli, D., (1997) Investigation of Xenopus laevis, assayed by the fetax procedure. Ann Clin & the developmental toxicity of hydrazine and its Lab Sci, 21: 381-391. metabolites. SETAC 18th Annual Meeting. San Francisco, Suzuki, K.T., Itoh, N., Ohta, K. and Sunaga, H. (1986) CA, Abstracts, p. 176. Amphibian Metallothionein Induction In the Frogs Rana- Schultz,T.W., Dumont, J. N. and Buchanan, M.V., (1983) Toxic Japonica Rana-Nigromaculata and Rhacophorus-Schlegelii. and teratogenic effects of chemical class fractions of a Comp Biochem Physiol C, 83: 253-260. coal-gasification electrostatic precipitator tar. Toxicology, Swann, J. M., Schultz,T.W. and Kennedy, J. R., (1996) The effects 29: 87-99. of the organophosphorous insecticides Dursban and Schuytema, G. S., Nebeker, A.V. and Griffis,W. L., (1995) Lorsban on the ciliated epithelium of the frog palate in Comparative toxicity of Guthion and Guthion 2S to vitro. Arch Environ Contam Toxicol, 30: 188-194. Xenopus laevis and Pseudacris regilla tadpoles. Bull Environ Tarzwell, C. M., (1950) Effects of DDT mosquito larviciding on Contam Toxicol, 54: 382-388. wildlife V. Effects on fishes of the routine manual and Semlitsch, R. D., Foglia, M., Mueller, A., Steiner, I., Fioramonti, E. airplane application of DDT and other mosquito larvicides. and Fent, K., (1995) Short-term exposure to triphenyltin Public Health Rep, 65: 231-255. affects the swimming and feeding behavior of tadpoles. Thurston, R.V., Gilfoil,T.A., Meyn, E. L., Zajdel, R. K., Aoki,T. I. Environ Toxicol Chem, 14: 1419-1423. and Veith, G. D., (1985) Comparative toxicity of 10 organic Seugé, J., Marchal-Ségault, D. and Bluzat, R., (1983) Toxicité chemicals to 10 common aquatic species. Water Res, aigue d’un fongicide dithiocarbamate, le thirame, vis-a-vis 19: 1145-1156. de plusiers especes animales d’eau douce. Environ pollut Thybaud, E., (1990) Acute toxicity and bioconcentration of Ser A, 31: 177-189. lindane and deltamethrin in tadpoles of Rana temporaria Shah, P.V., Monroe, R. J. and Guthrie, F. E., (1983) Comparative and the mosquitofish Gambusia affinis. Hydrobiologia, penetration of insecticides in target and non-target 190: 137-146. species. Drug Chem Toxicol, 6: 155-179. Tucker, R. K. and Crabtree, D. G., (1969) Toxicity of Zectran Smith,T. M. and Stratton, G.W., (1986) Effect of synthetic insecticide to several wildlife species. J Econ Ent, pyrethroid insecticides on non-target organisms. Residue 62: 1307-1310. Rev, 97: 93-120. Tyler Jones, R., Beattil, R. C. and Aston, R. J., (1989) The effects Snawder, J. E. and Chambers, J. E., (1989) Toxic and of acid water and aluminum on the embryonic developmental effects of organophosphorus insecticides in development of the common frog Rana temporaria. J Zool embryos of the South . J Environ Health (Lond), 219: 355-372. B, 24: 205-218. Tyler, M. J., (1972) An analysis of the lower vertebrate faunal Snawder, J. E. and Chambers, J. E., (1990) Critical time periods relationships of Australia and New Guinea. Pp 231-256. In and the effect of tryptophan in malathion-induced Bridge and Barrier:The Natural and Cultural History of development defects in Xenopus embryos. Life Sci, Torres Strait. ed by D.Walker. Publication BG3. 46: 1635-1642. Department of Biogeography, Australia National University, Canberra.

200 Tyler, M. J., (1994) Australian Frogs: a natural history. Reed Warner, S. C.,Travis, J. and Dunson,W. A., (1993) Effect of pH Books, Chatswood. variation on interspecific competition between two Tyler, M. J., (1997) The Action Plan For Australian Frogs. species of hylid tadpoles. Ecology, 74: 183-194. Wildlife Australia, a branch of Environment Australia, Watt, P.J. and Oldham, R. S., (1995) The effect of ammonium Commonwealth Department of Environment, Sport and nitrate on the feeding and development of larvae of the Territories, Endangered Species Program, Action Plan, smooth newt, Triturus vulgaris (L.), and on the behaviour of Canberra. its food source, Daphnia. Freshwater Biol, 33: 319-324. U.S.EPA, (1991) Methods for measuring acute toxicity of Webb, G. D., Sharp, R.W. and Feldman, J. D., (1979) Effect of efflents and receiving waters to freshwater and marine insecticides on the short-circuit current and resistance of organisms. United States Environment Protection Agency, isolated frog skin. Pestic Biochem Physiol, 10: 23-30. Washington, DC. Winchester, E. L., Moore, M.T. and Farris, J. L., (1996) The van den Bercken, J., (1977) The action of allethrin on the acute toxicity and developmental effects of propanil (3’,4’- peripheral nervous system of the frog. Pestic Sci, dichloropropionanilide(DCPA)) to Xenopus laevis and 8: 692-699. Rana sphenocephala. SETAC 17th Annual Meeting. van den Bercken, J., Akkermans, L. M. A. and van der Zalm, Washington, DC, Abstracts, p. 141. J. M., (1989) The effect of DDT and dieldrin on skeletal Wohlgemuth, E., (1977) Toxicity of endrin to some species of muscle fibres. Eur J Pharmacol, 21: 89-94. aquatic vertebrates. Acta Sc Nat Brno, 11: 1-38. Vardia, H. K., Rao, P.S. and Durve,V. S., (1984) Sensitivity of Woodall, C. and Maclean, N., (1992) Response of Xenopus toad larvae to 2,4-D and endosulfan pesticides. Arch laevis to cadmium administration. Comp Biochem Physiol C, Hydrobiol, 100: 395-400. 101: 109-116. Venturino, A., Gauna, L. E., Bergoc, R. M. and Pechen De Wyman, R. L., (1988) Soil acidity and moisture and the Dangelo, A. M., (1992) Effect of exogenously applied distribution of amphibians in five forests of southcentral polyamines on malathion toxicity in the toad Bufo New York USA. Copeia, 2: 394-399. arenarum Hensel. Arch Environ Contam Toxicol, 22: 135-139. Xu, Q. and Oldham, R. S., (1997) Lethal and sublethal effects Vijverberg, H. P.M., Zalm, J. M. and Bercken, J. v. d., (1982) of nitrogen fertilizer ammonium nitrate on Similar mode of action of pyrethroids and DDT on (Bufo bufo) tadpoles. Arch Environ Contam Toxicol, sodium channel gating in myelinated nerves. Nature, 32: 298-303. 295: 601-603. Yorio,T. and Bentley, P.J., (1973) The effects of methylmercury Vinson, S. B., Boyd, C. E. and Ferguson, D. F., (1963) Aldrin compounds on the skin (in vitro) of the leopard frog (Rana toxicity and possible cross-resistance in cricket frogs. pipiens). Comp Gen Pharmacol, 4: 167-174. Herpetologica, 19: 77-80. Zaffaroni, N. P., Arias, E., Capodanno, G. and Zavanella,T., Vismara, C., Bernardini, G., Bordone, L., Spinelli, O.,Teruzzi, A. (1978) The toxicity of manganese and Rossetti, C., (1995) Effects of chlorocresol (4-chloro- ethylenebisdithiocarbamate to the adult newt, Triturus 2-methyl phenol) administered during the fertilization and cristatus. Bull Environ Contam Toxicol, 20: 261-267. cleavage phases of Xenopus laevis. Bull Environ Contam Zaffaroni, N. P., Zavanella,T., Cattaneo, A. and Arias, E., (1986a) Toxicol, 55: 195-200. The toxicity of 2,4-dichlorophenoxyacetic acid to the adult Vismara, C. and Garavaglia, A., (1997) 4-Chloro-2- crested newt. Environ Res, 41: 79-87. methylphenoxyacetic acid containing compounds. Zaffaroni, N. P., Zavanella,T., Ferrari, M. L. and Arias, E., (1986b) Genotoxicity evaluation by Mutatox assay and comparison Toxicity of 2-methyl-4-chlorophenoxyacetic acid to the with acute (Microtox) and embryo (FETAX) toxicities. Bull adult crested newt. Environ Res, 41: 201-206. Environ Contam Toxicol, 58: 582-588. Zavanella,T., Pacces Zaffaroni, N. and Arias, E., (1979) Vismara, C., Rossetti, C., Bolzacchini, E., Orlandi, M., Luperini, Preliminary study on the carcinogenic activity of the A. and Bernardini, G., (1996) Toxicity evaluation of 4- fungicide manganese ethylenebisdithiocarbamate in the chloro-2-methylphenoxyacetic acid by microtox and adult crested newt Triturus cristatus carnifex.Tumori, comparison with FETAX. Bull Environ Contam Toxicol, 65: 163. 56: 85-89. Zavanella,T., Zaffaroni, N. P.and Arais, E., (1988) Evaluation of Vogiatzis, A. K. and Lombourdis, N. S. (1998) Cadmium the carcinogenic risk of the phenoxy herbicide MCPA to accumulation in liver and kidneys and hepatic an urodele amphibian. Ecotoxicol Environ Safety, metallothionein and glutathione levels in Rana ridibunda, 16: 114-122. after exposure to CdCl2. Arch Environ Contam Toxicol, 34: 64-68. Zavanella,T., Zaffaroni, N. P.and Arias, E., (1984) Abnormal limb regeneration in adult newts exposed to the fungicide Wang, C. and Murphy, S. D., (1982) Kinetic analysis of species Maneb 80. A histological study. J Toxicol Environ Health, differences in acetylcholinesterase sensitivity to 13: 735-746. organophosphate insecticides. Toxicol Appl Pharmacol, 66: 409-419. Warner, S. C., Dunson,W. A. and Travis, J., (1991) Interaction of pH, density and priority effects on the survivorship and growth of two species of hylid tadpoles. Oecologia, 88: 331-339.

201 Declines and disappearances of frogs: risk assessment and contingency strategies

Michael Mahony, John Clulow, Robert Browne and Melissa Pomering *

ABSTRACT involves a pro-active experimental approach to determining the cause of declines and the thresholds The recent decline and extinction of amphibian at which to initiate action. Contingency strategies species is a worldwide phenomenon without an involve a suite of methods which, when identified cause or solution. In the face of the implemented, form an insurance against inevitable further loss of species and population uncontrolled loss of species and populations.These diversity a more active program of risk assessment should involve captive husbandry, the development and implementation of contingency strategies is of reproductive technologies and the establishment required.The goal of such a program is simple and of a germplasm bank. Cryopreservation effectively directly useful, it prevents the total loss of lengthens the ‘genetic lifespan’ of valuable irreplaceable evolutionary genetic information that individuals, who can contribute to the genetic occurs with extinction. It involves two steps that are diversity of populations and species, or represent it, possible immediately.The first is a pro-active and allows higher levels of genetic variation to be experimental approach to identifying the causes maintained in smaller populations. and mechanisms of declines.The second is to take contingency actions as a form of insurance against the further loss of species and population diversity.

Risk assessment provides explicit means to recognise and prioritise species that are or may be susceptible to as yet unrecorded declines. It also

* Department of Biological Sciences,The University of Newcastle, University Drive, Callaghan NSW 2308.

202 INTRODUCTION Whatever the cause the outcome remains the same; isolation and reduction in population sizes to the point where they are The aim of this paper is to assess the major features of the susceptible to stochastic effects whether environmental, disappearances and declines among Australian frogs with a demographic, catastrophic or genetic. Secondary questions view to proposing methods to prevent further losses. Much of will most likely arise whether a single factor or multiple the paper is based on the experiences of the past eighteen factors are involved, and whether all declines/disappearances years, a time of considerable turbulence for those who work are due to the same factor(s). However, waiting for answers on frogs. Our perspective is not to be critical of the past but to these questions may leave a situation where further to use its lessons to move forward, to avoid making the same irreplaceable evolutionary diversity is lost. Solutions need to mistakes, and to prevent future loss of biodiversity. be sought whether the cause is known or not. It has been eighteen years since the disappearance of the Two steps are possible immediately.The first is a pro-active southern gastric brooding frog (Rheobatrachus silus) from south experimental approach to identifying the causes and east Queensland. Since that time another seven Australian mechanisms of declines.The second is to take contingency species have disappeared, with as many as a further 20 showing actions as a form of insurance against the further loss of signs of declines and regional disappearances (Mahony 1996). species and population diversity.The consequences of Field monitoring and surveys have shed light on various aspects inaction can be considered from a number of angles which of these declines and disappearances with various hypotheses relate to the instrumental and intrinsic values that are advanced.Tests of these hypotheses have begun but it is recognised for all living organisms, communities and apparent that we are a long way from a satisfactory explanation. ecosystem (Meffe and Carrol 1994). Unlike most cases of declines where the primary contributing factors are apparent (habitat loss, introduced species, overexploitation, and pollution) the cause of most amphibian WHAT IS RISK ASSESSMENT? declines remains unknown. Because the declines in Australia are Caughley and Gunn (1996) define risk assessment in recognised as being part of a global phenomenon (Blaustein and conservation biology as “estimating the probability of an Wake 1990;Tyler 1991), attention has been directed towards adverse event.”They divide assessment into, the probability agents that may act on this scale. Four major hypotheses have that the event will occur over a specified time interval, and been advanced: increased ultraviolet radiation, climate change, the probability that it will lead to the extinction of the action of a virulent pathogen, and pollution (for a review see population. Determining the events and their probabilities are Pounds et al. 1997). If any one of these postulated causes were therefore the prime elements in risk assessment.The greatest shown to be the primary “agent” what could be done to difficulty that faces the conservation biologist when dealing mitigate the impacts? It is highly unlikely in the short-term, that is with declining frogs is that the “agent” or “cause of declines” at least the next decade, that any effective action could be taken remains unknown.To use risk assessment in this standard to halt the declines. It simply would not be possible to protect all manner in the case of declining frogs is akin to asking the populations against increased ultraviolet radiation and the hole in process to cure a problem that it cannot logically diagnose. the ozone will not begin to decrease until about the year 2020 (UNEP 1992;WMO 1995). Likewise, it is not possible to The most common form of risk assessment used when vaccinate wild populations to deal with a pathogen. At best it dealing with endangered species is “population viability may be possible to put in place a range of protocols to slow the analysis” (PVA)(Gilpin and Soule 1986).There are various dispersal of such an organism if it is aided by human activities. techniques and software available to conduct PVA’s (for a What can be done to mitigate the possible harmful effects of review see Burgman et al. 1993) most of which rely on stable hormonal mimics that are spread around the world in the population demographic data.This form of data has rarely atmosphere? We argue that contingency strategies must be been collected for frogs in Australia, and the question that researched and put in place, regardless of whether or not the must be asked for the species’ that have already disappeared, “agents” of decline are identified. is whether knowledge of various demographic variables would have made any difference to the outcome? It is unlikely that The decline and disappearance of frogs in Australia and other detailed knowledge of life expectancy, birth and death rates continents from relatively undisturbed habitats provides the and measures of fecundity would have made any difference to inherent implication that habitat loss and fragmentation are predicting the declines, or to provide a solution.This was not responsible for all declines and that more subtle agents because the declines occurred so rapidly, and in several case are involved in many cases (Pounds et al. 1997).Whether were undetected (Mahony 1993a; Richards et al. 1993). some reported declines differ from those expected from natural population dynamics or the expected impacts of In the first five years of the disappearance of Australian frogs, various anthropogenic factors has been the subject of an because the “cause of decline” was unknown and therefore ongoing controversy (Pechmann et al. 1991, Pechmann and could not be foreseen its probability could not be Wilbur 1994; Pounds et al. 1997).The evidence from Australia determined. However, this should not have been an excuse is that natural fluctuations were not responsible. for the lack of risk assessment in the subsequent thirteen Disappearances in isolated populations were rapid (Trennery years that saw a further five species disappear and another et al. 1994; Laurance et al. 1996), and direct observations of 20 decline markedly (Mahony 1996). sick and dying individuals followed by population crashes were made at a number of geographically widespread localities (Dennis and Mahony 1994; Laurance et al. 1996; Berger et al. 1998). Finally, even if the cause could be identified mitigation is not likely to be possible in the short-term.

203 Risk assessment of the type discussed above is useful when Queensland. In the rainforests of this region six species of the cause of declines is known, whether it be due to intrinsic stream frogs disappeared from altitudes above 600 m. Species or extrinsic factors, however, it cannot work effectively in the of Cophixalus occupied habitats adjacent to these stream absence of any identified causal factor. Intrinsic factors, such as frogs, but none have apparently declined.They are still level of fecundity and death rates under “natural” conditions, commonly encountered in these habitats. A major difference can be included in risk assessment calculations. For example if between them and the species that disappeared was that the minimum habitat requirements of a species are they do not breed in streams and do not have free-living compromised the probability that a population will decline and tadpoles. Relying on restricted high altitude distribution as perhaps disappear can be calculated. If however an extrinsic factors would have led to a prediction that these species had factor, such as an introduced pathogen to which the native high probabilities of declining. fauna has not previously been exposed is responsible, then calculating the probability in the first instance is not possible. The probability of rapid declines from an unknown cause leaves biologists initially with very few options for action.This is even more so when base-line data on population EVIDENCE THAT THE CAUSE OF distribution and abundance or even the most basic DECLINES/DISAPPEARANCES IN demographic data have not been obtained. Caughley and AUSTRALIA WAS NOT DUE TO AN Gunn (1996) point out that PVA can hamper diagnosis of a INTRINSIC FACTOR conservation problem because it begins from the assumption To the best of our knowledge, no Australian frog had that the problem is low numbers;“It is not concerned with disappeared in the first 200 years of European settlement of why the population or species declined to small numbers, so the continent. In the late 1970’s, the probability of an “event” it does not prescribe how to get numbers back up to healthy occurring that would cause disappearances among frogs levels”.Viability analyses are at their most useful in estimating would have been very low. In 1980 if someone had asked the vulnerability of small populations to stochastic events, what the probability was that a species would go extinct, because the methodology treats the population at risk in the attention would have focused on those species with narrow context of a density-dependent single-species model.There is distributions, those reliant on special habitats, and those that ample evidence that the disappearances of populations and occupied habitats impacted by human activities.This was the species of frogs were not the result of a stochastic event in case for at least one species, the baw baw frog (Philoria frosti) the sense that populations were not reduced to small size which has an extremely narrow distribution, specialised followed by the action of stochastic events.The probability reproductive pattern, and its habitat coincided with the that this would happen to 20 or so species in a ten year development of ski slope facilities (Littlejohn 1963). period, across a variety of habitats and with a wide Accordingly the distribution and abundance of this species geographic range, when it had not happened in the past was recorded (Malone 1985), but its subsequent dramatic 200 years is very low. decline was not predicted (Hollis 1995).When assessing the Essentially the same conclusions have been reached for the status of Australian amphibians in 1978,Tyler considered that declines observed in Central America. Pounds et al. (1997) the most endangered species were those at montane addressed the basic question of whether the number of localities or with restricted distributions and he listed six disappearances in the Monteverde region of Costa Rica species, Arenophryne rotunda, Cophixalus concinnus, C. saxatilus, exceeded that expected for naturally fluctuating populations. Litoria longirostris, Philoria frosti, and Rheobatrachus silus. Using long-term studies of amphibian assemblages to develop He argued that their restricted geographical ranges were probabilistic null models for the number of disappearances, probably attributable to past climatic changes and not the they concluded that the observed pattern was highly effects of humans. He also pointed out that a high proportion improbable in the context of normal demographic variability. of the Australian frog fauna was only recently described and niche requirements and geographic ranges of many Failure to adopt even a minimal approach to risk assessment species were unknown. In such circumstances it was not has led to essentially status quo management when a series of possible to determine whether many species were urgent actions were necessary.With the knowledge that eight endangered or abundant. species had disappeared and a further 20 had clearly declined, the probability that other species may meet the Two of these species (L. longirostris, R. silus) are now same fate could be calculated.When combined with some presumed to be extinct (Richards et al. 1993), and a third means of prioritising species on a scale of susceptibility based (P. frosti) has declined markedly in range and abundance.The on a range a biologically meaningful correlates of past natural range of A. rotunda was found to be much larger than declines such as, size of total distribution, altitudinal range, originally known and there has been no suggestion that it has fecundity, breeding site and taxonomic affinities, a probabilistic declined in geographic range or abundance (Roberts 1990). scenario could be produced.To some extent this is what has The available evidence is that the two species of Cophixalus happened in a qualitative manner. Biologists aware of the have not declined (Mahony, pers. obs.). Using only restricted correlates have focused monitoring on species and distribution as a factor gave some predictive power of communities with a mix of these characteristics. However, this endangerment. However, it belied the unpredictable nature process did not prove to be effective because between 1984 and some of the most informative features related to the and 1993 species continued to disappear.This was due in disappearances that occurred. Several species of the genus part to two factors, the correlative approach did not provide Cophixalus, including the two listed by Tyler, occur in restricted high predictive value in all cases and there was an distributions at high altitude in the far north east of unwillingness to heed the warnings.

204 A PROPOSED APPROACH TO predictions about the potential for decline of the alpine RISK ASSESSMENT corroboree frog (Pseudophryne corroboree) or the tableland bell frog (Litoria castanea). Predictions were even more In the context of the disappearances of Australian frogs the difficult when it is appreciated that some sympatric rainforest aim of risk assessment should be to identify those species stream breeding frogs have not declined, and that no clear which are susceptible to whatever is the “cause of decline” phylogenetic pattern of susceptibility has been identified and thereby put into place appropriate strategies to prevent (Mahony 1996). their disappearance.This is possible even if the precise cause cannot be identified. Suspect species need to be prioritised We propose that a scoring system be adopted that combines and actions defined and implemented. Contingency strategies the predictive features of the species that have already form an essential part of the actions in the absence of threat declined/disappeared to produce a list of suspect species. abatement, because the precise threat has not been identified A simple example would be: and may not be easily mitigated. Predictive features Points In developing a risk assessment strategy several important points should be kept in mind.The species that are most 1. Altitudinal Range likely to be at risk are those with narrow geographic ranges, entirely above 300 m 10 occur at higher altitudes, with a concomitant narrow climatic Includes component above 300 m 5 tolerance.There have been few explicit studies of climatic and 2. Distribution physiological tolerance for Australian frogs, other than for some < 100 sq km 10 desert amphibians. Bioclimatic modeling has been conducted < 400 sq km 5 for some geographic regions and the close correspondence < 1 600 sq km 3 between climatic attributes and distribution was demonstrated 3. Breeding site (Nix and Switzer 1991). Secondly, certain species groups Stream breeding 10 whether they are delineated ecologically or by phylogenetic Upland swamp or bog 10 relationship have been susceptible. Attention should be 4. Fecundity directed to other ecologically similar and related species. < 50 eggs/female 10 The strategies are to identify susceptible species and use field <1 000 eggs/female 5 and laboratory experiments to identify “the cause” and 5. Phylogenetic affinity with species mechanism of declines. Means of insurance, such as captive already declined breeding and use of reproductive technologies (including Sibling species 10 germplasm banks) are considered as part of a broader Member of species group 5 conservation program.They have the added advantage of Monotypic genus 7 providing valuable information about various aspects of the reproductive biology, fecundity, diseases, parasites, behaviour, A score above 30 would place a species in the highly and physiology of the species. suspect category.

The finer details of such a scheme would require a greater PREVIOUS AND PROPOSED ACTIONS amount of detail and should be considered by a working The above discussion may give the false impression that group that involves biologists with a knowledge of the frogs disappearances have occurred with no action being taken. with assistance from those involved in PVA modeling. The response to date has been to establish four lines of research; (a) identification of the susceptible species, (b) b. Monitoring monitoring of declining and suspect species, (c) experimental Monitoring programs for several species that were shown or research to elucidate the cause of decline, and (d) captive expected to be in decline were established and several breeding and reintroduction of tadpoles. continue (McDonald 1990; Mahony 1993a,b; Hollis 1995; Osborne 1996). Monitoring is an essential action in any a. Identification of susceptible species contingency strategy. By itself however it is not likely to lead Previous attempts to identify susceptible species have focused to an identification of the cause, unless this happens by on identifying the ecological similarities of declining species chance observations.This approach has provided and their geographical distribution (Richards et al. 1993; confirmation of the speed and impact of the “event” in some Mahony 1996; Hero and Williams 1997), and while these have habitats while in others the pattern has been less obvious. proved to be valuable their predictive powers have been low. Declines in upland rainforest habitats in southern, mid and north Queensland were rapid (Richards et al. 1993; Laurance After the disappearance of the southern gastric brooding frog et al. 1996), whereas the decline of the baw baw and (R. silus) and the southern day frog (T. diurnus) in 1979-1980, corroboree frog have apparently been more gradual (Hollis it was reasonable to be wary and in a simple predictive 1995; Osborne 1996). manner keep a close watching brief on other sibling species. However, it was not possible to make the connection, on the We propose that monitoring be linked to established basis of biological similarities or phylogenetic affinities from thresholds for action.This is akin to having an explicit decision these upland rainforest stream dwelling species, with tree. Once again an example will make this point more understandable.

205 • If a population is estimated to have less than 5 000 The last of these studies had broader objectives. In 1993 adults (note that the effective population size is when it was conducted there was limited information on the considerably less)(Wright 1931, 1938): commence a aetiology of declines, for example it was not known which close monitoring program. part of the life stage was susceptible, whether reproduction was affected or if declines occurred at particular seasons • If a population is estimated to have less than 1 000 adults: and/or climatic conditions. Among declining species in the establish husbandry capabilities and techniques, sample for Wet Tropics region tadpoles remained in the streams well germplasm bank (i.e. sperm for 50 males and nuclei from after adults could no longer be detected (Richards et al. 50 females). 1993;Trennery et al. 1994), and in mid eastern Queensland • If a population is estimated to have less than 500 adults: McDonald (1990) noted that adults and juveniles were start captive breeding program immediately. Positive use present at the end of the summer season but did not of germplasm storage. reappear in the next spring. Using field enclosures to closely • If a population is estimated to have less than 100 adults: follow the fate of individuals of the sharp-snouted day frog Taudactylus acutirostris ensure contingency mechanisms such as germplasm bank ( ) Dennis and Mahony (1994) aimed to storage has been undertaken. determine at what stage, when and how animals died. At the time this study was conducted this frog had disappeared from c. Identification of the agent of decline about 90% of its former range and was restricted to several stretches of streams above 600 m on the Big Tableland in the Determination of the cause of decline has produced the north of the Wet Tropics region.The approach was to greatest level of debate (Laurance et al. 1996; Alford and reintroduce and observe adults and tadpoles exposed to the Richards 1997; Hero and Gillespie 1997). Initially the difficulty environment where they were once common and also in a was the uncertainty surrounding the reality of declines. Many control at the Big Tableland. All adults that were transferred biologists were suspicious of the reports of amphibian died including those in control conditions. Most tadpoles declines and pointed to the well-documented natural survived and many metamorphosed.With the exception of fluctuations in populations reported for various species, one translocation site, all juveniles died shortly after populations and communities in various habitats (Pechmann metamorphosis. Similar results were obtained with captive et al. 1991; Pechmann and Wilbur 1994). Indeed this initial colonies from the same source population; all adults held at caution, and appreciation of the variable climatic conditions of James Cook University Townsville died, tadpoles held by eastern Australia, led those who reported the first Taronga Zoo Sydney and Melbourne Zoo experienced levels disappearances (the southern gastric brooding frog and of mortality, and most juveniles died after metamorphosis southern day frog) to propose that it was the result of except for a small number that grew to adults at Melbourne drought (Tyler and Davies 1985). (North Queensland Threatened Frog Recovery Team 1995).

Another problem that has complicated assessment of the cause While establishing this experiment sick and dying individuals of declines is the multitude of observed factors associated with of a number of species that had been reported to be in declines among widespread species (Ferraro and Burgin 1993). decline (Litoria genimaculata, L. nannotis, L. rheocola) were Most cases are closely related to deterministic effects (land observed. Specimens were collected and dispatched for clearance, introduced predators, pollution). It is not surprising veterinary examination.Within three months of these that frogs are difficult to find in urban landscapes, where ponds observations all populations of these species had disappeared have been removed, or if they remain fish have been above 300 m altitude on the Big Tableland (Laurance et al. introduced, run-off that fills them is often polluted, and streams 1996).The evidence was that the frogs were involved in a have been turned into storm-water channels that are often rapid decline event. Adults were found moribund or dead in cement lined. Native vegetation that is required for foraging and the microhabitats they routinely occupied, L. nannotis were shelter has been removed or greatly modified. It is surprising found on ledges of a waterfall, L. rheocola on exposed rocks that frogs survive at all in some urban areas. within the stream or on small overhanging the stream. In the search for the cause of the catastrophic disappearances Dead and moribund individuals did not appear to be wasted, and declines among certain upland species the approach has the males had normally developed secondary sexual been poorly focused.We propose an explicit experimental characters (nuptial pads and spines), one female L. rheocola approach to identify the causal agent(s) which relies on the was gravid with large yolky eggs observable through the recognition that the frogs are the sensitive end point (Dennis ventral surface.The overwhelming impression was that the and Mahony 1994).Therefore it is the susceptible species that animals had died suddenly, in the microhabitats that they should be experimentally exposed to the postulated causes usually occupied during the warmer months of the year. of decline.This approach has been adopted in a number of A number of external signs of illness were common to most studies, sadly most of which have been of limited duration specimens, external lesions, clouded corneas, sloughing skin and scope. Examples included investigations of the impact of and lethargic responses.The proposed translocation/exposure ultraviolet radiation on early embryonic and larval experiment with adults from this population was unsuccessful, development in a small number of species (van de Mortel and it was presumed that this was because they were either and Buttemer 1996), impacts of introduced predatory fish already ill or that contact between individuals meant that (Harris 1995;White and Pyke 1996;Webb and Joss 1997; transmission was possible. Dennis and Mahony (1994) Gillespie and Hero 1999), and the role of various habitat concluded that a disease was active at the site, however, variables (Dennis and Mahony 1994). whether this disease was the proximal cause or a symptom of the decline could not be determined. It was highly unlikely that sick individuals would have been detected without close

206 observations, and it was not surprising that if a similar process In the absence of an identified cause, and the likely inability to had occurred in other populations that the event would go mitigate against any of the proposed causes there is little undetected.The “event” had all the hallmarks and inherent option but to examine possibilities that act as insurance difficulties of detection of what Underwood (1990) termed a against further losses. pulse event. Short duration, and apart from killing the animals, leaving no sign of its passing. Speare et al. (1993) postulated The aim of this section is to provide evidence that some that a virus was responsible. strategies should be put in place as insurance against the continued uncontrolled loss of species and populations.We consider three points that lead to this conclusion. Firstly, AN EXAMPLE OF RISK ASSESSMENT unlike most cases where the primary factors contributing to What do we mean by risk assessment? An example of the extinction are apparent (habitat loss, introduced species, concept will be presented (for brevity precise details of overexploitation, and pollution) the cause of most amphibian experimental design have not been included). declines remains unknown. Secondly, even if the cause was identified, mitigation is unlikely to be effective in the short- In south east Australia within the bell frog species group one term.Thirdly, apart from the species that have disappeared species (Litoria castanea) has disappeared and another numerous others have declined and become susceptible to a (L. aurea) has declined markedly (see Mahony 1999).Two new situation; populations become isolated and reduced to members of this species group (L. cyclorhynchus and L. moorei) the point where they are susceptible to stochastic effects occur in the south west of Western Australia.What is the whether environmental, catastrophic, demographic, or genetic. probability that these species may be susceptible and perhaps An example of the latter situation would be Fleayi’s barred decline/disappear? Let us assume that the cause of declines is a frog (Mixophyes fleayi) which in northern NSW and virulent pathogen or increased ultraviolet radiation. Our aim southeast Queensland occurs in small disjunct populations. would be to determine the susceptibility of these species to these factors. How do we proceed in the absence of an It remains unknown whether demographic and environmental identified pathogen? The approach would be to collect the stochasticity contributed to the demise of Australian frogs appropriate number of individuals and transport them to that have disappeared. However, once declines have eastern Australia (captive breeding could provide a large proceeded it is clear that these may contribute to further numbers of experimental individuals but this would use declines. Likewise, there is no evidence that the species that potentially valuable time). After a suitable quarantine period field declined/disappeared had reduced genetic variation. exposure studies, suitably replicated, should be conducted at Unfortunately, genetic measures were not available for any of sites where the bell frog once occurred but is now absent.The the species that have disappeared and are available on a most suitable location would be on the southern tablelands of limited basis for some species that have shown marked New South Wales.The ponds used should be free of decline (Colgan 1996; Osborne et al. 1996). Loss of genetic introduced fish and support a frog community. Field enclosures variation has been shown to lead to an increase in could be modelled on those used in numerous other studies. If susceptibility to disease and increased parasite load in wild ultraviolet radiation is suspected then associated laboratory populations (Vrijenhoek 1994), and these populations would studies should be conducted to produce dose response curves be expected to have reduced ability to cope with for ultraviolet radiation exposure, and accurate measure of field environmental change during evolution. However, the pattern values of ultraviolet radiation obtained.Veterinary pathologists of disappearances and declines do not bear the should be part of the research team. Laboratory experiments characteristics of species that become endangered as a to test whether the species are susceptible to already isolated consequence of long histories of small effective populations pathogens, such as Cythrid fungus (Berger et al. 1997) would sizes and consequent inbreeding and low genetic variation. also be highly desirable. Populations of some of these species were large and they extended over considerable range (McDonald 1992). Experiments similar to that above could be designed to Conversely, when species and populations decline rapidly determine whether the high altitude species of Tasmania from whatever cause (habitat loss, pollution or disease) levels (Bryobatrachus nimbus, Crinia tasmaniensis), NSW, Queensland of genetic variation will be normal.These tests for those and Victoria (e.g. Litoria subglandulosa, Mixophyes balbus, M. species that have declined markedly in Australia have not fleayi, Philoria kundagungan, P sphagnicolus and been conducted, and they are not possible for the species P. loveridgei,Taudactylus liemi) are susceptible to the agent that have disappeared.Whichever the case, inbreeding of decline. depression will exacerbate endangerment in populations that have been reduced to small sizes. Consequently, genetic CONTINGENCY STRATEGIES concerns are of major significance in the majority of endangered species and populations.The documented Declines/disappearances in isolated populations were rapid greater susceptibility to extinction of island populations is (Laurance et al. 1996; Alford and Richards 1997), and because related to the lower genetic variability of island populations of this most of the affected frogs were not subject to routine (Nevo et al. 1984). A decline in genetic variation can monitoring. Furthermore, many were in relatively remote preclude a species’ ability to respond to natural selection and locations. Hence reactions by biologists were cautious and consequently limit its evolutionary potential. In the long term, often too late. Cautious because amphibian populations are genetic variation can be a critical factor for persistence in a notorious for their seasonal fluctuations. Captive breeding changing environment. Small populations are often subject to was attempted only as a last ditch effort in one species, but others had disappeared before such efforts were contemplated.

207 the loss of alleles through genetic drift, or random amphibians, one of the major hurdles to ease of use of AI, fluctuations in allele frequency (Frankham 1995a,b,c). female reproductive physiology, is removed. Optimum methods Reduced population size can also result in inbreeding for semen handling and preparation have resulted from depression, or a decline in population fitness caused by research and development in agriculture, fisheries and clinical increased homozygosity and the unmasking of recessive laboratories. Looking forward, the latest developments in deleterious alleles (Leberg 1992; Hartl and Pucek 1994). reproductive technology might be employed to maximise use These problems can be reduced and overcome by the use of of genetic material. Intracytoplasmic sperm injection is currently gene banking technologies.To maximise the efficiency and used as an alternative to in vitro fertilisation (IVF) for infertile effect where endangered species are concerned a human couples. In principle, this offers a means of using poorly combination of methods and strategies are applied. preserved spermatozoa, providing a supply of oocytes is available. It would be potentially possible to restore a species WHAT CONTINGENCY STRATEGIES ARE using sperm only by hybrid backcrossing if a suitably close POSSIBLE TO AVOID CATASTROPHIC sibling species existed. Embryonic stem cells offer another LOSSES? potential avenue, these cells derived from the blastocyst when injected into enucleated eggs are capable of directing normal Captive breeding and the establishment of reproductive growth and development to adult frogs (Briggs and King 1952). technologies are the most plausible and achievable solutions. Cryopreservation of germ cells is the centrepiece in a range In parallel with a germplasm bank, a collection of associated of technologies involved. cells and sera would provide a useful backup. Any unforeseen questions about the origin or relatedness of individuals and CAPTIVE BREEDING evolutionary systematics of the taxa could be answered, and questions about the disease status of samples could be The advantages of captive breeding when a species faces investigated. extinction are obvious. However, there are some major hurdles to be overcome if the eventual aim is to return ADVANTAGES OF CRYOPRESERVATION species to the wild (Ralls and Ballou 1986a,b; Benford 1992; Holt 1992; Holt et al. 1996). In the context of the current The primary aim is to provide protection against situation where the cause of declines remains unknown and catastrophes. It is far better to know that a representative mitigation not possible it is uncertain how long stock may sample of the genome of a species is preserved in a manner need to be kept in captivity before release is possible.The that can be activated in the future than not to have this cost of captive breeding increases with the number of possibility. It is a tragedy that no preserved samples of the generations in captivity and the size of captive populations. genome of a unique evolutionary entity such as the gastric brooding frog were made when the species was common. It is widely recognised that captive breeding is a last ditch A tragedy repeated with the extinction of the second effort in the management of an endangered species. Despite species, five years after the loss of the first species. this there seems to be few guidelines or thresholds that are available to wildlife managers to indicate when captive Cryopreservation could enhance the breeding of animals in breeding should be considered. Presumably this is because captivity and increase the conservation biologist’s options for such thresholds differ between species and require the future. Single female animals could be mated with information on population demography, genetics and the multiple mates by IVF and the number of offspring produced cause of decline. Captive populations of endangered species by captive animals could be significantly increased, without the are most often founded from small numbers such that potential negative genetic effects of small population size. inbreeding becomes unavoidable. It is clearly desirable to Genome banks could be used in the same way as seed banks minimise inbreeding and loss of genetic variation in the genetic have been used for agricultural crops. Genetic diversity lost management of threatened species both in the wild and in in situ, in isolated populations, or in populations undergoing captivity.While close attention to pedigrees and schemes to declines and bottlenecks could be recovered and restored. avoid and/or reduce inbreeding depression have been devised The possibility of bringing back extinct species is remote and and applied with some success, this remains an area requiring in most cases nonexistent. One of the hurdles is that the close management. Finally it must be stressed that captive DNA in nuclei of tissue preserved in a museum is likely to husbandry techniques must be established if the reproductive have deteriorated with time.This is one of the most technologies discussed below are to be successful. important reasons to research cryobiology.

A range of potential benefits of germplasm banks for GENETIC RESOURCE BANKS conservation programs requiring metapopulation Genetic resource banking is the term given to the storage by management can be identified.Two adverse genetic changes cryopreservation of gametes and embryos from threatened which occur in captive breeding programs — inbreeding populations with the specific and deliberate intention to use depression and loss of genetic — variation can be overcome them in a breeding program at some future occasion (Moore by reproductive technologies. Many captive populations of et al. 1992; Holt et al. 1996). Genetic resource banks already threatened animals have been founded by only small exist and function successfully for domestic animals of numbers, because there is a natural caution about removing agricultural importance. Dairy cattle are routinely bred by the animals from the wild especially when the numbers are use of artificial insemination (AI) with frozen semen, and cattle decreasing. However, the measures we propose do not breeding centers with stocks of frozen semen exist throughout remove animals from isolated and/or small populations. the world. Because external fertilisation is the norm in Sperm can be collected unobtrusively within a short period

208 and the animals released immediately. Large numbers of breeding programs.The number of founding individuals rarely individuals can be assayed and samples stored indefinitely. exceeds ten adult pairs. Frankham (1995a,b,c) has argued that Secondly, the possibilities for extending the generation no finite population appears to be immune from inbreeding interval by means of cryopreservation are almost limitless. depression in the long-term, and to maintain evolutionary potential an effective population size at least in the hundreds GENERALLY PERCEIVED PROBLEMS is necessary. Cryopreservation can play a vital role to WITH OUR APPROACH overcome these difficulties by increasing the generation interval, thus reducing the number of populations likely to be It is generally accepted that the least expensive way of threatened. It is against this background that the long-term preventing extinction and loss of biodiversity is the advantages of cryopreservation should be viewed. maintenance of habitats.This argument is well established in the conservation biology literature (Caughley and Gunn ACKNOWLEDGEMENTS 1996), however, it does not consider or deal with a situation such as that which currently faces frogs in Australia and Field work by MM was supported by grants from the globally. One of the puzzling features is that species have Australian Museum Sydney, New South Wales National Parks disappeared from areas of pristine or near pristine habitat and Wildlife Service and Environment Australia. Laboratory and areas of large reserves where there are no indications of studies by JC, RB and MP have been supported by an ARC habitat destruction. Similarly, there is no evidence that an grant and the University of Newcastle.The manuscript was introduced competitor or predator is responsible, apart from critically reviewed by Stephen Donnellan. the hypothesis that an introduced pathogen is involved (Laurance et al. 1996). Preservation of habitat or declarations REFERENCES of new reserves would not have halted or prevented the loss Alford, R. A, and Richards, S. J., (1997) Lack of evidence for of the majority of species. Indeed, the only frogs that have epidemic disease as an agent in the catastrophic decline of disappeared in eastern Australia for which some indication Australian rainforest frogs. Conserv. Biol, 11:1026-29. that habitat change may be instrumental or played a contributory role are in the tableland bell frogs (L. castanea/ Blaustein, A.R., and Wake, D.B., (1990) Declining amphibian flavipunctata). Evidence that introduced predators (fish) have populations: A global phenomenon. Trends Evolutionary played a role with some declining species has been presented Ecology, 4(7):203-204. (Pyke and White 1996; Gillespie and Hero 1999). However, it Benford, G., (1992) Saving the “library of life”. Proc. Nat. Acad. remains uncertain whether this is the primary or a Sci. USA 89:11098-11101. contributing cause. Berger, L., Speare, R., Dasak, P., Green, D.E., Cunningham, A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., Biotechnology is not a panacea for the loss of species McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. And because species require habitat and interacting ecosystems. Parkes, H., (1998) Chytridiomycosis causes amphibian The general ethical issues raised by use of reproductive mortality associated with population declines in the technologies and cryopreservation are not new. For rainforests of Australia and Central America. Proc. Natl. conservation biologists they concern values placed on Acad. Sci. USA 95 : 9031-9036. wilderness and naturalness as well as the concepts, paradigms, Briggs, R., and King,T.J., (1952) Transplantation of living nuclei and approaches we use to conserve biodiversity (Holt et al. from blastula cells into enucleated frogs’ eggs. Proc. Nat. 1996). However, the problem of the disappearing frogs is not Acad. Sci. USA 38:455-463. one of habitat loss.The declines defy the conventional wisdom that most extinctions have deterministic causes. Gene banks Burgman, M., Cantoni, D., and Vogel, P., (1993) Risk assessment do not obviate the necessity of in situ conservation.That we in conservation biology. London: Chapman and Hall. need to develop these technologies reflects the difficulty of Caughley, G., and Gunn, A., (1996) Conservation biology in identifying the causes of decline and the likelihood that theory and practice. Blackwell Science, Cambridge, MA. mitigation will not be simple.The habitat from which many Colgan, D., (1996) Electrophoretic variation in the Green and species have disappeared remains in reserves, habitat Golden Bell Frog Litoria aurea. Aust. J. Zool, 30:170-176. restoration is not required, and reintroduction should not be Corben, C.J., Ingram, G. J., and Tyler.M. J., (1974) Gastric viewed as an overwhelming task, once the cause of declines brooding: unique form of parental care in an Australian has be identified and mitigated. frog. Science, 186:946-947. Finally, there is the view that the declines/disappearances Dennis. A., and Mahony, M. J., (1994) Experimental indicate the occurrence of subtle and widespread agents that translocation of the endangered Sharp-snouted day frog threaten biodiversity, the so-called ‘canary in the mine’ Taudactylus acutirostris, and observations on the cause of scenario. If this is the case then the need for development of declines among montane riparian frogs. Final Report to reproductive technologies is only strengthened. the Wet Tropics Management Authority. Pp. 1-25. Ferrero,T. J., and Burgin, S., (1993) Review of environmental The primary aim of researching the development of factors influencing the declines of Australian frogs. In reproductive technologies is to avoid species extinction and Herpetology in Australia: a diverse discipline. D. Lunney the loss of evolutionary potential in endangered populations. and D. Ayres (eds) Trans. R. Zool. Soc. NSW: Mossman. The first requirement is to act as insurance against the Frankham, R., (1995a) Conservation genetics. Annu. Rev. further loss of species.The second is to avoid inbreeding Genetics, 29:305-27. depression in reduced and isolated populations and retain their evolutionary potential.This is a major hurdle for captive

209 Frankham R., (1995b) Inbreeding and extinction: a threshold Mahony, M. J., (1993b) Status of the Yellow-spotted Tree Frog effect. Conserv. Biol, 9:792-99. and Peppered Tree Frog on the New England Tablelands of Frankham, R., (1995c) Genetic management of captive New South Wales. & The Status of Great Barred River populations for reintroduction. In Serena, M. ed. (1995 frogs in NSW. Unpublished Reports to NSW National Reintroduction biology of Australian and New Zealand Fauna. Parks & Wildlife Service and Australian Nature Chipping Norton, NSW Australia: Surrey Beatty. 264 pp. Conservation Agency. Gillespie, G.R, and Hero, J.M. (1999) Potential impacts of Mahony, M.J., (1996) The decline of the Green and Golden introduced fish and fish translocations on Australian Bell Frog (Litoria aurea) viewed in the context of declines amphibians. Pp 131-144 in Declines and Disappearances and disappearances of other Australian frogs. Australian of Australian Frogs ed by A.Campbell. Environment Zoologist 30(2). 237-247. Australia: Canberra. Mahony, M.J., (1999) Review of the declines and Gilpin, M. E., and Soule, M. E., (1986) Minimum viable disappearances within the bell frog species group (Litoria populations; the processes of population extinction. In: aurea species group). Pp 81-93 in Australia in Declines and Soule, M. E. ed. Conservation biology: the science of Disappearances of Australian Frogs ed by A.Campbell. scarcity and diversity. Sunderland, MA: Sinauer Associates. Environment Australia: Canberra. Harris, K., (1995) Is there a negative relationship between Malone, B.S., (1985) Status, distribution and ecology of the Gambusia and tadpoles on the northern tablelands? BSc baw baw frog (Philoria frosti). ARI technical report No. 36. (hons) Thesis. Department of Ecosystem Management, Arthur Rylah Institute, Department of Conservation, Univeristy of New England Armidale NSW, Australia. Forests and Lands,Victoria. Hartl, G.B., and Pucek Z., (1994) Genetic depletion in the McDonald, K. R., (1990) Rheobatrachus Liem and Taudactylus European bison (Bison bonasus) and the significance of Straughan and Lee (Anura: Leptodactylidae) in Eungella electrophoretic heterozygosity for conservation. Conserv. National Park, Queensland: distribution and decline. Trans. Biol, 8:167-74. R. Soc. South Australia 114:187-194. Hero J-M., and Gillespie G., (1997) Epidemic disease and McDonald, K. R., (1992) Distribution patterns and amphibian declines in Australia. Conserv Biol, 11:1-23-25. conservation status of north Queensland rainforest frogs. Qld Dept Envir. Her. Cons.Tech. Rep. 1:1-15. Hero J-M., and Williams, S.E., (1997) Ecological characteristics of declining amphibians: Are these frogs more susceptible McDonald, K.R., and Martin,W., (1996) Recovery Plan for the to extinction? The Australian Society of Herpetologists. stream-dwelling rainforest frogs of the wet tropics Newsletter 38. p.95. biogeographic region of north-east Queensland. Australian Nature Conservation Agency, Endangered Species Hollis, G. J., (1995) Reassessment of the distribution, Program. abundance and habitat of the baw baw frog (Philoria frosti): Preliminary findings. The Victorian Naturalist, 112: 190-210. Meffe, G. K, and Carroll, C. R., (1994) Principles of Conservation Biology. Sinauer Associates Inc. Sunderland, Holt,W.V., (1992) Advances in artificial insemination and Massachusetts. semen freezing in mammals. Symp. Zool. Soc. Lond. No. 64:19-35. Moore, H. D. M, Holt,W.V., and Mace, G. M., eds., (1992) Biotechnology and the conservation of genetic diversity. Holt,W.V., Bennett , P.M.,Volobouev,V., and Watson, P.F., Oxford Univ. Press, Oxford, England. (1996) Genetic resource banks in wildlife conservation. J. Zool. Lond, 238:531-544. Nevo, E., Bieles, A., and Ben-Shlomo, R., (1984) The evolutionary significance of genetic diversity: ecological, Ingram, G.J., and McDonald, K.R., (1993) An update on the demographic and life history correlates. In Evolutionary decline of Queensland’s frogs. Pp. 297-303. In: Herpetology Dynamics of Genetic Diversity, ed G.S . Mani, pp. 13-213. in Australia: a diverse discipline. D. Lunney and D. Ayres Berlin: Springer-Verlag. (eds) Trans. R. Zool. Soc. NSW: Mossman. Nix. H.A., and Switzer, M.A., (1991) Rainforest Animals. Atlas Laurance,W. F., McDonald, K. R., and Speare, R., (1996) of vertebrates endemic to Australia’s wet tropics. Kowari Catastrophic declines of Australian rain forest frogs: 1. Australian National Parks and Wildlife Service support for the epidemic disease hypothesis. in press. Publication, Canberra. Conserv. Biol, 77:203-212. Osborne,W.S., (1989) Distribution, relative abundance and Leberg P.L., (1992) Effects of population bottlenecks on conservation status of corroboree frogs, Pseudophryne genetic diversity as measured by allozyme electrophoresis. corroboree Moore (Anura: Myobatrachidae). Australian Evolution, 46:477-94. Wildlife Research 16:537-547. Littlejohn, M. J., (1963) The breeding biology of the baw baw Osborne,W.S., (1990) Declining frog populations and frog. Proc. Linn. Soc. N.S.W. 88:273-276. extinctions in the Canberra region. Bogong 11(4):4-7. Mahony, M. J., (1991) The status of frogs in New South Wales, Osborne,W.S., (1996) Recovery plan for the southern Australia. Unpublished Report to the Declining Amphibian corroboree frog P. corroboree. NSW National Parks and Taskforce IUCN/SSC. Wildlife Service, Sydney. Mahony, M.J., (1993a) The status of frogs in the Watagan Osborne,W.S., Zentelis, R. A. and Lau, M., (1996) Geographic Mountains area, the Central Coast of New South Wales. variation in corroboree frogs, Pseudophryne corroboree Pp. 257-264. In Herpetology in Australia: a diverse Moore (Anura: Myobatrachidae) L A reappraisal supports discipline. D. Lunney and D. Ayres (eds) Trans. R. Zool. Soc. recognition of P. pengilleyi Wells and Wellington. Aust. NSW: Mossman. J. Zool. 44:569-587.

210 Pechmann, J. H. K., and Wilbur, H. M., (1994) Putting declining Tyler, M.J., (1983) Miscellany. In The Gastric brooding Frog. amphibian populations into perspective: natural Croom Helm, Canberra. fluctuations and human impacts. Herpetologica,50:64-84. Tyler, M.J., (1991) Declining amphibian populations — a global Pechmann, J. H. K., Scott, D. E., Semlitsch R. D., Caldwell, J. P., phenomenon? An Australian perspective. Aly, 9:43-50. Vitt, L. J., and Gibson, J.W., (1991) Declining amphibian Tyler, M.J., and Davies, M., (1985) The gastric brooding frog populations: the problem of separating human impacts Rheobatrachus silus. In: G. Grigg, R. Shine and H. Ehmann from natural fluctuations. Science, 253: 892-895. (eds.) Biology of Australasian frogs and reptiles, Sydney, Pounds, J. A., and Crump, M. L., (1994) Amphibian declines Royal Zoological Society of New South Wales. and climate disturbance: the case of the golden toad and Underwood, A. J., (1990) Biological monitoring for human the harlequin frog. Conserv. Biol, 8:72-85. impact: How little it can achieve. Pp 105-123, In, Pounds, J. A, Fogden M. P.L, Savage J. M, and Gorman G. C., Proceedings of the 29th Congress of the Australian (1997) Test of null models for amphibian declines on a Society of Limnology, Jabiru N.T. tropical mountain. Conserv. Biol, 11:1307-22. UNEP (1992) Environmental effects of Ozone depletion: Pyke, G.H., and White, A.W., (1996) Habitat requirements for 1991 Update. United Nations Environment Programme. the Green and Golden Bell Frog Litoria aurea (Anura: van de Mortel,T. F., and Buttemer,W. A., (1996) Are Litoria Hylidae). Aust. Zool, 30:224-232. aurea eggs more sensitive to ultraviolet -B radiation than Ralls, K., and Ballou, J., (1986a) Captive Breeding programs for eggs of sympatric L . peroni or L. dentata. Aust. Zool, populations with a small number of founders. Trends Ecol. 30(2):150-157. Evol, 1:19-22. Vial, J.L., and Saylor, L., (1993) The status of amphibian Ralls, K., and Ballou, J., (1986b) Proceedings of a workshop on populations: a compilation and analysis.Working document genetic management of captive populations. Zoo Biol, 1, IUCN/SSC Declining Amphibian Task Force. 1:81-238. Vrijenhoek, R. C., (1994) Genetic diversity and fitness in small Richards, S.J., K.R. McDonald, and Alford, R. A., (1993) Declines populations. In Loeschcke V,Tomiuk J, Jain SK, eds. in populations of Australia’s endemic tropical rainforest Conservation Genetics. Basel: Birkhauser Verlag.37-53. frogs. Pacific Conservation Biology, 1:66-77. Webb, C., and Joss, J., (1997) Does predation by the fish Roberts, J. D., (1990) The biology of Arenophryne rotunda Gambusia holbrooki (Atheriniformes: Poeciliidae) contribute (anura: Myobatrachidae) a burrowing frog from Shark bay, to declining frog populations? Aust. Zool, 30(3): 316-324. Western Australia. Research in shark bay. Report of the Woodworth, L. M., Montgomery, M. E., Nurthen , R. K, Franc-Australe Bicentenary Expedition committee. Briscoe, D. A., and Frankham R. 1994. Modeling problems Speare, R., K. Field, J. Koehler, and K.R. McDonald., (1993). in conservation genetics using Drosophila: consequences “Disappearing” Australian rainforest frogs: have we found of fluctuating population sizes. Mol. Ecol. 3:393-99. the answer? Published abstract, Second International World Meteorological Organisation (1995) Scientific Conference of Herpetology, Adelaide, Australia. assessment of ozone depletion: 1994. Global ozone and Trennery, M.J.,W.F. Laurance, and K.R. McDonald., (1994) Monitoring Project — Report No. 37. Further evidence for the precipitous decline of endemic Wright, S., (1931) Evolution of Mendelian populations. rainforest frogs in tropical Australia. Pacific Cons. Biol. Genetics. 16:97-159. 1:150-153. Wright, S., (1938) Size of population and breeding structure Trounson, A., Peura, A., and Kirby, C., (1987) Ultrarapid in relation to evolution. Science 87:430-431. freezing: a low-cost and effective method of embryo cryopreservation. Fert. Steril, 48:843-450. Tyler, M. J., (1978) The impact of european man upon Australasian amphibians. In M. J.Tyler ed.The status of endangered Australasian wildlife. Proceedings of the Centeneray symposium of the Royal Zoological Society of Suth Australia.

211 Community involvement in threatened frogs surveys, monitoring and recovery in Australia

Harald Ehmann *

ABSTRACT To improve community involvement participants must feel welcomed, morally and materially Involvement to date by the non-professional community in frog conservation activities is encouraged, and integral to a broader research outlined. Frogs and their enthusiastic and community and ‘outcomes-oriented’ network. entertaining advocates have captivated Australians. Some projects that could be run with significant Schools and public involvement in Frogwatch, community involvement are suggested. FROG CENSUS, and similar programs have exceptional community support. Frog groups are INTRODUCTION widespread. Frogweeks, Frogdays, Frogfacts, recordings of calls, and information books and This summary of Australia-wide community involvement in frog conservation activities covers both threatened and booklets sustain strong community interest. currently non-threatened species.The wider community generally considers that all frogs are ‘threatened’, and, in the Frog and tadpole study groups, herpetological non-formal sense this is true for most species. societies, landcare groups, recovery teams, science This paper is not as comprehensive an overview of frog teachers and conservation groups participate in conservation activities by community groups in Australia as I habitat rehabilitation and survey/monitoring would like. My involvement is largely voluntary and therefore activities.They provide resources for schools and my resources are limited.There have been important additional community activities since the substantive wetlands projects, raise research funds, and mount completion of this paper in early 1998 eg Alcoa Frog Watch diverse public displays. in Western Australia, the establishment of additional

* Convenor, Australasian Affiliation of Herpetological Societies, P O Box R307 Royal Exchange Sydney New South Wales 2001.

212 herpetological groups in regional parts of Australia, the Frogweeks and Frogdays (see below) continue to raise posting of significant information on the Internet, and the community awareness. Resources such as Frogfacts sheets, establishment of the WWF Frogs! Project funded by Rio Tinto retailed tape recordings of frog calls (at both regional and Mining Co.. I focus on issues that I believe are important to state levels), and information books and booklets have consider for maximising community involvement. sustained strong community interest.

I have relied in part on my own direct experience with and Public displays of wide scope and appeal have been organised knowledge of several types of community organisations and run by frog and tadpole study groups in most states. namely frog and tadpole study groups (FATS groups), These have including public displays at shows, fairs, pet expos amateur herpetological societies, landcare groups and rural and open days. and pastoral landholder bodies. I have also drawn on the experiences and views of colleagues and informants many of Frogweeks and Frogdays whom are representatives of amateur herpetologists and Frogweeks and Frogdays are now fixtures in calendars at the landholders (see Acknowledgments). Summarising such views national and state level.The 1993 New South Wales Frog Week and experiences I have strived to write as a representative. (in the first week of November) was the first ever Frogweek in Problems that obstruct high community involvement in the world.The event was placed on the official state calendar by herpetological conservation activities. need honest and the then recently formed Frog and Tadpole Study (FATS) Group constructive regular checking if solutions are to be found and of the Australian Herpetological Society.This has since become successfully applied. the independent FATS Group of NSW Inc. In 1994 a national Frogweek was officially listed with the same INVOLVEMENTS TO DATE BY November time slot to get a week’s national “snapshot” akin COMMUNITY GROUPS to Waterweek’s snapshots which have strong public appeal. Most participating groups and states found it worked well to The community groups raise awareness. However areas that are usually without much The sustained high public profile of threatened frogs and their rain at this time (e.g. inland and South and southern Western enthusiastic and entertaining advocates have captivated world Australia) had limited success with frog survey-type activities. attention. Australia, Canada, Mexico, the USA and others all The uniform national week was subsequently replaced with have amphibian survey programs that involve volunteers. In state and regional time frames of one day to a week during Australia, much of the community involvement has come seasons with higher rainfall. Some groups extended their tape from the voluntary efforts of non-professionals and a small recording surveying well beyond the Frogweek time frame number of paid wildlife researchers (professionals). Some of and this gave improved species coverage. the latter have also contributed substantially with information, Frogweeks and frogdays are ideal to focus community publications and frog call recordings. Many frog and tadpole attention and weather permitting, survey activities by the study groups (which includes the amateur herpetological community can be a prominent feature of these events. societies) are organised and run by strongly committed To be more useful, the survey and monitoring activities by individuals.Their frog work has attracted good levels of public the community need to be spread throughout the frog participation and they are active in all states, and more are activity calendar for each region and state (see Australian forming at regional levels. Frog Count below). There are several other major types of community groups involved with frogs.These include landcare groups, conservation Survey and monitoring activities groups, the Australian Trust for Conservation Volunteers, science Surveying is the initial baseline work to establish the species teachers at secondary and primary level, and recovery teams present and at least to obtain an index of relative abundance. organised by Threatened Species Networks. Many of these Monitoring is repeating survey procedures at a site. groups include volunteers, contributors with professional skills, and experts (or at least access to them). Frog and tadpole study groups, in particular the strongly motivated members, have also become involved in frog Awareness raising and education survey and monitoring work both in an assisting capacity to researchers as well as in their own right with independent Secondary school classes were amongst the first large-scale funding. Of the latter, the most significant to date has been community involvements in frogwatch-type programs that the survey by the FATS Group of NSW (Inc) of the were developed in conjunction with school curricula. Some threatened frogs of New South Wales which was partly frog and tadpole study groups in some states have provided funded by the Australian Heritage Commission (Ehmann significant resources for school talks, presentations, and 1997). Professional survey and monitoring protocols and wetlands habitat projects. forms were developed for Australian frogs (Adair et al. 1994).

FROG CENSUS in South Australia and Frogwatch in Victoria The Australian Frog Count (see below) has evolved since (the latter was closed in 1995) are two statewide survey and 1993 with limited funding yet surprisingly good public monitoring programs funded by state environmental agencies. involvement (see Ehmann 1997).The Environmental They have raised widespread awareness and met with Protection Authority in South Australia adapted this model exceptionally high community support and involvement, to run the much better funded FROG CENSUS for that particularly in schools. Similar programs in other states have state’s community. run without agency funding- some for up to ten years.

213 Captive husbandry and breeding technique described by Clark et al. (1995) and used by the recovery team for the Eastern Barred Bandicoot in Victoria. Enthusiastic and committed members of all frog groups and many non-professionals take a strong interest in captive Most summarised issues are severe and may not occur or be husbandry and breeding. Many have had pleasing successes apparent in all projects involving the wider community. None- with some species and this has added to the body of the-less projects need to be designed and run to eliminate knowledge and experience. Some individuals are very keen to these issues. contribute their knowledge and experience to captive and ex-situ breeding programs for threatened species. It is The essence of satisfaction is when everyone is appreciated important that they be involved to maximise the chances of with positive regard and action. Each participant needs to success for these programs. Such contributions have not been know that the others involved believe in her/his ability to encouraged by some, information being ignored, trivialised or make a significant and meaningful contribution. It is about all- taken without due acknowledgment. inclusive attitudes, all-inclusive policies and an all-inclusive politic centred on the egalitarian versus elitist approach to In earlier years some individuals and groups were involved in the sharing of power and responsibility. It means human providing tadpoles to well-meaning people who wanted frogs equality and connection rather than hierarchy and isolation in their garden ponds and property dams. In some cases this (Sargent 1997). resulted in dispersals of tadpoles for hundreds of kilometres and a few species were sent beyond their natural ranges. Community ownership of an issue happens when the Such dispersal has not been done or condoned by any of the community itself speaks in accepting responsibility for the FATS groups for some time, but it may continue as it has for issue. Unrepresentative spokespersons for community groups decades as a private uninformed and unregulated activity, just (especially professionals and experts) are seen by the active as fishes are commonly translocated.The biological risks of community as elites dominating and controlling an issue.This unnatural translocations are unacceptable and this is now can discourage community involvement and ownership or, well-established and accepted by FATS groups. worse still, result in community backlash (Sargent 1997).

Rehabilitation projects Professionals and experts may ask “Why is the burden so heavy on us?”The answer, largely, is that as employed wildlife workers Many community groups and individuals have become with access to most of the resources, including the documented involved in frog habitat rehabilitation projects and stormwater wetlands are ideal for this. Planting aquatic and fringe information and technology, they are in positions of power.They vegetation, providing other shelter materials, and ensuring can therefore most easily affect the balance between resistance suitable water depths and bottom characteristics for breeding and acceptance (Sargent 1997).To do this professionals and and tadpoles have been the main activities. Frogs eventually experts obviously need to be adequately resourced in all facets find their way to most of these sites naturally and therefore of their pivotal involvement. Such resources are scarce in the translocations are not recommended or necessary. current economic climate in Australia.

Other activities Developing new directions for involvement The Restoring Australian Native Amphibians (RANA) group The individuals in participant groups identified in Table 1 need based in Queensland has been a national leader amongst to be involved to levels which they themselves set in community groups in raising money for declining frogs consultation with other participants.These levels are often research in Queensland. higher than some professionals are comfortable with.This discomfort may derive from the professional’s scientific The FATS Group of NSW (Inc) has been operating a Frogwatch training that emphasises the reduction of variables. Adverse Helpline for the past three years to respond to inquiries from experience and past unfulfilled expectations may also be schools and the public about Frogwatch and frogs.There are factors.There is a correlation between fearing, which results currently about 30 calls per month to this service. in a closing down of communication, and getting the worst outcome — expect the worst and it will happen.When this The essentials of community involvement is starting to happen it is the right time to rethink and put The public’s support and preparedness to get involved has more effort into resource development. An actively involved not yet been fully realised.This section examines how this community is one of the most valuable resources available. could be done.Three elements are discussed: 1) ensuring that Some recent approaches to organisational and fieldwork all participants are satisfied, 2) developing new directions for design that actively involve the community have proven very involvement, and 3) some survey, monitoring, recovery and effective.The use of hierarchical (classist) structures of rehabilitation projects that can be effectively implemented. academia, public service and politics does not work as well as Ensuring that all participants are satisfied adherents would like in harnessing voluntary community participation. Prototyping (Clark et al. 1995), active adaptive This often means overcoming some resistance, opinion, and/or management (Parma et al. 1998) and community focussed impression that some members of each participant group may wildlife management programs (Raven 1995; Ehmann and hold which gets in the way of whole-hearted commitment. Tynan 1997; the many Landcare projects in Australia) are Unless all the participants are “on side” and fully committed examples of approaches that work better. the arrest and reversal of frog declines will not happen as fast as it could.This is an essential feature of the Prototyping

214 TABLE 1: Participant groups and their potential negative issues that may compromise project outcomes

Participant group Resistance, opinion, and/or impression that may be held by some members. What is needed?

Landholders whose activities and land use is affected Will get interference with their usage and land management practices.What’s in it for us? Involvement and liaison from the onset. Volunteers who get very involved in ongoing awareness Will not get reasonable acknowledgment for their input. raising, monitoring and rehabilitation activities Give full recognition. Use each one’s name whenever possible in publicity. Let them speak publicly whenever possible. Public who send information, tapes, donations Never hear again. Give feedback, send outcomes report. Professionals who are employed to conduct monitoring, Having to deal with and win over all those participants, especially education and management actions. Usually poorly resourced. the “sticky” ones. See and treat them all as part of the whole valuable and equally concerned community.Talk over “What gets in the way of seeing and treating them all in this way” with someone skilled in listening well. Prune program to fit resources or increase resources. Non-professionals who become heavily involved in Will not get reasonable acknowledgment for their input and organising, education, resource preparation, fund raising, involvement.Will be out of pocket. captive breeding, rehabilitation (i.e. in addition to Include in as many facets of work as possible. Provide at least volunteer group above) out-of-pocket costs. Most sumarised issues are severe and may not occur or be apparent in all projects involving the wider community. None-the-less projects need to be designed and run to eliminate these issues.

It is very important that community members become Some frog survey, monitoring, recovery and involved in public speaking about projects. A well-placed and rehabilitation projects that could be implemented savvy “people’s” statement or call for community help can at the suburban, regional, state, and national level. result in high community involvement. Some examples of possible frog projects with the community and frog groups are: Existing community groups who may become involved in frogs programs include service organisations (e.g. Lions, Rotary 1. At the suburban level: adopt a site or species.This may be etc), the Guiding and Scouting movement, four wheel drive of interest to local government or schools.“Adopt a Swamp” or “Adopt a Wetland” will foster ownership of a clubs and even motor cycle (bikies) groups.While all of these local issue or public site.This is very important for species have their own specific agendas, they do have environmental and site rehabilitation. interests and concerns and can therefore involve themselves 2. At the regional level: adopt a river or a species.This may suit in projects set within a workable time frame. soil and catchment boards, regional planning authorities, local The liaison required to get effective community involvement government and private industry. Starting and supporting a “Friends of the Bell Frogs Group”, or “Friends of Croaky can be time consuming and needs to be allowed for in Creek” can help meet community expectations and planning and budgeting. It is also important to set goals and obligations. completion target dates on community projects.Without 3. At the state level: undertake survey/monitoring, captive them support will fade sooner rather than later. A project can (ex-situ) breeding, and rehabilitation. Set up a phone be extended with new targets if support continues or can be hotline to report sick and dying frogs. Any of these may be further recruited. run by fauna and environmental agencies and may easily attract funding from private industry. Those participants who are not paid for their involvement need and deserve support when they are out-of-pocket. FROG CENSUS is an existing successful program in South Some help with fuel and food costs goes a long way.They Australia.There is good potential for research and cannot be overloaded with duties they cannot sustain on an development of captive breeding programs that involve the ongoing basis along with their own life and work more experienced members of the many FATS groups. Community-based projects could adopt one to three commitments.They need the fun factor kept to the fore, and threatened species that are a focus in rehabilitation many dislike elitism, academic competition, pecking order and projects in large parts of the state.The dead and dying scientific “correct speak”. Many volunteers have an acute frogs telephone hotline (DaD Frogs Hotline?) can advise awareness of these factors. Ignoring any of these can result in finders what to do and provide direct contact details for diminished community involvement. researchers. A contact and forwarding network for sick and dying frogs would fit with several of the Frog Action Plan (Tyler 1997) recommendations.

215 4. At the national level: run projects of national significance method exactly every time they do the surveys. Otherwise and service and guide previous levels.There is a gaping the results are not comparable and valuable data is hole here that the federal agencies could fill, resources lost.”(NAAMP,the words in [ ] are my additions). So it comes permitting. An organisation called Frogs Australia similar to down to some training and survey/monitoring standards Birds Australia, has great potential. It seems an excellent which can be achieved, even if this may not have happened in way to set up,guide and optimise community involvement earlier attempts at wide-ranging surveys such as Victoria’s at all levels. now discontinued Frogwatch.

Key community projects for Frogs Australia could include: Tape recordings (hard evidence) are made by volunteers at • a national atlas or survey (and possibly a monitoring defined sites.The tapes are sent out to participants who extension) that could be called the Australian Frog Count record locality, date, time and collector contact details on the (AFC: see below); tape as a voice over. Additionally or alternatively the recorder • the monitoring of one or more widespread and sends a photocopy of a map (e.g. topographic, street threatened species (e.g. Pseudophryne bibroni, members of directory) with locations marked, time and date. A corporate the Litoria aurea species group) on a national basis or as sponsor may provide the tapes. focal species to get AFC data; The tapes are centrally processed i.e. they are listened to, • a DaD Frogs Hotline (see above); and scored for species and relative abundance, permanently • ensuring greater penetration of more specific and stored on CD-ROM for later rapid access and checking of regionally tailored frog habitat advice into land uncertain or important calls, and a reply and report is sent to rehabilitation literature and practice.To date a few stream the volunteer. A database is centrally maintained with ready and waterbody rehabilitation manuals and guides for access for participants to non-threatened species data. landcare groups have included mostly very general Various elements of this system have already been information about frogs’ needs.There is a need for better successfully carried out by the FATS Group of NSW, the information and this work needs to be resourced. South Australian Environmental Protection Authority FROG CENSUS program, the Frogwatch programs that ran in New South Wales and Victoria, and the Streamwatch and THE POTENTIAL AUSTRALIAN FROG Waterwatch programs. COUNT (AFC) IN MORE DETAIL Locality data associated with calling species can be added to Survey (atlassing) and monitoring programs are valuable and a GIS database for further analysis (e.g. distribution, habitat, feasible projects in which the public can get involved.They and disturbance), date and time data can be correlated with can provide scientific data in quantities unachievable by the weather data to determine calling/breeding activity (e.g. professional and expert surveys and at a fraction of the peaks, duration). As with any data gathering system there will resource costs (NAAMP). Much of the data will be from be errors and inaccuracies, but these can be controlled, common species and those that are not formally listed as minimised and allowed for in analyses.Volunteers may need threatened. However there will also be some really valuable some initial training which can be provided by regional species and locality records that would not be found without coordinators, via the Internet, and as a televised documentary community involvement. Furthermore this decade’s common made to recruit community involvement.Video copies of such species may become the next declining species, as we have material could be a major training resource. seen with the Bell Frogs (Litoria aurea species group). Contrary to some negative views about atlassing, information A nationally coordinated AFC should have an initial survey collected on the scale envisaged will provide a lot of useful phase that lasts for three years with a second phase taking on and cost-effective data. a monitoring role.This might start in the second year of the survey but could continue for ten or more years. The main objections to atlassing involving community volunteers come from those who raise concerns about, 1) It is possible to develop a “Frog Activity Forecast” system the comparability of subsequent monitoring activities and, 2) (providing FAF’s) that is based on Australian bio-regions, the reliability of the collected survey data. Both problems can known and suspected frog species, known calling/breeding be overcome.The former is also amongst the main problems seasons, and known weather requirements for calling. Regional associated with such data collected by professionals in the FAF’s can be broadcast with weather bulletins just like fog and past and even in the present.This problem is common to frost alerts. Initially these would be widely seen and/or heard many studies and is addressed in the next paragraph.Tape “hooks” in the community recruitment phase.“Frogs, fosts and recordings that are made at defined sites provide absolutely fogs” would join the blooper list! Regional FAF’s on television reliable data for the presence of most species of calling frogs. might be frog icons (with microphones and pulsating vocal Tapes may also provide relative abundance data. sacs!) on the state weather map to indicate something like “tomorrow night will be a good night to record in your Comparability can be achieved by using the same (few) region”. Species for which calling information is not available survey/monitoring personnel for all surveys, but in a are assumed to call at any wet time. continent with so few professionals and so many diverse species this seems impractical.“The methods for conducting surveys are many and must be chosen carefully based on the uniqueness of each monitoring effort [and species]. It is critical however that once the correct and most appropriate method is decided upon that the field worker follow that

216 Visually impaired people, many of whom have acutely The Australian community is poised to go much further with developed aural capabilities and keyboard skills, can process frogs. A national organisation like Frogs Australia is a logical the tape recordings. Organisations that service the visually next step that cannot be delayed much longer. Several impaired community have access to funding for appropriate projects involving the wider community and the non- training, provided that it leads to employment. Electronic professional frog and tadpole study groups have already been analysis could also be used, but there are several key steps in successfully trialed at regional and state levels.These can be tape decoding and information storage that could not be upgraded and improved to run nationally over an initial easily automated. three-year period. An increase in voluntary community effort is unlikely without further resources support and this needs CURRENTLY AVAILABLE RESOURCES to happen soon to maintain the growing momentum. A list of the resources available for Frogwatch and Frogweek ACKNOWLEDGMENTS activities in 1994 was widely distributed for free to over 90 frog study groups, researchers and Frogwatch contacts I thank George Appleby, John Barker, Brian Barnett, Elaine throughout Australia (Ehmann 1994). Baskett, Alistair Bax, Gary Beardsall, Darryl and Sharon Bell, Brian Bush, Peter Goonan, Shane Gow, Marc Hero, Marty Tape recordings of frog calls are available in all states, and a Hilton, Michael Mahony, Alana Manning, Geoff Mills, Rick few commercially produced compact discs of calls are also Nattrass, Mary Oldfield, Stan Orchard, Deborah Pergolotti, available for some regions (some commercially available ones Jacquie Recsei, Dale Roberts, Martyn Robinson,Tony are Barker and Grigg 1973, Gould League of South Australia, Robinson, Alma Serle, Phil Souter, Gerry Swan, Karen Thumm, Littlejohn 1990, Queensland Museum 1990, Stewart 1991). Lothar Voigt, Merinda Voigt, Steve Walker, Arthur White, Most tape recordings are available as retail items while the Wayne Winter and Bruce Worth for their generous and rest are made available on request and the payment of a helpful discussions, comments and input. I thank WWF for cost-covering fee (Ehmann 1994).The majority of Australia’s funding my airfare to contribute to and participate in the species have been recorded. Workshop.Thanks also to Stan Orchard and another Printed information including posters, pamphlets, books and anonymous person who provided reviews that have booklets are available.There are several posters of frogs, some improved the paper. are region-based (e.g. Murray-Darling), others are state-based, others are select Australian frogs (e.g. Queensland’s missing REFERENCES frogs). Pamphlets include titles like Keeping Green Tree Frogs, Adair, A., Ehmann, H. and Stace, J., (1994) Frog survey data Keeping frogs in your garden, Establishing frog habitats on your form and guidelines for completing form.Version 1.8.94. A property, Rainforest frogs, and Green and Golden Bell Frogs concise data form and protocol for Australian frog survey (Frogfacts, various dates). Many other pamphlets and leaflets work. Available from the FATS Group (NSW) Inc, PO Box have been produced (mostly at regional levels, with a few A2405, Sydney South NSW 2001. state and national ones) to raise public awareness and to aid in identification. Many excellent published books are Backhouse, G. N. and Clark,T.W., (1995) Case studies and commercially available that provide information, identification, policy initiatives in endangered species recovery in and student activities. Some booklets (which have been Australia: recommendations. Pp 110-116 in People and published by FATS groups and others) are also available Nature Conservation: perspectives on private land use including data forms and a protocol for Australian frog survey and endangered species recovery.Transactions of the work (Adair et al. 1994), how to raise native frogs (Bax 1993) Royal Zoological Society of New South Wales. and a waterproof guide to stream-dwelling frogs of the Wet Barker J., and Grigg G., (1973) Frog calls of south-eastern Tropics rainforest (Hero and Fickling 1994). Australia. Most calls from eastern New South Wales, Victoria and south-eastern Queensland, available from Within the large amount of frog information on the Internet Australian Museum Bookshop, Sydney). there is Australian material that is very useful for supporting school and community involvements. Frogwatch and survey Bax, A., (1993) Raising native frogs. A booklet with guidelines information is available at the NSW Streamwatch site. on raising tadpoles, metamorphs, food sources, frog shelters etc especially in coastal Queensland. Available from RANA, 42 Poinsettia Street, Inala QLD 4077. THE FUTURE OF COMMUNITY Clark,T.W., Backhouse, G. N. and Reading, R. P., (1995) INVOLVEMENT Prototyping in endangered species recovery programs: the While the wider community does not distinguish between Eastern Barred Bandicoot experience. Pp 50-62 in People threatened and “non-threatened” frogs they can contribute and Nature Conservation: perspectives on private land significantly to the recovery of threatened species.There is use and endangered species recovery.Transactions of the great potential for community involvement in threatened frog Royal Zoological Society of New South Wales. awareness raising, general and threatened species survey and Ehmann, H., (1994) 1994 Frogweek resources and the monitoring, threatened species captive breeding programs, Australian Frog Count (AFC). Frog and Tadpole Study frog habitat rehabilitation work and fundraising. Group of NSW Inc PO Box A2405 Sydney South NSW 2000.

217 Ehmann, H., (1997) (ed.) Threatened Frogs of New South Littlejohn, M., (1990) Calls of Victorian frogs. All calls of Wales: habitats, status and conservation. Published by Frog species in Victoria except Litoria spenceri. Available from and Tadpole Study Group of NSW Inc., PO Box A2405, The Herp Shop, 16 Suspension St Ardeer,VIC 3022. Sydney South 2000. NAAMP (North American Amphibian Monitoring Program). Ehmann, H. and Tynan, R., (1997) Wildlife management Internet site at http://www.im.nbs.gov/amphibs.html manual: the Gawler Ranges and Kingoonya Soil Parma, A., Amarasekare, P.,Mangel, M., Moore, J., Murdoch,W. Conservation Districts. A resource handbook. Pastoral W., Noonburg, E., Pascual, M. A., Possingham, H. P., Shea, K., Management Program, Department of Environment and Wilcox,W. and Yu, D. (1998).What can adaptive Natural Resources South Australian, and the Save the Bush management do for our fish, food, forests and biodiversity. Program. 186 pp and wall poster. Integrative Biology 1:16-26. Environment Canada, (undated) Backyard frog surveys. Queensland Museum, (1990) Songs of the Humpback and Canadian Wildlife Service, P O B5050, Burlington, frog calls of the Brisbane environs.The frog calls are Ontario. L7R 4A6 Canada. grouped into habitats and seasons, with an introduction by Environment Canada, (1995) Amphibian road call counts Glen Ingram, available from the Queensland Museum participant manual. Canadian Wildlife Service, P O B5050, Shop, Brisbane. Burlington, Ontario. L7R 4A6 Canada. Raven, L., (1995) Liaising with landholders. Pp 209-210 in Frogfacts, (five to now, more in progress, various dates and People and Nature Conservation: perspectives on private authors) Four page pamphlets available free by sending a land use and endangered species recovery.Transactions of stamped, self-addressed DL size envelope and a cost- the Royal Zoological Society of New South Wales. covering donation to the FATS Group (NSW) Inc, PO Sargent, M. J., (1997) A new sociology for Australians. Box A2405, Sydney South NSW 2001. Longmans, Sydney. Gould League of South Australia, (undated) has separate tape Stewart, D., (1991) Frog calls of north-east NSW.Tape recordings of frog calls for eight regions of South Australia recording of most calls from this species-rich area of New (to make tape use easier in schools). Available from 63 South Wales, also useful in south-eastern Queensland, Cungena Ave Park Holme SA 5043. Phone and fax: (08) available from David Stewart phone (02 6684 0127). 8277 2851. Tyler, M. J., (1997) The Action Plan for Australian Frogs. Hero, J-M. and Fickling, S., (1994) A guide to stream-dwelling Wildlife Australia. Endangered Species Program. From the frogs of the Wet Tropics rainforest. Available from Botanical Bookshop, PO Box351, Jamison ACT 2614.Or Queensland Book and Map Supplies, 37 Tully Street, South on the web Townsville QLD 4810. http://www.biodiversity.environment.gov.au/plants/threaten/ Lindemayer, D. and Possingham, H., (1994) The risk of frogap.htm extinction: ranking management options for Leadbeater’s Possum using population viability analysis. Centre for Resource and Environmental Research, Australian National University, Canberra.

218 Applications of assisted reproductive technologies (ART) to endangered anuran amphibians

John Clulow, Michael Mahony, Robert Browne, Melissa Pomering and Andrew Clark*

ABSTRACT and embryos, may not be available in the short to medium term.As an alternative to cryopreserving Attempts to conserve critically endangered whole embryos, cryopreservation of individual cells amphibian species by determining and addressing from early stages of amphibian embryos, followed the causes of decline in the field have failed.Assisted by reconstitution of complete organisms by cloning, Reproductive Technologies (ART) offer many has a high probability of success, and should be possibilities to assist the conservation of amphibian available in the future. In addition, the success of species, and, at least in the short term, in amphibian sperm cryopreservation means that combination with captive breeding, may provide the complete individuals may be reconstituted in the only means of preserving some species. Some ART future by androgenesis, a procedure that already has procedures have been available for some time (i.e. been shown to be achievable in amphibians. induced ovulation and sperm release, in vitro fertilisation (IVF), androgenesis and cloning), while others such as amphibian sperm cryopreservation have been developed within the last year.

Other procedures, such as the cryopreservation of eggs and embryos are not yet available, and based on problems with cryopreservation of fish oocytes

* Department of Biological Sciences, University of Newcastle, NSW 2308, Australia.

219 INTRODUCTION material in a form that is retrievable as living amphibians, and to improve the efficiency of captive-breeding programmes. There is now a strong consensus amongst many The discussion in this review is confined principally to a herpetologists and conservation biologists that the reported consideration of anuran amphibians, as there are no native decline and extinction of amphibian species in the last two urodeles or caecilians in Australia.Table 1 presents a summary decades (Mahony 1996; Pounds et al. 1997; other chapters of the ART procedures discussed. this volume) is a real phenomenon, rather than an apparent one with alternative explanations such as cyclical changes in population densities. Nevertheless, in spite of this consensus, THE REPRODUCTIVE SYSTEM OF identification of the causes of decline have eluded ANURANS investigators, and even more importantly, from the With few exceptions, anuran amphibians (frogs and toads) perspective of conservation, no effective means of reproduce by external fertilisation, generally followed by ameliorating threatening processes in the field have been development of the fertilised egg and larva (again, there are identified. Unfortunately, failure to identify causes or propose exceptions) in an environment external to the body of a effective solutions has meant that biologists have been parent.This greatly simplifies the application of assisted confined to a process of monitoring species into extinction. reproduction (AR), because it reduces the number of steps in While these studies are important, and should continue, the the process to a minimum, avoiding complications of AR in time has long past where alternative strategies should have higher vertebrates (mammals) where embryos must be been applied to limit the irreplaceable loss of amphibian reimplanted into an hormonally synchronized female to biological and genetic diversity. complete development in utero. Essentially, it means that AR for anurans may be as simple as combining motile, Fortunately, there are strategies which may be adopted to mature sperm with eggs in a fertilisation dish. Obviously, there limit these losses.These may broadly be classified as Captive will be differences between species in the ease with which Breeding and Assisted Reproductive Technologies (ART).This sperm and oocytes may be obtained, and later, in the paper is concerned with the second of these approaches (the requirements of the developing larva and metamorph, but scope of captive breeding is too large to be considered in the basic simplicity of the reproductive process is in favour of this paper, and is already the subject of a widely-dispersed the application of ART. amateur and professional literature). It considers existing and rapidly developing reproductive technologies which will play It is worth noting some basic features of the male and female an important role in the management of critically-endangered reproductive systems of anurans, as these have important amphibian species and populations in the future, provided bearings on how ART for anuran amphibians will be applied. that conservation managers avail themselves of the In males, the testis is both the organ of sperm generation opportunity to use it. ART will be most effective where they (the site of spermatogenesis), and the organ of sperm are integrated into captive and field-based conservation storage.This is quite different from higher vertebrates where programmes.The uses of ART will include the production of these processes are separated and sperm are stored in ducts stock for maintaining captive-bred populations (held as external to the testis. In higher vertebrates, sperm storage “insurance” against extinction of wild populations) and for re- outside the testis simplifies the process of sperm collection introduction programmes where extinct populations may be for AR, as sperm from these storage ducts may be collected re-established. As well, ART may be used in programmes to by electroejaculation, manual stimulation or artificial vaginae. introduce additional genetic diversity to populations which With male frogs, however, sperm will have to be accessed have declined to low numbers with consequent inbreeding from the testis (see below). depression. Cryopreservation, as a central component of these technologies, may be of use to retrieve “extinct” species in the event of the loss of wild populations (through retrieval of sperm, eggs or embryos held in cryo-banks), store genetic

TABLE 1: Summary of ART procedures that may be used with endangered amphibian species.

Procedure Application Available Prospects

Hormonal induction of sperm release IVF and sperm cryopreservation Yes Hormonal induction of oocyte release IVF Yes (ovulation) Sperm cryopreservation use with standard IVF or with androgenesis Yes Oocyte cryopreservation use with cryopreserved sperm in IVF No long term Embryo cryopreservation (whole embryo) Cryopreservation of whole organism No long term Cryopreservation of individual stem use with cloning to reconstitute whole organisms No short term (early embryonic) cells Androgenesis reconstitution of whole organisms from sperm only Yes Cloning reconstitution of whole organisms from single Yes embryonic cells

220 The reproductive system of female frogs also has features We have used HCG successfully to collect large numbers of which both assist and complicate AR.The ovarian cycle of sperm from anurans as diverse as cane toads (Bufo marinus) amphibians differs from that of higher vertebrates in that and Australian hylids (including Litoria chloris and L. aurea). primary oogonia (the germ cells that give rise to oocytes) The yields have been large, with hundreds of millions of persist in the adult, producing a new generation of oocytes sperm collected from individual animals (Pomering, Clulow, with each reproductive cycle (Lofts 1974). In contrast, most Mahony unpubl. data), more than enough for IVF or other vertebrates have only one population of oocytes held cryopreservation. Hormones that have been recorded as in arrested meiosis, and no oogonia persist in adults.The inducing sperm release in anurans include gonadotrophins adaptive advantage for amphibians of continually renewing and gonadotrophin releasing hormones (Bettinger and the stock of oocytes is that large numbers of eggs can be O’Loughlin 1950; Rugh 1962; Easley et al. 1979), and produced throughout the reproductive life of the female.The catecholamines such as adrenalin and dopamine (Minucci et female reproductive cycle (which may be seasonal or al. 1993 ). Fresh amphibian pituitary extracts (we frequently opportunistic) is characterised by the recruitment of a new use cane toads as a source), are an excellent source of generation of oocytes, which are very small, but grow rapidly amphibian gonadotrophins (Rugh 1962; Easley et al. 1979). through the process of vitellogenesis (the laying down of yolk Successful IVF has been reported using sperm aspirated from in the cytoplasm), a process under the control of oestrogen. the cloaca (McKinnell et al. 1976), although this is not a Vitellogenesis is followed finally by ovulation, which is caused commonly-performed procedure. Nevertheless, the by progesterone released in response to a surge in luteinizing important point for performing IVF in endangered species is hormone (LH) (Lofts 1974). Mature oocytes, capable of being that non-invasive collection of sperm by hormonal induction fertilised, may be held in the ovaries for an extended period, is a technique that works and is available. until the LH/progesterone surge occurs in response to environmental cues such as rain. Of most significance for AR, INDUCTION OF OVULATION pre-ovulatory oocytes (i.e. oocytes retained in the ovaries) are not capable of being fertilised, and it is only after As long as female anurans are fully gravid i.e. vitellogenesis is ovulation ie the release of oocytes into the oviducts, that complete, ovulation can be induced with a number of agents fertilisation may proceed. Consequently (see below), including gonadotrophins and gonadotrophin releasing procedures for the induction of ovulation are essential if hormones. Freshly extracted amphibian pituitaries (Creaser oocytes are to be available for in vitro fertilisation (IVF) or and Gorbman 1939; Rugh 1962) remain an excellent source other AR purposes. of gonadotrophins (probably the best, since the gonadotrophins are amphibian rather than from other vertebrates) for the induction of ovulation, and pituitaries COLLECTION OF GAMETES AND from one species can be used to induce ovulation in other FERTILISATION IN VITRO (IVF) species. However, there may be some concern over the The collection of gametes (sperm and oocytes) is necessary transmission of pathogens with the injection of pituitary glands for two AR procedures: IVF and cryopreservation. For many or pituitary gland extracts.This may be an important purposes, IVF will be the end point of the AR process and in consideration with the management of rare species, applications such as some captive-breeding programmes, particularly in the current context where unidentified collection of fresh gametes for IVF from individual animals pathogens are suspected of causing species declines (the same which have failed to breed naturally may be all that is considerations will apply to the use of pituitary glands for the required.The issue of gamete cryopreservation is considered collection of sperm). Consequently, commercial preparations further below. of LH-RH or gonadotrophins may be preferable, although in our hands these are less successful with a range of species INDUCTION OF SPERM RELEASE than are whole pituitary glands. Steroid hormones (Wright 1961; Schuetz 1971) also may be capable of inducing The standard procedure to obtain sperm from frogs for IVF ovulation on their own or in conjunction with gonadotrophins. and other procedures in the laboratory is to kill the male, then remove and macerate the testes in a dilute (hypotonic) Where pituitary glands are used, there is some argument for solution to activate the sperm (Rugh 1962).This is obviously collecting them from male frogs where testes are fully unacceptable when working with rare or endangered animals. developed. Enlarged testes are an indicator of elevated There are, however, a number of neuro — and reproductive androgen levels which may only be maintained by hormones which are capable of inducing spermiation (release gonadotrophin support from an active pituitary gland. In of sperm from the testis). Sperm released from the testes contrast, female pituitary gonadotrophin levels may vary enter the kidneys and subsequently the cloaca with urine (Lofts considerably with seasonal factors and stages of the 1974). Consequently, following induction of sperm release from reproductive cycle. the testis, sperm for IVF or cryopreservation may be collected non-invasively from the cloaca by aspiration with a soft plastic cannula.This procedure formed the basis of the earliest tests for human pregnancy because one of the hormones which induces spermiation in anurans is human chorionic gonadotrophin (HCG), a hormone present in high quantities in the urine of pregnant humans (Bettinger and O’Loughlin 1950).

221 IVF AND SPERM INJECTION FIGURE 1: Androgenesis. 1a shows the process of normal While standard IVF will be the most common means of fertilisation where the maternal (black) pronucleus combines with the paternal (grey) pronucleus to form a single diploid obtaining fertilisation after collection of oocytes and sperm nucleus. 1b shows androgenesis where only the sperm pronucleus (the process will not be considered further here, as it is persists, replicates and forms an entirely paternally derived reasonably simple: see Rugh 1962;Watson 1977; Hollinger diploid nucleus. and Corton 1980; Browne et al. 1998), fertilisation by sperm injection is a technique that also should be considered as having potential applications. Conventional IVF for amphibians, where sperm and oocytes are mixed in a dish requires large numbers of actively motile sperm, with optimal concentrations in the order of 105 to 106 ml-1 (Hollinger and Corton 1980; Browne et al. 1998). Sperm injection, on the other hand, requires only one sperm per oocyte, and sperm of poor quality (low motility) may be used. Although sperm injection is now a well-established technique in human IVF, and is reported for some other mammalian species (Ng et al. 1996), it has not been reported for amphibians except in Bufo arenarum (Zelarayan et al. 1995) . Nevertheless, there is good reason to believe that it is a feasible technique. In the future, sperm injection may be of benefit in reconstructing species (see androgenesis below) where only small numbers of sperm are available, or where cryopreservation of sperm has been poor e.g. as might occur with testes or whole frogs frozen without cryoprotectants.

CRYOPRESERVATION OF SPERM, EGGS AND EMBRYOS AND THE RECONSTITUTION OF DIPLOID ORGANISMS Cryopreservation has the potential to enhance conservation efforts aimed at preventing the complete extinction of amphibian species through ex-situ conservation.These include the possibility of indefinite storage of functional, retrievable genomes (either haploid or diploid) which can be used to reconstitute fully-functional diploid individuals in the case of Work is continuing to further optimise the protocols, but the extinction in the wild i.e. cryopreservation as “insurance” results to date are very encouraging, indicating that these against extinction. As well, cryopreservation would almost protocols are probably applicable to a wide range of certainly be a more efficient and cost-effective way of amphibian species. maintaining species ex-situ than is the maintenance of live, Of course, as desirable as sperm cryopreservation is, it only captive colonies. Cryopreservation may have other preserves a haploid genome, whereas the goal is to be able advantages as well including reduced costs and problems to retrieve functional diploid organisms. In the future, this may with transportation of specimens, and benefits in quarantine be achieved by the use of frozen-thawed sperm in management. conjunction with frozen-thawed oocytes (see below). However, in the event of failure to develop successful SPERM CRYOPRESERVATION protocols for cryopreserving amphibian oocytes, other Until recently, there has been little interest in cryopreserving possibilities may exist for the reconstitution of diploid amphibian sperm, and there are no amphibian sperm individuals from sperm only (see below). cryopreservation protocols currently reported in the literature. However, this situation is undoubtedly about to RECONSTITUTION OF DIPLOID change, as awareness of the amphibian crisis motivates INDIVIDUALS FROM HAPLOID GENOMES reproductive and cryo-biologists to investigate the problem. BY ANDROGENESIS Our laboratory has recently developed highly successful protocols for the cryopreservation of anuran sperm (Browne Androgenesis (the generation of haploid or diploid individuals et al. 1998) using the cane toad, Bufo marinus, as a model from sperm) has not been achieved to date with higher organism. Recovery of motility and fertility in B. marinus has vertebrates i.e. mammals (McGrath and Solter 1984). been high (with fertilisation rates up to 80% using thawed However, it has been achieved with several species of fish sperm, and a similar percentage of sperm recovering post- (Parsons and Thorgaard 1985; Corley-Smith et al. 1996;Young thaw motility). As well (Browne et al. unpubl. data), we have et al. 1996), and some urodele amphibians (Gillespie and shown that the protocols also result in the recovery of Armstrong 1980). The standard procedure (Figure 1) substantial percentages of frozen sperm in several species involves fertilisation of enucleated or irradiated eggs (usually of Australian hylids including Litoria fallax, L. lesueurii, of the same species, but sibling species or congeners can also L. phyllochroa, L. peronii, L. subglandulosa, and L. verreauxii. be employed) in which the female pronucleus is destroyed.

222 The result is a haploid zygote, as only the male pronucleus FIGURE 2: Production of a clone from an early embryo divides.The haploid nucleus can be converted to a diploid (blastula).The blastula is disrupted so that individual cells (clear nucleus by heat or pressure shock at the stage of the first cell circles) may be removed and individually inserted into the division. This treatment prevents cell division, but not the cytoplasm of an oocyte in which the female pronucleus (dark circle) has been destroyed or removed.The inserted cell nucleus replication of chromosomes in the nucleus. Androgenesis, replicates in the oocyte cytoplasm to form a new embryo. thus, results in a single step in the production of a completely homozygous individual, from which inbred lines can be established, or from which outbred lines can be re-established by natural matings. As an additional possibility, androgenetic production of diploids also may be achievable directly by di- spermy e.g. injecting two sperm into an enucleated oocyte, or inducing polyspermy (Elinson 1986).

Androgenesis will be most useful where the male is the heterogametic sex, as androgenesis will produce both male and female offspring.The limitation of androgenesis where males are homogametic (as some amphibians are) is that only male offspring will be produced. Nevertheless, the potential applications of androgenesis may be quite substantial in conserving critically-endangered, or resurrecting “extinct”, amphibian species. Androgenesis may prove to be particularly useful if either of two circumstances prove to be the case: (1) where attempts to develop methods for cryopreserving oocytes or embryos are unsuccessful (2) where frozen testes, but no live animals, are available. In the latter case, it is possible by combining sperm injection (using sperm from frozen testes held by museums or other institutions) with androgenesis, that live animals may be reconstituted. Bearing in mind this possibility, it is particularly important, where species have become extinct in the wild, that every effort is made to locate, and, for the time being, maintain frozen testes or whole males, in an unthawed state, preferably at -80oC or liquid nitrogen temperatures.

CRYOPRESERVATION OF OOCYTES AND EMBRYOS Successful cryopreservation of unfertilised oocytes, zygotes or developing amphibian embryos has not been reported by any laboratory but is a very important goal of future research.The barrier existing around the yolk (the yolk syncytial layer), rather capacity to cryopreserve unfertilised oocytes would than to problems with freezing the blastoderm (embryonic complement the now successful cryopreservation of amphibian cells). Whether such specific permeability barriers exist in sperm, and in combination, these two procedures would be amphibians remains to be determined. sufficient to insure against the complete extinction of any amphibian species in the future.The capacity to freeze zygotes, As an alternative to freezing early embryos, later embryos or larvae would obviate the need to use either developmental stages may be better targets for sperm or oocyte cryopreservation, and in many respects, cryopreservation. Rapid advances in cryopreservation of would be the most desirable of all the cryopreservation complex embyronic structures is now being reported, and scenarios. However, the probability of successful larvae of Drosophila (Steponkus et al. 1990) as well as highly- cryopreservation of either amphibian oocytes or embryos in differentiated larvae of a number of marine organisms (Olive the short to medium term is not high. Preliminary attempts in and Wang 1997) have been successfully cryopreserved. our laboratory using both vitrification (rapid freezing) and slow Nevertheless, it is likely that such a goal may be medium to cooling have not been successful.The progress, or rather lack long-term for amphibian larvae. of progress, in freezing fish embryos (Hagedorn et al. 1997a,b) may be an indicator of the difficulties to be faced in this area with amphibians. In many respects, fish oocytes and embryos are good models for amphibians, as in both groups, the oocytes are large and yolky. Attempts to establish protocols for fish embryo cryopreservation over many years have met with little success. Recent evidence obtained from zebra fish embryos (Hagedorn et al. 1997a,b) suggests that the problem may be due to a water and cryoprotectant permeability

223 CRYOPRESERVATION OF BLASTOMERES little attention until recently. Support for future research to AND RECONSTITUTION BY CLONING complete the suite of techniques is an urgent necessity (particularly to develop protocols for cryopreservation of If cryopreservation of intact embryos proves to be embryos or embryonic stem cells), as is support for the unachievable, an alternative approach exists which has a high transfer of procedures which currently are applied principally prospect of success.This is the combination of the in the laboratory situation to use with endangered species. cryopreservation of individual blastomeres (i.e. early Nevertheless, even without the full range of desired embryonic cells) with the regeneration of totipotent diploid procedures, valuable procedures such as sperm embryos by cloning (Figure 2). Cloning of amphibians has cryopreservation are available now to managers of been a reality for many years (Briggs and King 1952; conservation programmes and recovery teams to help McKinnell 1978), predating the recently much publicised conserve species which may otherwise disappear. cloning of a mammal from an adult somatic cell (Wilmut et al. 1997). In amphibians, cloning is achieved by disrupting early embryos mechanically, collecting individual cells into pipettes, ACKNOWLEDGMENTS and injecting those cells into recipient oocytes whose nucleus Support for work on amphibian sperm and embryo has been destroyed by irradiation or manually removed (i.e. cryopreservation has been received from the Research enucleated).The donor cell provides the nucleus, while the Management Committee, University of Newcastle, and from enucleated recipient oocyte provides the cytoplasmic the Australian Research Council. environment necessary to support the development of the single injected cell into a developmentally competent embryo. REFERENCES Cells collected from early embryonic stages (e.g. early blastulae) retain the capacity to develop into a complete Briggs, R. and King,T.J., (1952) Transplantation of living nuclei organism i.e. are totipotent, whereas cells derived from later from blastula cells into enucleated frogs’ eggs. Proc. Nat. developmental stages are less likely to be totipotent (Gurdon Acad. Sci. U.S.A., 38: 455-463. 1962; DiBerardino et al. 1986). Bettinger, H.F. and O’Loughlin, I., (1950) The use of the male toad, Bufo marinus, for pregnancy tests. Med. J. Aust., July 8, There needs to be developed, in conjunction with a pp. 40-42. demonstration in the laboratory that many endangered Browne, R.K., Clulow, J., Mahony, M. and Clark, A.K., (1998) amphibian species are capable of being cloned, a procedure Successful recovery of motility and fertility of to freeze early amphibian embryonic cells. It is, in fact, highly cryopreserved cane toad (Bufo marinus) sperm: a likely that such cells would be capable of cryopreservation. comparison of the effect of concentration DMSO or Cells at the stage of embryonic development from which glycerol as cryoprotectants in a sucrose base. Cryobiology, they would be taken for cloning, are much reduced in size in 37: 339-345. comparison to oocytes, and would be frozen as individual cells rather than complete embryonic organisms. In this Corley-Smith, G.E., Lim C.J. and Brandhorst, B.P., (1996) respect, their response to cryopreservation would probably Production of androgenetic zebrafish (Danio rerio). be similar to many tissue culture cell lines, which are routinely Genetics, 142: 1265-1276. frozen and thawed (Morris 1995). Creaser, C.W. and Gorbman, A., (1939) Species specificity of the gonadotrophic factors in vertebrates. Quart. Rev. Biol., One limitation of cloning as a means of regenerating diploid 14: 311-331. individuals may be that nucleo-cytoplasmic hybrids between DiBerardino, M.A., Orr, H.O. and McKinnell, R.G., (1986) different species (i.e. hybrids produced where the recipient Feeding tadpoles cloned from Rana erythrocyte nuclei. enucleated oocyte is of a different species to the donor Proc. Natl. Acad. Sci. USA, 83: 8231-8234. nucleus) are less likely to succeed than same species nucleo- cytoplasmic transplants.This may be a problem for Easley, K.A., Culley, D.D., Horseman, N.D. and Penkala, J.E., resurrecting “extinct” amphibian species which exist only as (1979) Environmental influences on hormonally induced frozen blastomeres, because same species recipient oocytes spermiation of the bullfrog, Rana catesbeiana. J. Exp. Zool., would not be available for the cloning process. Nevertheless, 207: 407-416. there is some evidence that nucleo-cytoplasmic hybrids Elinson, R.P.,(1986) Fertilisation in amphibians: the ancestry of produced between sibling species within the same genus may the block to polyspermy. Int. Rev. Cytol., 101: 59-100. be viable (McKinnell 1978, p. 164). Gillespie, L.L. and Armstrong, J.B., (1980) Production of androgenetic diploid axolotls by suppression of first CONCLUSIONS cleavage. J. Exp. Zool., 213: 423-425. Gurdon T.J., (1962) The developmental capacity of nuclei It is clear that attempts to conserve critically-endangered taken from intestinal epithelium cells of feeding tadpoles. amphibian species by determining and addressing the causes J. Embryol. Exp. Morphol., 10: 622-640 of decline in the field have failed. Alternative, pro-active approaches are needed if further complete extinctions of Hagedorn, M., Hsu, E., Kleinhans, F.W. and Wildt, D.E., (1997a) species are to be avoided. Assisted reproductive technologies New approaches for studying the permeability of fish offer many possible tools to assist with ex-situ conservation, embryos: toward successful cryopreservation. Cryobiology, and even to assist in management of species in-situ. Some 34: 335-347. procedures have been available for some time e.g. induced ovulation and sperm release, IVF, and cloning while others such as amphibian sperm cryopreservation have received

224 Hagedorn, M., Kleinhans, F.W.,Wildt, D.E. and Rall,W.F., Olive, P.J.W. and Wang,W.B., (1997) Cryopreservation of (1997b) Chill sensitivity and cryoprotectant permeability Nereis virens (Polychaeta, Annelida) larvae: the mechanism of dechorionated zebrafish embryos, Brachydanio rerio. of cryopreseration of a differentiated metazoan. Cryobiology, 34: 251-263. Cyrobiology, 34: 284-294. Hollinger,T.G. and Corton, G.L., (1980) Artificial fertilisation of Parsons, J.E. and Thorgaard, G.H., (1985) Production of gametes from the South African clawed toad, Xenopus androgenetic diploid rainbow trout. J. Hered., 76: 177-181. laevis. Gamete Res., 3: 45-57 Pounds, J.A., Fogden, M.P.L.,Savage, J.M. and Gorman, G.C., Lofts, B., (1974) Reproduction. Pp 107-218 in Physiology of (1997) Test of null models for amphibian declines on a the Amphibia.Volume II. Ed. B. Lofts, Academic Press, New tropical mountain. Conserv. Biol., 11: 1307-1322. Yo r k . Rugh, R., (1962) Experimental Embryology. 3rd Edition. Mahony, M.J., (1996) The decline of the green and golden bell Burgess, Minneapolis, USA. frog (Litoria aurea) viewed in the context of declines and Schuetz, A.W., (1971) In vitro induction of ovulation and disappearances of other Australian frogs. Aust. Zool., oocyte maturation in Rana pipiens ovarian follicles: effects 30: 237-247. of steroidal and non-steroidal hormones. J. Exp. Zool., 178: McGrath, J. and Solter, D., (1984) Completion of mouse 377-385. embryogenesis requires both the maternal and paternal Steponkus, P.L.,Myers, S.P.,Lynch, D.V., Gardner, L., Bronshteyn, genomes. Cell, 37: 179-183. V., Leibo, S.P.,Rall,W.F., Pitt, R.E., Lin,T.T. and MacIntyre McKinnell, R.G., (1978) Cloning. Nuclear Transplantation in R.J., (1990) Cryopreservation of Drosophila melanogaster Amphibia. Univ. of Minnesota Press, Minneapolis, USA. embryos. Nature, 345: 170-172. McKinnell, R.G., Picciano, D.J. and Krieg, R.E., (1976) Watson, G.F., (1977) Artificial hybridization techniques for Fertilisation and development of frog eggs after repeated anuran amphibians. Aust. Zool., 19: 239-241. spermiation induced by human chorionic gonadotrophin. Wilmut, I., Schnieke, A.E., McWhir, J., Kind, A.J. and Campbell, Labor. Anim. Sci., 26: 932-935. A.H.S., (1997) Viable offspring derived from fetal and adult Morris, C.B., (1995) Cryopreservation of animal and human mammalian cells. Nature, 385: 810-813. cell lines in Methods in Molecular Biology,Volume 38: Cryopreservation and Freeze-Drying Protocols, Wright, P.A., (1961) Induction of ovulation in vitro in Rana pp.179-187. Eds J.G. Day & M.R. McLellan. pipiens with steroids. Gen. Comp. Endocrinol., 1: 20-23. Minucci, S., Fasano, S., D’Antonio, M. and Pierantoni, R., (1993) Young,W.A.,Wheeler, P.A.,Fields, R.D. and Thorgaard, Dopamine regulation of testicular activity in intact and G.H., (1996) DNA fingerprinting confirms isogenicity of hypophysectomized frogs, Rana esculenta. Experientia, androgenetically derived rainbow trout lines. J. Heredit., 49: 65-67. 87: 77-80 Ng, S.C., Bongso, A. and Trounson, A., (1996) Update on Zelarayan, L.I., Oterino, J. and Buhler, M.I., (1995) Spontaneous micromanipulation techniques for assisted conception. maturation in Bufo arenarum oocytes: follicle wall Current Opinion in Obstet. Gynecol., 8: 171-177. involvement, respiratory activity and seasonal influences. J. Exper. Zool., 272: 356-362.

225 Appendix 1: Checklist of Australian Frogs Action Plan for Australian Frogs (Tyler 1997)

HYLIDAE coplandi Copland’s Rock Frog Cyclorana cyclorhyncha Spotted-thighed Frog australis Giant Frog dahlii Dahl’s Aquatic Frog brevipes Short-footed Frog dentata Bleating Tree-frog cryptotis Hidden-ear Frog electrica Electric Tree-frog cultripes Knife-footed Frog eucnemis Fringed Tree-frog longipes Long-footed Frog ewingii Brown Tree-frog maculosus Daly Waters Frog fallax Eastern Dwarf Tree-frog maini Main’s Frog freycineti Freycinet’s Frog manya genimaculata New Guinea Tree Frog novaehollandiae gilleni Centralian Tree Frog platycephala Water-holding Frog gracilenta Dainty Green Tree-frog vagitus Wailing Frog inermis Peter’s Frog verrucosus infrafrenata White-lipped Tree-frog jervisiensis Litoria latopalmata Gunther’s Frog adelaidensis Slender Tree-frog lesueuri Lesueur’s Frog alboguttata littlejohni Littlejohn’s Frog aurea Green & Golden Bell Frog longirostris Long-snouted Frog bicolor Northern Dwarf Tree Frog lorica Armoured Frog booroolongensis Booroolong Frog meiriana Rockhole Frog brevipalmata Green-thighed Frog microbelos burrowsae Tasmanian Tree Frog moorei Moore’s Frog caerulea Green Tree Frog nannotis Torrent Tree Frog castanea (flavipunctata) Yellow-spotted Tree Frog nasuta Rocket Frog cavernicola Cave-dwelling Frog nigrofrenata Bridle Frog chloris Red-eyed Tree Frog nyakalensis Nyakala Frog citropa Blue Mountains Tree Frog olongburensis Olongburra Frog,Wallum Sedge Frog cooloolensis Cooloola Tree Frog pallida Pale Frog

226 paraewingi Victorian Frog Kyarranus pearsoniana Pearson’s Frog kundagungan Mountain Frog peroni Peron’s Tree-frog loveridgei Loveridge’s Frog personata Masked Rock Frog sphagnicolus Sphagnum Frog phyllochroa Leaf-green Tree Frog Lechriodus fletcheri Fletcher’s Frog piperata Peppered Tree Frog raniformis Southern Bell Frog Limnodynastes revelata convexiusculus Marbled Frog rheocola Creek Frog depressus Flat-headed Frog rothii Roth’s Tree Frog dorsalis Bullfrog rubella Red Tree Frog d. dumerili Four-bob Frog, Eastern Banjo Frog spenceri Spotted Tree Frog d. insularis Four-bob Frog, Eastern Banjo Frog splendida Magnificent Tree-frog d. fryi Four-bob Frog, Eastern Banjo Frog subglandulosa Glandular Frog d. grayi Four-bob Frog, Eastern Banjo Frog tornieri Tornier’s Frog d. variegatus Four-bob Frog, Eastern Banjo Frog tyleri Tyler’s Tree-frog fletcheri Marsh Frog verreauxii Verreaux’s Frog interioris v. alpina Alpine Tree Frog ornatus wotjulumensis Wotjulum Frog peroni Brown-striped Frog xanthomera Orange-thighed Frog salmini Salmin’s Striped Frog spenceri Spencer’s Frog Nyctimystes dayi Day’s Frog tasmaniensis Spotted Grass Frog LEPTODACTYLIDAE terraereginae Adelotus brevis Tusked Frog Megistolotis lignarius Carpenter Frog Arenophryne rotunda Sandhill Frog nichollsi Nicholl’s Toadlet Assa darlingtoni Marsupial Frog Mixophyes balbus Stuttering Frog Bryobatrachus nimbus Moss Frog fasciolatus Great Barred Frog Crinia fleayi Fleay’s Frog bilingua Bilingual Froglet iteratus Southern Barred Frog deserticola Desert Froglet schevilli georgiana Tschudi’s Froglet Myobatrachus gouldii Turtle Frog glauerti Glauert’s Froglet insignifera Sign-bearing Froglet Neobatrachus parinsignifera Eastern Sign-bearing Froglet albipes White-footed Frog pseudinsignifera False Western Froglet aquilonius remota Remote Froglet centralis Trilling Frog riparia Streambank Froglet fulvus signifera Common Froglet kunapalari sloanei Sloane’s Froglet pelobatoides subinsignifera Small Western Froglet pictus Painted Frog tasmaniensis Tasmanian Froglet sudelli Sudell’s Frog tinnula Tinkling Froglet sutor wilsmorei Goldfields Bullfrog Spicospina flammocaerulea Sunset Frog Notaden bennetti Holy Cross Toad Geocrinia melanoscaphus Northern Spadefoot Toad alba White-bellied Frog nichollsi Desert Spadefoot Toad laevis Smooth Frog weigeli Weigel’s Toad leai Lea’s Frog lutea Saffron Frog Paracrinia haswelli Haswell’s Froglet rosea Roseate Frog Philoria frosti Baw Baw Frog victoriana Eastern Smooth Frog vitellina Orange-bellied Frog Heleioporus albopunctatus Western Spotted Frog australiacus Giant Burrowing Frog barycragus Western Marsh Frog eyrei Moaning Frog inornatus Plains Frog psammophilus Sand Frog

227 Pseudophryne MICROHYLIDAE australis Red-crowned Toadlet Cophixalus bibronii Bibron’s Toadlet bombiens Buzzing Frog coriacea Red-backed Toadlet concinnus Elegant Frog corroboree Southern Corroboree Frog crepitans Rattling Frog covacevichae Magnificent Brood Frog exiguus Scanty Frog dendyi hosmeri Hosmer’s Frog douglasi Douglas’s Toadlet infacetus Inelegant Frog guentheri Gunther’s Toadlet mcdonaldi McDonald’s Frog major Large Toadlet monticola occidentalis Western Toadlet neglectus Neglected Frog pengilleyi Northern Corroboree Frog ornatus Ornate Frog raveni peninsularis Cape York Frog semimarmorata Southern Toadlet saxatilis Rock Frog Rheobatrachus Sphenophryne silus Gastric Brooding Frog delphe Northern Territory Frog vitellinus Stream Frog, Eungella fryi Fry’s Frog Gastric Brooding Frog gracilipes Slender Frog Taudactylus pluvialis Rain Frog acutirostris Sharp-nosed Torrent Frog obusta Robust Frog diurnus Southern Day Frog eungellensis Eungella Torrent Frog RANIDAE liemi Liem’s Frog Rana daemeli Water Frog pleione Kroombit Tinker Frog rheophilus Tinkling Frog Uperoleia altissima Alpine Toadlet arenicola Jabiru Toadlet aspera Derby Toadlet borealis Northern Toadlet capitulata Small-headed Toadlet crassa Fat Toadlet fusca Dusky Toadlet glandulosa Glandular Toadlet inundata Floodplain Toadlet laevigata Smooth Toadlet lithomoda Stonemason Toadlet littlejohni Littlejohn’s Toadlet marmorata Marbled Toadlet martini Martin’s Toadlet micromeles Tanami Toadlet mimula Mimic Toadlet minima Small Toadlet mjobergi Mjoberg’s Toadlet orientalis Alexandra Toadlet rugosa Wrinkled Toadlet russelli Russell’s Toadlet talpa Mole Toadlet trachyderma Blacksoil Toadlet tyleri Tyler’s Toadlet

228 Appendix 2: Threatened Frogs Endangered Species Protection Act 1992

Schedule 1, July 1999, Amphibians

Endangered Geocrinia alba White-bellied Frog Litoria lorica Armoured Mistfrog Litoria nannotis Litoria nyakalensis Mountain Mistfrog Litoria spenceri Spotted Tree Frog Nyctimystes dayi Lace-eyed Tree Frog Rheobatrachus silus Gastric-brooding Frog Rheobatrachus vitellinus Eungella Gastric-brooding Frog Spicospina flammocaerulea Sunset Frog Taudactylus acutirostris Sharp-snouted Day Frog Taudactylus diurnus Southern Day Frog Taudactylus eungellensis Eungella Day Frog Taudactylus rheophilus Tinkling Frog

Vulnerable Geocrinia vitellina Orange-bellied Frog Litoria aurea Green and Golden Bell Frog Philoria frosti Mount Baw Baw Frog

Presumed Extinct None listed

229 Appendix 3: Threatened Frogs ANZECC list of threatened Australian Fauna

Amphibians, May 1999 Vulnerable Geocrinia vitellina Yellow-bellied Frog Extinct Heleioporus australiacus Giant Burrowing Frog Rheobatrachus silus Gastric-brooding Frog Litoria aurea Green and Golden Bell Frog Rheobatrachus vitellinus Eungella Gastric-brooding Frog Litoria littlejohni Littlejohn’s Tree Frog Taudactylus acutirostris Sharp-snouted Day Frog Litoria olongburensis Wallum Sedge Frog Taudactylus diurnus Southern Day Frog Litoria piperata Peppered Tree Frog Litoria raniformis Southern Bell Frog Extinct in the wild Litoria verreauxii alpina Alpine Tree Frog None Listed Mixophyes balbus Stuttering Frog Critically Endangered Pseudophryne pengilleyi Northern Corroboree Frog Litoria castanea Yellow-spotted Tree Frog Pseudophryne covaceviche Magnificent Brood Frog Litoria spenceri Spotted Tree Frog Taudactylus pleione Kroombit Tinker Frog Mixophyes fleayi Fleay’s Frog Conservation Dependent Philoria frosti Baw Baw Frog None listed Pseudophryne corroboree Southern Corroboree Frog Taudactylus eungellensis Eungella Day Frog Taudactylus rheophilus Tinkling Frog

Endangered Geocrinia alba Creek Frog,White-bellied Frog Litoria lorica Armoured Mistfrog Litoria nannotis Waterfall Frog Litoria nyakalensis Mountain Mistfrog Litoria rheocola Common Mistfrog Mixophyes iteratus Southern Barred Frog Nyctimystes dayi Lace-eyed Frog Spicospina flammocaerulea Sunset Frog

230 Appendix 4: Recommendations from ‘The Action Plan for Australian Frogs (Tyler 1997)

These recommendations are not ranked. terrestrial predators such as cats and foxes, subtle climatic changes and perturbations, and impacts of global changes Summary of the Problem to air and water quality. 27 species (13%) of Australia’s frog fauna are threatened, and of these, 8 species may have disappeared altogether. An Two geographic zones in particular show widespread additional 14 species give cause for concern. For most of patterns of decline: these species, the causes of decline are not known or are — montane rainforest areas in Queensland, poorly understood. — alpine and upland areas in south-eastern Australia. • Recommendation 1: That high priority be given to • Recommendation 4: That attempts be made to research and management action to address frog declines. determine if there are causal factors common to these declines and if any of these operate together, or even Causes of Decline synergistically. Frogs are known to be highly sensitive to aquatic pollutants (Tyler 1989, Bidwell and Gorrie 1995). Conservation Status • Recommendation 2: That research be continued into This Action Plan identifies 14 species for which information the toxicity of pollutants, particularly herbicides and their is inadequate to assign a conservation status, although it dispersants. It is also recommended that low-toxicity is recognised that there may be cause for concern for surfactants be required, particularly for herbicide use near these species. water bodies and drainage lines. • Recommendation 5: That high priority be given to survey and research necessary to clarify the distribution, Many factors have been suggested as contributing to frog abundance and conservation status of insufficiently decline but there is currently little evidence to implicate any known species. particular factors. • Recommendation 3: That research and analysis be Captive Breeding undertaken to clarify the possible contributing role of Many species declines have been recent and rapid, with these factors, including data already available from apparent extinction of some species. Successful rearing and Australian and international studies. Examples include: husbandry techniques need to be developed so that holding pathogens, local water quality, impacts of introduced fish action can be taken in the event of a sudden decline of a (eg. Gambusia holbrooki, trout, carp), impacts of introduced species in the wild. Such captive populations may provide

231 essential reservoir populations for later reintroductions. • Recommendation 6: That priority be given to improve expertise in captive breeding and husbandry techniques. This should involve conservation agencies, researchers, zoos and amateur groups.

Public Involvement There is considerable interest amongst the general public in frog decline, and many amateur groups, such as the Frog and Tadpole Study Group (FATSG) in NSW and the Victorian Frog Group (VFG), are already contributing valuable information in terms of field surveys and biological information. • Recommendation 7: That public involvement, particularly through amateur associations, be encouraged and incorporated into conservation assessment and recovery planning.

Documenting and Describing the Frog Fauna Several frog species have been described only recently, demonstrating the paucity of knowledge of the Australian frog fauna. • Recommendation 8: That an atlas of Australian frogs be developed and maintained, using existing and incoming data.

National Working Group for Frog Conservation Although an Australian branch of the Declining Amphibian Populations Task Force (DAPTF) of the Species Survival Commission was established in 1992, this is not functioning effectively. • Recommendation 9: That a national working group for frog conservation be established to facilitate regular communication between frog experts. This group should link to the DAPTF as appropriate.

232 Appendix 5:The Conservation Status of Australian Frogs (Tyler 1997)

Family HYLIDAE Litoria aurea Green and Golden Bell Frog ENDANGERED Litoria castanea (L. flavipunctata) Yellow-spotted Tree Frog ENDANGERED Litoria lorica Armoured Frog ENDANGERED Litoria nannotis Torrent Tree Frog ENDANGERED Litoria nyakalensis Nyakala Frog ENDANGERED Litoria olongburensis Wallum Sedge Frog VULNERABLE Litoria piperata Peppered Tree Frog VULNERABLE Litoria raniformis Southern Bell Frog VULNERABLE Litoria rheocola Creek Frog ENDANGERED Litoria spenceri Spotted Tree Frog ENDANGERED Litoria verreauxii alpina Alpine Tree Frog VULNERABLE Nyctimystes dayi Day’s Frog ENDANGERED

233 Family LEPTODACTYLIDAE Spicospina flammocaerula Sunset Frog ENDANGERED Geocrinia alba White-bellied Frog ENDANGERED Geocrinia vitellina Orange-bellied Frog VULNERABLE Mixophyes balbus Stuttering Frog VULNERABLE Mixophyes fleayi Fleay’s Frog ENDANGERED Mixophyes iteratus Southern Barred Frog ENDANGERED Philoria frosti Baw Baw Frog ENDANGERED Pseudophryne corroboree Southern Corroboree Frog ENDANGERED Rheobatrachus silus Gastric Brooding Frog ENDANGERED Rheobatrachus vitellinus Eungella Gastric Brooding Frog ENDANGERED Taudactylus acutirostris Sharp-nosed Torrent Frog ENDANGERED Taudactylus diurnus Mount Glorious Torrent Frog ENDANGERED Taudactylus eungellensis Eungella Torrent Frog ENDANGERED Taudactylus pleione Kroombit Tinker Frog VULNERABLE Taudactylus rheophilus Tinkling Frog ENDANGERED

INSUFFICIENTLY KNOWN SPECIES THAT MAY BE OF CONCERN HYLIDAE LEPTODACTYLIDAE Litoria booroolongensis Crinia tinnula Booroolong Frog Tinkling Froglet Litoria brevipalmata Heleioporus australiacus Green-thighed Frog Giant Burrowing Frog Litoria freycineti Kyarranus sphagnicolus Freycinet’s Frog Sphagnum Frog Litoria pearsoniana Pseudophryne australis Pearson’s Frog Red-crowned Toadlet Litoria subglandulosa Pseudophryne bibronii Glandular Frog Bibron’s Toadlet Pseudophryne pengilleyi Northern Corroboree Frog Taudactylus liemi Liem’s Frog Uperoleia marmorata Marbled Toadlet Uperoleia orientalis Alexandra Toadlet

234 235 236