Declines and Disappearances of
AUSTRALIAN
frogs
Edited by Alastair Campbell Biodiversity Group Environment Australia 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: Litoria rheocola, Creek Frog Environment Australia Library Photo by: Keith McDonald Designed by: Di Walker Design, Canberra
Contents Foreword Preface The Gordian Knots of the International Declining Amphibian Populations Task Force (DAPTF) Stan Orchard
9
A Review of Declining Frogs in Northern Queensland 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 Species Present and the Conservation of Threatened Species 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 Habitat in the Sydney 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 Amphibians 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 Frog (Pseudophryne 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 New Zealand Environment and Conservation Council list of threatened Australian fauna
230
Appendix 4: Recommendations from ‘The Action Plan For Australian Frogs’
231
Appendix 5: The Conservation status of Australian Frogs, ‘The Action Plan for Australian Frogs’
233
4
Foreword Experts continue to reaffirm that a significant proportion of the world’s frog fauna is in decline. While frogs have been great survivors for many millions of years, habitat loss and degradation are imposing themselves at increasing levels. More alarmingly frogs in pristine environments are being affected. Even in protected rainforest habitats 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 these signs especially important.
Australia formed a National Threatened Frogs Working Group in August 1997 and held a national workshop on the issue in November 1997. This collection of papers identifies many of the issues confronting our amphibian biologists, managers and enthusiasts. I commend it to the reader.
Stephen Hunter 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 frog assemblages in the world. For many Australian frog species however the prognosis is grim. Dramatic population declines in some Australian frog species have been reported since the 1980s, some of the more serious crashes occurring in pristine habitats. Frustratingly the causal factors for many declines remain elusive.
The purpose of the Workshop was to:
In April 1997, Environment Australia published the Action Plan for Australian Frogs.This plan identified 27 Australian frog species at threat and a further 14 species that may be of concern but which were poorly understood. Recovery outlines, identifying those research and management actions required, were presented for the 27 species believed to be at most threat.
4. recommend priorities for national effort for research, management and community attention;
In May 1997, the National Threatened Frog Working Group recognised a need to bring together specialists in research, management and policy making to discuss their understanding of the continuing declines to our frog fauna and to prioritise future action for addressing the issue. As a result a two day ‘National Threatened Frog Workshop’ was held at the University of Canberra in November 1997 sponsored by Environment Australia, the NSW National Parks and Wildlife Service, the University of Canberra and the Worldwide Fund for Nature.The Workshop brought together some 80 people from throughout Australia and overseas involved in frog research, management and policy development and included representatives from government agencies, educational institutions and non-government organisations. 6
1. provide an overview of the current status of knowledge about species which are declining; 2. provide an overview of what’s known about the decline and how it is being tackled; 3. identify gaps in knowledge;
5. help set future priority actions for the Working Group; and 6. establish on-going communication links between all participants and the wider community. Following the Workshop the Working Group met and emphasised: the need to maintain a broad perspective and to continue monitoring and research on a range of possible causal factors; the need to support research into new areas — particularly disease as a potentially major factor influencing global declines; the need to support strategic research and to develop protocols to address this issue; and the need to support research on fluctuating asymmetry as a possible early warning tool for predicting declines.They also requested that the workshop proceedings be published; that a brochure on declining frogs be prepared; that a web site to provide up to date information about declining frog issues be established and that a national but restricted internet discussion group on frogs should be established.
I am pleased to observe that the majority of these initiatives, along with the implementation of the recommendations of the Action Plan for Australian Frogs, are in hand and continue to be supported by a wide range of sponsor organisations. In particular, a number of important projects are supported through the Commonwealth Government’s Natural Heritage Trust.
I believe this symposium makes an extremely valuable contribution to our knowledge of our threatened frog fauna. While the recovery of threatened frogs and the search for our missing frog species continues to be highly problematic it remains a work of the highest priority.
The agreement by Environment Australia to publish the proceedings of that workshop has led to the development of 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.
Alastair Campbell Environment Australia
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 WardellJohnson, 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 I am not a member of the international Declining Amphibian Populations Task Force (DAPTF), but I am nevertheless a very keen observer of its activities and progress. My interest stems from my participation in Canada’s own well-established and well-regarded task force.Thus, my views are
are many sound reasons for the logical sequence of having scientific investigation precede conservation action, but the DAPTF must not lose sight of the need for real success on the conservation side of its mission. Should conservation be neglected at the expense of a relentless pursuit for more and more data, the logic of the DAPTF’s very existence could rightfully be brought into question.
personal, and I do not speak on behalf of the international DAPTF. Amphibian biologists owe a great debt to the international DAPTF because, against substantial cultural resistance, important and historic strides have been made in many positive, mostly scientific, directions — but what about the amphibians? Are they receiving as many benefits as the amphibian biologists?- well not necessarily at this stage.There
INTRODUCTION The international research effort to unravel the mysteries of declining amphibian populations has had considerable “success” since its formation in 1992 — but how do we measure its success? More scientists are involved in studying the conservation biology of amphibians than ever before.The communications links between amphibian researchers have never been better. Mountains of data are being gathered, analysed and the results published. Public awareness of the issues is improving and public sympathy appears to be growing. In spite of all this, there are still very few examples
* Canadian Amphibian and Reptile Conservation Network, 1745 Bank Street, Victoria, British Columbia, CANADA, V8R 4V7. 9
of how this newly acquired knowledge and public sentiment will be applied to successfully reverse declining population trends.Thus, while those of us involved in these research and monitoring efforts perceive success and growing momentum, from an amphibian survival perspective the future is not much brighter today than it was in 1992. One of the DAPTF’s espoused tenets is to promote means by which declines can be halted and reversed, but we may be largely ignoring this aspect of our work in our rush to make scientific discoveries. Ultimately, we should be striving to create a “frog-friendlier” world, and not simply adding more statistically defensible documents to the historical record of amphibian demise. At the heart of the declining amphibian populations conundrum is a complex of Gordian Knots that are ecological, psychological, sociological, political, logistical, and philosophical, and it may not be humanly possible to unravel them all. It is now so well-established that the amphibian populations crisis is real that I will not explore its history in any detail here.The phenomenon has been documented on several continents, most disturbingly in what appear to be remote, pristine habitats, and the problem is being actively if not intensively investigated by over 1200 scientists organised into 90 working groups worldwide (Halliday and Heyer 1997). Though mobilised and motivated, the DAPTF will only succeed if it can secure the tools to carry out and complete its mission. Much hinges on its ability or inability to impress upon a largely apathetic public and their legislative representatives that declining amphibian populations research is important, necessary and environmentally, culturally and economically beneficial.
WHY CARE ABOUT AMPHIBIANS? People are intrinsically self-indulgent, and thus most of us who gravitate towards studying amphibians do so because we find it to be personally rewarding. In other words, people who like and study amphibians are usually not conscripted but rather seem to generate spontaneously.There is, in fact, a very substantial cross-cultural resistance to these eccentric people with their curious predilections. It is probably fair to say that, historically, there has been little glory or professional advantage to specialising in amphibian biology. Fortunately, history also seems to prove that there will always be at least a small group of dedicated, but marginalised, biologists who care about the technical details and conservation of amphibians, but for reasons that may never be fully comprehensible except to others within their own subculture. I am sure that we all recognise that to the great mass of humanity amphibian survival has little meaning. In ecological terms, amphibian specialists properly view the current situation as being catastrophic. Amphibians are low on the food chain and, when abundant, are extremely important in energy flow, particularly through wetland, damp forest and aquatic ecosystems.There are analogous situations, from the past, where dramatic population declines in other vertebrate groups have been reversed through human intervention, but the greatest successes are inevitably tied to those species that have broad and perennial public appeal, or are easily convertible to cash on either the open or black markets. Amphibians are a hard sell, and may always be a hard sell. It is nevertheless the DAPTF’s mission to untangle these psycho-socio-political knots — but how? 10
The problem of how to popularise the amphibian conservation cause and thence foment a general public concern for their plight is far from being solved. Undaunted, Halliday and Heyer (1997) recently attempted to summarise what, in their view, are the most compelling reasons why people should care enough about amphibians to at least try to explain, arrest and reverse these horrendous population declines. Ethical reasons topped their list. Amphibians, they say, have a right to exist. People have a moral responsibility to help amphibian populations recover since environmental vandalism by people is probably, more often than not, at the root of the problem.Though not religious advocates, they nevertheless go on to invoke biblical references wherein it is said that “humans are a special creation of God and are given dominion over the rest of living organisms on Earth”.They argue that this biblical passage should be interpreted by us to mean that our interrelations with other organisms should be “a stewardship, not a slaughter.” However, this moral position is certainly not confined to Judeo-Christianity, it is a widely held tenet of many of the world’s philosophies. Unfortunately, it is almost universally ignored, at least by the most environmentally reckless and politically influential segments of each culture. Incidentally, what Halliday and Heyer failed to point out is that if you accept that “humans are a special creation of God” then logically so are all other species.Thus, to destroy a species is to profane a divine creation and to mock the wisdom of God for having created it in the first place. These species are after all, according to biblical text, some of the remnants of “Paradise” which God created before He created people — more food for ecumenical thought. Religion aside, Halliday and Heyer’s second reason to care about amphibians is because they are interesting organisms. They present the example of the poison dart frogs of South and Central America whose habits are complicated, unique, and whose extinction should therefore be perceived as a great loss to humanity. On the surface, this is a reasonable rationale for preventing extinctions but, on the other hand, how many of the 4,500 or so species of living amphibians are as “interesting” as poison dart frogs? How will this help us save all those frogs and salamanders who are insufficiently “interesting” to the general public? Interesting species have the distinction of also being the most desired commodities in the commercial amphibian pet trade. Ultimately, there may be a real battle to convince some modern, urban youth that any real frog can be as interesting as a digital or virtual frog. Getting down to business, the third point that Halliday and Heyer raise in defence of amphibians is the mercenary interest we should all have in preserving nature. For example, pharmaceutically important “amphibians may provide direct benefit to humans.” As in the previous argument, however, problems arise when one extends this line of reasoning. It is not easy to see how such discoveries will ever translate into protection of wild populations. Once a useful drug is discovered you may well be able to either synthesise it or to farm the species under laboratory conditions.Thus, in this scenario, it becomes important to preserve amphibians only because they may have some pharmaceutical benefit, but if a given species is thoroughly investigated and found to have no medicinal value — then what?
The last reason on Halliday and Heyer’s list is possibly their most compelling, or at least marketable. It is the proposition that due to their complicated semi-aquatic life-styles, amphibians may be especially vulnerable to the accumulation of environmental stresses.They say, “amphibians are important indicators of general environmental health” and thus they “may provide an early warning about deteriorating environments”. After all, their environments are our environments — contaminated or not, we ingest the same water and breathe the same air. Amphibians could, therefore, be a natural early-warning system that alerts people to a growing lethal contamination while its concentration is still sub-lethal to humans. People may even be moved and comforted by the notion that noble frogs are sacrificing themselves to save human lives. Accordingly, it is worth stressing the physiological similarities between amphibians and other vertebrates, including people. One thing that Halliday and Heyer did not discuss explicitly were the concepts of “shame” and “guilt”.These are the timehonoured psychological forces that organised religions have used to such great effect in controlling our baser motives. I think that many people are developing a stronger personal environmental ethic because they have been shamed into it. For example, they have been forced to know about and look at the senseless brutality of whaling and it is now political suicide, in most parts of the world, to advocate a commercial whale hunt. It is my firm impression that, at least in Canada, younger generations are beginning to grasp the magnitude of the environmental damage that has been wrought within the past century.They find it to be shameful, even if they are not able to fully grasp the data that documents each crisis or the subtleties of each scientific argument. All things considered, amphibians are harmless, defenceless, oddly beautiful, yet ecologically, culturally and economically significant. How can it be acceptable to ignore their plight? Thus groups such as the DAPTF might be more successful in getting the public to support their initiatives if the case for amphibians were presented more as a struggle between civility and brutality. Civilised people should be as alarmed about the extinction of species, even uninteresting amphibians, as they would be about a malicious assault on the unique treasures of a museum, art gallery or one’s own community. We need metaphors and similes and even paintings, poems and songs to help people, with no biological training, to conceptualise the magnitude of the problem and personalise the loss — the required political support should follow. Hitherto, the DAPTF has failed to garner the vast moral, financial, and political forces that it needs to adequately confront this crisis. In the main, this subject is still treated as a novelty in the popular media, and policy makers and major funding agencies remain, for the most part, fitfully interested but so far unmoved.
THE SEARCH FOR CAUSES Generally speaking, it is not too difficult to find reasons why amphibian populations are dwindling. For example, the commercial trade in live frogs and frog meat is huge. Acid rain has been implicated in the decline of some species.The commonly used fertiliser, ammonium nitrate, can be toxic to amphibians, but it is only one of myriad batrachicidal contaminants including herbicides, pesticides, and petroleum
products.Virulent pathogens are being found that may or may not have been introduced with non-native game fish, and assorted environmental stressors are well known to suppress the immune system and thereby induce infections. Perhaps amphibians are losing a natural ability to ward off infections from common microorganisms that they would normally not be susceptible to.The ozone layer has been seriously compromised, especially over Australia, and while early experimental results on the harmful effects of UV-B on amphibian embryos have not yet been replicated, UV-B still remains an interesting candidate. A 20 year study in northern Ontario, Canada has connected acidification of lakes to increased UV-B exposure by showing that acidity can dissolve free-floating organics that would otherwise effectively screen UV-B radiation. Studies are underway into hormone mimics and hormone blockers that can disrupt the endocrine system seriously enough to cause sex-reversal. While it is generally agreed that global warming is a fact, the effects that this will ultimately have, or may already be having, on amphibians remains mostly speculative. Another fertile area of scientific inquiry is how to contend with burgeoning populations of released, predatory exotics — such as the American bullfrog (Rana catesbeiana) and the cane toad (Bufo marinus). Can these be eradicated, and, if so, should they be? Finally, will wetlands and waterways ever be managed for amphibians with the zeal that they are today for predatory fish and waterfowl? Quite recently, malformations in amphibian populations suddenly became a hot topic.This is particularly true in North America where a minor media sensation occurred when grossly deformed leopard frogs were found in unprecedented numbers by school children in the State of Minnesota. Parasites and contaminants have been implicated but the basis of many amphibian malformations still remain mysterious. It is critical to differentiate between traumatic events and true developmental abnormalities, but there are very few specialists qualified to make this assessment. This year (1997) the United States National Aeronautics and Space Administration (NASA) became involved in DAP research when Ron Heyer (Smithsonian Institution) and Cynthia Carey (University of Colorado, Boulder) were awarded a grant from the NASA Mission to Planet Earth Program.The purpose of the project is to determine whether NASA has any climate change data, gathered by Earthmonitoring satellites, that could be useful in trying to account for known, well-documented amphibian declines scattered across the globe. A team of researchers from various parts of the world collaborated on this project and a report was due late in 1998. Increasingly, amphibian researchers are acknowledging the possibility that they themselves could be unwittingly transporting lethal agents and pathogenic organisms (e.g. viruses, viroids, bacteria, funguses) into amphibian habitats. While marking and measuring the animals they study, traditionally, researchers and their research subjects are in the most intimate contact. For this reason, some researchers now go to great lengths to minimise the possibility of contaminating critical study sites.This area of concern has not received much public discussion and is not mentioned in recent manuals on amphibian monitoring methodologies. However, in the summer of 1998 the DAPTF produced a fieldwork code of practice that is an important first attempt 11
to fill this gap (DAPTF et al. 1998). Historically, standardised methods of field study have presumed that it is harmless to amphibians for people to handle them directly — which may yet prove to be true. In the meantime however, especially when studying vulnerable populations, it would be most prudent to err on the side of caution and adhere to the code of practice as closely as possible.
HALTING AND REVERSING AMPHIBIAN DECLINES It is routinely argued that the results of research into the causes of declining amphibian populations will be applied to the problem of halting and reversing declining population trends — though earnestly stated, this is seldom the case. There are, sadly, very few instances where an intensive, remedial conservation program has been the logical consequence of research results from declining amphibian populations studies. Hitherto, the standard approach by scientists and wildlife managers in dealing with the declining amphibian populations phenomenon is to: a. attempt to identify the species and locations where populations are dwindling; b. attempt to identify the underlying cause, or causes, of population declines; c. establish extensive population monitoring programs; d. revise endangered species lists; e. produce problem analyses and management plans; f. launch public education campaigns. There is nothing inherently wrong with any of these actions, except that they are often used, for political reasons, as stall tactics and can consume a disproportionate amount of time and money. When faced with a crisis, at some point, the best information that is available must be applied and action must be taken or the species may be lost. Consequently, it is sensible to funnel the greatest proportion of available resources to the most vital need.This may perforce involve making fateful decisions and even leaving one’s office. It is supremely important for amphibian survival and for the future of the DAPTF to demonstrate that human intervention can not only explain declining population trends but can successfully reverse them. Some risks may have to be run in order to achieve this, and alas someone must shoulder the responsibility because in the process honest mistakes may be made. Examples of conservation recommendations that are commonly ignored include habitat creation and reclamation, the eradication of exotics, captive breeding, head-starting tadpoles, and re-introducing species back into localities within the historical range.The greatest care must be taken to act ethically and prudently. Nevertheless, as we now know, amphibian populations can crash so suddenly that there is no time for debate. A reasonably comprehensive plan (e.g.Tyler 1997) should be adopted and the logistical details of carrying out its provisions should be agreed upon well in advance of a worst case scenario so that time and effort are not wasted when the need suddenly arises.
12
BUREAUCRACY AND RED TAPE In our universe, red tape and bureaucracy are seemingly as inevitable as entropy. I was recently reminded of how commonly people can lose their perspective on a problem when they become totally immersed in the process of formulating committee decisions.Take for example the case of Dr Jaime Péfaur, Universidad de Los Andes,Venezuela, whose conservation efforts are being thwarted by, of all things, the newly revised criteria for listing a species to the World Conservation Union’s (IUCN), Red Data Book of endangered species around the globe. Dr Péfaur knows that species of frogs are disappearing rapidly in his country and he needs the resources to study this and try to arrest or reverse the trend. However, the government of Venezuela will not act to fund his research unless the species in question are declared “red-listed” by the IUCN, and the IUCN will not list them until conclusive statistics on population densities are available. Alas, Dr Péfaur finds himself trapped in a perfect paradox. He can not meet the IUCN’s demand for statistics because his government will not give him the resources and therefore the frogs will continue to disappear — unstudied and unprotected. The most valuable lesson to learn from Dr Péfaur’s experience is to never permit a doctrine or prevailing policy to take on more importance than the problematical situations that it was designed to prevent. Unpredictable problems will arise and it is best to say, “this is the problem — how do we fix it?”, rather than abandoning hope because, “this problem is not covered by the policy”.
POLITICS AND SCIENCE If the prevailing political environment is not to our liking, how can scientists go about changing it? In the past, we have tended to address our appeals to the established government bureaucracies assigned to environmental matters. Usually, these agencies are staffed by fellow biologists who we can relate to and who tend to be reassuringly sympathetic. Unfortunately, we are often disappointed when little action is taken, but we are content that everything that could have been done was. In many cases, our expectations and our assessments are naive, because government bureaucracies exist to preserve the current political condition.Thus, we must get our issues on the political agenda if we want to effect real change in public policies.To do this we need to demonstrate genuine support for our cause from a wide constituency.
HERPETOLOGICAL EDUCATION A novel development to spring from the Third World Congress of Herpetology, held in August 1997 in Prague, Czech Republic, was the formation of an International Committee on Environmental Education on Amphibians and Reptiles.The round-table discussion was attended by about 40 people representing: Hungary, the United States, Italy, Sweden, Slovenia, Israel, Switzerland, Canada, the Netherlands, Croatia, the United Kingdom, New Zealand, Denmark, Poland and Australia. It got me thinking about this long-neglected aspect of our work and how it relates to the success of the DAPTF.
One of the unique problems amphibians face in their struggle to coexist with people is the fact that many people express a phobic reaction to the very sight of them. Oddly, since we have no statistical benchmarks, herpetologists have only the vaguest of notions as to how innate, culturally pervasive, or extreme this attitude is today. Are attitudes changing? — or can they be changed? As critical as this information is to all aspects of the DAPTF’s work, we are woefully ignorant about how to convey our message to the general public for maximum effect. Consequently, it would be extremely valuable to understand more fully the psychological and cultural dimensions of herpetological education.The positive ramifications of this line of inquiry would be in: a) soliciting public sympathy and political support; b) developing more effective methods in herpetological education in schools; c) improving the effectiveness of conservation actions and associated public relations campaigns; d) providing a statistical benchmark to which future surveys may be compared in order to monitor shifts in the perceptions of the general public towards amphibians. Is it possible to popularise a universal eco-etiquette that includes amphibians and is lasting and cross-cultural? If education is the way to achieve this, we had better act quickly and it had better work because, in an already crowded world, the United Nations’ highest current projection is that the human global population will grow from 5.9 billion in 1998 to 11 billion people by 2050 (National Geographic Magazine, October 1998:5). Herpetological education can also inspire some people to try forging careers for themselves as amphibian specialists.This has always been a dicey proposition — but never more so than today. In North America, for example, the traditional objective scientific and moral authority of museum curators and university professors in directing conservation biology research is being appropriated by politically hamstrung and policy-driven government administrators.The DAPTF’s mission relies heavily upon authoritative scientific advocacy and a desire to integrate and infuse more of the details of 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.
it “has legs”.Thus, the external support and media attention that the DAPTF has received over the past six years very likely has at least as much to do with its entertainment and curiosity value as with altruistic public sentiment. So what happens some day, maybe soon, when the mysteries of the declining amphibian populations phenomenon are resolved and we are back to where we were prior to the formation of the task force — out of the limelight and tussling with the mundane problems of amphibian conservation. From a conservationist’s perspective, the issues will be no less grave and melancholy, but from the public’s view the glinting aura of the inexplicable will have been peeled away. The DAPTF will have to induce its own metamorphosis, since it will have come to strongly personify not a cause but a catastrophe. It will have to cultivate genuine goodwill and concern from the general public, and for this it will need to demonstrate successes in reversing declining population trends through the integration of science, remedial conservation actions and public education and participation. If the DAPTF fails in these arenas, its DAP research discoveries may come to little, and the task force and its great potential could tragically go the way of the golden toad.
REFERENCES DAPTF, (1998) The DAPTF Fieldwork Code of Practice. Copies of this leaflet are available by contacting John Wilkinson, Biology Department,The Open University, Walton Hall, Milton Keynes, United Kingdom, MK7 6AA (
[email protected]). Halliday,T.R. and Heyer, W.R., (1997) Are amphibian populations disappearing?: a task force status report 1996-1997.The Boreal Dip Net 2(1):2-6. Joel, L.S., (1998) Population. National Geographic, October 1998:5. Tyler, M.J., (1997) The Action Plan for Australian Frogs. Wildlife Australia, Canberra.
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 Eleven frog species have declined or disappeared since 1985 and 1989 in the Central Queensland Coast and Wet Tropics Biogeographic Regions respectively of northern Queensland. Declining species are in upland areas and declines occurred suddenly for highly susceptible species with close associations with streams. Despite considerable effort the causal factor(s) have not been determined although several hypotheses have been proposed. The taxa of concern are phylogenetically diverse and exhibit a variety of morphologies, reproduction, behaviour and microhabitat preferences. Species outside rainforests are under anthropogenic pressures especially in sugar cane expansion areas of the coastal lowlands. Current activities are focused on population monitoring by community groups, the Department of Environment and Heritage and James Cook
University. Recovery Plans have been compiled but have not been signed off by Government. Future efforts need to incorporate experimental ecology, management of water usage and wider communitybased participation in monitoring and surveys.
INTRODUCTION There has been a plethora of scientific and popular articles on declines of frog populations in northern Queensland since the 1980’s (e.g. Winter and McDonald 1986; McDonald 1990; Tyler 1991; Richards et al. 1993; Ingram and McDonald 1993; Dennis and Mahony 1994; McNellie and Hero 1994;Trenerry et al. 1994; Martin and McDonald 1995,1996, Laurance et al. 1996; Laurance 1996a, b; Mahony 1996; Hero 1996; Alford and Richards 1997; Alford et al. 1998; Berger et al. 1998). Numerous meetings, discussions, workshops and debates also have occurred since 1990, in an effort to determine causes of these frog declines in apparently secure areas of National Park and State Forest, some of which are further protected in a World Heritage Area. A wealth of unpublished reports and recovery plans have been devoted to the search for a solution to this problem. A benefit of these studies has been a vast increase in our biological knowledge of stream frogs in north Queensland.
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 Queensland Coast Biogeographic Regions including those shared.
Species
WT CMC Shared other bioregions
31 10 19
TOTAL
60
FIGURE 1: Wet Tropics and Central Queensland Coast Biogeographic Regions showing rainforest areas.
Total species
50 29
TABLE 2: Endemic, rare and threatened and declining frogs numbers in the Wet Tropics and Central Queensland Coast Biogeographic Regions.
Endemic Endangered Rare Declined
WT
CMC
23 (46%) 7 13 (2) 8 (7 End, 1 Rare)
3 (10.3%) 2 2(1) 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 (e.g., Laurance et al. 1996), or has attempted to test hypotheses using data that are too sparse to provide rigorous tests (e.g., Dennis and Mahony 1994; Laurance 1996 a,b). We argue that what is most needed at present is research carefully targeted at testing and potentially rejecting hypotheses. The rigorous rejection of some hypotheses will allow further research and monitoring efforts to be concentrated on refining and testing those hypotheses that remain. This paper reviews population declines in northern Queensland and proposes future directions for research and monitoring. Much of the information presented is derived from a current research program of the Queensland Department of Environment and Heritage, the Department of Zoology and Tropical Ecology at James Cook University and the Cooperative Research Centre for Tropical Rainforest Ecology and Management, and independent research funded by the Wet Tropics Management Authority.
STUDY AREAS AND SPECIES COMPOSITION This review covers rainforest areas in the Central Queensland Coast (CQC), Wet Tropics (WT), and Cape York Peninsula (CYP) biogeographic regions of northern Queensland (Figure 1). Population declines of rainforest stream-dwelling frogs have occurred in two of these 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).
A total of 60 frog species (50% of Queensland frogs) are found in these biogeographic regions; 19 of these species also occur in other regions. Fifty species have been recorded from the WT and 29 from the CQC (Table 1). The biogeographic regions have endemic species reflecting the influence of rainforest habitats (Table 2). Nearly 50% (23) of the WT frogs are endemic.This is reflected in the conservation status of WT species with high numbers of rare, geographically-localised species — a consequence of the interactive influences of the rainforest habitat and topography, climate and history. Many species are restricted to upland and mountain-top habitats. Only two of the WT species listed as rare under the regulations to the Queensland Nature Conservation Act 1992 are shared with other areas. All endangered species which have had population declines are endemic to the WT and CQC biogeographic areas. The declining species are representatives of the two most diverse families in Australia; Myobatrachidae and Hylidae. No declines have been reported for the other native frog families (the Microhylidae and Ranidae) within the areas. The eleven species which have experienced declines are from five genera (Table 3). Some have declined in parts of their distribution, while others have not been seen for some time (Table 3). Several declining species have very localised distributions while others were widespread in suitable habitat throughout the rainforest blocks within the biogeographic regions (Table 3).
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
Litoria Litoria Litoria Litoria Litoria Nyctimystes Taudactylus Taudactylus Adelotus Rheobatrachus Taudactylus
lorica nannotis nyakalensis rheocola genimaculata dayi acutirostris rheophilus brevis vitellinus eungellensis
Bio. Reg.
Declines
Action Plan Cons.Status
NCA Regs. Cons.Status
Distribution*
Decline period
WT WT WT WT WT+ WT WT WT CMC + CMC CMC
Y Y Y Y Y Y Y Y Y Y Y
E E E E
E E E E R E E E C E E
1 3 2 2 3 3 2 1 2 1 2
1991 1989–1993 1990 1989–1994 1990 –1994 1989–1994 1989–1994 1989–1991 1985–86 Jan 1985–Mar 85 1985–86
E E E E E
* % of latitudinal spread in biogeographic rainforest block 1. < 20% 2. 20–60% 3. 60–100%
REGIONAL ASSESSMENTS OF DECLINING FROGS Declines occurred in the CQC during 1985-86 and in the WT between early 1989 and early 1994.
Central Queensland Coast In the CQC, two species, Rheobatrachus vitellinus and Taudactylus eungellensis have declined precipitously (McDonald 1990). Another species, T. liemi, appears to be secure, although other Taudactylus species have suffered declines. Declines of T. eungellensis commenced in 1985 and the species was last seen in monitored sites in 1986 (McDonald 1990). An individual was observed in 1992 (Couper 1992) and small populations have been located since November 1993 in the area around Mt Dalrymple (McNellie and Hero 1994; Retallick pers. comm.). Populations have not been located in the southern and northern areas of the known distribution ( McNellie and Hero 1994; McDonald 1990; Retallick pers. comm.). The small populations located since 1993 are not as large or as extensive as those found in the 1970’s and 1980’s and only one population is nearing the numbers found at that time (Retallick pers. comm.).The three largest populations are found in one catchment. Other populations are smaller and appear to be stable with most of the existing populations at lower altitudes. Studies conducted by Richard Retallick indicate that the sizes of monitored populations are slowly increasing but it is not known if small populations in other catchments are increasing or if the ranges of populations are expanding. Rheobatrachus vitellinus has not been observed since March 1985, despite periodic intensive and extensive searches (McDonald, unpubl.;Retallick, pers. comm.; McNellie and Hero 1994).
16
Wet Tropics Eight species in the WT have experienced population declines (Table 3; Richards et al. 1993; Laurance et al. 1996; Martin and McDonald 1996; McDonald unpubl.).Three species (Litoria nannotis, L. rheocola and Nyctimystes dayi) have stable populations at lower altitudes (below approximately 400 m). Known declines in the WT occurred between 1988 and late 1994 (Table 3; Richards et al. 1993; Laurance et al. 1996). Taudactylus acutirostris, L. lorica and L. nyakalensis have not been observed in substantial numbers since late 1994 (Ingram and McDonald 1993, McDonald, unpubl.) with the most recent reports being of a single individual of T. acutirostris in Slaty Creek in 1994 (Roberts pers. comm.), a sub-adult at Big Tableland in January 1995 (McDonald unpubl.), and a report of the call of a single frog near Millaa Millaa. Small populations in two creeks of T. rheophilus has recently been relocated on the CarbineTableland and the Bellenden Ker Range ( Marshall 1998). In each case where population crashes affected monitored populations, tadpoles survived and metamorphosed after the adult population had crashed, indicating that the causal factor did not affect this stage of the frog life cycle (Richards et al. 1993; Dennis and Mahony 1994; Laurance 1996 a,b; Laurance et al. 1996; Martin and McDonald 1996).Tadpoles removed from declining populations successfully metamorphosed, but all individuals died before attaining adult size (Dennis and Mahony 1994). It is possible that post-metamorphic mortality was more rapid in individuals having close contact with stream water, but this conclusion is not strongly supported by statistical analysis (Dennis and Mahony 1994). At a monitored site at Big Tableland (15° 42’S 145° 16’ E, 620 m) four species of frog declined suddenly in late 1993 (Figure 2).The population of Litoria genimaculata at the site 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
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 which did not totally disappear were those having the largest number of pigmented eggs of all the declining species. At O’Keefe Creek, Big Tableland (400 m) species declined simultaneously with those at the primary monitoring site. These sites have been monitored every 4-6 weeks since mid 1992. Since June 1995 L. rheocola and N. dayi (but not T. acutirostris and L. nannotis) have occasionally reappeared near the 400 m site, but have not established resident populations.This site is immediately above a steep escarpment that forms a disjunction between upland and lowland populations.The lowland populations still persist. No colonisation of areas above the O’Keefe Creek site (400 m) has occurred, which may indicate that the factor which caused the declines is still operating. In December 1997 two N. dayi were heard calling below the primary monitoring site (620 m) indicating some individuals may be moving upstream although tadpoles and eggs have not been located. N. dayi activity along streams is very seasonal (McDonald unpubl.) and appears to spend more time away from streams than do the other species (McDonald unpubl.).
Cape York Peninsula No declines occurred among the Cape York rainforest stream frogs during 1993-1998 (McDonald unpubl.). Litoria longirostris, L. eucnemis and Rana daemeli populations monitored yearly in upland areas of the McIlwraith Range between 1993 and the present have remained stable. L. eucnemis and R. daemeli have considerable latitudinal and altitudinal ranges and both species occur in New Guinea. L. longirostris is endemic to, and patchily distributed in, rainforest streams above 400 m in the McIlwraith Range (McDonald and Davies unpubl.).
NATURAL HISTORY/ECOLOGY Declining frogs in north Queensland breed in rainforest streams in upland areas (Table 4). However they differ widely in taxonomic affinity, morphology, microhabitat preferences, breeding strategies and behaviour.
Habitat The declining frog species have a range of microhabitat preferences. In earlier reports (e.g., Richards et al. 1993) we referred to all species breeding in streams as stream-dwelling frogs. It is now clear that frogs which breed in streams differ in their fidelity to them.Those with the tightest association may be the most likely to disappear or decline (Dennis and Mahony 1994;Table 4 ). Rheobatrachus vitellinus, Taudactylus acutirostris, T. eungellensis and T. rheophilus and members of the Litoria nannotis group spend a greater proportion of their time in or immediately adjacent to streams than do L. genimaculata and Adelotus brevis (Table 4).
Daily active period Most declining species are nocturnal but several species are diurnal and there appears to be no link between activity periods and declines. Taudactylus eungellensis and T. acutirostris are diurnal while T. rheophilus is diurnal but can be active during the evening. All of the other species are nocturnal although during heavily overcast and rainy conditions some individuals become active and call during the day.
Morphology The declining frog species exhibit a range of morphologies.The species vary in size: the snout-vent length (SVL) of the smallest, 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 sexual dimorphism in 17
size, a characteristic which is most striking in L. genimaculata and N. dayi (Table 5). Adelotus brevis shows a reverse dimorphism, with males larger than females (Moore 1961). Weights reflect general appearance of the frogs. T. acutirostris is the lightest and the heaviest is R. vitellinus (Table 5). R. vitellinus is the most robust species with a weight to SVL ratio (g:cm) for males of 3.47:1 and for females of 4.81:1. T. acutirostris is the most gracile species with a weight: SVL
TABLE 4: Stream fidelity of declining and non-declining rainforest frogs.
Species
Stream-assoc.
Decline category
3b 3b 2b 2b 2b 1b 1b 1b 1b 1b 3b 3a 3b 2b 3a 1a 2a 1b 1b 1b
2 1 1 2 1 1 3 3 3 3 1 1 1 3 1 3 3 3 1 3
Adelotus brevis Litoria chloris L. eucnemis L. genimaculata L. lesueuri L. longirostris L. lorica L. nannotis L. nyakalensis L. rheocola L. xanthomera Mixophyes schevilli M. fasciolatus Nyctimystes dayi Rana daemeli Rheobatrachus vitellinus Taudactylus acutirostris T. eungellensis T. liemi T. rheophilus
The tadpoles of R. vitellinus do not have keratinised mouth parts, are unpigmented and are carried in the stomach of the female. Species of the L. nannotis group (except L. lorica which is unknown) and N. dayi have lotic suctorial tadpoles with large suctorial mouths for holding on to rocks in the swift flowing streams (Davies and Richards 1990; Richards 1992). Taudactylus acutirostris, T. eungellensis, Adelotus brevis and L. genimaculata have lotic-benthic form tadpoles with small oral discs.The tadpoles of T. rheophilus are unknown.
obligate
tight to moderate
loose to very loose
2 0 7
2 1 2
5 1 0
Fisher's Exact test, 2-tailed P = 0.019 Degree of association with streams: 1 a. obligate, aquatic.The species is found only in the water b. obligate, non-aquatic.The species is reliably found within the stream banks over the entire year 2 a. tight.The species is found within the stream banks over most of the year, but regularly ventures away from the stream b. moderate.The species is found with some reliability within the stream banks over an extended season 3 a. loose. Larvae occur predominantly in streams, but adults visit them irregularly b. very loose. Larvae do not occur predominantly in streams, adults may visit them irregularly none. Neither larvae nor adults occur within stream banks with any frequency
Decline categories 1 2 3
18
no declines some declines, no disappearances at least some disappearances
Most declining species, except for L. genimaculata, have relatively small numbers of large eggs (Table 6). Eight species 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 egg masses in clear jelly in pools (L. genimaculata); on or under rocks in streams for (T. acutirostris , L. rheocola, and T. eungellensis); under rocks in streams (N. dayi); under crevices on waterfalls (L. nannotis).The egg deposition site of R. vitellinus is unknown but the stomach certainly carries tadpoles (McDonald unpubl.). Egg deposition sites for T. rheophilus, L. nyakalensis and L. lorica are unknown.
Tadpoles of the declining species are highly variable (Table 6). Definition of tadpole forms follow Altig and Johnston 1989.
degree of association with streams
4
Eggs
Tadpoles
Summary:
No changes Declines Disappearances
ratio for males of 0.40:1 and for females of 0.53:1. No data are available on weights for A. brevis, L. lorica, L. nyakalensis, and T. rheophilus.
At least some tadpoles of each species except Rheobatrachus vitellinus and A. brevis overwinter in the stream. T. rheophilus and L. lorica are suspected to overwinter, while overwintering has been documented in the other seven species (Richards et al. 1993; Richards 1992; Retallick pers comm.).
NON-DECLINING FROGS Within the rainforests and their margins of the CQC and WT there is no indication of declines of populations of Litoria lesueurii, a species which has declined in the rainforests of south-east Queensland (Ingram and McDonald 1993).The widespread WT species, Litoria xanthomera and Mixophyes schevilli, have not declined at any site, although their populations have not been intensively and systematically monitored. None of the 14 species of litter-dwelling microhylids appear to have declined. However, relatively little is known of this group because standardised monitoring has been conducted at only one site, Paluma near Townsville. Presence/absence monitoring for microhylids at monitoring sites for stream frogs also indicates no loss of species (Dennis and Mahony 1994; McDonald unpubl.). In the CQC at Eungella, non-stream species such as L. chloris and M. fasciolatus appear secure. T. liemi has experienced no declines (McDonald pers. obs.; Retallick pers. comm.).
TABLE 5: Snout-vent length and weights of declining frogs in the Wet Tropics and Central Queensland Coast Biogeographic region.
Max SVL Min SVL (mm) (mm)
Species
Sex
No.
Litoria genimaculata
Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female
1021 102 7 3 756 68 99 41 6
54.4 86.6 33.1 37.3 36.4 41.2 51.8 59.1 32.3
32 53 29.6 32.9 27 28.4 41.2 48.8 29.7
265 4 14 7 48 59 99 39 490 640 12 9
41.9 58.8 43.5 38 58.3 72.2 25.5 31.2 32 37.11 27.4 30.5
32.7 54 34.2 29.1 42.6 47 18.8 22.7 23.8 25.5 24.3 24.1
Species
Sex
No.
Max WT (gms)
Min WT (gms)
Litoria genimaculata
Male Female Male Female Male Female Male Female Male Female Male Female Male Female
1021 102 756 68 99 41 265 4 48 59 99 39 489 640
9.6 38.4 3.3 4.55 12 17 3.9 11 26.5 46.5 1.6 2.5 3.7 5
2 7.5 1.2 1.8 6 8.2 1.9 8.4 11 13 0.6 0.9 0.8 1.9
L. lorica L. rheocola L. nannotis L. nyakalensis Nyctimystes dayi Adelotus brevis Rheobatrachus vitellinus Taudactylus acutirostris T. eungellensis T. rheophilus
L. rheocola L. nannotis Nyctimystes dayi Rheobatrachus vitellinus Taudactylus acutirostris T. eungellensis
POSSIBLE CAUSES OF DECLINES Although numerous possible causes have been proposed to explain declines of frog populations elsewhere, few of these have relevance to north Queensland populations. Habitat has remained intact with no clearing or logging of forests in the Wet Tropics World Heritage Area since 1988. Species were present even though logging and mining of the rainforest had occurred in the past (McDonald 1992). However, this does not rule out more subtle, unmeasured environmental changes affecting declining frogs. No obvious environmental changes were detected during stream monitoring at the times when declines occurred (Richards et al. 1993; McDonald unpubl.).Water characteristics including pH, conductivity, water temperature, concentrations of metal ions and biocides, and dissolved oxygen measured at a large number of sites have not significantly differed between locations which had declines and those that did not . Rainfall has remained within
Mean SVL Std Dev 42.93 68.75
3.04 6.06
Source McDonald and Martin unpublished Davies and McDonald 1979
31.52 36.27 47.04 53.63 31.6
1.56 2.54 2.09 2.27 2.975
37.22 57.2
1.65 2.2
McDonald and Martin unpublished McDonald and Martin unpublished Liem 1974 McDonald and Martin unpublished Moore 1961
52.11 59.78 20.62 24.6 27.63 32.33 26.1 28.5
4.04 6.31 1.48 1.62 1.2 1.72
McDonald and Martin unpublished R. Retallick pers comm. Liem and Hosmer 1973
Mean WT Std Dev 4.66 22.64 2 3.09 8.4 11.89 2.56 10.23 18.13 28.78 0.83 1.3 2.32 3.48
K. McDonald unpublished
1.06 6.45 0.27 0.66 1.05 2.17 0.3 1.24 3.8 8.91 0.19 0.27 0.39 0.57
Source McDonald and Martin unpublished McDonald and Martin unpublished McDonald and Martin unpublished McDonald and Martin unpublished K. McDonald unpublished McDonald and Martin unpublished R. Retallick pers comm.
expected ranges with no periods of prolonged drought. No declines were associated with periods of low rainfall (Laurance 1996 a,b). Attempts to correlate environmental changes with declines depend strongly on information about when the environment affected populations; recent evidence (Alford et al. 1997) suggests that environmental effects preceded some declines by 1–2 years, so many of these analyses may need to be repeated. However, data from other populations (SVL/weight ratios; McDonald 1990) failed to indicate any decline in health of monitored populations before declines occurred. There has been repeated speculation that UV-B light has caused declines. However this possibility appears to be unlikely in north Queensland. All declining species in this area are found in rainforest with dense canopy cover. Several are active only nocturnally, several lay eggs under rocks (exceptions are L. genimaculata and A. brevis), tadpoles are still present when adults disappear (most UV-B hypotheses 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
Taudactylus acutirostris
2.2–2.7
Egg no.
Pigmentation
Tadpole
25–40
Lotic benthic Yes
T. rheophilus Litoria nannotis
No
Lotic benthic? Lotic suctorial
No No
Lotic suctorial? Lotic suctorial
2.2–2.6
843 107 22 30–50
No Yes No No No
Lotic suctorial Lotic benthic Lotic suctorial In stomach Lotic benthic
Adelotus brevis
1.7–1.8
270
Yes
L. eucnemis L. longirostris
2.0–2.4
1000+ 28–60
Yes Yes
Lotic benthic/ lentic benthic Lotic benthic Lotic benthic
L.lorica L. rheocola L.nyakalensis L. genimaculata Nyctimystes dayi Rheobatrachus. vitellinus T. eungellensis
1.8–2.4 2.7–3.4 1.98–2.93 1.4–1.8 2.4–2.6 2.1–2.56 2.3–2.6
35–50 136–216
46–63
periods (Figure 2). All of these factors make it unlikely that declines have resulted from direct effects of UV-B on populations. Most importantly it is well established that there have been no significant changes in stratospheric ozone in the tropical areas which can be linked with UV-B increases as experienced in higher latitudes (Madronich and de Gruijl 1993; McPeters et al. 1996; Moise pers. comm.) A recent hypothesis suggests that a virulent pathogen, possibly a virus or Chytrid fungus, has decimated frog populations (Laurance et al. 1996; Berger et al. 1998). Circumstantial evidence from observations of rapid declines and location of sick and dying animals provides a strong link to a disease being the proximate cause of some declines (Dennis and Mahony 1994; Laurance et al. 1996; Berger et al. 1998). However, there are alternative interpretations (Alford and Richards 1997) for all lines of evidence suggesting that a disease might be the sole cause of declines (Laurance et al. 1996), and recent evidence (Alford et al. 1997) suggests that the health of adults began to decline long before populations crashed. Decreases in body condition were not observed in Rheobatrachus vitellinus prior to declines (McDonald 1990).The disease hypothesis clearly requires further testing and possibly refinement.
Reference Liem & Hosmer 1973 McDonald & Richards unpublished Liem & Hosmer 1973 Liem 1974 Hero & Fickling 1996 Davies & McDonald 1979 Liem 1974 Hero & Fickling 1996 Richards 1992 Davies 1989 Davies & Richards 1990 McDonald & Tyler (1984) McDonald pers. obs., Retallick pers comm., Liem & Hosmer 1973. Tyler 1994 McDonald unpublished McDonald unpublished, McDonald & Storch 1994
some cases (Alford unpubl.). Many coastal lowland habitats such as Melaleuca swamps and forests are important for the maintenance of frog populations. Species such as Notaden melanoscaphus which has isolated populations in the coastal areas near Townsville are disappearing at rates that are alarming because of loss of habitat through real estate expansion.
CURRENT ACTIONS Monitoring Monitoring studies have been conducted in the Wet Tropics and Eungella by James Cook University and the Queensland Department of Environment and Heritage for medium to long term periods.
Wet Tropics A transect at Birthday Creek, near Paluma has been monitored since 1987 and one at Big Tableland since 1992. Other sites have been visited less frequently at differing intervals throughout the WT over the last seven years that 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 on rainforest species, which do not appear to have suffered direct anthropogenic effects. Little systematic survey and monitoring effort has been devoted to species in other habitats. Monitoring around Townsville and in the Wet Tropics open forests indicates no declines other than those related to obvious anthropogenic causes such as clearing of habitat.There is local loss of habitat around real estate development or in the coastal lowlands through the sugar expansion program in the last five years.The construction of dams, quarries and road-side ditches may have enhanced breeding habitat for some species, although the effects of these structures on population dynamics are uncertain and may be negative in
At Eungella monitoring by the Queensland Department of Environment and Heritage occurred at selected sites during 1980–89 (McDonald 1990); since 1993 these and additional sites have been monitored by James Cook University.The current monitoring is conducted by Richard Retallick with assistance from national park rangers.
20
Public participation in the monitoring program has been in the form of assistance to researchers. Monitoring has also been conducted by the Cape York Herpetological Society at two locations in the WT for the last 18 months and it is expected that this program will expand to cover additional sites.
Because the results of Alford et al. (1997) suggest that developmental stability analysis may be a useful tool for detecting the effects of environmental changes before they lead to changes in population sizes, measurements of limb asymmetry have recently been incorporated into monitoring programs.The techniques used follow those outlined by Alford et al. (1998). The lengths of the hindlimbs, from heel to the outer end of the tibio-fibula, are measured three times on each side of the animal, using callipers and alternating between sides.The person handling the callipers gives them to a second person, who reads the length and records it.This gives three independent replicate measurements on each side of the animal, which can be analysed to examine asymmetry, following the methods suggested by Palmer and Strobeck (1986). Preliminary results of this monitoring indicate that asymmetry levels in populations of L. genimaculata are generally low.The technique is discussed in detail elsewhere in this volume (Alford et al. 1999).
RECOVERY PLANS Two recovery plans for the Wet Tropics and Eungella have been written (Martin and McDonald 1995, 1996) and approved by the Northern Threatened Frogs Recovery Team. The plans have not yet been adopted by the Federal or State agencies responsible for the environment. Recovery Plans need constant updating as more and better information comes to hand.There is a need to respond quickly to developments, especially with projects which exploit an event such as a species die off which occurs quickly (Figure 2). Organisational inertia can be a deterrent to implementing actions in these circumstances.This can be addressed by a yearly review and modification of action plans. A process should be in place to permit the rapid implementation by State and Federal environment agencies of such modifications.
FUTURE DIRECTIONS There are many challenges facing frog research and monitoring in the Wet Tropics and Eungella biogeographic regions. Future directions include: 1. Experimental work — experimentation and manipulation to test the hypotheses which have been advanced regarding possible causes of declines. 2. Management of water usage — such as dams and water extraction, and examination of impact on remnant frog populations. Experimental research and monitoring of impacts of water management regimes is essential as many water extraction processes are proposed in the high rainfall areas of Eungella and the Wet tropics, e.g., the proposed Finch Hatton Dam and water extraction in the Johnston River catchment. 3. Increased public participation — there is a need to have greater public participation, especially amongst herpetological clubs and conservation groups, because more information can be obtained through this process than by a few active scientists. Management staff in Government departments need to become aware of the declines and actively participate in monitoring programs particularly on national parks.Water resource managers should incorporate potential impacts of proposed water extraction and impoundment in any planning. 4 Community monitoring of common species — the public can assist in obtaining information on the distribution and current status of species currently thought to be common, especially
those in areas on the coastal lowlands which are under pressure from sugar expansion programs. There has been a preoccupation with the rainforest species but work is urgently required in the lowland open forests. Baseline data on the distribution, habitat preferences, breeding and population fluctuations of lowland species is urgently needed.This will require surveys and monitoring of frog populations. 5. Intensive monitoring of sites designed to address quantitative population assessment. Intensive studies are needed to address population changes and recruitment through mark recapture and monitoring for population stability.This will enable sites less frequently visited to be placed in the context of known population fluctuations determined from intensive long-term studies.
ACKNOWLEDGMENTS Thanks to P. Brooke, J. A. Covacevich, A. Campbell, A. Dennis, A.B Freeman, M. Hero, D.House, M. Mahony, W.E. Martin, A. Moise, S.J. Richards, R. Retallick and volunteers and students who have assisted sometimes in the most arduous of field conditions. Funding has been provided by our respective institutions, Environment Australia, the Australian Research Council, the Wet Tropics Management Authority, and the CRC for Tropical Rainforest Ecology and Management. Steve Richards and Dr Marg Davies made constructive criticisms of earlier manuscript drafts.This assistance is most gratefully acknowledged.
REFERENCES Alford, R. A., Bradfield, K.S. and Richards, S.J., (1997) Predicting declines in rainforest frog populations. Abstract, CRCTREM annual conference, Cairns, Queensland. Alford, R. A., Bradfield, K.S. and Richards, S.J., (1999) Measuring and analysing developmental instability as a tool for monitoring frog populations. Pp34–43 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Alford, R. A., and Richards, S.J., (1997) Lack of evidence for epidemic disease as an agent in the catastrophic decline of Australian rain forest frogs. Conservation Biology 11:1026-1029. Altig, R. and Johnston, G.F., (1989). Guilds of anuran larvae : relationships among developmental modes, morphologies, and habitats. Herpetological Monographs 3 : 81-109. Atkinson, R., (1996) Proc. Conf. on Health Consequences of Ozone Depletion. Cancer Forum Vol. 20 No. 3 Spec. Ed.168-173. Berger, L., Speare, R., Dasak, P., Green, D.E., Cunningham, A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. and Parkes, H., (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rainforests of Australia and Central America. Proc. Natl. Acad. Sci. USA 95 : 9031-9036.
21
Blaustein, A.R., Hoffman, P.D., Hokit, D.G., Kiesecker, J.M., Walls, S.C. and Hays, J.B., (1996) UV repair and resistance to solar UV_B in amphibian eggs : A link to population declines? Proc. Natl. Acad. Sci., USA 91 : 1791-1795.
biogeographic region of northeast Queensland. (submitted to Environment Australia, Canberra).
Couper, P.J., (1992) Hope for our missing frogs. Wildlife Australia. 29(4): 11-12.
Martin, W.E. and McDonald. K.R., (1996) Recovery plan for the stream-dwelling rainforest frogs of the Eungella area, mid-east Queensland. (submitted to Environment Australia, Canberra).
Covacevich, J.A. and McDonald, K.R., (1993) Distribution and conservation of frogs and reptiles of Queensland rainforests. Memoirs of the Queensland Museum 34(1):189-199.
McDonald, K.R., (1990) Rheobatrachus Liem and Taudactylus Straughan and Lee (Anura: Leptodactylidae) in Eungella National Park, Queensland: distribution and decline.Trans. R. Soc. S. Aust. 114(4): 187-194.
Davies, M. and McDonald, K.R., (1979) A new species of stream-dwelling hylid frog from northern Queensland. Transactions of the Royal Society of South Australia 103(7): 169-176.
McDonald, K.R., (1992) Distribution patterns and conservation status of north Queensland rainforest frogs. Conservation Technical Report 1., Queensland Department of Environment and Heritage, Brisbane.
Davies, M. and Richards, S.J. (1990) Developmental Biology of the Australian Hylid Frog Nyctimystes dayi (Gunther). Trans. Roy. Soc. Sth Aust. 114: 207-212.
McDonald, K.R., Covacevich, J.A., Ingram, G.J. and Couper, P.J., (1991) The status of frogs and reptiles. Pp 338-345 In Ingram, G.J and Raven, R.J. (eds), ‘An atlas of Queensland’s frogs, reptiles, birds and mammals’. (Queensland Museum, Board of Trustees: Brisbane). 391pp.
Dennis, A., (1982) A brief study of the Sharp-snouted Torrent Frogs Taudactylus acutirostris. North Queensland Naturalist 50: 7-8. Dennis, A. and M. Mahony, M., (1994) Experimental translocation of the endangered sharp-snouted day frog Taudactylus acutirostris and observations of the cause of declines among montane riparian frogs. Unpublished report prepared for Wet Tropics Management Authority, Cairns. Hero M-J., (1996) Where are Queensland’s missing frogs? Wildlife Australia 33(3): 8-13. Ingram, G. and McDonald, K.R., (1993) An update on the decline of Queensland’s frogs. Pp 297-303 in Herpetology in Australia: a diverse discipline eds D. Lunney and D. Ayers. Transactions of the Royal Zoology Society of New South Wales. Mosman.
McDonald, K. R., and Tyler, M. J., (1984) Evidence of gastric brooding in the Australian leptodactylid frog Rheobatrachus vitellinus.Trans. R. Soc. S. Aust. 108: 226. McNellie, M. and Hero, J.M., (1994) Mission amphibian.The search for the missing rainforest frogs of Eungella. Wildlife Australia 31(4): 22-23. McPeters, R.D., Hollandsworth, S.M., Flynn, L.E., Herman,J.R., and Seftor. C.J. (1996) Long-term trends derived from the 16 year combined Nimbus 7 / Meteor 3 TOMS version 7 record. Geophys. Res. Let. 23: 3699-3702. Moore, J.A. (1961) The Frogs of Eastern New South Wales. Bull Amer. Mus. Nat. Hist. 121: 149-386.
Laurance, W. F., (1996) Why are Queensland’s frogs croaking? Australian Nature 25 (4) : 56-62.
Palmer, A. R, and Strobeck, C., (1986) Fluctuating asymmetry: measurement, analysis, patterns. Annual. Review of Ecology and Systematics. 17:391-421
Laurance, W.F., (1996) Catastrophic declines of Australian rainforest frogs: is unusual weather responsible? Biol. Conser. 77 : 203-212.
Richards, S.J., (1992) The tadpole of the Australian frog Litoria nyakalensis (Anura, Hylidae), and a key to the torrent tadpoles of northern Queensland. Alytes 10(3): 99-103.
Laurance, W.F., McDonald, K.R. and Speare, R., (1996) Epidemic disease and the catastrophic decline of Australian rainforest frogs. Conserv. Biol. 10(2) : 406-413.
Richards, S.J., McDonald, K.R., Alford, R.A., (1993) Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conservation Biology 1:66-77.
Liem, D.S., (1974) A review of the Litoria nannotis species group and a description of a new species of Litoria from north-east Queensland, Australia. Memoirs of the Queensland Museum. 17(1): 151-168.
Speare, R., (1995) Preliminary study on diseases in Australian Wet Tropics amphibians. Deaths of rainforest frogs at O’Keefe Creek, Big Tableland. Final Report to Department of Environment and Heritage. Unpublished Report QDEH.
Liem, D.S. and Hosmer, W., (1973) Frogs of the genus Taudactylus with description of two new species (Anura: Leptodactylidae). Memoirs of the Queensland Museum 16(3): 435-457.
Stanton, J.P. and Morgan, M.G., (1977) The rapid selection and appraisal of key endangered sites. The Queensland case study. Uni. of New England, School of Nat. Res. Report No. PR 4.
Madronich, S. and de Gruijl, F.R., (1993) Skin cancer and UV radiation. Nature 366 : 23.
Trenerry, M. P., Laurance, W. F., and McDonald, K. R., (1994) Further evidence for the precipitous decline of endemic rainforest frogs in tropical Australia. Pacific Conservation Biology 1: 150-153.
Mahony, M., (1996) The decline of the Green and Golden Bell Frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust. Zool. 30 (2) : 237-247. Marshall, C., (1998) The reappearance of Taudactylus (Anura:Myobatrachidae) in north Queensland streams. Pacific Cons. Biol. 4: 39-41. Martin, W.E. and McDonald. K.R., (1995) Recovery plan for the stream-dwelling rainforest frogs of the Wet Tropics 22
Tyler, M.J., (1991) Where have all the frogs gone? Aust. Nat. Hist. 23 (8) : 618-625. Tyler, M.J., (1997) The Action Plan for Australian Frogs. Environment Australia, Canberra. Winter, J. and McDonald, K.R., (1986) Eungella, the land of cloud. Australian Natural History 22(1): 39-43.
Chytrid fungi and amphibian declines: Overview, implications and future directions Lee Berger1, 2, Rick Speare2 and Alex Hyatt1
ABSTRACT A recently described chytrid fungus, genus Batrachochytrium, killed free-living and captive amphibians in Australia, Central America and the USA. There is epidemiological, pathological, and experimental evidence that some amphibian populations suddenly declined due to mass mortalities caused by chytridiomycosis.
exposure. Tadpoles appear to be unaffected by the fungus which infects their keratinised mouthparts. Batrachochytrium can probably also survive and grow in the environment. Based on the epidemiology of the amphibian declines, chytridiomycosis appears to be an emerging disease causing mortality in many species of anurans and has caused the disappearance and presumed extinction of some species. These 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 Panama. Chytrid fungi caused a widespread
Batrachochytrium, as well as rendering them less able
infection 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 Amphibian declines in some regions have been attributed to habitat disturbance including pollution, cattle damage, fish introduction and habitat destruction such as logging and wetland degradation (Hayes and Jennings 1986,Tyler 1997). However, habitat disturbance does not explain the rapid disappearance of high-altitude stream-dwelling rainforest amphibians from many protected areas in Australia (Richards et al. 1993, Mahony 1996) and Central America (Lips 1998). There was no correlation between frog population declines and changes in ground level solar UV-B radiation in Queensland (Moise, unpubl. data). Several factors in the declines indicated that a waterborne infectious disease, of high virulence to adults of some species, had entered a population previously unexposed to it. These factors were: 1) sudden, severe declines occurred over a few months; 2) declines were asynchronous and spread as a front; 3) adults died while tadpoles survived and metamorphs died when they subsequently emerged; 4) no environmental changes were detected; 5) only stream dwelling frogs disappeared and, 6) in two intensively monitored sites, mass mortalities were observed at the time of significant population declines (Laurance et al. 1996, Lips 1999,Trenerry et al. 1994). In these two montane rainforest locations — Big Tableland, Australia (1993) and Fortuna, Panama (1997) — sick and dying anurans (including Taudactylus acutirostris, Litoria rheocola and L. nannotis) were collected for pathological examination and found to be infected with chytrid fungi in the skin (Berger et al. 1998). This fungus has been placed in a new genus, Batrachochytrium (Longcore et al. 1999). Along with the epidemiological evidence above, we present pathological and experimental data that demonstrate that chytrids are pathogenic to amphibians. The pattern of the population declines is consistent with being caused by Batrachochytrium as it is waterborne, is virulent to adults, does not kill tadpoles (Berger et al. 1999), prefers cooler temperatures (Longcore et al. 1999), and is not dependent upon the highly susceptible species for its continued existence. Similar waves of mass mortalities, described as the post-metamorphic death syndrome, have been reported in various amphibian populations in western North America (Scott 1993). Although the cause(s) was not determined, chytridiomycosis has recently been discovered in populations of endangered north American frogs, including leopard frogs (Rana yavapiensis and R. chiricahuensis) in Arizona (Nichols et al. 1998, Morell, 1999) and chytrids were also seen as incidental findings in six percent of a group of wild cricket frogs (Acris crepitans) in Illinois (Pessier et al. 1999). Although Batrachochytrium has a broad amphibian host range and is currently widespread, not all susceptible species have declined. The selectivity of the declines may be due to a combination of environmental factors and host biology that provide the necessary conditions for expression of disease, as well as rendering species less able to recover after population crashes. Declining species from high altitude rainforests have restricted ranges and smaller clutch sizes (Williams and Hero 1998).
24
In this paper we collate the data on the amphibian chytrid and expand on previously presented hypotheses, with a focus on Australian circumstances.
BIOLOGY OF BATRACHOCHYTRIUM AND THE CHYTRIDIOMYCOTA The amphibian chytrid has been placed in a new genus, Batrachochytrium (Phylum Chytridiomycota, Class Chytridiomycetes, Order Chytridiales) and an isolate from a captive blue poison dart frog (Dendrobates azureus) that died at the National Zoological Park in Washington has been described as B. dendrobatidis (Longcore et al. 1999). The ultrastructural morphology, amphibian host and 18S rDNA sequence of Batrachochytrium show that it is distinctly different from other chytrid fungi (Berger et al. 1998, Longcore et al. 1999). Chytridiomycete fungi are a large and diverse group and have been found in almost every type of environment, including rainforests, deserts and arctic tundra (Powell 1993). They are frequently found in soil and water where they digest substrates such as chitin from insect cadavers, cellulose from vegetable matter, keratin from hair and skin, or pollen. These species function as important primary biodegraders and are possibly vital to the ecosystem. Others are parasites of insects, fungi, algae, plants and nematodes and a few of these cause significant disease (Barr 1990, Powell 1993). Powell (1993) discusses the significance and inherent value of chytridiomycetes and reviews the ability of parasitic species to cause disease. The onset of chytridiomycete parasitism of phytoplankton is often correlated with a rapid decline in host population and so has a major impact on the ecology of the host. Synchytrium endobioticum causes black wart disease of potatoes in Europe and Canada, and was introduced to the USA in the early 1900’s but has since been eradicated. Coelomomyces has been considered for use in biological control of mosquitoes. Apart from species found among the normal rumen flora of ruminants, chytridiomycetes have not been found in vertebrates other than amphibians (Barr 1990, Berger et al 1998). Sparrow (1960) describes the evanescent nature of chytrid epidemics, with their sudden appearance, brief period of rapid multiplication and then decline and disappearance.This pattern is related to their virulence, ability for rapid reproduction, and the loss of optimal environmental conditions. Factors affecting the epidemiology of chytrid blooms include seasonal temperature changes, water pH, light, nutrition and dissolved oxygen (Sparrow 1968). These may be relevant considerations when attempting to isolate Batrachochytrium from the environment and when investigating the causes of outbreaks of chytridiomycosis. For example, epidemics in populations of Litoria caerulea in southern Queensland and northern NSW occurred in the winters of 1996, 1997 and 1998, demonstrating seasonal regularity (Table 1). Findings from studies of other aquatic zoosporic fungi may be pertinent here. The abundance of Saprolegniaceae in California was correlated with altitude (Sparrow 1968). As aquatic phycomycetes are probably very sensitive to contaminants they are considered good biological indicators of pollution (Sparrow 1968).
Most chytrids (i.e. members of the order Chytridiales) occur in aquatic habitats. They have motile flagellated zoospores which develop within a stationary sporangium. Sporangia of 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 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). Some chytrids have a thick walled, resistant resting spore stage which can survive for decades in extreme conditions (Powell 1993) but such a stage has not been observed in Batrachochytrium (Longcore et al. 1999), which may be a relatively fragile species. Culture media for the amphibian chytrid contained tryptone, gelatin hydrolysate and lactose (Longcore et al. 1999). In culture B. dendrobatidis developed most rapidly at 23C and grew at 28C, but did not grow significantly at 29C (Longcore et al. 1999). Cultures grew well at 15C and survived for more than three months at 4C (Longcore, unpubl. data). Chytridiomycetes do not generally survive freezing well, although some success with storage in liquid nitrogen has been achieved (Hohl and Iselin 1986). Species without resting spores are less able to be preserved in an inactive state (Hohl and Iselin 1986). No significant ultrastructural morphological differences were observed between isolates from Australia, the USA and Central America (Longcore et al. 1999, Berger et al. 1998) and DNA comparisons are needed to determine how many amphibian chytrid species exist. It is likely all isolates belong to a single species. The 18S rDNA sequence of chytrids from a wild caught Australian L. caerulea and a captive American D. azureus had only five base pairs different out of about 1700 bp sequenced, and four of these differences were deletions which may be due to error (James, Porter and Longcore, unpubl.). Preliminary sequencing of a more variable region, the rDNA internal transcribed spacers (ITS), demonstrates that similar strains (99% mortality rate of myxomatosis was not sufficient to exterminate rabbits from Australia (Fenner and Ratcliffe 1965), it is plausible that a similar mortality rate could wipe out frog species with limited distributions and relatively infrequent breeding. The introduction of avian malaria is suspected to have caused the extinction of birds in Hawaii (Warner 1968). A protozoan parasite of cats (Toxoplasma gondii) was probably introduced to Australia during the European invasion, and marsupials are among the most susceptible animals (Reddacliff et al. 1993). Phytophthora cinnamomi is an example of a pathogenic, introduced zoosporic fungus which threatens many native Australian plant species, and quarantine measures are recommended to prevent the invasion of the pathogen into new areas. Some plant species are highly susceptible, whereas others only become diseased after periods of stress such as a drought (Dawson and Weste 1985;Wills 1993). Infrastructure exists to prevent and manage exotic disease outbreaks in domestic animals, but little concern is shown for wildlife where many diseases are yet to be discovered and understood, and monitoring the disease status of populations currently appears to be the responsibility of no one.
Mycological studies on the chytrid are needed to learn more about its lifecycle, requirements and survival in the wild. Knowledge about the ecology and hosts of the fungus is essential to understanding the spread, and therefore to the management of the disease. Information about the particular conditions that encourage growth may help in understanding the factors which precipitate disease epidemics.
FUTURE DIRECTIONS Work on amphibian chytridiomycosis must continue — to confirm or reject the hypothesis that it is the primary cause of the declines, to determine how the epidemic began, to find ways to manage the problem in areas where the fungus is established, and to prevent it occurring in new regions. Knowledge gained from these investigations will be useful in preventing similar population crashes occurring in frogs again, and in developing management strategies for other wildlife species. We will continue to map the temporal and geographic 30
Further transmission experiments are required to confirm the pathogenicity in a range of species, and also to determine what environmental conditions e.g. temperature, are required for expression of the disease in frogs. Treatments for adults and tadpoles are also being tested which will aid captive breeding of endangered species. By producing large numbers of frogs in captivity, it may be possible to help species to survive and evolve immunity. Studies of the pathogenesis of chytridiomycosis are important to understanding this disease, and immunologists in the USA will investigate the innate and acquired immune response of frogs to Batrachochytrium. Further work will be done using DNA analysis to compare chytrids from various species from localities across Australia, Central America and the USA. The number of species of Batrachochytrium can be determined by using this information combined with morphological taxonomy. Molecular biology as a tool for molecular epidemiology can also be used to provide clues about the origins and spread of the fungus. Data are being collected to evaluate diagnostic tests. Histology and examination of skin scrapings are highly specific tests, but may not be very sensitive when used to detect chytrids on healthy specimens. Production of antibodies has commenced to enable more sensitive testing, and perhaps for use in detecting chytrids in the environment. Regulations regarding quarantine, testing, treatment and movement of amphibians need to be introduced to prevent further spread of Batrachochytrium within Australia and internationally. Although our understanding of the role of chytridiomycosis in amphibian declines is far from complete, we believe it is crucial to take immediate preventative measures rather than risk waiting for more scientific data to be accrued.
ACKNOWLEDGEMENTS Amphibian disease investigations would not have been possible without the tremendous support of herpetologists throughout Australia. A huge collaborative effort was involved in collecting
sick frogs from the field and urgently forwarding them to the lab.We are indebted to many people who have been involved, including those who provided sick or healthy specimens, allowed us access to their collections or gave advice, especially Gerry Marantelli, Harry Hines, Keith McDonald, Mike Tyler, Ken Aplin, Alastair Freeman, Craig Williams, Rebecca Short, Graeme Gillespie, Nick Sheppard, David Page, David Byrnes, Steve Williams, Craig Taylor, D.A. Stewart, Michael Mahony, Lance Tarvey, Michael Smith, Richard Retallick, Marc Hero, John Clarke, Michael Healey, Brent Dadds, Lothar Voigt, Adrian Wayne, Jeremy Morante, Danny Wotherspoon and members of the public.We are very grateful to Lee Skerratt, Deborah Middleton and Murray Littlejohn for comments on the manuscript. Thanks to Frank Fillipi for photography, Julia Hammond for bacteriology, and Megan Braun, Gail Russell,Terry Wise and Andrew Kent for the many tissue sections prepared. Peter Hooper and Mark Williamson are thanked for their support and pathology expertise.We also thank people who allowed us to cite their unpublished data. Lee Berger has been supported by a grant from Environment Australia.
REFERENCES Aplin, K., Speare, R., Berger, L. and Cowan, M. A., (1999) Outbreak of fungal disease in Western Australian Frogs. J Royal Soc WA (submitted). Barr, D. J. S., (1990) Phylum Chytridiomycota. Pp. 454-466 in Handbook of Protoctista. ed by L. Margulis, J. O. Corliss, M. Melkonian and D. J. Chapman. Lubrecht and Kramer, Monticello, New York. Bartnicki-Garcia, S. and Sing,V. O. (1986) Adhesion of zoospores of Phytophthora to solid surfaces. Pp. 279-283 in Zoosporic Fungi in Teaching and Research. ed by M. S. Fuller and A. Jaworski. Dept of Botany, University of Georgia, Athens GA. Berger, L. Speare, R. Daszak, P., Green, D. E, Cunningham, A. A., Goggin, C. L., Slocombe, R. Ragan, M. A. Hyatt, A. D., McDonald, K. R. Hines, H. B., Lips, K. R., Marantelli, G. and Parkes, H., (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rainforests of Australia and Central America. Proc Natl Acad Sci, 95: 9031-9036. Berger, L., Speare, R. and Kent, A., (1999) Diagnosis of chytridiomycosis in amphibians by histologic examination. Proc Frog Symposium: Frogs in the Community, Queensland Museum, Brisbane, February 1999 (submitted). Daszak, P.1, Berger, L., Cunningham, A. A., Hyatt, A. D., Green, D. E. and Speare, R., (1999) Role of Emerging Infectious Diseases in Amphibian Population Declines and Global Implications. Emerg Inf Dis, (in press). Dawson, P. and Weste, G., (1985) Changes in the distribution of Phytophthora cinnamomi in the Brisbane Ranges National Park between 1970 and 1980-81. Aust J Bot, 33: 309-315. Dobson, A. P and Carper, E. R., (1996) Infectious disease and human population history. Bioscience, 46: 115-126. Fenner, F. and Ratcliffe, F. N., (1965) Myxomatosis, Cambridge University Press, New York. Groff, J. M., Mughannam, A., McDowell,T. S., Wong, A. Dykstra, M. J., Frye, F. L. and Hedrick, R. P., (1991) An epidemic of cutaneous zygomycosis in cultured dwarf African clawed frogs (Hymenochirus curtipes) due to Basidiobolus ranarum. J Med Vet Mycology, 29: 215-223.
Hayes, M. P. and Jennings, M. R., (1986) Decline of ranid frogs species in Western North America: Are bullfrogs (Rana catesbiana) responsible? J Herpetol, 20: 490-509. Hero, J-M., Hines, H., Meyer, E., Morrison, C., Streatfeild, C. and Roberts, L., (1998) New records of “declining” frogs in Queensland, Australia (-February 1998). Froglog 29. Hillman, S. S., (1988) Dehydrational effects on brain and cerebrospinal fluid electrolytes in two amphibians. Physiol Zool, 61: 254-259. Hohl, H. R. and Iselin, K., (1986) Liquid nitrogen preservation of zoosporic fungi. Pp 143-145 in Zoosporic Fungi in Teaching and Research. ed by M. S. Fuller and A. Jaworski. Dept of Botany, University of Georgia, Athens GA. Laurance, W. F., McDonald, K. R. and Speare, R., (1996) Epidemic disease and the catastrophic decline of Australian rainforest frogs. Conserv Biol, 77: 203-212. Leberg, P. L. and Vrijenhoek, R. C., (1994) Variation among desert topminnows in their susceptibility to attack by exotic parasites. Conserv Biol, 8: 419-424. Lee, S. H., Lai, S. T., Lai, J.Y. and Leung, N. K., (1996) Resurgence of cholera in Hong Kong. Epidemiol Infect, 117: 43-49. Lips, K. R., (1998) Decline of a tropical montane fauna. Conserv Biol, 12: 106-117. Lips, K. R., (1999) Mass mortality and population declines of anurans at an upland site in western Panama. Conserv Biol, 13: 117-125. Longcore, J. E., Pessier, A. P. and Nichols, D. K., (1999) Batrachochytrium dendrobatidis gen. et sp. nov., a chytrid pathogenic to amphibians. Mycologia, 91: 219-227. Mahony, M., (1996) The decline of the green and golden bell frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust Zool, 30: 237-247. Marantelli, G., (1999) Husbandry: science or art ? — Are captive technologies ready to contribute to recovery processes for Australian frogs? Pp 168-176 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia, Canberra. May, R. M., (1988) Conservation and Disease. Conserv Biol, 2: 28-30. McDonald, K.R. and Alford, R.A., (1999) A review of declining frogs in northern Queensland. Pp 14-22 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia, Canberra. Morell,V. (1999) Are pathogens felling frogs? Science, 284: 728-731. Nichols, D.K., Pessier, A.P., and Longcore, J.E., (1998) Cutaneous chytridiomycosis: an emerging disease? Proc Am Assoc Zoo Vet, 1998:269-271. Pessier, A. P. Nichols, D. K., Longcore, J. E. and Fuller, M. S., (1999) Cutaneous chytridiomycosis in poison dart frogs (Dendrobates spp.) and White’s tree frogs (Litoria caerulea). J Vet Diag Invest, 11: 194-199. Powell, M. J., (1993) Looking at mycology with a Janus face: A glimpse of chytridiomycetes active in the environment. Mycologia, 85: 1-20. Reddacliff, G.L., Hartley, W.J., Dubey, J.P. and Cooper, D.W., (1993) Pathology of experimentally-induced, acute toxoplasmosis in macropods. Aust Vet J.70: 4-6. 31
Retallick, R., (1999) Translocations and experimental ecology of declining frogs. Update for the North Queensland Frog Recovery Team, August 1999. Unpublished report. Richards, S. J., McDonald, K R. and Alford, R. A., (1993) Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conserv Biol, 1: 66-77. Scott, M. E., (1988) The impact of infection and disease on animal populations: Implications for conservation biology. Conserv Biol, 2: 40-56. Scott, N. J., (1993) Post metamorphic death syndrome. Froglog, 7: 1-2. Sparrow, F. K., (1960) Aquatic Phycomycetes. 2nd re. ed. University of Michigan Press, Ann Arbor, Michigan. Sparrow, F. K., (1968) Ecology of Freshwater Fungi. pp 41-93 in The Fungi. ed by G. C. Gainsworth and A. S. Sussman, Academic Press, New York. Speare, R., (1994) Preliminary study on diseases in Australian Wet Tropics amphibians. Deaths of rainforest frogs at O’ Keefe Creek, Big Tableland. Final report to Queensland Department of Environment and Heritage. Unpublished report, Queensland Department of Environment and Heritage. Taylor, S. K., Williams, E. S.,Thorne, E. T., Mills, K. W., Withers, D. I. and Pier, A. C., (1999) Causes of mortality of the Wyoming toad. J Wildl Dis, 35: 49-57. Trenerry, M. P., Laurance, W. F. and McDonald, K. R., (1994) Further evidence for the precipitous decline of endemic rainforest frogs in tropical Australia. Pacific Conserv Biol, 1: 150-153. Tyler, M. J., (1997) The Action Plan for Australian Frogs. Wildlife Australia, Canberra. Viggers, K. L., Lindenmayer, D. B. and Spratt, D. M., (1993) The importance of disease in reintroduction programmes. Wildl Res, 20: 687-98. Warner, R. E., (1968) The role of introduced diseases in the extinction of the endemic Hawaiian avifauna. The Condor, 70: 101-120. Williams S. E. and Hero J-M., (1998) Rainforest frogs of the Australian Wet Tropics: guild classification and the ecological similarity of declining species. Proc Roy Soc Lond, B 265: 597-602. Wills, R. T., (1993) The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Aust J Ecol, 18: 145-159.
32
APPENDIX 1 What herpetologists can do to assist SURVEY OF SICK AND HEALTHY FROGS We wish to examine any diseased frogs or cane toads that are found in order to determine the cause of death and to screen for the presence of chytrids. Frog tissues deteriorate very rapidly after death, so if a sick frog is found that is likely to survive another 24 hours, it should be sent by courier to the Australian Animal Health Laboratory or James Cook University after contacting us. Frogs or tadpoles found dead should be fixed or frozen immediately to preserve the tissues. They should be fixed in 10% buffered neutral formalin, but 70% ethanol can also be used. It is important to slit open the belly, and to ensure the frog is well covered in fixative so that tissues are preserved rapidly. Details of what to do with sick or dead frogs have been posted on the World Wide Web at http://www.jcu.edu.au/dept/school/phtm/ PHTM/frogs/pmfrog.htm. Please send collection data with any frog submitted, and we will keep you informed about the results of the post mortem. Pathology is required for diagnosis, as the clinical signs of chytridiomycosis are not highly specific. We have prepared a frog mortality questionnaire (http://www.jcu.edu.au/school/phtm/PHTM/frogs/pmques.htm) which details the type of data that are important to observe and record if you encounter a mass mortality event. For our examination of archived frogs for chytrids, we need skin samples from amphibians from a wide range of localities and dates. We want skin from the pelvic areas and toes from any formalin-fixed or ethanol-fixed frog that has collection data. We especially require frogs from inland Australia, Northern Territory and northern WA, as none have been examined from these regions. A protocol is available at http://www.jcu.edu.au/school/phtm/frogs/pmskin.htm. People doing skeletochronology on histological sections of toes could simultaneously check the skin for chytrids (see Figure 4). In healthy frogs, the level of infection may be very low, with only occasional sporangia present along the skin surface. For detailed diagnostic histological features, see Berger et al. (1999). We are attempting to maintain a comprehensive list 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.
MANAGEMENT OF CHYTRIDIOMYCOSIS IN CAPTIVITY
PREVENTING SPREAD OF DISEASE IN THE WILD
If any epidemics of chytridiomycosis occur in captive collections, various antifungal drugs could be administered, and the results communicated.
To prevent the spread of chytrids or other diseases when performing field work, disinfection of equipment should be performed. We need more information on the resistance of this fungus to heat, desiccation and disinfectants; so at present are recommending measures for disinfection that have been proven against highly resistant organisms. (see protocol — http://www.jcu.edu.au/school/phtm/PHTM/frogs/prevent.htm).
Benzalkonium chloride is a disinfectant that has been used at 2 mg/l to successfully treat a similar superficial mycotic dermatitis in dwarf African clawed frogs (Hymenochirus curtipes) reported to be caused by Basidiobolus ranarum (Groff et al. 1991). The regime used experimentally was 30 minutes of bath treatment, on three alternate days. This was repeated in 8 days (i.e. 6 treatments in total). Oral itraconazole has also been used to treat B. ranarum infections (Taylor et al. 1999). One micro bead from 100mg itraconazole capsules was administered daily for 9 days to Wyoming toads (Bufo baxteri) at the first signs of disease. Benzalkonium chloride (1mg/L), amphotericin and fluconazole are effective against Batrachochytrium in vitro (Berger, unpubl.). In captivity, routine quarantine procedures (Marantelli 1999) have been adequate in restricting outbreaks to certain tanks, and no airborne transmission has been observed (Marantelli, unpubl.). Each group of frogs should be kept completely separate to ensure no water borne transmission of disease can occur. By changing and discarding gloves between every tank, avoiding splashing water between tanks, and disinfection of tanks and implements before reuse using 2% hypochlorite, many frogs have been housed in close proximity without transmission of disease.
Disease status should always be considered when translocating animals (Marantelli 1999,Viggers et al. 1993) and attempts should be made to reduce the chance of introducing disease to a naïve population. We recommend the screening of healthy frogs by histologic examination of toe clips, or by sacrificing a few in the group for more extensive skin examination (Berger et al. 1999). Testing for chytrids on sick frogs is more sensitive, so any deaths in a valuable group of animals should be submitted to a pathology laboratory for testing. To screen a group of healthy tadpoles, some need to be sacrificed so that their mouths can be examined histologically (see Berger et al. 1999). We have little information on the sensitivity of these tests, so it is impossible to recommend a statistically significant number of animals. Also, until we have more data on the distribution of chytrids in amphibian populations around Australia, it will be difficult to make decisions about the release of infected animals. New regulations are being proposed to control the 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 genotype to repeatedly produce precisely the same phenotype when exposed to the same environment during development (Zakharov 1992;Tracy et al. 1995; Moller 1996). Developmental stability is produced by the action of complex feedback mechanisms which ensure that random environmentally-induced deviations from the genetically determined phenotype are corrected as development progresses (Zakharov 1992; Graham et al. 1993a; Graham et al. 1994; Clarke 1995). These deviations are termed “developmental noise” (Palmer and Strobeck 1986; Graham et al. 1993a). It is thought that developmental stability is highest when well-integrated genomes are expressed in benign environments (Leary and Allendorf 1989; Graham 1992). Poorly integrated genomes or poor-quality environments can interfere with the ability of organisms to correct for the effects of developmental noise, leading to greater variability in the phenotypes produced. This increased variation is termed developmental instability (DI) (Palmer 1996).
measured by examining the degree of asymmetry of bilateral structures, increase as stress increases or health decreases in many species. Several studies have shown that DI levels increase before or during population declines. Increases in population DI levels may therefore indicate that populations are at risk of decline. This suggests that DI measurements should be incorporated in population monitoring programs. Because these techniques are relatively new, they have not been commonly used, and may be inaccessible to many field workers.We discuss how to measure DI in frogs, using blind measurement techniques that reduce the possibility of observer bias and maximise measurement precision.We also provide examples of data analysis, interpretation and presentation. The approach we outline will allow field biologists to incorporate DI measurement into their research programs.
Because the levels of developmental instability in populations of animals should reflect their genetic makeup and the quality 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; Leary and Allendorf 1989; Parsons 1990a, b; Freeman et al. 1996; Clarke 1994) that measurements of the degree of DI existing in populations should provide a sensitive indicator of their general health. Studies on a variety of taxa have shown that declining populations (Tsubaki 1998) or populations in stressful environments (Valentine et al. 1973; Moller 1996; Sarre 1996; Siimäki and Lammi 1998) exhibit increased levels of DI. Measuring DI requires the investigator to determine how far the phenotypes of a series of individuals deviate from those they would have had if their development had proceeded under ideal conditions. For most phenotypic attributes of most animals, the ideal state that would be produced by any particular genotype is unknown, making developmental instability difficult or impossible to measure. One class of features, however, have well-defined ideal states: Bilateral features of species that are normally bilaterally symmetrical should be perfectly symmetrical. Because of this, developmental stability analysis usually involves the examination of departures of traits from symmetry (Graham et al. 1993b).
populations of two frog species (Litoria nannotis and L. genimaculata) experienced large and prolonged increases in DI commencing nearly two years before population declines became apparent through standard monitoring techniques. Similar results were found for fruit flies by Tsubaki (1998). Alford et al. suggested that monitoring of DI levels should be incorporated into routine population monitoring for many species, because increases in DI may act as an early warning signal of stresses that may later lead to population declines, giving time to investigate and ameliorate the causes of stress before they adversely affect the population biology of species.
Most features of normally bilaterally symmetrical animals exhibit some degree of asymmetry, which can be of three types (Palmer and Strobeck 1986): directional asymmetry occurs when the mean of the left-right differences in sizes of bilateral structures is not equal to zero. Directional asymmetry is an element of the genetic and developmental program of certain structures in many species. A common example is the heart of mammals, in which the left side is normally larger than the right (Graham et al. 1993a). Characters that are directionally asymmetrical are not suitable for developmental instability analysis because their ideal state cannot be certainly known. Antisymmetry occurs when a structure is less than perfectly symmetrical in the majority of individuals, but in any one individual, either side is equally likely to be the larger one. This leads to a bimodal distribution of left-right differences in structure size, with a mean of zero. Antisymmetry is a part of the normal developmental program of some structures in some species. An example is the chelae of some species of lobsters (Palmer and Strobeck 1986), in which it is normal for the claw on one side to be larger than the other, but which side has the larger claw is determined by random stimuli. Antisymmetry may also occur when normal developmental processes, such as feedback systems controlling the allocation of resources between body parts during development, fail to function properly (Graham et al. 1994; Emlen et al. 1994). The causes of this failure may be either genetic or environmental. Environmentally-induced antisymmetry can be used to measure developmental instability, because the ideal phenotype, perfect bilateral symmetry, is known. The final form of asymmetry is fluctuating asymmetry (FA), in which the population mode and mean is perfect symmetry of bilateral structures, with left-right differences normally (or at least unimodally and symmetrically, Palmer and Strobeck 1986) distributed with a mean of zero. It is generally acknowledged that levels of FA provide a good index of levels of DI within populations (Palmer and Strobeck 1986; 1992).
• It must be as immune as possible to measurer bias.
Measurements of DI may provide a more sensitive indication of the general health of animal populations than can be obtained through routine monitoring of population sizes. Data collected by Alford et al. (1997) demonstrated that
The aim of this paper is to provide guidelines for measuring and analysing DI using techniques which should be practical to incorporate into regular monitoring programs for many species of frogs. Our technique for measuring DI is designed to meet the following criteria: • It must produce highly precise measurements. Differences of 1% or less in the sizes of structures can be important in DI analysis and, since the legs of frogs are small structures, precision on the order of 0.1 mm is critical for useful results.
• It must be possible to carry out the selected measurements on irreplaceable museum specimens and on living animals, so the measuring technique must not be invasive or destructive.
MEASURING METHODS Structures measured and effects of body size on measurements We recommend measuring the lengths of the most easily accessible long bones on the forelimbs and hindlimbs, plus adjoining tissue: for the forelimbs, this is the radio-ulna, and for the hindlimbs, the tibio-fibula. The forelimb measurements are made from the ‘elbow’ to the palm of the flexed ‘hand’. The hindlimb measurements are easier to make on live animals, as they are from the knee to the ankle, and can be made on most frogs very conveniently while they are in a normal resting posture. It is also possible to measure the length of the upper tarsal bones on the hindlimb, from the ankle to the base of the toes. The precise starting and ending locations of these measurements are not as important as is consistency in the choice of starting and ending locations. One possible problem in measuring and interpreting amounts of asymmetry is that the measured values of left-right differences will be correlated with the sizes of the parts measured. Comparing or combining asymmetry measurements taken in absolute units such as millimeters from animals over a wide range of body sizes can present difficulties. It may be possible in some cases to correct for size effects using standard morphometric techniques (see below and Palmer and Strobeck, 1986). A simpler approach is to take all measurements on animals of similar sizes. Fortunately, this is easily accomplished for most frog species, as the most commonly encountered individuals are adult males, which tend to occur in a limited range of body sizes. If animals measured vary by more than +/- 10% or so in body size, potential size effects should be checked using 35
scatterplots and correlation analyses comparing unsigned asymmetry with an overall linear size measurement such as snout-vent length (SVL). If a correlation is found, it may be possible to correct for it by performing subsequent analyses on residuals of regressions of asymmetry against body size, or by converting individual measurements to percentages of mean size for left and right structures (Palmer and Strobeck 1986).
The basic measurement protocol When measurements are carried out on whole animals rather than bones, there are likely to be measurement errors, particularly when the animals are alive. Results might also be influenced by the measurers’ expectations. To minimise these effects, we devised the protocol that follows. First, the SVL (snout-vent length) and the fore- and hindlimbs on the left side of the body are measured to the nearest 0.05mm, using calipers. These measurements are recorded along with an identifying number for each individual (if measuring animals that are not marked and will not be marked, house them in individually numbered containers such as click seal bags for the duration of measurements). This step is repeated for each series of individuals being measured. Second, taking care to choose animals in random sequence, and without data on left sides visible, the limbs on the right side of each animal’s body are measured and those data recorded. Ideally, the random sequence should be generated using random numbers. If this is not possible, care should be taken to ensure that the sequence is thoroughly mixed between repeat measurements, for example by physically shuffling animals or their containers. The two steps above are carried out a total of three times for each animal. Each new set of measurements is taken without reference to previously taken sets. Ideally, the three sets of measurements should be recorded on separate data sheets, so that seeing the values of previous measurements cannot possibly influence the measurer. This protocol is single-blind and, if carried out carefully by a conscientious operator, should ensure that precision is maximised by the repeated measurements, while measurer bias is virtually eliminated.
calipers and then hands them to the recorder, who lifts the tape, reads the measurement, and records it, then resets the calipers to zero. This eliminates the possibility of bias, since the measurer cannot set the calipers to previous measurements as he or she does not know their values. By measuring the limbs in sequence around the animal, each replicate measurement of each limb is taken as far as possible in the sequence from other replicate measurements of the same limb. This technique provides a basis for calculating levels of measurement error that should be as close as is practical to the technique outlined above. Another technique that works well in the laboratory but which we have not tried in the field is the use of digital electronic calipers with data outputs. This has the advantage that a single person can do measurements without a helper. A piece of opaque tape or other object is secured over the readout on the calipers so that the measurer cannot see the values obtained. The data are fed directly into the recording computer by activating the data output of the calipers after setting them for each measurement, in a sequence similar to that outlined above. The calipers are reset to zero after each measurement is sent to the computer. Again, this technique provides good estimates of the level of measurement error and eliminates the possibility of bias on the part of the person performing the measurements.
Variation among observers Data obtained by different observers may differ slightly. Measurements of asymmetry are relatively robust to small differences in technique or in judgement regarding the locations of landmarks for measurement. This is because the final item of interest is not the absolute sizes of structures, but the magnitude of differences between sides. Asymmetry measurements are sensitive to lack of care in measurement technique. All observers must therefore strive for precision in their measurements. The use of replicate measurements of each structure, as we recommend, will also increase precision of the final mean value. Observers should be carefully trained and supervised, and the possibility of observer effects should be taken into account when evaluating data sets collected by more than one observer.
DATA ANALYSIS Modifications to the basic protocol When measuring living animals in the field, it is important to minimise the amount of handling and the time that animals spend in bags, and to release animals at the point of initial capture after as short a time delay as possible. One technique that preserves the “blind” nature of replicate measurements but allows animals to be measured and released immediately after capture, is to work in two-person teams, which is also desirable for safety reasons any time field work is being performed in remote and potentially hazardous locations. One person acts as the measurer, and the other reads the calipers and records the data. A piece of opaque tape is used to cover the area of the calipers from which readings are taken. The measurer works around the animal three times, measuring the left fore- and hindlimbs and then the right fore- and hindlimbs each time. For each measurement, the measurer handles the animal and sets the
36
We use data analysis techniques similar to those outlined by Palmer and Strobeck (1986), with some modifications. This section presents the techniques used and sample outputs, using an artificial example data set we constructed. The data in that artificial data set are very similar, in levels of asymmetry and patterns of asymmetry, to data collected on real populations of rainforest frogs (Alford et al. 1997).
Within-sample tests of significance and examination of variance components We suggest initially analysing data using the ANOVA plus extraction of variance components technique detailed by Palmer and Strobeck (1986, pages 402-408). This allows separate examination of variance due to directional asymmetry, FA + antisymmetry, and measurement error. The ANOVAS can be used to test for whether there are significant levels of directional asymmetry, and for whether
the combination of FA + antisymmetry measured is significantly above the level produced by measurement error. Using the techniques suggested by Palmer and Strobeck, the mean squares for FA+antisymmetry and for error can be used to estimate a population index of FA. To perform these analyses, the data must be laid out as illustrated in Table 1. Separate fully-specified analyses of variance are performed for each character measured (forelimb length and hindlimb length) in each sample. TABLE 1: Format of data for analysis using modification of the ANOVA approach of Palmer and Strobeck (1986). Sample refers to the date or place of sampling; in this shortened example, all three animals are from the same sample. Individual number can be either sample-specific as in this table or can be any unique individual identifier, as from a toe-clipping or PIT tagging program. Replicate indicates which of the three replicate measurements of each limb is presented, side =1 for left, 2 for right (any coding could be used).
Sample Individual Replicate SVL Side Forelimb Hindlimb number length length 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3
1 1 2 2 3 3 1 1 2 2 3 3 1 1 2 2 3 3
47.7 47.7 47.7 47.7 47.7 47.7 44.9 44.9 44.9 44.9 44.9 44.9 48.1 48.1 48.1 48.1 48.1 48.1
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
11.7 11.5 11.7 11.4 11.8 11.5 11.4 11.4 11.4 11.4 11.3 11.3 12 12.1 12 12 12 12.1
15.8 16.1 15.8 16.2 15.8 16.1 14.4 14.3 14.4 14.3 14.3 14.3 16.6 16.1 16.6 16.3 16.6 16.1
TABLE 2: Analysis of variance table for data on forelimb asymmetry of 30 individuals in sample 1 of the example data set. Abbreviations: IND_NUM = individual number, REP = replicate measurement number.
Factor
DF
Sum of squares
Mean square
SIDE IND_NUM SIDE*IND_NUM REP SIDE*REP IND_NUM*REP SIDE*IND_NUM*REP
1 29 29 2 2 58 58
0.0642 60.9591 2.4491 0.0034 0.0081 0.1632 0.1186
0.0642 2.1020 0.0845 0.0017 0.0041 0.0028 0.0020
These ANOVAs must extract effects due to side of the body, individual, the interaction of side with individual, replicate, the interaction of side with replicate, the interaction of individual with replicate, and the three-way interaction of side, individual, and replicate. The ANOVAs are performed to construct the tables of sums of squares and mean squares needed to construct tests of hypotheses and estimates of FA+antisymmetry, not for testing hypotheses. 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 commands for SPSS version 6.1. These files are also available from the authors , as is a full set of sample data for testing the command files or testing other statistical packages, and full outputs from both SAS and SPSS. The results of an analysis of forelimb length in sample 1 of our example data set appear in Table 2. Most of the sums of squares and mean squares are not useful, although finding large effects of individual, for instance, might suggest that variation among individuals in body size is such that one of the techniques suggested by Palmer and Strobeck (1986) should be used to correct for body size effects, and the analysis should be redone on the corrected data. The important variance components extracted are the effects of side and the interaction of side with individual, which are related to the magnitudes of directional asymmetry and FA+antisymmetry, respectively, and the interaction of side, individual, and replicate, which serves as an error term for tests of the hypothesis that the FA+antisymmetry effect is not produced by measurement error. The mean squares for these components can be used to construct tests of significance as presented in Table 3. The SAS program in Appendix 1 does this automatically, while the SPSS code (Appendix 2) leaves it to the analyst to construct the tables by hand. Table 3 presents results of the tests of the hypotheses that 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 asymmetry (P = 0.3903), while there is a substantial degree of FA+antisymmetry (P < 0.0001). If the test for directional 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 of asymmetry cannot be used in any simple way as an index of DI, as the ideal condition (the ideal degree of asymmetry) cannot be known. Because the error term for the test of FA+antisymmetry is the interaction of the effects of side, individual, and replicate (= measurement error), the significant test for FA+antisymmetry indicates that measurement error 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) SIDE*IND_NUM (FA + antisymmetry)
1 29
0.0642 0.0845
SIDE*IND_NUM SIDE*IND_NUM*REP
29 58
0.0845 0.0020
0.76 41.32
0.3903 0.0001
37
would suggest that the species being examined has levels of FA+antisymmetry so low as to be indistinguishable from measurement error in the sample of individuals being examined. This result would not, however, necessarily indicate that DI analysis should not be incorporated in monitoring for this species; it might simply indicate that, at the present time in the population of animals measured, very little DI is being expressed. This might simply indicate that the population is healthy and experiencing little environmental stress. After performing significance tests, an index of FA+antisymmetry can be estimated using the following equation (Palmer and Strobeck 1986; page 406): FA+antisymmetry = [(mean square for side X individual number interaction) — (mean square for side X individual number X replicate interaction)] / number of replicate measurements. For the analysis presented in Table 2, this would be: FA+antisymmetry = (0.0845 — 0.0020) / 3 = 0.0275. This indicates that the mean squared asymmetry of the forelimbs of animals in sample 1, with differences caused by measurement error extracted, is 0.0275 mm. Taking the square root of this, the mean unsigned asymmetry, corrected for measurement error, is estimated as 0.166 mm.
Among-sample tests for significant differences in levels of FA+antisymmetry 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 populations, or comparisons of levels of expressed DI among populations. Such tests can be constructed using the same data employed for the within-sample tests and estimates discussed above. Code for testing for differences among samples (or calculating the MS necessary to construct tests for differences among samples) is incorporated in the SAS and SPSS programs presented in Appendices 1 and 2. A complete table of the analysis of variance carried out to determine whether levels of FA+antisymmetry of forelimbs differ between samples in our example data set is presented in Table 4. Although it can be interesting to look at the magnitudes of the variance components in this table, they do not mean a great deal, particularly if the data are unbalanced, that is, if different numbers of individuals were measured in the different samples. The F-statistic to test for differences between samples is calculated as: Fx,y = (mean square for the interaction of sample, side, and individual)/(mean square for the interaction of sample, side, individual, and replicate measurement). The x and y are the numerator and denominator degrees of freedom, respectively. For our example data set, this leads to: F29,58 = 0.1368/0.0020 = 68.4. The probability of an F as great or greater than this can be looked up in an F-table or can be calculated using a variety of available programs (including the FDIST(F,x,y) function of Microsoft Excel). In this case, it is less than 0.0001, so we 38
TABLE 4: Complete analysis of variance table for forelimb measurements of individuals in both samples in the example data set. Analysis was carried out to estimate the mean squares for the interaction of sample with side and individual, which is related to how much FA+antisymmetry change between samples, and the interaction of sample, side, individual, and replicate measurement, which serves as an error term for in the F-statistic examining whether the change in FA+antisymmetry between samples is significant. Abbreviations as in Table 2.
Factor
DF
Sum of squares
Mean square
SAMPLE SIDE SAMPLE*SIDE IND_NUM SAMPLE*IND_NUM SIDE*IND_NUM SAMPLE*SIDE*IND_NUM REP SAMPLE*REP SIDE*REP SAMPLE*SIDE*REP IND_NUM*REP SAMPLE*IND_NUM*REP SIDE*IND_NUM*REP SAMPLE*SIDE*IND_NUM*REP
1 1 1 29 29 29 29 2 2 2 2 58 58 58 58
9.4738 0.1440 0.5444 75.1879 42.9712 4.0977 3.9672 0.0029 0.0009 0.0080 0.0016 0.1488 0.1441 0.1103 0.1168
9.4738 0.1440 0.5444 2.5927 1.4818 0.1413 0.1368 0.0014 0.0004 0.0040 0.0008 0.0026 0.0025 0.0019 0.0020
would conclude that there is strong evidence that the level of FA+antisymmetry of the forelimbs differs between samples 1 and 2 of our example data set.
Graphs and statistics using means of replicate measurements Once the ANOVA techniques outlined above and discussed in greater detail by Palmer and Strobeck (1986) have been used to determine whether significant levels of FA+antisymmetry exist and to estimate mean levels corrected for the effects of measurement error, it can be very useful to examine the levels of asymmetry shown by the individuals in each sample. The first step in doing this is to take the mean of the three replicate measurements of the length of each limb. These means can then be used to calculate signed (right length — left length) and unsigned (absolute value taken of signed) asymmetry for the front and hind limbs of each individual. The unsigned asymmetry values for the front and hind limbs can be added together to get an estimate of the total limb asymmetry for each individual. Means and derived measurements for the 60 individuals in both samples of our example data set appear in Table 5. Signed asymmetry values can be used to directly examine the distribution of asymmetry in each sample by producing plots of the magnitude and direction of expressed asymmetry similar to Figure 1. This figure indicates that the asymmetry expressed in forelimb measurements in sample 1 of our example data set is true FA, rather than antisymmetry, as the distribution is unimodal (within the limits of resolution possible with a relatively small data set), and is symmetrical about a mean of zero. If the ANOVA test for significant levels of FA+antisymmetry had not been significant, examining a
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
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Individual
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Forelimb length Left Right
11.733 11.367 12.000 12.067 11.533 11.567 10.333 11.533 11.000 11.967 11.433 10.600 10.233 11.167 10.100 9.833 10.867 11.567 10.867 11.500 10.600 12.200 10.767 10.767 10.167 11.067 11.533 11.300 10.967 11.200 11.067 11.000 12.200 12.433 11.300 11.600 11.233 11.500 11.967 10.667 12.000 12.200 12.133 10.733 10.867 10.300 11.567 11.133 11.633 11.300 10.700 10.700 11.367 11.033 10.167 11.500 11.300 11.900 11.700 12.033
11.467 11.367 12.067 12.100 11.600 11.200 10.333 11.467 10.833 11.533 11.300 10.233 10.067 11.300 10.467 9.700 11.167 11.367 11.367 11.600 10.700 11.633 10.800 10.667 10.333 10.833 11.733 11.300 10.833 11.333 11.067 10.700 12.233 12.467 11.200 11.500 10.967 12.133 11.633 10.700 11.700 12.000 12.333 10.833 10.833 10.633 12.000 10.667 11.733 11.400 10.800 11.733 11.933 11.467 10.567 11.100 11.667 12.000 12.500 12.267
Hindlimb length Signed difference (R-L) Unsigned difference Left Right Forelimb Hindlimb Forelimb Hindlimb Total
15.800 14.367 16.600 16.167 15.367 15.367 13.667 15.400 14.600 15.333 14.700 13.533 13.567 14.767 13.500 12.900 14.367 15.500 14.600 15.133 14.100 15.867 14.600 14.600 13.833 14.400 15.200 15.200 14.200 14.667 13.867 15.400 16.600 15.667 14.667 13.833 15.833 14.867 15.500 13.400 15.300 16.900 16.600 14.000 15.033 13.100 15.467 14.600 16.100 14.900 14.167 14.400 16.433 15.000 14.167 14.500 15.933 16.500 16.200 16.167
16.133 14.300 16.167 15.900 15.700 15.033 13.833 15.633 14.667 15.467 15.033 14.233 13.800 15.067 13.433 12.533 14.767 15.100 14.767 15.567 15.133 15.300 14.100 14.733 13.933 14.733 15.533 15.333 14.667 15.067 14.967 13.867 16.867 16.133 14.300 14.867 15.100 15.133 15.333 14.467 15.467 16.167 16.700 14.367 15.733 14.700 16.033 13.900 16.100 15.000 13.900 14.400 16.333 13.967 14.233 15.067 16.000 17.033 16.100 16.633
-0.267 0.000 0.067 0.033 0.067 -0.367 0.000 -0.067 -0.167 -0.433 -0.133 -0.367 -0.167 0.133 0.367 -0.133 0.300 -0.200 0.500 0.100 0.100 -0.567 0.033 -0.100 0.167 -0.233 0.200 0.000 -0.133 0.133 0.000 -0.300 0.033 0.033 -0.100 -0.100 -0.267 0.633 -0.333 0.033 -0.300 -0.200 0.200 0.100 -0.033 0.333 0.433 -0.467 0.100 0.100 0.100 1.033 0.567 0.433 0.400 -0.400 0.367 0.100 0.800 0.233
0.333 -0.067 -0.433 -0.267 0.333 -0.333 0.167 0.233 0.067 0.133 0.333 0.700 0.233 0.300 -0.067 -0.367 0.400 -0.400 0.167 0.433 1.033 -0.567 -0.500 0.133 0.100 0.333 0.333 0.133 0.467 0.400 1.100 -1.533 0.267 0.467 -0.367 1.033 -0.733 0.267 -0.167 1.067 0.167 -0.733 0.100 0.367 0.700 1.600 0.567 -0.700 0.000 0.100 -0.267 0.000 -0.100 -1.033 0.067 0.567 0.067 0.533 -0.100 0.467
0.267 0.000 0.067 0.033 0.067 0.367 0.000 0.067 0.167 0.433 0.133 0.367 0.167 0.133 0.367 0.133 0.300 0.200 0.500 0.100 0.100 0.567 0.033 0.100 0.167 0.233 0.200 0.000 0.133 0.133 0.000 0.300 0.033 0.033 0.100 0.100 0.267 0.633 0.333 0.033 0.300 0.200 0.200 0.100 0.033 0.333 0.433 0.467 0.100 0.100 0.100 1.033 0.567 0.433 0.400 0.400 0.367 0.100 0.800 0.233
0.333 0.067 0.433 0.267 0.333 0.333 0.167 0.233 0.067 0.133 0.333 0.700 0.233 0.300 0.067 0.367 0.400 0.400 0.167 0.433 1.033 0.567 0.500 0.133 0.100 0.333 0.333 0.133 0.467 0.400 1.100 1.533 0.267 0.467 0.367 1.033 0.733 0.267 0.167 1.067 0.167 0.733 0.100 0.367 0.700 1.600 0.567 0.700 0.000 0.100 0.267 0.000 0.100 1.033 0.067 0.567 0.067 0.533 0.100 0.467
0.600 0.067 0.500 0.300 0.400 0.700 0.167 0.300 0.233 0.567 0.467 1.067 0.400 0.433 0.433 0.500 0.700 0.600 0.667 0.533 1.133 1.133 0.533 0.233 0.267 0.567 0.533 0.133 0.600 0.533 1.100 1.833 0.300 0.500 0.467 1.133 1.000 0.900 0.500 1.100 0.467 0.933 0.300 0.467 0.733 1.933 1.000 1.167 0.100 0.200 0.367 1.033 0.667 1.467 0.467 0.967 0.433 0.633 0.900 0.700
39
plot like Figure 1 would allow determining the magnitude of measurement error. A strongly bimodal distribution of signed asymmetry values would indicate that the forelimbs of animals in sample 1 showed antisymmetry. Unsigned asymmetry values can be added to give total asymmetries for each individual, or can be examined separately for front and hind limbs. Signed asymmetry values should never be totalled, as positive and negative asymmetries within animals could cancel out, leading to totals for highly asymmetrical animals that were similar to totals for highly symmetrical individuals. Figure 2 presents a plot of the means and standard deviations of unsigned asymmetry for the front and hind limbs of animals in samples 1 and 2 of our example data set, showing why the ANOVA detected
FIGURE 1: Frequency distribution of signed asymmetry of forelimbs of animals in sample 1 of the example data set.
The mean values for individuals can be used to calculate a variety of indices of asymmetry in addition to the signed and unsigned asymmetry of each structure and the total unsigned asymmetry. Palmer and Strobeck (1986) discuss these indices and their application in detail. For ease of interpretation, however, the simple signed or unsigned asymmetry is excellent, and in most practical applications is probably nearly as powerful as any of the other indices they discuss.
INCORPORATING DI MEASUREMENTS IN FIELD MONITORING PROGRAMS We hope that the techniques presented above, plus the 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 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 at the breeding sites where censuses are usually carried out can fluctuate greatly over short periods, complicating the determination of population sizes and status (Pechmann et al. 1991).
8
6 Number in class
significant differences between the samples: Sample 2 animals are substantially less symmetrical than are sample 1 animals. If sample 2 had been taken later in time, from the same population as sample 1, this would suggest that environmental stress levels experienced by the population were increasing, and would give cause for increased monitoring and efforts to determine the sources of the increased stress.
4
2
0 -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 Minimum difference in class
FIGURE 2: Mean unsigned asymmetry of the fore- and hindlimbs of animals in both samples of the example data set, illustrating the increase in asymmetry levels between the first and second samples.
Mean unsigned asymmetry +/-1 S.E.
0.65
0.45
0.25
0.05 Forelimb Hindlimb Sample 1
Forelimb Hindlimb Sample 2
Measurement Type and Sample
40
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 hour. After data on several samples of animals have accumulated, performing an ANOVA followed by 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 only a single measurement of each limb on each animal, which would greatly speed up data collection in the field. Measurements should always be taken “blind”, however, 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, 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. If the same individuals are measured repeatedly, data collected on older, less asymmetrical animals might obscure 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 (Figure 1) or unsigned asymmetry for each sample taken. A tendency towards spreading the tails of the distribution might provide cause for concern even if mean levels have not changed greatly.
CONCLUSION The incorporation of measurements of limb symmetry into regular amphibian monitoring programs, and their analysis using developmental stability analysis techniques, holds considerable promise for improving the ability of managers and biologists to detect declines in the health of populations before changes in population size become apparent. Techniques for the collection and analysis of data on DI, while available in the literature, have not been presented in a form that makes them readily accessible to many potential users. By presenting our techniques in detail we hope to make it possible for more field biologists and managers to use this technique, which should increase their ability to detect deterioration in the health of populations before drastic declines in animal numbers occur.
ACKNOWLEDGEMENTS Our research was supported by grants from the Australian Research Council and the Cooperative Research Centre for Tropical Rainforest Ecology and Management. Sascha Frydman provided many long, valuable discussions of the concepts and techniques of DI analysis and interpretation. Gary Fellers and Britta Grillitsch provided useful comments on much earlier drafts of this manuscript.
REFERENCES Alford, R. A., Bradfield, K. S., and Richards S. J. (1997). Increasing Fluctuating Asymmetry Precedes Frog Population Declines. Abstract of talk presented in the Declining Amphibian Symposium,Third World Congress of Herpetology, Prague.
M. A. Lewis, 136-58. Philadelphia, American society for testing and materials. Graham, J. H., Freeman D. C., and Emlen J. M. (1994) Antisymmetry, Directional Asymmetry, and Dynamic Morphogenesis. In Developmental Instability: Its Origins and Evolutionary Implications. 123-39. Dordrecht, Kluwer Academic Publishers. Leary, R. F. and Allendorf, F. W. (1989) Fluctuating Asymmetry As an Indicator of Stress: Implications for Conservation Biology. Trends in Ecology and Evolution, 4: 214-17. Moller, A. P. (1996) Parasitism and Developmental Instability of Hosts: a Review. Oikos, 77: 189-96. Palmer, A. R. (1996) Waltzing With Asymmetry. BioScience, 46: 518-32. Palmer, A. R. and Strobeck, C. (1986) Fluctuating Asymmetry: Measurement, Analysis, Patterns. Annual Review of Ecology and Systematics, 17: 391-421. Palmer, A. R. and Strobeck, C. (1992) Fluctuating Asymmetry As a Measure of Developmental Stability: Implications of Non-Normal Distributions and Power of Statistical Tests. Acta Zoologica Fennica, 191: 57-72. Pankakoski, E. (1985) Epigenetic Asymmetry As an Ecological Indicator in Muskrats. Journal of Mammalogy, 66: 52-57. Parsons, P.A. (1990a). Fluctuating Asymmetry and Stress Intensity. Trends in Ecology and Evolution 5: 97-98. Parsons, P. A. (1990b) The Metabolic Cost of Multiple Environmental Stresses: Implications for Climatic Change and Conservation. Trends in Ecology and Evolution, 5: 315-17.
Clarke, G. M. (1994) Developmental Stability Analysis: an Early-Warning System for Biological Monitoring of Water Quality. Australian Biologist, 7: 94-104.
Pechmann, J.H.K., Scott D.E., Semlitsch R.D., Caldwell J.P.,Vitt L., and Gibbons J.W. (1991). Declining Amphibian Populations: The Problem Of Separating Human Impacts From Natural Fluctuations. Science 253: 925-940.
Clarke, G. M. (1995) Relationships Between Developmental Stability and Fitness: Application for Conservation Biology. Conservation Biology, 9: 18-24.
Sarre, S. (1996) Habitat fragmentation promotes fluctuating asymmetry but not morphological divergence in two geckos. Researches in Population. Ecology. 38, 57-64.
Emlen, J. M., Freeman D. C., and Graham J. H.. (1994) Nonlinear Growth Dynamics and the Origin of Fluctuating Asymmetry. In Developmental Instability: Its Origins and Evolutionary Implications. 79-98. Dordrecht, Kluwer Academic Publishers.
Siimäki, P, and Lammi, A. (1998) Fluctuating asymmetry in central and marginal populations of Lychnis viscaria in relation to genetic and environmental factors. Evolution 52:1285-1292.
Freeman, D. C., Emlen J. M., Graham J. H.,. Mara R. L,Tracy M., and Alados C. L.. (1996) Developmental Instability as a Bioindicator of Ecosystem Health. In Proceedings: Shrubland Ecosystem Dynamics in a Changing Environment. Compilers J. R. Barrow, E. D. McArthur, R. E. Sosebee, and R. J. Tausch, 170-177. Ogden, UT, U.S. Department of Agriculture, Forest Service, Intermountain Research Station. Graham, J. H. (1992) Genomic Coadaptation and Developmental Stability in Hybrid Zones. Acta Zoologica Fennica, 191: 121-31. Graham, J. H., Emlen, J. M., and Freeman, D. C. (1993a) Developmental Stability and Its Applications in Ecotoxicology. Ecotoxicology, 2: 175-84. Graham, J. H., Freeman D. C., and Emlen J. M. (1993b) Developmental Stability: A Sensitive Indicator of Populations Under Stress. In Environmental Toxicology and Risk Assessment. Editors W. G. Landis, J. S. Hughes, and
Tracy, M., Freeman D. C., Emlen J. M., Graham J. H., and. Hough R. A. (1995) Developmental Instability As a Biomonitor of Environmental Stress. In Biomonitors and Biomarkers As Indicators of Environmental Change. Editor F. A. Butterworth, L. D. Corkum, and J. Guzman-Rincon, 313-37. New York, Plenum Press. Tsubaki,Y. (1998) Fluctuating asymmetry of the Oriental Fruit Fly (Dacus dorsalis) during the process of its extinction from the Okinawa Islands. Conservation Biology 12:926929. Valentine, D. W., Soulé, M. E. and Samollow P., (1973) Asymmetry analysis in fishes: a possible statistical indicator of environmental stress. Fisheries Bulletin 71: 357-370 Zakharov,V. M. (1992) Population Phenogenetics: Analysis of Developmental Stability in Natural Populations. Acta Zoologica Fennica, 191: 7-30.
41
APPENDIX 1 SAS code for full analysis of DI measurements taken on two samples of animals and stored as in Table 1. For additional samples, some of the code would be duplicated. Readers familiar with the SAS command language should be able to do this easily. Note that the data within each sample must be balanced: each animal must have three measurements taken of each limb. If the data are not balanced, the hypothesis tests and calculations of variance components done within each sample will be incorrect, since the Type I sums of squares will be order-dependent. The final section, which tests for differences in levels of FA+antisymmetry between samples, should produce correct results even if there are not the same numbers of animals in each sample. data all; /*change to path to your data file*/ infile ‘d:\sas\asym\example\data.prn’ end=eof; input; /*skips first line with variable names*/ do while(not eof); input date ddmmyy8. sample indid rep svl side fl hl; output; end; stop; /*if you have data for more than one sample, this sorts them so that SAS can do all the calsulations in one step*/ proc sort data=all; by sample; /*the following GLM does the main work, calculating the SS and MS needed to estimate FA + antisymmetry according to Palmer and Strobeck, and also testing for whether directional asymmetry and FA+antisymmetry are significantly greater than 0*/ proc glm data=all outstat=stats; by sample; class side indid rep; model fl hl=side|indid|rep; test h=side e=side*indid; test h=side*indid e=side*indid*rep; run; /*following is optional—it calculates and prints tables showing the FA+antisymmetry estimates for forelimb length and hindlimb length and the tables of the tests of hypothesis for directional and FA+antisymmetry for each sample*/ data temp; set stats; if (oldsamp ne sample) then do; flms0 = .; hlms0 = .; fldf0 = .; hldf0 = .; fldf1 = .; hldf1 = .; fldf2 = .; hldf2 = .; flms1 = .; flms2 = .; hlms1 = .; hlms2 = .; end; 42
if ((_source_ eq ‘SIDE’) and (_type_ eq ‘SS1’)) then select; when(_name_ = ‘FL’) do; flms0 = ss/df; fldf0=df; end; when(_name_ = ‘HL’) do; hlms0 = ss/df; hldf0=df; end; end; if ((_source_ eq ‘SIDE*INDID’) and (_type_ eq ‘SS1’)) then select; when(_name_ = ‘FL’) do; flms1 = ss/df; fldf1=df; end; when(_name_ = ‘HL’) do; hlms1 = ss/df; hldf1=df; end; end; if ((_source_ eq ‘SIDE*INDID*REP’) and (_type_ eq ‘SS1’)) then select; when(_name_ = ‘FL’) do; flms2 = ss/df; fldf2=df; end; when(_name_ = ‘HL’) do; hlms2 = ss/df; hldf2 = df; flfdir = flms0/flms1; flffa = flms1/flms2; flfdirp = 1-probf(flfdir,fldf0,fldf1); flffap = 1-probf(flffa,fldf1,fldf2); hlfdir = hlms0/hlms1; hlffa = hlms1/hlms2; hlfdirp = 1-probf(hlfdir,hldf0,hldf1); hlffap = 1-probf(hlffa,hldf1,hldf2); flasym = (flms1 — flms2) /3; hlasym = (hlms1 — hlms2) /3; output; end; end; oldsamp = sample; retain flms0 hlms0 fldf0 hldf0 fldf1 hldf1 fldf2 hldf2 flms1 flms2 hlms1 hlms2 oldsamp; run; proc print data=temp; var sample flasym hlasym; title ‘Palmer and Strobeck (1986) estimates of FA+antisymmetry for each character’; run; proc print data=temp; var sample flfdir fldf0 fldf1 flfdirp flffa fldf1 fldf2 flffap; title ‘tests for significance of directional asymmetry and FA+antisym for FL’; run; proc print data=temp; var sample hlfdir hldf0 hldf1 hlfdirp hlffa hldf1 hldf2 hlffap; title ‘tests for significance of directional asymmetry and FA+antisym for HL’; run; /*the following GLM is not according to Palmer and Strobeck but is legitimate—it tests for whether the FA+antisymmetry component of variance differs between samples*/ proc glm data=all; class sample side indid rep; model fl hl=sample|side|indid|rep; test h=sample*side*indid e=sample*side*indid*rep; title ‘analyses for difference in FA + Antisym component between samples’; run;
APPENDIX 2 SPSS code for full analysis of DI measurements taken on two samples of animals and stored as in Table 1. This code assumes that the data have already been read in or loaded. For additional samples, some of the code would be duplicated. Readers familiar with the SPSS command language should be able to do this easily. Note that the data within each sample must be balanced: each animal must have three measurements taken of each limb. If the data are not balanced, the hypothesis tests and calculations of variance components done within each sample will be incorrect, since the Type I sums of squares will be order-dependent. The final section, which tests for differences in levels of FA+antisymmetry between samples, should produce correct results even if there are not the same numbers of animals in each sample. Unlike the SAS code in Appendix 1, which carries out hypothesis tests and calculates estimates of corrected FA+antisymmetry values, all hypothesis tests and estimates would have to be constructed by hand using the MS reported by SPSS. See Table 3 and the text for examples. This may simply reflect the authors’ relative ignorance of the SPSS syntax language.
/ANALYSIS hll /DESIGN side ind_num ind_num*side ind_num*rep*side . FILTER OFF. USE ALL. EXECUTE . MANOVA fll BY side(1 2) ind_num(1 30) rep(1 3) sample(1 2) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL /ANALYSIS fll /DESIGN side ind_num sample side*ind_num side*ind_num*rep sample*ind_num side*sample*ind_num side*sample*ind_num*rep . MANOVA hll BY side(1 2) ind_num(1 30) rep(1 3) sample(1 2) /NOPRINT PARAM(ESTIM) /METHOD=UNIQUE /ERROR WITHIN+RESIDUAL /ANALYSIS hll /DESIGN side ind_num sample side*ind_num side*ind_num*rep sample*ind_num 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 The decline of frog populations is a worldwide phenomenon and a major conservation issue in Australia. In south-eastern Australia the understanding of the frogs affected, their ecology, patterns of decline and the causal agents is generally poor. The wet subtropical region encompasses the coast and ranges from about Newcastle north to Rockhampton. Twenty-three species of conservation concern from this region are examined. The extent of declines and known or likely threatening processes for each species, is investigated by comparing the distribution of
Although a lack of long term systematic baseline information on presence/absence or relative abundance has hampered our assessment, the species of concern fall into three groups, based on threats. Six species of stream breeding frogs have suffered declines from unknown causes. Two of these, endemic to the wet subtropics, have not been seen since the early 1980s. Eight species of mesic forest frogs, that are not dependant upon streams for breeding, show no evidence of declines. They are of conservation concern because of their rarity. The third group, frogs of the lowlands, are increasingly at risk due to loss or degradation of habitat.
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 Over the past decade there has been an increasing awareness of massive declines in frog populations in many parts of the world (e.g. Tyler and Davies 1985b, Blaustein and Wake 1990; Richards et al. 1993; Drost and Fellars 1996; Lips 1998). Despite the disappearance of several species and declines in populations of many others in south-eastern Australia since the 1970’s (Tyler and Davies 1985b; Osborne 1989; Czechura and Ingram 1990; Watson et al. 1991, Ingram and McDonald 1993), understanding of the problem is poor (Tyler 1997). For most species affected there has been little or no synthesis of information across their entire range, nor in a framework that allows easy comparison among species in a region. Reviews of the status of frogs in south-eastern Australia have either focused on small geographical areas (e.g.Watagan Mountains — Mahony 1993), single species (e.g. Litoria spenceri — Gillespie and Hollis 1996; Pseudophryne corroboree — Osborne 1989) or groups of closely related species (e.g. L. aurea complex — Pyke and Osborne 1996). Others have been constrained by political or other biologically artificial boundaries (e.g. Ingram and McDonald 1993; Ehmann 1997a;White and Pyke 1996). In this and other papers in these proceedings we examine the available data on the distribution and declines of species throughout their range, based on broad ecological boundaries of the rainforests of mid- and north-eastern Queensland (Qld) (McDonald and Alford 1999), the wet subtropical region of eastern Australia (this review), streams in the temperate zone of south-eastern Australia (Gillespie and Hines 1999) and the subalpine region (Osborne et al. 1999) to overcome these problems. Only a few species of concern, such as Adelotus brevis and L. lesueuri, span two or more of these biological regions. We define the wet subtropical region of Australia as those parts of the summer rainfall, subtropical climate zone (Bureau of Meteorology 1989, Figure 5), that receive in excess of 800 mm median annual rainfall (Bureau of Meteorology 1989, Figure 2), but excluding the upland temperate areas of the NSW northern tablelands. This region encompasses the escarpment, foothills and coastal lowlands south from Rockhampton Qld to about Newcastle NSW. The region includes large areas of subtropical 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). Over 50 species of frogs are known from the wet subtropical region, and about half of these are endemic. We have chosen to assess 23 species of conservation concern. These include 17 species that are listed as Rare,Vulnerable or Endangered under state or federal legislation (Table 1), three undescribed species which meet criteria for listing under one of these categories and three stream breeding species potentially in decline. Some of these species are listed because of their rarity and or restricted distributions, while others are threatened by loss of habitat. However, a large proportion are stream breeding species that have suffered population declines or have disappeared, without a clear understanding of the causal agent(s). Inclusion of three additional stream breeding species in this review was based on: a) 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
TABLE 1: Rare or threatened frogs of the wet subtropical region of Australia
Species
Assa darlingtoni Crinia tinnula Lechriodus fletcheri #Mixophyes balbus Mixophyes fleayi Mixophyes iteratus Kyarranus kundagungan Kyarranus loveridgei Kyarranus sphagnicolus Rheobatrachus silus Taudactylus diurnus Taudactylus pleione +Litoria aurea #Litoria booroolongensis Litoria brevipalmata Litoria cooloolensis Litoria freycineti Litoria olongburensis Litoria pearsoniana Litoria revelata #Litoria subglandulosa
Qld1 NSW2 Comm3 Action ANZECC5 Plan4
R V R E E R R
V V V V V V V V
E E V
R R V V E R V
I V E E
E E E E V
I E E V E I I
V
I V I
V
I
E E
Source: 1 Queensland Nature Conservation (Wildlife) Regulation 1994 2 New South Wales Threatened Species Conservation Act 1995 3 Commonwealth Endangered Species Protection Act 1992 4 Tyler 1997 5 The ANZECC list of Threatened Vertebrate Fauna (1995) Status abbreviation (refer to relevant source for definitions): E = endangered V = vulnerable R = rare I = insufficiently known The status of species listed in this table are reviewed here except for those marked: + reviewed by Mahony (this volume) # reviewed by Gillespie and Hines (this volume)
and monitoring i.e. A. brevis and L. lesueuri (Hines and Mahony unpubl. data); and/or c) anecdotal evidence of past declines i.e. A. brevis and L. lesueuri (Ingram and McDonald 1993), and Mixophyes fasciolatus (Corben in McDonald 1991). Four additional rare or threatened species known from wet subtropical Australia (Table 1) are not assessed here because the majority of their distribution is elsewhere. They are discussed in other papers in these proceedings (Gillespie and Hines 1999; Mahony 1999; Osborne et al. 1999). Our aim is to provide an overview of the status of frogs of conservation concern in the wet subtropical region of Australia. We examine available knowledge on the distribution, ecology and threats to these species, as a basis for identifying additional information needed for their conservation. For each species we assess:
45
1. recorded distribution, 2. current distribution with discussion of declines, 3. threats, 4. knowledge of biology, and 5. conservation status.
METHODS Distributional data for each of the 23 species of concern were collated from existing computer databases and augmented with records from unpublished reports, published literature and naturalists. Although not complete, this process enabled a large data set to be rapidly assembled. Digital data were received from the Australian Museum (AM), Queensland Museum(QM) and South Australian Museum (SAM), the National Museum of Victoria (NMV),The Australian National Wildlife Collection (ANWC), Atlas of NSW Wildlife (NSW Atlas), Atlas of Victorian Wildlife (Vic Atlas) and South East Queensland Regional Forest Assessment (SEQRFA) databases.Vetting of data sets for obvious species mis-identification and coarse georeference errors was aided by examining maps of the records for each species. Many records lacked georeferences so where an unambiguous locality description was provided a georeference was obtained from maps or gazetteer. To help assess regional population declines, maps for each species were produced showing the distribution of records pre-1990 and 1990 onwards. Nineteen-ninety was selected for two reasons. Apart from the documented disappearances 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
subtropical Australia is poor. Despite increasing awareness of declines during the 1980’s there was little survey or monitoring of frog populations. However, since 1989, large scale biodiversity inventories have been undertaken in the wet subtropical region such as regional forest assessments in south-eastern Qld and north-eastern NSW. During this period there were also targeted frog survey and monitoring programs (e.g. Ingram and McDonald 1993, Ehmann 1997a, Mahony 1996, Hines unpubl. data, McDonald unpubl. data) and a number of site or species specific studies (e.g. Mahony and Knowles unpubl. data on Mixophyes and Kyarranus). Major sources of information on the current distribution (up to December 1997) of frogs are presented in Table 2. Many sites where rare or threatened frogs were recorded prior to 1990 were revisited during this time. There are a number of potential limitations in the use of species records such as these including, biases in survey effort, undetected errors in georeferences or species identification and poor taxonomic resolution of some species groups. When necessary these issues are discussed in the species accounts, where we also draw on additional information (e.g. relative abundance measures) to aid in assessments of declines. A brief account is provided for each species, summarising or assessing available information on historical and current distribution (1990-1997), threats, biology and conservation status. Species have been grouped according to similarities in their ecology.
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 Service (QPWS)
survey, monitoring and research
threatened frogs, especially stream-breeding species (Hines unpubl. data) and Fraser Island (McDonald unpubl. data, QPWS unpubl. data)
Queensland Department of Natural Resources (QDNR)
monitoring
historical locations of R. silus and T. diurnus (Borsboom unpubl. data)
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 Wildlife Service
survey
biodiversity surveys, east coast NSW, particularly NE, for regional conservation planning
State Forests of NSW
survey
environmental impact assessments, east coast of NSW, particularly NE
University of Newcastle
survey, monitoring and research
most species, especially L. pearsoniana sensu lato, A. darlingtoni, 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 and research
various species, mostly in far NE NSW
Various biologists
survey, research and incidental records
all species, throughout
46
SPECIES ACCOUNTS
Cooloola sedgefrog
Coastal lowland species (“wallum” or “acid” frogs)
Distribution: Restricted to Qld, from Lake Woongeel, Fraser Island (24° 53´S 153° 14´E — QPWS unpubl. data), through the sandmasses of Cooloola, with a disjunct population on North Stradbroke Island (27° 32´S 153° 29´E — AM and QM specimens). The species has not been recorded on Moreton Island despite targeted surveys (McDonald and Stewart QPWS unpubl. data). Figure 2.
Wallum or acid frogs (Ingram and Corben 1975a, Kikkawa et al. 1979) inhabit low nutrient soils (mostly deep sands) of near coastal lowlands and sand islands.Vegetation types typical of these environments include heathland, Melaleuca swamp, sedgeland and Banksia woodland. Recent survey and monitoring results (Table 2) indicate that populations are relatively stable in protected habitat. However these species are at risk from continuing loss of habitat through clearing for agriculture, pine plantations, housing and infrastructure such as canal development, drainage projects and transport corridors. They occur in an area with the highest growth rate of human population in Australia. In the period 1974-1989, 33% of the 1974 bushland cover of coastal south-eastern Qld mainland was cleared, including 50% of Melaleuca forest and 34% of heathland (Catterall and Kingston 1993). Other threats include habitat degradation through changes to hydrological regimes, increased nutrients or sediments (e.g. storm water, sewage), weed invasion, inappropriate fire management, competition from invading frog species and predation from introduced fish (Gillespie and Hero 1999). Regular monitoring of these species has occurred on Fraser Island, where biyearly surveys of sites have been undertaken since 1992 (QNPWS unpubl. data). Wallum froglet
Crinia tinnula
Distribution: Lake Woongeel, Fraser Island SEQ (24° 53´S 153° 14´E — QPWS unpubl. data), south to Kurnell MENSW (34° 02´S 151° 13´E). Figure 1. The species is widespread on Fraser Island (QPWS and McDonald unpubl. data, QM specimens) contrary to Ehmann’s (1997c) statement that it was apparently absent from the island. There are records of C. tinnula on the mainland north of Woodgate, to Littabella National Park near Bundaberg SEQ (24° 41´S 152° 05´E — QPWS unpubl. data — indicated by a question mark in Figure 1). The identity of these populations requires confirmation. Current distribution and threats: As above, except for localised extermination of populations due to loss or fragmentation of habitat. The recent range extensions of this species (i.e. new records north and south of its previously known range) are almost certainly due to increased survey effort using improved field identification characters and aids (field guides and audio recordings), rather than recent migration or deliberate or accidental introductions, as suggested by Ehmann (1997c). Biology: Not well documented. Some information on breeding biology and habitat usage is presented by Straughan and Main (1966), Ingram and Corben (1975a) and Ehmann (1997c). Meyer (1997) investigated some physiological aspects of tadpole development in the acidic ion-poor waters typical of this species’ habitat. There is no information on population size, structure, dynamics or genetics, non-breeding habitat requirements or factors limiting distribution. Conservation status: The data presented here indicate that the current legislative status of C. tinnula in Qld and NSW, as Vulnerable, is appropriate. Accordingly its status nationally (currently not listed) should be reviewed. Crinia tinnula is not the subject of a recovery process.
Litoria cooloolensis
Current distribution: As above, except for localised extermination of populations due to loss or fragmentation of habitat. Biology and threats: Information on breeding biology and habitat usage of L. cooloolensis is provided by Liem (1974) and James (1996). Some physiological aspects of L. cooloolensis larval development in acidic, ion-deficient breeding waters have been investigated by Meyer (1997). Threats from high human visitation to the fresh water lake habitats important for L. cooloolensis include trampling of reed beds and pollution of water (e.g. increased nutrients, sunscreen residues) (James 1996). James (1996) examined the genetic structure of this species and found that populations on North Stradbroke Island diverged significantly. On Stradbroke Island it is restricted to reed beds in and around freshwater lakes and swamps. Water extraction for domestic purposes and sandmining has, and will continue to, significantly impact on this population through loss of habitat or degradation resulting from changes to hydrological processes and water quality. There is no information on population size, structure, or dynamics, non-breeding habitat requirements or factors limiting distribution. Conservation status: The species is currently considered Rare in Qld. Some populations (Stradbroke Island) of L. cooloolensis are under threat, so its legal status should be reviewed both in Qld and nationally (currently not listed). Litoria cooloolensis is not the subject of a recovery process. Wallum rocketfrog
Litoria freycineti
Distribution: Northern end of Fraser Island SEQ (25° 6´S 153° 11´E — QPWS unpubl. data), south to Jervis Bay MENSW (35° 08´S 150° 43´E — ANWC, NMV, QM specimens). Figure 3. Several inland records (from far NENSW and MENSW have not been included in Figure 3 as they require confirmation). Current distribution and threats: As above except for localised extermination of populations due to loss or fragmentation of habitat. Biology: Poorly known. There is general information on habitat use (Ingram and Corben 1975a) and breeding biology (Straughan 1966). Meyer (1997) investigated some physiological aspects of tadpole development in acidic iondeficient waters typical of this species’ habitat. There is no information on population size, structure, dynamics or genetics, non-breeding habitat requirements or factors limiting distribution.
47
Conservation status: The data presented here indicate that the current legislative status of L. freycineti in Qld as Vulnerable, is appropriate. Given the rate of habitat loss throughout its range, its status in NSW and nationally should be reviewed (currently not listed). Litoria freycineti is not the subject of a recovery process. Wallum sedgefrog
Litoria olongburensis
Distribution: Lake Woongeel, Fraser Island (24° 53´S 153° 14´E — QPWS unpubl. data), south to near Woolgoolga (30° 08´S 153° 11´E — Liem and Ingram 1977; AM, QM specimens) including Bribie, Moreton and North Stradbroke Islands, SEQ (Figure 4). The records from Woolgoolga (above),Yuraygir National Park and Yamba (NSW NPWS 1995, NSW Atlas, Hines unpubl. data) show that the distribution of this species extends much further south than stated by Ehmann (1997e). Current distribution and threats: As above except for localised extermination of populations due to loss or fragmentation of habitat. Biology: Limited information on breeding biology and habitat usage of L. olongburensis is provided by Liem and Ingram (1977), James (1996) and Ehmann (1997e). Meyer (1997) investigated some of the physiological aspects of larval development in acidic ion-deficient waters typical of this species’ habitat. Considerable genetic structuring occurs within L. olongburensis reflecting its disjunct distribution (James 1996). There is no information on population size, structure or dynamics, non-breeding habitat requirements or factors limiting distribution. Conservation status: The data presented here indicate that the current legislative status of L. olongburensis in Qld and NSW, as Vulnerable, is appropriate. Accordingly its status nationally should be reviewed (currently not listed). Litoria olongburensis is not the subject of a recovery process.
Sclerophyll forest species Green-thighed frog
Litoria brevipalmata
Distribution: Patchily recorded from low to mid altitudes from Cordalba State Forest SEQ (25° 10´S, 152° 10´E — Hines unpubl. data) south to Darkes Forest, MENSW (34° 14´S, 150° 55´E — AM specimens) (Figure 5). It is a difficult species to survey as it breeds for only a short period, usually after heavy spring and summer rains. Current distribution and threats: The species has declined in the Ourimbah Creek area since the time it was first detected in 1972 (Mahony 1996). This site is visited by many herpetologists from Sydney and the Central Coast and the large populations reported in the early 1970’s have not been observed in recent years. It is most likely that this decline is due to clearing for agricultural development (Mahony 1996). However, several other species (e.g. Mixophyes balbus, M. iteratus, and L. aurea) previously known from this and nearby areas have declined dramatically without any apparent reason (Mahony 1993). Therefore it is not possible to discount other, unknown threat(s) as the cause of decline of L. brevipalmata in the Ourimbah Creek area. Elsewhere there are no reports of declines or disappearances of this species.We are not aware of any recent surveys under 48
suitable conditions of the sites where McDonald (1974) first recorded this species in Queensland (the two most southwesterly records; Figure 5). Over the past five years additional populations have been located as a result of increased survey effort and better understanding of the factors affecting detectability of this species. The impacts of habitat disturbance on L. brevipalmata are not known. The lack of records from cleared environments and the decline at Ourimbah Creek suggests that L. brevipalmata is vulnerable to major habitat perturbations. Given the rate of habitat loss in the lowlands (e.g. Catteral and Kingston 1993), some localised populations must have been exterminated before their existence was recorded. Other disturbances to its habitat include inappropriate fire regimes, timber harvesting, grazing by domestic stock, weed invasion and changes to hydrological regimes and water quality. It is not possible to assess the effects of these disturbances as knowledge on the biology of L. brevipalmata is poor. Biology: Limited information. Anstis (1994) described larval development and Czechura (1978), McDonald (1974), McEvoy et al. (1979), Nattrass and Ingram (1993) and Aridis (1997) provide qualitative information on habitat at some sites. Ledlin (1997) provides information on larval morphology and ecology, and the habitat characteristics of the species in mid-eastern NSW. Genetic studies (allozymes and DNA) of eleven populations in north-eastern NSW and one in southeastern Qld (north from Watagan State Forest to Border Ranges National Park) revealed no cryptic species (Donnellan and Foster 1997). Additional samples are required from Qld to fully assess genetic diversity. There is no information on population size, structure, or dynamics, non-breeding habitat requirements or factors limiting distribution. Conservation status: The legislative status of L. brevipalmata in Qld is Rare and in NSW it is Vulnerable. Litoria brevipalmata is not the subject of a recovery process.
Mesic forest species Within subtropical Australia, mesic forest frogs are those that are predominantly dependent upon rainforest and wet sclerophyll forest communities. This vegetation is typically in the foothills and ranges of the Great Divide. In contrast to the habitat of the preceding frogs, the mesic forests have generally been cleared at a slower rate in recent times. In the past however, some large areas of rainforests on flat topography have been lost. Significant extensive disturbances to remaining habitat include timber harvesting, hydrological changes, cattle grazing, altered fire regimes, feral animals and weed invasion. The effects of these disturbances are largely unknown but from the knowledge of the frogs’ ecology they are likely to be highly significant in many situations. Within the frog fauna of the mesic forests there are three groups that differ in breeding biology; non-stream breeding, facultative stream breeding and obligate stream breeding. Species falling into each of these groups have undoubtedly declined due to localised loss or degradation of habitat. A number of stream breeding species have suffered additional declines from unknown causes, that are apparently not related to habitat disturbances. Unlike temperate riverine frogs (Gillespie and Hines 1999), there is no evidence of introduced fish causing declines of frogs in subtropical streams. Most streams either lack fish or support a small
FIGURE 1: Distribution of wallum froglet Crinia tinnula. Closed circles with ‘?’ are records 1990 onwards with questionable taxonomic identity.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 3: Distribution of wallum rocketfrog Litoria freycineti
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 2: Distribution of Cooloola sedgefrog Litoria cooloolensis
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 4: Distribution of wallum sedgefrog Litoria olongburensis
● 1990 onwards record ❍ pre-1990 record ■ capital cities
49
number of native species. However the impact of introduced fish is potentially very serious and we recommend strict controls on the introduction or movement of any fish species in south-eastern Australia (see Gillespie and Hero 1999 for detailed discussion). A fourth group, includes a single species, Taudactylus pleione, because its breeding biology remains unknown.
Mesic forest species — non-stream breeding Pouched frog
Assa darlingtoni
Distribution: Disjunct distribution from the Conondale and Blackall Ranges SEQ (26° 34´S 152° 52´E — NMV specimen from near Kondalilla) south to the Dorrigo Plateau NENSW (30° 22´S 152° 44´E — AM specimen, NSW Atlas, NSW NPWS 1994a), with populations on the D’Aguilar, Main, Gibraltar and Border Ranges (Figure 6). Current distribution: As above. The lack of recent records from Mount Nothophagus on the Qld/NSW border probably reflects a lack of survey effort. Under suitable conditions large numbers can be heard calling in the Conondale and Eastern Border Ranges (Hines unpubl. data, Ehmann 1997b). The small, recently discovered population on D’Aguilar Range, west of Brisbane, is regularly detected during monitoring of stream frogs undertaken by the University of Qld (Cunningham pers. comm.). Threats: In the past considerable clearing of this species’ habitat occurred (e.g. Dorrigo and Maleny plateaux) and much of its remaining habitat has been logged. However, the majority of this habitat is now reserved or excluded from timber harvesting. There is some evidence that timber harvesting may have negative impacts on this species (Lemckert 1999). Biology: The biology of A. darlingtoni is reasonably well documented, in particular its unusual breeding (Straughan and Main 1966;Tyler 1972; Ingram et al. 1975; Ehmann and Swan 1985; Mahony 1992). In north-eastern NSW its distribution was statistically modelled using a pool of 24 environmental predictors (NSW NPWS 1994a). The model highlighted the species’ preference for landscapes dominated by mesic forests, in areas with mild temperatures and moderate to high rainfall. This, in conjunction with information on local distribution and habitat association provided by Czechura (1991), Ingram et al. (1975), and Ehmann (1997b), dispels the misconception that A. darlingtoni is closely associated with Antarctic Beech (Nothophagus moorei) forest at high altitude (Straughan and Main 1966; Cogger 1996). There is no information on population size, structure or dynamics. A limited genetic study of the three major disjunct populations in NSW (Eastern Border Ranges, Gibraltar Range and Dorrigo Plateau) found no evidence of species level differences among populations (Donnellan and Mahony, unpubl. data). Conservation status: Listed as Rare in Qld and Vulnerable in NSW but not considered threatened nationally. We have not identified any additional information to suggest that its status in either state or nationally requires review. Assa darlingtoni is not the subject of a recovery process.
50
Mountain frogs
Kyarranus species
Three Kyarranus species have been formally described; K. kundagungan, K. loveridgei and K. sphagnicolus, all from south-eastern Qld and north-eastern NSW. Recent studies of allozyme variation (Knowles 1994; Mahony and Knowles 1994) revealed that there are six allopatric species. All are rare with restricted distributions; at least two species are known from less than 10 sites each. Members of the genus Kyarranus are sometimes placed in the genus Philoria with P. frosti. For the purposes of this study we have combined all records of Kyarranus because: a. taxonomic revision is pending; b. there has been considerable confusion over the identity of individuals or populations in the past and c. each species occupies similar ecological niches (i.e. very similar habitat requirements and breeding biology). Distribution: Six allopatric species occurring in the headwaters and seepage zones in mid to high elevation mesic forest from Mount Mistake SEQ (27° 53´S 152° 21´E — Ingram and Corben 1975b) south to the Comboyne Plateau in NENSW (31° 40´S 152° 18´E — AM specimens, NSW Atlas), Figure 7. For a detailed description of the distribution of each species refer to the “current distribution” section of species profiles 7 — 12 in Ehmann (1997a). Wotherspoon’s (1981) record from Barrington Tops National Park (151° 32’E 32° 10’S) is rejected because of probable mis-identification (Knowles and Mahony 1997d) and lack of corroborating records from the area despite intensive survey work (e.g. Knowles and Mahony 1997d; NSW NPWS 1994a). Current distribution: As above. There are no documented declines for the genus. Under suitable weather and seasonal conditions members of the genus have been readily located at sites where previously recorded (Hines unpubl. data; Knowles and Mahony unpubl. data). Targeted surveys for this genus have located many new populations (e.g. Smith et al. 1989a,b; NSW NPWS 1994a; Knowles and Mahony 1997a,b,c,d,e,f). Threats: In some areas considerable clearing of these species’ habitat occurred (e.g. Acacia, Dorrigo and Comboyne plateaux) and much of the remaining habitat has been logged. However, the majority of this habitat is now reserved or excluded from timber harvesting. These species are likely to be susceptible to upstream disturbances that affect hydrological regimes or water quality but there has been no research to assess this. Direct damage to breeding habitat by domestic stock has been observed at a number of sites (e.g. Knowles and Mahony 1997e; Hines pers. obs.). Biology: The factors affecting distribution have been investigated at local (Knowles 1994, Webb 1989) and regional scales (NSW NPWS 1994a). The eggs and larval development are well documented (Anstis 1981; deBavay 1993; Ingram and Corben 1975b). Diet in one population has been studied (Webb 1989). Little is known about non-breeding habitat requirements, population size, structure or dynamics. Conservation status: Kyarranus are listed as Rare in Qld and Vulnerable in NSW. The status of each species will need to be reviewed after the taxonomy of the group is revised and the distribution and threats to each are known. Recovery plans have not been prepared for any of the species of Kyarranus.
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
Distribution: South from Mistake Mountains SEQ (27° 58´S 152° 23´E — SEQRFA, Hines unpubl. data) to near Gosford MENSW (33° 25´S 151° 20´E — Covacevich and McDonald 1993). Figure 8. Moore (1961) referred to a specimen from Ravenshoe on the Atherton Tableland in north-eastern Qld. However McDonald and Miller (1982) argued that the locality recorded for this specimen (AM 19947) was probably erroneous. Current distribution: Corben (in McDonald 1991) felt that L. fletcheri had declined in abundance but lacked any quantitative data to support the suggestion. Recent regional surveys (e.g. Mahony 1996, NSW NPWS 1994a) and sitespecific monitoring (e.g. Hines unpubl. data) have recorded this species frequently throughout its known range. Threats: Lechriodus fletcheri is commonly encountered in wet sclerophyll forest as well as rainforest. In the past considerable clearing of this species’ habitat occurred (e.g. Comboyne, Dorrigo and Beechmont plateaux) and much of its remaining habitat has been logged. Although rainforest is now largely excluded from timber harvesting, logging of wet sclerophyll forest continues. The impact of this on L. fletcheri is unknown. Biology: The eggs and larval development of L. fletcheri are well documented (Moore 1961, Watson and Martin 1973, Martin 1967). The factors affecting this species’ distribution in north-eastern NSW (the vast majority of its range) were examined by modelling presence-absence data with a pool of 24 environmental predictors (NSW NPWS 1994a). The model showed that L. fletcheri had a marked preference for mesic forests in areas of high rainfall and deeper soils, on mid slopes or plateaux. These environments provide a greater abundance of ephemeral water bodies, an important breeding requirement for this species (Moore 1961, Hines unpubl. data). The role of larvae in the food web of water filled tree holes was investigated by Kitching and Callaghan (1982). There is no information on population size, structure, dynamics, or genetic variation. Conservation status: Lechriodus fletcheri is considered Rare in Qld because of its narrowly restricted distribution in that state. There is no evidence of declines or of major threats across its range so we believe there is no need to review its legal status. There is also no need for it to be included in a recovery process. Whirring Treefrog
Litoria revelata
Distribution: Three allopatric populations; Atherton Tableland, NEQ , Clarke Range MEQ, (Covacevich and McDonald 1993), and in the south from Mount Tamborine SEQ (27° 55’S 153° 10’E — QM specimens) to Ballina NENSW (28° 52’S 153° 34’E — QM specimens; Ingram et al. 1982). Figure 9. Records of L. revelata further south (e.g. Covacevich and McDonald 1993) are possibly records of other members of the L. ewingii complex (Mahony and Knowles 1994) and have not been included in this review. Current distribution (southern population only): No reports of declines. Under suitable conditions it has been found recently in abundance at several ponds on O’Reilly’s
52
Plateau (SEQ), and on the Tooloom Range and Acacia Plateau (NENSW) (Hines unpubl. data). It has been detected at most other sites in the remainder of its range in recent years (Mahony unpubl. data). Threats: Relative to the other mesic frogs considered in this review, L. revelata is more frequently detected in heavily disturbed sites (e.g. farm dams). Some of these sites continue to be disturbed through clearing, timber harvesting and associated activities, and cattle grazing. There have been no studies on the impact (negative or positive) of these disturbances on L. revelata and the long-term viability of these populations is not known. Biology: Poorly known. The description of the species (Ingram et al. 1982) provides basic information on distribution and habitat. Larvae from the Clarke Range population have been described by Hero et al. (1996). There is no information on population size, structure, genetics or dynamics. Resolution of taxonomic confusion in the southern population with other members of the L. ewingii complex is needed to accurately determine the distribution and ecology of L. revelata in that area. Conservation status: Currently considered Rare in Qld, due to its restricted and disjunct distribution. A review of its status in both states and nationally may be necessary once a better understanding of the taxonomy of the L. ewingii complex is gained.
Mesic forest species — facultative stream breeders Tusked frog
Adelotus brevis
Distribution: Disjunct population in the Clarke Range MEQ (21° 02´S 148° 36´E — 21° 19´S 148° 14´E — QM and AM specimens) then Shoalwater Bay MEQ (22° 40´S 150° 41´E — ANWC specimens) south along the coast and ranges to near Moss Vale MENSW (34° 37´S 150° 30´E — SAM specimen), inland to Blackdown Tableland (23° 43´S 149° 04´E — QM specimens) and Carnarvon Gorge (24° 55´S 148° 05´E — QM specimen), Figure 10. Current distribution: In south-eastern Qld there has been little recent survey work west of the Great Divide so the status of inland populations is not known.Within the subtropical zone A. brevis has not been recorded along the Great Dividing Range from the NSW border to the Bunya Mountains despite recent surveys and intensive monitoring in these areas (Hines unpubl. data). This area is climatically similar to the adjoining NSW northern tablelands where A. brevis appears to have suffered a serious decline (Gillespie and Hines 1999). Elsewhere in far south-eastern Qld the species is readily detectable in suitable seasonal and weather conditions (Ingram and McDonald 1993, Hines unpubl. data). In far north-eastern NSW, Mahony (1996) frequently located A. brevis in stream and pond situations but only below 400m. In that area there are few historical records above 400m making it difficult to assess whether declines have occurred. Adelotus brevis has apparently declined from elevated sites in the Clarke Range mid-eastern Qld (Ingram and McDonald 1993; McDonald and Alford 1999) and from the NSW northern tablelands (Gillespie and Hines 1999). Localised declines have also been reported from the Blackall and Conondale Ranges southeastern Qld (1978 to 1984 — Ingram and McDonald 1993) but no declines were apparent in the Watagan Mountains in mideastern NSW (Mahony 1993).
FIGURE 9: Distribution of whirring treefrog Litoria revelata in subtropical Australia
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 11: Distribution of great barred-frog Mixophyes fasciolatus
FIGURE 10: Distribution of tusked frog Adelotus brevis
● 1990 onwards record ❍ pre-1990 record ■ capital cities
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 urban Brisbane, the Lamington Plateau and D’Aguilar Range (Hines unpubl. data) and near Lismore (Tarvey NSW NPWS pers. comm.). Preliminary post-mortem examinations of specimens from Brisbane suggest that a chytrid fungus (Berger et al. 1998) is the cause of death (Berger unpubl. data). A major threat to A. brevis is loss and degradation of habitat, especially in the lowlands (see discussion on coastal lowland frogs), through agriculture and urban development. Although A. brevis persists in heavily disturbed sites the viability of these populations is unknown. Introduced fish, including Gambusia holbrooki, are also widespread in some lowland areas. Their impact on A. brevis is not known. Within forest catchments timber harvesting, cattle grazing and altered fire regimes may also affect this species. For example, Lemckert (1999) found that in the Dorrigo area (NENSW) A. brevis was apparently dependent upon patches of undisturbed forest. Biology: Poorly known. Katsikaros and Shine (1997) examined sexual dimorphism with respect to diet, habitat use and mating systems. Moore (1961), Watson and Martin (1973) and Daly (1995) described the nests, eggs and larvae. There is no information on population size, structure, genetics or dynamics. Conservation status: This is the first time that possible declines of A. brevis over large areas have been identified. It had previously been considered a secure species that was widespread and common (e.g. Tyler 1997; Covacevich and McDonald 1993; Ingram and McDonald 1993). It has therefore not been considered for listing as a threatened species. The information presented here and in the review by Gillespie and Hines (1999) should be used as a basis for reassessment of its status, particularly in NSW. Targeted surveys and inclusion of this species in monitoring and research programs for other declining species are recommended.
often uses unconnected pools near streams and also billabongs and dams well away from streams. Although M. fasciolatus is commonly found in disturbed environments it does not tolerate habitat clearing. In the Dorrigo area (NENSW), Lemckert (1999) found that M. fasciolatus was less abundant in more recently logged areas. The effect of other disturbances (e.g. fire, grazing) has not been examined. Berger et al. (1998) identified a chytrid fungal infection as the cause of death in a captive population and more recently in an animal from the wild (Mount Glorious SEQ — Berger unpubl. data). Biology: Poorly known. There are qualitative descriptions of broad habitat use (e.g. Cogger 1996; Barker et al. 1995; Straughan 1966) and its eggs and larvae are inadequately described. Current studies by University of Newcastle and QPWS are providing more detailed information on habitat requirements breeding biology and genetic population structuring. Captive husbandry techniques have been developed for this species at a number of institutions (Amphibian Research Centre, Melbourne Zoo, Lone Pine Koala Sanctuary). There is no information on population size, structure, dynamics or genetics, non-breeding habitat requirements or factors limiting distribution. Conservation status: Mixophyes fasciolatus in not listed as a threatened species in either state or nationally. It has not shown any significant reduction in range. Due to ecological similarities with stream dependant frogs that have declined or disappeared we recommend that this species be included in monitoring programs, particularly to assess changes in relative abundance. Mixophyes fasciolatus is and should continue to be used as a model for research into threatening processes and conservation management for the two species of Mixophyes currently considered endangered.
Mesic forest species — obligate stream breeders Great barred-frog
Mixophyes fasciolatus
Distribution: Clarke Range MEQ (20° 50’S 148° 28’E — 21° 17’S 148° 30’E — AM, QM and SAM specimens); Kroombit Tops SEQ (24° 22´S 151° 01´E — e.g. QM specimens) to Gosford MENSW (33° 26’S 151° 20’E — AM, MV and SAM specimens). Records from the Blue Mountains and further south (e.g. Cogger 1996) require confirmation. Figure 11. Current distribution: Frequently recorded, often in large breeding congresses throughout its known range (e.g. Mahony 1996; Ingram and McDonald 1993; NSW NPWS 1994a). Corben (in McDonald 1991) reported that M. fasciolatus declined markedly in the Conondale Range in the mid 1970’s, but no other researchers at the time reported declines in this area. Regular surveys and monitoring in this and other areas of south-eastern Qld (Hines unpubl. data) support Ingram and McDonald’s (1993) statement that there is no decline in this species. In the Conondale Range it is one of the few species that persisted at sites such as Booloumba Creek where four species of stream dependant frogs disappeared. Threats: The apparent resilience of M. fasciolatus to disturbance and factors causing the declines in other stream breeding species may be linked to its broader habitat utilisation and differences in breeding biology. Mixophyes fasciolatus is not dependant upon streams for breeding, but 54
Fleay’s barred-frog
Mixophyes fleayi
Distribution: Narrowly and disjunctly distributed in wet forests from the Conondale Range SEQ (26° 43´S 152° 35´E — Hines unpubl. data), south to Yabbra Scrub in NENSW (28° 37´S 152° 29´E — Mahony unpubl. data), Figure 12. Current distribution: Corben (in McDonald 1991) reported that M. fleayi declined in the Conondale Range in the late 1970’s. Ingram and McDonald (1993) reported that it had not been seen in the Conondale Range since the summer of 1990-91. Since Ingram and McDonald’s review, targeted surveys for M. fleayi have been undertaken. A population was found in the upper reaches of three neighbouring streams in the Conondale Range (Hines unpubl. data) despite surveys of historical sites lower down in the streams which failed to locate the species. In Qld other populations are currently known from Lamington plateau and the northern section of Main Range (Hines unpubl. data), Mount Barney area (Hero unpubl. data) and Currumbin (Dadds unpubl. data) and Tallebudgera Creeks (QM specimen, Hines unpubl. data) below Springbrook Plateau. There have been no records of M. fleayi from the extensively developed Mt Tamborine area since 1976, despite targeted surveys (Hines unpubl. data, Hero unpubl. data). In NSW it is known from Lever’s Plateau (Border Ranges — Cunningham pers. comm.), Yabbra and Tooloom Scrubs (Mahony unpubl. data, Dadds unpubl. data), Mt Warning (Southern Cross University unpubl.
data),Terania Creek in Nightcap Range and Sheepstation Creek in the Border Ranges (Mahony 1996, Mahony et al. 1997a). In the past two summer seasons there have been no sightings of M. fleayi at Terania Creek despite intensive searches, although prior to this only very low numbers had been observed (Mahony unpubl. data).
1994b, AM specimen). North of this there is currently a large population in the Dorrigo — Coffs Harbour area, North Washpool State Forest and Bungawalbin State Forest (Mahony unpubl. data, Southern Cross University unpubl. data). In far north-eastern NSW, M. iteratus is known from only three sites, despite intensive surveys.
Mixophyes fleayi has disappeared from some sites. Whether it has declined at others is difficult to assess due to a lack of historical records of relative abundance. The very low numbers recorded from many well surveyed sites suggest that declines in abundance may have occurred.
In south-eastern Qld M. iteratus is currently known from scattered locations in the Mary River catchment downstream to about Kenilworth, Upper Stanley River, Caboolture River and Coomera River (Hines unpubl. data, Marshall unpubl. data). During the early 1980’s M. iteratus declined and disappeared from at least two streams in the Conondale Range where it was well known (Corben in McDonald 1991, McDonald unpubl. data, Ingram unpubl. data). Recent surveys and monitoring (see R. silus) in this area failed to locate it despite a population being discovered in a nearby catchment (Parris unpubl. data). The Bunya Mountains (Straughan 1966) and Cunningham’s Gap (Straughan 1966, AM specimen, Corben pers. comm.) previously supported M. iteratus but these and nearby sites have recently been the subject of targeted survey or intensive monitoring (Hines unpubl. data), without locating the species. Assessing the extent of the decline is difficult because of the lack of baseline data on its distribution and abundance.
Threats: The reasons for declines or disappearance of M. fleayi are not known. Large areas of this species’ habitat have and continue to be degraded by feral animals (e.g. feral pigs in the Conondale Range), domestic stock (Main Range) and invasion of weeds. Upstream clearing, timber harvesting and urban development (e.g. Mt Tamborine) are all likely to have affected flow regimes and water quality. A chytrid fungal infection has been identified as the cause of illness and death of M. fleayi on Main Range and Lamington plateau (Berger et al. 1998). Biology: Poorly known. This species was first described in 1987 by Corben and Ingram, who provided very basic information of the species’ ecology. Current studies by Knowles (University of Newcastle) and QPWS are providing detailed information on habitat requirements, breeding biology, population structure and dynamics. Samples are being collected to enable examination of genetic structure. Conservation status: Mixophyes fleayi is currently listed as Endangered in both Qld and NSW and in the Action Plan for Australian Frogs (Tyler 1997), but it is not listed nationally. Based on its rarity, restricted distribution, possible declines and current threats its status nationally should be reviewed. A draft recovery plan has now been prepared and implemented (Hines 1997). A captive husbandry project has been initiated at Lone Pine Koala Sanctuary. Giant barred-frog
Mixophyes iteratus
Distribution: Belli Creek near Eumundi SEQ (26° 31´S 152° 49´E — Barden pers. comm.) south to Warrimoo MENSW (33° 43´S 150° 36´E — AM specimen), Figure 13. Cogger (1996) states that M. iteratus was distributed south “to about Narooma” (36° 13´S 150° 08´E), but there are no specimens nor other records this far south to substantiate the claim. Current distribution: Mixophyes iteratus has suffered major declines in the southern portion of its range. There are no recent records from the Blue Mountains (Mahony et al. 1997b), although there were only a few historical records in that area. In the Watagan Mountains M. iteratus is currently known from a small recently rediscovered population (Mahony unpubl. data). Although not common there in the past it was frequently recorded (Mahony 1993). Between the Hunter River and Macleay catchment there is currently only one known population, at Mount Seaview (Hobcroft pers. comm.), but survey effort in this area has been relatively low. There were only two confirmed historical records in this area (AM specimens from the Upper Allyn River and Middle Brother State Forest). A population was recently located in the southern Nambucca River catchment (NSW NPWS
Threats: Many sites where M. iteratus occurs are the lower reaches of streams which have had major disturbances such as clearing, timber harvesting and urban development in their headwaters. In the Dorrigo area (NENSW), Lemckert (1999) found that M. iteratus was less abundant in recently logged areas and at sites where there was little undisturbed forest. The impacts of feral animals, domestic stock, weed invasion and disturbance to riparian vegetation, all potential threats to current populations, are unknown. Biology: Poorly known with only basic information on its ecology (Straughan 1968, Mahony et al. 1997b). Current studies by Knowles and Mahony (University of Newcastle) are providing more detailed information on habitat requirements, breeding biology, population structure and dynamics. A study of genetic variation within the species has been completed (Donnellan and Mahony, unpubl. data) and no species level differences were found among the six NSW populations compared. Conservation status: Mixophyes iteratus is currently listed as Endangered in both Qld and NSW and in the Action Plan for Australian Frogs (Tyler 1997), but is not listed nationally. Based on declines in distribution and the current threats to M. iteratus, its status nationally should be reviewed. A draft recovery plan has now been prepared and implemented (Hines 1997). Southern gastric brooding frog
Rheobatrachus silus
Distribution: Restricted to elevations of 400 — 800m in the Blackall and Conondale Ranges SEQ between Coonoon Gibber Creek (26° 33’S 152° 42’E — QM specimens) and Kilcoy Creek (26° 47’S 152° 38’E — Covacevich and McDonald 1993, QM specimens), Figure 14.
55
Current distribution: Not sighted in the wild since 1981 despite continued efforts to relocate the species. Since Ingram and McDonald’s review (1993) the following surveys and monitoring for this species have been undertaken:
Current distribution: Not sighted in the wild since 1979 despite continued efforts to relocate the species. Since Ingram and McDonald’s review (1993) the following surveys and monitoring for the species have been undertaken:
a. Regular monitoring at Ingram’s (1983) study site — Beauty Spot 100 on Booloumba Creek, Bundaroo, Peters and East Kilcoy Creeks in the Conondale Range (Borsboom unpubl. data, QPWS unpubl. data) and at Picnic Creek (the type locality near Kondalilla) on the Blackall Range (QPWS unpubl. data);
a. T. diurnus was present at most sites R. silus occurred, so surveys and monitoring for that species (see above) were likely to detect T. diurnus.
b. 1995 intensive “frog search” of Conondale Range (Cunningham pers. comm.); c. 1997 “frog search” of the headwaters of Kilcoy, North Booloumba and Bundoomba Creeks, Conondale Range (Hines unpubl. data); d. since 1996 systematic surveys of many streams in the Conondale and Blackall Ranges. Some sections of streams were visited on many occasions over a range of weather conditions. Poorly surveyed streams in the Upper Stanley River were targeted (Hines unpubl. data, Dadds unpubl. data); and e. opportunistic surveys by various frog biologists. The species declined rapidly in late 1979, with only a single specimen located after that, in 1981 (Richards et al. 1993). Threats: The reason(s) for the disappearance of this species 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 threatened by feral pigs, invasion of weeds (especially mistflower Ageratina riparia), and altered flow and water quality due to upstream disturbances.
b. Regular (near fortnightly) diurnal monitoring at the type locality (Greene’s Falls) and nearby streams at Mount Glorious by Brisbane Frog Society for a year (1995-1996 — Holdway pers. comm.). c. A study on L. pearsoniana at nearby Love Creek (Cunningham pers. comm.) since September 1995 failed to detect T. diurnus despite some diurnal censuses and regular tadpole surveys. Threats: Like R. silus, the reason(s) for the disappearance for T. diurnus remains unknown (Martin et al. 1997). This species’ habitat is currently threatened by feral pigs, invasion of weeds (especially mist flower) and altered flow and water quality due to upstream disturbances. Biology: Czechura and Ingram (1990) reviewed knowledge of the biology of T. diurnus. There is a reasonable understanding of the factors affecting distribution of the species, its patterns of behaviour, breeding biology, diet, and thermal relations and water balance. There is no information on population size, structure, genetics or dynamics.
FIGURE 13: Distribution of giant barred-frog Mixophyes iteratus
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 critically endangered species (IUCN 1996) that is possibly extinct. A draft recovery plan has been prepared (Martin et al. 1997) but recovery is dependant upon locating an extant population. Monitoring of 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
● 1990 onwards record ❍ pre-1990 record ■ capital cities
Conservation status: It is listed as Endangered in Qld, nationally and in the Action Plan for Australian Frogs (Tyler 1997). Taudactylus diurnus is a critically endangered species (IUCN 1996) that is possibly extinct. A draft recovery plan has been prepared (Martin et al. 1997) but recovery is dependant upon locating an extant population. Monitoring of historical locations and surveys of potential habitat will continue as part of ongoing frog conservation actions in south-eastern Qld. Stony creek frog
Litoria lesueuri
Distribution: Endeavour Falls, Cooktown NEQ (15° 25’S 145° 05’E — Covacevich and McDonald 1993) to Lederderg Gorge Vic (37° 37’S 144° 25’E — Gillespie and Hines 1999). Figure 16. Current distribution: Within the wet subtropical region of Australia, L. lesueuri remains one of the most widespread, abundant and frequently recorded stream breeding frogs. There are however reports of declines. Ingram and McDonald (1993) reported declines from rainforest streams in the Blackall and Conondale Ranges of south-eastern Qld between the late 1970’s and early 1980’s, but the species remained common in wet sclerophyll forests. This period coincides with the disappearance of T. diurnus and R. silus, and declines of other species from the same area. Following their survey in 1993 Ingram and McDonald considered that populations of L. lesueuri were recovering along these FIGURE 14: Distribution of southern dayfrog Taudactylus diurnus (after Martin, McDonald and Hines 1997). Shaded areas are: light grey -state forest or timber reserve, dark grey - national parks and conservation parks. Some towns and larger streams are also shown.
● 1990 onwards record ❍ pre-1990 record ■ towns
streams. During recent surveys of mesic forests in northeastern NSW, L. lesueuri was infrequently recorded above 400m (5% of 36 sites surveyed — Mahony unpubl. data). This observation may be additional evidence for decline of this species from montane rainforest streams, but our assessment is limited by a lack of historical data on distribution and abundance in this area. There have been no reports of declines from lower altitudes or from open forests. In the temperate zone L. lesueuri is still frequently encountered at high altitude (Gillespie and Hines 1999). Threats and biology: Gillespie and Hines (1999) summarise known or likely threats to L. lesueuri as well as knowledge of its biology and taxonomy. Conservation status: Litoria lesueuri is not listed as a threatened species in any of the states in which it occurs. Given its past decline from some rainforest streams and its ecological similarity with other declining frogs, it is recommended that L. lesueuri be included in monitoring programs, to assess changes in relative abundance. Cascade treefrog
Litoria pearsoniana
The identity of various populations referred to this species are currently under review. Recent allozyme and DNA studies (Mahony et al. unpubl. data; McGuigan et al. 1998) indicate that the population at Kroombit Tops (24° 24´S 151° 01´E — Czechura 1986, QM specimens) may be an
FIGURE 15: Distribution of southern gastric brooding frog Rheobatrachus silus (after Martin, McDonald and Hines 1997). Shaded areas are: light grey -state forest or timber reserve, dark grey - national parks and conservation parks. Some towns and larger streams are also shown.
● 1990 onwards record ❍ pre-1990 record ■ towns
57
undescribed taxon. These studies will also more clearly delineate the southern range limit of L. pearsoniana. Distribution: Kandanga State Forest SEQ (26° 26´S 152° 24´E — SEQRFA) south to Gibraltar Range NENSW (29° 31´S 152° 25´E — Mahony et al. unpubl. data), Figure 17. Current distribution: Czechura (1991) and McDonald and Davies (1990) recorded declines of L. pearsoniana in the late 1970’s to early 1980’s from the Conondale and Blackall Ranges south-eastern Qld. Corben (in McDonald 1991) suggested that this species had not suffered a conspicuous decline in the Conondale Range, but that it had disappeared from some streams in Brisbane Forest Park. Ingram and McDonald (1993) found L. pearsoniana breeding in small numbers in the Conondale, Border and Main Ranges. During their survey only two were heard during ideal weather conditions at East Kilcoy Creek (McDonald unpubl. data) where it had previously occurred in hundreds during the study of McDonald and Davies (1990). Ingram and McDonald (1993) did not find it at Kondallila Falls in the Blackall Range, although it was common there in the 1970’s (McDonald unpubl. data). There are no reports of declines in NSW. More recent studies have found the species to be reasonably widespread (Mahony 1996; NSW NPWS 1994a) or “appear to have recovered at some sites” (McGuigan et al. 1998). Surveys and monitoring carried out over the past two seasons by QPWS in south-eastern Qld largely support these findings. It is one of the most frequently recorded species along streams in mesic forests. In May 1996 L. pearsoniana FIGURE 16: Distribution of stony creek frog Litoria lesueuri in subtropical Australia
● 1990 onwards record ❍ pre-1990 record ■ capital cities
58
was abundant over several kilometres of three streams in the Conondale Range; at Bundaroo Creek 61 individuals were counted on a 100m transect (Hines unpubl. data). However, at other sites where seemingly suitable habitat exists, L. pearsoniana is currently at low densities. For example only one individual was recorded during regular monitoring of a 1 200 m transect along a rainforest stream at Cunningham’s Gap south-eastern Qld (33 censuses totalling 24 km — Hines unpubl. data). Threats: The reasons for population declines are unknown. In the Blackall and Conondale Ranges they coincide with the period in which T. diurnus and R. silus disappeared. Large areas of this species’ habitat have and continue to be degraded by introduced animals, e.g. feral pigs and domestic stock, invasion of weeds and timber harvesting (see Parris and Norton 1997 for discussion). Upstream clearing and urban development have reduced habitat and are likely to have affected downstream flow regimes and water quality in some localities (for example Kondallila Falls). Infections of a chytrid fungus (Berger et al. 1998) have been found on dead individuals from Main Range south-eastern Qld and from the closely related taxon at Kroombit Tops (Berger unpubl. data). Other ill and dead L. pearsoniana have been found in the Conondale Range but have not yet been examined (Hines unpubl. data). Biology: Relatively well documented. McDonald and Davies (1990) described adult behaviour and breeding biology. Environmental and habitat preferences are provided by McDonald and Davies (1990), Covacevich and McDonald (1993), NSW NPWS (1994a) and Parris and Norton (1997). FIGURE 17: Distribution of cascade treefrog Litoria pearsoniana
● 1990 onwards record ❍ pre-1990 record ■ capital cities
Genetic structuring within L. pearsoniana has been reported by McGuigan et al. (1998) who found significant divergence among populations from different wet forest isolates. There is no published information on population dynamics or structure.
T. pleione are included in the unpublished documents of Cunningham and James (1994) and Borsboom et al. (1998). The breeding biology is unknown and there is no information on population size, structure or genetic variation.
Conservation status: Litoria pearsoniana is currently listed as Endangered in Qld but is not considered threatened in NSW or nationally. Reassessment of its status, based on a more thorough analysis of recent survey and monitoring data, is warranted. Resolution of taxonomic problems within the group is needed, particularly for the population at Kroombit Tops. A draft recovery plan for L. pearsoniana has been prepared and implemented (Hines 1997).
Conservation status: Taudactylus pleione is currently considered Vulnerable in Qld and in the Action Plan for Australian Frogs (Tyler 1997) but is not listed nationally. Its rarity, extremely restricted distribution, coupled with possible population declines and the identification of a number of likely threats suggests that the status of this species should be reviewed for listing as Endangered. A range of interim management measures, i.e. monitoring, handling protocols, stock exclusion and fire management are being implemented pending the finalisation of the recovery plan for the species (Borsboom et al. 1998).
Mesic forest species — breeding biology unknown Kroombit tinkerfrog
Taudactylus pleione
Distribution: Very restricted distribution. Confined to eight small patches of rainforest above 600m at Kroombit Tops, south-west of Gladstone SEQ (24° 22´S 151° 01´E) (Clarke et al. in press). Current distribution: Surveys in February 1997 greatly expanded the known distribution of T. pleione from three to eight sites with all additional sites in Kroombit Tops National Park (Clarke et al. in press). The only monitored population, in the head of Kroombit Creek, appears to have declined. The species was regularly encountered prior to 1997 but T. pleione was not heard or seen at this site during the 1997/98 season despite systematic monitoring (Clarke unpubl. data). During 1997/98 little other survey and monitoring work was undertaken in the area but T. pleione was heard calling at three of the recently discovered populations. Threats: Clarke et al. (in press) and Borsboom et al. 1998 list six main potential threats to T. pleione; timber harvesting, domestic and feral animals, visitor pressure, wild fire and the unknown agent(s) responsible for declines in other Qld frogs. Timber harvesting has ceased in the catchments above all known populations. A fence to exclude domestic stock has been constructed but stock continue to impact habitat at the head of Kroombit Creek. Feral pigs, which have the potential to prey upon T. pleione or destroy habitat, have recently been found nearby. A recent wildfire burned into many rainforest patches used by this species and may at least be partially responsible for decline of the species at the monitoring site. Modified fire management procedures have now been put in place. Four species of Taudactylus from similar habitat elsewhere in Qld declined dramatically or disappeared due to unknown causes (e.g. Ingram and McDonald 1993). In May 1998, dead L. pearsoniana (sensu lato see above) were found in Kroombit Tops (Hines and Clarke unpubl. data). Chytrid fungus, found by Berger et al. (1998) associated with frog deaths and declines elsewhere in Australia and Central America, was isolated from these animals (Berger unpubl. data). This, in conjunction with the probable decline of T. pleione at the monitoring site has heightened concern for this species. Biology: Knowledge of the biology of T. pleione is extremely poor. There are only two publications concerning this species (Czechura 1986 a,b) which provide qualitative information of habitat, calling and activity patterns, at the type locality. Additional information on these aspects of the biology of
DISCUSSION Taudactylus diurnus and R. silus, species endemic to the wet subtropical region of Australia remain missing and the factors causing their disappearance are unknown. Four other obligate or facultative stream breeding frogs, A. brevis, M. fleayi, M.iteratus and L.pearsoniana, have suffered significant local or regional declines. For L. pearsoniana, there is evidence of population recovery. In the case of A. brevis we (this review; Gillespie and Hines 1999) present data for the first time that indicates a major decline in populations from the northern tablelands of NSW north along the Great Divide into southeastern Qld. The magnitude and extent of declines in these four stream breeding species is difficult to quantify due to a lack of long term systematic baseline information. Determining the cause of these declines is made more difficult as many potentially threatening processes have operated over the range of these species. In many cases the impacts of these threats are unknown, let alone likely synergistic effects. More detailed discussion of threats to stream-breeding frogs is provided by Gillespie and Hines (1999). There is no evidence of regional declines in two facultative stream breeding species, M. fasciolatus and L. lesueuri although both have declined from high altitude rainforest streams in some areas. Given the serious declines in other stream breeding frogs in eastern Australia (see also Gillespie and Hines 1999, McDonald and Alford 1999) it is important that monitoring and research projects incorporate these species due to their potential vulnerability. Eight species of mesic forest frogs, that are not dependent on streams for breeding, show no signs of decline. They are of conservation concern because of their rarity and or restricted distribution. Five lowland species are increasingly at risk due to loss or degradation of habitat. There is an urgent need to establish recovery processes for the habitat for wallum or acid frogs — many threatened vegetation communities, plants and other animals are dependent upon these environments. Generally our knowledge about the distribution, population size ,structure, natural fluctuations, and biology of the frogs reviewed is poor. This needs to be greatly improved to assist in our understanding of the potential causes of declines and other threats and ultimately to ensure recovery of populations.
59
This review has concentrated on frog species currently considered rare or threatened, and additional stream breeding species for which there was evidence of declines. This approach is somewhat subjective and potentially overlooks species which may have declined. We recommend that a thorough analysis of the patterns of change in distribution of all frog species in south-eastern Australia be undertaken urgently. Such a review needs to address a number of limitations in the available data sets. The best historical information on species’ distributions lies in our museum collections. For each specimen the identity, locality and georeference need to be checked. This process has to be coordinated by both State and Federal conservation agencies. It will result in a much more robust data set for comparison with ongoing data collection. In order to complete this process taxonomic resolution of species complexes is needed. In this review we identified Litoria cooloolensis, L. lesueuri, L. pearsoniana and L. revelata and Kyarranus spp., as requiring further taxonomic studies. For many species more robust identification characters need to be determined. Whilst refinement of historic data sets is required this and other reviews in these proceedings have repeatedly identified major gaps in our understanding of species’ distribution, ecology and threatening processes. In eastern Australia priority in addressing these gaps must be given to stream breeding forest species (see also Gillespie and Hines 1999; McDonald and Alford 1999), to assist in the identification and management of the causal agent(s) of the declines. Ongoing investigations of ill and dead frogs from southeastern Australia and elsewhere have for the first time identified a pathogen (a chytrid fungus — Berger et al. 1998) apparently causing morbidity and associated with population declines. Much work is needed to fully test the disease hypothesis. Although it is necessary to continue this research urgently, investigations of other potential agents (pollutants, climate change, UV-B and synergistic effects) must also be carried out simultaneously. The greatest threat to lowland frogs is the rapid rate of habitat loss or degradation. Some ecosystems and many other species of plants and animals are dependent upon coastal environments. This is a major issue that conservation agencies must address. An important component in future research and management of all species is population monitoring. By carefully selecting target species, monitoring sites and methods, many of the gaps in our current knowledge will be filled. Monitoring will allow more rigorous quantitative assessments of declines, (see for example McDonald and Alford 1999) as well as assisting in the investigations of threatening processes. In south-eastern Australia monitoring has largely been ad hoc and recent, despite awareness of population declines for many years. State and Federal conservation agencies need to work together to ensure the establishment and continuation of a network of systematic monitoring throughout eastern Australia. 60
ACKNOWLEDGMENTS Species locality data were provided by the Australian, Queensland and South Australian Museums, Australian National Wildlife Collection,The National Museum of Victoria, Atlas of New South Wales Wildlife and the South-east Queensland Regional Forest Assessment Fauna Database. These data made up the vast majority of records used in the production of species distribution maps. We thank the relevant staff and curators of these collections and databases for their generous assistance. A. Borsboom, M. Cunningham, B. Dadds, R. Goldingay, M. Hero, R. Knowles, C. Marshall, E. Meyer, C. Morrison and K. Thumm provided additional species records and opinions on species’ status. Staff of the Parks and Wildlife Service at Moggill also assisted especially B. Dadds with data manipulation, R. Brown and H. Preece with geographical information system support and J. Angus who typed sections of the manuscript. We thank N. Campbell, W. Drake, S. Richards and A. White for their valuable comments. Financial support was provided by Queensland Parks and Wildlife Service, NSW National Parks and Wildlife Service, Environment Australia, the University of Newcastle and the Natural Heritage Trust.
REFERENCES Anstis, M. (1981). Breeding biology and range extension for the New South Wales frog Kyarranus sphagnicolus (Anura: Leptodactylidae). Australian Journal of Herpetology 1: 1-9. Anstis, M. (1994). The larval development of Litoria brevipalmata (Anura: Hylidae). Memoirs of the Queensland Museum 37: 1-4. Aridis, J.J. (1997). A range extension for Litoria brevipalmata (Anura: Hylidae). Memoirs of the Queensland Museum 42: 310. Barker, J., Grigg, G.C., and Tyler, M.J. (1995). A Field Guide to Australian Frogs. Chipping Norton, NSW: Surrey Beatty & Sons. Berger, L., Speare, R., Daszak, P., Green, D.E., Cunningham, A.A., Goggin, C.L., Slocombe, R., Ragan, M.A., Hyatt, A.D., McDonald, K.R., Hines, H.B., Lips, K.R., Marantelli, G. and Parkes, H. (1998). Chytridiomycosis causes amphibian mortality associated with population declines in the rain forests of Australia and Central America. Proceedings of the National Academy of Science USA 95: 9031-6. Blaustein, A.R. and Wake, D.B. (1990). Declining amphibian populations: A global phenomenon? Trends in Ecology and Evolution 5: 203-204. Borsboom, A., Clarke, J. and Cunningham, M. (1998). Draft recovery plan for the Kroombit tinker frog Taudactylus pleione 1997–2001. Queensland Department of Natural Resources and Queensland Department of Environment (May 1998). Bureau of Meteorology (1989) Climate of Australia. Australian Government Publishing Service, Canberra. Catterall, C.P. and Kingston, M. (1993). Remnant Bushland of South East Queensland in the 1990’s: Its Distribution, Loss, Ecological Consequences, and Future Prospects. Institute of Applied Environmental Research, Griffith University and Brisbane City Council, Brisbane, Australia.
Clarke, J.M., Borsboom, A.C., Cunningham, M. and Hines, H.B. (in press). The recovery process for the Kroombit tinkerfrog, Taudactylus pleione. WWF South-east Queensland Rainforest Recovery Conference Proceedings.
Ehmann, H. (1997b). 1. Marsupial Frog, Assa darlingtoni. pp. 42-47 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia.
Cogger, H.G. (1996). Reptiles and Amphibians of Australia. Reed Books, Chatswood, NSW.
Ehmann, H. (1997c). 2. Wallum Froglet, Crinia tinnula. pp. 48-55 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia.
Corben, C.J. and Ingram, G.J. (1987). A new barred river frog (Myobatrachidae: Mixophyes). Memoirs of the Queensland Museum 25: 233-237. Corben, C.J., Ingram, G.J. and Tyler, M.J. (1974). Gastric brooding: Unique form of parental care in an Australian frog. Science 186: 946-947. Covacevich, J.A. and McDonald, K.R. (1993). Distribution and conservation of frogs and reptiles of Queensland rainforests. Memoirs of the Queensland Museum 34: 189-199. Cunningham, M. and James, C. (1994). Taudactylus pleione, a report on the conservation status and distribution of the Kroombit tinker frog. Department of Zoology and Centre for Conservation Biology,The University of Queensland, unpublished report to Queensland Forest Service and Queensland Department of Environment and Heritage. Czechura, G.V. (1978). A new locality for Litoria brevipalmata (Anura: Pelodryadidae) from south east Queensland. Victorian Naturalist 95: 150-151. Czechura, G.V. (1986a). A new species of Taudactylus (Myobatrachidae) from southeastern Queensland, Australia. Memoirs of the Queensland Museum 22: 299-307. Czechura, G.V. (1986b). Kroombit Tops Torrent Frog Taudactylus pleione with a key to the species of Taudactylus. Queensland Naturalist 27: 68-71. Czechura, G.V. (1986c). Distant exiles: Frogs and reptiles recorded from Kroombit Tops, southeastern Queensland. Queensland Naturalist 27: 61-67. Czechura, G.V. (1991). The Blackall-Conondale Ranges: Frogs, reptiles and fauna conservation. Chapter 19. pp. 311-324 in G. Werren and P. Kershaw (eds) The Rainforest Legacy: Australian National Rainforests Study,Volume 2. Special Australian Heritage Publication Series, 7, Commonwealth Government Printer, Canberra. Czechura, G.V. and Ingram, G.J. (1990). Taudactylus diurnus and the case of the disappearing frogs. Memoirs of the Queensland Museum 29: 361-365. Daly, G. (1995). Observations on the Tusked Frog Adelotus brevis (Anura: Myobatrachidae). Herpetofauna 25: 32-35. de Bavay, J. (1993). The developmental stages of the sphagnum frog, Kyarrunus sphagnicolus Moore (Anura: Myobatrachidae). Australian Journal of Zoology 41: 151-201. Donnellan, S. C. and Foster, R. (1997) Genetic study of the Green-thighed Frog. Unpublished report for the Roads and Traffic Authority of NSW. Drost, C.A. and Fellars, G.M. (1996). Collapse of a regional frog fauna in the Yosemite Area of California Sierra Nevada, USA. Conservation Biology 10: 414-425. Ehmann, H.(ed) (1997a). Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia.
Ehmann, H. (1997d). 17. Green-thighed Frog, Litoria brevipalmata. pp. 156-163 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Ehmann, H. (1997e). 21. Wallum Sedgefrog, Litoria olongburensis. pp. 182-187 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Ehmann, H. and Swan, G. (1985). Reproduction and development in the Marsupial Frog, Assa darlingtoni (Leptodactylidae, Anura). pp. 279-285 in G. Grigg, R. Shine and H. Ehmann (eds) Biology of Australasian Frogs and Reptiles. Surrey Beatty and Sons/Zoological Society of New South Wales, Chipping Norton. Gillespie, G.R. and Hero, J.M. (1999) Potential impacts of introduced fish and fish translocations on Australian amphibians. Pp 131-144 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Gillespie, G.R. and Hines, H.B. (1999) Status of temperate riverine frogs in south-eastern Australia. Pp 109-130 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Gillespie, G.R. and Hollis, G.J. (1996). Distribution and habitat of the Spotted Tree Frog, Litoria spenceri Dubois (Anura: Hylidae), and an assessment of potential causes of population declines. Wildlife Research 23: 49-75. Hero, J.M., Fickling, S. and Retallick, R. (1996). The tadpole of Litoria revelata Ingram, Corben and Hosmer, 1982 (Anura: Hylidae). Transactions of the Royal Society of South Australia 120: 71-73. Hillis, D.M. and Moritz, C. (1996). Molecular Systematics, Sinauer Associates Inc., Sunderland. Hines, H.B. (1997). Draft recovery plan for southern barredfrog, Fleay’s barred-frog, giant barred-frog and cascade treefrog 1998-2002. Queensland Department of Environment (June 1997). Ingram, G.J. (1983). Natural history. pp. 16-35 in M.J. Tyler (ed) The Gastric Brooding Frog. Croom Helm, London. Ingram, G.J., Anstis, M. and Corben, C.J. (1975). Observations on the Australian leptodactylid frog, Assa darlingtoni. Herpetologica 31: 425-429. Ingram, G.J. and Corben, C.J. (1975a). The frog fauna of North Stradbroke Island, with comments on the ‘acid’ frogs of the wallum. Proceedings of the Royal Society of Queensland 86: 49-54. Ingram, G.J. and Corben, C.J. (1975b). A new species of Kyarranus (Anura: Leptodactylidae) from Queensland, Australia. Memoirs of the Queensland Museum 17: 335-339. 61
Ingram, G.J., Corben, C. and Hosmer, W. (1982). Litoria revelata: A new species of tree frog from eastern Australia. Memoirs of the Queensland Museum 20: 635-637. Ingram, G.J. and McDonald, K.R. (1993). An update on the decline of Queenslands frogs. pp. 297-303 in D. Lunney and D. Ayers (eds) Herpetology in Australia. Surrey Beatty and Sons, Sydney. IUCN (1996). International Union for the Conservation of Nature (IUCN) Red List Categories. Species Survival Commission, Gland, Switzerland. James, C. (1996). Conservation genetics of island and mainland populations of the sedge frogs Litoria cooloolensis and Litoria olongburensis. Department of Zoology and Centre for Conservation Biology,The University of Queensland, unpublished final report to Queensland Department of Environment and Heritage. Katsikaros, K. and Shine, R. (1997). Sexual dimorphism in the tusked frog, Adelotus brevis (Anura, Myobatrachidae) — the roles of natural and sexual selection. Biological Journal of the Linnean Society 60: 39-51. Kikkawa, J., Ingram, G.J. and Dwyer, P.D. (1979). The vertebrate fauna of Australian heathlands — an evolutionary perspective. pp. 231- in R.L. Specht (ed.) Ecosystems of the World 9A. Heathlands and Related Shrublands. Chapter 8. Elsevier Scientific Publishing Company, Amsterdam. Kitching, R.L. and Callaghan, C. (1982). The fauna of waterfilled tree holes in box forest in southeast Queensland. Australian Entomological Magazine 8: 61-70. Knowles, R. (1994). The systematics of the endangered montane rainforest frogs of the genus Philoria (Anura: Myobatrachidae), and implications for conservation management. Honours Thesis, Department of Biological Sciences, University of Newcastle. Knowles, R. and Mahony, M. (1997a). 7. Red and Yellow Mountain Frog, Philoria kundagungan. pp. 84-89 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Knowles, R. and Mahony, M. (1997b). 8. Masked Mountain Frog, Philoria loveridgei. pp. 90-95 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Knowles, R. and Mahony, M. (1997c). 9. Sphagnum Mountain Frog, Philoria sphagnicola. pp. 96-105 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Knowles, R. and Mahony, M. (1997d). 10. A Mountain Frog, Philoria sp. nov. 1. pp. 106-111 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Knowles, R. and Mahony, M. (1997e). 11. A Mountain Frog, Philoria sp. nov. 2. pp. 112-117 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia.
62
Knowles, R. and Mahony, M. (1997f). 12. A Mountain Frog, Philoria sp. nov. 3. pp. 118-123 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Ledlin, D. (1997). Ecology of the Green-thighed Frog (Litoria brevipalmata). B. Env. Sc (Honours) Thesis. The University of Newcastle. Lemckert, F. (1999). Impacts of selective logging on frogs in a forested area of northern New South Wales. Biological Conservation 89: 321-328. Liem, D.S. (1974). A new species of the Litoria bicolor species group from southeast Queensland, Australia (Anura: Hylidae). Memoirs of the Queensland Museum 17: 169-174. Liem, D.S. and Ingram, G.J. (1977). Two new species of frogs (Anura: Myobatrachidae, Pelodryadidae) from Queensland and New South Wales.Victorian Naturalist 94: 255-262. Lips, K.R. (1998). Decline of a tropical montane amphibian fauna. Conservation Biology 12: 106-117. Mahony, M. (1992). A frog in my pocket.Wildlife Australia 24-25. Mahony, M. (1993). The status of frogs in the Watagan mountains area the central coast of New South Wales. pp. 257-264 in D. Lunney and D. Ayers (eds) Herpetology in Australia. Surrey Beatty and Sons, Sydney. Mahony, M. (1996) Survey of the distribution and abundance of declining frog species in northern New South Wales. Unpublished report to Australian Nature Conservation Agency. Mahony, M. (1999) Review of the declines and disappearances within the bell frog species group (Litoria aurea species group) in Australia. P 81-93in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Mahony, M. and Knowles, R. (1994). A taxonomic review of selected frogs of north-east NSW forests. North East Forests Biodiversity Study Report No. 3g. NSW National Parks and Wildlife Service, unpublished report. Mahony, M., Knowles, R. and Pattinson, L. (1997a). 5. Silverblue-eyed Barred Frog, Mixophyes fleayi. pp. 72-77 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc, Sydney South, Australia. Mahony, M., Knowles, R. and Pattinson, L. (1997b). 6. Goldeyed Barred Frog, Mixophyes iteratus. pp. 78-83 in H. Ehmann (ed) Threatened Frogs of New South Wales: Habitats, Status 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. Weatherby (ed.) Australian Inland Waters and Their Fauna. ANU Press, Canberra, Australia. Martin, W.E., McDonald, K.R. and Hines, H.B. (1997). Recovery plan for southern platypus frog and southern dayfrog. Unpublished report to Environment Australia. McDonald, K.R. (1974). Litoria brevipalmata, an addition to the Queensland amphibian list. Herpetofauna 7: 2-4. McDonald, K.R. (1991). Report of a workshop on declining frog populations in Queensland. Queensland National Parks and Wildlife Service, unpublished report.
McDonald, K.R. (1992). Distribution patterns and conservation status of north Queensland rainforest frogs. Queensland Department of Environment and Heritage Conservation Technical Report 1: 1-51. McDonald, K.R. and Alford, R.A. (1999) A review of declining frogs in northern Queensland. Pp 14-22 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. McDonald, K.R. and Davies, M. (1990). Morphology and biology of the Australian tree frog Litoria pearsoniana (Copland) (Anura: Hylidae).Transactions of the Royal Society of South Australia 114: 145-156. McDonald, K.R. and Miller, J.D. (1982). On the status of Lechriodus fletcheri (Boulenger) (Anura: Leptodactylidae) in northeast Queensland.Transactions of the Royal Society of South Australia 106: 220. McEvoy, J.S., McDonald, K.R. and Searle, A.K. (1979). Mammals, birds, reptiles and amphibians of the Kilcoy Shire, Queensland. Queensland Journal of Agricultural and Animal Sciences 36: 167-180. McGuigan, K., McDonald, K., Parris, K. and Moritz, C. (1998). Mitochondrial DNA diversity and historical biogeography of a wet forest-restricted frog (Litoria pearsoniana) from mid-east Australia. Molecular Ecology 7: 175-186. Meyer, E.A. (1997). Acid tolerance in larval Litoria fallax and L. cooloolensis. B.Sc (Hons) Thesis, University of Queensland. Moore, J.A. (1961). The frogs of eastern New South Wales. Bulletin of the American Museum of Natural History 121: 149-386. Nattrass, A.E.O. and Ingram, G.J. (1993). New records of the rare Green-thighed Frog. Memoirs of the Queensland Museum 33: 348. NSW NPWS (1994a). Fauna of north-east NSW forests. North East Forests Biodiversity Study Report No. 3. NSW National Parks and Wildlife Service, unpublished report. NSW NPWS (1994b). Results of vertebrate fauna surveys of north-east NSW forests North East Forests Biodiversity Study Report No. 3a,Vol. 1, Site and transect based methods. NSW National Parks and Wildlife Service, unpublished report. NSW NPWS (1994c). Results of vertebrate fauna surveys of north-east NSW forests. North East Forests Biodiversity Study Report No. 3a,Vol. 2, Incidental records. NSW National Parks and Wildlife Service, unpublished report.
Pyke, G.H. and Osborne, W.S. (1996). The green and golden bell frog Litoria aurea: biology and conservation. Australian Zoologist 30: 132-258. Richards, S.J., McDonald, K.R. and Alford, R.A. (1993). Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conservation Biology 1: 66-77. Smith, A.P., Hines, H.B., Pugh, D. and Webber, P. (1989a). Mammals, reptiles and amphibians of the rainforests of the Focal Peak region. Department of Ecosystem Management, University of New England, Armidale. Smith, A.P., Hines, H.B. and Webber, P. (1989b). Mammals, reptiles and amphibians of the rainforests of the Mount Warning Caldera region. Department of Ecosystem Management, University of New England, Armidale. Straughan, I.R. (1966). An analysis of species recognition and species isolation in certain Queensland frogs. PhD Thesis, University of Queensland. Straughan, I.R. (1968). A taxonomic review of the genus Mixophyes, (Anura, Leptodactylidae). Proceedings of the Linnean Society of New South Wales 93: 52-59. Straughan, I.R. and Main, A.R. (1966). Speciation and polymorphism in the genus Crinia Tschudi (Anura, Leptodactylidae) in Queensland. Proceedings of the Royal Society of Queensland 78: 11-28. Tyler, M.J. (1972). A new genus for the Australian leptodactylid frog Crinia darlingtoni. Zoologische Mededelingen Rijksmuseum Van Naturolijke Historie Te Leiden 47: 193-201. Tyler, M.J. (1985). Gastric brooding: A phenomenon unique to Australian frogs. Search 16: 157-158. Tyler, M.J. (1997). The Action Plan for Australian Frogs. Wildlife Australia, Canberra, ACT. Tyler, M.J. and Davies, M. (1985a). A new species of Litoria (Anura: Hylidae) from New South Wales, Australia. Copeia 1: 145-149. Tyler, M.J. and Davies, M. (1985b). The Gastric Brooding Frog Rheobatrachus silus. pp. 469-470 in G. Grigg, R. Shine and H. Ehmann (eds) Biology of Australasian Frogs and Reptiles. Surrey Beatty and Sons/Royal Zoological Society of New South Wales, Chipping Norton.
NSW NPWS (1995).Vertebrates of upper north-east NSW. NSW National Parks and Wildlife Service, unpublished report to the Natural Resources Audit Council.
Watson, G.F., Littlejohn, M.J., Hero, J.M. and Robertson P. (1991). Conservation status, ecology and management of the spotted tree frog (Litoria spenceri). Arthur Rylah Institute Technical Report Series No 116. Department of Conservation and Environment,Victoria.
Osborne, W.S. (1989). Distribution, relative abundance and conservation status of Corroboree Frogs, Pseudophryne corroboree Moore (Anura: Myobatrachidae). Australian Wildlife Research 16: 537-547.
Watson, G.F. and Martin, A.A. (1973). Life history, larval morphology and relationships of Australian leptodactylid frogs. Transactions of the Royal Society of South Australia 97: 33-45.
Osborne, W.S., Hunter, D.A. and Hollis, G.J. (1999). Polulation 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.
Webb, G.A. (1989). Notes on the biology and conservation of Philoria sphagnicolus (Moore 1958) (Anura: Myobatrachidae). Herpetofauna 19: 1-6.
Parris, K. and Norton,T. (1997). The significance of State Forests for conservation of Litoria pearsoniana (Copland) and associated amphibians. pp. 521-526 in P. Hale and D. Lamb (eds) Conservation Outside Nature Reserves. Centre for Conservation Biology,The University of Queensland.
White, A.W. and Pyke, G.H. (1996). Distribution and conservation status of the green and golden bell frog Litoria aurea in New South Wales. Australian Zoologist 30: 177-189. Wotherspoon, D. (1981). Note on the range of the Sphagnum Frog Philoria sphagnicolus. Herpetofauna 2: 36.
63
The status of Rainforest Stream Frogs in north-eastern New South Wales: decline or recovery? Ross Goldingay, David Newell and Mark Graham*
ABSTRACT Rainforest stream-dwelling frogs have declined dramatically throughout eastern Australia.We collected baseline information on the distribution and relative abundance of these frogs in northeastern New South Wales.We established 100 m transects along rainforest streams at 54 sites and traversed these transects at night, counting all stream-dwelling frogs seen and heard
commonly encountered on forest tracks at night. Mixophyes fasciolatus was detected at 29 sites with an average abundance of 2.9 individuals per transect. The distribution and abundance of these species were not significantly influenced by elevation. Single surveys of sites in early 1999 confirmed the distribution data. Mixophyes iteratus and Mixophyes fleayi were each found at nine sites from 1997 to 1999. The former was confined to low
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 There has been much recent attention devoted to assessing the status of amphibian populations in many parts of the world (e.g. Richards et al. 1993; Drost and Fellers 1996). This has stemmed from the recognition that many species have experienced sudden and unexplained declines, and some species have disappeared totally (Carey 1993; Pounds and Crump 1994; Lips 1998). In Australia, as many as 9 frog species have disappeared in the last 15 years (Ingram and McDonald 1993; Laurance et al. 1996; Mahony 1996). The seriousness of this situation is reflected by the fact that no frog extinctions are known from the last 100 years in Australia. Many frog species are still in decline and there is a need to assess the status of frog communities throughout Australia. The importance of monitoring programs has been demonstrated by studies in north Queensland (Richards et al. 1993;Trenerry et al. 1994). Similar studies of rainforest stream frogs are urgently needed at other locations because this group seems especially vulnerable to population declines. The aim of the present study was to locate suitable sites and establish a monitoring program for rainforest stream-dwelling frogs in north-eastern New South Wales (NSW). This area included three species, Litoria pearsoniana, Mixophyes fleayi and Mixophyes iteratus, which have been reported to have suffered serious declines (Ingram and McDonald 1993; Mahony 1996) and which are recognised as threatened or endangered species (Lunney et al. 1996; Parris and Norton 1997).
METHODS Study Area The study area included forested areas in the Border Ranges National Park (NP), Mt Warning NP, Limpinwood Nature Reserve, Nightcap NP, Mebbin State Forest (SF), Nullum SF, Mooball SF,Whian Whian SF, Richmond Range NP and Toonumbar NP (Figure 1).Within all areas except Mebbin SF, where some streams may be ephemeral, transects were established along permanent streams with a rainforest verge. Between one and 12 transects measuring approximately 100 m in length were established in each area (Table 1), depending on stream availability and access. All but eight of the stream sites were located within 500 m of an access road.
Survey Methodology Site selection began in late 1996 when a small number of sites were selected and surveyed in Whian Whian SF and Border Ranges NP. Most sites were selected in 1997 and surveys were conducted between January 1997 and April 1998. Because this study aimed to provide baseline data for long-term monitoring, 2–3 surveys were conducted at night at each site under various weather conditions to ensure that we obtained appropriate data for a range of species. This included at least one night when rain occurred. Two sites in Nightcap NP and seven sites in Whian Whian SF were surveyed at least 14 times each between late November 1997 and mid-March 1998. The majority of these sites, in addition to a few extra sites, were surveyed on a single night in January-February 1999. A total of 54 transects were included in the study overall (Appendix 1); 47 were surveyed in 1997–98 while 49 were surveyed in 1999. Each transect was searched using a 30 watt spotlight for a minimum of 20 personminutes (number of observers x time).
Many species of frog were recorded during this study including Adelotus brevis, Assa darlingtoni, Lechriodus fletcheri, Litoria spp., Mixophyes spp. and Philoria spp. However, we only report on the stream-breeding species Litoria pearsoniana, Litoria lesueueri and three species of Mixophyes. We present the maximum counts for each species at each survey site for the 1997–98 period. An additional count was made at each creek site if a breeding aggregation was heard away from the fixed transect site. Searching was also conducted beyond the fixed transect at sites where there were historic records of M. fleayi in order to more fully assess its presence. A set of handling and disinfection procedures was followed. Boots were disinfected between catchments by means of a dilute bleach mixture. Handling of frogs was kept to a minimum and plastic bags (used only once) were everted over hands if frogs were handled.
RESULTS Status of Tree Frogs Litoria pearsoniana was detected at 39 sites in 1997–98 (Table 1) with an average abundance of 6.9 individuals per transect. Large numbers (16–40) of individuals were recorded at several sites. Litoria pearsoniana was detected at 35 sites in 1999 (Table 1) with an average abundance of 4.4 individuals per transect. Litoria lesueuri was detected at30 sites in 1997–98 (Table 1) with an average abundance of 4.3 individuals per transect. It was detected at 23 sites in 1999 with an average abundance of 2.7 individuals per transect. The lower abundance in 1999 of both species was likely a result of only a single survey being conducted at each site which would have included periods of lower activity. Litoria lesueuri was more widespread and abundant than indicated by these data. It was regularly observed crossing roads in the Border Ranges NP, Mebbin SF and Whian Whian SF when these areas were traversed at night, and the occurrence of large numbers of juveniles indicated that recent breeding had been successful. Each of these species were found at sites that ranged in elevation from 70 m to 780 m.
Status of the Barred Frogs Mixophyes fasciolatus was detected at 31 sites in 1997–98 (Table 1) with an average abundance of 2.9 individuals per transect. It was detected at 24 sites in 1999 (Table 1) with an average abundance of 3.5 individuals per transect. Sites where this species was found ranged in elevation from 70 m to 780 m. Mixophyes iteratus was found at only seven sites in 1997–98, all in Mebbin SF at low elevation (Table 1). It had an average abundance of 4.2 individuals per transect. Juvenile frogs were detected at three sites, indicating that recent breeding had been successful. At one site no adults were detected. In 1999, M. iteratus was again found at seven sites in Mebbin SF but this included two sites where it was not detected the previous year. It was also found at two other locations which were not included in the 1997-98 survey (Table 1). It had an average abundance of 3.4 individuals per transect.
65
FIGURE 1: Location of broad survey areas in north-eastern New South Wales.
Byron Bay
Key: 1. 2. 3. 4. 5. 6. 7.
Whian Whian State Forest Nightcap National Park Nullum State Forest Mount Jerusalem National Park Mount Warning National Park Wollumbin State Forest Mebbin State Forest 66
8. 9. 10. 11. 12.
Border Ranges National Park Mooball State Forest Toonumbar National Park Richmond Range National Park Bungdoozle Flora Reserve (Richmond Range State Forest)
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)
Mooball SF M1 (50) M2 (80) M3 (90) Mt Warning NP MW1 (430) MW2 (220) Nullum SF N1 (270) N2 (130) N3 (210) N4 (230) Mebbin SF Mb1 (135) Mb2 (145) Mb3 (200) Mb4 (170) Mb5 (155) Mb6 (170) Mb7 (165) Mb8 (165) Mb9(155) Mb10 (150) Mb11 (130) Whian Whian SF W1 (180) W2 (380) W3 (540) W4 (360) W5 (330) W6 (370) W7 (320) W8 (160) W9 (70) W10 (250) Nightcap NP Nc1 (200) Nc2 (180) Nc3 (550) Nc4 (600) Nc5 (780) Nc6 (520) Nc7 (460) Nc8 (470) Border Ranges NP B1 (570) B2 (750) B3 (150) B4 (730) B5 (240) B6 (675) B7 (880) Limpinwood NR L1 (560) L2 (500) L3 (380) Richmond Range NP RR1 (460) RR2 (560) RR3 (300) Tooloom NP T1 (710) T2 (690) Toonumbar NP T1 (210)
L. pearsoniana
A. – 1 1
L. lesueuri
B. – 2 1
A. – – –
B. 0 (2) – 5
M. fasciolatus
M. iteratus
M. fleayi
A. 1 5 –
B. 3 4 2
A. – – –
B. – – –
A. – – –
B. – – –
4 7
0 (2) 5
– 3
– –
– 1
– –
– –
– –
1 –
2 –
2 (6) 2 8 2 (6)
0 (4) – 0 (6) 2
– 0 (1) – –
– 0 (1) – –
2 0 (1) – –
– – 0 (2) –
– – – –
– – – –
– – – –
– – – –
– – – 2 1 – 2 (14) – 7 40 –
5 1 3 3 8 5 9 7 (1) 3 6 14 (8)
4 1 1 1 3 4 6
– – 1 – 1 1 1
5 – – – 1 2 – j 4 0 (1) 9
– – – 0 (1) 1 1 2 4 (7) – 1 (2) 8
– – – – – – – – – – –
– –
1 – –
2 0 2 – – – – 0 – 2 –
22 10 2 4 3 5 2 5 4 4
1 (4) 2 2 – 6 – – 2 – 16
4 2 – 5 4 3 7 2 4 3
8 2 – 2 4 5 2 4 4 3
– 1 – – – – – – – –
– – – – – – – – – –
– –
– –
4
– – – – – – – – – –
– – – – – – –
7 12 3 2 2 5 – nd
1 6 2 – nd – 2 6
3 13 – 2 2 – – nd
2 1 0 (3) 0 (3) nd – – –
3 3 3 2 3 2 3 nd
3 – 4 (3) 0 (2) nd 3 8 8
– – – – – – – nd
– – – nd – – –
– – – – – – – nd
3 – – – nd – 15 2
11 5 4 12 7 nd –
6 1 (2) nd nd – – –
3 – – – 5 nd –
0 (5) – nd nd 2 – –
– – 3 – 7 nd –
– – nd nd 3 – –
– – – – – nd –
– – nd nd – – –
2 1 – – – nd –
– 0 (5) nd nd – 16 –
9 10 8
nd 1 –
4 – 6
nd – 0 (2)
2 2 5
nd 2 2
– – –
nd – –
– – –
nd – – – – –
2 3 0 (1) –
– – 1 – – – 11 0 (2) – – 3 1 – – 1 1 – – – 2 3
–
(1) (3)
(1) (2)
– – – – –
18 nd –
2 (4) 7 –
– nd –
2 0 (1) 3
– nd –
0 (7) 2 8
– nd –
– – 6
– nd –
nd nd
– –
nd nd
– –
nd nd
– 3
nd nd
– –
nd nd
16
3 (4)
22
2
–
5
–
–
–
– – – – – – –
8 10 (10) –
67
Mixophyes fleayi was found at nine sites overall; three sites in 1997-98 and eight sites in 1999, ranging from 200 m to 765 m elevation (Table 1). This species was more abundant in 1999 with >5 individuals detected at four sites. These observations represent new records for this species at Mt Warning NP, at Tuntable Falls in Nightcap NP and at Kangaroo Ck in Tooloom NP. These results confirm that this species is rare in the study area and that populations are widely separated.
Influence of Elevation Elevation has been implicated as an influence on the decline of stream-dwelling frogs in north Qld (Richards et al. 1993). We investigated whether elevation influenced the distribution and abundance of our five target species by dividing up the sites into high (>300m) and low (0.1). Its abundance was not significantly different (t=0.44, P =0.65) at high (6.4 ± 1.1 individuals) versus low (7.4 ± 2.0) elevation sites. Litoria lesueuri was detected at 10 high and 20 low elevation sites but this distribution was independent of elevation (G=2.28, P >0.1). Its abundance was not significantly different (t=0.52, P =0.61) at high (3.8 ± 0.6 individuals) versus low (4.7 ± 1.2) elevation sites. Mixophyes fasciolatus was detected at 13 high and 18 low elevation sites but this distribution was independent of 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 low (2.9 ± 0.5) elevation sites. Mixophyes iteratus only occurred at sites ≤300 m elevation. In contrast, all but one of the nine sites where M. fleayi was found were >400 m elevation.
DISCUSSION Frog Declines Frogs and other amphibians have undergone sudden and pronounced declines in protected rainforest areas in many parts of the world (Heyer et al. 1988; McDonald 1990; Ingram and McDonald 1993; Richards et al. 1993; Pounds and Crump 1994; Lips 1998). In Australia, there has been speculation about the cause of the rainforest frog declines and its pattern of spread (see Laurance et al. 1996, 1997; Alford and Richards 1997; Hero and Gillespie 1997). The focus of this debate has been whether a water-borne pathogen is the most probable cause of the decline, and the environmental conditions over which such a pathogen may be virulent. There appears to be some consensus that the frog declines have spread from the Conondale Ranges in about 1979, based on the disappearance of the locally endemic Taudactylus diurnus and Rheobatrachus silus (Ingram and McDonald 1993; Laurance et al. 1996; Mahony 1996). Most attention has been devoted to the northward progression of disappearing species from this area in southeast 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
and penetrated the rainforest areas of north-eastern NSW by the early 1980’s and that it has been present here for well over 10 years. Indeed, Mahony (1993) reported the disappearance of several stream-dwelling frogs in the mid1980’s from the central coast of NSW. This may suggest that the most virulent phase, when species decline rapidly, has passed and that there may be hope to recover those species that have declined but are still present. However, it would be unwise to be too optimistic because species that are reduced to very low levels of abundance may be demographically unstable. Surveys such as those employed here should be continued to provide an on-going appraisal of the relative stability of affected species. In the present study, most sites were characterised by small numbers of individuals. There were only three sites where more than 20 individuals of any species were found. The absence of any studies reporting densities of these species prevent a resolution of how indicative these abundances are. These abundances may reflect a post-decline environment or continued decline, assuming that a pathogen has been present. Sick or dead frogs have been reported from a small number of locations throughout our study area such as Mebbin SF (Berger pers. comm.). Studies in Qld have suggested that species’ declines were more apparent at high elevation, but we were unable to find an influence of elevation on distribution and abundance for three species (L. pearsoniana, L. lesueuri, M. fasciolatus). Moreover, the two species of greatest conservation concern (M. iteratus, M. fleayi) were found almost exclusively at either low or high elevation.We are now entering a period when active management of many populations is needed and monitoring will be required to evaluate the success or otherwise of any management activities.
Survey Methodology This study aimed to employ a survey method equivalent to that used by Richards et al. (1993). It was not possible to survey streams frequently, but rather it was intended to obtain an index of frog abundance at each site. It became apparent that the activity levels of different species were not synchronised at a given site, largely due to different triggers for breeding. For example, the calling of male L. pearsoniana appears to peak during rainfall events while that of M. iteratus and M. fleayi appears to precede rainfall (pers. obs.). To accomodate these differences, we surveyed all sites during wet and dry periods in 1997-98. In early 1999, we were only able to survey a site once so these data should be interpreted more cautiously. One short coming of our survey methodology was that transects were located arbitrarily at a site. This may have led to the exclusion of important microhabitat components of a particular species. Often suitable microhabitats (e.g. fringing vegetation) occurred just outside the transect. The practical way to account for this was to include an additional count of the nearest breeding aggregation detected outside the transect if one was heard calling. This technique provided many useful data that produced a more accurate assessment of relative abundance and of the distribution across all survey sites.
Tree Frogs The most widespread and abundant species encountered during this survey were L. pearsoniana and L. lesueuri. Ingram and McDonald (1993) suggested that L. pearsoniana was a species of some conservation concern. They found that it had disappeared from the Blackall Range and was present in low numbers at other sites where it had previously been common. Laurance et al. (1996) suggested that it had declined by 90%. It is now listed as an endangered species in Qld (Parris and Norton 1997). Litoria lesueuri has also declined in many areas of south-east Qld but appears to be recovering (Ingram and McDonald 1993). These observations suggest that these two species will be important for evaluating the recovery of rainforest frog populations and may provide insights that can be applied to other species. We found that L. pearsoniana was widespread in the study area and relatively abundant at a number of sites. This situation offers good potential for assessing its population stability and using it as an indicator of any changes in management. Litoria lesueuri was also widespread in the study area but generally occurred at low abundance along the transects. This may not be an accurate reflection of the abundance of this species because it was commonly encountered when driving through the forest at night, under both wet and dry conditions. Therefore, the survey method may not be suitable for assessing changes in the population size of this species. Our surveys in 1997-98 suggest that the distribution and abundance of each of these species were not influenced by elevation.
Barred Frogs Mixophyes fasciolatus was the most common of the barred frogs, being found at 42 of 53 sites over the period of this study. Its distribution and abundance in 1997–98 was not influenced by elevation. This species is able to breed in ponds as well as streams (Mahony 1993) from which it may have derived some benefit which allows it to persist at many sites. Mahony (1993) suggested that M. fasciolatus was not as abundant on the central coast of NSW as it used to be. Whether its abundance has declined in north-east NSW is unknown but we have now established base-line data which can be used for comparison by future studies. Ingram and McDonald (1993) reported that M. iteratus may have disappeared from highland sites and that its status was uncertain. Mahony et al. (1997b) reported that this species had disappeared from much of its southern range. We only detected this species at low elevation sites, at three broad locations. Mahony et al. (1997b) recorded 5 adults at Byrrill Ck, in Mebbin SF. We recorded 9 adults there in 1998 and 8 adults in 1999. We found that M. iteratus was widespread in Mebbin SF. This species could be readily detected during most visits to this area, and the presence of juveniles at several sites suggests that this population may currently be stable. Mixophyes fleayi has a restricted distribution in north-east NSW and south-east Qld (Corben and Ingram 1987). Ingram and McDonald (1993) noted that M. fleayi was a species of conservation concern. Mahony et al. (1997a) noted the very restricted distribution and extreme rarity of M. fleayi in northeastern NSW. The current study included each of the sites where this species was detected by Mahony et al. (1997a) and
a further site (Byrrill Ck) where it was apparently detected by Manning (pers. comm.).We detected M. fleayi at the Border Ranges sites in both years of the study but only in 1999 at Terania Creek, Nightcap NP. We were unable to locate it at Byrrill Ck, Mebbin SF, despite >15 visits to this site.We detected M. fleayi at two new locations (Mt Warning NP and Tooloom NP). Each of these represent significant range extensions because these sites are well isolated from other locations. The abundance of M. fleayi was extremely low in 1997–98 with only 1–2 males detected at each site. However, larger numbers were detected in early 1999, which may have been due to the above-average rainfall experienced in that period and also in late 1998. Mahony et al. (1997a) reported six individuals from Terania Creek in 1994 but none in 1996. We detected none there between late 1996 and March 1998, despite 17 visits under various weather conditions. However, three individuals were detected in 1999. Mahony et al. (1997a) reported no individuals at Brindle Ck in 1994-1996 but we detected one individual in 1997-98 and five in 1999. Moderate numbers (8–16) of M. fleayi were detected at four sites in early 1999. This suggests that these may be more important sites and will require continued monitoring.
Management Implications There are several implications arising from this study: 1. that surveys of these sites must be continued on an annual basis to provide an on-going assessment of the stability of the populations of three key species (M. fleayi, M. iteratus, and L. pearsoniana), 2. that some assessment be made as to whether the factor that led to the decline of these species is still present and virulent, and 3. that key locations be identified for detailed management intervention. We assume that the rainforest stream-dwelling frogs of north-east NSW have been subject to some agent of decline for at least 10 years (see above). This suggests that these frogs have survived the most serious period of decline, unlike many species in Qld. Surveys must be continued to determine whether these species are undergoing a more gradual decline or whether they may be recovering. In the case of M. fleayi, it has only ever been known from a small number of sites in NSW which makes it difficult to assess whether it has declined here. The discovery of new sites is the result of more intensive survey, rather than an expansion in its range. The presence of M. fleayi at low and high density offers the opportunity to investigate factors that may limit its abundance. Detailed population studies at several sites should be initiated for M. fleayi and M. iteratus which are both listed as threatened in NSW. This would provide some calibration to the population indices generated by occasional monitoring surveys. Monitoring of populations is a first step to assessing the status of populations and identifying focal species for management. The second step is to determine whether an agent of decline is still present in the environment of these species. Thus, although the factor(s) causing declines in this frog assemblage is not resolved, it is still possible to ask whether the factor may still be operating. This could be assessed by conducting experiments using tadpoles of 69
affected species (e.g. L. pearsoniana). Tadpoles could be housed at a variety of sites to assess development and survival. If sites are identified where survival is high, then translocations could be considered to boost populations at those sites or assistance could be provided to reduce mortality of tadpoles born in situ. Any attempt at translocation should be mindful of mixing gene pools and should limit translocation to relatively short distances (e.g. 5–10 km). For species such as M. fleayi and M. iteratus, key sites should be identified where active management of their populations could be pursued. This may initially include trying to enhance tadpole survival in situ and ex situ, but may also involve the identification of sites for translocation studies. In the case of M. fleayi, it is likely that captive breeding will be needed to prevent the loss of remaining individuals in NSW. This study has shown that there are currently several sites available in north-east NSW for further studies directed at resolving management issues for each of these species.
ACKNOWLEDGEMENTS We thank Lance Tarvey (NPWS) for sharing our enthusiasm for this study. The NPWS Declining Frog Working Group provided financial support that facilitated the field surveys in early 1999. We thank Drs Steve Richards and Arthur White for many constructive comments on a draft of this paper.
REFERENCES Alford, R.A. and Richards, S.J., (1997) Lack of evidence for epidemic disease as an agent in the catastrophic decline of Australian rain forest frogs. Cons Biol, 11: 1026-29. Barker, J., Grigg, G.C. and Tyler, M.J., (1995) A Field Guide to Australian Frogs. Surrey Beatty & Sons, Chipping Norton. Carey, C., (1993) Hypothesis concerning the causes of the disappearance of boreal toads from the mountains of Colorado. Cons Biol, 7: 355-62. Corben, C.J. and Ingram, G.J., (1987) A new barred frog (Myobatrachidae: Mixophyes). Mem Qld Mus, 25: 233-37. Drost, C.A. and Fellers, G.M., (1996) Collapse of a regional frog fauna in the Yosemite Area of the California Sierra Nevada, USA. Cons Biol, 10: 414-25. Hero, J-M. and Gillespie, G.R., (1997) Epidemic disease and amphibian declines in Australia. Cons Biol, 11: 1023-25. Heyer, W.R., Rand, A.S., Goncalvez da Cruz, C.A. and Peixoto, 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
Lips, K.R., (1998) Decline of a tropical amphibian fauna. Cons Biol, 12: 106-17. Lunney, D., Curtin, A., Cogger, H. G. and Dickman, C. R., (1996) An ecological approach to identifying the endangered fauna of New South Wales. Pac Cons Biol, 2: 212-31. Mahony, M.J., (1993) The status of frogs in the Watagan Mountains area, the central coast of New South Wales. Pp. 257-64 in Herpetology in Australia: a diverse discipline. ed by D. Lunney and D. Ayers. Surrey Beatty & Sons, Chipping Norton. Mahony, M.J., (1996) The decline of the Green and Golden Bell Frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust Zool, 30: 237-47. Mahony, M., Knowles, R. and Pattinson, L., (1997a) 6. Goldeyed barred frog. Pp. 78-83 in Threatened frogs of New South Wales: habitats, status and conservation. ed by H. Ehmann. Frog and Tadpole Study Group of NSW, Sydney South. Mahony, M., Knowles, R. and Pattinson, L., (1997b) 5. Silverblue-eyed barred frog. Pp. 72-77 in Threatened frogs of New South Wales: habitats, status and conservation. ed by H. Ehmann. Frog and Tadpole Study Group of NSW, Sydney South. McDonald, K.R., (1990). Rheobatrachus Liem and Taudactylus Straughan and Lee (Anura: Leptodactylidae) in Eugella National Park, Queensland: distribution and decline. Trans Roy Soc South Aust, 114: 187-94. Parris, K.M. and Norton,T.W., (1997) The significance of State Forests for conservation of Litoria pearsoniana (Copland) and associated amphibians. Pp. 521-26 in Conservation Outside Nature Reserves. ed by P. Hale and D. Lamb. Centre for Conservation Biology,The University of Queensland, Brisbane. Pounds, J.A. and Crump, M.L., (1994) Amphibian declines and climate disturbance: the case of the golden toad and the harlequin frog. Cons Biol, 8: 72-85. Richards, S.J., McDonald, K.R. and Alford, R.A., (1993) Declines in populations of Australia’s endemic tropical rainforest frogs. Pac Cons Biol, 1: 66-77. Trenerry, M.P., Laurance, W.F. and McDonald, K.R., (1994) Further evidence for the precipitous decline of endemic rainforest frogs in tropical Australia. Pac Cons Biol, 1: 150-53.
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
Mooball SF M1 M2 M3 Mt Warning NP MW1 MW2 Nullum SF N1 N2 N3 N4 Mebbin SF Mb1 Mb2 Mb3 Mb4 Mb5 Mb6 Mb7 Mb8 Mb9 Mb10 Mb11 Whian Whian SF W1 W2 W3 W4 W5 W6 W7 W8
Site Name
AMG
Site Number
Palmvale Spur Barooka 1 Barooka 2
54630 686100 54600 685930 54600 685900
Breakfast Creek Rest Area
52780 685850 52930 685880
S Chowan 1 S Chowan 2 Sand Ridge Scrub
53850 53720 53480 53510
685030 685020 684870 684810
Swifts N Swifts S Tank S Kolonga O’Connors Bullocks Head Plantation N1 Plantation N2 Plantation SE Lemon Tree Byrill Ck
51620 51620 51630 51740 51750 51700 51740 51740 51810 51860 51870
685480 685460 685380 685400 685380 685300 685200 685190 685170 685140 685320
Big Scrub FR Duffs Tungun Rocky Ck 1 Boomerang Ck Boggy Ck Minyon Boomerang FR
53250 53600 53350 53530 53600 53760 53780 53220
683170 683130 683950 683770 683380 683520 683510 683620
Site Name
Whian Whian (cont.) W9 Quandong W10 Rocky Ck 2 Nightcap NP Nc1 Protestors Nc2 Terania Nc3 Bat Cave Nc4 McNamaras Nc5 Geebung Nc6 Mulgum Nc7 Tuntable 1 Nc8 Tuntable 2 Border Ranges NP B1 Sheepstation FR B2 Brindle Ck B3 Lynches B4 Gradys Ck B5 Sawpit Ck B6 Long Ck B7 Collins Ck Limpinwood NR L1 Hopping Dicks L2 Hidden Ck L3 Upper Oxley Richmond Range NP RR1 Cambridge Plateau RR2 Bungdoozle RR3 Peacock Ck Tooloom NP T1 Kangaroo Ck 1 T2 Kangaroo Ck 1 Toonumbar NP Tr1 Iron Pot Ck
AMG
53750 683300 53460 683530 53040 53040 53160 53190 53230 52730 52890 52910
683930 683970 683950 684030 684160 684180 684120 684110
50340 50670 50250 50730 48260 48570 51370
685790 686080 685540 686270 686170 686770 685420
51850 687400 51930 686720 51530 686800 47410 681230 47080 683530 47220 682950 44410 684980 44090 684950 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 Surveys for frogs were performed at water bodies in
at more than 30 sites, but the majority of frogs were recorded on less than 10 occasions. The five
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 forested areas.
zones. Recently, additional protection has been 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 State Forests of NSW Management Area report (Forestry Commission NSW 1980), but the general features are as follows. The DMA covers 213 338 hectares of an escarpment on the Great Dividing Range in northern NSW (see Figure 1). The land rises from 600m elevation in the south-east to over 1000m in its north-western corner, with annual rainfall decreasing along the same gradient from 2000 mm to 700mm (Forestry Commission NSW 1980). Temperatures are mild with a mean minimum of 14.5°C and maximum of 26°C. The majority of land is privately owned, but 95 886 hectares are located within state forest and crown timber lands and a further 46 885 hectares are held in National Parks and Nature Reserves found wholly or partly within the DMA (State Forests NSW 1995). The forest types in the area vary from warm temperate rainforests through to dry open sclerophyll forests. Logging of these forests commenced in the 1900s and increased in intensity until the 1930s and 1940s as a result of increasing mechanisation. Currently, logging is of a selective nature under a rotation cycle consisting of thinning operations at 15 and 30 years and a cut of saw quality logs at 60 years. Low intensity fires are prevalent in the dryer forests, both as a fire suppression mechanism and as a result of the activities of graziers, but the moister forests are rarely subject to fire.
maintenance of a minimum of 50% undisturbed habitat in harvested compartments has provided further protection to probable foraging and shelter habitats. Specific research is required to determine how forestry activities affect frogs and so determine if these protective measures will be as effective as planned through the long term.
INTRODUCTION Frogs are a poorly known component of the New South Wales (NSW) vertebrate fauna with relatively few comprehensive and systematic surveys having been performed for this group. This poor state of our knowledge on both the status of species and even their basic ecology is of concern given the recent noted serious declines in Australia (Mahony 1993; Richards et al. 1993) and the rest of the world (Pechmann and Wilbur 1994). In 1992, the initiation of the Threatened Species (Interim) Protection Act (TSIP Act) by the NSW government provided for the general protection of frogs and the targeted protection of threatened species which were listed under “Schedule 12” of the Act. Importantly, the process of determining which species required listing as threatened made clear the very limited knowledge available on the biology and status of most frog species (Lunney et al. 1996). It also provided a direction to determine the current status of threatened species, the processes threatening their survival and the means to conserve each species into the future. In 1993, as part of a series of Management Area impact assessments, the Forestry Commission of New South Wales (now State Forests of New South Wales or SFNSW) commenced a major amphibian survey within the Dorrigo Management Area (DMA) of northern NSW. This work covered the requirements of the TSIP and Environmental Protection and Assessment (EPA) Acts. The surveys in this study covered all land tenures within the DMA and were performed over an extended period of time in order to cover varying seasons and weather conditions, thus maximising the opportunities to locate frogs. The information gained was used to assess the habitat requirements of animals present within the DMA, and to assess the potential effects of forestry operations on all species present, particularly those considered to be threatened (see State Forests NSW 1995). This paper details the findings of these surveys. In particular, it provides an assessment of the current status in the Dorrigo area of species listed under the Threatened Species Conservation Act (which replaced the TSIP Act in 1995) and provides new information on their basic biology and habitat requirements. Finally, the current conservation measures developed to protect frogs within areas of forest subject to forestry operations are discussed.
SURVEY METHODS Surveys for amphibians were performed over three separate time periods: a. 26th of September to the 19th of December 1993 — when weather conditions were mostly warm and dry with occasional light showers; b. 11th to the 17th of February 1994 — when moderate to heavy rains fell before and during the survey period; and c. 21st to the 23rd of June 1994 — a period of dry and cool conditions. Survey techniques used followed those described in York et al. (1991) and included both surveys of water bodies and the traversing of roads to locate migrating frogs. The aim was to search as many water bodies as possible (with an emphasis on forest areas) to provide as broad and intensive a coverage as possible in the time available. Water bodies were located both from maps and by travelling along roads to find breeding sites visually or by listening for choruses of frogs. The selection of sites was not systematic in terms of stratifying the study or selecting equal effort for the different types of water bodies available, and so a measure of the relative abundance of each species can only be made with some caution. It was of greater importance to determine the distribution of species within the DMA and locate sites at which threatened species were present so that protective prescriptions could be applied for their conservation. Notably, this approach almost certainly reduced the opportunities to locate terrestrial breeding species as their breeding sites were less obvious as to their location (a stream is obvious whereas a small soakage is not).
73
Each site was surveyed at least once in each of the first two survey periods with a subset of sites being searched in the winter survey. Site surveys started with a three to five minutes listening period to identify the species calling and their approximate numbers. This was followed by a spotlight inspection of the water body and its surrounds to identify non-calling individuals which encompassed the entire bank area of ponds (both temporary and permanent) or an average of 50 metres of both sides of streams. Longer lengths of streams were not generally traversed in order to increase the number of sites which could be covered on each night (the aim was to cover as many sites as possible). Additionally, call playbacks were performed to elicit responses from noncalling males and so further increase the chances of locating all of the species present at each breeding site. Road searches were also performed after dusk during periods of rainfall to identify frogs crossing roads. Searches eventually totalled a minimum of 35 hours at pools, 45 hours at streams and 41 hours on road transects during this study. On each occasion frogs were located, the position was recorded to within the nearest 50 metres using speedometer readings. The start and end time of each search was recorded along with the air temperature, humidity, cloud cover, rainfall and wind strength.
RESULTS The Dorrigo Management Area proved to have a diverse frog fauna with 29 species located during these surveys. From the broad distributional maps of Cogger (1992) it was considered that 40 species may potentially occur within the DMA (see Table 1) and so 72.5% of those species potentially present were located. However eight species (marked in Table 1) are found in habitats not represented in the DMA or are apparently extinct in this part of their range. Thus, 29 of 32 of the species likely to be present (95%) were located. Notably, no “unexpected” species were located indicating that the predicted species list is relatively accurate. Records of frogs consisted of 174 riparian and 8 terrestrial breeding choruses and 176 incidental records of frogs on roads or in the forest (Figure 1). Thirty three records were obtained in rainforest, 90 in wet sclerophyll forest types, 136 in dry sclerophyll forest, 17 in eucalypt and native pine plantations and 82 within areas of cleared land. A species acquisition graph through time (Figure 2) demonstrates that half of the species located were recorded within two days of the commencement of surveys and that the majority of species (20) were collected within six days. Additional species encountered after this time were the result of heavy rainfall which stimulated some species to call, a change of survey season (for Pseudophryne bibronii — a late summer/autumn breeder) or specific searches for the otherwise difficult to locate species Philoria sphagnicolus. The number of different sites from which a species was recorded during the frog surveys was highly variable, ranging from 101 for Litoria lesueurii and 79 for Mixophyes fasciolatus to just one for Litoria caerulea, Litoria jervisiensis, Litoria tyleri and Pseudophryne bibronii. This is a misleading representation however, in that the majority of records for the two common species were of single individuals on roads. On the other 74
hand Pseudophryne coriacea (70 sites) and Crinia signifera (66 sites) were never recorded on roads and all site records represent calling individuals, often more than ten at a site, making them the most abundant and widespread species located during the surveys. The range of broad habitats from which each species were recorded is also indicated in Table 1. Most species are notable for utilising more than one habitat type, with a few species appearing to favour either dry or moist forest types. Eight of the species recorded on more than two occasions had 85% or more of their records fall within areas substantially covered by native forest (including plantations) and so are considered forest dependant. This includes four species of conservation significance (Mixophyes iteratus, M. balbus, Assa darlingtoni, Philoria sphagnicolus).
Endangered Species The surveys located a total of five species listed as threatened under the TSC Act with a greater concentration of these frogs being found on the eastern side of the DMA (Figure 3). Each of these frogs had previously been recorded in the area, but the number of records was increased for each species except the sphagnum frog. Little information has been published on the habits of these species and so general accounts of the records obtained are provided as follows: Hip-pocket Frog/Pouched Frog Assa darlingtoni The six populations located are at the southern end of this species’ range. All populations were within areas of wet sclerophyll forest or rainforest. Chorus sizes ranged from more than 100 males to approximately 20 males with calling being heard from September to November, although calling has been heard at other times of the year subsequent to this (Lemckert pers. obs.). Calling could be heard both during the day and at night, even during dry conditions. Calling sites were under the leaf litter or, on rocky scree slopes, from within spaces underneath or between rocks. The small size (2 metres wide), permanently flowing creek. Individuals of all sizes were easy to locate visually due to their large reflective eyes and habit of sitting above the leaf litter (as compared to M. balbus). Males were only heard calling sporadically. All populations were found in wet sclerophyll forest or rainforest. Seven of these records came from timber plantations indicating that this species is able to utilise disturbed habitats to at least some degree. The DMA and the forests to the east appear to now be the stronghold of this species as populations are present throughout the drainage systems of this area, but have disappeared or severely declined to the north and south (Mahony 1993). The reasons for this decline remain unclear as are the reasons for the continued abundance of this frog in the DMA. Sphagnum Frog
Philoria sphagnicolus
sclerophyll forest. It was evident during this study that generalised survey techniques are inappropriate in locating this species as they are often not associated with readily identified riparian areas. For this species, very specific microhabitats needed to be targeted which could only be reached by extensive foot travel in the DMA and so the status of this species could not be readily determined from this study.
DISCUSSION The Dorrigo Management Area, with 29 species, proved to be an area of relatively high frog diversity compared to other areas surveyed for forestry EISs. Surveys in the areas around Glen Innes (24 species — Smith et al. 1992) and Murwillumbah (23 species — Schodde 1996) recorded the next greatest numbers of species with between 2 and 22 species found in 13 other EISs (Lemckert 1996). This might indicate the DMA has a higher diversity of species than other nearby areas, but general distribution maps (Cogger 1992; Robinson 1993) would not suggest this. The majority of species found in the DMA should have been present in these other areas. Rather, the difference is more a reflection of the increased effort and flexibility directed at surveying frogs in this study. The survey periods in the other EIS studies were all constrained to a relatively narrow window of time (often just one month; Lemckert 1996). In the DMA, surveys were performed over a nine month period allowing various seasons to be covered and more opportunity to survey under optimal conditions. The DMA was also surveyed far more intensively, with the effort between two and five times greater that applied to other MAs. This emphasises the fact that any survey attempting to understand the status and 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 Litoria brevipalmata Mixophyes balbus Mixophyes fleayi Mixophyes iteratus Philoria kundagungan Philoria loveridgei Philoria sphagnicolus Assa darlingtoni *
5 hectare exclusion zone around known records 5 hectare exclusion zone around known records 30m wide buffer 200m upstream and downstream of known sites 40m wide buffer 200m upstream and downstream of known sites 30m wide buffer 200m upstream and downstream of known sites 50m radius exclusion zone around site 50m radius exclusion zone around site 50m radius exclusion zone around site Minimum 50m radius exclusion zone around southern populations
* Prescription applied voluntarily by SFNSW
The distributions and number of record sites for species provided some unexpected results. Species such as Litoria tyleri and L. caerulea and to a lesser extent Limnodynastes tasmaniensis and L. dumerilii were expected to be widespread and common, but were not. The surveys were conducted during both wet and dry conditions which should have covered all species preferred breeding conditions and the reasons for the relative rarity of these species are unclear. One possibility is that these frogs were at some form of cyclical low point in their abundance or detectability. Such variations in amphibian abundances have been documented overseas (e.g. Meyer et al. 1998). Such a cycle has obvious implications for environmental management and a cautious interpretation of survey results is required when determining the status of a species. Whatever is the case, this result does point out that surveys need to be intensive and cover several seasons in order to have confidence in locating even some apparently common species. Such time is rarely ever allowed for these sorts of studies. The habitats utilised by the species of conservation significance were generally the same as the general habitat requirements listed in standard reference texts, however there are a few notable differences. Barker et al. (1995) records that Mixophyes balbus is found in rainforest whilst Cogger (1992) generalises this to moist forests. In this study this species was also found within riparian areas in dry sclerophyll forest. Only Robinson (1993) provides any assessment of the breeding season noting that for all southern Mixophyes breeding appears to occur in late spring and early summer. In this study calling was heard from early spring to late summer for all three species of Mixophyes present in the DMA. Assa darlingtoni was recorded in both rainforest and wet sclerophyll forest although both Cogger (1992) and Robinson (1993) note it to be restricted to rainforest and antarctic beech forest. Most noticeably for Assa, all three texts indicate that the range of this frog extends approximately 50km into NSW, whereas Dorrigo is approximately 200km south of the border. Litoria subglandulosa is noted by both Barker et al. (1995) and Robinson (1993) as breeding from October to November. The calling season of this species did not appear to extend beyond mid-December, but it was heard calling strongly in September at all of the sites in the DMA and should be considered an early spring breeder.
The failure to locate Litoria brevipalmata was not a surprise as this species is known to breed over only a few days of each year coinciding with heavy rainfalls (Barker et al. 1995). Even regular checks of known breeding sites rarely result in this species being found outside of breeding events (Lemckert unpubl. data). The absence of Litoria aurea confirms the findings of Mahony (1996). Although this frog has been recorded immediately to the east of the DMA, Mahony (1996) found that Litoria aurea has disappeared from all upland areas within its historically known range.
Conservation of frogs in Timber Production Areas Historically, the accepted method of protecting frogs in forestry areas was the retention of “filter strips” of undisturbed vegetation along the banks of streams to conserve breeding habitats. Recently however, the SFNSW and the NSW National Parks and Wildlife Service (NPWS) have developed a set of agreed Conservation Protocols which provide multiple and broad ranging mechanisms to protect frogs within forests subject to logging operations (SFNSW/NPWS 1997 and see Table 2). A core part of these protocols remains the retention of (most usually) a 20m wide strip of undisturbed vegetation along the banks of streams (=40m strip in total) to protect breeding habitat and maintain water quality. This has however been extended to include both dams and soakage areas which protects nonstream breeding species. Additional protection has been granted to threatened species through the increase of the protective strips along streams from 20m to 30 or 40m for 200m upstream and downstream of sites (see Table 2). Water bodies utilised by Litoria aurea and L. brevipalmata are protected by the retention of 5 hectares of undisturbed vegetation centred around the site whilst known sites for species of Philoria are provided with a 50m buffer zone (Table 2). Specific prescriptions for the protection of Assa darlingtoni were not set under the Conservation Protocols, but SFNSW itself has applied a prescription whereby all areas of forest inhabited by this species and a 50m buffer zone are reserved from forestry operations. The need for dispersal routes has been considered in the Protocols by the provision of both the retained strips of stream-side vegetation and by the addition of cross 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 along streams and across ridges, although the effectiveness of such corridors as dispersal routes remains to be confirmed. The one aspect of the ecology of frogs that has been most difficult to address has been their foraging needs. It is generally perceived that logging results in the clear-felling of forest areas to leave a barren landscape. This may be true in other countries or even some areas of Australia, however in northern NSW logging has generally been of a selective nature. The result is that trees are removed in a patchier manner, with the area logged being dependent on what percentage of an area had commercial value and the amount of riparian area that required protection. So even up to recently, an average of 30% of any compartment remained unlogged (Forestry Commission NSW 1980) and so significant areas of undisturbed vegetation were available for foraging if a frog required such habitat. The Conservation Protocols have provided additional protection and restrictions such that now no greater than 50% of the available logging area can be harvested, further increasing the area of undisturbed forest available for use. The newly increased protection available for frogs in general and threatened species in particular has been provided to increase the likelihood that this group of vertebrates will persist in logging areas into the foreseeable future. However, the effectiveness of the Conservation Protocols remains unproved. Monitoring of populations is to be undertaken over the long-term, but if there are significant impacts over the long term declines may be detected only after they are difficult to reverse. There is a need for immediate, explicit studies to examine the effectiveness of these prescriptions. Such studies need to concentrate on each stage of the frog life-cycle to determine how and when impacts occur from forestry and so what changes can be made to management practices to reduce the impacts of future operations. This program should most specifically target the species of conservation significance which appear to be the species most vulnerable to disturbance and/or have the most specific habitat requirements. Effective prescriptions in combination with the development of a comprehensive and adequate reserve system are critical to ensure the continued survival of frog populations in the Dorrigo Management Area and in other timber production forests.
ACKNOWLEDGEMENTS Firstly, our thanks go to Garry Daly, Mark Fitzgerald, Liz Kelso, Allan Manning, Stefanie Pidcock and Jacqui Richards for their help during the surveys. Thanks also go to the staff of the Dorrigo Office of SFNSW, and particularly Paul Roberts, in providing help in various aspects of the surveys. Finally, we would like to thank Jacqui Recsei, Karen Thumm, Dr. Michael Mahony and an anonymous reviewer for their critical reviews of this document.
REFERENCES Andrews, S. P., Gration, G. L., Quin, D. G. and Smith, A. P., (1994) 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
Barker, J., Grigg, G. C. and Tyler, M. J., (1995) A Field Guide to Australian Frogs. Surry Beatty and Sons, Chipping Norton, NSW. Cogger, H. G., (1992) Reptiles and Amphibians of Australia (Revised Edition). Reed. Sydney, NSW. FCNSW., (1980) Forestry Operations in the Dorrigo Management Area. Forestry Commission of New South Wales. Sydney, NSW. Lemckert, F. L., (1996) Surveys for the Green and Golden Bell Frog (Litoria aurea) by State Forests of NSW. Aust. Zool. 30: 208-213. Lunney, D., Curtin, A., Ayers, D., Cogger, H. G. and Dickman, C. R., (1996) An ecological approach to identifying the endangered fauna of New South Wales. Pac. Cons. Biol. 2: 212-231. Mahony, M.J., (1993) The status of frogs in the Watagan Mountains area, Central Coast of New South Wales. Pp 257-264 in Herpetology in Australia: A Diverse Discipline. ed by D.Lunney and D.Ayers. Surrey Beatty and Sons with the Royal Zoological Society of NSW, Sydney. Mahony, M.J., (1996) The decline of the Green and Golden Bell Frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust Zool, 30: 237-247. Meyer, A. H., Schmidt, B. R. and Grossenbacher, K., (1998) Analysis of three amphibian populations with quarter century long time-series. Proc. R. Soc. Lond. 265: 523-528. Pechmann, H. K. and Wilbur, H. M., (1994) Putting declining amphibian populations in perspective: natural fluctuations and human impacts. Herpetologica 50: 65-84. Richards, S. J., McDonald, K. R. and Alford, R. A., (1993) Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conservation Biology 1, 66-77. Robinson, M., (1993) A Field Guide to Frogs of Australia: from Port Augusta to Fraser Island including Tasmania. Reed. Sydney, NSW. Schodde, R., (1996) Murwillumbah management area fauna survey. Report to State Forests of New South Wales. CSIRO. Canberra, ACT. SFNSW., (1995) Environmental Impact Statement for the Dorrigo management area. Unpublished report by State Forests of NSW. Sydney, NSW. SFNSW/NPWS (1997) Threatened species protocol: Survey design and potential habitat. Joint unpublished report by State Forests of New South Wales and the New South Wales National Parks and Wildlife Service. Sydney, NSW. Smith, A.P., Moore, D.M. and Andrews, S.P., (1992) Proposed forestry operations in the Glen Innes Management Area — Fauna Impact Statement. Report to State Forests of New South Wales. Austeco Pty Ltd, Armidale, NSW. York, A., Binns, D. and Shields, J., (1991) Flora and Fauna Assessment in NSW State Forests. Survey Guidelines. Procedures for Sampling Flora and Fauna for Environmental Impact Statements.Version 1.1. Report for the Forestry Commission of NSW Sydney, NSW.
Review of the declines and disappearances within the bell frog species group (Litoria aurea species group) in Australia Michael Mahoney *
ABSTRACT Declines and disappearances of species and populations of three species of bell-frogs have been reported, Litoria castanea has not been observed in the wild since the mid 1970’s, and L. aurea and L. raniformis have declined from parts of their former distribution. All populations of these three species formerly found on the Great Dividing Range in New South Wales have disappeared.
Victoria nor have there been reports of widespread declines in the other three species (L. cyclorhynchus, and L. moorei, in southern Western Australia, and L. dahlii in the north of the continent).The cause of declines remains unknown; however, the introduced fish Gambusia holbrooki may be responsible for, or play a role, in the declines in some regions.The bell frog group has a number of features that make it a suitable model to investigate the wider issue of 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 The purpose of this paper is to review the decline and disappearance of members of the bell frog species group (Litoria aurea species group,Tyler and Davies 1978; Maxson et al. 1982; King 1993). For each species there will be a brief review of the historical and present distribution, habitat requirements, biology, and the pattern of declines and disappearances where they have been documented, followed by a general discussion on the possible causes of declines and the research that has been conducted.The paper aims to bring together the current state of understanding of the declines and disappearances within this group so that they may be compared with those in other amphibians. I have relied on published accounts and various technical reports (fauna surveys, environmental impact statements, management plans, fauna impact statements and species impact statements). Consequently, the review does not present the most recent findings from a number of research projects that have been in progress for some years and are ongoing.
SYSTEMATICS AND ECOLOGY Bell frogs as they are commonly known, are a distinctive group within the Australo-Papuan hylid genus Litoria.The species group comprise six species L. aurea, L. castanea, L. cyclorhynchus, L. dahlii, L. moorei, and L. raniformis (Tyler and Davies 1978; Barker and Grigg 1977; Barker et al. 1995). All members of the group have rich green dorsal surfaces FIGURE 1: Distribution of members of the Litoria aurea species group.
82
with mottling of gold or bronze, all are relatively large, adult males reach to about 90 mm and adult females 100 mm snout vent lengths. All occur in mesic environments and they are usually associated with water, their general distributions are presented in Figure 1.The species group has one representative in northern Australia, two representatives in the south-west of Western Australia, and three in southeastern Australia. No species penetrates the arid zone. As might be expected for members of a species group they also share similar habitat preferences and behaviour. Adults are usually associated with water, and they are active by day basking in the sun. Individuals can usually be observed on emergent vegetation above water or on the edge of water bodies.When disturbed they drop into the water and remain submerged for a short period. Breeding generally occurs in still water bodies ranging in size from ephemeral pools to large lakes, coastal floodplains and billabongs (Oxbow lakes) (Copland 1957; Main 1965; Humphries 1979;Tyler and Davies 1978; Barker et al. 1995). None of the species breeds in fast flowing streams, but they may use still pools in streams. A full account of the variation in breeding sites of Litoria aurea is provided by Pyke and White (1996), and details of the breeding habitat of the other species can be obtained from various sources (see Barker et al. 1995 for a general description). These frogs are largely terrestrial, although termed “tree-frogs”; they rarely climb trees, and have only small adhesive pads on their fingers and toes.While they are generally closely associated with water when observed by day, by night adults and juveniles
may move widely in terrestrial situations when foraging particularly during rainy periods. Individuals have often been found considerable distances from the nearest water body (White 1995a), but evidence from a limited mark-recapture study of L. aurea indicated that some animals remain at or return to a breeding site when adult (Murphy 1996). However, the importance of movement (immigration, emigration and distances travelled when foraging) for different ages and sexes relative to edaphic features remains a matter of some debate. Seasonal activity varies among the species depending largely on the regional climate. In coastal areas of New South Wales L. aurea is active at all times except in the midst of winter. At Homebush Bay in Sydney and at Kooragang Island in the Hunter Valley individuals overwinter in crevices in scree slopes and piles of rubble (Greer 1994, Hamer unpubl.). In the north L. dahlii may be observed at all times of the year basking on vegetation around wetlands, or in the dry months it hides in cracks in clay areas (Tyler and Davies 1986).The tableland species L. castanea had a much shorter period of activity from October to April, and adults required suitable sites for winter torpor (Humphries 1979; Courtice 1972; Courtice and Grigg 1975). Many of the sites where the species was found experience frost during the winter months, but reports of brumation sites are few. High altitude areas of the northern and southern tablelands have a frost period that exceeds a median duration of 150 days per year (Hobbs and Jackson 1977). Courtice and Grigg (1975) report that individuals of the New England tableland population were found buried in soil beneath a tree stump, and under tin, and Humphries (1979) reports that on the southern tablelands groups were found under reeds at the edge of the breeding site. In the south west of Western Australia L. cyclorhynchus and L. moorei breed in spring and summer, and the species occur near swamps and permanent water (Main 1965;Tyler et al. 1984). I have found L. moorei in late autumn in moist forest habitat sheltering under sheet metal. Litoria raniformis breeds in spring and summer and is active in the warmer months. Near Murray Bridge and Mannum on the Murray River in South Australia animals have been found during winter in groups beneath thick beds of reeds on the edge of wetlands that occur adjacent to the river (Mahony unpubl.).
THE GREEN AND GOLDEN BELL-FROG Litoria aurea The biology and conservation of the green and golden bellfrog was the subject of a workshop held in Sydney at the Australian Museum and Taronga Zoo in 1994 and contributions were published in a subsequent edition of the Australian Zoologist Volume 30(2) May 1996 edited by Pyke and Osborne.This publication has 17 papers that deal with various aspects of the biology of the frog including reproductive biology, habitat requirements, former and present distribution and abundance, cause of decline and conservation management.
Distribution The historical distribution of the green and golden bell frog has been reasonably well documented (Courtice and Grigg 1975). White and Pyke (1996) compared the historical and current distribution prior to 1990 with the known distribution in 1996, and details of the former distribution and disappearance of the southern tableland population were presented and reviewed by Osborne et al. (1996).There has been a contraction in
range and apparent reduction in abundance, although the latter has been more difficult to verify.The species can no longer be found on the southern and central tablelands, or on the western slopes (see Figure 2). It has contracted from its former northern extent along the New South Wales coast and extant populations are apparently small and widely spaced (Clancy 1996). No populations have been detected on the eastern hinterland of the Great Dividing Range in the past ten years (Lemckert 1996). Formerly it ranged in altitude from sea level to about 700 m, but currently there are no populations known above 150 m (Ravensworth in the Hunter Valley)(Interagency Advisory Group 1994; Wellington and Wells 1994; Resource Planning 1994; Mahony 1997). Most extant populations occur close to the coast. Since 1990 only 38 localities have been recorded and 19 of these are in the Sydney Basin biogeographic region. Many of the populations are small, estimates range from as few as 5 to 15 adults at some isolated sites (Cogger 1993; Fanning and White 1994;White 1993, 1997;White and Pyke 1996).The largest known populations are at Homebush Bay, the site of the year 2000 Olympics development, and on the Kurnell Peninsula to the southeast of Botany Bay. At Homebush the species occurs and breeds in semi-permanent pools in the uneven base of a large disused brick quarry (about 15 hectares and 20 m deep with almost vertical sides (Greer 1994; Pyke 1995). Populations occur in the land surrounding the quarry and have been recorded breeding at several human-made permanent and ephemeral water bodies (Pyke 1995; Pyke and White 1996). Because of the large developments proposed for the Homebush site, fauna surveys were conducted and fauna impact statements prepared, including estimates of population sizes, habitat use and an assessment of potential threatening processes (Greer 1994; Pyke 1995).With the aim of retaining an evolutionarily viable population in the area inside and outside of the quarry, frog populations have subsequently been the subject of intensive monitoring, demographic studies, and investigation of microhabitat and habitat use by a postgraduate student Michelle Christie from the University of Sydney, Drs Graham Pyke and Arthur White from the Australian Museum Sydney and personnel from the Australian Museum Business Services. Wetlands on the Kurnell Peninsula provide significant habitat for the frog.The peninsula is primarily a series of pleistocene coastal sand dunes with formerly extensive holocene freshwater wetlands (> 10 000 years of age)(Roy and Crawford, 1978, 1979). Natural habitats in this area have been greatly disturbed since first European settlement with at various times, cattle grazing, removal of vegetation, sand mining and industrial and urban development. Large tracts of the peninsula are currently covered by transgressive sand dunes destabilised by earlier activities (Pickard 1972; Skinner 1973; Roy and Crawford 1979). Adduction sand mining has resulted in mirror lakes, which are used by the frog for breeding. Some natural wetlands occur in swales between dunes on the northern end of the peninsula and in the remnants of the Holocene freshwater wetland on the mid northern end of the peninsula (Mahony 1998).The continued persistence of this population is threatened by an array of developments that are currently in progress or proposed for the peninsula, and there are few chances that interconnected natural habitat will be set aside (Gunninah 1997, 1998; Mahony 1997, 1998). 83
The remaining populations in the Sydney basin are all apparently small with estimates of less than 20 adults per site (White and Pyke 1996). Outside this area larger populations are known to the south on the Illawarra coastline and to the north in the Hunter Valley, however measures of the sizes of these populations are based on counts taken at various times and not based on values for which estimation errors are available. Several large populations are known in coastal swamp habitat south of Kiama and Nowra respectively (Daly 1995, 1996; Morgan and Buttemer 1996; Goldingay 1996; Murphy 1996, NSW NPWS 1996a; Mahony 1997). In the Hunter Valley the largest population occurs on Kooragang Island near the mouth of the river. Much of the island is developed for industrial purposes but several large natural and human constructed wetlands occur (Markwell 1984; NSW NPWS 1996b). Disjunct populations also occur on pastoral land 38, 98 and 124 kilometres up the valley. These populations are currently the most inland known in the former distribution of the species. Populations on the north coast are widely spaced and the number in each population is apparently small (Clancy 1996; Pyke and White 1996). In Victoria this species is restricted to east Gippsland (Brook 1979a) where it occurs predominantly at low altitudes along the coast and hinterland. Gillespie (1996) compared its current distribution with historical records, and found little evidence of a decline in the distribution or abundance of the species in this region. He also observed that the landscape of East Gippsland was relatively intact and possibly provides a greater continuity of habitat with a higher density of breeding or refuge areas than in regions where the species had declined further north in New South Wales. He reported that it was usually associated with stationary, mostly permanent water bodies in both forested and cleared habitats. He also noted that apart from isolated occurrence in farm dams, Gambusia was absent from the region east of the Snowy River in Victoria. Populations occur on two offshore islands in New South Wales, Bowen Island in Jervis Bay (Osborne and McElhinney 1996) and Broughton Island north of Port Stephens (NSW NPWS Atlas 1998). It remains unknown whether these populations are relictual or the result of assisted translocation. Bowen Island is about 300 m and Broughton Island about 2 km respectively from the mainland.The species has been successfully introduced to New Zealand (Robb 1980; Ford 1986,1989; Pickard and Towns 1988; Bell 1982), and New Caledonia (Cree 1984) and Vanuatu (Tyler 1979).
Habitat Requirements The natural habitat requirements of the species have proved difficult to define because it has been associated with almost every type of water body except fast flowing streams (Pyke and White 1996). Many of the sites where the species was formerly known have been disturbed, and the species still occurs at many sites that have experienced long periods of disturbance.This has led some to suggest that the frog is opportunistic or occupies a successional role (Pyke and White 1996). Few historical records have come from within forest in New South Wales (Pyke and White 1996). Lemckert (1996) detailed the results of numerous field surveys in forested and adjacent habitats in New South Wales and concluded that the species was rarely found within forested 84
lands. At the same time he noted that the frog was not detected in adjacent agricultural lands where it would have been expected.This is at odds with the report of Gillespie (1996) who indicates that the species uses forest habitats in East Gippsland. However, Pyke and White (pers. comm.) consider that the species is essentially a coastal and nearcoastal riparian wetland species in this region and is only found in forest sites after moving from these breeding sites. They also note that the species does not use forest habitats in New Zealand. A question remains whether the absence of the frog in forest habitats in New South Wales is a reflection of its decline rather than its preference for other habitats. Habitat features at 74 sites, 18 where the species was extant with the breeding site identified, and 56 sites where the species no longer occurs and the breeding site was not identified, were tabulated by Pyke and White (1996).The result was a description of the habitat at the extant sites compared to historic sites, and therefore does not necessarily identify the requirements of the species prior to declines. They found that for a site to support a breeding population it should contain water bodies which are still, shallow, ephemeral, unpolluted, unshaded, with aquatic plants and free of Gambusia and other predatory fish, with terrestrial habitats that consist of grassy areas and vegetation no higher than woodlands, and a range of diurnal shelter sites. Breeding occurred in a significantly higher proportion of sites with ephemeral ponds rather than sites with fluctuating or permanent ponds, and where predatory fish were absent. According to their analysis both the level of water fluctuation and presence/absence of Gambusia have significant effects on whether or not the frog breeds at a site.The coupling of these two features may provide insight into the habitats that continue to support this species. Fish are purged from ephemeral ponds during periods of extended dry and these ponds are often isolated, whereas once established in permanent ponds or along drainage lines they remain and are able to recolonise.The possible predatory role of Gambusia holbrooki in the decline and continued pressure placed on populations of bell frogs will be discussed later. At this stage it is worthy of note that use of ephemeral breeding sites was not a feature associated with members of the bell frogs group in earlier habitat descriptions (Courtice and Grigg 1975; Dankers 1977; Humphries 1979; Cogger 1992).
Pattern of Decline The precise time at which populations began to decline and disappear is difficult to assess. White and Pyke (1996) provide evidence that declines have occurred since 1990, but the pattern before that date is unclear. On the southern tablelands populations disappeared some time after 1977 (Osborne et al. 1996). Assessment of the current distribution reveals that all remaining populations are at low altitude, the great majority are within a few kilometres of either the coast or an estuary, and those few populations that are inland, such as those at Ravensworth and Bayswater in the Hunter Valley are in close proximity to a large lake.The significance of these distribution patterns relative to the broader phenomenon of declining amphibians remains unknown, but they provide numerous avenues for further research into the causal agents.
Future Prospects Because of its occurrence in the large urban area of Sydney and the fact that one of the largest remaining populations in the Sydney Basin occurs on the year 2000 Olympic development site, this species has received considerable media attention.The plight of this species has therefore drawn public attention and awareness to the conservation needs of frogs, and while it may not be as critically endangered as other members of the species group, it acts as a flagship species for all frogs.There are several ongoing and long-term studies on this species that focus on its ecology, causes of decline, and its potential for relocation and reintroduction. Captive husbandry has been reported by several groups (Robinson 1993a) and two “A Class” zoos (Taronga Zoo, Mosman Sydney NSW; Australian Reptile Park, Somersby, NSW) have breeding programs dealing with specific populations. Hence we can expect that conservation management of this species should continue to be enhanced and based on detailed field ecology.
THE SOUTHERN AND NORTHERN TABLELANDS BELL-FROG Litoria castanea Some uncertainty surrounds the taxonomic status of the various populations of what is referred to here as Litoria castanea.Thomson et al. (1996) have provided a detailed 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.
NORTHERN TABLELANDS POPULATION OF Litoria castanea Distribution The northern tableland population (L. flavipunctata of Courtice and Grigg 1975) was known from a relatively restricted distribution centred around the town of Guyra on the New England Tableland (see Figures 1 and 2).This area is on the higher section of the tableland and most historic sites were above 1 000 m (ADH).There are 13 sites in this region all of which have been verified by examination of museum specimens or photographs (Mahony unpubl.).This population was allopatric from the southern tableland population separated by a distance of about 500 km, it was not sympatric with any other member of the group. There are no verified records of this population after 1975 (Ehmann and White, 1997), and the last specimen to be placed in a museum was collected in 1973 (Australian Museum register).The earliest record of this population on the northern tablelands, is a specimen from “Hillgrove” about 25 km south-east of Armidale (see Figure 2) which until recently was misidentified as a specimen of L. aurea. Details of the precise date of collection and the collector are unknown but this specimen was registered with the Australian Museum prior to 1890.
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 General features of the habitat and breeding biology were provided by Courtice and Grigg (1975).
Pattern of Decline In the early 1970’s individuals of this population were apparently common in farm dams, near large tableland lakes and in still pools of streams. Professor Grigg (Department of Zoology, University of Queensland) has a photo which shows eight individuals sitting on reeds basking in the manner typical of bell-frogs.The Australian Museum has a series of 42 individuals collected in the vicinity of Little Llangothlin Lagoon on 13/8/1971.There are also a number of anecdotal reports, which indicate the species was once common. Apparently staff from the Zoology Department at the University of New England were able to collect reasonable numbers to use for teaching purposes, and the landholder of a property (Millievale) south west of Guyra recounted that students conducted field studies on the property and one feature was the bell-frogs around the farm house dam. Photographs in life show that members of this population were large and vividly marked and it may be expected that its disappearance would have been apparent. However, the extent of the demise of this population did not become apparent until the mid 1980’s when several herpetologists noted that it had not been observed since the mid 1970’s (Mahony 1996).There had been no systematic surveys of historic sites and other suitable habitat in the region in the period between 1973 and the early 1990’s.The first detailed systematic surveys were commenced in the summer of 1992 (Mahony and Knowles 1993), and surveys continued for another three breeding seasons, and the region covered was expanded. Unfortunately no population of the frog was found.The New South Wales Frog and Tadpole Study Group also conducted a survey for this population in the summer of 1993-1994 and 1994-1995 which focused on historical sites and potential habitat in the region, however no specimens were observed (Ehmann and White 1997). To a great extent the knowledge that we have of this species is based on the collections made by students, staff and visiting researchers to the Department of Zoology at the University of New England. In the late 1960’s and early 1970’s there was a group of herpetologist working in the department with an active field program in which series of animals were collected and placed into the department museum (Heatwole et al. 1995).The collection and register is now held by the Australian Museum Sydney. Not only do we know of the existence of this population and some details of its distribution because of their work, in addition the register provides invaluable information when trying to understand its pattern of decline and disappearance.Thus unlike many other regions in New South Wales there is a reasonable base-line upon which to base comparisons of distribution, abundance and former community composition. During the same period of active collection in which the bell frogs were encountered, large series of the stream hylid Litoria boorolongensis were registered.This species has now also disappeared from the tablelands. Similarly, there are numerous occasions in the register’s field notes where the various species making up a frog community were recorded. 86
Bibron’s toadlet (Pseudophryne bibronii) was encountered across the tablelands in many habitats, while recent surveys have failed to detect it at any site in the region except on the eastern fringes.These species were far more common, as far as can be ascertained by the number of records in the field register, and more widespread in distribution than the bell frog. Sometime after 1973, along with the bell frogs, these species disappeared from across large areas of the northern tablelands, to the point that they can now be regarded as regionally extinct. At the same time several other species such as Crinia signifera, C. parinsignifera, Limnodynastes dumerilli, Lim. peroni, Lim. tasmaniensis, Litoria fallax, L. lesueuri, L. peroni and L. verreauxi, that were also recorded in the register, remain widespread and common in the region wherever suitable habitat occurs.The available evidence indicates that the decline of the bell-frog was rapid, and that similar declines affected two other species on the tablelands, however, these species remain in populations at low altitudes. Habitats used by these three species are markedly different; L. castanea breeds in still water situations, L. boorolongensis is closely associated with fast flowing rocky streams, and P. bibronii breeds in terrestrial soaks and swampy situations. Gambusia is widespread on the New England Tablelands however there is an absence of data on the time of introduction and rate of dispersion in the region (Harris 1995).
Future Prospects In the spring and summer of 1996-1997 a public awareness program was conducted by the NSW National Parks and Wildlife Service in the hope that some remnant population may be reported and to obtain historical information about the species. A colour brochure with pictures of the frog, typical habitat and maps of distribution was distributed, and a travelling poster display was mounted at schools, libraries and shire offices in the New England region. No population was reported and no new information on the historic distribution was obtained.The prognosis for this species is not good unless an unknown population is discovered.
SOUTHERN TABLELANDS POPULATION OF L. castanea Distribution Details of the decline and disappearance of the southern tablelands population of L. castanea have been presented and reviewed by Osborne et al. (1996).The species apparently had a restricted distribution between Canberra and Bombala on the southern tablelands at altitudes between 700 and 800 m. It was broadly sympatric with L. aurea in the north of its range and with L. raniformis in the southwest of the region.
Habitat Requirements Details of its biology and ecology are available from a study of the dynamics of a breeding frog community conducted by Humphries (1979) at Oakdale (700 m alt.) near the NSW/ACT border between 1973 and 1977.The study provides some of the best evidence of the relative abundance of the southern population and its habitat requirements. Litoria castanea was sympatric with L. aurea at the study site.The most significant finding was that both L. aurea and L. castanea (dealt with by him under the name
L. raniformis, although he was aware that the taxon he was studying differed from L. raniformis sensu stricto), were residents at the permanent pond where they bred. By comparison with other species in the community he regarded the bell frogs as the only residents, “during the breeding season they are semi-aquatic, with males spending a major fraction of their time in or near water. Few adults emigrated, and if they did the movement was not related to reproduction. Many juveniles left the pond after metamorphosis, but a considerable number remained, and reached sexual maturity within the area bounded by the drift fence”. Of the six ponds in the study area the one most intensely studied had been constructed as a stock water pond in a drainage line, it was roughly circular in shape with a diameter of about 20 m, and had extensive emergent and floating vegetation.
Pattern of Decline
Bell frogs occurred in much lower densities than the other “migratory” species, which used the pond for breeding, presumably because of their larger size (Humphries 1979). The total number of adult L. aurea of both sexes at the study pond was 21, 26 and 27 over successive seasons. Comparable figures for L. castanea cannot be stated precisely because an experimental introduction of 38 males was made into this enclosed pond in November of 1975. Prior to this the number in the first year was 6 adults, and the number of adult females did not rise above 3. Recapture rates within and between seasons were high for both species.
At the time the species disappeared no immediate concern was raised when a population or species was not observed during field work.Variable climatic conditions, including drought, coupled with the cryptic nature of some species were assumed to explain the failure to observe some species. It was often assumed that a species was absent because field work did not coincide with the preferred climatic conditions of the particular species. Often local anthropogenic impacts were also considered responsible for the absence of a species and searches over wider areas were rarely initiated. In Australia there have been few close ecological studies of frogs that have considered temporal patterns of reproduction with seasonal or climatic condition. Hence, when frogs first disappeared no alarm was raised and no wider or systematic surveys were conducted to determine the scale of declines.
At Oakdale the period of breeding was shorter than for populations of bell frogs on the coast, starting a month or two later, but adults were active until mid-April, when low temperatures inhibited activity. After this time communal brumation occurred each year in the permanent pond, “Frogs of both sexes and species were found together, torpid beneath leaf litter at the base of large Carex appressa tussocks. Mild, wet weather occasionally permitted emergence, and some feeding.”(Humphries 1979). Fidelity of the resident adults at the pond was facultative; the soil near two other ponds dried during autumn, and frogs of both species “emigrated from these ponds to overwinter”. It is not indicated to where these animals emigrated, and migration towards these ponds was observed during the breeding season. Newly-metamorphosed juveniles foraged later in the year than adults, and used the sheltered belt of littoral vegetation. At the well vegetated study pond many juveniles reached sexual maturity and were recruited into the breeding population, while many others dispersed, and a few individuals born elsewhere entered the pond. Population age class structure indicated that males and females reached maturity in their third year and lived for at least six breeding seasons. Furthermore, individuals, marked in October 1973, remained at the one breeding pond over six breeding seasons. Measures of fecundity were limited, two clutches of L. castanea had 1 885 and 3 893 eggs.These values are considerably lower than reported for L. aurea at low altitudes and maturity has been reached in the first year in this species under artificial conditions (White 1995a; Hamer and Mahony unpubl. data).
In retrospect it is apparent that the disappearance of the northern and southern tablelands population of L. castanea occurred in a similar time frame, however, the pace at which the decline and disappearance occurred cannot be accurately assessed, because no monitoring was conducted.The available evidence suggests that the decline was rapid, to the extent that biologists were not aware that it was occurring. Osborne et al. (1996) concluded that the declines of members of this group on the southern tablelands was rapid and occurred sometime between 1978 and 1981 and “did not involve a prolonged stage when the frogs were in low numbers”. Surveys, discussions with experienced local herpetologists, and comparison of historical field records confirmed that all three species of bell frogs on the southern tablelands suffered an extensive decline, with no confirmed records since 1980.
THE SOUTHERN BELL-FROG Litoria raniformis Distribution This species was distributed across a large area of south east Australia including Tasmania (Figure 1). In New South Wales and the Australian Capital Territory the range of the species was centred on the Murray and Murrumbidgee River valleys and their tributaries. It occurred throughout the Southern Tablelands.The species was also recorded on the central tablelands as far north as Bathurst (Ehmann and White 1997).The species was widespread across Victoria being only absent from the western desert regions and the eastern alpine regions (Littlejohn 1963,1982; Brook 1979a; Hero et al. 1991). In South Australia the species is known to occur along the lower Murray River Valley, the lower south-east to near Keith, and a small, apparently introduced population, in the Adelaide Hills (Tyler 1978). In Tasmania the species occurred broadly across the north and east of the island, and on the Bass Strait Islands (Brook 1979b).
Habitat Requirements Habitat requirements are broadly similar to L. aurea, although several authors mention that it was associated with “permanent” water bodies, hence there may be some question whether this species differs in habitat requirements 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 considerable ambiguity associated with the word ephemeral it is difficult to compare habitats of bell frog species. It occurred across a greater altitude than all other members of the species group, with populations ranging from sea level to 1 300 m on the Southern Tablelands (Osborne et al. 1996).
Pattern of Declines Declines have occurred in sections of the range of this species. Consistent with the pattern already reported in other members of the species group, all tableland populations have disappeared. In a review of the species distribution and status, Ehmann and White (1997) noted that in New South Wales the species has disappeared from sites in the central and southern highlands.The disappearance of populations from the Southern Tablelands of NSW/ACT has been reviewed by Osborne et al. (1996). It is currently known from the Monaro district near the Victorian border and is widespread throughout the Murray River valley but has disappeared from a number of sites along the Murrumbidgee River. They noted that in Victoria the species “still has a widespread distribution ....although some declines have been reported from central Victoria (Gillespie pers. comm.)”. To my knowledge there are no published accounts which document declines in populations of L. raniformis apart from those in the upland areas on the southern tablelands of NSW (Ehmann and White 1997; Osborne et al. 1996).There are however numerous informal reports of declines of this species in southern Victoria and in Tasmania which require validation by systematic survey and monitoring.
Litoria cyclorhynchus, L. moorei and L. dahlii To my knowledge there are no reports of declines of either of the Western Australian or northern species in this species group. Reynolds (1995) examined the impact of Gambusia on the eggs and larvae of six species of frog, including L. moorei, on the Swan Coastal Plain Western Australia. He found that L. moorei coexisted with Gambusia in nine lakes in this region and he suggested that the fish was responsible for reductions in frog population sizes in some instances.
CAUSES OF DECLINES IN THE BELL FROG GROUP Declines among Australian frogs are recognised as being part of a global pattern of declines among amphibians. A major question is whether the cause of most declines is a single widespread factor, which could explain the global phenomenon, or a range of factors that are involved at regional and local levels? On the global scale a range of factors have been suggested as possible causes, including increased ultraviolet radiation related to ozone layer depletion, pollutants dispersed in the atmosphere, climate change and the action of a virulent pathogen (Blaustein and Wake 1990; Barinaga 1990; Pounds et al. 1997). On a regional scale there have been examples of declines due to pollution (insecticides and herbicide use), reduced water quality (Warner et al. 1993), habitat destruction and isolation (Barinaga 1990), altered water flow regimes (Hero 1991), and 88
the introduction of exotic fauna particularly predators (Bradford 1989; Blaustein and Wake 1990; Bronmark and Edenham 1994; White and Pyke 1996). In this latter category, it has been suggested that the introduced fish Gambusia may have been responsible for the decline of members of the bell frog group in eastern Australia (Mahony 1993; Webb 1994; White and Pyke 1996; Morgan 1995; Morgan and Buttermer 1996; Webb and Joss 1997). Investigations of disappearances among bell frogs have focused primarily on the declines in L. aurea and L. castanea.Two major directions of research have been pursued, the role of increased ultraviolet radiation, and the impact of Gambusia. In several cases the impact of Gambusia has also been linked with investigation of habitat requirements of the species. Attention to these issues has been required because of the need to actively manage populations in certain areas of Sydney.
THE ROLE OF GAMBUSIA Results of several studies are consistent with the hypothesis that Gambusia contributes to the decline of frog populations. The first study to indicate the impacts of this introduced fish on native frog communities was conducted by Dankers (1977) who found that the number of metamorphosing frogs was much higher in the part of the study pond from which the fish was excluded, compared with the remainder in which Gambusia activity was not reduced. Litoria aurea was present in the pond community studied and it has subsequently disappeared. More recently several studies have shown that Gambusia will attack and eat tadpoles including those of the green and golden bell frog (Harris 1995; Morgan and Buttermer 1996; Pyke and White 1996; Webb 1994; Webb and Joss 1997). However, the importance of Gambusia as a predator relative to other factors in causing the decline of bell frogs remains unclear.The role of introduced predators in the decline of native species cannot be underestimated, along with habitat destruction they are particularly responsible for the extinction of island faunas (Clout 1996). Heyer (1973, 1976) demonstrated that predation and larval habitat preferences that evolved during speciation were significant in permitting coexistence. If fish predators were absent during evolution, as may be the case with bell frogs, the frog would not be expected to have defensive adaptations, or it may have adaptations to cope with native fish but not with introduced fish. Studies on the impact of Gambusia have taken two lines, first to show that eggs and tadpoles of bell frogs and other frog species are palatable to the fish and that it is an active predator of tadpoles, and secondly, investigations of the extent of the distribution of the fish, the habitats it occupies, and the relation between its distribution and that of frogs. Gambusia was found to be a voracious predator on the tadpoles of green and golden bell frogs (Morgan and Buttermer 1996; Pyke and White 1996) and a number of other native frogs (Harris 1995; Morgan and Buttermer 1996; Webb and Joss 1997). Studies have dealt with various factors including predator prey ratios, density of predators and the effect of available cover or substrate. In all cases the fish rapidly killed tadpoles.
A criticism that can be leveled at the studies that have examined the impact of Gambusia is that rarely have they measured field densities of the fish and of tadpoles in the absence of fish and constructed experiments that deal with these natural densities. In general the laboratory densities of fish are probably higher than encountered in the field and therefore the results unrealistic. Gambusia are present in all states of Australia except Tasmania (Arthington and Lloyd 1989), however, there is little documented information on their spread in various regions. Harris (1995) examined a large number of waterbodies in the New England region around the town of Armidale and the fish was found to be widespread in streams, dams and lakes, but it did not occur in all waterbodies. Pyke and White (1996) reported the number of historic sites used by the green and the golden bell frog that had populations of fish. There is a great deal that remains to be understood about the impact of Gambusia on individual species and on frog communities.These studies will have important implications for the management of habitats to maintain natural frog community composition and the abundance of frogs.The variation found among water bodies makes extrapolation of ecological conclusions difficult. Future studies need to focus on the relationship between the distribution and number of ephemeral water bodies and distance from nearest permanent water bodies. For species that are particularly susceptible to Gambusia, the occurrence of ephemeral waterbodies surrounding permanent water bodies may provide valuable refuge sites. Gambusia and frogs could coexist during periods of high rainfall due to the presence of ephemeral waterbodies.These waterbodies provide a refuge from the fish. While various studies have revealed that the eggs and tadpoles of bell-frogs are palatable to Gambusia under laboratory conditions, to date there have been no similar experiments under field conditions. It is unknown if in ponds with submerged weeds and cover whether some egg masses and tadpoles escape predation. Could the introduction and dispersal of Gambusia explain the disappearance of the tableland populations of bell-frogs and the decline of the low altitude populations of the green and golden bell frog? Several important pieces of information required to answer this question are absent.The dates of introduction of Gambusia to many regions are not documented, and even if they were, the impact may have been gradual and the apparent rapid disappearance of the frogs the final phase of a long process. Differences in declines between high and low altitude populations may be due to differences in the length of the activity period of frogs which is determined by climatic factors coupled to impacts such as predation. Finally, it should be borne in mind that Gambusia have not been implicated in the disappearance of any of the seven species of rainforest stream frogs in eastern Queensland (see review by McDonald and Alford 1999), and they are absent from numerous former breeding sites on the southern and northern tablelands where various members of this group have disappeared (see Osborne et al. 1996, Mahony 1993).
The relationship between the decline and disappearance of bell frogs and the role that Gambusia has played leads to a conundrum.There are sites where the frog has disappeared but where the fish is absent, and there are sites where the frog can be found and the fish are present. As an example of the first case, I have compared two subcatchments of the Hunter River near Ravensworth, one Bettys Creek (about 30 sq km) is free of any Gambusia, while the adjacent catchment Mine Creek (about 40 sq km) has a large population of the fish. Each catchment has much the same aspect, experiences the same climate, includes a variety of waterbodies including permanent ponds, lotic stream habitats, and ephemeral sites, and both have a long history of clearing followed by pastoral use.The area is currently the site of several large open cut coal mines. Bell frogs have been known in the area for some time, the Museum of Victoria and the Australian Museum have specimens collected at Ravensworth and nearby Lake Liddel in 1971.Two property owners who have lived in the Mine Creek catchment for over twenty years were able to describe the frog and recognised that it was once common in dams on their properties. A small population is extant in one section of the Bettys Creek catchment (Wellington and Wells 1994; Mahony 1995, 1997). No extant population of the frog has been found in the Mine Creek catchment in the past three years despite intensive systematic searches, while a small population has been detected in only two ponds during each summer season in the Bettys Creek catchment. Gambusia cannot be the limiting factor to the distribution and abundance of the frog in the Betty’s creek catchment, because the fish is absent from all water bodies.This argument has numerous caveats because it may be that there are more than one limiting factor to the distribution and abundance of the frog. Examples of the second case, where the fish and frog coexist, are less common. I am aware of at least two sites, which have the latter characteristics, Coomaditchy Lagoon south of Wollongong (van der Mortel and Goldingay 1998), and on Kooragang Island near the mouth of the Hunter River (Hamer and Mahony in prep.). In both cases there are a variety of waterbodies available for reproduction and tadpole development.Van de Mortel and Goldingay (1998) reported that the lagoon contained high densities of the fish however tadpole survival and adult recruitment at this site appeared high. Unfortunately quantitative measures of fish and tadpole densities at sites have not been reported. A study of the comparative recruitment success from different water bodies on Kooragang Island has not been completed. It may be that in such situations the population of the frog is maintained by recruitment from ephemeral pools and that these represent a situation where some sites are sinks and other sources in a metapopulation sense. Finally, while the importance of Gambusia relative to other factors in causing the decline of the bell frogs remains unclear, there seems little doubt that at sites where the frog is being managed the control of Gambusia is something for which strategies can be developed and implemented. Both the decline of the species in places where Gambusia was absent and the 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 RADIATION Experiments to test the hypothesis that increased ultravioletB (280–320 nm) radiation was adversely affecting L. aurea populations were conducted by van de Mortel and Buttermer (1996).They conducted the experiment twice, firstly using three clutches of eggs (late December 1994) and secondly using only one clutch (mid February 1995). In the first experiment they found no significant effect on hatching success at two elevations (sea level and 600 m) between sunlight filtered for UV-B, unfiltered sunlight and a filtered control (p=0.749). In the second experiment hatching success was significantly higher under the UV-B filter than under unfiltered sunlight (p=0.017).The first experiment was conducted under cloudy conditions and the second under clear skies.These experiments were in several ways preliminary. Impacts were investigated on hatching success and did not follow effects on further growth and development. Unfortunately, they were unable to measure field UV-B levels, and they observed that it would be desirable to construct the dose-response relationship of L. aurea eggs to UV-B radiation to assess their results.There is a vital need to further this line of investigation and build on the results already obtained.
DISEASE The occurrence of sick and moribund individuals has not been associated with the declines and disappearances of members of this group. Absence of sick individuals does not rule out the possibility of disease being a factor, it simply may be that because there was no active research on these animals at the time disappearances occurred no sick individuals were encountered. While this was the case in the period 1977 to 1985, it is less so in recent years during which there has been considerable attention given to low altitude populations of the green and golden bell frog. Some sick and dead individuals of L. aurea have been collected and deposited for pathology, but to date no pattern consistent with other cases where a pathogen has been implicated has emerged. We have proposed elsewhere (Mahony, 1995 ) that the bell frog group provides a useful model in which to test hypotheses about the possible role of disease.
CONCLUDING REMARKS It is clear that many issues surrounding the decline and disappearances of bell frogs require in depth investigation. I have postulated elsewhere (Mahony 1996) that in one respect, unexplained rapid disappearances from upland populations, the bell frogs are similar to other species that have disappeared, while on the other hand they have marked differences in reproductive biology and habitat. Several aspects of the biology of the green and golden bell frog make it an ideal candidate to study the cause of disappearance in other Australian frogs. Study of this species, as an exception to the general situation, may increase our understanding of the broader pattern and as a result improve our abilities to conserve all endangered frogs.This frog has several advantages as a research model. Successful captive husbandry is possible, something that has not been achieved for any other Australian species listed as critically endangered or endangered. When coupled with high levels of fecundity, bell frogs lay between 4 000 and 8 000 eggs, this enables large numbers of eggs and tadpoles to be available for experimental studies. Hence, stock has been available to conduct experiments on UV-B tolerance and on the impact of predators. Adults are large enough to be individually marked using passive transponders enabling population demographic studies (Christie 1996), and a number of populations of green and golden bell frog in the Sydney basin have been compared genetically and measures of genetic diversity obtained (Colgan 1996).Two populations are known from offshore islands and the species occurs on several pacific islands where it was introduced.This enables studies on population demographic structure, dispersal and metapopulation structure. Several programs are currently in progress to manage populations of the green and golden bell frog, including habitat restoration and enhancement, reintroduction, and relocation.These provide invaluable opportunities to obtain information on the biology of this frog and should therefore provide insight into the declines and disappearances among other Australian frogs.
ACKNOWLEDGMENTS Special thanks to Steve Donnellan and Arthur White for critically reading the manuscript.
DROUGHT
REFERENCES
Attention has been drawn to the possible role of drought in the disappearance of bell frogs.The period between 1977 and 1982, when the upland species and populations disappeared, represented a period of below average rainfall and drought conditions in the northern and southern tablelands (Osborne et al. 1996; Mahony unpubl. data). However, examination of long-term records indicate that drought is a regular feature of these regions (Osborne et al. 1996). On the northern tablelands several of the large mountain lagoons have contracted, but these sites have rarely completely dried out (Casanova and Brock 1996). Similarly, on the southern tablelands several of the sites where members of the group occurred did not become dry during the drought in the early 1980’s (Osborne et al. 1996).
Arthington, A.H. and Lloyd, L.N., (1989) Introduced Poeciliids in Australia and New Zealand. In G.K. Meffe and F.F. Snelson J.R. (eds) Ecology and Evolution in Livebearing Fishes (Poeciliidae). pp. 333-348. Prentice Hall, New Jersey.
90
Barinaga, M. (1990) Where have all the froggies gone? Science, 247:1033-1034. Barker, J., Grigg,G., and Tyler, M. J. (1977) A Field Guide to Australian Frogs. Rigby, Adelaide. Barker, J., Grigg,G., and Tyler, M. J. (1995) A Field Guide to Australian Frogs. Surrey Beatty and Sons. Chipping Norton. Bell, B. D. (1982) The amphibian fauna of New Zealand. In: New Zealand Herpetology. NZ Wildlife Service Occasional Publication No. 2. (Ed: Newman, DG) N.Z. Wildlife Service, 27-89.
Blaustein, A.R., and Wake, D.B. (1990) Declining amphibian populations: A global phenomenon. Trends Evolutionary Ecology, 4(7): 203-204. Bradford, D. F. (1989) Allopathic distribution of native frogs and introduced fish in high Sierra Nevada lakes of California: Implications of the negative effects of fish introductions. Copeia 1989(3): 775-778. Bronmark, I.C., and Edenham, P., (1994) Does the presence of fish affect the distribution of tree frogs (Hyla arborea)? Conserv. Biol. 8:841-845. Brook, A.J., (1979a) Atlas of Frogs of Victoria. Department of Zoology, University of Melbourne. Publication No. 2. Brook, A.J., (1979b) Atlas of Frogs of Tasmania. Department of Zoology, University of Melbourne. Publication No. 3. Cassanova, M.T., and Brock, M. A. (1996) Wetland flooding regime and plant community establishment, Abstract. Wetlands for the future INTERCOL V Conference Abstract Book: p. 109. Clancy, G.P. (1996) The Green and Golden Bell Frog Litoria aurea in the Station Creek area of Yuraygir National Park. Aust. Zool, 30: 214-217.Christy, M.T. (1996) The efficacy of using Passive Integrated Transponder (PIT) tags without anaesthetic in free-living frogs. Aust. Zool, 30(2):139-142. Clout, M. (1996) Introduced species:The greatest threat to global biodiversity? Species 34-35. Cogger, H. G. (1992) Reptiles and Amphibians of Australia. 4th ed. Reed Books, Sydney. Cogger, H. G. (1993) Faunal Impact Statement. Green and Golden Bell Frog (Litoria aurea) occurring on property at the corner of Dalmeny street and Kimberley Grove, Rosebery. Colgan, D. (1996) Electrophoretic variation in the Green and Golden Bell Frog Litoria aurea. Aust. J. Zool, 30:170-176. Copland, S. J. (1957) Presidential address. Australian tree frogs of the genus Hyla. Proc. Linn. Soc. N.S.W. 82:9-108. Courtice, G. (1972) A taxonomic revision of the Litoria aurea complex (Anura: Hylidae). B.Sc. (Hons) Honours Thesis, University of Sydney, Sydney. 48 p. Courtice, G. P., and Grigg, G. C. (1975) A taxonomic revision of the Litoria aurea complex (Anura: Hylidae) in southeastern Australia. Aust. Zool, 18:149-163. Cree, A. (1984) Breeding biology, respiration and larval development of two introduced frogs (Litoria raniformis and L. ewingi). New Zealand Journal of Zoology 11: 179-188. Cree, A. (1988) Water balance responses of the hylid frog Litoria aurea. J. Exp. Zool, 247:119-125. Daly, G. (1995) Observations on the Green and Golden Bell Frog Litoria aurea (Anuran: Hylidae). Herpetofauna, 25:2-9. Daly, G. (1996) Some problems in the management of the Green and Golden Bell Frog Litoria aurea (Anura: Hylidae) at Coomonderry Swamp on the South Coast of New South Wales. Aust. J. Zool, 30: 233-236. Dankers, N. M. J. A. (1977) The ecology of an anuran community. Ph.D., University of Sydney. Ehmann, H., and White, A. (1997) Frog and Tadpole Study Group of NSW. (1997).Threatened Frog Species of New South Wales. Ed. By H.Ehmann. Frog and Tadpole Study Group of NSW: Sydney.
Environment Australia (1995).Thackway, R. and Cresswell I. D. eds. An interim biogeographic regionalisation for Australia: a framework for establishing the national system of reserves,Version 4.0. Australian Nature Conservation Agency, Canberra. Fanning, F. D., and White, A.W. (1994) The Green and Golden Bell Frog Litoria aurea. Management plan for the Rosebery population. Prepared in accordance with requirements of the Section 120 license for the site. Ford, M. A. (1986) Some observations of reproductive behaviour in captive smooth-backed green frogs (Litoria aurea). Tane 31:113-120. Ford, M. A. (1989):Vocalisations of the green frogs Litoria aurea and Litoria raniformis, including species-specific male calls. N.Z. J. Zool. l. 16:7-23. Gillespie, G.R. (1996) Distribution, habitat and conservation status of the Green and Golden Bell Frog Litoria aurea (Lesson 1829)(Anura: Hylidae) in Victoria. Aust. Zool. 30: 199-207. Goldingay, R.L. (1996) The Green and Golden Bell Frog Litoria aurea — from riches to ruins: conservation of a formerly common species. Aust. J. Zool, 30: 248-256. Greer, A. E. (1994) Faunal Impact Statement for the proposed development works at the Homebush Bay Brick Pit. Prepared for Property Services Group. Gunninah Environmental Consultants. (1997) Assessment of the flora and fauna issues for a proposed development at Cronulla (The North Cronulla integrated resort). Gunninah Environmental Consultants. (1998) Proposed sand overburden removal. Lot 8, DP586986 Captian Cook Drive Kurnell. Fauna and Flora Assessment. Harris, K. (1995) Is there a negative relationship between Gambusia and tadpoles on the Northern Tablelands? BSc (Hons) Department of Ecosystem Management, University of New England, Armidale. Heatwole, H. H., de Bavay, J. M., Webber, P., and Webb, G. (1995) Faunal survey of New England. IV the frogs. Memoirs of the Qld. Mus, 38(1):229-249. Hero, J-M. (1991) A froggy forecast. Wildlife Australia Autumn: 14-15. Hero, J-M., Littlejohn, M., and Marantelli, G. (1991) Frogwatch Field Guide to Victorian Frogs.Victorian Department of Conservation and Environment, Melbourne. Heyer, R. W. (1973) Ecological interactions of frog larvae at a seasonal tropical location in Thailand. J. Herpetol. 7:337-361. Heyer, R. W. (1976) Studies in amphibian habitat partitioning. Smithsonian Contributions to Zoology, 242:1-27. 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). pp.75-93. Armidale, NSW Department of Geography. University of New England. Homebush Bay Corporation (1994) Homebush Bay Development Guidelines.Volume 1. Environmental Strategy. Draft 24/10/94. Humphries, R. B. (1979) Dynamics of a breeding frog community. PhD dissertation, the Australian National University, Canberra.
91
Interagency Advisory Group (1994) Plan of Management. Revegetation & wildlife management in relation to the proposed Mount Owen Mine. Prepared for Interagency Advisory Group in relation to proposed expansion of Mount Owen Coalmine.
Moore, J. A. (1961) Frogs of Eastern New South Wales. Bull. Amer. Mus. Nat. Hist, 121:149-386. Morgan, L. A. (1995) Analysis of the impact of predation by Gambusia holbrooki on Litoria aurea tadpoles. B.Env.Sc.(Hons) Thesis. University of Wollongong: Wollongong.
King, M. (1993) A cytotaxonomic analysis of Australian hylid frogs of the genus Litoria. In: Herpetology in Australia. A Diverse Discipline. (Eds: Lunney, D., Ayers, D.) Royal Zoological Society of NSW, Sydney, 169-175.
Morgan, L. A. and Buttemer, W. A. (1996) Predation by the non-native fish Gambusia holbrooki on small Litoria aurea and L. dentata tadpoles. Aust. J. Zool, 30:143-149.
Lemckert, F. (1996) Surveys for the green and golden bell frog (Litoria aurea) by state forest of New South Wales. Aust. Zool, 30:208-213.
Murphy, M. J. (1996) A capture/recapture study of the endangered hylid frog Litoria aurea. Herpetofauna 25:19-21.
Littlejohn, M.J. (1963) Frogs of the Melbourne area. Vict. Nat, 79:296-304.
Myers, G.S. (1965) Gambusia, the fish destroyer. Aust. Zool, 13:102.
Littlejohn, M. J, (1982) Amphibians of Victoria.Vict.Yearbook 85:1-11.
NSW National Parks and Wildlife Service (1996a) Seven Mile Beach National Park and Comerong Island Nature Reserve. Draft Plan of Management.
Mahony, M. J. (1993) The status of frogs in the Watagan mountains area of the Central Coast of New South Wales. Pp. 257-64 in Herpetology in Australia: A Diverse Discipline ed. by D. Loony & D. Ayers. Royal Zoological Society of New South Wales: Mosman, Australia. Mahony, M. J. (1995) Experimental translocation of the endangered sharp-snorted day frog Taudactylus acutirostris, and observations on the cause of declines among montane stream frogs.The Australian Society of Herpetologist Inc. Newsletter 37:113. Mahony, M. J. (1996) The decline of the Green and Golden Bell Frog Litoria aurea viewed in the context of declines and disappearances of other Australian frogs. Aust. Zool, 30:237-247. Mahony, M. J. (1997) Land and Environment Court Proceedings No. 10710 of 1997, Australian Properties-vSutherland Shire Council. Cronulla Integrated Resort: Fauna assessment, Species Impact for the endangered green and golden bell frog (Litoria aurea). Mahony, M. J. (1998) Land and Environment Court Proceedings No. 10349 of 1998, Kurnell Development Pty Ltd -v- Sutherland Shire Council. Fauna assessment, in particular issues dealing with the impact on the endangered green and golden bell frog (Litoria aurea). Mahony, M. J., and Knowles, R. (1993) Research plan for the Yellow-spotted bell frog and Peppered tree-frog. Annual Report to Australian Nature Conservation Agency. Unpublished Report. Main, A. R. (1965) Frogs of Southern Western Australia. Handbook No. 8. Western Australian Naturalists’ Club — Perth. Markwell, K. W. (1984) A biographical study of the frog fauna of Hexham Swamp. BA (Honours) Thesis, University of Newcastle, Newcastle, NSW. Maxson, L. R.,Tyler, M. J., and Maxson, R. D. (1982) Phylogenetic relationships of Cyclorana and the Litoria aurea species-group (Anura: Hylidae): A molecular perspective. Aust. Zool, 30:643-651. McDonald, K.R. and Alford, R.A. (1999) A review of declining frogs in northern Queensland. Pp 14-22 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra.
92
NSW National Parks and Wildlife Service (1996b) Kooragang Nature Reserve and Hexham Swamp Nature Reserve. Draft Plan of Management. NSW National Parks and Wildlife Service (1998) Wildlife Atlas. Osborne, W.S. (1990) Declining frog populations and extinction’s in the Canberra region. Bogong, 11: 4-7. Osborne, W.S. (1992) Declines and extinctions in populations of frogs in the A.C.T.: a discussion paper. Report 92/8, Australian Capital Territory Parks and Conservation Service: Canberra. Osborne, W. S., Littlejohn, M. J., and Thomson, S.A. (1996) Former distribution and apparent disappearance of the Litoria aurea complex from the Southern Tablelands of New South Wales and the Australian Capital Territory. Aust. Zool, 30: 190-198. Osborne, W.S. and McElhinney, N.A. (1996) Status, habitat and preliminary observations on calling of the Green and Golden Bell Frog Litoria aurea on Bowen Island, Jervis Bay National Park. Aust. J. Zool, 30: 218-223. Pechman, J. H. K., Scott, D. E., Semlitsch, R. D., Caldwell, J. P.,Vitt, L. J. and Gibbons, J. W. (1991) Declining amphibian populations:The problem of separating human impacts from natural fluctuations. Science, 253: 892-895. Pechmann, J. H. K, and Wilbur, H. M. (1994) Putting declining amphibian populations into perspective: natural fluctuations and human impacts. Herpetologica, 50:64-84. Pounds, J. A, Fogden M. P. L, Savage J. M, and Gorman G. C. (1997) Test of null models for amphibian declines on a tropical mountain. Conserv. Biol, 11:1307-22. Pickard , J. (1972) Rate of movement of transgressive sand dunes at Cronulla, New South Wales. J. Geol. Soc. Aust, 19(2):213-216. Pickard, C. R. and Towns, D. R. (1988) Atlas of the Amphibians and Reptiles of New Zealand. Conservation Sciences Publication Number 1. Science and Research Directorate, Dept of Conservation, Wellington. Pyke, G. H. (1995) Fauna Impact Statement for proposed development works at The Homebush Bay Development Area, excluding the Brickpit. Prepared for Olympic Coordination Authority.
Pyke, G.H. and White, A.W. (1996) Habitat requirements for the Green and Golden Bell Frog Litoria aurea (Anura: Hylidae). Aust. Zool, 30: 224-232. Resource Planning Pty Ltd. (1994) Fauna Impact Statement Proposed Extension of the Mt Owen Colliery near Ravensworth, New South Wales. Prepared for Hunter Valley Coal Corporation Pty Ltd. Reynolds, S.J. (1995).The impact of the introduced mosquitofish (Gambusia holbrooki) on mortality of premetamorphic anurans. Unpublished Honours thesis, Dept of Zoology, University of Western Australia. Robb, J. (1980) New Zealand Amphibians and Reptiles. Collins, Auckland. Collins, Auckland. Robinson, M. (1993a) Raising the Green and Gold, or Smooth Swamp Frog Litoria aurea to maturity in captivity. Herpetofauna 23:15-16. Robinson, M. (1993b) A Field Guide to Frogs of Australia. From Port Augusta to Fraser Island including Tasmania. Australian Museum/Reed Books, Sydney. 112 pages. Roy, P. S., and Crawford, E. A. (1978) Holocene geological evolution of the southern Botany Bay — Kurnell region, Central New south Wales Coast. Engineering, Marine and Environment Section, NSW Department of Minerals and Energy, 100 pp. Roy, P. S., and Crawford, E. A. (1979) Holocene geological evolution of the southern Botany Bay — Kurnell region, Central New south Wales Coast. Records of the Geological Survey of New South Wales 20(2):159-250. Skinner, J. E. (1973) A barrier system, Kurnell peninsula, Botany Bay, N.S.W. B.Sc (Hons) school of Earth Sciences, Macquarie University, 158 pp. Thomson, S. A., Littlejohn, M. J., Robinson, W. A., and Osborne, W.S. (1996) Taxonomy of the Litoria aurea complex: a reevaluation of the Southern Tableland populations of the Australian Capital Territory and New South Wales. Aust. Zool, 30: 158-169. Tyler, M. J. (1976) Frogs. Collins, Sydney. Tyler, M. J., (1978) Amphibians of South Australia. South Australian Museum, Adelaide. Tyler, M. J., ed. (1979) The Status of Endangered Australasian Wildlife. Proc. Centenary Symposium of the Royal Zoological Society of South Australia, Adelaide, 21–23 Sept 1978. Published by Royal Zool. Soc. of S. Aust. Tyler, M. J. (1997) The Action Plan for Australian Frogs. Wildlife Australia, Canberra. 77 pp. Tyler, M. J. and Davies, M. (1978) Species-groups within the Australopapuan Hylid Frog Genus Litoria Tschudi. Aust. Zool. Supplementary Series No. 63:1-47. Tyler, M. J. and Davies, M. (1978) Frogs of the Northern Territory. Tyler, M.J. and Davies, M. (1986) Frogs of the Northern Territory. Conservation Commission of the Northern Territory.
Tyler, M. J. and Davies, M. (1993) Family Hylidae. In: Fauna of Australia. (Eds: Glasby, C. J., Ross, G. J. B., Beesley, P. L.) Australian Government Publishing Service, Canberra, 58-63. Tyler, M. J., Smith L. A., and Johnston, R.E. (1984) Frogs of Western Australia. Western Australian Museum. van de Mortel,T. F. and Buttemer, W. A. (1996) Are Litoria aurea eggs more sensitive to ultraviolet-B radiation than eggs of sympatric L. peronii or L. dentata? Aust. J. Zool, 30: 150-157. van de Mortel,T. F., and Goldingay, R. (1998) Population assessment of the endangered Green and Golden Bell Frog Litoria aurea at Port Kembla, New South Wales. Aust. Zool. 30(4): 398-404. Warner, S. C.,Travis, J., and Dunson, W. A. (1983) Effect of pH on interspecific competition between two species of Hylid tadpoles. Ecology, 74:183-194. Watson, G. F. (1993) Diversity and distribution of Australian amphibians. In: Herpetology in Australia. A Diverse Discipline. (Eds: Lunney, D., Ayers, D) Royal Zoological Society of NSW, Sydney, 3. Watson, G. F., and Littlejohn, M. J. (1985) Patterns of distribution, speciation and vicariance biogeography of southeastern Australian amphibians. In:The Biology of Australasian Frogs and Reptiles. (Eds: Grigg,G. C; Shine, R; Ehmann, H.) Surrey Beatty & Sons Pty Ltd, Sydney, 91-97. Webb, C. E. (1994) Does predation by holbrook Gambusia holbrooki (Atheriniformes: Poecilidae) contribute to declining frog populations? B.Sc. (Hons) Thesis. Macquarie University: Sydney. Webb, C. E. and Joss, J. (1997) Does predation by the fish Gambusia holbrooki (Atheriniformes: Poeciliidae) contribute to declining frog populations? Aust. Zool, 30(3):316-324. Wellington, R., and Wells, R. (1994) Herpetofaunal Survey of the Ravensworth State Forest area for the proposed extension of the Mount Owen Coal Project, Hebden, New South Wales. Report prepared for Resource Planning Pty Ltd. and included with FIS for proposed extension of Mount Owen Colliery. Wells, R. W., and Wellington, C. R. (1985) A synopsis of the Amphibia and Reptilia of New Zealand. Aust. J. Herpet. Suppl. Series 1:62-64. White, A. (1993) Faunal Impact Statement. Proposed development of Telecom site at Roberts Road, Greenacre. White, A.W. (1995a) The Green and Golden Bell Frog (Litoria aurea). Frog Facts 5:1-4. White, A.W. (1995b) Disappearing frogs. Aust. Zool, 30:1-9. White, A.W. (1997) M5 East Proposal. Strategy for compensatory habitat, monitoring and long-term maintenance of the Marsh Street wetlands, Arnecliffe. Report to the Roads and Traffic Authority. White, A.W. and Pyke, G.H. (1996) Distribution and 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 We conducted surveys for the green and golden bell frog (Litoria aurea) in north-eastern NSW, north of
A management plan is urgently needed for the bell frogs in Yuraygir National Park in order to address disturbances to breeding sites and to establish 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 substantially in NSW during the last 20 years that it has been listed in NSW as a threatened species since 1992 (Lunney et al. 1996) and was recently (late 1997) listed by the Commonwealth as a vulnerable species (Commonwealth of Australia 1997). The historic range of the species in NSW extended from the Victorian border to as far north as Byron Bay. It was known from 92 sites where it occurred in large numbers (White and Pyke 1996). It has now disappeared from about 50 sites and at only five sites are there known to be populations of more than 20 individuals (White and Pyke 1996; van de Mortel and Goldingay 1998).
1996 and 1998. Bell frogs were recorded at only two of these sites, both within Yuraygir National Park where the total number of frogs detected was 10. No evidence of breeding was found. We also conducted surveys for the introduced plague minnow (Gambusia holbrooki) and found it at nine of the 15 sites. This fish was absent from the sites occupied by bell frogs. This study suggests there may only be one population of bell frogs remaining in this part of northern NSW but we recommend further surveys to verify this.
A fundamental component of developing a conservation strategy for a species is to be able to conserve it throughout its geographic range (e.g. Murphy and Noon 1992). For 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 (D:N)
Frog Surveys Present
1. Ocean Shores E 55220, N 684500 2. Tyagarah Nature Reserve E 55580, N 683290 3. Byron Sewerage Works E 55630, N 683290 4. Lake Ainsworth E 55760, N 681580 5. Ballina Football Fields E 55550, N 680590 6. Ballina Bicentennial Park E 55350, N 680850 7. Boundary Creek Road E 54670, N 679275 8. Evans Head Sewage Works E 54150, N 678140 9. Broadwater NP E 53890, N 678210 10. Wendoree Lagoon E 53340, N 676230 11. Primitive Area Lagoon E 53350, N 676080 12. Diggers Camp E 55280, N 670000 13. Station Creek E 52380, N 668640 14. Station Creek Road E 52300, N 668630 15. Blue Lake E 52320, N 668480 P = present;
– = absent.
Bell Frog (D:N)
4:3 6:4
1986
2:1
Fish Surveys Density
Fish
3:2
2.3
3:2
0.6 *
–
P
4:4
1985
3:2
0.2 *
4:3
1972
3:2
12.0
5:2
3:2
2.8 *
1:1
1:1
-
3:2
2:1
1.1 *
3:3
1:1
1.8 *
3:3
1:1
0.4 *
2:1
1:1
–
3:2
1993
1:1
–*
5:4
1995
2:2
–
4:3
BF
2:2
–*
6:4
1994, BF
2:2
–*
* = the presence of firetail gudgeons.
range into appropriate management units (Goldingay 1996). Conserving species in this way will preserve any genetically distinct populations, reduces correlation of environmental factors and provides security against catastrophes. One of the key areas for conserving the green and golden bell frog will be in the north-eastern portion of its range. Peripheral areas of species’ ranges may contain populations that are genetically distinct from those in other portions of the range (e.g. Lesica and Allendorf 1995). White and Pyke (1996) provide evidence that just two distinct populations of green and golden bell frogs now occur in north-eastern NSW, although several others were present historically. The aim of this study was to further assess the status of the green and golden bell frog on the far north coast of NSW. This will allow recovery planning for this species to begin in this region.
METHODS We surveyed 15 sites (Figure1) between Red Rock (40 km north of Coffs Harbour) and Ocean Shores (15 km north of Byron Bay) based on those described in the published literature and the presence of suitable habitat. Historic records of bell frogs occurred at five of these sites (Clancy
1996; White and Pyke 1996). Another site (Ocean Shores) which we surveyed was mistakenly reported by White and Pyke (1996) as an historic location (White pers. comm.; Tarvey pers. comm.). Bell frogs have never been detected at the dam site at Ocean Shores despite regular surveys during the 1980’s (van Beurden pers. comm.). The other eight sites contained suitable habitat (Pyke and White 1996) and occurred within the historic range. One of these sites was located approximately 1 km from Blue Lake where bell frogs had been found most recently (Clancy 1996). Sites were surveyed for frogs at night and during the day, between March 1996 and March 1998. During each visit to a site at night, a hand-held spotlight was used to survey all suitable habitat. A minimum of 30 minutes was spent searching each site. This survey method was used extensively to detect bell frogs at sites near Wollongong (van de Mortel and Goldingay 1998; Goldingay and Lewis unpublished data). All bell frogs seen and heard calling were counted; frogs were sexed based on the presence of dark nuptial pads in males >5 cm snout-vent length. A dip-net was used to search for bell frog tadpoles and involved continual random sweeping whilst wading through the site. Different zones of the water column were surveyed to increase the surveys’ effectiveness.
95
FIGURE 1: Location of study area.
Survey sites are: 1. 2. 3. 4. 5.
Ocean Shores Tyagarah Nature Reserve Byron Sewerage Works, Lake Ainsworth, Ballina Football Fields, 96
6. 7. 8. 9. 10. 11.
Ballina Bicentennial Park, Boundary Creek Road, Evans Head Sewerage Works, Broadwater National Park Wendoree Lagoon Primitive Area Lagoon
12. 13. 14. 15.
Diggers Camp Station Creek Station Creek Road Blue Lake.
The plague minnow (Gambusia holbrooki) (McDowell 1996) has been implicated in the demise of the green and golden bell frog in NSW (Morgan and Buttemer 1996; Pyke and White 1996). Surveys were conducted at all but one site to document the presence and density of this fish (Table 1). Fish were surveyed using a dip-net (diameter 35 cm) to scoop through approximately 30 litres of water in the water column. This was repeated three times in adjacent water columns (open water, ecotone between inshore and open water areas, and the inshore zone). Fish surveys were conducted during the day and at night and averaged. Data are reported only for the inshore microhabitat at each site because other microhabitats gave an inadequate representation of fish density.
RESULTS Despite multiple surveys across the 15 sites, we were able to detect bell frogs at just two sites (Table 1). The sites with the older bell frog records were surveyed extensively without success. We were able to confirm the historic presence of bell frogs at Lake Ainsworth, near Lennox Head. This observation was made in about 1985 by Dr E. van Beurden who had continued his extensive surveys for cane toads in the area (see van Beurden and Grigg 1980). The two sites where bell frogs were detected are in Yuraygir National Park (NP) and about 1 km from each other. At Blue Lake, two bell frogs were seen in September 1996 and five were seen at this site during warm conditions in August 1997. These initial surveys focussed on the southern end of Blue Lake because of the ease of access and small area (ca. 0.3 ha) compared to the whole lake (ca. 5 ha). In December 1997, a canoe was used to survey the entire perimeter of Blue Lake at night. This included regular stops to mimic the call of the frog and to illuminate the extensive beds of cumbungi. Nine male bell frogs were detected. Two bell frogs (at least one was a female) were detected during a similar census in March 1998. Thus, we have direct evidence of 10 individuals at this site. We detected bell frogs during all visits to Blue Lake. At the Station Creek Road dam, no frogs were detected in 1996 but two males were detected by call in August 1997. This new location is approximately 1 km from Blue Lake. Diggers Camp was also surveyed during the 1997-98 season but no frogs were detected despite surveys being undertaken in both wet and dry periods. We were unable to detect any evidence of breeding (i.e. amplecting frogs, tadpoles, juveniles) by bell frogs at either of the two occupied sites. Searches for bell frog tadpoles were also conducted at the other sites but none were found. The plague minnow was found at 9 of the 15 sites (Table 1). The densities observed were generally low except for one 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.
DISCUSSION This study has provided further evidence of the grim situation for the green and golden bell frog in north-eastern NSW. We were able to confirm its presence at just one of six sites in our study area where there was an older record of its occurrence. We located one new site close to that historic site. Our surveys were repeated several times (including day and night) at most sites and spanned many months, so we believe that this is a true reflection of the bell frog’s status at these sites. Many further sites must be surveyed across the study area before a conclusion can be reached that the species’ distribution has contracted to Yuraygir NP. 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 Clancy (1996). Similar surveys at sites near Wollongong have detected up to 89 adult bell frogs (Goldingay and Lewis unpubl. data). Only adult frogs were detected at Blue Lake and there were no signs of breeding, suggesting that breeding activity may be suppressed. Our surveys in Wollongong have often detected juvenile and subadult frogs and occasionally tadpoles. We were unable to detect bell frogs in the vicinity of Diggers Camp. However, single bell frogs were found here in November 1993 at two sites about 2 km apart (Hines pers. comm.) and two bell frogs were found in October 1997 (Hines pers. comm.) in the swamp we searched. None were present there in September 1998. These observations suggest a very small number of bell frogs has held on at this site which is located 15 km north of Blue Lake. There was a high level of disturbance at most sites except those in Bundjalung NP (Wendoree and Primitive Area Lagoons), although these lagoons are the legacy of sandmining in the past. The wetlands in southern Yuraygir NP were subject to disturbance from wild horses which trampled the edges of the wetlands and fed on the aquatic plants. Such disturbance has been present for many years (Clancy 1996). The presence of exotic predators such as plague minnows, cane toads and foxes may have influenced the distribution of the bell frog. There is an apparent negative association between the plague minnow and the bell frog in our study area which is consistent with the findings of Pyke and White (1996). However, this fish was absent from two sites without frogs. Bell frogs occur at one site near Wollongong where plague minnows are abundant but coexistence may be facilitated by submerged vegetation (van de Mortel and Goldingay 1998). Thus, it is difficult to fully implicate this exotic fish in the decline of the bell frog in northern NSW but it may have contributed to the decline. For example, the Ballina Football Fields were constructed on a large freshwater wetland and the remaining ponds now have high densities of plague minnows. Management of sites with bell frogs should always endeavour to exclude this fish.
97
It is not known whether the cane toad has had an adverse effect on the bell frog but toads occur at high densities at Lake Ainsworth where a small number (ca. 2) of bell frogs was recorded over several years up until about 1985 (van Beurden pers. comm.). Toads have been present in this area since about 1968 (van Beurden and Grigg 1980). At present, toads do not occur south of Evans Head in the study area. The sites where bell frogs were detected in Yuraygir NP require urgent management. Feral horses which have affected the aquatic vegetation and trampled the edges of the water bodies must be removed from the National Park. The microhabitats present on pond edges are required by bell frogs for shelter, foraging and breeding (van de Mortel and Goldingay 1998). There is no direct evidence of any breeding activity by bell frogs in Yuraygir NP over the last 5 years (see also Clancy 1996). Therefore, an attempt should be made to assist frogs to breed. This could include either collecting of amplecting pairs in order to collect spawn masses (see van de Mortel and Buttemer 1996) or providing artificial breeding ponds. The latter has been used as a technique for sampling the tadpoles of hylid frogs (e.g. Zimmerman and Simberloff 1996) and should be trialed as a way of collecting bell frog spawn. In either case, tadpoles would be reared to a large size or to metamorphling stage before release. Small breeding ponds may need to be constructed as a more permanent solution to aid breeding at this site. The discovery of two frogs in a dam approximately 1 km from Blue Lake suggests that bell frogs will readily colonise artificial breeding sites in this region as has been documented elsewhere (Pyke and White 1996). Further surveys are required throughout the study area over the next two years to confirm the results of this study and to provide continued monitoring of the size of the bell frog population in Yuraygir NP. The current absence of breeding may reflect several years of poor summer rain. Whatever the case, the small size of this population shows that it will be very vulnerable to local extinction and a broader strategy should be considered to increase the number of occupied sites (Goldingay 1996) in order to conserve this most northern population. The population in Yuraygir NP now apparently represents all that remains of the northern population of the green and golden bell frog. There is a major discontinuity in range between this area and the nearest southern populations (see White and Pyke 1996). This population is therefore likely to be genetically distinct (e.g. Lesica and Allendorf 1995; Colgan 1996) from other bell frog populations and must be given a high priority for conservation.
ACKNOWLEDGMENTS 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
REFERENCES Clancy, G.P., (1996) The green and golden bell frog in the Station Creek area of Yuraygir National Park. Aust. Zool, 30: 214-17. Colgan, D., (1996) Electrophoretic variation in the Green and Golden Bell Frog Litoria aurea. Aust. Zool, 30: 170-76. Commonwealth of Australia, (1997) Declaration under s18.(1) of the Endangered Species Protection Act 1992. Goldingay, R.L., (1996) The Green and Golden Bell Frog (Litoria aurea) — from riches to ruins: conservation of a formerly common species. Aust. Zool, 30: 248-56. Lesica, P. and Allendorf, F.W., (1995). When are peripheral populations valuable for conservation. Cons. Biol, 9: 753-60. Lunney, D., Curtin, A., Cogger, H. G., and Dickman, C. R., (1996) An ecological approach to identifying the endangered fauna of New South Wales. Pac. Cons. Biol, 2: 212-31. McDowell, R., (1996) Freshwater Fishes of South-eastern Australia. Reed Books, Chatswood. Morgan, L.A. and Buttemer, W.A., (1996) Predation by the non-native fish Gambusia holbrooki on small Litoria aurea and L. dentata tadpoles. Aust. Zool, 30: 143-49. Murphy, D.D. and Noon, B.R., (1992) Integrating scientific methods with habitat conservation planning: reserve design for northern spotted owls. Ecol. Appl, 2: 3-17. Pyke, G.H. and White, A.W., (1996) Habitat requirements for the Green and golden Bell Frog Litoria aurea (Anura: Hylidae). Aust. Zool, 30: 224-32. van Beurden, E.K. and Grigg, G.C., (1980) An isolated and expanding population of the introduced toad Bufo marinus in New South Wales. Aust. Wildl. Res, 7: 305-10. van de Mortel,T.F. and Buttemer, W.A., (1996) Are Litoria aurea eggs more sensitive to Ultraviolet-B radiation than eggs of sympatric L. peroni or L. dentata? Aust. Zool, 30: 150-7. van de Mortel,T. and Goldingay, R., (1998) Population assessment of the endangered green and golden bell frog Litoria aurea at Port Kembla, New South Wales. Aust. Zool, 31: 48-54. White, A.W. and Pyke, G.H., (1996) Distribution and conservation status of the Green and Golden Bell Frog (Litoria aurea) in New South Wales. Aust. Zool, 30: 177-89. Zar, J.H., (1984) Biostatistical Analysis. Prentice-Hall, London. Zimmerman, B.L. and Simberloff, D., (1996) An historical interpretation of habitat use by frogs in a central Amazonian forest. J. Biogeog, 23: 27-46.
Loss and Degradation of Red-Crowned Toadlet habitat in the Sydney Region Karen Thumm and Michael Mahony *
ABSTRACT The entire distribution of the red-crowned toadlet (Pseudophryne australis), a frog listed as vulnerable under the New South Wales Threatened Species Conservation Act 1995, is within the geological region referred to as the Sydney Basin. Mapping of site
six extant sites from across the Basin were studied and were generally within the top 40% of the slope below the ridge-top. Breeding sites were in ephemeral situations.We postulate that the position of the breeding sites is related to the geology and geomorphology associated with eroded Hawkesbury 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 The red-crowned toadlet, Pseudophryne australis, is a small 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 larval development are restricted to non-perennial watercourses that generally form below the ridge line. The egg mass is usually terrestrial but within the bed of a non-perennial watercourse. Typically the eggs are placed in a moist, but not flooded situation, under some form of debris such as thick leaf litter or under rock. The clump of eggs has often been referred to as a nest (Harrison 1922; Woodruff 1976; Barker and Grigg 1977). Following rain, water flowing down the non-perennial watercourses passes through the egg mass and releases the embryos from the egg capsules, sweeping them into small ponds within this watercourse, where the tadpole stage and metamorphosis occur (Parker 1940; Barker and Grigg 1977). These ponds are ephemeral and high embryonic and larval mortality due to desiccation leads to a low recruitment rate (Thumm unpubl. data). Woodruff (1976, 1978) considered that the habitat requirements and reproductive biology of the red-crowned toadlet differed markedly from the other south-eastern species of Pseudophryne (P. bibronii, P. dendyi, P. semimarmorata) which he reported to be seasonal breeders. He noted that in these species the sites selected for breeding flooded and retained water for extended periods. He also reported that this distinctive seasonal breeding pattern coincided with the regular autumnal rainfall. Apart from the general descriptions of breeding sites provided by Harrison (1922) and Woodruff (1978) there is limited data on the habitat features of the breeding sites of P. australis. Most descriptions of the distribution of the red-crowned toadlet simply note that it is restricted to the Sydney Basin of New South Wales although some descriptions have considered underlying geological substrates (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 4 million (Department of Urban affairs and Planning 1995). It covers an area of about 26 478 km2 and the area of dense urban development within this Basin is about 4 000 km2 (Sydway 1996) or 15%. This figure does not include large towns in coastal areas north and south of Sydney. A considerable portion of the urban areas to the west of the central business district occurs on the Cumberland Plain which is primarily on Wianamatta shale substrates, but to the north, south and west of this plain urban development occurs mainly on Hawkesbury Sandstone substrates. It is in these areas where urban development has the potential to destroy the habitat of the toadlet and it is in these areas where current and future urban development is concentrated. Continued destruction and degradation of this habitat is inevitable as Sydney grows to accommodate an increasing population. This paper gives an overview of the characteristics of the red-crowned toadlet’s habitat, its distribution and its breeding biology. We argue that the combination of these features make this species particularly sensitive to changes related to urban development. An examination of the habitat
100
preferences of the species at a regional scale is also presented to assist in gauging the impact of future development on this species.
METHODS Distribution To determine the historic and current distribution of the redcrowned toadlet, location records of Pseudophryne australis were collated from museums, from the New South Wales National Parks and Wildlife Service (NPWS) Atlas of NSW Wildlife and from literature. Following collation of all records, multiple records were removed. The geological strata within the Sydney Basin were mapped and the areas covered by each were calculated (1:500,000 Geological Sheet, NSW Dept. of Mines 1969). The red-crowned toadlet records were superimposed onto this map in order to calculate the relative number of records found in each geological formation. The extent of dense urban development was estimated using a street directory (Sydway 1996).
Habitat To determine the habitats occupied by the red-crowned toadlet at the landscape level the topographical features of 56 extant red-crowned toadlet breeding sites including slope, altitude (to the nearest 5 metres), distance below the ridgetop measured as a vertical distance, and position in relation to cliffs were recorded. The breeding area was defined by surveying upslope and downslope from the areas where calls were first heard until no more calling males were located.
Rainfall Daily rainfall statistics were obtained for the period between 1992 and 1996 for the Turramurra weather station (Bureau of Meteorology). This was the closest station with a complete rainfall data set to the site where reproductive biology was studied (approx. 10 km distance).
Breeding Biology In order to assess the broad features of the reproductive biology of the red-crowned toadlet a breeding site was studied over a six year period. The site chosen consisted of an ephemeral watercourse that contained several breeding sites and ponds. Data was collected on the hydroperiod of breeding sites, oviposition frequency and recruitment. The site in Hornsby Heights (Australian Metric Grid (AMG) 56H 324000 6274500), was in the core area of the distribution of this species (Figure 1). It contained two depressions, which filled after heavy rain to form ephemeral ponds. The lower pond had three 0.9 m sides and was 0.13 m deep and the upper pond was 0.3 m x 0.2 m x 0.05 m. There was also one semipermanent pond of approximately 0.30 m in diameter and 0.12 m depth, which was located on a rock shelf. To assess whether the hydroperiod of this site was representative of red-crowned toadlet breeding sites, sixteen other ephemeral watercourses known to have populations of red-crowned toadlets were visited, and a total of 31 ponds measured. The area used for the long-term study was visited 239 times between September 1992 and August 1995 in order to collect data on oviposition frequency and recruitment.
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 month in the first year, 5 visits per month in the second year, and 5.4 per month in the third year averaging out at 6.4 visits per month over the three years. Individual females were identified by reference to their unique ventral pattern. The reliability of this method was tested using specimens in the Australian Museum, by drawing the ventral markings of 20 adult frogs, recording their registration number separately, and then asking Museum staff to use the drawing to identify individuals. To test the assumption that pattern does not change with time, a series of 12 captive held individuals were compared over a period of up to nearly three years. The frequency of visits to the main study site made it possible to record any new clutches within a few days of being laid, to follow through the persistence of ephemeral ponds and to record successful metamorphosis. To prevent disturbance affecting the reproductive recruitment at this study site, counts of egg mass size (numbers of eggs/clutch) were made at another nearby site in Berowra Heights (AMG 56H 328400 6278650), approximately 4 km distant.
Threats to Red-Crowned Toadlets from Urbanisation To assess the impact of urban development fifty-six sites were visited over 4 years (Figure 2). Disturbance indicators resulting from urbanisation were routinely recorded: siltation, soil pH, weed presence and species composition, presence of stormwater outlets, quantities of remaining bushrock as a percentage of the surface area, and fire hazard reduction activities.
RESULTS Distribution A total of 431 records were obtained from museums (292), the NPWS Atlas of NSW Wildlife (130) and from the literature (9). Based on these records, there were 141 localities for which P. australis was recorded (Figure 1). The limits of the distribution were defined as between Mt.Victoria in the west to Pokolbin in the north and near Barren Grounds in the south (Figure 1). The majority of records fell within the Sydney Basin (417 of the 431 records). Details relating to the fourteen records that fell outside this region were closely scrutinised. Those held by museums were inspected, the accession register checked and in a few cases the collector of the specimen was contacted.
Three specimens and one Atlas record (AM R9407 Jaffa Texas Ashord Downs, AM R70161 Lightning Ridge, Smith Lakes AM R78699 and the (NPWS) Border Ranges) were found to be misidentifications. One specimen from Point Lookout in the New England region (AM R45704) had external features consistent with identification as a specimen of P. australis. The specimen was faded but a distinct head mark typical of P. australis was apparent. This specimen was transferred to the Australian Museum from the University of New England Department of Zoology collection. Comparison of the registers revealed that errors had occurred in transcription of the collectors’ identity and date. Further, there were three specimens listed in the University of New England register but only one specimen was in the Australian Museum. When the collectors (de Bavay and Frazier) were contacted, they did not recall collecting the specimen in that area. There were similar problems with a specimen collected in “Armidale” (AM R45705). Firstly the data was incorrectly transposed to the AM register, and secondly, the collector (Frazier) did not recall this specimen. A third specimen (AM R 45703) from the region “20 mls W of Armidale” cannot be verified as an accurate record because none of the data in Pengilley’s (the collector listed in the NE and AM register) logbook is similar to the Museum register’s records. It is doubtful whether it will ever be possible to discount these three records from the NE of NSW completely, because it is not known who and where the specimens were collected. However, considering the large number of inaccuracies relating to these specimens, it appears as if these records should be given little weight. Three other specimens from outside the Sydney Basin could not be found in museums (Bowning AM R337, Gapstead Railway Station MOV D6889 and Oberon AM R12269). One specimen from Tom Groggin (AM R12982) has been sent to the Philippines National Museum. It is assumed that this specimen was a P. dendyi, because in some parts of their distribution, this species has a similar pattern to P. australis except that the crown is yellow. Specimens from Kiandra (R03350 — A/B, R03351) were sighted by Mark Hutchinson and “are certainly not australis” (correspondence from Hutchinson, curator of herpetology in the South Australian Museum). There were two coding or transcription errors in the NPWS Wildlife Atlas which were ascertained by the authors. There were historical records for red-crowned toadlets from areas of Sydney that are now most densely populated, e.g. Chatswood. A few populations remained in natural bushland remnants within urban areas, e.g. Bradley’s Head, Manly Dam Reserve, Lindfield.
TABLE 1: Number of red-crowned toadlet locations within each of the geological formations of the Sydney Basin.
Geological Formation Sydney Basin Quaternary Sands and Volcanics Wianamatta Shale Hawkesbury Sandstone Narrabeen Group Coal Measures
102
km2 26 2 2 7 11 2
478 863 843 045 144 583
% of total
No. of localities
% of total
100% 10.8% 10.7% 26.6% 42.1% 9.8%
141 2 2 125 11 1
100% 1.4% 1.4% 88.7% 7.8% 0.7%
Association with Geological Strata
Breeding Biology
The relationship between distribution and geological strata is presented in Figure 1. Eighty-nine percent of locations were associated with the Hawkesbury Sandstone which is exposed in 26.6% of the total area of the Sydney Basin (Table 1). There were 7.8% of records within the areas dominated by the Narrabeen Group of sandstones which comprise 42.1% of the Basin (11 144km2). Only two records were located within the Wianamatta Shale formation of the Cumberland Plain, and both were based on relatively old (1889, 1923) museum specimens. It can be assumed that the locality data supplied with these specimens only mentioned the closest settlement. One of the two records assigned to Quaternary Sands is an older specimen (1894) and one lacks collection data. The one record assigned to coal measures is on the scree slope below the escarpment on the western rim of the Basin. Four localities could not be assigned to a geological formation due to the vagueness of the locality data (e.g. “Sydney”).
Low Recruitment Rate
There were records from the rim of the Basin, from Mt Victoria in the west, from near Barren Grounds in the south and from Pokolbin in the north (Figure 1). There were more records in the north of the distribution area than in the south. Only 12% of all records were found in the southern half of the distribution area (south of Botany Bay, 34°Lat S) although this area makes up approx. 40% of the entire area of the Basin.
Habitat Characteristics Geomorphological Relationship A diagrammatic representation of the relationship between red-crowned toadlet breeding sites and the geomorphology based on the examination of 56 extant sites is presented in Figure 2. Breeding sites were generally located just below the first escarpment (60% of sites studied) or in areas where there were large rock outcrops (80%) (Figure 3). All except two were located within the first 100 m altitude from the ridge-top (Figure 4). Sites were generally (87% of 56 sites) within the top 40% of the slope (Figure 5), and only four sites occurred on plateaus. Most breeding sites (67% of 56 sites) were associated with a non-perennial natural drainage lines. No breeding sites were situated within swamps that occasionally form on the plateaus in the Sydney Basin area (referred to as hanging swamps by Keith and Benson 1988 and Keith and Myerscough 1993), although red-crowned toadlets were found in the small drainage lines that fed into or out of hanging swamps (e.g. Mt.Victoria). Ephemeral ponds had a mean size of 0.85 x 0.51 x 0.07 m. Lengths ranged from 0.22–3.50 m, widths from 0.11 m–2.0 m and depths from 0.05–0.25 m.
Rainfall The rainfall pattern for Turramurra for three years is presented in Figure 6. Two broad features were apparent. Firstly, there was no distinctive seasonal pattern, and secondly, total rainfall varies considerably between years with heavy rainfall or drought occurring in any month of the year.
At the long-term study site at Hornsby Heights, during the period from Sept 1992–August 1995, there was water in the ephemeral ponds on 18 occasions. Of 57 egg masses recorded from the site, 11 tadpoles metamorphosed successfully. A large proportion of the terrestrial egg masses dried up before the tadpoles hatched and were swept into ponds (Thumm unpubl. data). On two occasions the flow of water through the water course from heavy rain was so strong that all tadpoles were swept out of the ponds to be stranded in the leaf litter and die. Clutch sizes averaged 22 eggs/clutch (average taken from 55 clutches); hence the recruitment rate was less than 1% averaged across all clutches at this site.
DISCUSSION Distribution Red-crowned toadlets showed a strong association with Hawkesbury Sandstone substrates which covers about 27% (7 045 km2) of the Sydney Basin (Table 1). The general perception (Barker and Grigg 1997) that red-crowned toadlets are found on “Sydney sandstone” (including the Narrabeen Group) is misleading. Records on the Narrabeen Group (including sandstones) made up only 7.8% of all records, although the Narrabeen Group comprises approximately 42% of the area of the Sydney Basin. Only two red-crowned toadlets were found in the Wianamatta Shale area of the Cumberland Plain (10.7% of the Sydney Basin). Hawkesbury Sandstone, the geological formation on which Red-crowned toadlets were most frequently found, is closer in to the Central Business District than the Narrabeen Group of sandstones (Figure 1). For this reason it is more susceptible to development. The observed difference in the abundance of sites between the north and the south of the Basin has not been previously recognised. The lower density of records in the south of Sydney may be attributable to a number of factors. This region contains a number of water catchment areas that are closed to access and accordingly there have been fewer collections made. Another possibility is that the topography of the Hawkesbury Sandstone in the south of Sydney is less rugged and perhaps provides less suitable habitat for the frog. This area also has a larger percentage of Wianamatta Shale outcropping than north of 34°S Lat. To determine whether the difference observed is real requires an unbiased sampling strategy. A similar approach would also be needed to understand the link between geomorphology and habitats between these areas. Until this matter is resolved consideration of the impact of habitat alteration in the south of Sydney should take into account the lower density of localities in this part of the distribution area. 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 Sandstone soil landscape illustrating that red-crowned toadlet breeding sites are found below the first escarpment on the talus slope. (Redrawn from Chapman, G.A. and Murphy, C.L. (1989) Soil Landscapes of the Sydney 1:100 000 Sheet. Soil Conservation Service of N.S.W., Sydney).
generalised locality data were presented with many museum specimens, making many records unsuitable for the purpose of precise mapping. Thirdly the scale of the geological maps leads to a level of inaccuracy when considering the intricate 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
regions of the Basin have been subject to relatively close scrutiny and collection while others have been poorly studied. This was apparent in the number of records for the frog from the north shore region of Sydney (area bounded by 33°40’ — 33°50’ Lat, 150° 05’ — 150°15’ Long ) an area which has a long history of urban development. The distribution data show clearly that the species once occurred abundantly in this region. By contrast there were few records from the north west region of the Sydney Basin. This area consists of deeply dissected plateaus that have few roads and access points. Another possible source of error was that only
There is a considerable amount of evidence that shows the redcrowned 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. This pattern differed from the situation described by Woodruff (1976) for the three species of Pseudophryne in south eastern Australia where eggs are laid in a position where a rising groundwater table will inundate the eggs. Furthermore, the ponds in which the tadpoles develop were short-lived when compared with these inundated areas used by other species (Martin 1967;Woodruff 1976).We postulate that the reproductive strategy of red-crowned toadlets is adapted to take advantage of ephemeral waterbodies that support few predators. Perennial situations are not exploited in spite of a low 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 For a frog that relies on ephemeral water bodies for breeding, rainfall patterns and quantities would be predicted to be of importance. Rainfall in Sydney is unpredictable, not seasonal, and often falls as heavy spring/summer thunderstorms or East Coast (winter) Cyclones (Bureau of Meteorology 1991), with heavy falls or droughts occurring in any month of the year. The Sydney Climatic Survey states that “an examination of the records reveals that the totals from month to month and from year to year are quite variable and the rain tends to fall in concentrated bursts”. There is “an average of 6 storms each year” (Bureau of Meteorology 1991). It appears that the opportunistic reproductive strategy of the red-crowned toadlet in all seasons of the year has been shaped by the unpredictable rainfall in the Sydney Basin and the features of the habitats it occupies. Due to the lack of seasonality it is difficult to establish guidelines for habitat management.
Breeding Biology A combination of the ephemeral breeding habitat and the uncertain nature of rainfall in the region were observed to be responsible for the low levels of recruitment in this species. A combination of observations made at the long-term study site and the 56 sites examined for habitat features indicated that disturbance to breeding sites or their hydrology is likely to significantly affect recruitment and hinder the recovery potential of populations of this frog by disturbing the finely balanced breeding strategy which has evolved. Frequent disturbance associated with urbanisation may jeopardise reproductive success leading to the extirpation of populations.
Threats to Red-Crowned Toadlets from Urbanisation The impacts of urban development include direct destruction of habitat and the adverse effects of housing and infrastructure which extend into areas beyond the urban fringe. Housing in Sydney’s suburbs frequently follows ridges. Development is in the headwater of catchments and is accompanied by degradation of the bushland surrounding the houses e.g. pollution, an increase in the amount of water reaching the soil, soil nutrients and weed propagules (Clements 1983). About 15% of the original distribution area available to redcrowned toadlets has been developed, and housing is continuing to expand, especially in the “outer ring” which includes many local government areas within predominantly Hawkesbury Sandstone areas. The Department of Urban Affairs and Planning (DUAP) expects an increase in population to Sydney of 804 000 to 4.5 million by the year 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 and 56% in “release areas” (DUAP Population Predictions 1995). A recent ESD Study (Ecological Surveys and Planning 1998) pertaining to the proposed development of “Landcom” housing in the Hornsby Shire identified 18 areas with a total of approximately 563 lots. All sites are on Hawkesbury Sandstone , and “most...are ridge-tops and spurs or the upper slopes and benches” on the “fringes” of the Hornsby plateau. (Ecological Surveys and Planning 1998). Inevitably the need for increased housing will put pressure on red-crowned toadlet habitat as the edges of the city expand. 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 leading to a development of the talus slope below the first steep escarpment. In established suburbs, land on the fringes of the ridges which have previously been excluded from development due to the topography, are now being “in-filled”. Siltation is generally an “edge effect” of new housing developments, leading to an alteration of the characteristics of the watercourse and to weed invasion (Buchanan 1989). Siltation entering a nesting area has been seen to displace red-crowned toadlets from a preferred nest site (Thumm, unpubl. data). One site visited on eight occasions over a period of two years, and where red-crowned toadlets were frequently observed to be present and to lay eggs under a particular rock, was no longer used for breeding, after large quantities of silt entered the site and surrounded the rock. It has taken 4 years for this site to be used again, but there is still no breeding under that particular rock (Thumm and Mahony unpubl. obs.). It is generally accepted that many household chemicals, e.g. insecticides, herbicides and fungicides, are deleterious to frogs (Tyler 1989). The increased nutrient load in the soils near suburbs (Clements 1983, Buchanan 1989) caused by chemicals such as fertilisers and detergents, results in weed invasion down the water courses coming off the ridge-tops. Stormwater from urban areas is typically directed into the non-perennial water courses used by red-crowned toadlets. We have not observed this species breeding in polluted conditions. Watercourses affected by stormwater adjacent to pristine watercourses supporting red-crowned toadlet populations are not used by this species. It therefore appears likely that the decline in water quality associated with stormwater leads to the loss of this species.
Housing developments also alter the quantities of water flowing through the ephemeral water courses used for breeding. When undisturbed these “upper laterals” (Harrison 1922) only flow after heavy rainfall, after which they are observed to dry up, leaving a few ephemeral ponds. Housing developments typically seal the catchment and concentrate the flows into a few watercourses. Clements (1983) reported an increase in water entering the bushland in suburban areas of 30–50%. It is suggested here that the change in flows created by the redirection of stormwater may affect levels of recruitment. Even slight alterations to the hydrology of a breeding site due to sealing of the catchment or diversion of flows within a development, are likely to alter the balance of conditions within which a terrestrial nest site will be successful. There is the potential for conflict between the requirement to protect residents in Sydney from bushfire and the need to conserve fire-sensitive flora and fauna. Fire hazard reduction activities are generally carried out in the areas just below the escarpment, in the preferred breeding habitat of the redcrowned toadlet. Frequent disturbance and degradation of the habitat of this species, clearing and fire hazard burning is likely to reduce the size of populations, due to the low recovery potential of this species. A colony with 15 calling males monitored prior to a wild fire in 1994 was revisited after the fire. Only one male was calling on a day on which there was a lot of activity at a nearby control study site. No leaf litter remained at all in the breeding area at the site of the fire. It was evident that there would be no nest sites available, nowhere to refuge and no foraging areas. No calling was heard at a second site three years after a fire in 1994. Osborne (1991) stated that fire was a “potential threat” to 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 crevices of cliffs in dry times, but that they are found just below the leaf litter in wetter periods. It is not known which weather conditions should be chosen for fire hazard reduction burns in order to create the type of fire which would have the least effect on population numbers. This species does not have an “off-season” (hibernation or aestivation) making forward planning difficult. Due to their low recovery rate, it is considered that a minimum time span between prescriptions of all fire management activities within an individual redcrowned toadlet site of about 10 years including burns, fire trail maintenance, turbo-mowing and clearing is kept in order to avoid the cumulative impacts of each activity. Another more subtle change associated with urban development are the impacts on nearby adjacent natural habitats. Impacts of activities such as the removal of bush rock for gardens and the invasion of exotic plants are difficult to assess. Because redcrowned toadlets lay eggs beneath rocks and seek refuge beneath them, reduction in the quantities of bushrock causes degradation of their habitat. Typically bushrock is removed from ridge roads and tracks (Schlesinger and Shine 1994), as truck access is required to remove the rock. Areas adjacent to ridges are therefore primarily affected. Bushrock removal has been listed as a Key Threatening Process under the NSW Threatened Species Conservation Act 1995, and red-crowned toadlets are listed as one of the species impacted by this process.
ACKNOWLEDGMENTS Thank you to the following museums and organisations for locality data: Australian Museum, South Australian Museum, Museum of Victoria, Museums and Art Galleries of the Northern Territory, Queensland Museum, Western Australian Museum, CSIRO and NPWS. Thanks also to all naturalists who generously provided us with locality data. Thank you to Ross Sadlier from the AM for checking the reliability of using ventral patterns for identification of individual red-crowned toadlets, to Jacquie Recsei and Marion Anstis for reviewing the manuscript, to John de Bavay for his help in researching the New England specimens, and to Thomas Thumm for help with mapping. The study of the geological preferences of red-crowned toadlets was supported by a Peter Rankin Trust Scholarship. The manuscript was improved by the suggestions of two anonymous referees.
REFERENCES Barker J. and Grigg G., (1977) A Field Guide to Australian Frogs. Rigby, Adelaide. Buchanan, R. A., (1989) Bush Regeneration. Recovering Australian Landscapes. TAFE Student Learning Publications NSW. Bureau of Meteorology, (1991) Sydney Climatic Survey, Australian Government Publishing Service, Canberra. Chapman, G.A. and Murphy, C.L., (1989) Soil Landscapes of the Sydney 1:100 000 Sheet. Soil Conservation Service of NSW, Sydney. Clements, A., (1983) Suburban development and resultant changes in the vegetation of the bushland of the northern Sydney region. Australian Journal of Ecology 8: 307-319. 108
Cogger, H. G., (1992) Reptiles and Amphibians of Australia. 5th Edition. Reed Books, Sydney. Cogger, H.G., Cameron, E.E., Sadlier R.A. and Eggler P., (1993) Action Plan for Australian Reptiles. Australian Nature Conservation Agency, Canberra. Endangered Species Program Project No. 124, 156. Department of Urban Affairs and Planning, (1995). Population Predictions. New South Wales Government. Duellman W. E. and Trueb L., (1986) Biology of Amphibians. McGraw-Hill, New York. Ecological Surveys and Planning (1998) Landcom ESD Study 1. Ed. S. Douglas. Unpublished report for the Total Environment Centre Inc. Sydney. Harrison, L.,(1922) On the breeding habits of some Australian frogs. Australian Zoologist, 3: 17-34. Keith D.A. and Benson D.H., (1988) The natural vegetation of the Katoomba 1:100 000 map sheet Cunninghamia 2 (1): 107-144. Keith D. A. and Myerscough P.T., (1993) Australian Journal of Ecology 18: 325-340. Mahony, S., (1997) Efficacy of the “Threatening Processes’ Provisions in the Threatened Species Conservation Act 1995 (NSW): Bushrock removal and the endangered broad-headed snake. Environmental and Planning Law Journal, 14 (1): 3-15. Martin, A. A., (1967) Australian Anuran Life Histories: Some Evolutionary and Ecological Aspects. Pp. 175-191 in Australian Inland Waters and the their Fauna. A.H. Weatherley ANU Press. Canberra. NSW Dept. of Mines, (1969) Sydney Basin 1:500 000 Geological Sheet. Osborne, W. S., (1991) The biology and management of the corroboree frog (Pseudophryne corroboree). In NSW NSW National Parks and Wildlife Service Species Management Report Number 8 NSW National Parks and Wildlife Service, Sydney. Parker, H.W., (1940) The Australasian frogs of the family Leptodactylidae. Novit. Zool. 42: 1-106. Richards, S. J. and Bull C.M., (1990) Size limited predation on tadpoles in three Australian frogs. Copeia 4: 1041-1046. Schlesinger C.A. and Shine R., (1994) Choosing a rock: Perspectives of a bush-rock collector and a saxicolous lizard. Biological Conservation 67: 49-56. Shine, R. and Fitzgerald, M., (1989) Conservation and reproduction of an endangered species: the broad-headed snake, Hoplocephalus bungaroides (Elapidae). Australian Zoologist, 25(3): 65-67. Sydway (1996) Greater Sydney Street Directory 3rd Edition. Thumm, K and Mahony M., (1997) The red-crowned toadlet, Pseudophryne australis: Pp. 143-156 in Threatened Frogs of NSW Habitats, Status and Conservation Ed. Harald Ehmann, published by the Frog and Tadpole Study Group of NSW Inc. Tyler, M. J., (1989) Australian Frogs.Viking O’Neil. Australia. Woodruff, D. S., (1976) Courtship, reproductive rates, and mating system in three Australian Pseudophryne (Amphibia, Anura, Leptodactylidae) Journal of Herpetology. 10 (3):13-318. Woodruff, D. S., (1978) Hybridization between two species of Pseudophryne (Anura, Leptodactylidae) in the Sydney Basin Australia. Proc. Linn. Soc. of NSW, 102 (3):131-147.
Status of Temperate Riverine Frogs in South-eastern Australia Graeme Gillespie1 and Harry Hines2
ABSTRACT Knowledge of the distribution and population declines of temperate riverine frogs in southeastern Australia is reviewed. Patterns, nature and potential causes of declines are examined, and level of current knowledge of demography and biology of species for addressing declines is assessed.There are nine obligate lotic species and three facultative riverine species currently recognised in temperate south-eastern Australia.
factors limiting distribution and abundance, or identifying causative agents of declines.A range of potentially threatening processes exists throughout the region, some of which are involved in some population declines. Systematic surveys are required to fully ascertain the current status and distribution of riverine species in the region.A strategic monitoring program is required to document ongoing trends in populations of both declining and non-declining species. Research is required on the 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 frog species in Australia (Osborne 1989, 1990; Czechura and Ingram 1990; McDonald 1990;Tyler 1991; Watson et al. 1991; Ingram and McDonald 1993; Mahony 1993; Richards et al. 1993;Trenerry et al. 1994; Hollis 1995; Gillespie and Hollis 1996; White and Pyke 1996;Tyler 1997). A large proportion of species reported to have declined in Australia consists of riverine species from the eastern sea-board region. Of the 41 species listed as Endangered,Vulnerable or Insufficiently Known in Australia by Tyler (1997), 20 are lotic species.This number represents roughly 60% of the lotic frog species in Australia. For many of these species the causes of decline are poorly known.
The region under consideration encompasses the mesic south-eastern region of the continent, as defined by Littlejohn (1981), south of 29° S (Figure 1).This includes much of the coast and ranges of South Australia and Victoria, and the coast and ranges of New South Wales as far north as the Northern Tablelands.This region generally falls within the warm temperate climate type identified by Walter and Lieth (1967 cited in Bridgewater 1987), which has a noticeable winter, with year-round rainfall. Most of this region receives a median annual precipitation in excess of 500 mm, and the annual average evaporation is less than 2400 mm (Bureau of Meteorology 1975, 1989).The region has a predominantly winter maximum rainfall regime in the south, and uniform rainfall in the east (Bureau of Meteorology 1975, 1989).The north has tropical influences and receives a summer maximum rainfall regime.
In the temperate regions of south-eastern Australia there are at least nine lotic species and several facultative streambreeders (Barker et al. 1995). Six of these are listed as threatened by State or Federal agencies, either because of population declines and/or rarity (Table 1).This paper reviews the current status of lotic and facultative stream-breeding species in temperate south-eastern Australia. Many biologists and organisations have contributed to the current level of knowledge of species in this region. Whilst we acknowledge this contribution, it is not an aim of this review to present a 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.
Species considered in this review are either restricted to, or have significant proportions of their distributions within this region.There is not a discrete latitudinal or altitudinal boundary between temperate and sub-tropical regions, and some species occur in both. One exception is the inclusion of the Stream-bank Froglet Crinia riparia from the Flinders Ranges, South Australia, which falls within the xeric Eyrean zoogeographical subregion (after Littlejohn et al. 1993).This species has been included to provide a complete review of lotic species in south-eastern Australia.
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 Booroolong Frog Litoria booroolongensis Lesueur’s Frog L. lesueuri Blue Mountains Tree Frog L. citropa Leaf-green Tree Frogs L. phyllochroa complex Peppered Tree Frog L. piperata Spotted Tree Frog L. spenceri New England Tree Frog
L. subglandulosa
Endangered (NSW), Insufficiently known (Tyler 1997) not listed not listed not listed Vulnerable (NSW),Vulnerable (Tyler 1997) Endangered (Commonwealth); Endangered (Tyler 1997); Critically Endangered (Vic); Endangered (NSW) Vulnerable (NSW), Insufficiently Known (Tyler 1997)
Myobatrachidae Stuttering Frog Stream-bank Froglet
Mixophyes balbus Crinia riparia
Vulnerable (Tyler 1997); Endangered (Vic);Vulnerable NSW) not listed
Facultative stream breeders Myobatrachidae Tusked Frog Adelotus brevis Giant Burrowing Frog Heleioporus australiacus Banjo Frog
110
Limnodynastes dumerilii dumerilii
not listed Insufficiently known (Tyler 1997);Vulnerable (Vic); Vulnerable (NSW) not listed
FIGURE 1: The temperate region of south-eastern Australia, showing main geographic features
DATA COMPILATION AND ASSESSMENT OF POPULATION DECLINES Distributional data for each species were compiled from records held in the Australian Museum (AM), Museum Victoria (MV), National Wildlife Collection (NC), South Australian Museum (SM), New South Wales National Parks and Wildlife Service Wildlife Atlas (NSW Atlas), and the Department of Natural Resources and Environment Atlas of Victorian Wildlife (Vic Atlas).These data were augmented with additional records from the literature, unpublished reports and personal observations/communications from herpetological authorities where available. Data were vetted for obvious mistakes or doubtful records. From this information distribution maps were compiled for each species, showing records pre-1990 and 1990 onwards. Although somewhat arbitrary, many declines were first reported during the 1980’s.The 1990’s roughly incorporates the recent period of general heightened concern about declining frogs and increased attention to conservation status.There has been a relatively high level of surveying activity for frogs throughout the region during this time.This has included general fauna surveys conducted by NSW National Parks and Wildlife Service Biodiversity Study, pre-logging and Regional Forest Assessment surveys in Victoria and NSW, and targeted frog surveys (e.g., CNR 1993; Gillespie and Hollis 1996; Mahony 1996 unpubl.; Ehmann 1997; Holloway and Osborne 1996 unpubl.; Hunter and Gillespie 1999). Many sites where species were historically known to occur were revisited
during this time. Information for the period from 1990 onwards therefore provides a useful comparison of current distributions and status with historical information. Published and unpublished literature was reviewed for assessments or reports of population declines, to assess current knowledge of distribution and population status and level of knowledge of biology of each species; and to identify potential threatening processes.This review was augmented where possible by discussions with other recognised authorities on particular species or those with local experience or expertise in particular regions.
GUILD COMPOSITION AND BIOGEOGRAPHY Lotic species are defined as those which invariably breed in permanent-flowing streams or associated stream-channel habitats. Numerous frog species breed in riparian habitats to varying degrees in south-eastern Australia. In this review, facultative stream-breeders are defined as those species which often breed in permanent-flowing stream habitats, but which also commonly reproduce in lentic habitats away from streams. Nine lotic species occur within this region (Table 1). Most of these are hylids and two are myobatrachids.The hylids comprise two species groups, which are distinctive on morphological and phylogenetic bases (see Tyler and Davies 1978; Hutchinson and Maxson 1987; McDonald and Davies 111
1990).The systematics of some species complexes within these groups are still being resolved (Donnellan, Mahony, Knowles, Foster unpubl. data). The Litoria citropa group (Tyler and Davies 1978) includes the Blue Mountains Tree Frog (L. citropa), the Cascade Tree Frog (L. pearsoniana), the Leaf-green Tree Frog (L. phyllochroa complex), the Peppered Tree Frog (L. piperata) and the New England Tree Frog (L. subglandulosa).The Spotted Tree Frog (L. spenceri) arguably also resides in this group, based upon larval and adult morphology and call structure (Watson et al. 1991; Hero and Gillespie 1993).The L. citropa group is restricted to temperate south-eastern Australia, with the exception of L. pearsoniana. Litoria pearsoniana also occurs in the temperate zone but is predominantly distributed within subtropical regions (McDonald and Davies 1990) and is discussed by Hines et al. 1999. Species in the L. citropa group are predominantly restricted to streams draining the eastern fall of the Great Dividing Range, with the exception of L. spenceri (Watson et al. 1991). The L. lesueuri group in south-eastern Australia comprises Lesueur’s Frog (L. lesueuri) and the Booroolong Frog (L. booroolongensis) (see Barker et al. 1995; Anstis et al. 1998). Litoria booroolongensis is predominantly restricted to streams draining the western slopes of the Great Divide in New South Wales, while L. lesueuri is widespread along both sides of the Great Divide, from Victoria up the east coast to northeast Queensland (AM, NSW and Vic Atlas records). Its status is also discussed by Hines et al. (1999) and McDonald and Alford (1999). Only two obligate lotic myobatrachid species occur in temperate south-eastern Australia: the Stuttering Frog (Mixophyes balbus) and Crinia riparia (Barker et al. 1995). Mixophyes balbus is restricted to the eastern fall of the Great Dividing Range. Crinia riparia is restricted to streams draining the Flinders Ranges in South Australia (Odendaal and Bull 1983). Three facultative stream breeders (all myobatrachids) occur in this region (Table 1). Adelotus brevis is a monotypic genus, occurring predominantly along the eastern fall of the Great Divide, with a limited distribution in some western drainages (AM, QM, and NSW Atlas records). It also occurs in subtropical areas, and its status in those areas is discussed by Hines et al. (1999) and McDonald and Alford (1999).The Giant Burrowing Frog (Heleioporus australiacus) is restricted to the eastern fall of the Great Divide, and is the only eastcoast member of this genus (Lee 1967; Gillespie 1990). Five subspecies of the Banjo Frog (Limnodynastes dumerilii) are recognised (Watson and Littlejohn 1985), only one of which, L. dumerilii dumerilii, appears to commonly reproduce in streams (Martin 1972; Gillespie, pers. obs.).This subspecies is widespread, predominantly on the tablelands and western slopes of the Great Divide, from South Australia to southeast Queensland (Martin 1972). Several other frog species occasionally breed along streams in south-eastern Australia, such as the Broad-palmed Frog (Litoria latopalmata), Peron’s Tree Frog (L. peronii), species of the L. ewingii group (Holloway 1997; Gillespie pers. obs.) and the Common Froglet (Crinia signifera) (Odendaal et al. 1982; Holloway 1997). However, these species rarely breed in flowing streams. Where they do breed along water courses, 112
these species mostly oviposit in lentic stream-side pools (Gillespie pers. obs.) or slow-moving sections (Odendaal et al. 1982).There are no reported declines in these species, and they are not considered in this review.
SPECIES REVIEWS The nature and magnitude of population declines vary greatly between temperate riverine species, as does the knowledge base for making informed assessments of current population status or causes of decline. Available information on demography and nature of population declines is described for each species below:
Booroolong Frog (Litoria booroolongensis) Restricted to New South Wales, predominantly along the western-flowing streams of the Great Divide from 200 to above 1000 m above sea level (asl), from catchments draining the Northern Tablelands to the Tumut River in the southern Highlands (Heatwole et al. 1995; Anstis et al. 1998; Hunter and Gillespie 1999; NSW atlas records).This species was formerly abundant along streams draining the Northern Tablelands (Heatwole et al. 1995) (Figure 2). Further south, there are relatively few historical records (AM records; NSW Atlas).There are problems with accurate identification of L. booroolongensis which have confounded assessments of current status, especially in the south of its range where the species is superficially similar to L. lesueuri (Gillespie 1999). Some records, mostly from south-eastern New South Wales, are not supported by specimens (e.g., NSW Atlas), and require confirmation. Litoria booroolongensis has not been recorded from the Northern Tablelands during the past 15 years despite extensive fauna surveys in recent years by the North-east Forest Biodiversity Study (NSW NPWS 1994), Regional Forests Assessment Program and others (Harris, University of New England, pers. comm. in Anstis et al. 1998; Hines pers. obs.; Mahony, University of Newcastle pers. comm.).There are several recent records of the species on the eastern slopes of the Great Divide, north of Newcastle and south of Wollongong (NSW Atlas), but specimens have not been lodged with any museums. Intensive fauna surveys in north-eastern NSW over the past eight years (e.g., NSW NPWS 1994) have failed to locate this species, suggesting that its current status is extremely rare.The species may have also declined at former sites in the Blue Mountains (Recsei pers. comm.). The species is recently known from near Tamworth (Mahony pers. obs.); this is the only known extant population in northern New South Wales. Further south the species persists in the Turon River and Winburndale Creek, Winburndale Nature Reserve (Macartney, Ranger at Winburndale Dam, pers. comm.). Recent extensive stream surveys targeting lotic species in the Southern Highlands have not located the species in the Tumut River and Yarrangobilly Creek, where it historically occurred (NC records), and only located it in the Goobarragandra River, which may be close to the geographic limit of the species (Hunter and Gillespie 1999). Overall, there have been very few records of the species in the past 5 years, contrasting markedly with the 1980’s when the species was perceived to be abundant (Heatwole et al. 1995).This situation may reflect, in part, a lack of survey
effort in many regions; however, this is not the case in northeastern NSW, nor in the Northern Tablelands (NSW NPWS 1994).The species may have declined over its entire range.
Mountains Scheme structures affecting stream flow. Such disturbances to natural flow regimes may displace L. lesueuri from streams.
Knowledge of the ecology of this species is limited. Broad habitat associations and reproductive biology are described by Anstis et al. (1998).The species persists in a range of habitat types and geographic regions. For instance, in the Goobarragandra River, the species occurs in relatively undisturbed reaches within Kosciuzsko National Park, as well as highly modified reaches in farmland downstream (Gillespie and Hunter unpubl. data).
Both dead and moribund specimens of L. lesueuri which had a chytrid fungal infection have recently been found along some streams in south-east Queensland and north-east Victoria (Berger et al. 1998; Gillespie and Berger unpubl. data).The impact of this disease upon populations of this species is unknown.
Several potentially threatening processes have operated, or are operating, in various parts of the range of L. booroolongensis. Introduced fish occur in many streams in which the species has been recorded (Hunter and Gillespie 1999; Anstis, Berowra Heights, NSW, pers. comm.). Although tadpoles of L. booroolongensis appear to be less palatable to trout compared with some other temperate lotic species (Gillespie unpubl. data), introduced fish may still exert significant predatory pressure upon populations of this species (Gillespie and Hero 1999). Land clearance, forest grazing and timber harvesting have occurred adjacent to many streams, or in the headwaters of catchments in which the species has been recorded (Gillespie and Hines pers. obs.). Flow modification has occurred in many streams and weed invasion, particularly by willows, has grossly modified riparian habitats (Anstis et al. 1998; Hunter and Gillespie 1999).
Lesueur’s Frog (Litoria lesueuri) Litoria lesueuri is widespread from Lerderderg Gorge in central Victoria, along the east coast as far as Cooktown, north-eastern Queensland and along the western fall of the 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 and VIC Atlas records). It is widely thought that this species comprises a number of taxa, with two in temperate Australia (Barker et al. 1995; Mahony 1996 unpubl.; Donnellan and Mahony unpubl. data). Litoria lesueuri (sensu stricto) occurs in the south, from Victoria and along the east coast and ranges to Sydney, and along the western slopes of the Great Divide to the Australian Capital Territory (Donnellan and Mahony unpubl. data). Hybrids of L. lesueuri and the taxon to the north may occur in the upper reaches of the Murrumbidgee River and tributaries around Canberra (Donnellan and Mahony unpubl. data).These taxa are treated in this review as one species. Litoria lesueuri remains abundant at numerous localities, and there is little evidence of any decline (Gillespie and Hollis 1996; Mahony 1996 unpubl.). Notably, this species persists at many sites, in some cases in abundance, where other species have declined (Mahony 1993; Richards et al. 1993; Gillespie and Hollis 1996). Litoria lesueuri is absent, or occurs in very low numbers, in some montane streams in north-eastern Victoria and southern NSW, which may indicate declines in these regions prior to surveys, but this is unknown (Gillespie and Hollis 1996; Hunter and Gillespie in 1999). In the southern highlands, Hunter and Gillespie (1999) found that L. lesueuri and other lotic frog species were absent from many streams below impoundments or other Snowy
Several aspects of the biology of L. lesueuri have been examined. Information has been gathered on microhabitat use of larvae, adult activity patterns, life history, and larval predatorprey relationships (Martin et al. 1966; Richards and Alford 1992; Gillespie 1997a unpubl.; Holloway 1997). Litoria lesueuri possesses several ecological attributes that distinguish it from most other temperate lotic species. It is frequently observed away from streams, sometimes on dry ridges several kilometres from water courses (Mahony 1996 unpubl.; Gillespie pers. obs.). Despite a predominantly lotic reproductive strategy, L. lesueuri also often breeds in isolated streamside lentic pools (Anstis et al. 1998), and occasionally breeds in lentic habitats away from streams, such as quarries and dams (Mahony 1996 unpubl.; Anstis et al. 1998). In contrast to most other lotic species within the region, L. lesueuri persists along some streams through cleared pastoral areas and urban fringes (AM, NSW and Vic atlas records).This may suggest greater tolerance for habitat alteration or disturbance.Tadpoles of Litoria lesueuri are relatively unpalatable to introduced trout, which may afford them more protection against predation from introduced fish compared with other sympatric lotic species (Gillespie in review).
Blue Mountains Tree Frog (Litoria citropa) Litoria citropa has a widespread distribution along streams east of the Great Divide, from the Mitchell River in eastern Victoria (VIC Atlas records), north to the Hunter River in NSW (Anstis and Littlejohn 1996) (Figure 4). It occurs up to 1000 m asl in the north of its range at Blackheath, NSW (Anstis and Littlejohn 1996), but is more widespread at lower altitudes (AM, NSW and VIC Atlas records). In southern NSW and Victoria it is restricted to below 500 m asl (NSW and Vic Atlas; AM records). Littlejohn et al. (1972) report the south-western limit of this species at Aberfeldy,Victoria, based on an Australian Museum specimen with this locality. However, recent surveys found L. citropa to be abundant along the Mitchell River and streams further east in Gippsland, but absent in all streams west of the Mitchell River, including the Valencia, Avon, Macallister, Aberfeldy and Thompson Rivers (Gillespie and Hollis 1996; Gillespie unpubl. data).The Mitchell River catchment is also the south-western biogeographical limit of a number of other warm temperate elements, including several riparian and rainforest plant species (Cameron, pers. comm. Flora Section, Department of Natural Resources and Environment,Victoria).The origin of the Aberfeldy specimen is therefore doubtful; it is more probable that the Mitchell River catchment is the south-western limit of the range of this species.
113
Litoria citropa occurs along slow-flowing pool sections of permanent or semi-permanent streams through dry sclerophyll forest, temperate rainforest and coastal woodland habitats (Tyler and Anstis 1975; Littlejohn 1981; Holloway 1997). It is restricted to streams with intact riparian and adjacent forest vegetation (Gillespie pers. obs.; Anstis pers comm.). Consequently, this species has been displaced from areas cleared for pasture or urbanisation in parts of its historic range. Elsewhere, there is no evidence that this species has declined. However, few surveys have been conducted through most of the range of this species, and only limited monitoring of populations has occurred. Mahony (1993) visited several sites in the Watagan Mountains at the northern limit of L. citropa between 1977 and 1992 and reported that, although at low abundance, there was no evidence of a decline in numbers of L. citropa. The species is well represented within existing reserve systems of southern NSW and eastern Victoria (VIC and NSW Atlas records). Many catchments in which this species occurs are subject to intensive timber harvesting.This species persists along some streams in catchments in which logging has occurred (Gillespie pers. obs.); however, the health or stability of these populations is unknown. Forest grazing and land clearance up-stream also occur in some catchments. Again, the impact of these disturbances is unknown. Several aspects of the biology of L. citropa have been examined. Information has been gathered on microhabitat use of larvae, adult activity patterns, life history, and larval predator-prey relationships (Tyler and Anstis 1975; Anstis and Littlejohn 1996; Gillespie 1997a unpubl.; Holloway 1997; Gillespie unpubl. data).
Leaf-green Tree Frog complex (Litoria phyllochroa (sensu stricto), L. nudidigitus and L. barringtonensis (sensu stricto)) The L. phyllochroa complex appears to comprise three species (Mahony, Knowles and Donnellan, unpubl. data). Litoria phyllochroa (sensu stricto) occurs from the Sydney region to the Coffs Harbour region. Litoria barringtonensis (sensu stricto) (Copland 1957) occurs north from the Hunter River to near Gibraltar Range.The name L. nudidigitus was applied to the southern taxon of Littlejohn (1967), after the subspecies L. phyllochroa nudidigitus identified by Copland (1962). Litoria nudidigitus occurs from the Sydney region along the eastern slopes of NSW to the Thompson River in eastern Victoria. It also has a limited distribution on the north and western sides of the Great Divide in tributaries of the Mitta Mitta River in north-eastern Victoria (Gillespie and Hollis 1996;VIC Wildlife Atlas), and the Upper Murray and Goodradigby Rivers in southern NSW, and the Cotter River in the ACT (Gillespie and Osborne 1994; Hunter and Gillespie in press). The specific taxonomic identities of specimens and records of this complex held in Museums and wildlife atlas databases have not yet been resolved, precluding the presentation of 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
Members of the complex are common and widespread in low to mid-elevation forests. Litoria phyllochroa (sensu stricto) occurs up to 1100 m asl in Barrington Tops. Litoria barringtonensis (sensu stricto) also occurs in the Barrington Tops area but the geographic distribution of this taxon in relation to that of L. phyllochroa (sensu stricto) is currently being resolved (Mahony pers. com.). Litoria nudidigitus occurs up to 1100 m asl in the upper Murray catchment (Gillespie and Hollis 1996). As with L. citropa, these species have been displaced from areas cleared for pasture or urbanisation in parts of their historic range. Elsewhere there is little evidence of any widespread decline of these species (Mahony 1993; Figure 5). Fauna surveys in eastern Victoria during the past 10 years have found L. nudidigitus to be common and widespread along many streams in lowland and coastal areas (VIC Atlas records). However, surveys have not been comprehensive throughout the range of the species, and monitoring of populations has only occurred at a few sites. Mahony (1993) found no evidence of declines of L. phyllochroa (sensu stricto) at several lowland sites monitored in the Watagan Mountains in NSW between 1977 and 1993. Some upland populations of L. barringtonensis (sensu stricto) may have declined in the north of its range (Mahony pers. comm.). Otherwise, L. phyllochroa (sensu lato) has been frequently recorded in recent surveys in north-eastern NSW (NSW NPWS 1994). Above 500 m asl L. nudidigitus is relatively uncommon, with a very patchy occurrence (Gillespie and Hollis 1996; Hunter and Gillespie in press), and this may be due to population declines prior to surveys in these areas.The tadpoles of L. nudidigitus are palatable to introduced trout, which are common and widespread in upland streams of this region (Gillespie 1997a, unpubl.).Trout are probably a major factor in determining the low density of this species in upland steams. Several populations of L. nudidigitus have been regularly monitored in north-eastern Victoria and the Southern Highlands of NSW since 1994 (Gillespie unpubl. data). Most of these populations have remained relatively stable during this time. However, in 1996 one population suddenly declined at its upper limit of distribution (1100 m asl) in a trout-free section of Bogong Creek, Kosciuszko National Park (Gillespie 1997a unpubl.; Gillespie 1998 unpubl.). Monthly surveys during the summers of 1996/97 and 1997/98 located very few frogs or tadpoles in this section of the stream. No frogs or tadpoles were located in the summer of 1998/99. Further downstream, below 800 m asl, the population has remained relatively stable.The cause of this localised decline is unknown. The biology of L. nudidigitus has been extensively studied. Information has been gathered on microhabitat use by adults and larvae, movement and activity patterns, population agestructure and life history, larval competition and predatorprey relationships (Hero and Gillespie 1993; Gillespie 1997a, unpubl.; Holloway 1997; Gillespie in review).The biology of L. phyllochroa (sensu stricto) and L. barringtonensis (sensu stricto) has not been examined specifically, but is likely to be very similar to that of L. nudidigitus.
FIGURE 2: Distribution of the Booroolong Frog Litoria booroolongensis. Open circles indicate records collected pre1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 4: Distribution of the Blue Mountains Tree Frog Litoria citropa. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 3: Distribution of Lesueur’s Frog Litoria lesueuri. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 5: Distribution of the Leaf-green Tree Frogs of the Litoria phyllochroa complex. Open circles indicate records collected pre1990; 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 ❍ pre-1990 record ■ capital cities
115
Peppered Tree Frog (Litoria piperata) Litoria piperata was formerly known from five streams draining the east of the Northern Tablelands, from 800 — 1120 m asl, from Gibraltar Range to Armidale, northern NSW (Tyler and Davies 1985) (Figure 6). Despite searches of the historic localities and other streams with similar habitat within the region (Mahony et al. 1997a; Hines unpubl. data.), the species was last sighted in 1973. Little is known about the biology of L. piperata.The morphological similarity of L. piperata to L. pearsoniana and L. phyllochroa suggests that ecological similarities are also likely. Causes of the apparent disappearance of L. piperata are unknown; however, most of the historic sites and other streams in the region have undergone substantial alteration and suffered significant habitat disturbance through land clearance, grazing and timber harvesting (Hines pers. obs.). Introduced predatory fish species (Eastern Gambusia Gambusia holbrooki and salmonids) also occur in these streams and may have displaced frog populations by predation upon larvae. Populations of frogs which closely resemble L. piperata, were located on the Northern Tablelands north of the known historic range of the species, in 1992 (NSW NPWS 1994) (Figure 6).The advertisement calls of males in these populations were similar to that of L. pearsoniana (Mahony pers. comm. in Tyler 1997).The advertisement call of L. piperata is not known but the species is morphologically
FIGURE 6: Distribution of the Peppered Tree Frog Litoria piperata. Open circles indicate records collected pre-1990; closed circles with ‘?’ are records 1990 onwards with questionable taxonomic identity.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
similar to L. pearsoniana (Tyler and Davies 1985; McDonald and Davies 1990). It is possible that L. piperata represents morphologically distinct outlying populations of L. pearsoniana. Further genetic and morphometric studies are required to resolve the systematics of these populations before the conservation status of L. piperata can be properly assessed. Further targeted surveys are also required of other streams in the region to try and locate any other remaining populations.
Spotted Tree Frog (Litoria spenceri) Litoria spenceri is restricted predominantly to the western fall of the Great Divide, from Lake Eildon to Mount Kosciuszko, from 280–1100 m asl (Gillespie and Hollis 1996) (Figure 7). Extensive systematic surveys have been conducted for this species throughout eastern Victoria and southern NSW (Watson et al. 1991; Gillespie and Hollis 1996; Hunter and Gillespie 1999; Gillespie 1998 unpubl.). Litoria spenceri has only ever been found in 19 streams, and has always been considered to be rare (Watson et al. 1991).The species is now believed to be extinct in four of these streams, and has declined substantially in distribution and abundance along most others (Gillespie and Hollis 1996).These declines occurred in the 1970’s and early 1980’s (Watson et al. 1991); but based upon the known demography of the species, it probably suffered population declines over a wider area earlier in this century, and possibly late in the last century (Gillespie and Hollis 1996).The remaining streams comprise 12 discrete isolated populations (Gillespie and Hollis 1996; Hunter and Gillespie 1999). Based upon density estimates from mark-recapture studies, surveys, population monitoring, and habitat modelling, the largest of these populations is estimated to contain approximately 1000-1500 adults in the upper Goulburn River (Gillespie unpubl. data).The sizes of all other populations are estimated to be less than 1000 adults. 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 as introduced fish and stream disturbances, have also been examined. A national Recovery Plan has been prepared for 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 1994, and was recently expanded to include all populations (Robertson and Gillespie in review). Censuses have been conducted annually along 1 km transects on each stream. Most adult populations have remained relatively stable during the five years of monitoring (Gillespie unpubl. data). However, the population of L. spenceri at Bogong Creek, Kosciuszko National Park, suffered a precipitous decline in 1996 (Gillespie 1997a unpubl.).This population is highly unusual. It is mostly confined to a short (1.6 km) stream reach above a waterfall and below a pondage and aqueduct for the Snowy Mountains Scheme, which excludes trout (Gillespie 1997a unpubl.). It is the highest elevation population, and was the only high-density population known prior to this decline (Gillespie and Hollis 1996). During the three years prior to the decline, the population was intensively studied and estimated to comprise 700–1000 adults and several thousand immature frogs (Gillespie 1997a unpubl.). In the summer of 1996/97 only six adults, three juveniles, and one clutch of tadpoles were located (Gillespie 1998 unpubl.). Over 130
metamorphs emerged from this clutch. In 1997/98 only two of these juveniles were located; no other frogs were found (Gillespie 1998 unpubl.). In 1998/99 one of these juveniles and one adult male, located in the summer of 1996/97, were found (Gillespie unpubl. data). During the season prior to the population decline (1995/96), several dead and one moribund frog were located.The moribund frog was found to be infected with a chytrid fungus, recently identified by Berger et al. (1998).This fungus has been associated with other declining frog populations (Berger et al. 1998), but its role in these declines is not known. An alternative cause of the decline may have been the record flood in Bogong Creek in October 1996 (Snowy Mountains Authority records). However this event occurred prior to any breeding in that season, so no eggs or tadpoles were present in the stream.The species is adapted to living in mountain streams with high energy spring snow melt (Hughes and James 1989) and the adult population is expected to be able to cope with these events. Numbers of 117
L. nudidigitus also declined from this section of stream, but the population further down stream remained stable (Gillespie 1998 unpubl.).This suggests that the causative agent of decline was restricted to the upper reaches. A chytrid fungus was detected in several dead and moribund adults of L. spenceri from three other populations in Victoria during the summer of 1997/98 (Gillespie and Berger unpubl. data). Subsequent monitoring has not detected any declines of these populations (Gillespie unpubl. data).The role of the chytrid fungus in the population dynamics of this and other frogs remains to be resolved.
New England Tree Frog (Litoria subglandulosa) Litoria subglandulosa is restricted to the eastern fall of the Great Divide from Fal Brook in the Barrington Tops National Park, NSW, to near Stanthorpe, just north of the Queensland border, at altitudes from 500 to 1400 m asl (Tyler and Anstis 1975; Heatwole et al. 1995; Anstis and Littlejohn 1996; NSW Atlas; AM and QM records) (Figure 8). Recent taxonomic studies suggest that there may be two sibling species within L. subglandulosa, separated north and south near the latitude 30o 30’ S (Mahony, Knowles, Foster and Donnellan unpubl. data) but in this review it is treated as a single species. Knowledge of the historical distribution of L. subglandulosa is limited. Prior to 1975, this species was known from only three localities (Tyler and Anstis 1975), and few other localities were reported until the 1990’s (Anstis and Littlejohn 1996). Most records of this species are from surveys since 1990 (Figure 8) (see Anstis 1997).These surveys have not been comprehensive and more populations may exist. Consequently, there is a limited historical base for assessment of population declines.The species may have disappeared or suffered a drastic decline in three streams in the vicinity of the type locality, near Point Lookout (Anstis and Littlejohn 1996; Anstis 1997).These authors also report that other lotic species historically common in these streams, L. booroolongensis, L. pearsoniana, and M. balbus, have also declined (Anstis and Littlejohn 1996; Anstis 1997). Anstis (1997) also reports that the type locality has undergone significant alteration since the 1970’s, when L. subglandulosa was common there. Much of the fringing riparian vegetation has gone, presumably from cattle grazing (Anstis and Littlejohn 1996; Anstis 1997). Introduced trout species have been released into these streams (Anstis 1997). Litoria subglandulosa persists at other localities documented in the 1970’s and 1980’s (Anstis and Littlejohn 1996; Anstis 1997). However, surveys and monitoring are currently inadequate to assess population trends. Litoria subglandulosa occurs along slow-flowing pool sections of permanent streams, through dry and wet sclerophyll forest, rainforest, montane forest and heathland (Anstis and Littlejohn 1996; Anstis 1997).The species has also been recorded along streams through semi-cleared grazing land (Anstis 1997). Knowledge of the biology of L. subglandulosa is limited.The reproductive biology and some aspects of life history have been documented (Tyler and Anstis 1975; Anstis and Littlejohn 1996). Apart from broad habitat associations, the ecological requirements of adults and larvae are unknown.The species has a distinctive larval morphology, unique amongst Australian hylids (Tyler and Anstis 1975), which may reflect specialised diet and microhabitat requirements.
118
Several potentially threatening processes operate in catchments containing populations of L. subglandulosa. The species persists along streams in catchments in which timber harvesting has occurred (Anstis 1997); however the health and stability of these populations are unknown. Forest grazing and aerial spraying has occurred at some sites in the vicinity of the type locality (Anstis 1997).Trout have been introduced into many streams, and may be preying on tadpoles (see Gillespie and Hero 1999).
Stuttering Frog (Mixophyes balbus) Mixophyes balbus is restricted to the eastern slopes of the Great Divide, from the Cann River catchment in far East Gippsland,Victoria; to tributaries of the Timbarra River near Drake, NSW (Mahony unpubl. data, AM records; NSW atlas; VIC atlas records) (Figure 9).The species occurs over an altitude range of 20 to over 1400 m asl; from low to high altitudes from south to north (Straughan 1968; AM records; NSW atlas;VIC atlas records). It is typically found in association with permanent streams through temperate and sub-tropical rainforest and wet sclerophyll forest (Mahony et al. 1997b), and also in moist gullies in dry forest (Gillespie pers. obs.; Hines unpubl. data). Mixophyes balbus was formerly more frequently encountered in the northern parts of its range (AM records; NSW atlas; VIC atlas records) than south of Sydney (Figure 9). It may have been uncommon in the south, or this may reflect limited historic searches in that region. Webb (1991) reported the species as ‘common’ in streams at Bondi State Forest, in particular, below 800 m asl.The species has only been found in Victoria on three occasions (Tennyson Creek, Cann River and Jones Creek) (VIC Atlas records), and has always been considered rare in that State. Tyler (1997) reports that the species has declined during the past ten years in many areas where it was historically known. In the northern part of its range Mahony (1993) reported that M. balbus was common in the Watagan Mountains area in the early 1980’s, but had declined by the 1990’s.The species has declined at several sites in the vicinity of the type locality near Point Lookout, NSW, where it was formerly common in the 1960’s and 1970’s (Anstis and Littlejohn 1996; Anstis 1997). Elsewhere the species has only been found in low numbers (Mahony 1996 unpubl.; Mahony et al. 1997b). Surveys in south-eastern NSW since 1990 by State Forests NSW and NSW NPWS have located individuals of M. balbus at only a few sites (Lemckert et al. 1997; Daly 1998; Lemckert unpubl. data). Intensive fauna surveys in East Gippsland conducted in the past 15 years have not located this species (see CNR 1993). Extensive searches at historical sites and streams in close vicinity over the past seven years have not located the species (Holloway and Osborne 1996 unpubl.; Gillespie, Osborne and Holloway and unpubl. data). Mixophyes balbus may now be extinct in Victoria and some other parts of its range in NSW. However, the species is highly cryptic and small populations may remain undetected. Elsewhere the species is now mostly extremely rare.The collective observations of M. balbus (Mahony et al. unpubl. data) suggest that the species tends to have a patchy distribution along water courses. While actual estimates of population size are not available, where populations have
been recorded recently, the species appears to be in very low numbers (Mahony and Knowles pers. comm.). Further surveys are required in the region to properly ascertain the status of the species. Knowledge of the biology of M. balbus is limited. It has more ecological similarities with its sub-tropical and tropical congeners, rather than other temperate zone lotic frog species. Its reproductive biology is very similar to M. fleayi, a congeneric species in north-eastern NSW and south-eastern Queensland (Knowles et al. 1998 unpubl.). Eggs are deposited in very shallow, gently running water either in a shallow excavation in the stream bed or pasted directly onto bed rock (Knowles et al. 1998 unpubl.).Tadpoles are free-swimming, developing in pools and runs (Knowles et al. 1998 unpubl.). The specific ecological requirements of adults and larvae are poorly known.The species has been found away from riparian habitats in the north of its range (State Forests NSW 1995), and adults may routinely disperse into surrounding forests outside of the breeding season. No information is available on population structure and dynamics. In northeastern NSW, statistical modelling was used to investigate the relationship of M. balbus with 24 environmental predictors (NSW NPWS 1994).The species showed a preference for the interiors of large forest tracts in areas with relatively cool mean annual temperatures. Several potentially threatening processes have operated at sites where M. balbus has been found, or up-stream in catchments. Logging and associated forest management practices have been carried out in some catchments where M. balbus historically occurred, or currently occurs (Mahony et al. 1997b).The health or stability of extant populations in these disturbed catchments is unknown. Forest grazing and land clearance for pasture up-stream have also occurred in some catchments (Mahony et al. 1997b). Mahony et al. (1997b) report that the species is not known from any localities with disturbed riparian vegetation or significant human impacts up-stream. However, populations of this species have also disappeared in catchments with seemingly minimal human disturbance (Mahony et al. 1997b).This may indicate that the species is highly sensitive to perturbations in its environment. Tadpoles have been found in sympatry with native fish (Knowles, Mahony and Hines unpubl. data), and probably have survival strategies to avoid predation from them (see Gillespie and Hero 1999). However, the impact of introduced fish, such as Eastern Gambusia (Gambusia holbrooki), carp (Cyprinus spp.) and salmonids, is unknown. Introduced fish (salmonids) have been recorded at sites where M. balbus has declined (Anstis 1997). Mahony et al. (1997b) did not observe introduced fish at any sites where they found M. balbus. Other introduced fish, such as G. holbrooki and Cyprinus spp., may also occur in some streams within the range of this species (see Gillespie and Hero 1999). However, M. balbus has also disappeared from many streams which do not contain introduced fish species (Knowles and Mahony unpubl. data; Hines and Gillespie pers. obs.). A National Recovery Plan is in preparation for this species.
Giant Burrowing Frog (Heleioporus australiacus) The species is confined to eastern slopes of the Great Dividing Range and coastal regions from the southern end of the Olney State Forest north of Sydney, NSW (AM records), to Walhalla, in the Central Highlands of eastern Victoria (Littlejohn and Martin 1967;VIC Atlas) (Figure 10).The species has been found near sea level on the coast, and almost 100 km inland, along the escarpment of the Great Dividing Range, up to 1000 m asl. (Gillespie 1990, AM Records; Webb, pers. comm. in Daly 1996; Recsei 1997). Most records are concentrated at the northern end of the range, in the Sydney region on the Hawkesbury sandstone formation. Most other records are from the southern part of the range, in eastern Victoria and the south-east corner of NSW (Gillespie 1990). However, these constitute only 40 sparsely-distributed documented records (Gillespie 1990;VIC Atlas; NSW Atlas).There is a notable disjunction in the distribution of records between Jervis Bay and the Eden District (Figure 10). One individual has been reported from Narooma State Forests, NSW (Wellington and Wells 1994, unpubl.).The paucity of records in this region may be due to the rarity of the species, although it may also reflect the limited survey effort for frogs in south-eastern NSW. Available information indicates that H. australiacus is rare (Webb 1987; Gillespie 1990; Recsei 1997; NSW Atlas). However, it is widely regarded as a highly cryptic species, usually only detected at night after heavy rains (Gillespie 1990; Daly 1996).This characteristic has hampered the detection of the species during fauna surveys and assessments of relative abundance. It also makes any rigorous assessment of changes in local abundance or population trends impossible. Historically, H. australiacus appears to have been locally common in the Sydney region (Barker and Grigg 1977). Choruses of the species have been reported as “common” after summer rain storms, particularly in and around Royal National Park, NSW, and other fringes of suburban Sydney in the 1960’s and 1970’s (Lee pers. comm. in Gillespie 1990; Grigg, Dept. Zoology, University of Queensland, pers. comm.). There are relatively numerous records of the species from the Sydney region in recent years (Figure 10), but this may reflect an increase in reports in response to heightened interest in frog-population declines. With few exceptions, these recent records have been of single or few individuals. Concerns have been raised that H. australiacus has declined throughout many parts of the Sydney — Hawkesbury sandstone region (Recsei 1997; Mahony pers. comm.). Local declines have mostly been in areas which have suffered habitat fragmentation, along with direct and indirect impacts from urbanisation, which is placing increasing pressure upon many remaining populations in the region (Recsei 1997). Elsewhere, particularly in the southern part of the range of the species, only one or a few adults have been located at any one time (Gillespie 1990; Daly 1996; NSW Atlas). In Victoria, the most individuals ever located were four over several kilometres on a wet road (Gillespie pers. obs.). Webb (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 during this time, have generated most of the records in the state. However, the species has been detected infrequently. 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 conducted by Lunney and Barker (1986) near Bega, and by 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. A range of threatening processes operate across the range of 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 (Gillespie 1990; Daly 1996; Recsei 1997; Gillespie 1997b unpubl.).
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
FIGURE 8: Distribution of the New England Tree Frog Litoria subglandulosa. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 9: Distribution of the Stuttering Frog Mixophyes balbus. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
FIGURE 10: Distribution of the Giant Burrowing Frog Heleioporus australiacus. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
Within the temperate zone A. brevis appears to have declined from the Northern Tablelands of NSW.There are no recent 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, pers. comm.).Targeted surveys for this species throughout its known range in the temperate zone are required to more 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 are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
Eastern Banjo Frog (Limnodynastes 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 subspecies, 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 environments (NSW and VIC Atlas records).The sub-species breeds in both ephemeral and permanent ponds, both 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
FIGURE 12: Distribution of the Stream-bank Froglet Crinia riparia. Open circles indicate records collected pre-1990; closed circles are those recorded 1990 onwards.
● 1990 onwards record ❍ pre-1990 record ■ capital cities
Odendaal and Bull (1982). It is therefore possible to assign most records to a species based on their geographic location (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 range. Given its specialised lotic larval stage it is likely to be susceptible to the types of threats causing declines in other 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 NSW (NSW NPWS unpubl. data), and remains common on the Northern Tablelands (Hines pers. obs.). A limited herpetofauna survey of the western slopes of the Great Divide in southern NSW in 1994-1995 found the sub-species to be relatively common (Lemckert 1998). Populations are currently known from the Main Range in south-eastern Queensland (Hines unpubl. data).
Stream-bank Froglet (Crinia riparia) This species is restricted to streams draining the Flinders Ranges in South Australia, from 400 to over 1000 m asl (Figure 12).The species typically occurs along swift-flowing rocky streams in this region (Odendaal and Bull 1982, 1983; Odendaal et al. 1982). It is the only species of Crinia with a specialised lotic reproductive mode (Odendaal et al. 1982; Odendaal and Bull 1983). The distribution of C. riparia was surveyed extensively from 1977 to 1979 by Odendaal and Bull (1982).The species was found to be common along numerous streams within the region.There have been no formal surveys conducted for this species in recent years. However, during water-sampling surveys conducted in the Flinders Ranges over the past four years, the South Australian Environmental Protection Authority documented frog species (SA EPA unpubl. data). These records do not distinguish between C. riparia and C. signifera. However, these species are largely allopatric with a narrow zone of sympatry, which was surveyed in detail by 122
Knowledge of Species Distributions Over 90 % of records for temperate riverine species have been collected since 1960 (all museum and wildlife atlas records combined). Prior to 1960 the distributions of most species were very poorly known.The assessment of historical distribution is therefore based primarily on records collected between 1960 and 1990. For most species the general distribution and broad habitat associations have been identified. Generally the data are inadequate for modelling specific habitat associations or patterns of decline. Some species have been surveyed extensively across their ranges; but for most, detailed surveys have been conducted in only parts of their ranges. With the exception of L. spenceri, detailed information on the number and sizes of remaining populations of each declining species are not known. Species such as L. citropa, L. phyllochroa, L. subglandulosa and H. australiacus lack adequate base-line survey data across their ranges to determine whether or not general population declines have occurred. Most surveys have been non-systematic or at too low an intensity to provide more than presence-only data. Surveys of presence or absence of a species, or its relative abundance have been undertaken for only a few species and usually at only a few sites.This information deficit limits assessments of magnitude of declines, or analyses of patterns of population declines. Comprehensive surveys ascertaining distribution and relative abundance have only been completed for one species, L. spenceri (Gillespie and Hollis 1996; Hunter and Gillespie in press; Gillespie 1998 unpubl.).This has enabled analyses of
TABLE 2: Summary of population declines of temperate riverine frog species. See text for further explanations.
Species
Disappeared
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
Major decline Major regional Some local Some Local across most of decline; otherwise population population species’ range persistent declines; declines, throughout status otherwise remainder of otherwise persistent and range indeterminate widespread
No evidence of declines
* * * *
patterns of distribution, which has shed light on causes of decline. In addition, this has provided information on population size and distribution, which has then been used to target management actions for protection of the species (Robertson and Gillespie in review).
Population Monitoring
* * * * * * * *
population change have been identified, and the on-going status of each population is known.This has also enabled quick detection of new population declines (Gillespie 1997a unpubl.). Monitoring programs have also recently commenced in NSW and south-eastern Queensland on some other declining species (e.g., Mixophyes species), but these programs need to be expanded to incorporate multiple sites across the distribution of each species so that regional patterns and variation in population trends can be observed. Other declining and “non-declining” riverine species also need to be systematically monitored. In view of the number of species within the guild which have already suffered declines, it cannot be assumed that those species currently considered secure will not suffer declines in the future. Further to this, comparisons of the population dynamics and demography of declining and non-declining species may be informative of causes of decline.
Population monitoring is essential to examining patterns of change in populations, and when ascertaining whether changes are natural fluctuations or unusual declines (Pechmann et al. 1991; Pechmann and Wilbur 1994).The value of monitoring programs has been demonstrated in several studies (Osborne 1989; Richards et al. 1993; Gillespie 1997a unpubl.; Lips 1998; Osborne et al. 1999). Monitoring may provide information not only on magnitude and time of declines, but also on environmental factors influencing population changes. Repeated sampling also provides information on the biology of species, such as movement and seasonal activity patterns.This information is also necessary for survey design, interpretation of survey results and other field observations. For most temperate lotic species, no systematic population monitoring has occurred, or has only recently commenced, despite awareness of declines of a number of species for many years (see Tyler 1997). Only a few populations of some species have been visited repeatedly and over broad time scales (10–20 years) (Mahony 1993; Anstis and Littlejohn 1996).These studies provide broad comparisons of historic and current status at sites, but sampling has generally been too infrequent to enable detailed assessment of the causal factors.
The basic biology of most riverine species in south-eastern Australia, such as life history, general reproductive phenology and broad habitat associations of larvae and adults, is known. However, information on the detailed ecology and demography of most species is inadequate for interpreting or determining causes of declines (Table 3). For some species, such as L. booroolongensis, L. subglandulosa and H. australiacus, this information is relatively limited, or even non-existent in the case of L. piperata. Factors which limit distribution and abundance are unknown for most species.
A population-monitoring program for L. spenceri commenced in 1990. Intensive monitoring with mark-recapture studies have been conducted on two populations since 1992. Annual low intensity broad-scale monitoring has been conducted across most populations since 1994. In 1998 monitoring was expanded to include all extant populations. Populations of L. lesueuri and L. phyllochroa are also monitored at some of these sites.This program has provided detailed information on the behaviour and dynamics of individual populations, and on general patterns of population behaviour. Extraneous factors influencing
Habitat associations and larval ecology of C. riparia have been examined in some detail (see Odendaal and Bull 1980, 1982, 1983; Odendaal et al. 1982, 1984, 1986). Studies of habitat use and larval ecology of M. balbus are under way (Mahony and Knowles unpubl. data). Only one declining species, L. spenceri, has been extensively studied. Some information on life history, habitat use and larval ecology of L. booroolongensis, L. lesueuri, L. citropa, and L. phyllochroa has also been collected through this work. Knowledge of L. spenceri should provide a useful model for aspects of the
Adequacy of Knowledge of Ecology of Temperate Riverine Species
123
ecology and population dynamics other species within the L. citropa group, and a basis for developing some hypotheses to address population declines.
DIFFERENCES BETWEEN DECLINING AND NON-DECLINING SPECIES Reported declines in temperate riverine species are not associated with any particular phylogenetic assemblage. Species from two families and different species groups have declined. Similarly, no clear geographic pattern is evident in declines. Populations of various species have declined on both sides of the Great Dividing Range, from the Central Highlands of Victoria to the Queensland Border. Most population declines have been at altitudes above 500 m asl.The only species to have disappeared, L. piperata, is restricted to this altitude range. Some species have also declined at low altitude sites, e.g., M. balbus (Mahony 1993; Holloway and Osborne 1996 unpubl.).The decline of L. spenceri was found not to be associated with altitude (Gillespie and Hollis 1996). Declines are not restricted to one broad ecological guild of species, although obligate lotic species appear to be affected more than facultative lotic species.The following conclusions can be made about the four species which have clearly suffered major declines (Table 2). 1. They are all obligate stream-breeders, although L. booroolongensis is known to reproduce in isolated streamside pools as well as in the flowing stream (Gillespie pers. obs.). 2. They are species which are mostly restricted to riparian habitats, rather than dispersing widely into surrounding terrestrial habitats.
Most of the species that have not suffered major declines and remain common are either facultative stream breeders, or are known to venture considerable distances from riparian habitats.These species tend to have larger geographic ranges and all occur at low altitudes down to sea level. With the exception of declines of tableland populations of A. brevis, there is little evidence of decline of those species which appear to tolerate gross habitat modification.
POTENTIAL CAUSES OF DECLINES The distributions of most species within south-eastern Australia have undoubtedly been drastically affected by extensive clearing of the lowlands and tablelands since European settlement. For many species which are mostly restricted to forested environments, these changes undoubtedly removed considerable areas of suitable habitat. However, declines have occurred in catchments with minimal gross disturbance or alteration.The causes of these declines are more subtle and difficult to resolve. A range of potential threatening processes has been identified (Table 4).
Introduced Fish Several species of introduced fish occur in streams within the region (see Gillespie and Hero 1999). Species of trout are able to exert significant predation pressure upon larvae of L. spenceri, L. citropa and L. nudidigitus, and have been strongly implicated in the decline of L. spenceri (Gillespie 1997a unpubl.; Gillespie in review).Trout are also present in several upland streams where L. booroolongensis, L. piperata, L. subglandulosa and M. balbus, have declined or disappeared. In view of the vulnerability to trout predation of other members of the L. citropa group, these fish are likely to have had a significant impact on L. piperata and L. subglandulosa, 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
Litoria piperata L. booroolongensis L. spenceri L. subglandulosa L. lesueuri L. phyllochroa complex L. citropa Mixophyes balbus Heleioporus australiacus Limnodynastes dumerilii Crinia riparia Adelotus brevis
Systematic assessment of distribution and abundance
Systematic monitoring of populations
Knowledge of ecology and assessment of potential threatening processes
Completed
All
Advanced
Limited
Some
P A
None
A A A A A P P P
None * *
* * A P P A
* P
Limited
* *
A
A data available on population size or relative abundance P presence/absence data only • existent
124
None
* * * * *
* * * *
* * *
TABLE 4: Summary of threatening processes likely (L) or potentially (P) contributing to population declines of temperate riverine frogs.
Species
L. piperata L. booroolongensis L. spenceri M. balbus L. subglandulosa H. australiacus L. lesueuri L. phyllochroa complex L. citropa Limnodynastes dumerilii Crinia riparia Adelotus brevis
Habitat Forestry Destruction Practices
L L L L L L
P L L P L
Stock Grazing
L P P P L
L
L
P
Introduced Introduced Pesticides Fish Mammals pollutants
L L L P L P L L P P P P
L L P P
P L P
P P
Increased UV*
Disease*
P P P P P P P P P P P P
P P P P P P P P P P P P
* 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 Foxes and cats are common and widespread throughout south-eastern Australia and are potentially a major threatening process to terrestrial frog species such as H. australiacus and M. balbus. Frog bones have been detected in fox scats (Triggs, Dead Finish,Victoria, pers. com.). Specimens of adult Limnodynastes dumerilii have been found in the stomachs of foxes (Marks, Department of Natural Resources and Environment,Victoria, pers. com.).The impact of foxes and cats on frog populations is unknown and warrants investigation.
Forestry Activities Timber harvesting and associated forest management practices have occurred with varying intensities in many catchments throughout the ranges of both declining and nondeclining species in south-eastern Australia.Timber harvesting, associated road construction, and burning practices can significantly affect stream water temperature, sediment loads, turbidity, nutrient levels and flow regimes (Boughton 1970; Langford and O’Shaughnessy 1980; Flinn et al. 1983; Clinnick 1985; Cornish and Binns 1987; Campbell and Doeg 1989). Growth and development of tadpoles are directly affected by these factors (reviewed by Duellman and Trueb 1994). In contrast to species which are able to opportunistically breed in standing temporary water bodies, in which fluctuations in temperature, nutrient and oxygen levels may be extreme, the larvae of species that rely on streams are likely to be less tolerant of such changes because the normal stream environment is more stable. Recent investigations by O’Shaughnessy (1995 unpubl.) found that forest roads were the main sources of sediment input into streams in eastern Victoria. Increased sediment loads may reduce the availability or quality of oviposition sites through filling of interstitial spaces in the stream bed and blanketing substrates, resulting in increased mortality of eggs from predation, desiccation or flooding.This may be particularly important for species, such as L. booroolongensis and L. spenceri, which oviposit in rock crevices (Gillespie pers. obs.), or M. balbus which oviposit in the stream bed (Knowles et al. 1998 unpubl.). Larval survival may similarly be reduced
by limiting sheltering sites and food availability. Experiments conducted with tadpoles of L. spenceri found that deposited fine sediment can retard development rates of tadpoles, thus reducing their chances of survival (Gillespie 1997a unpubl.).
Agriculture and Grazing Land clearance for agriculture has adversely impacted upon the headwaters of some catchments in which declines have occurred, especially on the tablelands. Apart from direct effects of habitat removal, other down-stream impacts are likely, such as pesticide and fertiliser run-off, increased nutrient loads from grazing stock, and increased sediment loads from erosion. Forest grazing is also likely to have significant effects upon some populations through destruction or modification of riparian vegetation, increased nutrient levels, trampling of stream edges, and increased erosion damaging oviposition sites and larval habitats (Parris and Norton 1997; Knowles et al. 1998 unpubl.).
Hydrological Changes Flow regimes of some catchments have been modified for irrigation or hydro-electric power generation. In the southern highlands, Hunter and Gillespie (1999) found that L. lesueuri and other lotic frog species were encountered less frequently in streams below impoundments or aqueducts affecting stream flow. Increased stream flows during the warmer months are likely to have severe adverse impacts upon temperate riverine frog populations. Significant rises in water level and velocity during this period are likely to flush eggs and larvae downstream. Reduced temperatures of subsurface releases of water from dams during the summer months (SMEC 1997) are likely to inhibit larval growth and development.These reduced temperatures may also favour introduced trout (which may result in increased predation pressure). Sub-surface waters may also be anoxic and have different pH and higher concentrations of activated heavy metals (Doeg 1987; Ligon et al. 1995; Erskine 1996). Such alterations to water chemistry may be detrimental to tadpoles. Reduced peak flows resulting from dams may allow build up of sediments and colonisation of the stream channel by vegetation. Increased entrained sediments will reduce availability of oviposition sites and refugia for larvae by 125
blanketing the stream bed and in-filling of crevices between rocks. Encroachment of vegetation will reduce basking sites for adult frogs, and shading lower stream temperatures, which may reduce larval growth.
Climate Change Climatic change and increased levels of ultra-violet radiation have been suggested as causative agents in the declines of frog populations (Blaustein et al. 1995; Broomhall 1997). It is possible that upland populations of some species may be more susceptible to UV than lowland populations, which may explain some of the upland population declines in southeastern Australian temperate riverine species.The effects of increased UV might be expected to be most pronounced in upland southern populations of species which bask, such as L. spenceri and L. lesueuri (Gillespie and Hollis 1996). However, the current distributions and abundances of these species do not support this.
Disease Trenerry et al. (1994) and Laurance et al. (1996) suggested that a pathogen may be responsible for some frog declines in Australia. Criticism of this hypothesis was largely based on their inability to isolate and identify the pathogen (Hero and Gillespie 1997; Alford and Richards 1997). Ongoing studies of ill and dead frogs, including some collected during dramatic declines in north Queensland and Central America, now provide strong evidence that a chytrid fungus is the proximal cause of death (Berger et al. 1998, Berger et al. 1999). Several species of temperate riverine frog have been found infected with the chytrid fungus (Berger et al. 1998, Berger et al. 1999). Some populations of frogs infected with this fungus have declined while others have not. It is not known if the chytrid fungus is responsible for these declines, or is an indication of other environmental stress.The fungus may be a normally innocuous opportunistic natural pathogen, which is able to kill frogs under certain environmental conditions. Alternatively, this fungus may be a novel pathogen in populations of Australian frogs. Clearly there is a range of potentially threatening processes operating with varying impact throughout streams and across catchments of temperate south-eastern Australia.The interactions of these changes and disturbances within streams and catchments are likely to be complex. Assessment of these impacts upon frog populations is likely to be further complicated by interactions with other stochastic environmental processes.The persistence of populations in streams affected by timber harvesting or some other disturbance process does not mean that they have not been significantly adversely impacted upon. Many of these processes may serve only to reduce the ability of populations to cope with other catastrophes or environmental loads.The mediumterm and long-term influences of past and current management practices on the viability of riverine frog populations in south-eastern Australia cannot be dismissed. An applied research program is required, targeting a range of threatening processes, so that their importance can be judged.
126
FUTURE DIRECTIONS A more strategic approach is required for assessing the nature and causes of population declines within this guild throughout the region, rather than examining single species separately.The current approach to recovery planning is likely to result in duplication of research, adoption of inconsistent methods, and lack of coordination between separate programs which otherwise have similar objectives. Systematic surveys are required to ascertain the magnitude and nature of observed declines of most species. Surveys need appropriate sampling intensity, methods, scale and stratification. In conjunction with these surveys, information should be gathered on abundance of introduced fish, levels of habitat disturbance, both to streams and catchments, and other potentially threatening processes.This data should be quantifiable and suitable for appropriate multivariate analyses of distribution and abundance. Patterns of current distribution should also be examined at a landscape scale. In conjunction with these surveys, appropriate monitoring sites should be established to assess population dynamics and to detect future changes in abundance. Strategic monitoring should include ‘non-declining’ species, as there is no guarantee that these species are secure in the medium term. Monitoring sites should be stratified throughout the region to incorporate environmental variation across the ranges of all species. Studies are required to examine comparative ecology and habitat requirements of species and to ascertain factors which limit distribution and influence abundance. Information is required on aquatic and terrestrial habitat use, population structure and demography, and factors which influence recruitment and mortality. Both declining and non-declining species should be examined as differences between these taxa may be informative on causes of decline. Processes potentially responsible for observed declines require specific examination, such as the impact of introduced aquatic and terrestrial predators, impacts of various stream disturbances, disease and climate change.
ACKNOWLEDGMENTS We thank the Arthur Rylah Institute, Department of Natural Resources and Environment,Victoria, the Queensland Parks and Wildlife Service, NSW National Parks and Wildlife Service, and Environment Australia for financial support. Museum records and other distribution data were gladly supplied by the Australian Museum, the Queensland Museum, the Museum of Victoria, the South Australian Museum, the Australian National Wildlife Collection, the South Australian Environmental Protection Agency, the NSW Wildlife Atlas, and the Victorian Wildlife Atlas. R. Brown and H. Preece, Queensland Parks and Wildlife Service, provided geographical information system support.This manuscript was greatly improved by discussions with M. Mahony, M. Anstis, F. Lemckert, D. Ayers and J. Recsei. R. Loyn provided comments on the manuscript.The manuscript was greatly improved by referee reports from M. Littlejohn and W. Osborne, and comments from N. Clemann.
REFERENCES Alford, R. A. and Richards, S. J. (1997) Lack of evidence for epidemic disease as an agent in the catastrophic decline of Australian rainforest frogs. Cons. Biol., 11: 1026-29. Anstis, M. (1997) The glandular frog Litoria subglandulosa. Chapter 25 In H. Ehmann (Ed.).Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW, Sydney. pp. 213-221. Anstis, M. and Littlejohn, M. J. (1996) The breeding biology of Litoria subglandulosa and L. citropa (Anura: Hylidae), and a re-evaluation of their geographic distribution.Trans. Roy. Soc. S. A., 120: 83-99. Anstis, M., Alford, R. A. and Gillespie, G. R. (1998) Breeding biology of Litoria booroolongensis Moore and L. lesueuri Dumeril and Bibron (Anura: Hylidae).Trans Roy. Soc. S. A., 122:33-43. Barker, J. and Grigg, G. (1977) A Field Guide to Australian Frogs. Surrey Beatty, Chipping Norton, NSW. Barker, J., Grigg, G. and Tyler, M. J. (1995) A Field Guide to Australian Frogs (2nd Ed.). Surrey Beatty, Chipping Norton, NSW. Berger, L., Speare, R., Daszak, P., Earl Green, D., Cunningham, A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., McDonald, K. R., Hines, H. B., Lips, K. R., Marrantelli, G. and Parkes, H. (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rainforests of Australia and Central America. Proc. Nat. Acad. Sci., 95: 9031-9036. Berger, L., Speare, R., and Hyatt, A.D. (1999) Chytrid fungi and amphibian declines: overview, implications and future directions. Pp 23-33 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Blaustein, A. R., Edmond, B., Kiesecker, J. M., Beattie, J. J. and Hokit, D. G. (1995) Ambient ultraviolet radiation causes mortality in salamander eggs. Ecol. Appl., 5: 740-743. Boughton, W. C. (1970) Effects of Land Management on Quantity and Quality of Available Water. A Review. Water Resources Laboratories Report No. 120, University of NSW, Sydney. Bridgewater, P. B. (1987) The present Australian Environment — Terrestrial and Freshwater. In D. W. Walton and G. R. Dyne (Eds.). Fauna of Australia Volume 1A General Articles. Burea of Flora and Fauna, Australian Government, pp. 69-100. Broomhall, S. D. (1997) Comparative effects of ultra violet -B (UV-B) radiation on two sympatric species of Australian anurans, Litoria verreauxii alpina and Crinia signifera. Unpublished B. Sc. Honours thesis. Applied Ecology Research Group, University of Canberra. Bureau of Meteorology (1975) Climatic Atlas of Australia. Map Set 5. Rainfall. Australian Government Publishing Service, Canberra. Bureau of Meteorology (1989) Climate of Australia. Australian Government Publishing Service, Canberra. Campbell, I. C. and Doeg, I. J. (1989) The impact of timber production and harvesting on streams: a review. Aust. J. Mar. Freshw. Res., 40: 519-539.
Clinnick, P. F. (1985) Buffer strip management in forested areas: a review. Aust. For., 48: 34-45. CNR (1993) Statement of Resources, Uses and Values, East Gippsland Forest Management Area. Department of Conservation and Natural Resources, Orbost,Victoria. Copland, S. J. (1957) Presidential address. Australian tree frogs of the genus Hyla. Proc. Linn. Soc. N.S.W., 82: 9-108. Copland, S. J. (1962) Hyla phyllochrous Gunther (Amphibia) as an addition to the fauna of Victoria, with the description of a new race and a note on the name of the genus. Proc. Linn. Soc. N.S.W., 87: 137-40. Cornish, P. M. and Binns, D. (1987) Streamwater quality following logging and wildfire in a dry sclerophyll forest in south eastern Australia. For. Ecol. Manage., 22, 1-28. Covacevich, J.A. and McDonald, K.R. (1993) Distribution and conservation of frogs and reptiles of Queensland rainforests. Mem. Qld. Mus., 34: 189-199. Czechura, G.V. and Ingram, G. S. (1990) Taudactylus diurnus and the case of the disappearing frogs. Mem. Qld. Mus., 29: 361-365. Daly, G. (1995) Observations on the Tusked Frog Adelotus brevis (Anura: Myobatrachidae). Herpetofauna, 25: 32-35. Daly, G. (1996) Observations on the eastern owl frog Heleioporus australiacus (Anura: Myobatrachidae) in southeastern New South Wales. Herpetofauna, 26: 33-42. Daly, G. (1998). Review of the status and assessment of the habitat of the stuttering frog Mixophyes balbus (Anura: Myobatrachidae) on the south coast of New South Wales. Herpetofauna, 28: 2-11. Doeg,T. J. (1987) Response of the macroinvertebrate fauna of the Mitta Mitta River,Victoria, to the construction and operation of the Dartmouth Dam. 2. Irrigation release. Occ. Pap. Mus.Vic., 1: 101-128. Duellman and Trueb (1994) Biology of Amphibians. (2nd Ed.). John Hopkins University Press: Maryland, USA Ehmann, H., (1997) (ed.) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW Inc., Sydney South. Erskine, W. D. (1996) Downstream hydrogeomorphic impacts of the Eildon Reservior on the mid-Goulburn River, Victoria. Proc. Roy. Soc.Vic., 108: 1-15. Flinn, D. W., Farrell, P. W., Stuart, H.T. L., Leitch, C. J. and Hopmans, P. (1983) The effects of fire in eucalypt forest on soils, nutrient cycling, tree growth and catchment hydrology: a review with particular reference to fuel reduction burning. In Proceedings of the Symposium: Fighting Fire with Fire. Graduate School of Environmental Studies, Monash University, Melbourne, pp. 146-85. Gillespie, G. R. (1990) Distribution, habitat and conservation status of the giant burrowing frog, Heleioporus australiacus (Anura: Myobatrachidae), in Victoria.Vict. Nat., 107: 144-153. Gillespie, G. R. (1997a, unpublished) The Biology of the Spotted Tree Frog (Litoria spenceri) and Examination of Factors Responsible for Population Declines. Arthur Rylah Institute, Department of Natural Resources and Environment, Heidelberg,Victoria. Unpublished Report to the Biodiversity Group, Environment Australia, Canberra.
127
Gillespie, G. R. (1997b, unpublished) Survey Design and Management Prescriptions for the Giant Burrowing Frog (Heleioporus australiacus) and the Stuttering Frog (Mixophyes balbus). Arthur Rylah Institute, Department of Natural Resources and Environment, Heidelberg,Victoria. Unpublished report to the New South Wales National Parks and Wildlife Service, Queanbeyan. Gillespie, G. R. (1998, unpublished) Spotted Tree Frog Recovery Program — Annual Report to April 1998. Unpublished report to the Biodiversity Group, Environment Australia, Canberra. Gillespie, G. R. (1999) The booroolong frog Litoria booroolongensis Moore (Anura: Hylidae): an addition to the frog fauna of Victoria.Vict. Nat., 116: 112-114. Gillespie, G. R. (in review) The role of the introduced trout in the decline of the spotted tree frog Litoria spenceri. Cons. Biol. Gillespie, G.R. and Hero, J.-M. (1999) Potential Impacts of Introduced Fish and Fish Translocations on Australian Amphibians. Pp 131-144in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Gillespie, G. R. and Hollis, G. J. (1996) Distribution and habitat of the spotted tree frog Litoria spenceri Dubois (Anura: Hylidae), and an assessment of potential causes of population declines. Wildl. Res., 23: 49 — 75. Gillespie, G. R. and Osborne, W. S. (1994) Update on the status of the spotted tree frog (Litoria spenceri) in the Australian Capital Territory.Vict. Nat., 111: 182-183. Heatwole, H., De Bavay, J., Webber, P. and Webb, G. (1995) Faunal survey of New England. IV.The frogs. Mem. Qld. Mus., 38: 229-49. Hero, J.-M. and Gillespie, G. R. (1993) The tadpole of Litoria phyllochroa (Anura, Hylidae). Proc. Roy. Soc.Vic., 105: 31-38. Hero, J.-M. and Gillespie, G. R. (1997) Epidemic disease and amphibian declines in Australia. Cons. Biol., 11: 1023-25. Hines, H.B., Mahony, M.J., and McDonald, K.R. (1999) An assessment of frog declines in wet subtropical Australia. Pp 44-63 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Hollis G. J. (1995) Reassessment of the distribution, abundance and habitat of the baw baw frog Philoria frosti Spencer: Preliminary Findings.Vict. Nat., 112: 190-201. Holloway, S. E. (1997) Survey protocols for the streambreeding frogs of far East Gippsland: the implication of habitat modelling and an assessment of techniques. Unpublished M. Sc.Thesis, University of Canberra, Canberra. Holloway, S. E. and Osborne, W. S. (1996 unpublished) The Stream-breeding Frog Fauna Associated with Montane Forests in Far East Gippsland. Unpublished Report to the Department of Natural Resources and Environment, Orbost,Victoria. Hughes, J. M. R. and James, B. (1989) A hydrological regionalization of streams in Victoria, Australia, with implications for stream ecology. Aust. J. Mar. Freshwater Res., 40: 303-326. Hunter, D. and Gillespie, G. R. (1999) The distribution, abundance and conservation status of riverine frogs in Kosciuszko National Park. Aust. Zool., 31: 198-209 128
Hutchinson, M. N. and Maxson, L. R. (1987) Phylogenetic relationships among Australian tree frogs (Anura: Hylidae: Pelodryadinae): an immunological approach. Aust. J. Zool., 35: 61-74. Ingram, G. J and McDonald, K. R. (1993) An update on the decline of Queensland’s frogs, In D. Lunney and D. Ayers (Eds.) Herpetofauna in Australia a Diverse Discipline.Trans. Roy. Zool. Soc. NSW, Mosman, pp. 297-303. Katsikaros, K. and Shine, R. (1997) Sexual dimorphism in the tusked frog, Adelotus brevis (Anura, Myobatrachidae) — the roles of natural and sexual selection. Biol. J. Linn. Soc., 60: 39-51. Knowles, R., Hines, H.B.,Thumm, K. Mahony, M., and Cunningham, M. (1998 unpublished) Oviposition of the barred-frogs (Mixophyes species) in southeastern Australia with implications for management. Unpublished abstract of a talk presented to the Australian Society of Herpetologists meeting, February 1998. Langford, K. J. and O’Shaughnessy, P. J. (Eds.) (1980) Water Supply Catchment Research. Second Progress Report, Coranderrk. M.M.B.W. Report W0010. Melbourne and Metropolitan Board of Works, Melbourne. Laurance, W. F. K., McDonald, K. R. and Speare, R. (1996) Epidemic disease and the catastrophic decline of Australian rain forest frogs. Cons. Biol., 10: 406-13. LCC (1985) Report on the East Gippsland Area Review. Land Conservation Council, Melbourne. Lee, A. K. (1967) Taxonomy, ecology and evolution of the genus Heleioporus (Gray) (Leptodactylidae). Aust. J. Zool., 15: 367-439. Lemckert, F. (1998) A survey for threatened herpetofauna of the south-west slopes of New South Wales. Aust. Zool., 30: 492-500. Lemckert, F., Potter, M., Smith, B. and Brues,T. (1997) Recent records of the southern barred frog (Mixophyes balbus) from the south coast of NSW. Herpetofauna, 27: 43-45. Ligon, F. K., Dietrich, W. E. and Trush, W. J. (1995) Downstream ecological effects of dams. Bioscience, 45: 183-192. Littlejohn, M. J. (1967) Patterns of zoogeography and speciation in south-eastern Australia Amphibia. Ch. 6 In A. H. Weatherley (Ed.) Australian Inland Waters and their Fauna. Australian National; University Press, Canberra, pp. 150-174. Littlejohn, M. J. (1981) The amphibia of mesic southern Australia: a zoological perspective. In A. Keast (Ed.) Ecological Biogeography of Australia. Dr W. Junk,The Hague, pp. 1304-1330. Littlejohn, M. J., Loftus-Hills, J. J., Martin, A. A. and Watson, G. F. (1972) Amphibian fauna of Victoria. Confirmation of records of Litoria (= Hyla) citropa (Tschudi) in Gippsland. Vict. Nat., 89: 51-54. Littlejohn, M. J. and Martin, A. A. (1967) The rediscovery of Heleioporus australiacus (Shaw) (Anura: Leptodactylidae) in eastern Victoria. Proc. Roy. Soc.Vic., 80: 31-36. Littlejohn, M. J., Davies, M., Roberts, J. D. and Watson, W. F. (1993) The Myobatrachidae. In C. J. Glasby, G. J. B. Ross and P. L. Beesley (Eds.). Fauna of Australia.Vol. 2A Amphibia & Reptilia. Australian Government Publishing Service, Canberra, pp. 41-57.
Lips, K. R. (1998) Decline of a tropical montane amphibian fauna. Cons. Biol., 12: 106-116. Lunney, D. and Barker, J. (1986) Survey of the reptiles and amphibians of the coastal forests near Bega, New South Wales. Aust. Zool., 22: 1-9. Mahony, M. J. (1993) The status of frogs in the Watagan Mountains area, the central coast of New South Wales. In D. Lunney and D. Ayers (Eds.) Herpetology in Australia a Diverse Discipline.Trans. Roy. Soc. NSW, Mossman, pp. 257-264. Mahony, M. J. (1996, unpublished) Survey of the Distribution and Abundance of Declining Frog Species in Northern New South Wales. Unpublished report for the New South Wales National Parks and Wildlife Service, Hurstville. Mahony, M., Knowles, R. and Pattinson, L. (1997a) Peppered tree frog Litoria piperata. Chapter 22 In H. Ehmann (Ed.) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW, Sydney, pp. 189-194. Mahony, M., Knowles, R. and Pattinson, L. (1997b) Stuttering frog Mixophyes balbus. Chapter 4 In H. Ehmann (Ed.) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW, Sydney, pp. 65-71. Martin, A. A. (1972) Studies in Australian amphibia III.The Limnodynastes dorsalis complex (Anura: Leptodactylidae). Aust. J. Zool., 20: 165-211. Martin, A. A., Littlejohn, M. J. and Rawlinson, P. A. (1966) A key to the anuran eggs of the Melbourne area, and an addition to the anuran fauna.Vict. Nat., 83: 312-15. McDonald, K. R. (1990) Rheobatrachus Liem and Taudactylus Straughan and Lee (Anura: Leptodactylidae) in Eungella National Park, Queensland: distribution and decline.Trans. Roy. Soc. S. A., 114: 187-194. McDonald, K.R. and Alford, R.A. (1999) A review of declining frogs in northern Queensland. Pp 14-22in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. McDonald, K. R. and Davies, M. (1990) Morphology and biology of the Australian tree frog Litoria pearsoniana (Copland) (Anura: Hylidae).Trans. Roy. Soc. S. A., 114: 145-56. Norris, K. C., Mansergh, C. A., Ahern, L. D., Belcher, C. A., Temby, I. D. and Walsh, N. G. (1983) Vertebrate Fauna of the Gippsland Lakes Catchment Victoria. Occasional Paper Series No. 1, Fisheries and Wildlife Division,Victoria, Ministry for Conservation, Melbourne. NRE (1999) Threatened Vertebrate Fauna in Victoria — 1999. Department of Natural Resources and Environment, Victoria. NSW NPWS (1994) Fauna of North-east NSW Forests. North-east Forests Biodiversity Study Report No. 3, Unpublished Report, NSW National Parks and Wildlife Service, Hurstville.
Odendaal, F. J. and Bull, C. M. (1982) A parapatric boundary between Ranidella signifera and R. riparia (Anura: Leptodactylidae) in South Australia. Aust. J. Zool., 30: 49-57. Odendaal, F. J. and Bull, C. M. (1983) Water movements, tadpole competition and limits to the distribution of the frogs Ranidella riparia and R. signifera. Oecologia, 57: 361-67. Odendaal, F. J., Bull, C. M. and Nias, R. C. (1982) Habitat selection in tadpoles of Ranidella signifera and R. riparia (Anura: Leptodactylidae). Oecologia, 52: 411-14. Odendaal, F. J., Bull, C. M. and Richards, S. J. (1984) Interactions during feeding between tadpoles of Ranidella signifera and R. riparia. J. Herp., 18: 489-92. Odendaal, F. J., Bull, C. M. and Telford, S. R. (1986) Influence of acoustic environment on the distribution of the frog Ranidella riparia. Anim. Behav., 34: 1836-43. Osborne, W. S. (1989) Distribution, relative abundance and conservation status of corroboree frogs, Psuedophryne corroboree Moore (Anura: Myobatrachidae). Aust. Wildl. Res, 16: 537-47. Osborne, W. S. (1990) Declining frog populations and extinctions in the Canberra region. Bogong, 11: 4-7. Osborne, W.S., Hunter, D.A. and Hollis, G.J. (1999) Population 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. O’Shaughnessy, P. and Associates (1995, unpublished) Water Quality Protection Measures for the Conservation of the Spotted Tree Frog. Unpublished report to the Forest Service,Victorian Department of Natural Resources, Melbourne. Parris, K. and Norton,T. (1997) The significance of State Forests for conservation of Litoria pearsoniana (Copland) and associated amphibians. In P. Hale and D. Lamb (Eds.) Conservation Outside Nature Reserves. Centre for Conservation Biology,The University of Queensland, St Lucia, pp. 521-526. Pechmann, J. H., Scott, D. E., Semlitsch, R. D., Caldwell, J. P.,Vitt, L. J. and Whifield Gibbons, J. (1991) Declining amphibian populations: the problem of separating human impacts from natural fluctuations. Science, 253: 892-95. Pechmann, H. K. and Wilbur, H. M. (1994) Putting declining amphibian populations in perspective: natural fluctuations and human impacts. Herpetologica, 50: 65-84. Recsei, J. (1997) The giant burrowing frog Heleioporus australiacus. Chapter 3 In H. Ehmann (Ed.) Threatened Frogs of New South Wales: Habitats, Status and Conservation. Frog and Tadpole Study Group of NSW, Sydney, pp. 55-64. Richards, S. J. and Alford, R. A. (1992) Nest construction by an Australian rainforest frog of the Litoria lesueuri complex (Anura: Hylidae). Copeia, 1992: 1120-23. Richards, S. J., McDonald, K. R. and Alford, R. A. (1993) Declines in population of Australia’s endemic tropical rainforest frogs. Pac. Cons. Biol., 1: 66-77.
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 Tree Frog Recovery Plan. Prepared for the Biodiversity Group, Environment Australia, by Department of Natural Resources and Environment, Melbourne,Victoria. SFNSW (1995). Queanbeyan and Badja Management Area Environmental Impact Statement. State Forests of New South Wales. SMEC (1997) Snowy Mountains Data Compilation. Prepared by SMEC Australia Ltd for the Department of Land and Water Conservation, Sydney. Straughan, I.R. (1968). A taxonomic review of the genus Mixophyes, (Anura, Leptodactylidae), 1968. Proc. Linn. Soc. NSW. 93:52-59 Trenerry, M. P., Laurance, W. F. and McDonald, K. R. (1994) Further evidence for the precipitous decline of endemic rainforest frogs in tropical Australia. Pac. Cons. Biol., 1, 150-53. Tyler, M. J. (1991) Declining amphibians — a global phenomenon? An Australian perspective. Alytes, 9, 43-45. Tyler, M. J. (1997) The Action Plan For Australian Frogs. Biodiversity Group, Environment Australia, Canberra. Tyler, M. J. and Anstis, M. (1975) Taxonomy and biology of frogs of the Litoria citropa complex (Anura: Hylidae). Rec. S. A. Mus., 17: 41-50. Tyler, M. J. and Davies, M. (1978) Species groups within the Australopapuan hylid genus Litoria Tschudi. Aust. J. Zool. Suppl. Ser., 63: 1-79. Tyler, M. J. and Davies, M. (1985) A new species of Litoria (Anura: Hylidae) from New South Wales, Australia. Copeia, 1985:145-49.
130
Walter, H. and Lieth, H. (1967) Klimadiagramm — Weltatlas. Jena, Fisher (Maps only). Watson, G. F. and Littlejohn, M. J. (1985) Patterns of distribution, speciation and vicariance biogeography of southeastern Australian amphibians. In G. Grigg, R. Shine and H. Ehmann (Eds) Biology of Australasian Frogs and Reptiles. Royal Zoological Society of New South Wales, Sydney, pp. 91-7. Watson, G. F., Littlejohn, M. J., Hero, J-M. and Robertson, P. (1991) Conservation Status, Ecology and Management of the Spotted Tree Frog (Litoria spenceri). Arthur Rylah Institute for Environmental Research Technical Report Series No. 116. Department of Conservation and Environment, Melbourne. Webb, G. A. (1987) A note on the distribution and diet of the giant burrowing frog, Heleioporus australiacus (Shaw and Nodder 1795)(Anura: Myobatrachidae). Herpetofauna, 17(2): 20-22. Webb, G. A. (1991) A survey of the reptiles and amphibians of Bondi State Forests and surrounding areas, near Bombala, New South Wales. Aust. Zool., 27: 14-19. Wellington, R. and Wells, R. (1994). Fauna Survey of the Morisset District Central Coast, New South Wales, Reptiles and Amphibians. Unpublished Report to State Forests of New South Wales. White, A. W. and Pyke, G. H. (1996) Distribution and conservation status of the green and golden bell frog Litoria aurea in New South Wales. Aust. Zool. 30: 177-89.
Potential Impacts of Introduced Fish and Fish Translocations on Australian Amphibians Graeme Gillespie1 and Jean-Marc Hero2
ABSTRACT This review examines the potential impact of introduced fish on amphibians, with particular emphasis on Australian freshwater systems. Firstly,
subject to thorough research. Many Australian amphibian assemblages, including several threatened species, are potentially threatened by a variety of introduced fish species. Future research priorities and
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 introduced fish on amphibians both overseas and within Australia is then reviewed. Evidence in the literature strongly suggests that introduction of exotic fish or translocation of native species could have enormous impacts on the amphibian assemblages of Australian freshwater systems. Introduced fish have been implicated in the decline of several anuran species, though few cases have been
INTRODUCTION The reported declines of many amphibian populations both in Australia and around the world are now recognised as a very real phenomenon.These declines pose a serious threat to global amphibian diversity, and may result from recent global environmental change associated with human activities. The cause(s) of many species declines, particularly in apparently pristine tropical forests of Central America and Australia, remain obscure (but see Lips 1998). However, in many cases one or other anthropogenic impacts, commonly identified as key threatening processes in the decline or 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 declines.These include habitat destruction (Laan and Verboom 1990; Johnson 1992; Wardell-Johnson and Roberts 1991; Gillespie and Hollis 1996; Dubuis 1997; Waldick 1997), pollution (Bishop 1992; Bidwell and Gorrie 1995; Bertram and Berrill 1997), over-exploitation (Jennings and Hayes 1985) and introduction of exotic predators (Orchard 1992; Bradford et al. 1993; Lannoo et al. 1994). In Australia, anthropogenic impacts on amphibian populations are only now beginning to be appreciated.The relative significance of various potentially threatening processes to the maintenance of many amphibian communities is poorly understood.This is a reflection in part of the inherent difficulties associated with studying amphibian population dynamics, the high diversity of amphibian assemblages present in Australia, and the small number of ecologists and amount of resources available for studying them. Decisions of research and management priorities and allocation of resources must therefore be based upon careful assessments of current knowledge about the importance of all potentially threatening processes, from both within Australia and overseas. After over-exploitation and habitat destruction, introduced predators have been identified as the main cause of mammal and bird species extinctions in modern times, particularly in Australasia (Meffe and Carroll 1994; Leakey and Lewin 1995). It is likely that introduced predators may also play a significant role in the decline of some amphibian species.The following review examines the potential impact of introduced fish on amphibians, with particular emphasis on Australian freshwater systems. Firstly, we examine the ecological relationships between fish predators and their amphibian prey, and how these may be altered when non-native fish are introduced. We then review current knowledge and research on the impacts of introduced fish on amphibians both overseas and within Australia. Finally, we outline future research priorities and guidelines for examining the impact of introduced fish on Australian amphibians, and identify key management objectives.
SIGNIFICANCE OF FISH PREDATORS IN AMPHIBIAN COMMUNITIES Predation is considered to be a major factor regulating the distribution of amphibian larvae (e.g. Calef 1973, Heyer et al. 1975, Duellman 1978, Scott and Limerick 1983, Smith 1983, Woodward 1983, Wilbur 1984, Hayes and Jennings 1986, Kats et al. 1988). Heyer et al. (1975) suggested that predation by aquatic predators, primarily fish, was the most important biotic factor influencing the temporal and spatial composition of tadpole communities.The combined direct and indirect effects of fish predators on the local distribution of individual species of tadpole consequently influence local and regional amphibian assemblage structure. Recent studies have demonstrated the importance of fish predators in determining tadpole species-composition (species present) and tadpole species richness (number of species) in temperate (Hecnar and M’Closkey 1997) and tropical systems (Fickling 1995; Hero et al. 1988). In this section, we examine the ecological relationships between fish predators and their amphibian prey and how these may be altered when non-native fish are introduced.
132
Fish may directly impact amphibian species by predation on larvae (Macan 1966; Heyer et al. 1975; Sih et al. 1988) or eggs (Grubb 1972). Consequently fish predators are capable of eliminating larval amphibian species from some habitats (Tyler 1963, Macan 1974, Petranka 1983). In a comprehensive study by Petranka (1983) the larvae of the salamander Ambystoma texanum were found to be restricted to the fish-free, upper portions of breeding streams and this was attributed to predation on larvae by endemic species of fish. Kats et al. (1988) identified fish predation as a primary factor influencing marked differences between larval amphibian assemblages in ephemeral and permanent aquatic habitats. Similarly, Fickling (1995) found that Litoria nannotis and L. rheocola were restricted to streams without predatory fish in the Tully gorge, northern Australia.
TADPOLE SURVIVAL STRATEGIES Amphibian larvae can physically evade fish predators through spatial or temporal avoidance (Petranka 1983, Bradford 1989, Holomuzki 1995, Hecnar and M’Closkey 1997, Hero et al. 1998). Many species of amphibian breed only in temporary water bodies in which fish are absent. Recent studies have shown that females of some species of amphibian choose oviposition sites in waterbodies without fish (Resetarits and Wilbur 1989; Kats and Sih 1992; Bronmark and Edenhamn 1994; Hopey and Petranka 1994; Holomuzki 1995). Alternatively, amphibians can avoid fish predators by reproducing in a waterbody at times when fish predators are absent (e.g. streamside ponds that are isolated from the stream at some times of the year). The larvae of many amphibian species occur in habitats containing predatory fish, such as permanent lakes and streams. Survival or anti-predator strategies allow these species to coexist with fish predators.These strategies include cryptic colouration (Wasserug 1971), behavioural responses 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 al. 1978; Kruse and Stone 1984; Kats et al. 1988; Werner and McPeek 1994). In contrast to species which typically occur in fish-free habitats, larvae of species which coexist with fish predators may possess one or a combination of these survival traits (Kats et al. 1988). Many amphibian larvae which coexist with predacious fish are unpalatable or noxious (Liem 1961;Voris and Bacon 1966; Wasserug 1971; Brodie et al. 1978; Walters 1975; Kruse and Stone 1984; Kats et al. 1988; Hero 1991; Werner and McPeek 1994). Amphibian larvae which do not respond behaviourally to predatory fish are typically toxic or unpalatable to fish (Voris and Bacon 1966; Kruse and Fransis 1977; Kruse and Stone 1984; Kats et al. 1988). Antipredator strategies used by tadpole species against invertebrate predators, such as immobility (Azevedo-Ramos et al. 1992; Chovanec 1992; Werner and McPeek 1994) are not usually effective against fish predators that use visual cues (Hero 1991; Werner and McPeek 1994).Therefore the distribution of each larval amphibian species is related to the survival strategies it possesses and is strongly influenced by the distribution of predatory fish.
PREDICTED IMPACTS OF INTRODUCED PREDATORS
OVERSEAS EVIDENCE FOR IMPACTS OF INTRODUCED FISH ON AMPHIBIANS
Predator-prey relationships are maintained by a constant evolution of both the predator to capture and the prey to avoid capture, commonly described as the “evolutionary arms race” (Dawkins and Krebs 1979). Survival strategies tend to be predator specific and are unlikely to be effective against all predators. For example, female amphibians may not be able to recognise the chemical cues produced by introduced fish species and may inadvertently oviposit in a water body with exotic fish predators, resulting in levels of predation that preclude survival of the species. Palatability of a species of tadpole can differ among different species of fish predator (Hero 1991; Holomuzki 1995). Hence, a species of tadpole may be unpalatable to the native fish predators with which it coexists, but may not be unpalatable to a novel fish predator. Prey species may not identify introduced fish as predators and hence fail to use the appropriate survival strategies (temporal or spatial isolation), or the species of tadpole may not have the necessary antipredator defences that allow them to coexist with introduced fish species.The introduction of an exotic predator is therefore likely to disrupt the arms race in favour of the predator.
The consequences of introducing fish into breeding habitats for amphibians have been well documented overseas. A number of studies in Europe, North and South America have implicated or demonstrated that introductions of predatory fish are responsible for the decline or extinction of some amphibian species.
Fish predators can also influence tadpole assemblages indirectly by consuming invertebrate tadpole-predators, such as dragonfly naiads and predacious diving beetles (Wilbur and Fauth 1990; Werner and McPeek 1994; Hero et al. 1998). Along an environmental gradient Werner and McPeek (1994) found that only Rana catesbiana tadpoles were found in waterbodies with fish predators while R. clamitans was found primarily in fishless ponds with high densities of invertebrate predators. Furthermore, the presence of fish predators can reduce densities of some species of tadpole, and this may release other species from competition, thus enhancing their survival (Morin 1986; Werner and McPeek 1994).This predator mediated release from competition may result in a shift in species composition from species that are competitively dominant to species that are competitively inferior but have the survival strategies that allow them to coexist with fish. The general pattern observed in natural systems is that species of tadpole that are vulnerable to predation by invertebrate predators survive in waterbodies with fish (where the density of invertebrate predators is low due to predation by fish), and species of tadpole that are vulnerable to predation by fish survive in waterbodies where predacious fish are absent (Hecnar and M’Closkey 1997; Hero et al. 1988).The introduction of predacious fish species will potentially result in the elimination of some tadpole species and a shift in the species composition to those species which have the survival-strategies that allow them to coexist with the introduced predator. Theory therefore predicts that the introduction of exotic fish to aquatic systems may lead to the elimination of some species of tadpole, resulting in changes in the species composition of natural tadpole assemblages.These changes may be extremely detrimental to the long term survival of some species, undermine amphibian communities and disrupt natural aquatic systems.
Brönmark and Edenhamn (1994) suggest that, in Europe, the widespread introduction of various fish species into farm dams and ponds has contributed to the decline of Hyla arborea.They found that H. arborea in Sweden was predominantly restricted to ponds in which fish had not been introduced. No reproduction was recorded during a three year period in ponds containing either pike (Esox lucius), perch (Perca fluviatilis), roach (Rutilus rutilus), Crucian carp (Carassius carassius), rudd (Scardinius erythrophthalmus) or tench (Tinca tinca). Pike (Esox lucius), perch, and Crucian carp have been shown in laboratory studies to readily feed on H. arborea tadpoles and metamorphs (Brönmark unpublished, in Brönmark and Edenhamn 1994). Macan (1966) reported a dramatic decrease in numbers of bufonid and ranid tadpole species following the introduction of brown trout (Salmo trutta) into a British tarn. Braña et al. (1996) found that amphibian species’ numbers and amphibian abundance were significantly lower in lakes of northern Spain containing introduced fish: brown trout, rainbow trout (Oncorhynchus mykiss), tench, roach and European minnow (Phoxinus phoxinus).They concluded that the presence of these introduced species was responsible for the almost complete disablement of large permanent waterbodies for amphibian reproduction and subsequent decline of amphibian species in the region. In North America the introduction of salmonids into previously fishless habitats has impacted upon numerous amphibian species. Burger (1950) reported the wide scale elimination of tiger salamander (Ambystoma tigrinum) larvae from ponds in Colorado following stocking with trout. Fish introductions, primarily trout, have been suggested as an important factor contributing to the decline of ranid frog species (Hayes and Jennings 1986; Liss and Larson 1991; Hecnar and M’Closky 1996). Several species of introduced salmonids have profoundly affected the distribution of the Mountain Yellow-legged Frog (Rana mucosa) within the past century by eliminating the species from nearly all waters where fish have been introduced (Grinnell and Storer 1924; Bradford 1989; Bradford et al. 1993). Hayes and Jennings (1986) noted that the abundance of endemic Rana species in California was inversely correlated with densities of introduced fish species, primarily trout.Tyler et al. (1998) demonstrated that larval salamanders (Ambystoma macrodactylum) were found in much higher densities in alpine lakes without fish than in lakes that contained introduced trout populations. In Canada Liss and Larson (1991) reported the decline of amphibian species in naturally fishless lakes after stocking with trout. Hecnar and M’Closkey (1996) concluded that the presence of introduced predatory fish was responsible for the decline of amphibian species in south-western Ontario.They 133
found that amphibian species richness was significantly lower in ponds containing introduced predatory fish. However, those amphibian species with either large larval body size or large clutch size were less adversely affected than others, and occurred more frequently with predatory fish. The introduction or translocation of other species has also been implicated in the decline of some amphibian species in North America. Introduced mosquitofish (Gambusia spp.) have been identified as the most likely cause of localised declines of Californian newts (Taricha torosa) in southern California (Gamradt and Kats 1996). Petranka (1983) documented decimation of small-mouthed salamander (Ambystoma texanum) larvae in local pools in streams following colonisation by Green Sunfish (Lepomis cyanellus), and Sexton and Phillips (1986) noted a dramatic reduction in species richness after the introduction of this species. Semlitsch (1983) reported almost complete mortality of Rana esculenta tadpoles following the addition of Pike to experimental ponds. Declines of some amphibians in South America have also been attributed to introduced fish.The introduction of various fish species: salmonids, European carp (Cyprinus carpio), Odonthestes bonariensis and catfish (Ictalurus spp.), is thought to be a principal factor leading to the decline of
amphibians in southern Chile (Formas 1995). Introduced salmonids are also thought to be responsible for the extinction of several Atelopus species in Costa Rica (Pough et al. 1998). In most of the above cases, fish introductions have occurred for recreational purposes. Hence, the frequent reports involving trout species, which have been widely introduced in lakes and streams throughout both hemispheres due to their popularity with anglers. It should be emphasized that translocation of native fish species into aquatic systems that have not previously contained the species could have similar impacts on amphibian fauna.
INTRODUCED FISH IN AUSTRALIA The list of fish introduced into Australia is extensive (Table 1). At least 24 exotic species have established self-sustaining populations in Australian freshwater systems to date. In addition, several native species have been translocated into aquatic systems in which they did not naturally occur.These include Murray cod (Maccullochella peelii), golden perch (Macquaria ambigua), Macquarie perch (M. australasica), bass (M. novemaculeata), barramundi (Lates calcarifer), catfish (Tandanus tandanus) and rainbow fish (Melanotaenia spp.) (Raadik, Arthur Rylah Institute (ARI),Victoria, pers. comm.).
TABLE 1: List of exotic fish which have established populations in Australian waters.
Species
Salmonidae Rainbow Trout Chinook Salmon Brook Trout Atlantic Salmon Brown Trout Cyprinidae Goldfish European Carp Rosy Barb Roach Tench Percidae Redfin Perch Poecilidae gambusia (Mosquito Fish) One-spot Livebearer Sailfin Molly Guppy Swordtail Platy Cyprinodontidae American Flag Fish Cobitidae Oriental Weatherloach Cichlidae Blue Acara Convict Cichlid Mozambique Mouthbrooder Black Mangrove Cichlid Zilles Cichlid
Origin
Occurrence in Australia
Oncorhynchus mykiss O. tshawytscha Salmo fontinalis S. salar S. trutta
Nth. America Nth. America, N.E. Asia Nth. America Europe, Nth America Europe,West Asia
NSW;Vic;Tas; s.w.WA; s.e. SA s.w.Vic s.e. NSW;Tas s.e. NSW;Vic;Tas NSW;Vic;Tas; s.w.WA; s.e. SA
Carassius auratus Cyprinus carpio Puntius conchonius * Rutilus rutilus Tinca tinca
East Asia Europe Asia Europe Europe, Central Asia
s.e. Qld; NSW;Vic; SA; s.w.WA s.e. Qld; NSW;Vic; SA; s.w.WA s.e.Qld Vic sth. NSW;Vic;Tas
Perca fluviatilis
Europe, Nth. Asia
NSW;Vic; east SA;WA;Tas
Gambusia holbrooki Phalloceros caudimaculatus Poecilia latipinna P. reticulata Xiphophorus helleri X. maculatus
Nth. America Sth. America Cent. America Cent. America, Caribbean Cent. America Cent. America
Qld; NSW;Vic; SA;WA;Tas s.w.WA s.e. Qld s.e. Qld s.e. Qld s.e. Qld
Jordanella floridae
Nth. America
n.e. Qld
Misgurnus anguillicaudatus
East Asia
s.e. Qld; s.e. NSW;Vic
Aequidens pulcher Heros nigrofasciata Oreochromis mossambicus Tilapia mariae T. zillii *
Cent. America Cent. America East Africa West Africa Africa
e. Qld Vic e. Qld, s.w.WA Vic 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 fishing purposes, or through releases of species from the aquarium trade.The one exception is the eastern gambusia or mosquito fish (Gambusia holbrooki), which was misguidedly introduced to control mosquitoes (Myers 1965). Salmonids have been widely introduced into streams and lakes for recreational fishing.The brown trout and rainbow trout are the most successful and widespread species.These occur in mainland streams along the Dividing Range from Victoria up to northern NSW, in the Adelaide Hills and in south-western Western Australia (Allen 1982, 1989; McDowall 1996). In south eastern Australia, brown trout are abundant in all upland streams and are only excluded from a few small tributaries where waterfalls have blocked their upstream passage (Cadwallader and Backhouse 1982; Faragher and Harris 1993; McDowall 1996). Rainbow trout have a more patchy distribution but are in high abundance in many small upland water courses (McDowall 1996;Victorian Fish Database, ARI,Victoria). Both species also occur in many lakes and reservoirs in these regions. Stocking of lakes and streams occurs extensively in NSW, and several lakes are stocked in Victoria by Fisheries authorities.These species are also stocked in farm dams. Brook trout have established in Tasmania in lakes of the Tyndall Ranges (McDowall 1996), the Clarence Lagoon on the Central Plateau (Swain, University of Tasmania, pers. comm.), and are currently restricted to one stream on the mainland in Kosciuszko National Park, NSW (Harris, NSW Fisheries, pers. comm.). Chinook salmon are currently restricted to several lakes in south-western Victoria, maintained by stocking (McDowall 1996). Atlantic salmon are stocked in lakes and reservoirs in south-eastern NSW, central Victoria and Tasmania (McDowall 1996).The species has escaped from hatcheries into the Rubicon and Latrobe Rivers in Victoria (McDowall 1996). Goldfish and European carp were originally introduced as ornamental fish and have spread throughout the MurrayDarling system and other inland and coastal waterways in south-eastern Australia (Cadwallader and Backhouse 1982; Faragher and Harris 1993; McDowall 1996).They also occur in south-western Western Australia (Allen 1982, 1989) and Tasmania (Swain pers. comm.).Tench and roach were introduced in the late 1800’s into lakes and rivers in Victoria for fishing and have both spread. Movements of roach have also been recorded up rivers feeding lakes, such as the Howqua and Big Rivers in the catchment of Lake Eildon, Victoria (Victorian Fish Database, ARI,Victoria). Introductions of roach still occur illegally for caurse fishing.Tench have also been stocked for fishing in lakes and reservoirs in southern NSW and Tasmania. Redfin have been widely introduced throughout the MurrayDarling River system, lakes and farm dams in south-eastern mainland Australia,Tasmania and south-western Western Australia as a popular angling species (Allen 1982, 1989; Rowland 1989; McDowall 1996).This species has also penetrated up major tributaries of some lakes, in some cases considerable distances, such as Eildon and Glenmaggie in Victoria (Victorian Fish database, ARI,Victoria).
Gambusia are widespread throughout south-eastern Australia, including the Murray-Darling system, and extend up the east coast as far as Townsville. It also occurs in south-western WA and some water courses near Alice Springs (Allen 1982, 1989; McDowall 1996).The species is exceptionally hardy and is able to tolerate an extremely broad range of environmental conditions (McKay 1984). It is a highly invasive species inhabiting marshes, lakes and dams, slow-flowing streams and associated billabongs and aqueducts. It is most abundant in modified habitats and areas near human settlement (Allen 1989; McDowall 1996). The remaining species have all originated from the aquarium trade (McKay 1984; Allen 1989; Ryan 1995; McDowall 1996). Only three of these have so far established extensive distributions in the wild.The oriental weatherloach (Misgurnus anguillicaudatus) occurs in streams along the east coast of NSW and several south-flowing catchments in Victoria, such as the Yarra and Latrobe Rivers. It also occurs inland in south-eastern Australia, in the Murrumbidgee, Ovens and Murray Rivers (McDowell 1996;Victorian Fish database, ARI,Victoria).The Mozambique mouthbrooder (Oreochromis mossambicus) has established populations in the lowland reaches of several coastal rivers in Queensland between Brisbane and Cairns, and has been reported in several rivers in south-western Western Australia (McKay 1984; Allen 1989; McDowall 1996).The guppy (Poecilia reticulata) is widespread from Brisbane to north-east Queensland (Ryan 1995).The one-spot livebearer (Phalloceros caudimaculatus) has been recorded in ponds and drains around Perth (Allen 1989).The sailfin molly (Poecilia latipinna), swordtail (Xiphophorus helleri) and platy (X. maculatus), are restricted to a few streams around Brisbane (McDowall 1996).The rosy barb (Puntius conchronius) also established itself in one stream in the Brisbane area but has apparently died out (Brumley 1991; McDowall 1996).The American flag fish (Jordanella floridae) has been recorded near Cairns (Allen 1989).The convict and black mangrove cichlid (Tilapia mariae) are restricted to the Hasellwood Pondage in Morwell Victoria, which contains warm water outflow from the power station. Zilles cichlid (T. zillii) was recorded in tributaries of the Swan River estuary in Western Australia. in 1975, but is believed to have been successfully eradicated (Allen 1989).The blue acara (Aequidens pulcher) has been recorded in one stream in Brisbane (Ryan 1995). Many of Australia’s inland waters, particularly in south-eastern regions, contain one or more introduced fish species. In some cases these species have completely displaced native fish and substantially modified aquatic ecosystems. In addition to those species already established, there is continual interest from the recreational and commercial fishing industry to establish hatcheries or introduce more species. Several hundred species of exotic ornamental fish have been imported into Australia for the aquarium trade (McKay 1984).The potential for more of these species to become established in natural waters is high, particularly in tropical and subtropical regions (McKay 1984; McDowall 1996). With the possible exception of the Oriental Weatherloach, all of the introduced species have the potential to prey upon amphibian eggs and larvae, and many species may also prey upon adults. As indicated in the literature reviewed earlier, this has already been demonstrated for most of the more widely introduced species, such as the salmonids, cyprinids, redfin perch and gambusia, on other continents. 135
REVIEW OF IMPACTS OF INTRODUCED FISH ON AMPHIBIANS IN AUSTRALIA Few studies have been conducted to investigate the relationships between introduced fish and amphibians in Australia.To date, the impacts of only three introduced species, brown and rainbow trout, and gambusia, have been investigated. Collectively these studies have assessed impacts on only 16 species of frog to any degree (Table 2). Some information is presented on redfin perch and carp; however, appropriate research is required to further examine the impacts of these species.
Impact of Fish Predation on Adult Frogs It is common knowledge among the fishing fraternity that frogs make good bait for trout and redfin perch (Baxter, Victorian Fisheries, pers. comm.; Harris, NSW Fisheries, pers. comm.; Lake, Department of Biological Sciences, Monash University, pers. comm.).This suggests that trout and redfin perch may readily attack frogs in the wild. Collection of frogs for bait may place excessive pressure on some frog populations (Watson et al. 1991).The use of frogs for bait is now banned in some States. However, it is likely that fish are able to exert their greatest impact on frog populations by preying upon larval stages.
Impact of Predation by Trout Species Species most at risk from predation by trout are those which breed exclusively in streams in south-eastern Australia.There are eight such species, several of which have declined in recent years and are considered endangered or vulnerable (Gillespie and Hines 1999).The spotted tree frog (Litoria spenceri) has always been considered rare (Ahern 1982); however, declines were observed in most of the few known populations in the 1970s and 80s (Watson et al. 1991), and the species is now listed as endangered (Tyler 1997). Watson et al. suggested that introduced trout may be contributing to this decline.Trout are present in all the streams in which the species is known to have occurred (Victorian Fish database, ARI,Victoria). Surveys of the distribution and relative abundance of L. spenceri and other upland riverine species have found that L. spenceri only occurred in abundance in one reach of stream which was above a waterfall which trout could not negotiate (Gillespie and Hollis 1996; Hunter and Gillespie 1999). Only a few high density upland populations of the leaf-green tree frog (L. phyllochroa) have been located, most of which have also been above waterfalls in trout-free streams (Gillespie pers. obs.). In contrast, lesueur’s frog (L. lesueuri) remains widespread and is abundant along many streams where trout are present (Gillespie and Hollis 1996; Hunter and Gillespie 1999; Gillespie pers. obs.). Gillespie (1997, unpubl.) examined the relative palatabilities of five riverine frog larvae, L. booroolongensis, L. citropa, L. lesueuri, L. phyllochroa and L. spenceri, to two sympatric native fish, mountain galaxias (Galaxias olidus) and two-spined blackfish (Gadopsis bispinosis), and introduced brown trout. All fish readily consumed tadpoles of Limnodynastes peronii which occur in lentic habitats without fish; however, only trout ate a significant proportion of tadpoles of any riverine tadpole species. Further in-stream experiments demonstrated that despite the availability of alternative food
136
sources for trout, and refuge microhabitats for tadpoles, trout were able to impose a significant predation pressure on L. spenceri and L. phyllochroa. Brown and rainbow trout are now considered to be the primary cause of decline of L. spenceri (Robertson et al. 1998, unpublished; Robertson and Gillespie, 1998, unpubl.).These findings suggest that upland populations of the other species within the L. citropa complex, i.e. L. subglandulosa and L. pearsoniana, may also be highly vulnerable to predation from trout. Although L. lesueuri complex species were less palatable, L. booroolongensis has also declined (Gillespie and Hines 1999; NSW NPWS Scientific Committee Determination Advice No. 97/27). Other factors such as habitat degradation may be involved but it remains unclear what impact trout may have on these species. For example, an egg mass of L. lesueuri was found in the stomach of a brown trout (Rardik, ARI,Victoria, pers. comm.).This fish was able to take out most of the annual reproductive investment of a single female frog (several hundred eggs) in one sitting. It is expected that predation pressure by trout on these species will be high when alternative food resources are limited.
Impact of Predation by Gambusia Gambusia has so far received the most scrutiny regarding its potential impact upon Australian frog populations.The broad distribution and wide range of habitats occupied by gambusia means that it may potentially impact upon many lentic and lotic frog populations across a large area of Australia. Only one study has examined predation by gambusia upon anuran eggs; Reynolds (1995) found that eggs of Crinia insignifera and C. glauerti were unpalatable. Preliminary trials also suggested that eggs of Litoria adelaidensis, L. moorei and Crinia georgiana may also be unpalatable (Reynolds 1995). However, several studies have shown experimentally that gambusia are capable of preying on small larvae of a number of Australian anuran species: Limnodynastes tasmaniensis, Litoria lesueuri and L. dentata (Harris 1995); Crinia insignifera and C. glauerti (Reynolds 1995); Litoria aurea and L. dentata (Morgan and Buttemer 1996); Limnodynastes peronii and Crinia signifera (Webb and Joss 1997). A number of studies have identified negative associations between the presence of gambusia and frog species. Dankers (1977) found that tadpole numbers of several species were drastically reduced in ponds containing gambusia after early December, coinciding with a seasonal increase in fish biomass. McGilp (1994) found a negative correlation between the occurrence of Brown Tree Frog (Litoria ewingii) and that of gambusia in waterbodies along the Yarra River in Melbourne. Blyth (1994) compared survival and recruitment of three species of Western Australian anuran larvae, Crinia glauerti, C. insignifera and Heleioporus eyrei, in the presence or absence of gambusia in experimental field enclosures.Tadpole survival of all three species was significantly lower in the presence of gambusia at the end of the experimental period. However, the design of the enclosures allowed access for oviposition by local frog populations, as evidenced by increases in numbers of experimental animals in some enclosures. Other potential predators of premetamorphic stages also had access, such as invertebrates and birds. Furthermore, each species/fish treatment was not replicated.These factors limit interpretation of the results of this study.
TABLE 2: List of introduced fish and native frog species-interactions that have been examined in Australia.
Fish species
Frog species
Brown Trout (Salmo trutta)
Booroolong Frog Blue Mountains Tree Frog Lesueur’s Frog Leaf-green Tree Frog Spotted Tree Frog Striped Marsh Frog Leaf-green Tree Frog Spotted Tree Frog Slender Tree Frog Green and Golden Bell Frog
Litoria booroolongensis L. citropa L. lesueuri L. phyllochroa L. spenceri Limnodynastes peroni L. phyllochroa L. spenceri Litoria adelaidensis L. aurea
Kerferstein’s Tree Frog
L. dentata
Moor’s Frog Lesueur’s Frog Tschudi’s Froglet Glauert’s Froglet Sign-bearing Frog Common Froglet Moaning Frog Striped Marsh Frog Spotted Marsh Frog Spotted Marsh Frog
L. moorei L. lesueuri Crinia georgiana C. glauerti C. insignifera C. signifera Heleioporous eyrei Limnodynastes peroni L. tasmaniensis L. tasmaniensis
Rainbow Trout (Oncorhynchus mykiss) Eastern gambusia or Mosquito Fish (Gambusia holbrooki)
Goldfish (Carassius auratus)
Webb and Joss (1997) examined amphibian species richness and abundance in relation to gambusia density and cover of emergent aquatic vegetation in ten ponds near Sydney.They found a significant negative relationship between fish density and frog abundance but no relationship for species richness. The descriptions provided for each waterbody indicate a high degree of variability in habitat among pond sites. Unfortunately additional factors such as pool size and native vegetation cover, which may strongly affect frog abundance, were not considered in their analyses.Tadpole density is easier to sample systematically than adult frog density in pond habitats (Heyer et al. 1994). Given that tadpoles are one of the life stages on which gambusia potentially preys upon, a measure of their relative abundance, rather than that of adult frogs, will provide a more reliable indicator of the impact of gambusia. Reynolds (1995) examined the occurrence of six anuran species with gambusia in water bodies near Perth, Western Australia. In contrast to the above studies, he found no relationship between the presence/absence of fish and individual anuran species, with one exception, Crinia insignifera, which was found infrequently with gambusia. However, he observed that most of the sites used by C. insignifera were ephemeral and unsuitable for gambusia. Species richness was generally lower at sites occupied by gambusia, but many of these sites were also degraded, contributing to their unsuitability as frog breeding habitats. In addition Reynolds (1995) experimentally examined predation by gambusia on several tadpole species in Western Australia.Trials with tadpoles indicated that gambusia were able to attack and kill tadpoles of L. adelaidensis, C. georgiana and H. eyrei. Controlled palatability experiments showed that survival of L. moorei tadpoles was significantly reduced in the
Source
Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Gillespie (1997), unpublished Reynolds (1995) Morgan and Buttemer (1996); Pyke and White (1996) Harris (1995); Morgan and Buttemer (1996) Reynolds (1995) Harris (1995) Reynolds (1995) Blyth (1994); Reynolds (1995) Blyth (1994); Reynolds (1995) Webb and Joss (1997) Blyth (1994); Reynolds (1995) Webb and Joss (1997) Harris (1995) M. Healey (unpublished data)
presence of gambusia. However, gambusia showed a strong preference for invertebrate prey (Daphnia sp. or mosquito larvae). Both groups were consistently consumed completely before tadpoles in all trials. In a field enclosure experiment, in which tadpoles were also exposed to invertebrate predators, Reynolds (1995) found no significant difference in survival in the presence or absence of gambusia.These results, in conjunction with his field survey data, suggest that the impact of gambusia upon populations of these frog species is influenced by several factors, and under natural conditions may be limited. Gambusia cannot consume large prey as these small fish are gape-limited predators. Webb and Joss (1997) conducted predation experiments examining the impact upon survival of different size classes of C. signifera and Limnodynastes peroni tadpoles by hungry and pre-fed gambusia.They found significant differences between predation rates due to tadpole size class and hunger status of fish.Tadpole species which are able to rapidly attain moderate to large size may therefore minimize the impact of predation (Caldwell et al. 1980; Crump 1984). Several studies have reported damage to the fins of larger tadpoles from gambusia attack (Dankers 1977; Blyth 1994; Harris 1995).This could result in reduced survival of larger tadpoles due to reduced mobility and feeding, inability to escape other predators, or reduced metamorphic fitness. However, some tadpole species have been found to survive tail loss (Harris 1995). Wilbur and Semlitsch (1990) reported tail regeneration by tadpoles of Rana catesbeiana even after considerable loss, and suggest that this may be a general mechanism to reduce the impact of predation. 137
Concerns for the role of gambusia in the decline of amphibian species, particularly members of the L. aurea complex, have been expressed by several authors (Mahony 1993; Daly 1995; Morgan and Buttemer 1996; White and Pyke 1996; White and Ehmann 1997). However, evidence linking gambusia to declines of frog populations in the L. aurea complex is limited, due in part to conflicting findings and methodological limitations of some studies. For example, Morgan and Buttemer (1996) conducted controlled predation experiments examining the impact upon survival of tadpoles of L. aurea and L. dentata by gambusia. The influence of macrophytes on the predatory impact of gambusia was also examined.They found that in the absence of macrophytes gambusia were able to significantly reduce tadpole survival of both species within 24 hours. In the presence of macrophytes, the effect was substantially reduced and no significant impact of gambusia could be detected on L. aurea after three days. However, survival of L. dentata was still significantly reduced after two days.These findings indicate that presence of gambusia may significantly influence the survival of tadpoles, but that this is likely to be strongly influenced by habitat structure and tadpole behaviour. Litoria aurea larvae have also been found in sympatry with native predatory fish (pers. obs.). In the absence of comparative data on the impact of these natural predators upon larval survival, it is difficult to assess the relative ecological significance of gambusia predation. Pyke and White (1996) surveyed waterbodies in the Sydney region for L. aurea, and examined associations between evidence of breeding, occurrence of introduced fish, and habitat.They found that breeding was most strongly associated with ephemeral rather than permanent or ‘fluctuating’ ponds, followed by the absence of introduced fish, primarily gambusia, and speculated that this fish was a major cause of decline of L. aurea (Pyke and White 1996). However, examination of their data reveals that pond permanency and occurrence of gambusia are highly correlated and so the results could also be explained in terms of unmeasured features of pond permanency, or abundance of other predators. White and Ehmann (1997) suggest that gambusia is also implicated in the decline of L. flavipunctata, a closely related species to L. aurea. However, Osborne et al. (1996) point out that many of the sites from which this species has disappeared do not contain gambusia. Furthermore, both L. aurea and L. raniformis, an ecologically similar species which hybridises with L. aurea (Watson and Littlejohn, 1985), have been recorded in abundance at some sites containing gambusia (van de Mortel and Goldingay 1998; Gillespie pers. obs.; Pyke, Australian Museum, pers. comm.). The role of gambusia in the decline of L. aurea is unclear. Other factors require careful consideration, such as pond duration, habitat quality, presence of other aquatic predators and availability of refugia. Evidence of gambusia having a major impact on the abundance of other Australian amphibians is also unclear. However, many of the studies to date have demonstrated that gambusia are capable of killing a variety of tadpole species and eggs. Considering the wide distribution of gambusia, it probably does have significant impacts on some native amphibian species, particularly in the eastern states where seasonal peak fish abundance coincides with the larval stages of many species 138
(Reynolds 1995). Further research is required to ascertain the role of gambusia in the decline of amphibian species assemblages with respect to other threatening processes. For instance, as gambusia occur in areas which are mostly disturbed or modified in other ways, the relative impacts of these habitat changes upon amphibian populations need to be differentiated from those wrought by the fish.
Impacts of Predation by Redfin Perch and Carp Species Research overseas suggests that redfin perch and carp species may be major predators of some tadpole species. No published studies have addressed the impact of carp species or redfin perch upon frogs in Australia. Leslie (1995) has attributed the decline of frogs in some wetlands in the Murray-Darling Basin in part to predation on premetamorphic stages by carp species. Healey et al. (1997) observed no evidence of frogs breeding in four billabongs on the Murrumbidgee floodplain and suggested that this may have been explained by the presence of carp species. However, no observations were made at sites where carp were absent and the absence of tadpoles could be explained by a number of alternative abiotic and biotic hypotheses. Laboratory predation experiments have shown that tadpoles of Limnodynastes tasmaniensis are palatable to goldfish and redfin perch (Healey, Charles Sturt University, Wogga Wogga, pers. comm.), indicating the potential for these species to consume tadpoles. However, it is unknown whether they prey on tadpoles in the wild when alternative food is available. Carp are able to significantly modify the physical habitat of aquatic systems, by uprooting aquatic vegetation and increasing turbidity (Roberts et al. 1995).These changes may have indirect impacts on tadpoles through loss of food resources, cover for protection from other predators, and loss of oviposition sites.
Broader Implications of Introduced Fish for Australian Amphibian Species The evidence presented here strongly suggests that introduction of exotic fish or translocation of native species could have an enormous impact on the amphibian assemblages of Australian freshwater systems. However, in many cases the impacts have not been investigated. For example, the impact of carp species on Australian amphibian assemblages has not been examined, despite their widespread distribution and frequently-raised concerns about their adverse effects upon freshwater systems. Introduced fish within mainland Australia are currently generally restricted to the eastern sea board, Murray-Darling system and south-west Western Australia.This distribution also overlaps with regions of high amphibian species richness (see Barker et al. 1996). Consequently a large proportion of Australian anurans are potentially affected by one or more introduced fish species. Species most likely to be affected are those which breed in permanent aquatic habitats, such as streams and wetlands. However, many which breed in more ephemeral habitats, such as billabongs and temporary pools along flood plains of rivers, may also be affected as these habitats are seasonally colonised by introduced fish when water courses swell. Changes to the rural landscape within these regions have
resulted in removal of many natural ephemeral aquatic habitats and the expansion of more permanent habitats by way of stock dams.These are the only breeding habitats in some areas for species which would otherwise breed in ephemeral water bodies. Farm dams are often stocked with introduced and native angling species which are likely to impact these amphibian assemblages.These habitats may have become ecological sinks for some species. A large proportion of Australia’s threatened amphibian species breed in habitats currently occupied by, or within the range of, introduced or translocated fish. Lotic species assemblages are particularly vulnerable.The range of introduced trout species includes part or all of the distributions of ten south-eastern Australian lotic species, five of which have declined (Tyler 1997; NSW NPWS Scientific Committee Determination Advice No. 97/27; Gillespie and Hines 1999). Some populations of these species are probably exposed to redfin perch and gambusia as well. The three species currently recognised within the L. aurea complex have all declined. Gambusia occur throughout much of the range of these species. Redfin perch and carp species occur throughout most of the range of L. flavipunctata, L. raniformis and in part of the range of L. aurea. Gambusia have already been implicated in the decline of this species group; redfin perch and carp species may also be contributing. Other regions of Australia which contain significant amphibian assemblages, but are currently free of introduced fish, such as the Wet Tropics, may be at risk in the future if further introductions of other exotic fish species occur.
Potential for Introducing Exotic Pathogens Recent studies have suggested that an introduced pathogen may be responsible for amphibian declines in Australia and Central America (Blaustein et al. 1994; Laurance et al. 1996; Lips 1998; Berger et al. 1998).The potential for the introduction of disease into Australian freshwater systems via the importation of fish for the aquarium trade has been clearly identified (Mckay 1984; Laurance et al. 1996). Laurance et al. (1996) has suggested that a pathogen introduced in this way might be responsible for frog declines in north-east Queensland, but at this time there is no evidence to support this (Hero and Gillespie 1997; Alford and Richards 1997). However, disease risk imposed by the continual importation of live freshwater fish into the country cannot be ignored.
MANAGEMENT SOLUTIONS The importation of exotic fish for the aquarium trade should only be acceptable following rigid quarantine protocols that eliminate the possibility of introducing pathogens either with the fish or the water they are transported in.The aquarium trade should be advised of the potential impact of introduced pathogens and fish species and a shift towards the use of native fish species for the pet-trade encouraged. Similarly, gambusia should not be introduced into new systems for mosquito control; alternatively native fish species local to the area may be more suitable. Once fish have been introduced into an aquatic system and established self-sustaining populations, they are extremely difficult to remove. Small, confined water bodies, such as dams,
can be drained to effect 100% removal. However, this option is usually not available. Most introduced fish in Australia occur in streams or larger waterbodies which cannot easily be drained.There have been numerous attempts in the United States of America to eradicate unwanted fish populations from streams and lakes, using a variety of techniques, such as electrofishing, netting and poisoning.The only demonstrated successful approach for complete removal of fish from these systems is with a toxicant.This approach has become a standard management technique throughout the USA (Eschmeyer 1975), mainly as a fishery technique to improve populations of recreational over non-recreational species (Ryan 1977). More recently this has expanded to aquatic conservation to protect threatened fauna from introduced fish species. However, examples of treatments designed to accomplish a complete kill, as required for long-term exclusion, are few and, of these, only few have been successful (Rinne et al. 1981; Gresswell 1991; Stefferud et al. 1992). There has been only one successful eradication of salmonids from any Australian waters.This was conducted in several small mountain streams in eastern Victoria as part of the implementation of the barred galaxias Recovery Plan (Raadik 1993). Artificial trout barriers were established across the streams and all fish above the barriers and below remaining native fish populations were killed with rotenone, allowing the native species to recolonise the rehabilitated zones (Raadic, ARI, unpublished data). Expanding this approach to larger watercourses is problematic. It is more difficult to effect a complete eradication due to an exponential increase in stream length and complexity with increased catchment size, barrier construction becomes increasingly more difficult and expensive on larger streams, and the risk of re-introduction also increases. Saddlier and Gillespie (1997) assessed the feasibility of excluding trout from streams to protect populations of L. spenceri. Of the 13 streams examined, exclusion was considered feasible only on reaches of three streams because of the above constraints. If successful this would afford protection to approximately 7 % of the current range of the species. The environmental and socio-economic costs of eradicating fish must also be measured against the longer-term benefits to conservation. Several problems arise with eradication programs. 1. Native fish species and some invertebrate groups are also affected by rotenone, which disables gill function. 2. Most large waterbodies such as lakes and streams are also used to supply water for human consumption and recreational purposes, including fishing; hence poisoning may risk human health. Furthermore opposition by the recreational fishing community is likely to influence the political decision-making process. 3. For waterbodies which are used for angling, there is a high risk of reintroduction of popular angling species by members of the public.These factors further restrict the range of circumstances in which eradication of introduced fish is feasible. In the future it may be possible to develop biological agents to control introduced fish populations. However, this would be extremely costly and take many years to develop. Clearly, in many instances removal of introduced fish for maintenance of amphibian populations is not feasible at this time. Management should focus therefore on identifying and acting 139
on those habitats where the feasibility of removal of introduced fish is high, and restricting the spread of existing exotic species and further introductions of more species. There is a strong fishing culture in Australia which has a large focus on introduced species. Many introduced species, such as salmonids and redfin perch, continue to be considered desirable alien species by State agencies. However, servicing this culture continues to erode the biotic integrity of Australian freshwater systems.There needs to be a shift in emphasis by State fisheries managers from introduced fish to native species.The ongoing commercial stocking programs for some introduced species in some States, such as salmonids, pose a major threat to several significant amphibian assemblages. Discontinuation of stocking programs, especially in regions where the fish populations are not self-sustaining, will greatly benefit the conservation of some amphibian species.The needs of recreational fishing must be balanced with the benefits derived from maintaining natural fish assemblages. Control of exotic fish stocks may enhance remaining native fish stocks which are also suited to recreational fishing pursuits, while maintaining natural assemblages of native invertebrates and amphibians. It is important to emphasise that native species should not be released into systems in which they did not occur naturally.
CONCLUSIONS AND FUTURE DIRECTIONS In summary, fish are a major influence on amphibian assemblage structure. Hence, they play a major role in determining the distribution and abundance of amphibian species.The introduction of exotic fish to aquatic systems has the potential to eliminate amphibian species. Additionally there is potential to introduce disease or pathogens into freshwater systems.These changes may be extremely detrimental to the long-term survival of some species, undermine amphibian communities and disrupt natural aquatic systems. In view of the large number of introduced fish species and extensive distribution of some of these within Australia, many amphibian communities are currently vulnerable to impacts from exotic fish. Limited research has been carried out in Australia on the impact of introduced fish upon amphibian assemblages. However, there is strong evidence from both overseas and within Australia that those fish species which have been introduced pose a serious threat to a range of anuran species, a number of which have already declined. In particular, trout have been shown to be responsible for the decline of at least one threatened species (L. spenceri), and gambusia has been suggested in the decline of others.The impact and management of introduced fish therefore warrants serious consideration in the development and implementation of recovery plans for declining frog species. Further information is required to assess the impact of introduced fish upon amphibian assemblages throughout the range of habitats and regions in which they have spread. This is essential to gain a proper understanding of the role of introduced fish in frog declines, and identify management objectives. Priority should be given to the following areas of investigation: Determine which introduced or translocated fish species are impacting upon frog communities, and which frog species and communities are most at risk. 140
Information is required for most fish species which have established self-sustaining populations and a range of exotic and native species which are readily introduced in frog habitats. Information is urgently required on the impacts of redfin perch and carp species, which are widespread and potentially affect numerous frog species, particularly in the Murray Darling system and New England Tablelands region of NSW where several frog species have disappeared. More information is required to ascertain the impact of gambusia in Eastern Australia, particularly on the L. aurea complex.The impact of trout upon all upland temperate lotic anuran species in south-eastern Australia also requires further investigation. Broad-scale surveys are required to determine relationships of occurrence of frog species in relation to the distribution of introduced and native fish, and a range of other biotic and abiotic variables.The ability of fish species to impact frog communities should be tested experimentally.The relative effectiveness of tadpole survival strategies amongst species in the community to native sympatric predators and introduced species should be compared. Predation experiments should include adequate replication and use of known palatable and unpalatable (where available) tadpole species as controls. Fish density, fish size and tadpole sizes that replicate field observations should also be factored into experimental designs.The ability of fish to prey upon eggs should also be examined if possible as this may be a more vulnerable life stage for some species. Ascertain the relative importance of the role of introduced fish in frog population declines with relation to other biotic or abiotic factors. Relative impact of introduced fish must be examined in conjunction with other factors potentially limiting survival, such as habitat degradation. Other biotic or abiotic factors may either exacerbate or ameliorate the impacts of introduce fish on frog populations. Surveys to determine relationships between occurrence of fish and frogs should incorporate collection and analysis of confounding biotic and abiotic habitat variables (e.g. hydroperiod, water quality, aquatic vegetation, adjacent adult habitat, native predator abundance, etc.).These should be designed where possible with adequate power in sample size (i.e. adequate number of waterbodies) to assess the relative contributions of introduced fish and other factors which significantly influence occurrence of frog species. For some species which are rare or have limited distributions, surveys of this kind are likely to have inadequate power.The relative impact of introduced fish can be examined in field experiments conducted in stream or pond enclosures which closely mimic conditions experienced in natural breeding bodies, incorporating natural levels of cover, sympatric predators and hydroperiod. One or more variables can be manipulated, in conjunction with predator levels to assess their relative contributions to larval mortality. The threats imposed by introduced fish to Australian amphibian assemblages require immediate and on-going attention by conservation and fisheries managers.The possibility of eradication of introduced species should be assessed on a case by case basis; however, this is currently expected to have limited feasibility in many instances.There is
a strong need for development of improved effective eradication techniques. However, the immediate priority for managers should be prevention of further translocations and introductions of fish species. For those species under serious threat from introduced fish, all extant populations currently free of introduced predators should be identified and appropriate steps taken to ensure that fish are not introduced. Current policies and management of introduced fisheries and the aquarium-trade require review and need to take into consideration the potential impact upon amphibian assemblages. Enhancement of native fisheries, rather than those based upon exotic species should be encouraged. Stocking programs for introduced fish should be discontinued in aquatic systems known to support vulnerable amphibian species.Tighter control of the importation and maintenance of ornamental species is required, particularly of those species with potential to establish self-sustaining wild populations and impact upon native biota.The public also need to be educated about adverse effects of releasing or translocating fish on amphibians and other biota.
ACKNOWLEDGMENTS We thank the Arthur Rylah Institute, Department of Natural Resources and Environment,Victoria, and Griffith University for financial support. H. Hines, B. Magnusson, R. Loyn,T. Raadik, R. Swain and W. Osborne provided comments on the manuscript.
REFERENCES Alford, R. A. and Richards, S. J. (1997). Lack of evidence for epidemic disease as an agent in the catastrophic decline of Australian rainforest frogs. Conservation Biology, 11: 1026-29. Ahern, L. D. (1982) Threatened Wildlife in Victoria and Issues Related to its Conservation. Ministry for Conservation, Fisheries and Wildlife Service Paper 27. Allen, G. R. (1982) A Field Guide to Inland Fishes of Western Australia. Western Australian Museum, Perth, Australia. Allen, G. R. (1989) Freshwater Fishes of Australia.T. F. H. Publications, New Jersey. Arthington, A. H. and Blühdorn, D. R. (1995) Improved Management of Exotic Aquatic Fauna: Research and Development for Australian Rivers. Land and Waters Resources Research and Development Corporation. Azevedo-Ramos, C.,Van Sluys, M., Hero, J.-M. and Magnusson W. E. (1992) Influence of tadpole movement on predation by odonate naiads. Journal of Herpetology, 26: 335-38. Barker, J., Grigg, G. and Tyler, M. J. (1995). A Field Guide to Australian Frogs. 2nd Edition. Surrey Beattie, Chipping Norton, NSW. Berger, L., Spear, R., Daszak, P., Green, D. E., Cunningham, A. A., Goggin, C. L., Hines, H. B., Lips, K. R., Marantelli, G. and Parkes, H. (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rainforests of Australia and Central America. Proceedings of the National Academy of Sciences, 95: 9031-9036.
Bertram, S. and Berrill, M. (1997) Fluctuations in a northern population of gray treefrogs, Hyla versicolor. In D. M. Green (ed.) Amphibians in Decline Canadian Studies of a Global Problem. Chapture 6, Herpetological Conservation No. 1, Canadian Association of Herpetologists, Montréal, pp. 57-63. Bidwell, J. R. and Gorrie, J. R. (1995) Acute Toxicity of a Herbicide to Selected Frog Species: Final Report. Western Australian Department of Environmental Protection, Perth. Bishop, C. A. (1992) The effects of pesticides on amphibians and implications for determining causes of declines in amphibian populations. In C. A. Bishop and K. E. Pettit (eds.) Declines in Canadian Amphibian Populations: Designing an National Monitoring Strategy. Occasional Paper No. 76, Canadian Wildlife Service, Ottawa, pp. 67-70. Blaustein, A. R., Hokit. D. G., O’Hara, R. K. and Holt, R. A. (1994) Pathogenic fungus contributes to amphibian losses in the Pacific Northwest. Biological conservation, 67: 251-254. Blyth, B. (1994) Predation by Gambusia holbrooki on Anuran Larvae at the RGC Wetlands Centre, Capel Western Australia. RGC Wetlands Centre Technical Report No. 22, Capel, W.A. Bradford, D. F. (1989) Allopatric distribution of native frogs and introduced fishes in the high Sierra Nevada lakes of California: implication of the negative effect of fish introductions. Copeia, 1989: 775-8. Bradford, D. E.,Tabatabai, F. and Graber, D.M. (1993) Isolation of remaining populations of the native frog, Rana mucosa, by introduced fishes in Sequoia and Kings Kanyon National Parks, California. Conservation Biology, 7: 882-8. Braña, F., Frechilla, L. and Orizaola, G. (1996) Effect of introduced fish on amphibian assemblages in mountain lakes of northern Spain. Herpetological Journal, 6: 145-148 Brodie Jr, E. D., Formanowicz, D. R. and Brodie III, E. D. (1978). The development of noxiousness of Bufo americanus tadpoles to aquatic insect predators. Herpetologica, 34: 302-306. Brönmark, C. and Edenhamn, P. (1994) Does the presence of fish affect the distribution of tree frogs (Hyla arborea)? Conservation Biology, 8: 841-5. Brumley, A. R. (1991) Cyprinids in Australia, In: I. J. Winfield and J. S. Nelson (eds) Cyprinid Fishes: Systematics, Biology and Exploitation, Chapman and Hall, London. pp. 265-283. Burger, W. L. (1950) Novel aspects of the life history of two Ambystomas. Journal of the Tennessy Academy of Science, 25: 252-257. Cadwallader, P. L. and Backhouse, G. N. (1982) A Guide to the Freshwater Fish of Victoria.Victorian Government Printing Office, Melbourne, Australia. Caldwell, J. P.,Thorp, J. H., and Jervey,T. 0. (1980) Predator-prey relationships among larval dragonflies, salamanders, and frogs. Oecologia, 46: 285-9. Calef, G. W. (1973) Natural mortality of tadpoles in a population of Rana aurora. Ecology, 54: 741-58. Chovanec, A. (1992) The influence of tadpole swimming behaviour on predation by dragonfly nymphs. AmphibiaReptilia, 13: 341-9.
141
Crump, M. L. (1984) Ontogenetic changes in vulnerability to predation in tadpoles of Hyla pseudopumma. Herpetologica, 40: 265-71.
Grubb, J. C. (1972) Differential predation by Gambusia affinis on the eggs of seven species of anuran amphibians. The American Midland Naturalist, 88: 102-108.
Daly, G. (1995) Observations on the green and golden bell frog (Litoria aurea) (Anura: Hylidae) in southern New South Wales. Herpetofauna, 25: 2-9.
Harris, K. (1995) Is there a negative relationship between gambusia and tadpoles on the Northern Tablelands? B.Sc. 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. Dawkins, R. and Krebs, J. R. (1979) Arms races between and within species. Proceedings of the Royal Society of London, 205: 489-511. Dubuis, L. A. (1997) Effects of logging on terrestrial amphibians of coastal British Columbia. Chapture 20 In D. M. Green (ed) Amphibians in Decline Canadian Studies of a Global Problem. Herpetological Conservation No. 1., Society for the Study of Amphibians and Reptiles, Saint Louis, Missouri, pp. 185-90. Duellman, W. E. (1978) The biology of an equatorial herpetofauna in Amazonian Ecuador. University Kansas Museum of Natural History Miscellaneous Publication, 65: 1-352. Eschmeyer, P. H. (1975) (ed.) Rehabilitation of Fish Populations with Toxicants: A symposium. North Central Division Special Publication 4. American Fisheries Society, Bethesda, Maryland, USA. Faragher, R. A. and Harris, J. H. (1993) The historical and current status of freshwater fish in New South Wales. Australian Zoologist, 29(3-4): 166-76. Fickling, S. (1995) The influence of fish predation on the local distribution of frogs and tadpoles in rainforest streams of the Tully River. Unpublished Honours Thesis, James Cook University,Townsville, Australia. Formas, J. R. (1995) Amphibians. Chapt. 44, In J. A. Simonetti, M.T. K. Arroyo, A. E. Spotorno and E. Lozada (eds) Diversidad Biologica de Chile, pp.314-25. Gamradt, S. C. and Kats, L. B. (1996) Effect of introduced crayfish and mosquitofish on Californian newts. Conservation Biology, 10: 1155-62. Gillespie, G. R. (1997) The Biology of the Spotted Tree Frog (Litoria spenceri) and Examination of Factors Responsible for Population Declines. Unpublished Report to the Biodiversity Group, Environment Australia, Canberra. Arthur Rylah Institute, Department of Natural Resources and Environment, Melbourne,Victoria. Gillespie, G. R. and Hines, H.B. (1999) Status of temperate riverine frogs in south-eastern Australia. Pp 109-130 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Gillespie, G. R. and Hollis, G. J. (1996) Distribution and habitat of the spotted tree frog Litoria spenceri Dubois (Anura: Hylidae), and an assessment of potential causes of population declines. Wildlife Research, 23: 49-75. Gresswell, R. E. (1991) Use of antimycin for removal of brook trout from a tributary of Yellowstone Lake. North American Journal of Fisheries Management, 11: 83-90. Grinnell, J. and Storer,T. I. (1924) Animal life in the Yosemite. University of California Press, Berkeley, California.
142
Hayes, M. P. and Jennings, M. R. (1986) Decline of Ranid frog species in Western North America. Are bullfrogs (Rana catesbeiana) responsible? Journal of Herpetology, 20: 490-509. Healey, M.,Thompson, D. and Robertson, A. (1997) Amphibian communities associated with billabong habitats on the Murrumbidgee floodplain, Australia. Australian Journal of Ecology, 22: 270-78. Hecnar, S. J. and M’Closkey, R.T. (1997) The effects of predatory fish on amphibian species richness and distribution. Biological Conservation, 79: 123-31. Hero, J.-M. (1991) Predation, palatability and the distribution of tadpoles in the central Amazon rainforest. Unpublished PhD thesis, Griffith University, Brisbane, Australia, 234 pp. Hero, J.-M., Gascon, C. and Magnusson, W. E. (1998) Direct and indirect effects of predation on tadpole community structure in the Amazon rainforest. Australian Journal of Ecology, 23: 474-482. Hero, J.-M. and Gillespie, G. R. (1997). Epidemic disease and amphibian declines in Australia. Conservation Biology, 11: 1023-25. Heyer, W. R., Donnelly, M. A., McDairmid, R. W., Hayek, L.-A. C. and Foster, M. S. (eds.) (1994) Measuring and Monitoring Biological Diversity. Standard Methods for Amphibians. Smithsonian Institution Press, Washington. Heyer, W. R., McDiarmid, R. W. and Weigmann, D. L. (1975) Tadpoles, predation and pond habitats in the tropics. Biotropica, 7: 100-11. Holomuzki, J. R. (1995) Oviposition sites and fish-deterrent mechanisms of two stream anurans. Copeia, 1995: 607-613. Hopey, M. E. and Petranka, J. W. (1994) Shorter contributions: ichthyology. Copeia, 1994(4): 1023-1025. Hunter, D. and Gillespie, G. R. (1999) The distribution, abundance and conservation status of riverine frogs in Kosciuszko National Park. Australian Zoologist, 31: 198-209. Jennings, M. R. and Hayes, M. P. (1985) Pre-1900 overharvest of the Californian red-legged frog (Rana aurora draytonii): the inducement for bullfrog (Rana catesbeiana) introduction. Herpetologica, 41: 94-103. Johnson, B. (1992) Habitat loss and declining amphibian populations. In C. A. Bishop and K. E. Pettit (eds) Declines in Canadian Amphibian Populations: Designing an National Monitoring Strategy. Occasional Paper No. 76, Canadian Wildlife Service, Ottawa, pp. 71-75. Kats, L. B., Petranka, J. W. and Sih, A. (1988) Antipredator defences and the persistence of amphibian larvae with fishes. Ecology, 69: 1865-1870. Kats, L. B. and Sih, A. (1992) Oviposition site selection and avoidance of fish by streamside salamanders. Copeia, 1992: 468-473.
Kruse, K. C. and Francis, M.G. (1977) A predation deterrent in larvae of the bullfrog, Rana catesbeiana.Transactions of the American Fisheries Society, 106: 248-252. Kruse, K. C. and Stone, B.M. (1984) Largemouth bass (Micropterus salmoides) learn to avoid feeding on toad (Bufo) tadpoles. Animal Behaviour, 32: 1035-1039. Laan, R. and Verboom, B. (1990) Effects of pool size and isolation on amphibian communities. Biological Conservation, 54: 251-262. Lannoo, M. J., Lang, K., Waltz,T. and Phillips, G. S. (1994) An altered amphibian assemblage: Dickinson county, Iowa, 70 years after Frank Blanchard’s survey. American Midland Naturalist, 131: 311-319. Laurance, W. F., McDonald, K. R. and Speare, R. (1996) Epidemic disease and the catastrophic decline of Australian rain forest frogs. Conservation Biology, 10: 406-413. Leakey, R. and Lewin, R. (1995) The Sixth Extinction. Doubleday, New York. Leslie, D. J. (1995) Moira Lake: a case study of deterioration of a River Murray natural resource. Unpublished M. Forestry Sc. thesis, University of Melbourne. Liem, K. F. (1961) On the taxonomic status and the granular patches of the Javanese frog Rana chalconota Schlegel. Herpetologica, 17: 69-71. Lips, K. R. (1998) Decline of a tropical montane amphibian fauna. Conservation Biology, 12: 106-116. Liss, W. J. and Larson, G. L. (1991) Ecological effects of stocked trout on North Cascades naturally fishless lakes. Park Science, 11: 22-23. Macan,T.T. (1966).The influence of predation on the fauna of a moorland fish pond. Archiv Für Hydrobiologie, 61: 432-452. Macan,T.T. (1974) Freshwater Ecology. Longmans, London. Mahony, M. J. (1993) The status of frogs in the Watagon Mountains area, the central coast of New South Wales. In D. Lunney and D. Ayers (eds.) Herpetology in Australia a diverse discipline.Transactions of the Royal Society of New South Wales, Mossman, NSW, pp. 257-264. McDowall, R. M. (ed) (1996) Freshwater Fishes of Southeastern Australia. 2nd ed., Reed, Sydney. McGilp, E. (1994) Distribution of anuran amphibians in the lower Yarra River valley. Unpublished B. Sc. Honours thesis. University of Melbourne, Parkville,Victoria. McKay, R. J. (1984) Introductions of exotic fishes in Australia. In: W. R. Courtenay Jr. and J. R. Stauffer Jr. (eds.) Distributions, Biology and Management of Exotic Fishes. John Hopkins University Press, Baltimore & London pp. 177-199. Meffe, G. K. and Carroll, C. R. (1994) Principles of Conservation Biology. Sinauer Associated, Sunderland, M. A. Morgan, L. A. and Buttemer, W. A. (1996) Predation by the non-native fish Gambusia holbrooki on small Litoria area and L. dentata tadpoles. In G. H. Pyke and W. S. Osborne (eds) The Green and Golden Bell Frog (Litoria aurea): Biology and Conservation. Royal Zoological Society of NSW, pp. 143-49. Morin, P. J. (1986) Interactions between intraspecific competition and predation in an amphibian predator-prey system. Ecology, 67: 713-20.
Myers, G. S. (1965) Gambusia, the fish destroyer.Tropical Fish Hobbyist, 13: 31-32. Orchard, S. A. (1992) Amphibian population declines in British Columbia. In C. A. Bishop and K. E. Pettit (eds) Declines in Canadian Amphibian Populations: Designing an National Monitoring Strategy. Occasional Paper No. 76, Canadian Wildlife Service, Ottawa, pp. 10-13. Osborne W. S, Littlejohn, M. J. and Thomson, S. A. (1996) Former distribution of the Litoria aurea complex from the southern tablelands of New South Wales and the Australian Capital Territory. In G. H. Pyke and W. S. Osborne (eds) The Green and Golden Bell Frog (Litoria aurea): Biology and Conservation. Royal Zoological Society of NSW, pp. 190-98. Petranka, J. W. (1983) Fish predation: a factor affecting the spatial distribution of a stream-breeding salamander. Copeia, 1983: 624-28. Pough F. H., Andrews, R. M., Cadle, J. E., Crump, M. L., Savitzky, A. H. and Wells, K. D. (1998) Herpetology. Prentice-Hall, New Jersey. Pyke, G. H. and White, A. W. (1996) Habitat requirements of the green and golden bell frog Litoria aurea (Anura: Hylidae) In G. H. Pyke and W. S. Osborne (eds) The Green and Golden Bell Frog (Litoria aurea): Biology and Conservation. Royal Zoological Society of NSW, pp.224-32. Raadik,T.A. (1993) A Research Recovery Plan for the Barred Galaxias, Galaxias fuscus Mack 1936, in South-eastern Australia. Unpublished report to the Australian National Parks and Wildlife Service; Department of Conservation and Natural Resources,Victoria. Resetarits, W. J. Jr and Wilbur, H. M. (1989) Choice of oviposition site by Hyla chrysoscelis: role of predators and competitors. Ecology, 70: 220-28. Reynolds, S. J. (1995) The impact of introduced mosquitofish (Gambusia holbrooki) on the mortality on premetamorphic anurans. Unpublished B. Sc. Honours thesis, University of Western Australia, Perth, Western Australia. Rinne, J. N., Minkley, W. L., and Hanson, J. N. (1981) Chemical treatment of Ord Creek, Apache County, Arizona, to reestablish Arizona Trout. Journal of Arizona-Navada Academy of Science, 16: 74-78. Roberts, J., Chick, A., Oswald, L. and Thompson, P. (1995) Effect of carp, Cyprinus carpio L., an exotic benthivorous fish, on aquatic plants and water quality in experimental ponds. Marine and Freshwater Research, 46: 1171-80. Robertson, P. and Gillespie, G. R. (1998 unpublished report) Draft Spotted Tree Frog Recovery Plan. Prepared for the Biodiversity Group, Environment Australia, by Department of Natural Resources and Environment, Melbourne,Victoria. Robertson, P., Lowe, K. and Gillespie, G. R. (1998 unpublished report) Flora and Fauna Guarantee Action Statement for the Spotted Tree Frog, Litoria spenceri. Flora and Fauna Branch, Department of Natural Resources and Environment, Melbourne,Victoria. Rowland, S. J. (1989) Aspects of the history and fishery of the murray cod Maccullochella peelii (Mitchell) (Percichthyidae) Proceedings of the Linneaen Society of New South Wales, 111 (3): 201-13.
143
Ryan, J. H. (1977) Non-game fish control at Frenchman Reservoir, Plumas County, California. Department of Fisheries and Game, Inland Fisheries Administration Report 78-1, Sacramento, California, USA. Ryan, M. (Ed.). (1995) Wildlife of Greater Brisbane. Queensland Museum, Brisbane Saddlier, S. R. and Gillespie, G. R. (1997) Feasibility Study of Trout Eradication for Spotted Tree Frog Conservation. Unpublished Report for New South Wales National Parks and Wildlife Service, Jindabyne, Arthur Rylah Institute, Department of Natural Resources and Environment, Victoria. Scott, N. J. and Limerick, S. (1983) Reptiles and amphibians. In D. H. Janzen (ed.) Costa Rican Natural History. University of Chicago Press, Chicago, pp. 351-416. Semlitsch, R. D. (1983) Effects of different predators on the survival and development of tadpoles from the hybridogenetic Rana esculenta complex. Oikos, 67: 40-46. Sexton, O. J. and Phillips, C. (1986) A quantitative study of fish-amphibian interactions in three ponds.Transactions of the Missouri Academy of Sciences, 20: 25-35. Sih, A., J. Petranka, W. and Kats, L. B. (1988) The dynamics of prey refuge use: a model and tests with sunfish and salamander larvae. American Naturalist, 132: 463-83. Smith, D. C. (1983) Factors controlling tadpole populations of the chorus frog (Pseudacris triseriata) on Isle Royale. Ecology, 64: 501-10. Strefferud, J. A., Propst, D. L. and Burton, G. L. (1992) Use of antimycin to remove rainbow trout from White Creek, New Mexico. In Hendrickson, D. A. (ed.). Proceedings of the Desert Fishes Council.Volumes 22 and 23, 1990 and 1991, Annual Symposia and Index for Volumes 16 through 23. Desert Fishes Council, Bishop, California, USA, pp. 55-66. Tayler, J. (1983) Orientation and flight behaviour of a neotenic salamander (Ambystoma gracile) in Oregon.The American Midland Naturalist, 109: 40-49. Tyler, M. J. (1963) A taxonomic study of amphibians and reptiles of the central highlands of New Guinea. Transactions of the Royal Society of South Australia, 86: 105-30.
Waldman, B. (1982) Sibling association among schooling toad tadpoles: field evidence and implications. Animal Behaviour, 30: 700-713. Walters, B. (1975) Studies of interspecific predation within an amphibian community. Journal of Herpetology, 9: 267-79. Wardell-Johnson, G. and Roberts, J. D. (1991) The survival status of the Geocrinia rosea (Anura: Myobatrachidae) complex in riparian corridors: biogeographical implications. In D. A. Saunders, and R. J. Hobbs (eds) Nature Conservation 2:The Role of Corridors, Surrey Beatty & Sons, Chipping Norton, NSW, pp. 167-75. Wasserug, R. (1971) On the comparative palatability of some dry-season tadpoles from Costa Rica. American Midland Naturalist, 86: 101-9. Watson, G. F.and Littlejohn, M. J. (1985) Patterns of distribution, speciation and vicariance biogeography of southeastern Australian amphibians. In G. Grigg, R. Shine and H. Ehmann (eds) Biology of Australasian Frogs and Reptiles. Royal Zoological Society of New South Wales, Sydney, pp. 91-7. Watson, G. F., Littlejohn, M. J., Hero, J.-M. and Robertson, P. (1991) Conservation Status, Ecology and Management of the Spotted Tree Frog (Litoria spenceri) Arthur Rylah Institute for Environmental Research Technical Report Series No. 116. Department of Conservation and Environment, Melbourne. Webb, C. and Joss, J. (1997) Does predation by the fish Gambusia holbrooki (Atheriniformes: Poeciliidae) contribute to declining frog populations? Australian Zoologist, 30(3): 316-24. Werner, E. E. and McPeek, M. A. (1994) Direct and indirect effects of predators on two anuran species along an environmental gradient. Ecology, 75: 1368-82. White, A. and Ehmann, H. (1997) Southern Highlands Bell Frog. In H. Ehmann (ed) Threatened Frogs of New South Wales. Frog and Tadpole Study Group of New South Wales, Sydney, pp. 171-76.
Tyler, M. J. (1997) Action Plan for Australian Frogs. Wildlife Australia Endangered Species Program, Canberra.
White, A. W. and Pyke, G. H. (1996) Distribution and conservation status of the green and golden bell frog Litoria aurea in New South Wales. In G. H. Pyke and W. S. Osborne (eds) The Green and Golden Bell Frog (Litoria aurea): Biology and Conservation. Royal Zoological Society of NSW, pp. 177-189.
Tyler,T., Liss, W. J., Ganio, L. M., Larson, G. L., Hoffman, R., Deimling, E. and Lomnicky, G. (1998) Interaction between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conservation Biology, 12: 94-105.
Wilbur, H. M. (1984) Complex life cycles and community organisation in amphibians. In P. W. Price, C. N. Slobodchikoff and W. S. Gaud (eds.) A New Ecology: Novel Approaches to Interactive Systems. John Wiley, New York. pp. 195-224.
van de Mortel,T. and Goldingay, R. (1998). Population assessment of the endangered green and golden bell frog Litoria aurea at Port Kembla, New South Wales. Australian Zoologist, 30: 398-404.
Wilbur, H. M. and Fauth, J. E. (1990) Experimental aquatic food webs: interactions between two predators and two prey. American Naturalist, 135: 176-204.
Voris, H. K. and Bacon, J. P. (1966) Differential predation on tadpoles. Copeia, 1966: 594-98. Waldick, R. (1997) Effects of forestry practices on amphibian populations in eastern North America. Chapter 21 In D. M. Green (ed) Amphibians in Decline Canadian Studies of a Global Problem. Herpetological Conservation No. 1. Society for the Study of Amphibians and Reptiles, Saint Louis, Missouri. pp. 191-205. 144
Wilbur, H. M. and Semlitsch, R. D. (1990) Ecological consequences of tail injury in Rana tadpoles. Copeia, 1990: 18-24. Woodward, B. D. (1983) Predator-prey interactions and breeding-pond use of temporary-pond species in a desert anuran community. Ecology, 64: 1549-55.
Population declines and range contraction in Australian alpine frogs William Osborne1, David Hunter1 and Greg Hollis2
ABSTRACT Over a period of six years (1993-1998) we undertook
Most sites where L. v. alpina were still present were at the lower altitudinal limit of the species and almost all
an extensive, broad-scale field assessment of the
were associated with artificial water bodies such as
status of frogs in the Australian Alps.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 Recent dramatic declines and extinctions of amphibians in many countries have raised international concern about the likely causes of the declines (see reviews by Stebbins and Cohen 1995; Phillips 1994).The reported disappearances and 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).
FIGURE 1: Location of main areas with subalpine and alpine environments in mainland south-eastern Australia. We undertook recent surveys for frogs (1995-1998) in the following locations: Bimberi Range, Snowy Mountains, Davies Plain, Buffalo Plateau, Bogong High Plains, Dargo High Plains, and Baw Baw Plateau. The mountainous region to the east of Mt Buller has not been subject to recent surveys. Some symbols overlap several sites.
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 in wildlife atlas records for some species in recent years. In this paper, we report on our extensive field surveys that assessed the relative abundance and distribution of four threatened subalpine and alpine species: Litoria verreauxii alpina (Alpine Tree Frog), Pseudophryne corroboree (Southern Corroboree Frog), P. pengilleyi (Northern Corroboree Frog) and Philoria frosti (Baw Baw Frog). We also report on the results of long-term monitoring of P. corroboree, P. pengilleyi and P. frosti.
146
METHODS Description of the study area The study was undertaken in the Australian Alps; the mountainous region stretching from immediately west of Canberra in the ACT to Mt Baw Baw in Victoria (Figure 1). Within this region our surveys were restricted to areas above 1000 m altitude, with most surveys being conducted in the subalpine zone (between about 1400-1800 m) and alpine 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,
particularly at higher altitudes, and frequently falls as snow during the cooler months (Costin 1954; Land Conservation Council 1977; Brown and Millner 1989).The physical and biological features of the Alps have been described in detail elsewhere (Costin 1954, 1957; Land Conservation Council 1977; Costin et al. 1979; McDougal 1982; Barlow 1986; Good 1989, 1992; Green and Osborne 1994; Green 1998) and will not be repeated here.
Determining former distribution Lists of specimens held in the Australian Museum, Museum of Victoria and the Australian National Wildlife Collection, Canberra, were obtained and the records noted if locality details appeared to be reliable.The date of collection of each specimen was also recorded, if available, and the altitude of the collection site determined from topographic maps. Information obtained from the Victorian Wildlife Atlas (Department of Natural Resources and Environment, Victoria) was also examined (this included many of the detailed records of L. v. alpina and P. frosti obtained from the field notes of Littlejohn, Watson, Coventry and Malone). The distribution of P. corroboree and P. pengilleyi was mapped previously by Osborne (1989 and unpubl. data), who recorded the distribution of both species in surveys conducted in 1986 and 1987. Detailed surveys of the distribution of P. frosti on the Baw Baw Plateau were also made in 1983 and 1984 by Malone (1985). Malone prepared detailed maps indicating the extent of the populations surveyed. Although there has been no previous effort to assess the distribution of L. v. alpina, Osborne previously made extensive field notes recording the distribution of L. v. alpina in the Snowy Mountains region from 1978 to 1987. In addition, there are many museum records defining the occurrence of many historic locations for this sub-species on the Bogong High Plains in Victoria and in the Kosciusko region.
Surveys of current distribution and relative abundance Following the recommendation of Gillespie et al. (1995), our surveys concentrated on four species of frogs believed to be of concern for conservation at higher altitudes in this region: Pseudophryne corroboree, P. pengilleyi, Philoria frosti and Litoria verreauxii alpina.These species have reasonably well-defined breeding seasons (Pengilley 1971; Green and Osborne 1994; Hero et al. 1991; Hollis 1995) that allowed us to undertake field work at times appropriate for detecting each species. Two approaches to survey were adopted: (1) auditory censuses (call counts) (Osborne 1989; Zimmerman 1994), undertaken during the known calling season of each species; and (2) tadpole sampling (Shaffer et al. 1994), searches of potential breeding pools undertaken by a combination of visual search, spotlighting and dip-netting. Detailed descriptions of survey techniques are given elsewhere: Osborne (1989, 1991) (P. corroboree and P. pengilleyi), Hollis (1995) (P. frosti) and Hunter et al. (1998) (L. v. alpina). We conducted an extensive, broad-scale survey throughout the known distribution of each species.The surveys were conducted over five summers between September 1994 and March 1998. Surveys at Mt Baw Baw were conducted by Hollis and Hunter and surveys elsewhere in Victoria and NSW
were conducted by Hunter and Osborne. We surveyed most known historic sites (locations obtained from museum and literature records, and from field notes of reliable observers) as well as hundreds of new sites. At each site, details of the location and the number of pools, seepages or creek-lines surveyed were recorded. At sites where frogs were found the number of frogs and the number of pools containing frogs or tadpoles were recorded together with descriptive details of the habitat. Potential breeding sites were visited during the day (Pseudophryne spp., Philoria) and early evening (all species). No surveys were conducted during cold or very wet weather. Pseudophryne corroboree and P. pengilleyi were surveyed specifically by their calling response to a loud human shout (Osborne and Hunter unpublished observations; see also Osborne 1989, 1991).This technique was also used with P. frosti to confirm the calls of individuals calling sporadically. Surveys for P. frosti involved listening for their advertisement calls (Hollis 1995). Litoria v. alpina was surveyed at night by listening for its characteristic call (Smith 1998). Calls of individuals at every location where calling was heard were tape-recorded and compared to calls of known L. v. alpina (Smith pers. comm.). At some sites the males calling could be counted individually, but where abundance was high, an estimate was made of the number calling. During later analysis of the data the following groupings of numbers calling were used: 1-5, 6-10, 11-25, 26-50, 51-100, greater than 100. Tadpoles (L. v. alpina) were sampled with a dip-net or by direct visual count. Each pool in a potential breeding site was searched visually and by dip-netting in daylight, or was checked at night using a small spotlight. If no tadpoles were observed, ten sweeps were made with a dip-net. Each sweep was made along the bottom of the pool, and through any aquatic vegetation, to collect hidden tadpoles. Identification was made in the field or, where identification was uncertain, several individuals were collected and raised to metamorphosis. Surveys for L. v. alpina and both species of Pseudophryne were conducted throughout the Snowy Mountains, Fiery Range and Brindabella Range in New South Wales. In the ACT and adjacent areas of NSW several locations were surveyed along the Bimberi Range and Brindabella Range (specifically Ginini Flats, Snowy Flats, Blackfellows Gap, Rolling Ground Gap, Leura Gap, Mt Bimberi, and Brumby Flats). In Victoria, surveys were conducted at Davies Plain, Mount Buffalo and on the Bogong High Plains.
Long-term monitoring Three species were subject to long-term monitoring: P. corroboree and P. pengilleyi have been monitored at varying intervals since 1986 by Osborne and Hunter (1986-1995); Philoria frosti has been monitored annually for the last five years (1993-1997) by Hollis, and a decade earlier for two years (1983-1984) by Malone (1985).There was no monitoring of L. v. alpina. In most cases monitoring sites were established subjectively at strategic locations throughout the known range of each species. Because of the rugged and remote nature of much of the study area, it was necessary to locate many monitoring sites within walking distance of nearby roads and walking trails.
147
FIGURE 2: Former and current distribution of Pseudophryne corroboree. Open circles, sites where P. corroboree were recorded by Osborne (1989) during the period 1986-1987; closed circles, sites where the species was recorded by Osborne (1989) and still occurs; closed square, additional breeding sites at which the species was detected during recent surveys (1995-1998). Some symbols overlap several sites.
still present at most of these sites (Figure 3) (he was unable to find the frogs in the vicinity of Hume Sawmill at the northern extremity of its range). Osborne (1989) also found the species to be widely distributed and common throughout the Fiery Range and Bogong Mountains (Figure 3). 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
Fifteen sites were monitored for P. corroboree, eight sites for P. pengilleyi and 24 for P. frosti. For convenience, sites are grouped into the following sub-populations: P. corroboree — northern Snowy Mountains (7 breeding sites), and southern Snowy Mountains (8 sites); P. pengilleyi — Fiery Range (2 sites), Brindabella Range (3 sites) and Bimberi Range (3 sites); P. frosti — plateau (21 sites), escarpment edge (4 sites).
RESULTS Former distribution Pseudophryne corroboree Prior to the detailed survey undertaken by Osborne (1989) there were museum records of P. corroboree from only seven 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
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, Robertson and Malone pers. comm.), Mt Buller region (3), Mt Cobberas (1), Mt Hotham (3) and Mt Wellington (1).The Australian Museum provided records from the Kosciusko region (Charlottes Pass, Lake Albina, Etheridge Range, Rams Head Range, Perisher and Smiggin Holes). We also obtained field note records of the subspecies from 41 locations in the
FIGURE 3: Former and current distribution of Pseudophryne pengilleyi. Open circles, sites where P. pengilleyi were recorded by 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.
148E
Snowy Mountains (Osborne, field notes) (Figure 5).There are also field records of L. v. alpina from four high altitude locations in the ACT (Rolling Ground Gap, Leura Gap and Ginini Flats, and Snowy Flats) (Osborne, field notes). Based on these records it is obvious that L. v. alpina was widespread and abundant throughout much of the high country of south-eastern Australia.
FIGURE 4: Former and current distribution of Philoria frosti breeding aggregations on the Baw Baw Plateau. Closed circles, breeding sites where P. frosti still occurs, open circles, sites previously found by Malone (985) to have P. frosti present, but which no longer support the frogs (for details see Hollis 1995, 1998). Some symbols overlap several sites.
Current distribution and abundance Pseudophryne corroboree A total of 170 potentially suitable breeding sites were surveyed for P. corroboree. During the four summers of survey work P. corroboree was detected at 63 different sites.These 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 eastern edge of the former distribution (Figure 2) and only a single specimen was found at low-altitude sites near Tooma Dam in the northern Snowy Mountains.The central portion of the former range, in the region just north of Mt Jagungal (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 were recorded in 1984 compared with 19 in 1993. We (Hollis 1995) only recorded 2.2% and 2.1% of the number of calling males recorded by Malone in 1983 and 1984 respectively. 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 montane forest in a catchment running north (Figure 4).
FIGURE 6: Changes in the relative abundance of Pseudophryne. corroboree and P. pengilleyi over a 12 year period (19986/1987, open columns, and 1996/1998, closed columns). (a) P. corroboree; (b) P. pengilleyi, Brindabella – Bimberi Range population; (c) P. pengilleyi, Fiery Range population.
(a)
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 recorded at only seven locations in NSW.The subspecies was only detected at three high altitude locations, all in Kosciusko National Park — the highest sites were at Charlotte Pass Village, 1780 m, and near Jacky’s Lookout, 1760 m. All other sites with the sub-species were between 1200m and 1500m altitude. Despite surveying a number of historic locations on the Kosciusko Main Range (Figure 5) no frogs were located in the alpine zone. Of the seven locations where extant populations of L. v. alpina were located in Kosciusko National Park, four were associated with artificial water bodies (see Hunter et al. 1998 for details).These ranged from small dams to reservoirs. A similar absence of tadpoles of L. v. alpina from small pools was also noted during extensive surveys in the Snowy Mountains that we conducted for P. corroboree during January and February 1997. No L. v. alpina were found during our surveys at Baw Baw Plateau, Davies Plain, and Bogong High Plains (Hunter and Osborne unpubl. data; Figure 7) in Victoria. We did, however, locate several small populations to the south-east of Mt Hotham near Dinner Plain (altitudinal range 1300 to 1600 m), and found a more extensive population on the Dargo 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 Ginini Flats and Snowy Flats during spring 1996 and no frogs were recorded (a single L. verreauxii was heard calling at Snowy Flats in November 1995, however the specimen could not be found to determine sub-specific status). During 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 Whistling Tree Frog (L. v. verreauxii) was heard calling near Coree Flats, and tadpoles were subsequently found at this site.
Pseudophryne pengilleyi Long-term monitoring of P. pengilleyi was only undertaken in the Brindabella Range. Only one population, Ginini Flats — a subalpine site (1600 m) in the ACT was subject to annual monitoring. Less-regular monitoring was undertaken at Coree Flats (980 m) in NSW. Numbers present at Ginini Flats declined substantially during the first few years of monitoring and have remained low ever since (Table 1). By contrast, the Coree Flats population has supported a larger number of calling males (at least during the years surveys were carried out,Table 1). However, monitoring at Coree Flats commenced after the major drop in numbers had occurred at other sites (Table 1). Earlier collecting and observations by Pengilley (1966 and pers. comm.) at this site indicated that the adult frog population was very large (perhaps well over 500 individuals).The low numbers detected in 1998 are likely to be a direct response to the extreme drought conditions prevailing during the breeding season. Philoria frosti Twenty-four sites surveyed on the Baw Baw Plateau originally by Malone (1985) were re-surveyed annually for the numbers of calling males present (Table 2). At all sites there has been a very large reduction in the numbers recorded. The mean number of calling males recorded at 24 sites
FIGURE 7: Location of sites surveyed for Litoria verreuxii alpina on the Bogong High Plains in 1996 and 1997. Closed circles, sites where L. v alpina were found; open circles, potential breeding habitat surveyed but no frogs of this subspecies found; open squares, sites of historic occurrence determined from examination of museum specimens, but not surveyed during this study. Some symbols overlap several sites.
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
– 12 – 35 – – – 12 – – – – 0 3 0 0 0 2 1 7 0 0 0
– – – – – – – – – – – – – 2 0 5 2 2 0 0 0 0 0
– 8 0 13 1 – – – – – – – – 0 0 0 0 0 0 0 0 0 0
– 0 0 3 2 – – – – – – – – 3 0 0 0 0 0 2 0 0 0
– 1 0 2 7 – – 6 20–50 0 – – 0 3 0 0 0 0 0 4 0 0 0
– 5 0 10–20 – – – – 20–50 – 20–50 – 0 0 0 0 0 0 0 12 0 0 0
– 8 0 0 1 – – – 6–10 – 20–50 – 0 0 0 0 0 0 0 3 0 0 0
10–15 0 0 0 0 30 3 1 8 0 20–50 2 0 0 0 0 0 0 0 3 0 0 0
14 0 0 0 1 44 7 2 7 0 0 0 0 0 0 0 0 0 0 1 0 0 0
10 0 0 0 0 32 9 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
19 0 0 1 0 95 13 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
140 –
65 50–100
– –
25 –
28 –
32 –
Pseudophryne corroboree Maragle Range Round Mountain 1 Round Mtn 2 Ogilives Ck 1 Ogilives Ck 2 Dargals Jagumba Fire Trail Toolong Range 1 Toolong Range 2 Mt Jagungal 1 Mt Jagungal 2 Mt Jagungal 3 Happy Jacks Plain Guthega Blue Cow Link Road Link Road Pipers Ck 1 Pipers Ck 2 Pipers Ck 3 Pipers Ck 4 Pipers Ck 5 Smiggin Holes
15 20–50 0 0 7 – 15 14 50–100 5 >100 – 5–10 – 50–100 5 100–500 10–20 50–100 – – – 20–50 – 1 0 5 0 4 0 7 5 2 0 11 4 6 0 – – 1 0 2 0 1 0
Pseudophryne pengilleyi Ginini Flats (1600m) Coree Flats (980m)
500+ 100–500 – –
declined from 124 individuals in 1983 to between 1.5 and 3.3 individuals over the five years 1993 to 1997 (Table 2). In 1983-1984 the maximum population size recorded at a monitoring site was 667 individuals, by contrast between 1993 and 1997 the largest population recorded was 41 individuals. In 1997 frogs were recorded at only six of the 25 monitoring sites (Table 2). Monitoring of newly-discovered sites in montane forest has only been recently commenced and is not reported here in detail.
DISCUSSION On a broad scale, our surveys provide compelling evidence that at least three taxa of high-altitude frogs (Litoria verreauxii alpina, Pseudophryne corroboree, and Philoria frosti) have suffered serious population declines. A fourth species, Pseudophryne pengilleyi, has declined at higher altitudes (above 1400 m) but remains common at montane altitudes in the Fiery Range.These results confirm previous concern about the conservation status of these endemic taxa (e.g. Gillespie et al. 1995;Tyler 1997). As with frog declines in other parts of Australia, there does not appear to be any phylogenetic relationship between the species that have declined and those that have not declined (e.g. Richards et al. 1993; Mahony 1996).The frogs in decline in the highlands of south-eastern Australia do not comprise a natural taxonomic group; they include one hylid (L. v. alpina) and three myobatrachids (P. corroboree, P. pengilleyi, P. frosti).
152
10 54 100–500100–500
24 15 50–100 50–100
5–10
Only P. corroboree and P. pengilleyi are closely related (Osborne and Norman 1991; Osborne et al. 1996b). A feature of all declining taxa in the high-country is that they breed in lentic situations such as in shallow pools or seepages. This is in contrast to other declining frog species in Australia, most of which are riverine species (e.g. see Gillespie and Hines 1999).The range of pool types varies considerably: P. 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 1990); P. pengilleyi breeds in bog pools at higher altitudes and in shallow seepage pools in gullies at lower altitudes (Osborne 1990); L. v. alpina breeds in deeper pools, which include fens, stream cut-offs, lakes and reservoirs (Hunter et al. 1998). Similarly, the range of life-history strategies employed varies across these species.Three of the species produce relatively small clutches of eggs (P. corroboree, P. pengilleyi, P. frosti), a demographic feature that may be important in their response to changing environmental conditions. All have a tadpole stage; the tadpole of P. frosti is non-feeding, depending instead on a large supply of yolk, tadpoles of both species of Pseudophryne develop within the egg capsule and do not hatch until the terrestrial oviposition site floods in winter or the following spring, L. v. alpina lays its eggs in pools and the free-swimming larvae hatch within a few days of laying and the tadpoles complete development in the pools.Thus, no single aspect of the field biology of these frogs stands out as a feature in common that may help to explain the declines.
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 2. Access Road 2 3. Access Road 3 4. Chairlift 5. Village Flat 6. Neulyne Plain 7. La Trobe Plain 8. Macallister Plain 9. Pudding Basin 10. Moondarra Flat 11. Baragwanath Flat 12. Currawong Flat 13. Currawong Flat 14. Creek Corner 15. Creek Corner A 16. Tanjil Plain 17. East Tanjil 18. McMillians Flat 19. The Morass 20. Freeman’s Flat 21. Wombat Flat 22. Tyers River 1 23. Tyers River 2 24. Mustering Flat 25. Gwinear Flat
43 30 0 6 183 24 206 82 101 225 167 536 174 52 9 120 71 64 667 41 – 14 7 57 93
21 26 0 8 149 – – – – – 245 – 231 49 12 – – – – 104 18 9 3 0 2
2 0 0 0 3 2 3 5 2 0 11 8 2 0 – 9 1 0 30 0 1 0 0 0 0
3 0 0 0 0 2 2 2 1 0 4 4 3 0 – 3 0 0 26 0 – 0 0 0 0
1 0 0 4 0 1 0 0 1 0 4 5 0 2 5 3 0 0 10 0 – 0 0 0 0
2 0 0 0 0 0 4 1 0 0 0 2 0 4 0 4 2 0 23 0 – 0 0 0 0
0 0 0 0 0 0 3 0 0 0 1 0 0 0 1 1 0 0 41 0 23 0 0 0 0
M M M M P P P P P P P P P P P P P P P P P P P P P
Total count for the 13 sites monitored annually
867
847
18
10
11
6
1
Do the declines represent natural populations fluctuations or real declines? There is debate as to whether the population declines observed in amphibians in recent years represent temporary population fluctuations in response to variation in seasonal conditions, particularly precipitation, or are real catastrophic declines (Blaustein 1994; Pechmann and Wilbur 1994). Determining whether an observed decline conforms to one or the other of these situations ideally requires long-term population monitoring and an understanding of the species demography (Pechmann and Wilbur 1994). In the last few years we have undertaken detailed ecological studies of each of the declining species that allow us to now consider this question. Population declines in alpine and subalpine frogs in south-eastern Australia have continued over a long enough period (greater than a decade) that we can speculate that they are no longer related to normal cycles in weather patterns linked to the Southern Oscillation.The populations have shown no sign of recovery despite there being a lengthy period of more favourable weather conditions (1988-1996) (Osborne and Davis 1997). On a broader scale, it is obvious that the declines have also been extensive in a biogeographical sense.Three species (L.v. alpina, P. corroboree and P. frosti) have undergone partial contractions of their geographic ranges. In some situations the frogs have completely disappeared from some areas separated by major biogeographic barriers (e.g. deep
montane valleys) that would prevent re-establishment of populations by means of normal dispersal.These observations strongly suggest that the observed declines in these alpine frogs are not typical of what would be expected from normal population fluctuations such as would result from decreased recruitment or local extinctions (e.g. Semlitsch et al. 1996). Moreover, the declines are similar to frog declines elsewhere that have been inferred from comparisons of historic and current distributions (Gillespie and Hollis 1996; Osborne et al. 1996a; Fisher and Shaffer 1996). The concurrence of population declines in areas as widely separated as the Baw Baw Plateau and Snowy Mountains indicates that the causes are not likely to be localised. We hypothesise that the factors causing the declines are due to factors operating on a broad rather than a local scale. Possible causes are discussed below.
Is there a global influence that affects high-altitude populations? Existing hypotheses that have been suggested as contributing to high-altitude population declines include unusual weather patterns (Corn and Fogleman 1984; Pounds and Crump 1994), acid precipitation (Dunson et al. 1992), increased UV-B radiation resulting from ozone depletion (Blaustein et al. 1994b), deposition of pesticides (Colborn and Clement 1992) and virulent disease (Laurance 1996; Berger et al.
153
1999).There is no evidence to date that acid precipitation is a problem in most parts of the Australian Alps, however, there is no information available on the levels of possible contaminants in precipitation and dust in this region. The catastrophic disappearance of organisms frequently indicates the action of a particularly virulent pathogen. Laurance et al. (1996) proposed that a water-borne disease may have caused declines and extinctions in populations of stream-breeding montane frogs in the wet tropics of Australia.They further postulated that the pathogen was a virus that occurred at a range of elevations but became highly virulent only in cool upland habitats. More recently, a fungal pathogen has gained some favour as a possible cause of ongoing declines in some tropical and temperate Australian frogs (see Berger et al. 1999). Uncertainty in identifying the likely causal agent emphasises the need for further research on populations of frogs in Australia. At least some declines have been attributed to changes in long-term weather patterns (Osborne 1989; Pounds and Crump 1994; Stewart 1995; Osborne et al. 1996a), however other studies have ruled-out the possibility of changes in climate being a possible cause of catastrophic declines (Laurance 1996). High-elevation regions, particularly at high latitudes, have experienced increasing solar UV-B radiation (Broomhall 1998) caused by depletion of the ozone layer. Because potentially harmful intensities of UV-B can penetrate several metres in clear freshwater (Schindler et al. 1996), well below the depths used for egg laying and tadpole development, there is growing concern that declines in alpine species may relate to increased UV-B radiation (Blaustein et al. 1994b), particularly given the increases in ultraviolet radiation experienced in alpine areas (Bluthaler and Ambach 1990). A recent study conducted in the Snowy Mountains (Broomhall et al. in press) showed that developing embryos of L. v. alpina are significantly more sensitive to ambient UV-B radiation than a non-declining species, Crinia signifera. In artificial pools at high altitudes L. v. alpina experienced very high mortality rates unless shielded from UV-B. In the discussion below we consider in more detail the possibility of prolonged dry weather preventing breeding or dehydrating frogs over a period long enough to have caused a complete disappearance from most of the region. Consideration of the hypothesis that alpine frogs have suffered adversely from increased levels of ultraviolet radiation is discussed elsewhere in this volume (Broomhall et al. in press).
Is there a link between long-term weather patterns and frog declines? Given the reliance of amphibians on moist environments for both physiological maintenance and reproduction (Duellman and Trueb 1986) and the increasing concerns about global climate change, it is not surprising that climate has been implicated as a possible factor driving amphibian population declines in a number of species through out the world (see above references). Several long term studies have correlated population fluctuations in amphibians with climatic variables, particularly annual variation in rainfall patterns (Stewart 1995; Pechmann et al. 1991).The direct effect of climate on amphibians may include desiccation of both the larval and adult
154
phases, while indirect effects include prevention of breeding activity and the lowering of the immune system making the frog more susceptible to pathogens (Pounds and Crump 1994). Despite the regions high precipitation, marked annual fluctuations are a feature of the long-term rainfall pattern in the southern Australian Alps, and moderately severe droughts have occurred in this area previously. Records of weather measured since late last century in the Snowy Mountains indicate that annual precipitation has fluctuated between periods marked with both drier and wetter conditions (Osborne and Davis 1997).These oscillations generally relate to variability in the Southern Oscillation Index (McBride and Nicholls 1983) and overall there has been no long-term trend of declining precipitation.The last extended dry period (1979-1987) coincides with the period when the frogs declined. Osborne and Davis (1997) noted that features of the 1979-1987 dry period that distinguished it from earlier extended dry periods identified at Kiandra included that it was two years longer and the mean annual precipitation during this period was slightly lower (less than 10 %). Annual maximum temperatures were above average for six of the nine years in this period (based only on the 34-year record at Cabramurra in NSW).These trends are also evident in data obtained for the Victorian highlands and Erica near the Baw Baw Plateau (Smith et al. in prep.). Despite these observations, we consider that it is unlikely that such slight changes in the long-term weather patterns could result in frog declines on the scale observed; especially because the frogs were observed to be in much higher population levels after earlier droughts.The possible interaction of lower precipitation with increased atmospheric temperatures is considered below. Global warming has been considered a threat to high altitude species (e.g., Busby 1988; Bennett et al. 1991; Brereton et al. 1995), however in south-eastern Australia there has been only a very slight increase in annual maximum temperatures since 1951 (less than 0.1°C). Osborne and Davis (1997) found no appreciable trend in the annual temperature at Cabramurra in the Snowy Mountains (the longest temperature record available). By contrast, Davis (1998) notes that there has been a very slight increase in mean winter maximum and minimum temperatures (May–Oct) since the 1970’s, however this increase represents about 0.5°C (calculated from data in Davis 1998). Particularly warm years occurred both prior to, and during, the period of the present declines. The species in the group that would be expected to be most sensitive to a deterioration in long-term weather patterns would be those with a prolonged larval development period that might be affected by early pond-drying, or lowered water tables. In both P. corroboree and P. pengilleyi, the premetamorphic stage usually lasts for ten to twelve months. Philoria frosti may also be sensitive to reduced moisture availability because of its habit of breeding in very small depressions in seepages. Possible response to changing weather patterns is considered further below. If alpine frogs were responding to drier and warmer conditions it is likely that they would have disappeared earlier from the lower-altitude, drier margins of their range. This feature is in fact partially evident for two species,
P. corroboree and P. frosti, but the reverse is true for P. pengilleyi and L. v. alpina. Osborne and Davis (1997) noted that the decline in P. corroboree appears to have been progressive in its effect, occurring first in areas with lowest predicted rainfall (such as on the far eastern edge of the distribution) and most recently in the wetter parts of the species range (Mt Jagungal,Toolong Range and Round Mountain). Interestingly, the largest remaining population (in the Dargals Range) occurs in an area predicted to be the wettest site. All low-altitude populations (below about 1400 m) have disappeared, but so too have most of the highest altitude populations (above 1700 m) (such as at the many former sites near Mt Jagungal) (Osborne and Davis 1997; Hunter and Osborne unpubl. data).The situation with P. frosti is somewhat different. Although the species has declined and disappeared from the drier eastern end of its former distribution (Hollis 1995; Figure 4), on the western side of the plateau, it has also declined at higher altitudes with the largest remaining breeding aggregations occurring in dense montane forest at the lower limit of the species distribution. Both P. pengilleyi and L. v. alpina have also shown greatest decline at the highest altitudes, sites that are also exposed to significantly higher precipitation. It is not at all clear whether there is a relationship between the extent of population declines in alpine frogs and long-term trends in precipitation.The potential influence of climatic change requires further analysis, and underscores the need for continued monitoring of these highly threatened species.
CONCLUSION Our surveys indicate that in the highlands of south-eastern Australia four species of frogs have experienced pronounced population declines.These frogs are Litoria verreauxii alpina (Alpine Tree Frog), Pseudophryne corroboree (Southern Corroboree Frog), P. pengilleyi (Northern Corroboree Frog) and Philoria frosti (Baw Baw Frog). Pseudophryne pengilleyi is still widespread and abundant at lower altitudes, but there are few remaining substantial populations of the other three species, which are faced with the likelihood of extinction in the short term if the current trends continue. There appears to be an altitudinal influence on the extent of the declines, with the most serious declines all occurring at higher altitudes, particularly in the subalpine and alpine zones. This apparent relationship between altitude and the extent of the population decline has been observed in other frog species in Australia (Richards et al. 1993; Osborne et al. 1996; Mahony 1996) and in other countries (Corn and Fogleman 1984; Bradford 1991; Carey 1993; Fellars and Drost 1993). Tyler (1997) recognised the significance of this altitudinal relationship, and recommended that further research be undertaken into the factors causing it. We support this view and recommend increased multi-disciplinary research effort and urgent experimental management aimed at stemming the decline in these endemic alpine amphibians.
ACKNOWLEDGEMENTS We are particularly grateful to Environment Australia, the NSW National Parks and Wildlife Service, the Victorian Department of Natural Resources and Environment, the Australian Alps National Parks Liaison Committee and the University of Canberra for providing the funding to support
the various components of this study. Maxine Davis assisted with analysis of long-term weather records and Michael Smith confirmed all call records obtained for L. v. alpina. We thank Craig Smith, Ken Green, Graeme Enders, Marjo Rauhala, Brett MacNamara, Graeme Gillespie and Gerry Marantelli for their assistance and logistical support. John Coventry, Graeme Gillespie, Murray Littlejohn, Brian Malone, Peter Robertson, Graeme Watson and John Wombey have provided much encouragement and advice over the years. We also thank members of the Baw Baw Frog Recovery Team and the Corroboree Frog Recovery Team.
REFERENCES Barlow, B.A. (ed), (1986) Flora and Fauna of Alpine Australasia. Ages and Origins. CSIRO, Melbourne. Bennett, S., Brereton, R., Mansergh, I., Berwick, S., Sandford, K. and Wellington, C., (1991) The potential effect of the enhanced greenhouse climate change on selected Victorian fauna. Arthur Rylah Institute for Environmental Research,Victoria. Berger, L., Spear, R. and Hyatt, A.D. (1999) Chytrid fungi and amphibian declines: overview, implications and future directions. Pp 23-33 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Blaustein, A.R., (1994). Chicken Little or Nero’s fiddle? A perspective on declining amphibian populations. Herpetologica, 50: 85-97. Blaustein, A. R., and Wake, D. B., (1990) Declining amphibian populations: a global phenomenon? Trends in Ecology and Evolution, 5:203-204. Blaustein, A.R., Wake, D.B. and Sousa, W.P., (1994a) Amphibian declines: judging stability, persistence, and susceptibility of populations to local and global extinctions. Conservation Biology, 8: 60-71. Blaustein, A.R., Hoffman. P.D., Hokit, D.G., Kiesecker, J.M. Walls, S.C. and Hays, J.B., (1994b) UV repair and resistance to solar UV-B in amphibian eggs: a link to population declines? Proceedings of the National Academy of Sciences USA, 91: 1791-1795. Bluthaler, M. and Ambach, W., (1990) Indication of increasing solar ultraviolet-B radiation flux in alpine regions. Science, 248: 206-208. Bradford, D. F., (1991) Mass mortality and extinction in a high-elevation population of Rana muscosa. Journal of Herpetology, 25(2): 174-177. Broomhall, S., (1998) The implications of ozone depletion for the Australian Alps: a review. Pp 224-247 In K. Green (ed.) Snow. A natural history; an uncertain future. Australian Alps Liaison Committee, Canberra. Brown, J.A.H. and Millner, F.C., (1989) Aspects of the meteorology and hydrology of the Australian Alps. Pp 297-332 in Good, R. (ed.) The Scientific Significance of the Australian Alps.The Australian Alps National Parks Liaison Committee, Canberra. Brereton, R., Bennett, S. and Mansergh, J., (1995) Enhanced greenhouse climate change and its potential effect on selected fauna of south-eastern Australia: a trend analysis. Biological Conservation, 72: 339-354.
155
Broomhall, S., (1998) The implications of ozone depletion for the Australian Alps: a review. Pp 224-247 In K. Green (ed.) Snow a Natural History; an uncertain future. Australian Alps Liaison Committee, Canberra.
Gillespie, G.R. and Hines, H.B. (1999) Status of temperate riverine frogs in south-eastern Australia. Pp. 109-130 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra.
Busby, J.R., (1988) Potential impacts of climate change on Australia’s flora and fauna. Pp 387-398 In G.I. Pearman (ed.) Greenhouse: Planning for Climate Change. CSIRO, Melbourne.
Gilpin, M.E. and Soule, M.E., (1986) Minimum viable populations: Processes of species extinction. Pp 19-34 In M.E. Soule (ed.) Conservation Biology: The science of scarcity and diversity. Sinauer Assoc., Sunderland, Massachusetts.
Carey, C., (1993) Hypothesis concerning the causes of the disappearance of Boreal toads from the mountains of Colorado. Conservation Biology, 7(2): 355-362.
Good, R. (ed.), (1989) The Scientific Significance of the Australian Alps. Australian Alps National Parks Liaison Committee, Canberra.
Caughley, G. and Dunn, A., (1996) Conservation Biology in Theory and Practice. Blackwell, Cambridge, Massachusetts.
Good, R.B., (1992) Kosciusko Heritage. The conservation significance of Kosciusko National Park. National Parks and Wildlife Service of New South Wales, Hurstville, Sydney.
Cody, M.L and Smallwood, J.A. (eds), (1996) Long-term Studies of Vertebrate Communities. Academic Press, San Diego. Colborn,T. and Clement, C. (eds), (1992) Chemically-induced alterations in sexual and functional development: The wildlife/human connection. Princeton Scientific Publishing Co., Princeton, N.J. Congdon, J.D. and Dunham, A.E., (1994) Contributions of long-term life history studies to conservation biology. Pp 181-194 In G.K. Meffe, C.R. Carroll (eds.) Principles of Conservation Biology. Sinauer Assoc., Sunderland, Massachusetts. Corn, P.S. and Fogleman, J.C., (1984) Extinction of montane populations of the northern Leopard Frog (Rana pipiens) in Colorado. Journal of Herpetology, 18: 147-152. Costin, A.B., (1954) A Study of the Ecosystems of the Monaro Region of New South Wales. Government Printer, Sydney. Costin, A.B., (1957) The high mountain vegetation of Australia. Australian Journal of Botany, 5: 173-189. Costin, A.B., Gray, M.,Totterdall, C.J., and Wimbush, D.J., (1979) Kosciusko Alpine Flora. CSIRO/Collins, Melbourne. Crump, M.L., Hensley, F.R., and Clark, K.L., (1992) Apparent decline of the Golden Toad: underground or extinct? Copeia, 1992(2): 413-420. Davis, C.J., (1998) Meteorological aspects of snow. Pp 3-34 In Green, K. (ed.) Snow. A natural history; an uncertain future. Australian Alps Liaison Committee, Canberra. Duellman, W.E. and Trueb, L., (1986) Biology of Amphibians. Johns Hopkins University Press, U.S.A. Dunson, W.A., Wyman, R.L. and Corbett, E.S., (1992) A symposium on amphibian declines and habitat acidification. Journal of Herpetology, 26: 349-352. Fellers, G.M. and Drost, C.A., (1993) Disappearance of the Cascades Frog Rana cascadae at the southern end of its range, California, USA. Biological Conservation, 65: 177-181. Fisher, R.N. and Shaffer, H.B., (1996) The decline of amphibians in California’s Great Central Valley. Conservation Biology, 10(5): 1387-1397.
Green, K. (ed), (1998) Snow. A natural history; an uncertain future. Australian Alps Liaison Committee, Canberra. Green, K. and Osborne, W.S., (1994) Wildlife of the Australian Snow-country: A comprehensive guide to alpine fauna. Reed Books, Chatswood, N.S.W. Hero, J-M., Littlejohn, M. and Marantelli, G., (1991) Frogwatch Field Guide to Victorian Frogs. Department of Conservation and Environment,Victoria. Heyer, W.R., Donnelly, M.A., McDiarmid, R.W., Hayek, LC. and Foster, M.S. (eds), (1994) Measuring and Monitoring Biological Diversity. Standard Methods for Amphibians. Smithsonian Institution Press, Washington. Hollis, G.J., (1995) Reassessment of the distribution, abundance and habitat of the Baw Baw Frog Philoria frosti Spencer: preliminary findings. The Victorian Naturalist, 112: 190-201. Hollis, G.J., (1997) Recovery Plan for the Baw Baw Frog (Philoria frosti). Unpublished report to the Endangered Species Program, Environment Australia, Canberra. Hunter, D.A., Osborne, W.S. and Smith, M.J., (1998) Distribution and abundance of the alpine tree frog (Litoria verreauxii alpina) in the Australian Alps National Parks. Report on the first seasons survey (1996-97). Unpublished Report to the NSW National Parks and Wildlife Service. Applied Ecology Research Group, University of Canberra, Canberra. Land Conservation Council, (1977) Report on the Alpine Study Area. Land Conservation Council,Victoria. Laurance, W.F., (1996) Catastrophic declines of Australian rainforest frogs: is unusual weather responsible? Biological Conservation, 77: 203-212. Laurance, W.F., McDonald, K.R., and Speare, R., (1996) Epidemic disease and the catastrophic decline of Australian rainforest frogs. Conservation Biology, 10: 406-413. Mahony, M., (1996) The decline of the Green and Golden Bell Frog Litoria aurea viewed in the context of declines in other Australian frogs. Australian Zoologist, 30: 237-247.
Gillespie, G.R., Osborne, W.S. and McElhinney, N.A., (1995) The Conservation Status of Frogs in the Australian Alps: A Review. Unpublished Report to the Australian Alps Liaison Committee, Canberra.
Malone, B., (1985) Status, distribution and ecology of the Baw Baw Frog (Philoria frosti). Arthur Rylah Institute Technical Report No. 36. Department of Conservation, Forests and Lands,Victoria.
Gillespie, G.R. and Hollis, G.J., (1996) Distribution and habitat of the spotted tree frog, Litoria spenceri Dubois (Anura: Hylidae), and an assessment of the potential causes of population declines. Wildlife Research, 23:49-75.
McBride, J.L. and Nicholls, N., (1983) Seasonal relationships between Australian rainfall and the Southern Oscillation. Monthly Weather Review 111, 1998-2004.
156
McDonald, K.R., (1990) Rheobatrachus Liem and Taudactylus Straughan and Lee (Anura: Leptodactylidae) in Eungella National Park, Queensland: distribution and decline. Transactions of the Royal Society of South Australia, 114: 187-194. McDougall, K.L., (1982) The alpine vegetation of the Bogong High Plains. Environmental Studies Publication No. 357, Ministry for Conservation,Victoria. Osborne, W.S., (1988) A survey of the distribution and habitats of Corroboree Frogs, Pseudophryne corroboree in Kosciusko National Park: with a reference to ski resort development. Report prepared for NSW National Parks and Wildlife Service, Snowy Mountains Region, Jindabyne. Osborne, W.S., (1989) Distribution, relative abundance and conservation status of the corroboree frogs, Pseudophryne corroboree Moore (Anura: Myobatrachidae): Australian Wildlife Research, 16: 537-547. Osborne W.S., (1990) The conservation biology of Pseudophryne corroboree Moore (Anura: Myobatrachidae): A study of insular populations. Ph.D. thesis, Australian National University, Canberra. Osborne W.S., (1991) The biology and management of the Corroboree Frog (Pseudophryne corroboree) in NSW. Species Management Report Number 8, NSW National Parks and Wildlife Service, Sydney. Osborne, W.S. and Davis, M.S., (1997) Long-term variability in temperature, precipitation, and snow cover in the Snowy Mountains: is there a link with the decline of the southern corroboree frog (Pseudophryne corroboree)? Report to NSW National Parks and Wildlife Service, Snowy Mountains Region, Jindabyne. Osborne, W.S. and Norman, J.A. (1991) Conservation genetics of Corroboree Frogs, Pseudophryne corroboree: population subdivision and genetic divergence. Australian Journal of Zoology 39:285-297. Osborne, W.S., (1998) Draft Recovery Plan for the Southern Corroboree frog (Pseudophryne corroboree). NSW National Parks and Wildlife Service, Sydney. Osborne, W. S., Littlejohn, M. J. and Thomson, S. A., (1996a) Former distribution and apparent disappearance of the Litoria aurea complex from the Southern Tablelands of New South Wales and the Australian Capital Territory. Australian Zoologist, 30: 190-198. Osborne W. S., Zentelis R. A. and Lau, M., (1996b) Geographical variation in corroboree frogs, Pseudophryne corroboree Moore (Anura: Myobatrachidae): A reappraisal supports recognition of P. pengilleyi Wells and Wellington. Australian Journal of Zoology, 44: 569-587. Pechmann, J. F. K., Scott, D. E., Semlitsch, R. D., Caldwell, J. P., Vitt, L. J. and Gibbons, J. W., (1991) Declining amphibian populations: the problems of separating human impacts from natural fluctuations. Science, 253: 892-895. Pechmann, J.H.K., and Wilbur, H.M., (1994) Putting declining amphibian populations in perspective: natural fluctuations and human impacts. Herpetologica, 50: 65-84.
Pengilley, R.K., (1971) Calling and associated behaviour of some species of Pseudophryne (Anura: Leptodactylidae). Journal of Zoology, London, 163: 93-103. Pengilley, R.K., (1966) The biology of the genus Pseudophryne (Anura: Leptodactylidae). Master of Science thesis, Australian National University, Canberra. Phillips, K., (1994) Tracking the Vanishing Frogs. St Martin’s Press, New York. Pounds, J.A., and Crump, M.L., (1994) Amphibian declines and climate disturbance: the case of the Golden Toad and the Harlequin Frog. Conservation Biology, 8(1): 72-85. Richards, S.J., McDonald, K.R. and Alford, R.A., (1993) Declines in populations of Australia’s endemic tropical rainforest frogs. Pacific Conservation Biology: 1: 66-77. Semlitsch, R.D., Scott, D.E., Pechmann, J.H.K. and Gibbons, J.W., (1996) Structure and dynamics of an amphibian community. Evidence from a 16-year study of a natural pond. Pp 217-248 In Cody, M.L and Smallwood, J.A. (eds) Long-term Studies of Vertebrate Communities. Academic Press, San Diego. Schindler, D.W., Curtis, P.J., Parker, B.R., and Stainton, M.P., (1996) Consequences of climate warming and lake acidification for UV-B penetration in North American boreal lakes. Nature, 379: 705-708. Shaffer, H.B., Alford, R.A., Woodward, B.D., Richards, S.J., Altig, R.G. and Gascon, C., (1994) Quantitative sampling of amphibian larvae. Pp130-141 In Heyer, W.R., Donnelly, M.A., McDiarmid, R.W., Hayek, LC. and Foster, M.S. (eds) Measuring and Monitoring Biological Diversity. Standard Methods for Amphibians. Smithsonian Institution Press, Washington. Smith, M.J., (1998) Intraspecific variation in the male advertisement call and morphology of Litoria verreauxii. Bachelor of Applied Science, Honours Thesis. Applied Ecology Research Group, University of Canberra. Spellerberg, I. F., (1991) Monitoring Ecological Change. Cambridge University Press, Cambridge. Stebbins, R.C. and Cohen, N.W., (1995) A Natural History of Amphibians. Princeton University Press, Princeton, New Jersey. Stewart, M.M., (1995) Climate driven population fluctuations in rainforest frogs. Journal of Herpetology, 29:437-446. Trenerry, M.P, Laurance, W.F. and McDonald, K.R., (1994) Further evidence for the precipitous decline of endemic rainforest frogs in tropical Australia. Pacific Conservation Biology, 1: 150-153. Tyler, M.J., (1997) The Action Plan for Australian Frogs. Environment Australia, Endangered Species Program, Canberra. Zimmerman, B.L. (1994) Audio strip transects. Pp 92-97, In Heyer, W.R., Donnelly, M.A., McDiarmid, R.K., Hayek, L.C. and Foster, M.S. (eds) Measuring and Monitoring Biological Diversity — Standard Methods for Amphibians. 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 Marantelli 2 and Ken Green3
ABSTRACT Over the past 18 years the Southern Corroboree
survivorship to metamorphosis of captive-reared and field embryos and tadpoles. Differences in the
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. It involves increasing the size of several small breeding populations by means of experimental field and captive management intending to reduce egg and tadpole mortality. Initially, an assessment of our ability to reduce mortality during these early life-history stages is being undertaken by comparing the level of
INTRODUCTION The Southern Corroboree Frog (Pseudophryne corroboree) is a strikingly marked species restricted to high montane and sub-alpine bog habitats in the Snowy Mountains at altitudes 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 hatch and the tadpoles to move into an aquatic environment (Barker et al. 1995). In the early to mid 1980’s, P. corroboree 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 range of the species (Osborne 1989, 1991). As documented by a long-term monitoring program, P. corroboree has failed to show any signs of recovery since the initial population crash, with many of the monitored populations continuing to decline to extinction (Osborne 1998; Osborne et al. 1999). In 1996 an endangered species recovery team was established to implement the first phase of the recovery process as outlined in a draft Recovery Plan for this species (Osborne 1996), and in 1997 P. corroboree was assessed as being critically endangered using the 1994 IUCN criteria (Tyler 1997). A primary aim of endangered species recovery programs is to identify the processes threatening the particular species so that threat abatement can then be addressed in the development and implementation of recovery actions (Dickman 1996). As such, the uncertainty behind the factors causing the decline in P. corroboree has proven problematic for the recovery process. Extensive surveys conducted over the past few years have failed to locate P. corroboree at 85% of known historic localities (Osborne et al. 1999). Furthermore, 77% of extant populations consist of fewer than five calling males while only three populations consist of more than 15 calling males (Hunter and Osborne unpubl. data).These results are particularly concerning given that the long term monitoring program has demonstrated a high propensity for populations consisting of fewer than five calling males to become extinct within a few years (Osborne 1998). Apart from the likelihood that the factors which initiated the decline in P. corroboree populations are still operating, the detrimental effects associated with small population size may also be an important process working against the recovery of this species. In an attempt to aid the recovery of small populations of P. corroboree, a population augmentation project (direct manipulation of recruitment through to the terrestrial frog stage to increase the adult population size) was commenced in 1997.There are two aims to this: 1. to assess whether it is possible to increase recruitment through to the metamorphic stage via a combination of captive-rearing and active field management to prevent tadpole mortality in the field; and 2. to determine whether an increase in the level of recruitment through to metamorphosis will result in an increase in the breeding adult population size in small remnant populations of P. corroboree. While population repatriation programs involving captive breeding or rearing have been undertaken for a number of bird and mammal species (for a review see Griffith et al. 1989), there have been considerably fewer projects involving amphibians (for a review see Dodd and Seigel 1991). Such projects are often received favourably by the public. However, their limited success has led a number of authors to question the value of repatriation projects as a conservation tool (Dodd and Seigal 1991; Snyder et al. 1996).This limited success has been largely attributed to a failure to remove the processes that caused the initial decline and a lack of knowledge of the biological requirements of the species. This was particularly well demonstrated with the Natterjack Toad (Bufo calimita) recovery program in England where
substantial success at re-establishing populations at extinct sites, and enhancing small populations, was only achieved after knowledge of the breeding site requirements and the factors which caused the decline in this species were identified and mitigated (Denton et al. 1997). Further criticisms have also targeted poor experimental design and lack of adequate follow-up monitoring to assess the success of repatriation projects (Hein 1997). Given the difficulty in determining the causal factors of recent frog declines (refer to papers in this symposium), it is unlikely that the reasons behind the decline in P. corroboree populations will be identified prior to further local extinctions (see Osborne et al. 1999 for a better discussion of hypotheses relating to declines in alpine frogs).There is a very real possibility of the species becoming extinct in the near future. This situation was the impetus behind the Recovery Team for P. corroboree supporting a population augmentation project. Furthermore, information obtained on the demography of P. corroboree should aid both management and further efforts to determine the causal factors of decline in this species. In this paper we present the methods employed in the project, discuss its limitations and the relevance of the project to the recovery process for P. corroboree, and present the results of the first year of this study.
METHODS Choice of sites At the outset of the study in March 1997 we were faced with considerable logistical constraints that limited the number of populations available as experimental populations. Only three populations were deemed suitable for population augmentation during the first year of this project.These included one population in the Dargal Range which had 32 calling males (this site will be referred to as Site A), one population from the Jugumba Range which had 13 calling males (Site B) and one population near Round Mountain which had two calling males (Site C).The exact location of these sites will not be disclosed in this paper.The number of calling males refers to the number of males detected during the 1997 breeding season. While Site A would not be considered a small population in comparison with other extant populations of P. corroboree, it was included so as to increase the sample size of clutches for assessing whether we are capable of artificially reducing embryonic and tadpole mortality. After this process, which is expected to take two to three years, Site A will probably be removed from the experiment. As such, it was not deemed necessary to manipulate every clutch at this site and so a limit of ten clutches was allocated for egg collection. Control populations were chosen by randomly selecting sites from a list of other known extant remnant populations.
Field procedures In January 1997 at each breeding location the calling sites of males were marked with flagging tape.These calling sites were then inspected in early March (after breeding had finished) in order to collect half of each clutch for captive rearing at the Amphibian Research Centre (ARC) in Melbourne. All eggs found were removed and placed in a clean polypropylene container (Genfac Plastics) moistened
159
with water, and the numbers of fertile eggs, dead eggs and empty capsules were counted. Half of the viable eggs were replaced in the oviposition site and the nest carefully returned to its original condition. In order to monitor the survival of tadpoles from the half clutches left in the field, plastic enclosures (1.0m x 0.5m x 0.4 m) were constructed to link each oviposition site with an enclosed portion of the adjacent experimental pool.The bottom and top were open to allow access to the pond substrate, and to allow light and precipitation to enter the enclosed part of the pool.The base of each enclosure was embedded deeply in the silt and moss in the pool, and extended into the moss bank beyond the position of the hidden eggs.Two walls of the enclosure had large, rubberised fibreglass mesh-covered openings to allow for water movement between the enclosure and the rest of the pool. All materials used were inert and resistant to ultra-violet radiation.To investigate the effect of the field enclosure on the water temperature, Hobo data loggers (Onset Computer Corporation) and Tinytag data loggers (Gemini Data Loggers, Chichester, England) were placed inside and outside enclosures in a number of experimental pools. Prior to snow fall in June 1997 the enclosures were checked regularly to determine whether the eggs left in the field had hatched. A sub-sample of experimental nest sites was also checked for signs of excessive embryonic mortality.
Captive husbandry The eggs removed from the field were placed in small plastic containers packed with sphagnum moss and immediately transported by car to the ARC (an approximately six hour drive). Upon arriving at the ARC, the eggs were removed from containers, further divided and placed into either empty clean containers or containers containing natural nest material.These containers were each floated inside a slightly larger container containing sterilised water which in turn was placed in a water-bath.The eggs were maintained in a moistened terrestrial state, with the contact of each container within the water bath providing thermal profiles similar to that which might be expected in natural nesting sites. The entire system was enclosed within a controlled temperature room (a modified shipping container) with ambient temperature set to approximate recorded field temperatures based on data collected from Site A using two Hobo data loggers in winter 1996.The 1996 recordings did not start until the beginning of June so that temperatures in the container up until June 1997 were based on spot temperature records from the field and updates from the field data loggers when they could be accessed. On hatching the tadpoles were moved to glass aquaria (1 000 X 300 X 750mm) containers within the constant temperature room.Tadpoles from each pool were housed together and each clutch was separated by screens made from anodised aluminium and rubberised fibreglass mesh. To mimic the field environment, individual tadpole enclosures 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
no more than 50 in each 180 litre tank. A drip irrigation system provided for water exchange at a rate of approximately 20% per day.
Post-winter tadpole monitoring After snow had melted in mid-spring the captive-reared tadpoles were returned to the field where they were released into field enclosures (as described above) within their natal pools.To allow for a direct comparison of survivorship within clutches, all tadpoles released into enclosures represented individuals from the same clutch. As a precaution against the possibility of high mortality occurring during the release process, the captive-reared tadpoles were released in two batches. Each batch represented half the tadpoles collected from each experimental clutch.The two batches were released one month apart so that the first batch could be monitored for signs of failing to adjust to the field environment. Field tadpoles, early release tadpoles and late release tadpoles were housed in separate enclosures. For assessing survivorship differences between the captive-reared and field-reared tadpoles after snow melt, a census of tadpole numbers in each enclosure was conducted once a fortnight until metamorphosis in mid summer.Tadpole counts were conducted at night by torch light.The number of tadpoles in an enclosure was assessed by dip-net removal until five minutes had elapsed without further tadpoles being found. During each census, Gosner stage (Gosner 1960) and snoutvent length were recorded for five randomly chosen tadpoles from each enclosure, or fewer if fewer tadpoles were present. Chi-squared contingency tests were used to test for significant differences between mortality levels both between the treatments within sites, and between sites within treatments. A two-way ANOVA was used to compare differences in the size of tadpoles at Gosner stage 41 for both between-treatments and between-sites.To meet the assumption of homogeneity of variance, the measurements were log transformed and three outliers were removed.Two analyses were undertaken.The first was to test for size differences between treatments and involved removing Site B from the analysis as this site lacked the field treatment due to 100% winter mortality.The second was conducted to test for differences in tadpole size between sites and involved removing the field treatments of Site A and C from the analysis, again due to the lack of a field treatment at Site B.
Disease prevention protocol A strict set of protocols was maintained for the purpose of minimising the potential spread of pathogens both from the captive environment to the field, and between experimental sites. All equipment used at the experimental sites was either sterilised prior to use or was a new item that had never been used in the field or with captive frogs. Prior to undertaking fieldwork in the area of the experimental sites footwear was sterilised in bleach. Further precautions were taken at Site A where, prior to entering the sites, footwear were changed for gumboots which were housed adjacent to the site.
To prevent contamination of ex-situ work, this project was maintained in isolation from the rest of the frog collection at the ARC. It had an independent water supply, a set of dedicated equipment and was housed at a separate site. Only new, sterilised or field sourced materials were taken inside the room, and a routine of showering, changing into clean clothes and dedicated boots was practised before entry. As a further precaution before release, 10 tadpoles (one from each of the 10 natal pools represented in the study) were killed, preserved in 10% neutral buffered formalin and sent to the CSIRO Australian Animal Health Laboratory for examination.This material was sectioned, stained and examined for signs of disease. Only seven of the tadpoles could be fully examined.These and the available sections of the other three showed no evidence of disease (Berger pers. comm.).
Maintenance of pool water levels The first year of this project coincided with the drought during the spring/summer of 1997/98.To avoid the drying of pools and subsequent loss of tadpoles, a number of techniques were used to maintain water in pools at the experimental sites. At Site A, both a solar and petrol water pump were used to pump water from a nearby stream into a large polyethylene tank. From this tank the water was then fed into the various pools that contained experimental enclosures by gravity.The breeding pools at Site B are slow flowing seepage lines which allowed adequate water levels to be maintained by damming the out-flow end of the pool. At Site C, gravity-fed water from a nearby stream maintained an adequate water level in the single breeding pool with enclosures.
RESULTS Comparison of survivorship between field and captive treatments Collection of embryos from the field took place on 15 and 16 March, 1997, approximately six weeks from the end of the breeding season. Of the 25 male nests investigated at the three sites, 16 contained eggs (11 from Site A, 4 from Site B and 1 from Site C). From these 16 nests, 374 eggs were collected for captive rearing while 324 eggs were left in the male nests for the field comparison. The greatest level of field mortality at all three sites occurred during the over-winter stage, with Site B experiencing total mortality (Figure 1). On the other hand, the greatest level of mortality for the captive-reared tadpoles occurred during the post-winter tadpole stage (Figure 1). After adjusting for the number of individuals removed from the captive stock for pathogen screening and further captive husbandry research, the total number of animals which survived from collection to metamorphosis was significantly higher for the captivereared tadpoles than those left in the field at Site B (X2 = 43.8, P < 0.01, d.f. = 1) and Site C (X2 = 15.86, P < 0.01, d.f. = 1) whereas there was no significant difference between the captive-reared and field survivorship for Site A (X2 = 0.25, P > 0.1, d.f. = 1). As this analysis used the total number of tadpoles which survived within a site, the lack of a significant difference between the captive-reared and field tadpoles for Site A was largely due to one large clutch attaining a high level of survivorship in the field.The mean clutch survivorship 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 compared to 13% survivorship to metamorphosis in the field-reared animals. The field tadpoles displayed a higher level of survivorship than the captive-reared tadpoles during the post-winter stage (Figure 1).The greatest level of mortality for the captivereared tadpoles was during the first two weeks after their release (Figure 2). Furthermore, there was a greater level of survivorship for the early-release tadpoles than the laterelease tadpoles (Figure 2). The post-winter developmental rate of tadpoles showed a similar pattern across the three experimental sites (Figure 3), with the late-release tadpoles appearing to develop at a faster rate than the early-release tadpoles, which in turn developed faster than the field tadpoles. Date of metamorphosis at all
162
three sites was usually about two weeks apart between the field, early and late-release tadpoles (Figure 2).The exception to this was the similar date at metamorphosis for the early and late-release tadpoles at Site C. The water temperatures recorded inside and outside enclosures from four pools are given in Table 1.There were only minor differences between the temperature inside and outside the enclosures with the greatest differences being observed between enclosures. There was a significant difference in tadpole size (measured at Gosner stage 41) between the three experimental sites (Figure 4) with tadpoles at Site C being the largest and Site B being the smallest (Table 2).There were no significant difference observed between the field, early release or late release tadpoles (Table 2).
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 Comparison between field-reared and captive reared tadpoles The monitoring results from the first year of this project have confirmed our ability to increase recruitment significantly through to the terrestrial frog stage via captive rearing.The difference in survivorship between the captive-reared and field animals was largely due to the higher level of field mortality experienced at all three sites during the winter (Figure 1). Whether winter mortality in the field occurred in the egg or tadpole stage is unknown. Given that the low autumn precipitation resulted in very few nest sites being flooded prior to snowfall, and considering that remains of dead eggs were found in a number of nest sites the following spring, much of this mortality may have occurred prior to the eggs hatching. Further experiments investigating different
levels of mortality between over-wintering eggs compared to over-wintering tadpoles are currently being undertaken. The low level of recruitment through to the post-winter tadpole stage observed in this study may have been a feature of other P. corroboree populations as a survey conducted in spring at seven other remnant populations failed to locate tadpoles at five of these (Hunter and Osborne unpubl. data). Research conducted during the mid 1960’s on both P. corroboree and the Northern Corroboree Frog P. pengilleyi (then considered to be P. corroboree) documented high annual variability in early life history mortality as a result of varying climatic conditions between seasons (Pengilley 1992). Given that the winter field mortality observed during this study may have been a product of the poor climatic conditions experienced during the 1997 winter (ie. low autumn rainfall and late snow cover), there is a need to establish our ability to reduce mortality artificially during climatically good years. 163
While over winter mortality was greater in the field than in captivity, the opposite occurred during the post-winter stage where a greater level of mortality was exhibited by the captive-reared tadpoles (Figure 1). A number of factors may have contributed to this result, including possible negative effects of the captive rearing process.The greater mortality exhibited by the late-release tadpoles than the early-release tadpoles (Figure 2) indicates that a greater level of survivorship may be attained by releasing tadpoles earlier. Considerable variation in mortality levels was also observed between sites in both the field and captive-reared treatments (Figure 1). While this variation between sites may have been due to both the low sample size of clutches being monitored and local environmental and genetic effects, it also indicates the potential for high levels of early life-history mortality within these small remnant populations.
TABLE 1: Comparison between temperatures recorded inside and outside of enclosures. All figures are average temperatures with minimum and maximum in parenthesis.
Round Mt. Enclosure Round Mt. Free Snakey A Enclosure Snakey A Free Snakey 9 Enclosure Snakey 9 Free Snakey B Enclosure Snakey B Free
Oct/Nov
Dec
8.4 (4.7/22.5) 8.7 (-0.4/23.9) 13.6 (3.9/30.5) 13.8 (2.9/33.8) 13.1 (0.9/44.1) 13.3 (5.5/33.8) 11.4 (5.0/20.5) 13.5 (-1.7/32.7)
9.2 (5.0/32.9) 9.1 (5.2/34.3) 15.5 (3.9/34.2) 14.7 (-2.7/35.3) 15.1 (5.8/46.2) 15.8 (9.9/34.8) 13.9 (7.5/22.4) 14.5 (2.6/35.2)
The relationship between date and size at metamorphosis in amphibians is considered to be of importance to both juvenile survivorship and adult fitness (Semlitsch et al. 1988). As such, any differences between captive-reared and fieldreared tadpoles for these features may result in differences in the fitness of individuals between the two treatments. While there was no significant difference between the size of captive-reared and field tadpoles within sites there was a significant difference in the size of tadpoles between the three sites (Figure 3).Tadpoles often exhibit a high level of phenotypic plasticity for a number of traits, including size (Duellman and Trueb 1994) and as such it would be expected that size differences may be observed between sites as a results of local environmental effects. 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 developmental rate (Figure 3) and temperature (Table 1) as 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 (Figure 3).This difference was due to the captive-reared tadpoles being slightly less developed than the field tadpoles at the time of release. It is difficult to speculate how this difference in date at metamorphosis may affect the fitness of the captive-reared animals. In any case, date at metamorphosis for the captive-reared tadpoles occurred within the period tadpoles were observed metamorphosing at other non-experimental sites (Hunter pers obs.).
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.
DF
Sum of Squares
Mean Square
Analysis (a) Among cells Treatment Site Treatment*Site Within Total
5 2 1 2 59 64
0.0114 0.0012 0.0079 0.0005 0.0322 0.0437
Analysis (b) Among cells Treatment Site Treatment*Site Within Total
5 1 2 2 63 68
0.0296 0.0013 0.0260 0.0002 0.02552 0.05512
Source
Assumptions and limitations of this project The rationale behind the approach taken in our research has been based on generalisations about amphibian population dynamics. Assuming that the process of captive-rearing does not have a negative impact on the fitness of released individuals, the results will be strongly influenced by the extent to which mortality at the egg and tadpole stage contributes to the current regulation of population size. However, even if egg and tadpole mortality is a key factor in determining population size, an increase in recruitment through to the terrestrial frog stage in the experimental populations is greatly restricted by the low number of breeding adults within these sites and the relatively small clutch size for this species. As such, both the control and experimental populations may remain equally susceptible to stochastic levels of juvenile and sub-adult mortality, despite reduced egg and tadpole mortality in experimental populations. Our reluctance to extend the use of captive rearing to bridge mortality levels during the juvenile and sub-adult stages is a response to a lack of knowledge of the husbandry requirements of the post-metamorphic stages. Moreover, returning individuals to the field as early stage tadpoles reduces the possibility of other complications resulting from the captive environment, such as artificial selection or the need to imprint on the natal pool (Bloxam and Tonge 1995, Dodd and Seigel 1991). While information on age-specific schedules of mortality for P. corroboree is lacking, demographic information obtained since the commencement of this project does lend some 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.
F Value
P>F
0.0023 0.0006 0.0079 0.0003 0.0005
4.17 1.09 14.41 0.60
0.0026 0.3430 0.0003 0.6052
0.0060 0.0013 0.0132 0.0001 0.0004
14.62 3.23 32.68 0.27
0.0001 0.0771 0.0001 0.7637
Furthermore, both this study, and other research conducted prior to the decline of P. corroboree (Pengilley 1992), have indicated the potential for high levels of within-season embryonic mortality, with some populations experiencing total embryonic mortality.These findings suggest that early life-history mortality may significantly contribute to population regulation and, as such, support our attempt to reduce mortality during this stage as a means of increasing the size of adult populations. In comparing survival between the captive-reared and field tadpoles within the same sites we have compromised our capacity to relate future changes in the size of experimental populations to the efforts of captive rearing.This is because we will not be able to differentiate between individuals that were captive-reared and individuals that were left in the field. Ideally, marking cohorts would be undertaken so that upon reaching maturity they could be identified as either captivereared or field individuals, however, this would be very difficult because P. corroboree tadpoles become extremely cryptic upon approaching metamorphosis. A more appropriate approach may have been to compare captive versus field survivorship between sites rather than within sites.This, however, would have required the use of a greater number of sites than is currently available. Also the comparison of tadpole survivorship between treatments would not control for the variation in survivorship between sites (Figure 1). At this stage, this limitation is considered an inherent part of the initial phase of this project as priority has been given to refining the methodology and obtaining information to aid other aspects of the recovery process.
165
Relevance of this project to the recovery process The need to identify and mitigate the processes causing the decline in a species before attempting repatriation projects has been strongly emphasised in the literature (Dodd and Seigel 1991; Griffith et al. 1989). While this approach is sensible, it may not always be an appropriate prerequisite as it does not recognise the potential contribution of repatriation experiments to other research and recovery actions (Armstrong et al. 1994, Soderquist 1994). With respect to the recovery process for P. corroboree, hypotheses relating to the decline of this species have been formulated in the absence of any information on which life history stage(s) has suffered an increase in mortality.This greatly limits our ability to allocate priorities of research into the causal factors of decline. Even so, those hypotheses which have been proposed are either virtually impossible to test using controlled scientific experiments (ie. those examining habitat or climate change) or require relatively high numbers of individuals for experimentation which are currently not available (ie. those examining disease or increased levels of UV-B radiation).
As a result of locating several additional populations of P. corroboree in suitable areas during the 1998 breeding season, an additional four experimental populations and four control populations were incorporated into the project during the second year.This will greatly enhance our ability to relate changes in the size of experimental populations to the process of reducing early life-history mortality. Our ability to obtain meaningful results may also be influenced by the duration of this project.This is due to the possibility that P. corroboree may live for up to six years as a breeding adult (Hunter unpubl. data), and, as such, successive years of population augmentation may have an accumulative effect on the breeding population size. The process of increasing survivorship through to metamorphosis needs to continue for a further three years. In addition, the development of a long term monitoring program for P. corroboree by the New South Wales National Parks and Wildlife Service should provide the commitment and continuity necessary for assessing the outcomes of this project.
ACKNOWLEDGMENTS
Given the constraints currently imposed on the recovery process for P. corroboree, the potential contribution from the population augmentation project will be much broader than just attempting to increase the size of small remnant populations.The development of husbandry techniques and reduction of mortality will allow more scope for conducting experiments and establishing a captive breeding colony without having to exert undue levels of harvesting pressure on extant populations. Data obtained on the current levels of embryonic and tadpole mortality will also assist in determining whether early life history mortality is significantly contributing to the continued decline in P. corroboree.This information is particularly important for guiding research and management actions, because the current factors regulating the population size of P. corroboree are not necessarily those that caused the initial decline in this species.
This project has been funded by the Endangered Species Program of Environment Australia, the New South Wales National Parks and Wildlife Service, the University of Canberra and the Amphibian Research Centre. A special thanks to members of the Victorian Frog Group, particularly Rebecca Hirst, for undertaking further fundraising to assist this project. For assistance with either fieldwork and or the captive husbandry we thank Joanne Doherty, Graeme Enders, Narelle Freestone, Jason Kirby, Fiona Morrow, Dave Reznick, Griff Rose, Craig Smith, Mike Smith and Quinton Smith. Thanks also to David Judge for statistical advice. We are indebted to Dave Lawrence and his staff at the NSW National Parks and Wildlife Service office in Khancoban for their tireless efforts in maintaining water in breeding pools. Helpful comments on this manuscript were provided by Mani Berghout, Keith McDonald and Alastair Campbell. This project is the result of continued support and guidance by the Southern Corroboree Frog Recovery Team.
FUTURE DIRECTIONS
REFERENCES
To account for annual variation in field mortality rates, and hence our ability to decrease embryonic and tadpole mortality artificially below field levels, further comparisons of field versus captive survivorship will be undertaken during the second field season. In addition to this comparison, techniques aimed at reducing early life-history mortality without having to remove eggs from their natal pools will be investigated.The successful development of such a technique will hopefully reduce the need to remove eggs into captivity and allow greater allocation of resources in this area to developing a breeding program for P. corroboree.The production of high numbers of tadpoles, through captive breeding, for release into experimental populations may reduce the possibility that stochastic processes will override our efforts of increasing survivorship. Successful breeding of P. corroboree may also provide greater potential for conducting experiments into the causal factors of decline and attempting to establish P. corroboree at former breeding locations.
Armstrong, D. P., Soderquist,T., and Southgate, R. (1994). Designing experimental reintroductions as experiments. Pp. 27-29 in Reintroduction Biology of Australian and New Zealand Fauna, ed by M. Serrena. Surrey Beatty & Son, Chipping Norton.
166
Barker, J., Grigg, G.C. and Tyler, M.J. (1995) A Field Guide to Australian Frogs. Surrey Beattie & Sons. Bloxam, Q. M. C. and Tonge, S. J. (1995). Amphibians: suitable candidates for breeding-release programs. Biodiversity and Conservation, 4: 636-644. Denton, J. S., Hitchings, S. P., Beebee,T. J. C. and Gent, A. (1997). A recovery program for the Natterjack Toad (Bufo calamita) in Britain. Conservation Biology, 11: 1329-1338. Dickman, C. R. (1996). Incorporating science into recovery planning for threatened species. Pp. 63-73. in Back from the Brink: Refining the threatened species recovery process. Ed. S. Stephens and S. Maxwell. Surrey Beatty & Son, Chipping Norton.
Dodd, C. K. and Seigel, R.A. (1991). Relocation, repatriation, and translocation of amphibians and reptiles: are they conservation strategies that work? Herpetologica, 47: 336-350.
Osborne, W. S. and Norman, J. E. (1991). Conservation genetics of corroboree frogs, Pseudophryne corroboree: population subdivision and genetic divergence. Australian Journal of Zoology, 39: 285-297.
Duellman, W.E. and Trueb, L. (1994) Biology of Amphibians. (2nd Ed.). John Hopkins University Press: Maryland, USA. Gosner, K. L. (1960). A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica, 16: 183-190.
Osborne W. S., Zentelis R. A. and Lau, M., (1996) Geographical variation in corroboree frogs, Pseudophryne corroboree Moore (Anura: Myobatrachidae): A reappraisal supports recognition of P. pengilleyi Wells and Wellington. Australian Journal of Zoology, 44: 569-587.
Griffith, B., Scott, J. M., Carpenter, J. W. and Reed, C. (1989). Translocation as a species conservation tool: status and strategy. Science, 245: 477-480.
Pengilley, R. K. (1992). Natural history of Pseudophryne spp. (Anura: Myobatrachidae) in the Southern Highlands of NSW, Australia. Sydney Basin Naturalist, 1: 9-29.
Hein, E. W. (1997) Improving translocation programs. Conservation Biology, 11: 1270-1274.
Rienert, H.K. (1991).Translocation as a conservation strategy 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 corroboree Moore (Anura: Myobatrachidae). Australian Wildlife Research 16: 537-547.
Semlitsch, R. D., Scott, D. E. and Pechmann, H. K. (1988).Time and size at metamorphosis related to adult fitness in Ambystoma Talpoideum. Ecology, 69: 184-192.
Osborne W.S., (1991) The biology and management of the Corroboree Frog (Pseudophryne corroboree) in NSW. Species Management Report Number 8, NSW National Parks and Wildlife Service, Sydney.
Snyder, N.F.R., Derrickson, S.R., Beissinger, S.R. Wiley, J.W., Smith,T.B.,Toone, W.D. and Miller, B. (1996). Limitations of captive breeding in endangered species recovery. Conservation Biology, 10: 338-348.
Osborne, W. S. (1996). Draft Recovery Plan for the Southern Corroboree Frog. (Pseudophryne corroboree). Unpublished report to the NSW National Parks and Wildlife Service.
Soderquist,T. R. (1994).The importance of hypothesis testing in reintroduction biology: examples from the reintroduction of the carnivorous marsupial Phascogale tapoatafa. Pp. 159-164 in Reintroduction biology of Australian and New Zealand fauna, ed by M. Serrena. Surrey Beatty & Son, Chipping Norton.
Osborne, W. S. (1998). Recovery Plan for the Southern Corroboree Frog (Pseudophryne corroboree) 1998-2001. NSW National Parks and Wildlife Service, Sydney. Osborne, W. S., Hunter, D.A. and Hollis, G.J. (1999) Population 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.
Tyler, M. J. (1997). Action Plan for Australian Frogs. Wildlife Australia, Endangered Species Program: Canberra.
167
Husbandry: science or art? — Are captive technologies ready to contribute to recovery processes for Australian frogs? Gerry Marantelli *
ABSTRACT The history of uses for frog husbandry in Australia is extremely varied. Frogs are held in captivity for a wide range of reasons. Such holding leads to both a need for, and the development of, husbandry technologies.This history, and its combined captive achievements are reviewed.The risks and benefits of captive technologies as tools in conservation are summarised and priorities and protocols for the use of husbandry in conservation are proposed. The potential for captive husbandry to contribute to the conservation of Australia’s frogs is discussed, along with the roles of stakeholders, and suggestions for maximising their cooperative contributions with illustrations drawn from case studies. It is concluded that we should accelerate our pace in this area of research as the needs for husbandry technologies * Amphibian Research Centre, PO Box 424, Brunswick, Victoria 3056. 168
have already outstripped our supply of information. It is suggested that efforts be made to rigorously investigate husbandry of as many species as possible, and priority groups are outlined. Finally it is determined that care must be taken to ensure risks are mitigated and needs are real before embarking on captive conservation programs.
INTRODUCTION The need to utilise all available resources in the effort to stem amphibian declines has seen a significant amount of attention focused on husbandry and the use of captivity as a tool in the conservation process.The usefulness of husbandry, or more specifically captive breeding, as a conservation tool has been the subject of much debate, with some authors suggesting that many complex issues are simply overlooked (see Gippoliti and Carpaneto 1997; Snyder et al. 1996; Snyder et al. 1997). A number of projects employing captive breeding or rearing to reestablish or bolster wild populations of frogs have been attempted with varying levels of success (Banks 1996; Denton et al. 1997; Hunter et al. 1999;Tonge and Bloxam 1989). By their
nature, conservation projects incorporating captive care achieve among the highest profiles of all conservation efforts.They place conservation actions within population centres and provide the public and media with easy access to otherwise ‘invisible’ species. Such access and exposure brings with it the opportunity to educate and influence; as well as to secure funds and resources not readily available to other less popularised or more remotely located conservation efforts (Kleiman and Mallinson 1998; Snyder et al. 1996). If not carefully handled, this exposure may also lead to an unrealistically optimistic estimation of the true value of husbandry in such conservation projects (Snyder et al. 1996). Potential stakeholders in captive conservation represent a multitude of different interests and disciplines.The effective networking and correct utilisation of all such parties remains the most effective way to ensure beneficial outcomes from the use of husbandry in the conservation of Australia’s frogs.
THE ROLE OF HUSBANDRY IN CONSERVATION AND RESEARCH The Action Plan for Australian Frogs (Tyler 1997) estimates a total of A.$5 million dollars should be committed over five years to aid the recovery of Australia’s threatened frogs. 20% – $1 million, has been earmarked for husbandry-related actions.There are also additional connections between husbandry and other research-based actions listed, including toxicology and disease research. The need for husbandry technology in conservation extends well beyond its role in breeding of threatened species for release. In addition to increasing numbers of wild frogs by release, captive supply can also minimise harvesting for other uses. Good husbandry can reduce abnormal behaviours associated with confinement.This can lead to less variables in other research, a better understanding of, and ultimately an increased capacity to conserve, frogs. Captive frogs are usually the most accessible to the public and the media, providing opportunities to educate as well as promote conservation programs and generate financial assistance.The ease of monitoring captive frogs has also often led to observations not readily accessed in the field. Husbandry can be broken down into three functional components: Holding (temporary care of animals), Rearing or single generational maintenance (for conservation purposes this is usually across a mortality barrier, with the intention of achieving greater than field survival e.g. recruitment enhancement projects) and Breeding (production of animals from captive-held stock). Captive breeding, rearing and holding can contribute to conservation and conservation research by the following means: 1. Reducing pressures on wild populations by provision of captive-bred frogs for end uses: I. II. III. IV.
Pets. Medical, biological and conservation research. Education and display. Food, or other animal products.
2. Providing for and contributing to conservation education by: I.
Increasing public exposure to frogs, thereby drawing attention to conservation issues relating to frogs.
3. Contributing to our knowledge of frogs by: I.
Performing specific investigations into the husbandry of frogs.
II. Conducting investigations to assist in ecological research or recovery plans. III. Incidental exposure to observations not readily accessible to field biologists. IV. Producing technologies to reduce effects related to confinement that may otherwise confound research. 4. Generating funding for conservation and research by: I.
Sales of frogs, tadpoles and expertise to end users of frogs. II. Attracting sponsorship and public support for high profile conservation programs. III. Indirectly generating funds for all frog research by increasing community awareness and exposure. 5. Captive breeding or raising of declining frog species to produce stock for: I. Decline hypothesis testing. II. Translocation experiments. III. Re-release to the wild.
RISKS ASSOCIATED WITH HUSBANDRY RELATED CONSERVATION ACTIONS Any manipulation of natural systems is associated with risk. Simply collecting specimens could put pressure on wild populations in some cases. Introduction of additional animals and manipulation of mortality schedules, which are often the aim of captive programs, pose yet more risks (for a review see Snyder et al. 1996). As populations decline, the possible damage done by each of the above processes escalates. Unfortunately, it is also when populations decline that the above interventions are given higher priority. Careful attention must be given to all potential risks when any conservation measure involves manipulation of field situations. To follow the medical profession’s caveat, ‘above all do no harm’, would be a wise philosophy (Myers 1993).
COLLECTION AND GENETIC REPRESENTATION Captive conservation requires founder stock. Procurement of stock should not exceed a population’s capacity to be harvested (Stevens and Goodson 1993).To reduce genetic impact, it is desirable to proceed with collection of individuals as soon as it is recognised that collection is required and well before the population is at risk. Animals collected should be considered as no longer available to the wild population; too many instances of failed reintroductions exist ( Wolf et al. 1996) to place reliance on individuals being used to develop husbandry protocols ever contributing to wild populations.Wherever possible, collection should concentrate on animals of low ecological value to the population and higher value in captivity. In most cases this would constitute the earliest life history stages available, as larval and juvenile mortality are usually high in nature and significantly reduced in captivity. Collection of early life history stages also necessitates forward planning, as adult captive animals will take longer to achieve.Where protocols for long-term holding or establishment of breeding colonies are being developed, a staggered approach is recommended. A small number of adult frogs can be used to develop techniques for achieving reproductive success, while larger numbers of juveniles are raised to form a colony. By the time a colony is achieved, techniques for raising as well as breeding should have been developed and any difficulties overcome. 169
Adequate representation of a population or species should be considered in cases where a colony is formed to secure a species against extinction and in cases where captive stock will be used to reestablish or bolster wild populations. What constitutes adequate representation must be determined for each situation and in extreme cases such representation may not even be feasible.The loss of genetic representation in captive populations (Briscoe et al. 1992), along with extinctions or gene loss attributed to low founder numbers or inbreeding (Frankham 1995; Robinchaux et al. 1997), are well documented.
RELEASE A number of issues and some suggested solutions have been raised with respect to the release of animals which have spent time in captivity. Genetic considerations rate highly (Backus et al. 1995; Ryman et al. 1995), while the potential to release disease must be carefully assessed before release (Viggers et al. 1993).The relative fitness of released animals should also be assessed to ensure efforts are not being wasted (Crayford and Percival 1992; Hunter et al. 1999).This is best done by monitoring animals post-release. The above precautions have not always been taken and sometimes the consequence of such omissions have led to very expensive failures and an inability to identify the reasons, or establish techniques to improve later attempts. Recently a number of releases of captive-bred Green and Golden Bell Frogs, Litoria aurea failed completely. Although numerous explanations including: human disturbance, under-developed habitat, introduction of predatory fish, and predation by waterbirds and feral carnivores are offered (Hobcroft 1998); the absence of adequate post-release monitoring has left these explanations without adequate support or quantification (Meikle pers. comm.). Given the financial and ecological costs of captive projects, the greatest mistake we can make is not failing to be successful, but failing to design a project that gives us the greatest possibility of identifying the reasons for our failure. The recent discovery of the amphibian chytridiomycete fungus that may be the proximal cause of numerous frog declines (Berger et al. 1998), highlights the need to consider the risk of disease being introduced with any introduction of frogs. Not only known disease but all diseases should be controlled. Screening, where available, may be able to eliminate some possible pathogens from release. Adequate quarantine, however, not only provides preventative control but protects against unknown or undetectable pathogens. Fitness has always proven difficult to quantify. Failing to include such assessments in any release, however, could lead to erroneous predictions about, or measures of, the success of release programs. It has been demonstrated that crowding can cause developmental delays, smaller adult size and lower fecundity and fertility. Rarely are captive animals raised at field densities and so the question begs, are we really producing stock suitable for release? Smooth Froglets, Geocrinia laevis, raised in crowded captive conditions took much longer to mature, reached sexual maturity at a significantly smaller size and produced smaller clutches (Marantelli unpubl. data) than the greatest extremes measured for this species in nature (Scroggie, pers. comm.).These and other as yet unknown 170
effects of captivity could create lasting implications for the populations into which captive raised animals are released. All release programs should be backed by sound monitoring and, wherever possible, by comparative analysis of fitness (measured by success in development and reproductive output) of released versus wild individuals.
REMOVAL OF CAUSAL AGENTS The IUCN guidelines recommend that release should not 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 be identified.Two considerations may give cause to reconsider these guidelines with respect to frogs: most frogs are considerably cheaper, both financially and ecologically, to produce than are the majority of other terrestrial vertebrates, and; limited, planned experimental reintroduction aimed at hypothesis testing may in many cases be a cost effective and viable technique for identifying causal agents. Once the necessary precautions have been taken and plans developed that maximise the chance of success, while providing opportunities to identifying any failings, programs aimed at using husbandry to produce frogs for wild release can be considered. At this point it is valuable to look at the history of frog husbandry in Australia and what our current knowledge has to offer.
AN HISTORICAL PERSPECTIVE As with most areas of natural history discovery, the early period of frog biology was dominated by taxonomists. During this time taxonomists had cause to temporarily house live frogs in research collections, providing some of our earliest insights into frog husbandry.The need to describe life histories saw considerably more attention focused on techniques for achieving frog spawn and raising tadpoles. Soon other uses of captive frogs; for research, as pets and as display animals began to add to our knowledge. Each had their limitations: taxonomists rarely raised frogs past metamorphosis and the supply of frogs for research, pets and display was typically replenished from wild sources. Although exhibit-based collections sometimes kept individual specimens for very long times, there was seldom the need to breed replacement stock. Amateur collectors often saw breeding as a yardstick by which to measure their achievements and were among the first to breed Australian frog species (pers. obs.). Unfortunately, many such firsts were achieved outside Australia, where the difficulty in obtaining more specimens must have contributed to the efforts expended on achieving breeding success. More popular literature on husbandry of common Australian frogs has been published in North America than in Australia (e.g. de Vosjoli et al. 1996). In Australia it was rare for there to be any need to culture frogs for long periods of time, to raise young frogs to sexual maturity, or to house the larger numbers of individuals necessary to give robust results. Frog husbandry was, and in many cases still is, an art rather than a science -a means to an end, be it for taxonomy, research, pet keeping or display. Declines in numerous species of Australian frogs coupled with the apparent success of captive conservation programs for other vertebrate animals (for a review see Wolf et al. 1996) has seen a focus placed on captive programs for a number of
Australia’s frogs. Unfortunately the development of husbandry for amphibians has not been subject to the same level of interest or need as it has for other vertebrate groups. Large commercial interests in mammals, birds, fish and to a lesser extent reptiles have driven the development of husbandry technologies; frogs have been largely overlooked. Dietary information, medications, reproductive technologies and other husbandry techniques developed for commercially important vertebrates and humans, are routinely applied with success to captive care of non-commercial vertebrate species. Significantly less commercial value has been placed on frogs. The resulting comparative reduction in our warehouse of knowledge no doubt contributed to the poor success rate of some early attempts at captive care of declining amphibians. While Australian zoos had on the whole been successful with captive conservation of other vertebrate groups, they experienced considerably more difficulty with frogs. With few exceptions frogs had only quite recently been kept in Australia’s zoos and were little more than a minor inclusion in predominantly reptile driven herpetofauna displays.This lack of experience coupled with the absence of information from commercial research placed us in a very precarious position when attempts were first made to employ captivity in efforts to salvage declining frog populations. Over 100 of the now presumed extinct Taudactylus acutirostris died in captive collections during attempts to secure that species in captivity (pers. obs.); while over 50 tadpoles and metamorphs of the critically endangered Litoria spenceri were taken into captivity at Melbourne Zoo, only one remained alive when the project to develop husbandry protocols for that species was moved elsewhere (Gillespie pers. comm.). Despite the fact that these projects did include some successes and did increase our knowledge, and that other projects have been successful, they are stark evidence that we were not prepared to deal with last ditch efforts when it came to amphibians. Australia’s track record numbers at least four species Rheobatrachus silus, R. vitellinus,Taudactylus diurnus and T. acutirostris that, while not all held specifically for conservation, possibly the last individuals of their species died in the hands of their human keepers. It has only been recently that any attention has been focused on more disciplined attempts to establish protocols for Australian frog husbandry. Links between exhibit based collections and universities have recently become more common, promoting increased scientific rigour in this area of husbandry research.The establishment of the Amphibian Research Centre (ARC) in 1994 (the first facility in Australia driven purely by frog husbandry) has enabled the commercial uses of frogs to fund the development of husbandry protocols for a number of common native frog species.
The commercial need to constantly improve techniques, coupled with the high volumes of individuals produced, has led to significant advances not otherwise able to be achieved by holding small numbers of specimens.
CAPTIVE ACHIEVEMENTS FOR AUSTRALIAN FROGS During 1996 and 1997 a survey of 14 institutions and 12 private collectors known to have kept significant numbers of frogs was conducted by The Amphibian Research Centre. The survey asked respondents to indicate success or lack thereof for a number of possible achievements in husbandry, for each species of Australian frog which they had held in captivity. Items included: the introduction of various life history stages into captivity, maintenance of various life history stages for defined periods of time or across defined developmental events, achievement of captive spawning under various sets of circumstances and the number of generations achieved in captivity. Most respondents were visited and interviewed. In some cases survey sheets were distributed. In other cases data derived from communications, publications or personal observations were used. Although not exhaustive, this survey gives a good impression of the relative success Australia’s herpetologists and zookeepers have had in keeping Australian frogs. At least 150 species of Australia’s 205+ species of frogs have been held in captivity although at least 25 of these were never held for more than two months.The number of species and genera for which significant benchmarks have been achieved is included in Table 1.The number of species which are listed in the Australian Frog Action Plan that have reached each benchmark in captivity are also included (Table 1.) Much of the information collected by our survey was anecdotal. Few respondents were able to quote figures for exact periods of time that specimens were kept, number of specimens kept or raised, or even in some cases the number of times breeding had been recorded. It should also be noted that much data was contributed by biologists with no specific interest in long-term husbandry, or who only kept species long enough to achieve spawn for life history descriptions. That species bred were seldom raised beyond F1 is more due to lack of need than to failure.Those studying life histories simply did not need to attempt to proceed any further. Despite these limitations, the results indicate that very few species of Australian frogs have been studied extensively in captivity.The typical spawning event (excluding those species spawned prior to one month in captivity) was
TABLE 1: Captive achievements for Australian frogs.
Achievement
Adults kept > 2 months Adults kept > 12 months Spawned in captivity F1 progeny raised to at least metamorphosis F2 or beyond achieved *
Number of species
Number of genera
Number of listed species. (Frog Action Plan)
125 90 41* 32 14
20 16 15 10 5
13 5 3 4 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. This must raise the question of what we are doing wrong which can cause a group of adult frogs to largely fail in their primary biological function — reproduction!
THE PLAYERS In all fields of endeavour there are usually a number of individuals and organisations who contribute to any successful operation.The nature and popularity of conservation ensures that a multitude of players are involved in most actions and the level of success of such actions usually depends on the precise use and coordination of these contributors (Cannon 1996; Kleiman and Mallinson 1998). Husbandry and its highly popularised subordinate — captive breeding, attracts perhaps the widest diversity of participants and therefore involves considerable challenges in terms of effective coordination of contributions.The primary focus is considered here to be husbandry, irrespective of its objective being commercial, research or conservation.The players and their connections are illustrated and discussed below and in Figure 1.
PUBLIC The public can be both consumers of frogs and providers of financial and political support for recovery actions.There is demand for pet frogs and tadpoles, so much so that the child with a tadpole in a bucket is almost an Australian icon.The Australian public grew up with frogs, their sounds and the experience of raising a tadpole.Their desire to ensure this right for their children can be counted amongst the most powerful of all the recovery tools at our disposal. Husbandry and captive breeding can ensure the public continued access to and contact with frogs and tadpoles.The Australian people are not only potential direct financial supporters of recovery actions, but the voice of frogs to others, including corporate sponsors and government.
COMMUNITY ORGANISATIONS
Recovery teams administer conservation decisions and actions for specific species.They are valuable sources for communication of information pertinent to each species’ conservation.
Community organisations galvanise the sentiments of the public and allow those involved in recovery processes direct access to a sympathetic audience. Frog groups around the country not only have the networks necessary to raise funding and political pressure, but act to ensure that those who take an interest remain interested.The provision of information, in the form of talks and newsletter articles from conservation scientists, is an integral part of ensuring that these groups remain effective in their popularising of frog conservation issues. Such groups may also provide assistance with some actions and are in a unique position to provide remuneration to sponsors in the form of direct contact with an interested consumer base. It should also be noted that in some cases members of such groups have contributed and continue to contribute significant amounts of information to our knowledge of husbandry.
UNIVERSITIES/RESEARCH FACILITIES
MEDIA
If frog husbandry in Australia is to secure credibility as a science, it will be necessary for links to be firmly established between scientists and those who are investigating husbandry. In some cases these links already exist and have seen rigorous experimental investigations into husbandry (e.g. Hunter et al. 1999). Universities and research collections are also end users of frogs for non-conservation research and could be provided with stock, use of facilities and husbandry expertise from those investigating husbandry. As a newly developing area of research, husbandry offers a multitude of possibilities for student projects and since facilities to conduct such research are expensive, their efficient use depends on maximising their output.
Frogs held in captivity provide the media with easy access to interesting stories.The media responds to the public, who in turn rejuvenate their interest with each new story.The media is in the unique position of being the main method available to conservation biologists for rewarding the public and sponsors for their efforts, while generating more interest from both.
ENVIRONMENT AUSTRALIA Environment Australia (EA) administers Australian frog recovery plans, the National Threatened Frog Working Group and provides funding for conservation actions.
RECOVERY TEAMS
CORPORATE SPONSORS Captive programs place threatened species in population centres and provide access to the public and the media.This access, coupled with the public’s perception of the importance of such programs and their desire to see action being taken, places such programs in a unique position with respect to generating funds not otherwise available to less visible or less popularised forms of conservation. Sponsorship of such projects can provide resources beyond the scope of normal recovery plan budgets, while providing valuable marketing opportunities for the sponsor.The effective sale of recovery actions to corporate sponsors almost invariably requires visibility of the species concerned and, as a consequence, husbandry.
172
EXHIBIT BASED COLLECTIONS House, display and in some instances breed amphibians.They are in some cases active participants in husbandry research. Husbandry research ideally should be able to provide exhibitbased collections with expertise and “ecologically dead” stock for display.The placement of small numbers of specimens in collections specialising in public education offers us a further avenue for recognising contributors and promoting our conservation objectives to the public.
DEVELOPMENT OF HUSBANDRY PROTOCOLS The need to ensure a scientific approach to the development of husbandry techniques will require the production of a set of guidelines. Only a rigid experimental approach is likely to yield the results that are needed in the available time. Lack of communication and co-ordination will continue to lead to
FIGURE 1: Relationships between stakeholders in the use of frog husbandry for conservation.
Recovery plans/teams
Environment Australia
Universities/res earch facilities
Provided with access to husbandry as a recovery and experimental tool. Provide funds and direction for work.
Provided with high profile and effective recovery actions. Provide assistance in maintaining links and funding through recovery plans.
Provided with use of stock, facilities and expertise. Provide projects with student labor and collaborative grant seeking.
Exhibit based collections
Community organisations Provided with regular information on projects of interest to members. Provide increased awareness and acknowledgment of sponsors through publications & newsletters etc.
Husbandry Experimental, recovery and/or commercial frog husbandry
Provided with expertise and ecologically dead stock for display. Provide public access to frogs, information about frog conservation projects and acknowledgment to sponsors.
Public Provided with access to pet/schoolroom stock as well as education of issues. Provide increased profile and incentive for sponsors as well as direct financial input into projects.
Corporate sponsors Provided with high profile wide ranging recognition of contributions. Provide financial support for projects.
Media Provided with easy access to good stories. Provide public access and recognition of sponsors.
173
repetition and wasted resources. In husbandry, two processes should be seen as vital to the development of protocols for species reintroduction: provision of appropriate ecological parameters and quarantine. Administratively, the processes of communication and cooperation are critical.
PROVISION OF ECOLOGICAL PARAMETERS Failure to provide an environment comparable to that naturally experienced by each species can lead to a multitude of problems. While many species can be maintained, bred and raised in quite sterile environments, the absence of natural conditions can lead to selection for captivity (Snyder et al. 1996) and production of animals which are ecologically (Lyles and May 1987; Page et al. 1989) or immunologically (Viggers et al. 1993) naive. Efforts to develop protocols should begin with attempts to provide for all environmental parameters before minimisation of care requirements is attempted. Some considerations that should be included in the establishment of captive facilities for each species are: stocking densities; spatial and temporal contact between individuals; environmental variants including microenvironmental variants and their daily, annual or other cycles (e.g. temperature, photoperiod, humidity, precipitation, water levels and flow rates, barometric pressure and food availability); and availability of specific microenvironments used by each species (e.g. sites used for basking, calling, oviposition, sheltering, hunting, feeding and thermoregulation). Common omissions in captive efforts with frogs include failure to provide non-breeding environments (most captive frogs are housed permanently in environments which replicate the habitat in which they reproduce), housing females with males (exposure of females to reproductively-primed males does not usually occur in nature until the female is ripe), holding too few males to elicit chorusing, and housing animals at densities which modify normal reproductive behaviour.
QUARANTINE Although disease has been flagged as a general risk in wildlife conservation (May 1986, Simonetti 1995, Cunningham 1996), the recent publication of evidence supporting the possibility of disease as a causal agent in Australian frog declines (Berger et al. 1998 and Berger et al. 1999) places special emphasis on disease control in captive conservation efforts for frogs. Only one method exists to adequately control the effect of both known and unknown disease — quarantine. Disease and parasites play an important role in natural systems.The inclusion or exclusion of certain disease organisms in captive programs can only be achieved by combining quarantine, pathology and exposure to natural materials. Effective quarantine involves isolation of captive animals from exposure to materials, other individuals or species which have either not come from the same location, or have not themselves been subjected to the same level of isolation (Snyder et al. 1996). Practically this involves the minimisation of contact and includes use of sterilised water, separate implements and separate housing at enclosure, room or even facility level. Animals should not be exposed to any biological material not sourced from their point of origin.
174
S.P.F. quarantine procedures are well documented in husbandry literature, and should be practiced. Functional quarantine for frog facilities has been achieved by wearing and disposing of latex gloves, avoiding contact of other clothing or implements with enclosure contents, use of separate implements and water supply, splash guards and disinfection of enclosures and implements before reuse with 1% sodium hypochlorite (followed by rinsing). Snyder et al. (1996) propose single species facilities separate from multispecies collections. Individuals used in captive conservation measures which may ultimately include release, should be sourced directly from the wild. Materials sourced from their natal environment should be used to provide for their ecological requirements.These precautions should ensure that exposure to pathogens exotic to the species range will not occur, at the same time as facilitating the possibly necessary exposure to natural pathogens. As an additional precaution, samples should be taken for full pathological investigation prior to release of any stock (Viggers et al. 1993). Pathology should not be seen as a suitable precaution in its own right and extreme caution should be exercised if releases of animals, which have not been adequately quarantined, are being considered. No release of animals held without adequate quarantine should occur into extant populations or sites that may facilitate exposure of extant populations.
COMMUNICATION AND COOPERATION It is recommended that a network be established representing the stakeholder groups (Figure 1) to co-ordinate husbandry research and its application to conservation of Australia’s frogs.The establishment of such a network remains the most likely method of ensuring the required measure of success to secure husbandry as a contributor to conservation, and to prevent failures due to lack of coordination or design resulting in the demise of this potentially useful conservation tool.
PRIORITIES FOR HUSBANDRY RESEARCH IN AUSTRALIAN FROGS Our survey indicated that husbandry knowledge for a number of ecological and phylogenetic groups of Australian frogs is still lacking. Despite some 70% of Australia’s frogs being kept at one time in captivity less than 20% have been bred.Those successfully bred represent primarily lowland species with lentic tadpoles and few if any unusual life history characteristics. Species for which significantly less husbandry success have been recorded belong to groups which include stream-dwelling species, high altitude species, arid-adapted species and species exhibiting specialised life history characteristics (e.g. direct development or parental care). Most of Australia’s threatened frog species belong to one or more of these ecological groups. Limited experience or success has been reported for declining taxa or close allies. As the costs of using captive technologies can be high and their value has not yet been adequately assessed, it is recommended that efforts to develop protocols for threatened taxa be viewed as a priority, but efforts to secure sustainable populations of threatened species in captive collections should not be seen as a routine step in species recovery. Efforts to develop husbandry protocols should focus
on ecological and phylogenetic analogues of threatened species and be tested and modified to suit threatened species only once demonstrated to be successful on such analogue species. Captive breeding for reintroduction should not necessarily be seen as the ultimate goal of husbandry research. As knowledge improves it should be applied to the conservation measures already discussed, very few of which actually include captive production of frogs for release into natural habitat. Techniques such as recruitment enhancement should be seen as useful tools which focus on husbandry skills already developed for Australian amphibians and which also mitigate many of the risks associated with longer term captive programs.The role of husbandry in education and generation of both public interest and funds remains a significant reason for holding threatened species in captivity.
A CASE IN POINT, Pseudophryne corroboree AT THE AMPHIBIAN RESEARCH CENTRE A recruitment enhancement program for the Southern Corroboree Frog Pseudophryne corroboree has been operating since early 1997 (Hunter et al. 1999).The captive component of this project has been carried out at the ARC. The project has been housed in a dedicated facility on a site separated from other ARC frog colonies. Husbandry protocols, provision of ecological requirements, isolation and quarantine procedures, as well as routine pathological screening of specimens prior to each release, have been followed as prescribed in this paper. Post-introduction monitoring to assess relative fitness of captive-reared animals is described and preliminary results reported in Hunter et al. (1999). Methodology for captive care was based on three years of successful commercial breeding of P. semimarmorata. This project has been supported by a wide range of participants and is presented here as an appropriate model for the involvement of relevant stakeholders in captive conservation.The project was supported by funds provided to the New South Wales National Parks and Wildlife Service (NPWS) by Environment Australia (EA). Over two years, $21 000 has been contributed to the captive component of the P. corroboree project by the recovery team. Cooperation between the ARC and the University of Canberra has ensured the development of rigid experimental design following scientific principles. Further cooperation between the ARC and Victoria University of Technology has provided student labour to conduct two small research projects within the larger husbandry program.The involvement of the Victorian Frog Group (VFG), by regular inclusion of information about the project in their newsletter, led to the involvement of Australian Geographic (AG) as a corporate sponsor leading to over $5 000 being contributed to the project. Media coverage and support generated by this publicity has led to a further $2 000+ in public donations to the project.The volunteer resource drawn to the ARC by the presence of this and other projects has contributed greatly to the ARC’s development and continued success. As a consequence of this and its commercial operations, the ARC has contributed some $15 000+ in material resources and $30 000+ in labour and infrastructure resources to the project. In just under two years a project valued at over $70 000 has been carried out with less than a 30% contribution from government. Over $50 000 has been contributed largely by
means not readily accessible to other conservation actions. Such non-transferable funding has been the backbone of this project and allowed for other financial resources to be directed elsewhere within the P. corroboree recovery plan. Similar programs are in progress and should be achievable by most projects involving threatened frogs in captivity
CONCLUSIONS With rising interests in husbandry and a number of conservation actions involving husbandry already underway, there is clearly a need to utilise husbandry as a tool in conservation processes. The need to move quickly to cater to the numerous actions listed in recovery plans which require husbandry is real, as is the need to recognise the limitations of husbandry as a developing science.While legitimate questions must still be raised about genetics, fitness and disease risks associated with the release of animals exposed to captivity, the need for solutions in what seem to be otherwise unwinnable situations must be respected. Where possible it would be prudent to direct the focus to those aspects of husbandry that have demonstrated repeated success, to ensure we do not overextend ourselves — the consequences of which are dire. Programs aimed at tadpole rearing and manipulating mortality schedules across single generations could be considered as other techniques are being perfected. Efforts to develop techniques for longer term management should be commenced immediately with analogue species, focusing on those phylogenetic and ecological groupings for which we have the least knowledge. Maximum use must be made of all the potential contributors to husbandry research. Skilful coordination of multidisciplinary groups within recovery teams remains our best hope of effecting viable use of husbandry as a tool. We must recognise the public perception of captive programs and make use of the support they can give, while taking care not to promote the idea of captivity as a panacea.The effective use of media and the public must be explored to provide alternative sources to traditional funding. Above all there is a need for pragmatism. While programs relying on husbandry may not always be the cheapest or best conservation actions, we cannot deny the support afforded such visible conservation measures and the level of public awareness and sense of involvement they can bring. Such programs should be able to generate much of their own funding, reducing competition for other conservation dollars. Sustainable commercial use of frogs must be encouraged as a free mechanism for driving research.The situation that exists; end users with capital while husbandry research is severely under-resourced, begs the consummation of a very convenient marriage. Realistic attitudes to commercialisation of wildlife conservation may provide the only solution to a rapidly drying government funding well.
ACKNOWLEDGMENTS Thanks to all those who have by their collective efforts contributed to the knowledge base for Australian frog husbandry as described in this paper, especially: Margaret Davies, staff at Taronga Zoo and Currumbin Sanctuary, Lothar Voigt, Harvey Vaux, Karen Thumm, Jacqie Recsie, Will Osborne, Mark Cowen, Ross Alford and Shane Gow.
175
Staff at the ARC who assisted in conducting the frog husbandry survey, or who assisted with maintenance of the collections while I was absent visiting interstate collections or preparing this paper: Natalie White, Mason Hill, Joanne Doherty, Fiona Morrow, Michelle Love, Anne Gaskett and Lindley Makay. To all those who donated their time to assist with the P. corroboree project: Fiona Morrow, Kristy Penrose, Lee Berger, Mason Hill, Joanne Doherty, David Reznick, Raelene Hobbs, Rebekah Hirst, Australian Geographic and the members of the VFG. And to my wife Sally, whose tolerance of my work habits has been pivotal in all of the above.
REFERENCES Backus,V. L., Bryant, E. H., Hughes, C. R. and Meffert, L. M., (1995) Effect of migration or inbreeding followed by selection on low-founder-number populations: implications for captive breeding programs. Conservation Biology, 9: 1216-1224.
(Pseudophryne corroboree). Pp 158-167 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Hobcroft, D., (1998) Conservation program for the Green and Golden Bell Frog Litoria aurea at Taronga Zoo: a cooperative recovery plan. Proceedings of 1997 ARAZPA/ASZK Conference. Auckland Zoo, Auckland NZ. IUCN (1998) Guidelines for re-introductions. IUCN/SSC Re-introduction Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. Kleiman, D. G. and Mallinson, J. C., (1998) Recovery and management committees for Lion Tamarins: partnerships in conservation planning and implementation. Conservation Biology, 12: 27-38. Lyles, A. M. and May, R. M., (1987) Problems in leaving the ark. NATURE, 326: 245-246. May. R., (1986) The precautionary tale of the Black-footed Ferret. Nature, 320: 13-14. Myers, N., (1993) Biodiversity and the precautionary principle. Ambio, 22: 74-79.
Banks, C.B., (1996) A Conservation Program for the Threatened Romer’s Tree Frog (Philautus romeri). Advances in Herpetoculture 1: 5.
Page,G., Quinn, P., and Warriner, J., (1989) Comparison of the Breeding of Hand- and Wild-Reared Snowy Plovers. Conservation Biology, 3: 198-201
Berger, L., Speare, R., Daszak, P., Green, D. E., Cunningham, A. A., Goggin, C.L., Slocombe, R., Ragan, M. A., Hyatt, A. D., McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. and Parkes, H., (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rainforests of Australia and Central America. Proc. Natl. Acad. Sci., 95: 9031-9036.
Robinchaux, R. H., Frier, E. A. and Mount, D. W., (1997) Molecular genetic consequences of a population bottleneck associated with reintroduction of the Mauna Kea Silversword (Argyroxiphium sandwicense ssp. sandwicense [Asteraceae]). Conservation Biology, 11: 1140-1146.
Berger, L., Speare, R. and Hyatt, A.D. (1999) Chytrid fungi and amphibian declines: overview, implications and future directions. Pp 23-33 in Declines and Disappearances of Australian Frogs ed by A.Campbell. Environment Australia: Canberra. Briscoe, D. A., Malpica, J. M., Robertson, A., Smith, G. J., Frankham, R., Banks, R. G. and Barker, J. S. F., (1992) Rapid loss of genetic variation in large captive populations of Drosophila flies: implications for the genetic management of captive populations. Conservation Biology, 6: 416-425. Cannon, J.,R. (1996) Whooping Crane recovery: a case study in public and private cooperation in the conservation of endangered species. Conservation Biology 10: 813-821. Crayford, J. and Percival, S., (1992) Born captive, die free. New Scientist, 8: 21-25. Cunningham, A. A., (1996) Disease risks of wildlife translocations. Conservation Biology, 10: 349-353. Denton, J., Hitchings, S., Beebee,T. and Gent, A. (1997) A recovery program for the Natterjack Toad (Bufo calamita) in Britain. Consevation Biology, 11: 1329-1338. de Vosjoli, P., Mailloux, R., and Ready, D., (1996) Care and Breeding of Popular Tree Frogs. Advanced Vivarium Systems, inc., Santee, California. Frankham, R., (1995) Inbreeding and extinction: a threshold effect. Conservation Biology, 9: 792-799. Gippoliti, S. and Carpaneto, G. M., (1997) Captive breeding, zoos and good sense. Conservation Biology, 11: 806-807. Hunter, D., Osborne, W., Marantelli, G. and Green, K., (1999) Implementation of a population augmentation project for remnant populations of the Southern Corroboree Frog 176
Ryman, N. and Laikre, L., (1991) Effects of supportive breeding on the genetically effective population size. Conservation Biology, 5: 325-329. Ryman, N., Jorde, P. E. and Laikre, L., (1995) Supportive breeding and variance effective population size. Conservation Biology, 9: 1619-1628. Simonetti, J. A., (1995) Wildlife conservation outside parks is a disease-mediated task. Conservation Biology, 9: 454-456. Snyder, N. F. R., Derrickson, S. R., Beissinger, S. R., Wiley, J. W., Smith,T. B.,Toone, W. D. and Miller, B., (1996) Limitations of captive breeding in endangered species recovery. Conservation Biology, 10: 338-348. Snyder, N. F. R., Derrickson, S. R., Beissinger, S. R., Wiley, J. W., Smith,T. B.,Toone, W. D. and Miller, B., (1997) Limitations of captive breeding: reply to Gippoliti and Carpaneto. Conservation Biology, 11: 808-810. Stevens, D. R. and Goodsen, N. J., (1993) Assessing effects of removals for transplanting on a high-elevation Bighorn Sheep population. Conservation Biology, 7 :908-915. Tonge, S. J. and Bloxam, Q. (1989) Breeding the Mallorcan Midwife Toad Alytes muletensis in captivity. Int. Zoo Yb 28:45-53. Tyler, M., J. (1997) The Action Plan for Australian Frogs. Wildlife Australia, Canberra. Viggers, K. L., Lindenmayer, D. B. and Spratt, D. M., (1993) The importance of disease in reintroduction programs. Wildl. Res., 20: 687-698. Wolf, C. M., Griffith, B., Reed, C. and Temple, S. A., (1996) Avian and mammalian translocations: update and reanalysis of 1987 survey data. Conservation Biology, 10: 1142-1154.
Conservation status of frogs in Western Australia Dale Roberts1, Simon Conroy1 and Kim Williams2
ABSTRACT Seventy eight frog species are known from Western Australia. Only three have been the subject of recent systematic survey to assess status. For eleven others there are some objective data (post 1988) suggesting they are still widespread and abundant.Three frog species are of conservation concern in south-western Australia: Geocrinia alba, G. vitellina and Spicospina flammocaerulea. All three have small extents of occurrence (< 200 km2), small areas of occupancy and fragmented distributions. Twenty three sub- populations of G. alba have gone extinct since 1983 and there are continuing threats from grazing, vegetation clearing and fire. Geocrinia vitellina has a range of 6 km2 but there are no major threats except fire. Spicospina flammocaerulea is known from 13 populations over a range of
STATUS OF FROGS IN WESTERN AUSTRALIA Seventy eight frog species are known from Western Australia from fifteen genera (Tyler et al. 1994; Roberts et al. 1997). The Kimberley region (regions sensu Tyler et al. 1994) with summer rainfall has 28 species, the arid zone with irregular but tending to summer rainfall maxima has 17 species and the southwest with predominantly winter rainfall has 30 species. Several species overlap two zones. Although there has been sporadic work on some species in the Kimberley and the arid zone (e.g. Roberts 1996; Watson and Gerhardt 1997) neither area has been the subject of any systematic survey to determine the status of anurans. Further, before the period of major frog declines starting in Australia in the mid seventies (Mahony 1996), many species in these areas were either undescribed, uncollected or their distributions were poorly known so shifts in distribution, local extinctions or declines are unlikely to have been detected. For example,Tyler et al. (1994) reported eight species of Uperoleia from the Kimberley and arid zone of Western Australia — six were described after 1980 and there has been little published work since then.
194 km2, but has a highly fragmented range and population sizes are uncertain. Its true status is unclear.
In the south-west, winter rainfall zone, there has been more intensive work which collectively suggests most widespread 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 frogs recorded and the number of sites where species were heard during recent studies on geographic variation in male advertisement call. One of us (JDR) worked on eight species from 1988-1992 and Littlejohn and Wright (1997) reported on a ninth.These data are summarised in Table 2.These data are further supported by population or population genetics studies on Heleioporus psammophilus (Berry 1996), H. albopunctatus, Neobatrachus kunapalari, N. pelobatoides and Pseudophryne occidentalis (Davis 1997) and Crinia georgiana (McDonald 1998) which all reported abundant populations of the species studied. Although there is some evidence that land clearing, increasing salinity and fragmentation of vegetation may have affected population size and local occurrence of some widespread species (e.g. Main 1990; Davis 1997), there is no evidence of general, recent declines or extinctions in anuran populations in Western Australia comparable with those in other parts of Australia (Mahony 1996;Tyler 1997). However, it is also important to note that, with three exceptions listed below, there has been no systematic survey of the status of any frog species in Western Australia since 1992. Three species, Geocrinia alba and G. vitellina (Wardell-Johnson and Roberts 1989; Roberts et al. 1990) and Spicospina flammocaerulea (Roberts et al. 1997) are currently declared as “threatened” pursuant to section 14 (2) (ba) of the West Australian Wildlife Conservation Act 1950. Species listed as “threatened” generally fall into the IUCN (1994) categories of extinct, critically endangered, endangered or vulnerable. In the most recent rankings of species according to IUCN (1994) Red List criteria (released in May 1998), the West Australian Threatened Species Scientific Committee (TSSC) listed G. vitellina as vulnerable (criterion D2), G. alba as endangered (criterion C2a) and S. flammocaerulea as vulnerable (criterion D2)1. Concern about the status of these species came from initial observations of small range, low population number or low population density, and subsequently from detailed studies of range, fragmentation and population size.
THREATENED FROG SPECIES — WESTERN AUSTRALIA Geocrinia alba and G. vitellina occur north and west of the Blackwood River between Margaret River and Augusta (Figure 1).They are members of a complex of four closely related species (G. alba, G. lutea, G. rosea and G. vitellina) that all have direct developing eggs, breed in spring and into early summer and have population structures consistent with very low levels of dispersal among even adjacent local populations (Driscoll 1997, 1998a, b). Geocrinia alba has an extent of occurrence (range) of approximately 130 km2 (Figure 1, minimum convex polygon method, IUCN 1994). Wardell-Johnson and Roberts (1993) estimated that 70% of creek systems suitable for breeding (where suitability was defined by geomorphology shared with
1 data reported in this paper were not available to the TSSC in late 1997 when decisions on status were last considered. Status assessments discussed below represent new evaluations based on data reported here. 178
known population sites) had been cleared since European settlement with likely consequent loss of populations. Geocrinia alba had a small natural range which has been radically reduced and is now severely fragmented (Figure 1). Fifty six discrete populations of G. alba have been located in the period 1983 to 1997.These populations can be further subdivided into 80 sub-populations based on variation in adjacent land use (e.g. cleared and grazed, natural forest or tree plantations), or division by physical features such as major roads crossing creek systems (Wardell-Johnson et al. 1995 define populations and sub-populations). Most populations occur on privately owned land (Wardell-Johnson et al. 1995; Figure 1). Based on IUCN (1994) criteria, G. alba is critically endangered (criterion B, area of occupancy less than 2.5 km2; B1, severely fragmented, and continuing decline observed in, B2 (b) area of occupancy and B2 (c) number of sub-populations; Figure 1,Tables 2 and 3; Wardell-Johnson et al. 1995). Geocrinia vitellina has a much smaller range at about 6 km2 with only six known populations but all of these are in State Forest or are proposed as conservation reserves and under no immediate threat from clearing or logging activity (Appendix 2 of Wardell-Johnson et al. 1995; Figure 1). Geocrinia vitellina is vulnerable based on IUCN (1994) criterion D2. Spicospina flammocaerulea was discovered in 1994 and described in 1997 (Roberts et al. 1997). Field work in 1997 raised the number of known populations to 13.This species has a limited extent of occurrence (around 194 km2), small area of occupancy (0.78 km2 calculated by assuming an area of occupancy of 0.06 km2 (probable area at Mountain Road one of the largest sites) at each site and multiplying by 13)) and a fragmented range (Figure 2). Based on these data it is clearly vulnerable (criterion D2). However, this evaluation takes no account of possible variation in population size. In Table 4, we report data on number of calling males at several sites from 1994 to 1997. At Mountain Road, numbers of 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 may be worse. Field work in 1998 and 1999 is designed to assess population size more directly using mark-releaserecapture techniques and survey of tadpole populations. Spicospina flammocaerulea is found in peat based swamps near the West Australian south coast, east and north-east of Walpole in an area of moderate relief with granite outcrops and associated ranges of hills rising to 300 — 400 m (e.g. Mt Frankland 411 m, Mt Roe 340m; Mt Lindesay 440 m; Roberts et al. 1997). Five populations are on private property north-west of Bow Bridge, with the remainder in the Mount Frankland National Park or on land designated to form part of the Mount Roe-Mount Lindesay National Park but not yet declared (Figure 2).
POPULATION SIZE: Geocrinia Driscoll (1996) reported counts of calling males and from mark-release-recapture studies counted total numbers of males calling over a season in breeding populations for G. alba and G. vitellina.These data indicate that for G. alba,
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 found in the population over the whole season. For G. vitellina this figure is slightly lower at 83%.This means counts of calling males are a reasonably accurate estimator of 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.
FIGURE 2: Distribution of Spicospina flammocaerulea in relation to land tenure. Shaded areas are privately owned land. Remainder is managed as state forest, or national park (see text). Arrow on inset indicates approximate location of area covered by main map.
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 WardellJohnson 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
TABLE 1: Number of frogs recorded (total over all sites) and number of locations where species heard, for eight south-west frog species with ranges largely or wholly in the intermediate rainfall zone (the wheatbelt) of Western Australia. Data collected 1988 - 1992 by Roberts. The complete range of all species was covered but not all species were sampled with equal intensity.
Species
No. recorded
No. locations heard
Neobatrachus kunapalari pelobatoides albipes
72 27 36
102 9 98
Pseudophryne occidentalis guentheri
264 333
90 192
Crinia pseudinsignifera subinsignifera
535 83
230 12
Heleioporus albopunctatus
106
77
Spearwood Creek was last deliberately burnt in a spring, fuel reduction burn in 1983. As part of the management plan for this species, Spearwood creek south of Denny Road and Geo Creek (the second and third most easterly records on 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 made at Spearwood North in October 1997 (Figure 3), a wild fire (an escape from a fuel reduction burn in an adjacent block) burnt 50% of the total area of occupancy for this species, including 85% of the fire exclusion zone. Within the fire exclusion zone, 23.6% of breeding habitat burnt at mild to moderate intensities (i.e. scorched foliage retained on Leptospermum and Agonis species; cf. Wardell-Johnson and Roberts 1993) and a further 25.6% was subject to an intense fire (total defoliation of the same plant species and almost complete removal of surface litter). Other areas were either not burnt or burnt at lower intensities.
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 privately versus publicly owned land, 1983 to 1997. 75 subpopulations known 1983–1996. There was no significant difference in extinction rates on privately and publicly owned land ( x2 = 0.05, d. f. = 1, p > 0.05)
Public Private
Extinct 1997
Extant 1997
6 17
15 37
TABLE 3: Extinction rates for sub-populations of G. alba on privately owned land related to adjacent vegetation. Cleared covers total clearing, partial clearing and sites cleared and developed for tree plantations. Intact includes all sites where there has been no or minor modification of adjacent upland vegetation. No data available for two new sites discovered in 1997.There was a significantly lower extinction rates at sites with intact, upland vegetation ( x2 = 6.32, d. f. = 1, p
