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 (<2% sequence divergence) can infect a range of Australian frog species (Morgan, unpubl.). More studies are needed to define significant variations.
DISTRIBUTION In Australia, Batrachochytrium has been found in frogs since 1989 and has been observed in various regions — including rainforests of southern, central and northern Queensland and
Figure 1: Map of Australia with stars indicating where Batrachochytrium has been found on amphibians. This distribution correlates with searching intensity.
AUSTRALIA
northern NSW, montane and foothill forests of Victoria, and suburban/semi-rural populated areas of Brisbane, Adelaide and Perth (Figure 1,Table 1). Captive axolotyls from Townsville (Speare, unpubl.) and Perth (Aplin and Kirkpatrick, unpubl.) were also found with chytridiomycosis. Overseas, chytrids have been found in rainforest anurans in montane Panama in 1997 (Berger et al. 1998), in captive frogs in National Zoo, Washington DC and other American zoos since 1988 (Nichols et al. 1998, Pessier et al. 1999), and in wild frogs in Arizona (Nichols et al. 1998) and Illinois (Pessier et al. 1999). Recently, chytridiomycosis was diagnosed in anurana that died in high altitude regions of Equador, including adults of Telmatobius niger collected in1989 from the Azuay province, adults of Atelopus sp. (bomolochos complex) collected in 1991 from the Riobamba province and captive metamorphs of Gastrotheca pseustes that died in 1999 (Merino-Viteri and Coloma, unpubl.). As investigations into frog disease become more common, the known distribution of the amphibian chytrid is likely to greatly expand, and by examining archived specimens the dates of first occurrence will be pushed back. Investigating the origin and historical range of the amphibian chytrid may determine whether it has been introduced or is endemic. Apparently healthy frogs can exist as carriers; clinically normal frogs (including L. genimaculata) collected from the wet tropics in Queensland have been found with low levels of infection (Speare and Freeman, unpubl.). Although Batrachochytrium has not yet been found in non-amphibian hosts, it is possible that other cold-blooded animals with keratinised surfaces could harbour infections and play a role in the persistence and spread of the disease. As the chytrid is susceptible to desiccation (Berger, unpubl.), transportation of the disease long distances is most likely to have occurred via the movement of amphibians, either deliberately (for example in the pet trade), or accidentally (such as in fruit boxes). Both these scenarios have been observed and epidemics in captive collections have occurred after the arrival of a few infected frogs (Marantelli, pers. comm.). Cane toads are also potential 25
TABLE 1: Species, locations and dates of amphibians in Australia found infected with Batrachochytrium.
Hylidae Litoria adelaidensis Litoria caerulea
Litoria Litoria Litoria Litoria
chloris citropa ewingi genimaculata
Litoria infrafraenata Litoria lesueuri
Litoria moorei
Litoria nannotis Litoria pearsoniana Litoria raniformis Litoria rheocola (tadpoles & adults) (tadpoles & adults) Litoria spenceri Nyctimystes dayi
Perth,WA Casino, NSW Brisbane, Qld Brisbane, Qld Rockhampton, Qld Mackay, Qld Bellingen, NSW Lanitza, NSW Dorrigo Plateau, NSW Captive, Sydney Woodville, SA Kirrama State Forest, Qld Windsor Tableland, Qld Topaz, Qld Wright’s Creek, Qld O’Keefe Creek, Qld Captive, Adelaide, SA Mary River, Qld Eungella NP, Qld Goomburra, Qld Buckwong Ck,Vic Perth,WA Ferndale, Perth,WA Manjimup,WA Mt Barker,WA Mt Helena,WA Sawyers Valley,WA Witchcliffe,WA Big Tableland, Qld Mebbin State Forest, NSW Kroombit Tops, Qld Captive, Adelaide, SA Big Tableland, Qld Westgid Ck, Qld Frenchman Ck, Qld Bogong Ck,Vic Taponga River,Vic Tully, Qld
Aug 1998 July1996 July–Sep1996 June–July 1997 Aug 1997 Sep 1997 July 1998 Aug 1998 Dec 1997–Feb 1998 May 1999 Oct 1998 Nov 1989 April 1993 July 1998 Oct 1998 Dec 1998 Oct 1998 May-June 1996 June 1996 April, 1997 Feb 1998 Sept 1998 Nov 1998 Nov 1998 Nov 1998 Nov 1998 Nov 1998 Nov 1998 Oct 1993 Dec 1997 May 1998 Sep 1998 Oct 1993 May 1999 July 1999 March 1996 Feb–March 1998 Dec 1998
Brisbane, Qld Brisbane, Qld Lamington NP, Qld Sawyers Valley,WA Springwood, NSW Captive, Perth,WA Woodlands, Perth,WA Goomburra, Qld, Mt Compass, SA Captive Melbourne,Vic Valley View, SA Adelaide, SA Captive, Adelaide Cunningham’s Gap, Qld Captive Melbourne,Vic Goomburra Captive Melbourne,Vic Mt Glorious, Qld Cunningham’s Gap, Qld Captive, Melbourne,Vic Carbine Tableland, Qld Big Tableland, Qld Captive Townsville, Qld Eungella National Park, Qld
Nov 1995 Aug 1998 Jan 1999 Nov 1998 May 1998 Nov 1998 May 1998 Dec 1996 Oct 1998 Nov 1998 Jan 1999 May–June 1996 Aug 1998 Jan 1999 June–Aug 1996 Dec 1996 Dec 1996–Feb 1997 Aug 1998 Aug–Oct 1996 Nov 1997 Nov 1990 Nov 1993–Jan 1994 Dec 1993 Oct 1995
Captive Geelong,Vic
June–Aug 1996
Captive Townsville, Qld
Feb 1999
# * * *
*
*
# # # # # # # # *
* ® ® *
J A 96/692 A 96/961 A 97/845 A 97/845 A 97/845 A 98/871 A 98/871 A 97/1754 A 99/232 A 98/1331 A 98/927 J A 98/1231 J 98/398 J 99/28 A 99/232 A 96/570 A 96/1136 A 97/574 A 98/320 A 98/927 J 99/12 J 99/12 J 99/12 J 99/12 J 99/12 J 99/12 J 93/290 A 97/1747 A 98/320 A 98/1331 J 93/290 J J A 96/370 A 98/320 A 98/1810
Myobatrachidae Adelotus brevis
Crinia pseudinsignifera Heleioporus australiacus Heleioporus eyrei Limnodynastes dorsalis Limnodynastes dumerilii
Limnodynastes tasmaniensis Lechriodus fletcheri Neobatrachus kunapalari Mixophyes sp. (tadpoles) Mixophyes fasciolatus Mixophyes fleayi Pseudophryne pengilleyi Taudactylus acutirostris
Taudactylus eungellensis
* * *
A 99/927 A 99/927 A 99/951 J A 98/871 J J A 96/1429 A 98/1331 A 98/1469 A 99/232 A 96/495 A 98/871 A 99/951 A 96/1431 A 96/1428 A 96/1431 A 98/1159 A 96/ 962 A 97/1510 A 98/927 J 93/290 J 93/290 A 96/657
*
A 96/673
# # # *
*
* *
Bufonidae Bufo marinus Ambystomatidae Ambystoma mexicanum
Accession numbers are provided: "A" refers to specimens deposited at AAHL, "J" refers to specimens deposited at JCU. * denotes records from Berger et al. 1998. # denotes records from Aplin et al. 1999 “ denotes records from Retallick 1999
26
J
carriers as they continue to expand their distribution in Australia, as evidence suggests that free-ranging toads are infected with chytrids1 (Parkes, pers. comm.). We hypothesise that Batrachochytrium was introduced to Australia in the 1970’s around Brisbane (where the first precipitous declines occurred), although as yet there are no hard data to support this assumption. Batrachochytrium has since become established in many areas on the east coast, around Adelaide and in south west Western Australia.
FIGURE 2: Captive-bred metamorph of Mixophyes fasciolatus (AAHL accession no. 96/1431) from the Amphibian Research Centre with naturally acquired chytridiomycosis. This frog is in the terminal stages of the disease - note depressed attitude, half closed eyes and accumulations of sloughed skin over the body (arrow head). Bar = 0.5 cm.
CHYTRIDIOMYCOSIS:THE DISEASE Typical clinical signs in Australian frogs with chytridiomycosis were lethargy, inappetence, skin discolouration, presence of excessive sloughed skin, and sitting unprotected during the day with hind legs held loosely to the body (Berger, Speare and Marantelli, unpubl.). Frogs became moribund in terminal stages with loss of righting reflex, and death usually occurred a few days after the onset of clinical signs. Most diseased frogs were in reasonable body condition and 45 out of 69 were considered to have moderate or large fat bodies. Gross pathology of internal organs was generally unremarkable. Some notable species variation in the clinical signs was observed. The rainforest frogs from Big Tableland, northern Queensland (Speare 1994) predominantly showed neurological signs; most commonly, abnormal sitting posture with hind legs adducted, lethargy, and slow response to tactile stimuli. When handled, these frogs (Litoria rheocola, L. nannotis, Taudactylus acutirostris) became rigid and trembled with extension of the hindlimbs and flexion of the forelimbs. Less common signs included minute skin ulcers and haemorrhage in skin, muscle or eye. Many frogs appeared anaemic, with pale muscles and internal organs. In contrast, individuals of Litoria caerulea often became intensely red on the belly, legs and feet. These frogs had marked congestion and reddening of organs internally, and skin ulcers were occasionally seen (Berger and Hines, unpubl.) Captive metamorphs of Mixophyes fasciolatus became slow and sloughing skin accumulated over the body (Figure 2). The presence of shedding skin on the body is a nonspecific sign which is seen in other diseases where frogs become lethargic and fail to wipe themselves, but with chytridiomycosis, excessive shedding may occur in response to infection of the skin. In frogs with chytridiomycosis, light microscopy of unstained pieces of sloughing skin revealed masses of round to oval refractile fungi (Figure 3), enabling rapid diagnosis. On histologic examination, chytrid fungi were observed inhabiting the outer keratinised layers of the epidermis (Figure 4) and were associated with thickening or erosions and occasional ulcerations of the skin (Berger et al. 1998, Pessier et al. 1999). Different frogs varied greatly in their intensity of infection, but skin of the ventral body, limbs and feet were most consistently infected. Bacteria often colonised the layers of sloughing keratin. There was negligible inflammatory response in the dermis. Scanning electron microscopy of diseased skin revealed fungal discharge tubes emerging through the surface of the skin and confirmed
FIGURE 3: Unstained skin slough from a dead adult of Litoria caerulea from Thagoona, Qld (AAHL accession no. 97/845) examined by light microscopy. Note refractile round and oval chytrid fungi. Most are empty, but one contains developing zoospores (arrow). Occasionally empty sporangia are seen divided by thin septae. E = epidermal cell. Bar = 20 µm.
FIGURE 4: Histological section of heavily infected skin from the toe of a wild-caught adult of Litoria caerulea (AAHL accession no. 97/845). Batrachochytrium does not invade through the epidermis but occurs just under and within the superficial keratinised layer which becomes thickened. Solid, immature sporangia are present (arrow), as well as mature sporangia containing numerous dark zoospores (Z). Many sporangia are empty after the flagellated zoospores have swum out through the discharge tube (below the D). Immature stages are present in the superficial layer of viable epidermal cells and older, empty stages are commonly seen in the outer dead layers of keratin which are sloughed. Batrachochytrium does not form hyphae. (Haematoxylin & eosin). E = epidermis, K= keratin. Bar = 50 µm.
1 Fourteen out of 63 wild-caught juvenile toads from Lennox heads, NSW which were brought into AAHL in june 1996 died between 1 and 44 days after arrival and chytridiomycosis was diagnosed in 5 which were examined histologically (Parkes, unpubl. data). 27
FIGURE 5: Scanning electron micrograph of skin from an infected adult of Litoria lesueuri from Goomburra, Qld (AAHL accession no. 97/5741), showing fungal discharge tubes protruding through the surface of epidermal cells. Bar = 10µm.
hyperkeratosis and extensive sloughing of the epidermis (Figures 5 and 6). The internal structure of Batrachochytrium was revealed by transmission electron microscopy (Figure 7). The lack of inflammation in the skin could be due to a lack of stimulation of the host immune system – perhaps due to the superficial site of infection, insufficient epidermal damage, or the chytrid may have low inherent antigenicity. There were few specific internal lesions in sick frogs, suggesting that the ultimate cause of death was metabolic or toxic. Focal necroses, vacuolation, or cloudy swelling were sometimes apparent in a range of internal organs (Speare and Berger, unpubl.).
FIGURE 6: Scanning electron micrograph of skin from the toe of an infected adult of Litoria lesueuri (AAHL accession no. 97/5741) with extensive peeling and degeneration of the keratinised skin surface. The surface of healthy skin from control frogs appeared smooth. Bar = 100µm.
FIGURE 7: Transmission electron micrograph of a mature sporangium cultured from the skin of a sick Nyctimystes dayi from Tully, Qld (AAHL accession no. 98/18103). It is filled with developing zoospores, each with a single flagellum (arrow head). The discharge tube is closed with a gelatinous plug which deliquesces when the zoospores are ready to be released. Bar=3µm.
Histologic examination of organs involved in immunity — i.e. spleen and bone marrow, revealed no evidence of immunosuppression, apart from in one frog. Tests (including viral and bacterial culture, electron microscopy and haematology) for infectious organisms did not detect any other significant pathogens. Of 147 wild and captive frogs with chytridiomycosis, concurrent disease was diagnosed in 16 (11%). However, apart from one frog with immunosuppression, these diseases were not considered to be the primary cause of death. The concurrent diseases included septicaemia (four frogs), microsporidial hepatitis (one frog) and hyphal mycotic dermatitis (two frogs), all of which may have occurred secondary to chytridiomycosis. The other diseases - biliary hyperplasia and/or fibrosis (six frogs), foreign body myositis (one frog) and mild, localised mucormycosis (one frog) — all appeared chronic and inactive and may not have contributed to the deaths. A variety of helminth and protozoan parasites were identified as incidental findings (Berger and Speare unpubl.). During our survey of amphibian diseases in Australia, 272 wild and captive sick frogs were examined between 1989 and 1999, but the only disease that was found consistently was chytridiomycosis which occurred in 54% and accounted for almost all cases of unusual mortality in the wild. A range of diseases was diagnosed in individual sick frogs that did not have chytridiomycosis, including viral, bacterial, protozoal, fungal, tapeworm, neoplastic, traumatic and congenital diseases, and inappropriate ingestion (Berger and Speare, unpubl.). Batrachochytrium is not a ubiquitous parasite. In a histological survey of toeclips from 348 apparently healthy frogs from Queensland and Victoria collected between 1989 and 1999 only 7 (2.0%) were infected (Speare, Berger and Kent, unpubl.). In experimental infections in Australia using Mixophyes fasciolatus a terminal illness was reached in 6/6 frogs 10–18 days after exposure to infected skin scrapings at 24C (Berger et al. 1998). In the USA, 2/2 Dendrobates tinctorius died 23 and 31 days after being exposed to broth culture of B. dendrobatidis (Longcore et al. 1999). Small doses of the pathogen have now been shown to cause fatal chytridiomycosis in metamorphs of Mixophyes fasciolatus — 3/3 frogs each exposed to an estimated 1000 zoospores died or became terminally ill between 23 and 38 days post exposure, 3/3 frogs exposed to approximately 100 zoospores died between 35 and 47 days post exposure, however 3 frogs exposed to approximately 10 zoospores did not succumb to chytridiomycosis and 2 have remained healthy for over 3 months (the other frog died by misadventure) (Marantelli and Berger unpubl.). The metamorphs were
28
housed individually and were infected by bath exposure to zoospores (cultured from a captive Limnodynastes dumerilli) in 10ml water for 24 hours.
Once a new pathogen has emerged, the rapid global transport by air of people, goods and animals will assist in its dissemination both within and between continents.
In epidemics in captive collections of M. fasciolatus and Bufo marinus, almost all tadpoles survived to metamorphose, but suffered high mortality rates 2-3 weeks after metamorphosis (Berger et al. 1998). Tadpoles were found infected with Batrachochytrium, but were never found dead or dying as a consequence of infection (Berger et al. 1998). Tadpoles were infected in the mouthparts (see Berger et al. 1999), the only keratinised area of tadpole skin. After metamorphosis, the skin over the body becomes keratinised allowing the fungal infection to spread. The resistance of tadpoles to chytrid fungi is consistent with the epidemiology of the amphibian declines where tadpoles have been seen after adults have disappeared (McDonald and Alford 1999, Lips 1999). During the declines in Panama in 1997, many abnormal tadpoles were seen with partial or complete loss of keratinised mouthparts, although tests for fungal infection in these animals were not reported (Lips 1999). Tadpoles with abnormal mouthparts have not been observed in Australia.
If long-term degradation of the environment was the key problem in declines of amphibian populations, we would expect reproductive and nutritional status to be affected before fatal immunosuppression occurs. However, moribund frogs were found which were gravid (Mahony 1996), and we have seen many with adequate fat reserves. Also, with severe immunosuppression a range of opportunistic infections is likely to be involved, rather than a solitary chytrid fungus. The experiments in captivity demonstrate that the chytrid fungus can cause 100% mortality in conditions where uninfected animals remained healthy. There is no evidence that predisposing immunosuppression is necessary for epidemics of chytridiomycosis to occur.
Two theories have been proposed to explain how a fungus restricted to the superficial epidermis could kill frogs (Berger et al. 1998, Pessier et al. 1999). The first hypothesis is that the chytrid releases proteolytic enzymes or other active compounds that are absorbed through the permeable skin of the frog. The second hypothesis is that damage to skin function results in disturbance of water or electrolyte balance resulting in death.We tend to favour the first explanation. Anurans have evolved to manage water stress by shifting water and electrolytes from the brain and cerebrospinal fluid and can withstand levels of dehydration that would be fatal to other classes of vertebrates (Hillman 1988). The epidermal damage caused by chytridiomycosis does not appear severe enough to result in the major changes in water and electrolyte balance required to cause death.
DISCUSSION AND IMPLICATIONS The observations of high mortality rates resulting in the disappearance of populations suggests there has been some recent change in the balance between host and pathogen. Three explanations are suggested: 1. Batrachochytrium has been introduced recently to these areas, 2. it has suddenly become more pathogenic to the host, or 3. the frogs have a lowered resistance due to environmental changes allowing the endemic parasite to cause mass mortalities. The first explanation seems the most plausible, due to the epidemiological factors already discussed, and acute or chronic environmental problems in these protected rainforest areas were not detected (Richards et al. 1993). The amphibian chytrid fungus may have originated by the escape of a pathogen from a local environment, transfer to new hosts, followed by rapid progression through these non-essential hosts. Two recent examples of emerging infectious diseases following this pattern are Ebola virus in Africa and Vibrio cholerae 0139. The first has had several episodes of escape, from an as yet undetected reservoir host into the human population of Zaire (Dobson and Carper 1996), while the second, an environmental bacterium, arose from the Bay of Bengal in 1992 and has spread rapidly through the human population of Asia (Lee et al. 1996).
Although Batrachochytrium has a wide distribution and broad host range, we suspect only some species had the necessary combination of characteristics to render them vulnerable to decimation by disease. The declining species from high altitude rainforests in eastern Australia have significantly smaller clutch sizes, occupy restricted geographic ranges, have aquatic larvae associated with streams, and many spend a large proportion of their time in or adjacent to streams (Williams and Hero 1998, McDonald and Alford 1999). Populations of these species are therefore less able to recover from declines due to any cause, and also inhabit environments that would support Batrachochytrium i.e. cooler, riparian habitats. Other aspects of host biology, genetic factors (see Leberg and Vrijenhoek 1994), humidity and complex ecological factors affecting the lifecycle of the chytrid, could be important in determining the selectivity of the declines. As Batrachochytrium can probably exist as a free-living organism in the environment (Longcore et al. 1999), can be carried by healthy tadpoles and has a broad host range (Berger et al. 1998), this may explain an ability to persist and cause disease even when the density of adults from particular species has been greatly reduced (Daszak et al. 1999). Although most of the frog species that have disappeared from Queensland rainforest since 1979 (e.g. Rheobatrachus vitellinus, R. silus, Taudactylus diurnus, T. rheophilus, Litoria lorica and L. nyakalensis (Tyler 1997)) were not found dying and tested for disease, the epidemiological evidence suggests that chytridiomycosis caused mass mortalities of these species, as was observed in declining frogs (including T. acutirostris) at the Big Tableland site in north Queensland in 1993 (Berger et al. 1998). The possible role of chytridiomycosis in frog declines in southern Australia is less clear. The episodes of population declines described above now appear to have passed in many areas. Numbers of some species, decreased significantly when other species disappeared, but are now increasing.These species include T. eungellensis (McDonald and Alford 1999), L. genimaculata on Big Tableland (McDonald unpubl.) and L. pearsoniana from southeast Queensland (Hero et al. 1998). As residual populations are increasing after significant declines although chytrids are still present (Berger et al. 1998), this suggests resistance is present in some members of the population. A balance between Batrachochytrium and the frogs may be 29
developing in areas where we suspect it has been present for at least five years, and although disease still occurs, it does not devastate the populations. Batrachochytrium may now behave as an endemic pathogen with outbreaks of disease occurring when conditions are optimal. The long term prognosis may be good for species which have survived and are recovering, as long as remaining habitats are protected to allow damaged populations to reestablish. However, since we currently know little about the interaction between the amphibian chytrid and hosts, and many frog species are currently considered threatened or their status is insufficiently known (Tyler 1997), we should not be complacent. Also, if uninfected areas still exist in Australia then locally endemic species in those areas may be at high risk.
distribution of chytridiomycosis by examination of frogs preserved in museum collections to determine whether chytrids were introduced to Australia or to these habitats and to determine from where and how chytrids have spread. Information on the current distribution may also help in formulating management plans to prevent new outbreaks — for example, in deciding where exposed or uninfected frogs should be released after captive breeding. This survey relies on a collaborative approach (see Appendix 1). As many healthy tadpoles may carry chytrids for extended periods, sampling of tadpoles may provide a more sensitive means of assessing a location. Examination of sick frogs is the most sensitive way to detect chytrids, but these are not always available.
Infectious disease is important in the population biology of wild animals, as it is in humans and domestic animals (May 1988). A review of infectious disease and animal populations concluded that disease is an important factor affecting survival, reproduction, dispersal, community structure and genetic diversity, and should therefore be considered by ecologists examining host population-dynamics (Scott 1988). Disease becomes a threatening force when environmental degradation puts pressure on populations, and international trade and smuggling continually threaten to introduce new pathogens to which the native fauna has no resistance (Scott 1988, Daszak et al. 1999). Exotic diseases can have effects similar to those of feral predators, with susceptible native species facing extinction while the ecosystem readjusts, and are another example of how increased global homogeneity leads to reduced biodiversity. Although the initial >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 (