I ’ t I ( jr.^ Journal of the Royal Society of Western Australia CONTENTS Recent Advances in Science in Western Australia Royal Society of Western Australia Medal Recipients 1993 Medal Recipient: Professor J R De Laeter Page 1 3 4 A question of time: Royal Society Medallist's Lecture for 1993 J R De Laeter The impact of prolonged flooding on the vegetation of Coomalbidgup Swamp, Western Australia R H Froend and P G van der Moezel Rottnest Island artifacts and palaeosols in the context of Greater Swan Region prehistory C E Dortch and P A Hesp Volume 77 Part 1 March 1994 ISSN 0035-922X The Royal Society of Western Australia To promote and foster science in Western Australia PATRON Her Majesty the Queen VICE-PATRON His Excellency Major General Michael Jeffrey AO MC Governor of Western Australia COUNCIL 1993-94 President W A Cowling B Agric Sc (Hons) PhD Immediate Past President V Semeniuk BSc (Hons) PhD Vice-Presidents D I Walker BSc (Hons) DPhil S Hopper BSc (Hons) PhD Joint Hon Secretaries L N Thomas BSc MSc P Gardner BEng (Hons) GDipCSci DipEd V Hobbs BSc (Hons) PhD Hon Treasurer JDodd BA MSc PhD Hon Editor P C Withers BSc (Hons) PhD Hon Journal Manager B T Steer BSc PhD Hon Librarian M A Triffitt BA ALIA Members BDell BSc (Hons) PhD D K Glassford BA (Hons) PhD D Gordon BSc (Hons) PhD N G Jablonski AB PhD L E Koch BSc MSc PhD K McNamara BSc (Hons) PhD The Royal Society of Western Australia was founded in 1914. The Society promotes exchange among scientists from all fields in Wests ^ Australia through the publication of a journal, monthly meetings where interesting talks are presented by local or visiting scientists, occasional symposia or excursions on topics of current importance. Members and guests are encouraged to attend meetings on the thi Monday of every month (March-December) at 8 pm. Kings Park Board offices. Kings Park, West Per^, WA 6005. Individual membership subscriptions for the 1993/94 financial ial year are $40 for Ordinary Members and $20 for Student and re $60 for the 1994 calendar year. For membership forms, contact ripl Members. Library, company and institution subscriptions are _ __ Membership Secretary, c/o WA Museum, Francis Street, Perth, WA 6000. The Journal of the Royal Society of Western Australia was first published in 1915. Its circulation exceeds 600 copies. Nearly 100 of th^L are distributed to institutions and societies elsewhere in Australia. A further 200 copies circulate to more than 40 countries. The also has over 350 personal members, most of whom are scientists working in Western Australia. The Journal is indexed and abstra^^ internationally. Coucr Design: Mangles' kangaroo paw {Anigozanthos manglesii) and the numbat {Mxfnnecobiusfasciatus) are the floral and faunal of Western Australia. Alsodepictedisacollectionoflivingstromatolites which areof particular significance in Western Australian go^^ ^ The three subjects symbolize the diversity of sciences embraced by the Royal Society of Western Australia. (Artwork: Dr Jan Tayl^^^^' Journal of the Royal Society of Western Australia CONTENTS Recent Advances in Science in Western Australia Convergent evolution in the dentitions of grazing macropodine marsupials and the grass-eating cercopithecine primate Theropithecus gelada. N Jablonski Re-examination of the Murchison Downs meteorite: A fragment of the Dalgaranga mesosiderite? A W R Bevan and B J Griffin Invertebrate community structure related to physico-chemical parameters of permanent lakes of the south coast of Western Australia D H D Edward, P Gazey and P M Davies Page 33 Volume 77 Part 2 June 1994 ISSN 0035-922X The Royal Society of Western Australia To promote and foster science in Western Australia PATRON Her Majesty the Queen VICE-PATRON His Excellency Major General Michael Jeffery AD MC Governor of Western Australia COUNCIL 1993-94 President Immediate Past President Vice-Presidents Joint Hon Secretaries Hon Treasurer Hon Editor Hon Journal Manager Hon Librarian Members W A Cowling V Semeniuk D I Walker S Hopper L N Thomas P Gardner V Hobbs B Agric Sc (Hons) PhD BSc (Hons) PhD BSc (Hons) DPhil BSc (Hons) PhD BSc MSc BEng (Hons) GDipCSci DipEd BSc (Hons) PhD J Dodd BA MSc PhD P C Withers B T Steer M A Triffitt BDell BSc (Hons) PhD BSc PhD BA ALIA BSc (Hons) PhD D K Glassford BA (Hons) PhD D Gordon BSc (Hons) PhD N G Jablonski AB PhD L E Koch BSc MSc PhD K McNamara BSc (Hons) PhD AustraHa^hrouJh^P^i^hHr inl914. The Society promotes exchange among scienHsts from all fields in VVested occasional ^ ^ purnal monthly meetings where interesting talks are presented by local or visiting scientists, an^ Minda^of mnnlh c“'^''enOmportance. Members and guests are encouraged to attend meetings on the thir‘J Monday of every month (March-December) at 8 pm. Kings Park Board offices, Kings Park, West Per^, WA 6005 i ' —o -- —4 oi rv, T * c::>i I Cl III, rvrvuuuj. subsciy tions for the 1993/94 financial year are $40 for Ordinary Members and $20 for Student and Associa‘^ 'xif **^^^*^^*^*^ sub^nptions are $60 for the 1994 calendar year. For membership forms, contact t*^ MnQPiim Cl-Mfs*- r>r.«^<.U lAfA /:nnri * ' Members. Library, company ancf institution subscriptions are >60 tor I Membership Secretary, c/o WA Museum, Francis Street, Perth, WA 6000. ar JdfstriWed^m^ VVestern Australia was first published in 1915. Its circulation exceeds 600 copies. Nearly 100 of th^s also has over 350 ^ K elsewhere in Australia. A further 200 copies cirailate to more than 40 countries The Societ internatiCnally^^ personal members, most of whom are scientists working in Western Australia, The Journal is indexed and abstract^' of Weste^!Sstraha^AIsm?p^ (Aw/^02flnf/jps and the numbat (Myrmecohius fasciatus) are the floral and faunal emhienj suD)eas symbolize the diversity of sciences embraced by the Royal Society of Western Australia. (Artwork: Dr Jan Taylor)- CONTENTS Page Biosystematics of Australian mygalomorph spiders: I Description of a new species of Anarne and its aerial 1 tube (Araneae: Nernesiidae) ' B York Main 65 I Wet heathlands of the southern Swan Coastal Plain, Western Australia: A phytosociological study R S Smith and P G Ladd 71 Seed dispersal of Hibbertin hi/pericoides (Dilleniaceae) by ants. A Schatral, S G Kailis and JED Fox Holdings in the Library of The Royal Society of Western Australia OF VICTORIA 53596 Volume 77 Part 3 September 1994 ISSN 0035-922 X The Royal Society of Western Australia To promote and foster science in Western Australia PATRON Her Majesty the Queen VICE-PATRON His Excellency Major General Michael Jeffery AD MC Governor of Western Australia COUNCIL 1994-95 President D I Walker BSc (Hons) DPhil Immediate Past President W A Cowling B Agric Sc (Hons) PhD Vice-Presidents S Hopper BSc (Hons) PhD M G K Jones MA PhD Joint Hon Secretaries L N Thomas BSc MSc PG Dip EIA P Gardner BEng (Hons) GDipCSci DipEd V Hobbs BSc (Hons) PhD Hon Treasurer R Froend BSc (Hons) PhD Hon Editor P C Withers BSc (Hons) PhD Hon Journal Manager J E O'Shea BSc (Hons) PhD Hon Librarian M A Triffitt BA ALIA Members BDell BSc (Hons) PhD JDodd BA MSc PhD D K Glassford BA (Hons) PhD D Gordon BSc (Hons) PhD K McNamara BSc (Hons) PhD V Semeniuk BSc (Hons) PhD The Royal Society of Western Australia was founded in 1914. The Society promotes exchange among scientists from all fieldsin j Australia through the publication of a journal, monthly meetings where interesting talks are presented by local or visiting scientists, occasional symposia or excursions on topics of current importance. Members and guests are encouragecf to attend meetings on the tni Monday of every month (March-Deccmoer) at 8 pm. Kings Park Board offices, Kings Park, West Perth, WA 6005. Individual membership subscriptions for the 1994/95 financial year are $40 for Ordinary Members and $20 for Student and Associ^*^ Members. Library, company ana institution subscriptions are $60 for the 1994 calendar year. For membership forms, contact Membership Secretary, c/o WA Museum, Francis Street, Perth, WA 6000. The Journal of the Royal Society of Western Australia was first published in 1915. Its circulation exceeds 600 copies. Nearly 100 are distributed to institutions and societies elsewhere in Australia. A further 200 copies circulate to more than 40 countries. The also has over 350 personal members, most of whom are scientists working in Western Australia. The Journal is indexed and abstra internationally. Cover Desipi: Mangles' kangaroo paw (Anigozanthos man^lesii) and the numbat (Myrmecobiusfasciatus) are the floral and of W^tern Australia. Also depicted is a collection of living stroma tolites which are of particular significance in Western Australian g*?^ f The three subjects symbolize the diversity of sciences embraced by the Royal Society of Western Australia. (Artwork: Dr Jan Tayl^^^^ JOURNAL OF THE ROYAL SOCIETY OE WESTERN AUSTRALIA CONTENTS VOLUME 77 PART 4, DECEMBER 1994 Diseases in Ecosystems: pacts in south-western Australia Page 97 99 101 103 105 107 109 113 Foreword. W A Cowling and R T Wills Symposium Summary. S H James Session 1: Biology. K Old Session 2: Impact on Ecology. S Hopper Session 3: Impact on Industry. L Mattiske Session 4: The Future. S H James Ecosystem pathogens: A view from the centre (east). P Bridgewater and B Edgar The major plant pathogens occurring in native ecosystems of south-western Australia. B L Shearer Role of environment in dieback of jarrah: Effects of waterlogging on jarrah and Phytophthora cinnamomi, and infection of jarrah by P, cinnamomi. E M Davison 123 Ecological impact of plant disease on plant communities. R T Wills and G J Keighery 127 Smut and root rots on native rushes (Restionaceae) and sedges (Cyperaceae). K A Websdane, I M Sieler, K Sivasithamparam and K W Dixon 133 Impact of plant diseases on faunal communities. B A Wilson, G Newell, W S Laidlaw and G Friend 139 Disease and forest production in Western Australia with particular reference to the effects of Phytophthora cinnamomi, D S Crombie and F J Bunny 145 The impact of plant disease on mining. IJ Colquhoun and A E Petersen 151 Threats to flora-based industries in Western Australia from plant disease. R T Wills and C J Robinson 159 Management of access. K Gillen and A Napier 163 Control options of plant pathogens in native plant communities in south-western Australia. G E St J Hardy, P A O'Brien and B L Shearer 169 Future ecosystems — use of genetic resistance. J A McComb, M Stukely and IJ Bennett. 179 Future ecosystems — ecological balance (ecological impact of disease causing fungi in Western Australia). G J Keighery, D J Coates and N Gibson 181 The future — effects of plant diseases on society. J T Young 185 iii ^ ^ a t» . « ‘. i *ij- Ti « # 1 , b ' .■ * v* ^ ^" f, *•'■ « '* !•* * •■' - ^S^MfiS|fl&4i> ■ wy w- >»\ *. . iw*«’ - > „ * < -3^ ' - ■-.■ y \ ’ -•, 4.-,- '• ■■ ■ ■ " ^ -K. .. -#■ , V’'-. .-^ 7^. ■*•*-. ■' ,,, j • , » ■' ^•V sOi € .. rr-» ;-■.: '* ■# .;•';t^vM^ #» "S.. «>!. ... , ^ Li.i-”'’' Ai^ i. '■ '4¥^ jinh jii«>t*<- ■_ ’^‘* .' ^ - 1- .-■'’ ^-T«S< W^i: V.vj'" ^ jp,,^ t-1 • ■^H'W’i'.i^ ‘I'j ,-i;< X. ■“>.'IT '*»«!«*^r itl* ' y^USi*' •■ • • ■ ^icfSE^iV ; Aajf i 5,tJ^ Sr ih- ‘ 1 f ‘ r-^42f - “it ' • . . .^-i •* ^ ♦ V-ll- -» .1i>'I / .1 ?S- vrRiyj*»‘ - ■ ‘-J*. ?. » '^' .'f* ‘ *r > ■ , V * ^ ' A" '•mt. t ■» ■ {'* ’ ■« p9 ^, v3w ^ ‘j« ' %■* \: ’ . . Ik ^ ^ ■ »_ - • .. I'f ^ _■ • V. S '^' ^ Jr- .LH; n* ■ 's ■' ... rtf; .1 r '.’• ^ %«£»•' . A-.- M ^ . 'S; T 4 *7' Journal of the Royal Society of Western Australia, 77 (1), March, 1994 || / . i — i- iA^ i Recent Advances in Science in Western Australia Earth Sciences W K Witt of the Geological Survey of Western Australia, Perth, describes how the Menzies-Kambalda region has produced 1700 tonnes of gold, with most production accounted by mafic rocks, especially Fe-rich basalt and differentiates of fractionated sills. Mineralization occurs in diverse structural settings, related to the latest stages of regional deformation or emplacement of comtemporaneous granitoid intrusions, that were active during gold-related hydrothermal activity. Witt W K 1993 Gold mineralization in the Menzies-Kambalda region. Eastern Goldfields, Western Australia. Geological Survey of Western Au.stralia. Report 39. Two iridium anomalies have been identified by R S Nicoll (AGSO Canberra) and P E Playford (GSWA Perth) near the Frasnian-Famennian boundary in the Canning Basin reef complexes, one above and one below the actual boundary; neither were associated with the extinction event in the Zone. The anomalies were associated with bedscontainingabundantFnitote microstromatolites, indicating that the iridium was probably organically concentrated, rather than being associated with an impact event. Nicoll R S & Playford P E 1993 Upper Devonian iridium anomalies, conodont zonation and the Frasnian-Famennian boundary in the Canning Basin, Western Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 104:105-113 Coworkers from Durham and Oxford (U. K.) used laboratory studies of quantifiable rock geotechnical parameters to understand slope form and development. Highly concave slopes have formed in limestone, which has little deformation before yield, whereas convexo-concave slopes are characterized by material with a greater strain before yield, a relatively low modulus of elasticity, and a pronounced discontinuity pattern. Allison R J, Goudie, A S, & Cox N J 1993 Geotechnical properties of rock ma.sscs: their control on slope for, and mechanisms of change along the Napier Range, Western Australia. Geomorphology 8:65-80 A simplified scheme is presented by J R Vearncombe, of the University of Western Australia, to categorize quartz vein morphology from gold deposits, based on the growth direction of quartz or pseudomorphed chalcedony in the veins, into seven types; face-control, displacement-control, parallel-control, radiating, non-directional-control, replacement, and modified. Crustal depth is a major control in the Yilgarn Craton deposits, with parallel and radiating textures near the surface, displacement and non-directional typical of mid-crustal, frontal from near surface to midcrustal, and replacement and modified at all crustal levels. Vearncombe J R 1993 Quartz vein morphology and implications for formation depth and classification of Archean gold-vein deposits. Ore Geology Reviews 8:407-424 Ten articles concerning a variety of topics have been published in the latest Professional Paper volume of the Geological Survey of Western Australia, including five of © Royal Society of Western Australia 1994 hydrogeological interest; Salinity control at Lake Toolibin (M W Martin); Hydrogeology of the Cervantes-Lancelin Region (A M Kern); Point sources of groundwater contamination in the Perth Basin (K-J Hirschberg); Municipal waste disposal in Perth and its impact on groundwater quality (K-J Hirschberg); Hydrogeology of the Collie Basin (] S Moncrieff). Professional Papers. Geological Survey of Western Australia. Report 34 Life Sciences A survey of the tick load in natural populations of sleepy lizards in South Australia, by M. Bull and D Burzacott of Flinders University, showed that lizards do not appear to be adversely affected by high tick loads, as neither size nor longevity was negatively correlated with tick load, and mating pairs had higher tick loads than non-mating individuals. Bull C M & Burzacott D 1993 The impact of tick load on the fitness of their lizard hosts. Oecologia 96:415-419 The diets of arid zone dasyurid marsupials are shown by D Fisher and C Dickman, of the University of Sydney, to consist primarily of beetles, spiders, scorpions and centipedes. Small dasyurids avoided beetles with hard cuticles, and generally preferred prey 5 to 7.5 mm long over smaller (<2.5 mm) prey. Fisher D O & Dickman C R 1993 Diets of insectivorous marsupials in arid Australia; selection for prey type, size or hardness? Journal of Arid Environments 25:397-410 The pattern of torpor of the Eastern pygmy possum was shown by F Geiser of the University of New England to be regular in the laboratory, with increasing occurrence at lower air temperatures. The duration of torpor lengthened with lower air temperature, to about 17 days at 5 C, and the minimum metabolic rate declined to less than 2% of the basal rate. Geiser F1993 Hibernation in the Eastern pygmy possum, Cercartetus mamms (Marsupialia: Burramyidae). Australian Journal of Zoology 41:67-75 A model of water regulatory efficiency developed for granivorous parrots by R MacMilien, of the University of California at Irvine, and R Baudinette of Flinders University, suggests that small size imparts a higher water efficiency, but requires them to rely on small seeds rich in carbohydrates, as these provide maximal yields of metabolic water. MacMilien R E & Baudinette R V 1993 Water economy of granivorous birds: Australian parrots. Functional Ecology 7:704- 712 The collembolan faunas of rehabilitated bauxite mines arc shown by P Greenslade (CSIRO, Canberra) and J Majer (Curtin University) to have increased species richness in larger plots, with greater plant species richness and percentage plant cover; their results provide directions for improving rehabilitation practices. Greenslade P & Majer J D 1993 Recolonization by Collembola of rehabilitated bauxite mines in Western Australia. Australian Journal of Ecology 18:385-394 15517—1 1 Journal of the Royal Society of Western Australia, 77 (1), March, 1994 The standard metabolic rate of Western Australian frogs was shown by P Withers of the University of Western Australia to be similar to that predicted for other anurans. Species of Neobatrachus and Cyclorana had a depressed metabolic rate (to 20-30% of standard) during aestivation. Withers P C 1993 Metabolic depression during aestivation in the Australian frogs, Neobatrachus and Cyclorana. Australian Journal of Zoology 41:467-473 Physical Sciences Chemists from the School of Mathematical and Physical Sciences at Murdoch University have developed a potentiometric titration method and determined copper (I) equilibrium constants in aqueous solution for cyanide and D-penicillamine complexes. The copper (I) solutions are prepared by reduction of the common copper (II) state with excess copper metal, and are stabilised by chloride. Hcfter. G T, May P M & Sipos P 1993 A general method for the determination of copper (I) equilibria in aqueous solution. Journal of the Chemical Society Chemical Communicalions:l704-1706 An experimental and theoretical study, by physicists of theUniversity of Western Australia and Murdoch University, of the angular correlations of sequential cascading photons in an atomic hydrogen system with a defined scattering plane demonstrates that the measured correlations and the deduced multiple moments show an order-of-magnitude agreement with various theoretical models. Williams J F, Kumar M & Slelbovics A T 1993 Angular correlations between sequential cascading photons from n=3 atomic hydrogen. Physical Review Letters 70:1240-1243 Polarized neutron diffraction experiments at low temperatures in a high magnetic field have been used by researchers at the University of Western Australia, the Royal Institution (UK) and the Institut Laue-Langevin (France) to study the anisotropy of the orbital moment of the low spin hexacyanoferrate (III) ion. The data are consistent with a cubic crystal field model in which the magnetization is dominated by the orbital moment. Day P. Delfs C D, Figgis B N, Reynolds P A & Tasset F 1993 Polarized neutron diffraction from Cs,KFe(CN)^ Molecular Physics 78:769-780 The structure of hydrogenated amorphous silicon, an important photovoltaic material, has been modelled by hypothetical silanemoleculeswith diamond or similar lattices using the semi-empirical quantum mechanical AMI method by scientists at Murdoch University. Densities of states and infrared spectra were calculated and compared with experimental data. Clare B W, Jennings P J, Cornish J C L, Talukder G, Lund C P & Hcfter G T 1993 Simulation of the electronic and vibrational structure of hydrogenated amorphous silicon using cluster models. Journal of Oimputational Chemistry 14:1423-1428 Note from the Hoii Editor: This column helps to link the various disciplines and inform others of the broad spectrum, of achievements of WA scientists (or others writing about WA). Contributions to "Recent Advances in Science in Western Australia" are welcome, and may include papers that have caught your attention or that you believe may interest other scientists in Westem Australia a nd abroad. Papers in refereed journals, or books, chapters and reviews will be accepted. Abstracts from conference proceedings will not be accepted. Please submit short (2-3 sentences) summaries of recent papers, together with a copy of the title, abstract and authors' names and addresses, to the Honorary Editor (c/o Western Australian Museum) or a member of the Publications Committee: Dr S D Hopper (Life Sciences), Dr A E Cockbain (Earth Sciences), and Assoc. Prof. G Hefter (Physical Sciences). Final choice of articles is at the discretion of the Hon Editor. "Letters to the Editor" concerning scientific issues of relevance to this journal are also published at the discretion of the Hon Editor. Please submit a word processing disk witfv letters, and suggest potential reviewers or respondents to your letter. PC Withers, Hon Editor, Journal of the Royal Society ofWA 2 Journal of the Royal Society of Western Australia, 77 (1), March, 1994 The Royal Society of Western Australia Medal Recipients 1924-1993 The Medal of the Royal Society of Western Australia was instituted in 1924 to mark the centenary of the birth of Lord Kelvin (26 June, 1824). The Royal Society Medal (originally referred to as the “Gold Medal” and then subsequently sometimes as the “Kelvin Medal” due to the association of the inaugural award with the centennial Kelvin celebration and because the medal bears in relief on its obverse side the head of Kelvin) is awarded approximately every four years for distinguished work in science connected with Western Australia. The original dye for the medal, first struck in 1924, remains in the safe-keeping of the Society. The first three medals were struck in gold, and all subsequent medals in silver. The first medallist of the Royal Society was Dr William J. Hancock, Government Electrical Engineer and Honorary Medical Radiographer at Perth Hospital, who in 1924 was the recipient of the Medal of the Royal Society of Western Australia, in recognition of his pioneering work in the medical application of X-rays. The most recent medallist. Professor John R. de Laeter, was presented with the seventeenth Medal of the Royal Society of Western Australia, in recognition of his contributions to geophysics and geochronology. Recipients of the Royal Society Medal of Western Australia are: (reference to Journal of the Royal Society notice of medal award in parentheses) 1924 1929 1933 1937 194* 1945 1949 1955 1959 1966 1970 1979 1979 1979 1983 1986 1993 Dr W J Hancock: radiography; medical application of X-rays (10:xvii) Dr E S Simpson: mineralogy and geology of Western Australia (15:iv) Mr W M Came: plant pathology; the bitter pit of apples (19:xi) Mr A Gibb Maitland: geology; Pilbara survey and artesian water supplies (23:xi) Prof E de C Clarke: geology of Western Australia (27:v) Mr L Glauert: natural sciences (31:vi) Mr C A Gardner: botany; the flora of Western Australia (35:v) Dr H W Bennetts: veterinary science; live slock diseases (40:1) Prof E J Underwood: animal nutrition and husbandry (43:67) Mr C F H Jenkins: agricultural entomology and natural history (49:91) Prof R T Pridcr: geology; petrology and mineralogy (53:95) Prof R M Bemdl: anthropology; aborginal studies (63:29) Emer Prof B J Grieve; botany; ccophysiology and the flora of WA (63:29) Dr D L Serventy: zoology; ornithology and nature conservation (63:29) Dr J S Beard: botany; vegetation classification and mapping (65:93) Prof C A Parker: soil biology Prof J R De Laeter: geophysics and geochronology (77:4) © Royal Society of Western Australia 1994 3 Journal of the Royal Society of Western Australia, 77 (1), March, 1994 The Royal Society of Western Australia Medallist, 1993. John de Laeter, AO CitWA. BSc(Hons), BEd(Hons), PhD, DSc(WA), FTS, CPhys, FInstP, Hon FAIP. Professor of Physics, Deputy Vice-Chancellor, and Dean of Graduate Studies, Curtin University of Technology. The Medallist for 1993, Professor John de Laeter, AO Cit WA, BSc(Hons), BEd(Hons), PhD, DSc(West Aust), FTS, CPhys, FInstP, Hon FAIP, Professor of Physics, Deputy Vice- Chancellor of Research and Development and Dean of Graduate Studies at Curtin University of Technology, was elected by the Council of the Royal Society of Western Australia because of his contributions to the scientific knowledge of the geophysics of Western Australia. The Medal was presented at the Inaugural Royal Society of Western Australia Medal Lecture by His Excellency the Governor of Western Australia, Major General Michael Jeffrey, AO MC, Vice-Patron of the Royal Society of Western Australia, on Monday 8 November, 1993. Professor John R. de Laeter was born at South Perth in 1933, and educated at Perth Modem School and The University of Western Australia from which he holds First Class Honours degrees in Physics and Education, and Doctorates in Philosophy and Science. He is a Fellow of the Australian Academy of Technological Sciences and Engineering, and the Australian and British Institutes of Physics. He is presently a member of the Board of CSIRO, the Higher Education Council, the Research Training and Careers Committee, and the Institutional Grants Committee of the Australian Research Council. He has held visiting research appointments at McMaster University in Canada, the Australian National University, Pennsylvania State © Royal Society of Western Australia 1994 University, Cambridge University, and the Central Bureau of Nuclear Measurements in Belgium. He is past chairman of the Commission of Atomic Weights and Isotope Abundances, and is a member of the Council of the Inorganic Chemistry Division of the International Union of Pure and Applied Chemistry. His major research interest is the application of mass spectrometry to a range of astrophysical, chemical, geological and nuclear problems. He has played a major role in the development of geochronology in Western Australia, in analysing Western Australian meteorites, in radioactive waste containment studies, and in the accurate determination of the isotopic abundances of numerous elements, many of which led to new values of their atomic weights. He has a long-standing interest in nuclear astrophysics and the origin of the chemical elements. He is also recognised as a science educator, with a particular concern for ensuring that the public is informed of the role of science and technology in society. He was one of the key persons involved in the concept of the Scitech Discovery Centre in Perth, and presently serves as Deputy Chairman of the Board. His other major concern has been to improve the relationship between Universities, Government and industry, and was instrumental in the establishment of Technology Park, adjacent to Curtin University, where he serves as Chairman of the Advisory Board. Professor de Laeter received the ANZAAS Medal in 1992. 4 Journal of the Royal Society of Western Australia^ 77 (1), March 1 994 A question of time: Royal Society Medallist's Lecture for 1993 J R De Laeter Department of Applied Physics, Curtin University of Technology, Perth, Western Australia, 6001 Manuscript received November 1993 The cheeseboard is the world, the pieces are the phenomena of the universe, the rules of the game are what we call the laws of nature. The player on the other side is hidden from us. We know that his play is always fair, just, and patient. But also we know, to our cost, that he never overlooks a mistake, or makes the smallest allowance for ignorance. T H Huxley, 1868 A Liberal Education Introduction I am honoured that the Royal Society of Western Australia has conferred on me the Royal Society Medal for 1993. As a member and Past President of the Society, 1 am pleased that my colleagues have seen fit to recognise my research, which has almost in itsentirety been carried out in Western Australia, first at the University of Western Australia, then at the Western Australian Institute of Technology and subsequently at Curtin University. I dedicate this lecture to my wife and family and research colleagues who have made this award possible. I would also like to thank His Excellency, the Governor of Western Australia, Major General Michael Jeffrey, for presenting the Royal Society medal to me on the occassion of this lecture. I have chosen "A Question of Time" as my topic for this Medallist's address, and I would like to explore with you some of the questions which all of us have asked:- How old is the Universe? How long did the Sun take to form? How old is the Solar System? How old is this rock? The concept of time has never ceased to intrigue and puzzle those who think about it. We instinctively feel that time goes on unceasingly, and that there is nothing we can do to halt its inexorable progress. However, Albert Einstein has taught us that time is not immutable; rather it is relative and depends on the observer. I wish to consider time as the order in which events occur. How then are we tomeasure geological time? For centuries various people tried to estimate the age of the Earth by heat flow (Lord Kelvin), by tidal interaction (George Darwin), by the saltiness of the oceans (Edmond Halley) and by the accumulation of sediments (Charles Walcott). However, none of these "clocks" was particularly accurate, and the physicist Lord Kelvin fell into disrepute with geologists because his estimated age of 10 million years, which he obtained by examining the cooling of the Earth from a molten body, was far too short as far as the geologists were concerned. Then in 1896 came the breakthrough that was needed to measure geological time. Henri Becquerel discovered that certain minerals were radioactive, and Marie and Pierre Curie then showed that these radioactive atoms change into other atoms at regular and constant rates. After a certain period of time, exactly half the radioactive parent atoms decay to daughter atoms - this period is the half-life of the radioactive parent. Provided one knows the value of the half-life and the proportion of parent to daughter atoms, we can calculate the period of time over which the parent has been decaying, subject to the fact that the system has been "closed" over the time interval concerned. In 1902 two scientists at McGill University in Canada, the New Zealand physicist Ernest Rutherford and the English chemist Frederick Soddy, investigated the radioactive decav of uranium and showed thatitdecayedtoa daughter product (lead) and also produced helium from the alpha particles emitted. In 1905 in the Silliman lectures at Yale, Rutherford suggested the possibility of using radioactivity as a geological timekeeper, on the basis that if one could measure either of the daughter products helium or lead, then one could measure geological time by this uranium nuclear clock (Rutherford, 1906). Unfortunately the only means available at that time to measure helium and lead were chemical techniques, and both suffered from serious shortcomings. Helium, being a gas, leaked out of uranium-rich ores, especially if they were weathered, and thus the calculated ages were very much minimum estimates. The measurement of the quantity of lead could not distinguish the daughter-product lead from primordial lead, and so calculated ages were overestimated. In the early part of this century, before the establishment of the University of Western Australia, Perth Technical College offered undergraduate degrees in science, in collaboration with the University of Adelaide. In 1904 a group of Western Australians, in an endeavour to gain support for a University, persuaded Frederick Soddy who by then was based in England, to make the long sea-voyage to Australia during his University vacation to give a series of lectures in Perth, Fremantle and Kalgoorlie on the latest scientific results. One of Soddy's lectures concerned radioactivity and the determination of the age of the Earth. © Royal Society of Western Australia 1994 5 nf the aee of the Earth was HearguedthatKelvin'ses^eo ” bsoLtely wrong ,ecay Genkin 1985). by the energy re Chemist and Assayer of the ES Simpson, who ^ ^ interest in Soddy's Geological Survey J the President of the Royal lectures. Simpson aK Society Medal in Society of WA and was ^ bnght-yellow 1929. In 1910 gn^atileat WodginainthePiIbara. uranium mineral from helium in He named it ‘ ' be 13 Ma although he pointed out it, and calculated Its ag to helium leakage that this age was " .J Comically analysed two other (Simpson 1910)- H - jte which he identified as minerals from Qrogummite. Although Simpson mackintoshite and t g minerals, he did not measured the amount oH^ad publish their ages a . the University of Sydney, Professor L A Cotto , | calculate the uranium-lead used Simpson s J (Cotton 1926). These published ages of these fiiree ^he mackintoshite, ages were 1475, . specimens respectively. Holmes thorogummiteand^ & Lawson (19^/J re ... sample was included in a popular book in tl ^^ok, and so the widespread S5'S»ra. W.6Kr6 Australia possussud the Earth’s oldest material. • 1 ,, fbink that in 1910, a few years after the H “ -%Tofth! uranium nuclear clock by Rutherford, that a discovery of the u measured the age Western Austr Uo^^ever Simpson did not maintain his of a Pilbara “^ns; and in a letter in 1927 politely "ra'SestionfromSirDougla^ geologist, that he should continue such work. Mass spectrometry and the modern era of geochronology The chemical method of determining uranium-lead ages was fraught with many errors, and little progress was made ^ rVfinim: such ages until the discovery of isotopes by J J ^lioi^onat theCaLdish Laboratory atCambridge in 1912. F W Aston, using a vastly improved mass ^ able to show that lead had at least three isotopes. In 19 9, he showed that a uranium-rich sample of brojgerite was hig y enriched in “*Pb, and calculate on age of Ma (Aston 1929). ThVs heralded a new era in geochronology, based on physical rather than chemical methods. Since that time, the mass spectrometer has become the tool of every practising aeochronologist - a veritable time machine which enables us to explore the past. An even more significant conclusion from Aston's isotopic results was drawn by Rutherford, who calculated the age of the Earth to be 3.4 x 109y (Rutherford 1929). This marked the advent of cosmochronology. In 1940, Alfred O Nier of the University of Minnesota designed a simple mass spectrometer which could measure the isotope abundances of elements with good accuracy (Nier, 1940). His sector field mass spectrometers are now common in geochronological laboratories around the World. He has rightly been called the “father of modern mass spectrometryNier was able to calculate a U"Pb age of 2570±70 Ma for a monazite (Nier et al 1941), and later pioneered the K-Ar geochronological technique (Aldrich & Nier 1948). A mass spectrometer measures the relative abundances of the isotopes of an element by separating the various masses in a transverse magnetic field (Fig 1). The sample is mounted on a filament in the ion source of the mass spectrometer. Thermal energy ionises the atoms, and the resulting positive ions are accelerated from the source into the magnet, and subsequently the dispersed ions arc collected in a detector. Varying the magnetic field enables ions of different mass to be brought to a focus in the detector, and hence the relative isotopic abundances of the element in the particular sample can be measured. Figure 1. Schematic diagram of a Nier 60 magnetic sector mass spectrometer showing the ion source, magnet and collector assembly. The Physics Department at the University of ' Western Australia j In the late 1940's two young physics lecturers at the University of Western Australia (UWA) - Peter Jeffery and Hilary Morton - were building a small nuclear accelerator. On his way to take up a foundation position at the Australian . National University (ANU), Sir Marcus Oliphant visited the * Physics Department and suggested to Jeffery and Morton | that they should build a mass spectrometer and initiate a geochronological research program. Oliphant was a student I of Rutherford and was familiar with the physical technique ^ used in geochronology. He referred to the widespread belief that Western Australia contained the Earth's oldest minerals, | and that geochronology would be a good field of research for the University to undertake. I 6 Journal of the Royal Society of Western Australia, 77 (1), March 1994 Jeffery and Mortonbuilt a massspectrometeroutof copper tubing and other odds and ends, but by 1953 had decided to abandon the project unless financial support could be found. Morton in fact left Perth to go to ANU, but Jeffery persevered with a grant from the Carnegie Geophysical Institute in Washington, DC. Jeffery (1976) describes the effect of these funds in the following terms: "The Carnegie funds provided the Perth group with a new sixty degree Nicr-lype mass spectrometer and also permitted the original home-made spectrometer to be upgraded by the replacemcntofitswaterpipesections with fabricated stainless steel. Such exotic devices as a chart recorder to replace a wall galvanometer, and commercial diffusion pumps were all very acceptable. In spite of these improvements in equipment however, ion currents were still being measured using 'acorn' 954s as electrometers" I joined the mass spectrometry group at UWA in 1954 as an honours student in physics, and worked with Peter Jeffery and a PhD student - Bill Compston - on carbon isotopes using the "copper tube" mass spectrometer. I was fortunate to be a co-author of the first paper ever published by the mass spectrometry group which described this work Qeffery et al. 1955). The second machine was commissioned by another PhD student, David Greenhalgh in 1955, and the first U-Pb age date was published in 1959 (Greenhalgh & Jeffery 1959). Peter Jeffery spent 1955 at the Carnegie Institute in Washington DC, and on his return commenced a program of Rb-Sr dating. Bill Compston, who had subsequently completed a Postdoctoral Fellowship at the California Institute of Technology (Cal Tech), became a lecturer in Physics at UWA, and they were joined by a PhD student (Glen Riley) to develop the Rb-Sr technique - which is based on the decay of the radioactive parent ^'Rb to the stable daughter ®^Sr. The half life of ^T^b is 4.88 x 10^^, and this chronometer is therefore ideally .suited for old rocks. Towards the end of the 1950's, the validity of the Rb-Sr method was being questioned since mineral separates from whole rock samples gave different results. A granite from the Boya quarry was analysed to give ages of 2430 Ma for the whole rock specimen, 650 Ma for the biotite, and 1290 Ma for the microcline extracted from the sample. Compston & Jeffery (1959) argued that the mineral separates had lost a proportion of their radiogenic 87Sr some time after crystallisation, presumably through a mctamorphic event, but that the 87Sr was not lost to the whole rock and was simply redistributed within it. Dr Alan Wilson from the Geology Department at UWA joined forces with Jeffery, Compston, Greenhalgh and Riley to exploit these geochronological techniques in the southern Yilgarn Block and the Albany Fraser Province. A paper by the group in 1960 reported 36 Rb-Sr, 16 K-Ar and 21 U-Pb ages (Wilson et al 1960). The collaboration of physicists and geologists represented an important advance, in that the combined talents of field geologists and laboratory physicists were available to tackle this new field of scientific research which was of enormous potential value to mineral exploration in this State. After the success of the 1950s, one might have assumed that theUni versity of Western Australia would have become oneoftheworld's leading centres for geochronology, situated as it was in a mineral-rich State with extensive Precambrian terrains. This was not the case. By 1961, the geochronological program at UWA was non-existent. Bill Compston took up a position at ANU, Glen Riley went to the Australian Institute of Nuclear Science and Engineering, Alan Wilson moved to Queensland, David Greenhalgh became a science teacher, and Peter Jeffery developed new interests in nuclear astrophysics in cooperation withjohn Reynolds at the Physics Department in the University of California at Berkeley. When Dr Jeffery returned to Perth in 1962,1 became his first PhD student in nuclear astrophysics. Geochronology at the Western Australian Institute of Technology On my appointment as Inaugural Head of the Department of Applied Physics at the Western Australian Institute of Technology (WAIT) in 1968, after a Post Doctoral fellowship at McMaster University in Canada, a research program to search for isotopic anomalies in meteorites was establi.shed using a 30cm radius of curvature solid source mass spectrometer. Inlatel968,Dr AlecTrendallof theGeological Survey of Western Australia (GSWA) and Bill Compston argued that some mass spectrometer time should be devoted to geochronology, because of the necessity to place time constraints on the rock units being mapped by geologists at the GSWA, and the inability of Compston's ANU laboratory to handle the immense amount of work that needed to be done. The Rb-Sr technique was chosen because it was relatively simple and ideally suited to the old Archaean rocks which constituted a major portion of the State, and which was the focus of much of the geological mapping at that time. The first geochronological paper from the WAIT laboratory was published in 1970 (De Laeter & Trendall 1970). In the early 1970s we decided to develop a new geochronological technique based on the decay of ’^^Lu to with a half life of 3.54 x 10^^. Although the mass spectrometry of hafnium presents some difficulties, the Lu- lif chronometer has some characteristics which made it an attractive research project. However, our interests were diverted from geochronology to cosmochronology and thus the Lu-Hf geochronometer was not developed at that time (McCulloch et al 1976). Malcolm McCulloch later was instrumental in developing the Sm-Nd geochronological technique at Cal Tech (McCulloch & Wasserburg 1978a), before taking an appointment at ANU. The Sm-Nd technique is based on the decay of *^^Sm to *‘*-^Nd with a half life of 10.6 x lO^'y. As parent and daughter are both rare earth elements, natural fractionation processes do not favour a separation as occurs for many pairs of elements used in geochronology. This technique was introduced at WAIT in 1979 (Fletcher & Rosman 1980), and has been used successfully to solve a number of geochronological problems. By the mid-1970s, geochronology was a major component of the work of the mass spectrometer laboratory at WAIT. In 1974 geologists from the University of Western Australia became associated with the mass spectrometry laboratory, and an extensive survey of Yilgarn ages was carried out using the Rb-Sr technique. With the appointment of Dr M T Bickle to the Geology Department of UWA in 1979, the tempo of the geochronology program accelerated, and a 7 of the Royal Society of Western Australia, 77 (1), March 1994 (Bickleeffl/.1983)- ^ the U-Pb and Pb-Pb was established at U ^-^^^^j^^^^^^^theexistingRb- geochronologicaltecni 4 Sr and Sm-Nd techniques. f Dr R T Pidgeon at WAIT in lysi The appointment or ^ geochronology as yet heralded the introdujo ^j^hnique available in Western another geochronolo^c ^ ^^^^g^pt,nding increase in the Australia, and there Uronology group which now scientific output *®| hysicists from GSWA,UWA and comprised geologists an p y^^^^^^^ealth Scientific and WAIT. Scientists from industryhave Industrial Research O g research projects. A also been “'‘^^^Lonologists at ANU has also been a close liaison with 8^°^ recently Dr Neal McNaughton feature of ^ and he and his students have been was appointed o ^WA Together with Ian actively assoaated w h h^^^^ McNaughton has Sped the K-S^eochronological technique. .979 .be Roy»> ^»“o.Sr9«rwt.eTAt"a,i' reviews describing asp , ^he 150"’ Anniversary of from 1829 to 1979 on ^ One of these concerned the founding Figure2shows from 1910 to 1977 taken from that pape . For the Yilgarn Block - the time of major generation of gneisses and granitoids was remarkably consistent over almost the entire sampled area of the block, and peaked at about 2700 my; - there is little evidence for any wide spread about this peak, and evidence for much older ages is still not compelling; - over a large area of the Block, the granitoids had a relatively short prior-crustal history, and by implication were generated over a relatively brief period; - the time interval between the formation of the greenstone belts and the peak of granitoid emplacement is of unknown length, but the available evidence suggests that it was small. For the Pilbara Block - the main period of gneiss and granitoid generation was also uniform over the area of the Block, and was earlier than that in the Yilgarn Block; - the error limits attached to granitoid ages make it uncertain whether emplacement was restricted within a narrow time range; - the highly fractionated small-volume granitoids ("tin granites") are significantly younger; - greenstone belt ages are about 300-500 Ma older than the peak granitoid age. F’ re 2 Chronological summary up to 1977 of all publications including first reports of ages of Western Australian Precamb'rian rocks and minerals by methods based on radioactive decay. Selected second reports are also included where these augment or upgrade the first. In 1981, on the occasion of the 2"*^ Archaean Conference in Perth, a review of the geochronological data on the two major Archaean terrains of Westem Australia was published (De Laeter et al 1981). At that time most of the dates were from Rb-Sr analyses (837 whole-rock and 142 mineral Rb-Sr ages), 137 mineral Pb isotope analyses, and 15 K-Ar analyses, together with some U/Pb data. The judgement was made that, despite the large database, disappointingly few firm conclusions could be enunciated. Nevertheless, the following conclusions were drawn from these data (De Laeter et ul. 1981): Zircons are Forever One of the most intriguing episodes in Western Australian geochronology has been the search for the oldest rocks. In 1981, Denis Gee and his colleagues at the GSWA argued that a predominantly gneissic terrain, which forms an area around thewe„slemmarginoftheYilgam Block, wasofgreatantiquity {GeeetaL 1981). A study of banded gneisses near MtNarryer in the northern part of the Western Gneiss Terrain gave a Rb- Sr whole rock isochron age of 3348±43 Ma with an initial ®^Sr / ®^r ratio of 0.7037 (De Laeter et al. 1981). A prior crustal 8 Journal of the Royal Society of Western Australia, 77 (1), March 1994 history of approximately 200 Ma could be inferred, and this was supported by Sm-Nd model ages of 3620 Ma to 3710 Ma from samples of this Meeberrie gneiss (Fig 3). Further work gave a Pb-Pb isochron of3357±70 Ma, whereas Rb-Sr ages for granites intruding the gneisses gave the classic Yilgarn age of approximately 2600 Ma (De Lacter ei al 1985). The Sm-Nd ages were interpreted as the time of extraction of the protoliths of the gneisses from a chrondritic source, whilst the Rb-Sr age is thought to represent a time of intracrustal reworking Figure 3. Simplified geological map of the Mt Narryer region. Western Australia. It was about this time that the Sensitive High Resolution IonMicroProbemassspectrometer(SHRlMP)in the Research School of Earth Sciences at ANU became available for zircon geochronology. Designed by Steve Clement, SHRIMP was the''brainchild" of Bill Compston. After years of development it had already produced some exciting results using the zircon method. However, there was a certain amount of scepticism in various parts of the world as to its real capabilities. Zircons from one of the Meeberrie gneiss samples gave a U-Pb age of 3300 Ma for the rims of the zircons whilst the interior portions gave ages between 3560-3690 Ma (Kinny et al. 1988). These older ages have been interpreted as minimum estimates for the original magmatic ages of the xenocrystic cores. These data were in excellent agreement with the conventional geochronology carried out at WAIT, and were convincing evidence of the power of ion probe mass spectrometry. U-Pb studies by the ion microprobe on detrital zircons from quartzite adjacent to the banded gneiss showed that most of them formed between 3500 and 3750 Ma, although some of them gave ages of about 3300 Ma (Froude et al. 1983). These ages suggest that the zircons may have been derived by erosion of the adjacent gneisses or their protoliths. In addition Froude et al. (1983) reported the existence of four zircons from the same quartzite which have nearly concordant U-Pb ages between 4100 and 4200 Ma. These results suggest that pre-3800 Ma silica-saturated rocks were present in the Earth's crust. It is possible that intact remnants of these rocks may have survived in this region. The Jack Hills metasedimentary Belt is a narrow curvilinear east to north-east trending belt approximately 60 km north¬ east of Mount Narryer. It is composed of minor metabasalts and substantial thicknesses of chert and banded iron formation interleaved with pelitic and psammitic metasediments (Compston&Pidgeon 1986). Detrital zircons from the Jack Hills metasedimentary bell analysed at ANU using the ion microprobe mass spectrometer, has revealed the oldest ages so far determined. One zircon grain registers an age of 4276±6 Ma. which is a minimum estimate for its original age (Compston & Pidgeon 1986). Sixteen other grains have the same or slightly younger age, similar to the zircon ages measured at Mount Narryer. The frequency of occurrence of the old zircons is 12±5%. Kober et al. (1989) report the analyses of thirty zircon crystals from the Jack Hills metaconglomerate using the single zircon, direct evaporation, thermal ionization technique. Four of the thirty zircons gave ages in excess of 4000 Ma, confirming the microprobe analyses of Compston & Pidgeon (1986). Approximately 50% of the analysed zircons yielded an age of 3380±20 Ma, whilst other crystals gave ages of 3300 Ma, 3440 Ma and 3570 Ma. As had been observed at Mount Narryer, some of the zircons demonstrated a more complex age structure with intergrowth of the old phases with younger domains. Figure 4. Time-space plots of (a) the Perth and (b) Harvey traverses. Localities have been projected parallel to the western edge of the transition zone, the traverse line being normal to the transition zone. 15517—2 9 j urnal of the Royal Society of Western Australia, 77 (1), March 1994 .crps in excess of 4000 Ma have Although parentrockswith ^. that Western not been found, long material has been Australia possesses the Eart vindicated. hronological publication from the The. most recent [t Curtin was in the last issue mass spectrometry YSciences (De Laeter & of the Australian ^ ° hie of Rb-Sr biotite ages from Libby 1993). It from Harvey to Kuhn (Fig 4). The Perth to Kellerbernn and I Rb-Sr system forbiotites between rocks exceeds most area, then climb quickly 430-500 Ma in the west roughly constant at through a transition zone ^^^j^^^^p^^^jj^^ 3 ,,^presenting 2300Mato theeast JV a period of uplift m the eany Cosmochronology One of the ”^"SentShat"lave occurred in the 3 to place a time seal System and on the ormation and elen^ Although this is a lucleosynthesis of the search by the occurrence Hunting task, we are ^hh a wide array of half- ,,.,„p„„n„bero Ives, *« P™"* ^ To dblain accurals daling, a decay ichieve this ^ j ^ith a half-life that is of the same .ystem should be se ^ measured. Thus, long- d' TdfoaSve dX schemes such as U-. Th-Pb, Sm-Nd, Ibl t Ar Re-S) U-Hf and K-C, have been used lo Ib-Sr, K Ar, Ke , occurred early in the histor>' Oh "sollr sStem whereas short-lived radionuclides such ; ^Be %, ^H,' and the U-series disequilibrium system "ave been ^sed to study more recent events. Geochronology utilizes isotopic dating techniques to Tieasme he age of terrestrial materials. The radioachvity of measure ti g ^les them to be used as isotopic edrgSl « differcnliallon -rustal re ationships. Isotope geoiogy ima Sh degree of sophistication, which has only been made poliblebythedevelopmentofsensitivemassspect^^^^^^^^ [hat are capable of routine measurements of high precision and accuracy. Conventional gas source and solid source mass spectrometers have been successfully applied to the long-lived chronometers, but the short-lived chronometers have required instrumentation with the best possible sensitiviw. Ion probe and accelerator mass spectrometers arehvosuchrecentdevelopments. Unfortunately the dating of terrestrial materials does not provide a great deal of informaHon on the time of formation of the Solar System since the Earth is an active planet where a variety of geological processes have combined to destroy the early record of its There are three major epochs in the history of the cosmos. The first epoch T extends from the "Big Bang" to that time when the solar nebula was isolated from galactic nucleosynthesis. The second epoch A is of relatively short duration and represents the time interval for the collapse of a gas cloud leading to the formation of the Solar System. The third epoch represents that period from the formation of the Solar System to the present (Figure 5). A Figure 5. A schematic diagram of theliistory of theUniverse. The first epoch T represents the time period from the “Big Bang" to the isolation of the solar nebula from galactic nucleosynthesis. The second epoch A is the time interval between the termination of nucleosynthesis and the solidification of Solar System bodies, while the third epoch represents the age of the Solar System. The shape of the production rate for nucleosynthesis, pfr), is for illustrative purposes only. The Age of the Solar System Meteorites and lunar samples provide us with alternative avenues to determine the age of the Solar System T . Meteorites are the debris of small planetary bodies that solidified early in the history of the Solar System and have been closed isotopic systems retaining evidence of early Solar System processes. Neither have they been subjected to the long planetary accretion processes associated with larger bodies. The first isotopic determination of the age of the Solar System was carried out in 1953 on primordial meteoritic lead (Patterson 1956), resulting in an age of formation for planetary bodies of -4.5 x 10^ years. Exacting chronological studies of a variety of meteoritic materials show that a fine- scale separation of events took place early in the Solar System during planetary formation, and these events may be resolved down to a few million years (Wasserburg 1987). Numerous chronological investigations of lunar materials were carried out on samples obtained during the various missions to the Moon, from which an excellent chronology has been determined (Wasserburg et al. 1977). Figure 6 shows a representation of lunar chronology. Crystallisation ages of approximately 4.5 x 10** years have been obtained from lunar rocks, so that there is a good correspondence between the meteoritic and lunar data for the time of planetary formation. The melt rocks derived from impact metamorphism give crystallisation ages of (3.85-4.05) x 10^ years, and there is good agreement between the U, Th-Pb, Rb-Sr and K-Ar data. This is interpreted as the result of a major bombardment of the moon by meteorites which created the lunar basins. This has been called the "terminal lunar cataclysm", and it is possible that this bombardment a ffected the Solar System as a whole (Tera et al. 1974). After the termination of the bombardment phase some 3.85 x 10** years ago, there was continued but decreasing volcanic activity until approximately 3.0 x 10^ y ago, after which the lunar surface appears quiescent. There is no evidence of recent igneous activity on the moon. 1 10 Journal of the Royal Society of Western Australia^ 77 (1)^ March 1994 Figure 6. A schematic diagram showing the chronology of major lunar events. My involvement with the chronology of lunar samples was quite peripheral. In 1971 a fire in a chemical preparation room adjacent to the mass spectrometer laboratory at WAIT badly damaged the mass spectrometer which had been installed in 1968. Although the cost of replacement was covered by insurance, there was a time lag before a new machine could be installed. Fortuitously, Bill Compston was taking up a six month appointment at the Lunar Science Insti tute in Houston, Texas, and he asked me to supervise his laboratory at ANU during his absence. It was the time of the Apollo missions, and geochronologists were trying to measure the ages of the returned lunar samples. There was intense competition to determine the ages by a variety of techniques, and national as well as personal prestige was at stake. Bill Compston's ANU laboratory had reported Apollo 14 Rb-Sr ages which were systematically greater by a few percent than other reported measurements. Compston was to present a paper at the Lunar Conference which would attempt to justify the ANU results. However, there was a possibility that the elemental Rb /Sr ratios measured at ANU on the lunar samples were wrong, because of systematic errors in the calibrations of the Rb and Sr isotopic spikes. The US National Bureau of Standards had just produced some certified stoichiometric Rb Cl (SRM 984) and Sr CO^ (SRM 987) standards and it was decided to recalibrate the ANU spike solutions against these new standards. Working around the clock, we managed to complete the work just before the Lunar Science lecture where Bill Compston was able to announce a 1.8% error in the calibrations which necessitated a decrease in all the previous ANU ages for the Apollo samples. The revised ages were then in good agreement with Wasserburg's group at Cal Tech (De Laeter et at. 1973). I also worked on the Apollo 15 lunar samples at ANU before returning to Perth in mid 1972 (Compston et al. 1972). Extinct Radionuclides A number of unsuccessful mass spectrometric attempts were made to demonstrate the existence of short-lived radionuclides early in the evolution of the Solar System before John Reynolds showed that a large enrichment of *^’Xe existed in Xe extracted from the Richard ton meteorite (Reynolds 1960). In 1961 Peter Jeffery and Reynolds proved that the excess ^^^Xe was correlated with iodine in the meteorite, thus proving that the decay of the radionuclide ^^^I (whose half-life is 17 x 10*), had taken place within the meteorite itself (Jeffery & Reynolds 1961). It was possible to calculate a time A~10®years between the synthesis of and the formation of the Solar System. The value of A-IO® years created somewhat of a dilemma, because astrophysical models predicted a much shorter time interval for the formation of the Solar System. One reason behind the search for evidence of extinct radionuclides in meteorites was to identify the heat source that melted some of the meteorite parent bodies. One of the most logical candidates was “Al, which decays to “Mg with a half-life of 0.72 x 10* years. I was associated with the first mass spectrometric study to identify the presence of this potential heat source by the measurement of the isotopic composition of Mg in meteoritic feldspars whilst I was at McMaster University in 1967 (Clarke et al. 1970). Although this initial attempt was unsuccessful, a study of calcium- aluminium rich inclusions from the Allende meteorite by Chris Gray and Bill Compston gave an excess in “Mg, which was shown to be correlated with Al, in an isochron-type array (Gray & Compston 1974). Figure 7 shows an “Al / “Mg isochron from Lee et al. (1977). A plausible interpretation is that “Al was present in the meteorite and that the decay was in-situ. If it is assumed that the minerals were isotopically homogeneous at the time of solidification, a value of A of - 10*years can be calculated. This is significantly lower than the value of -10® years derived from systematics, and much closer to astrophysical estimates. Figure 7. An "isochron" diagram of “Mg/“Mg versus ^^Al / “Mg for minerals from an Allende inclusion. The linear dependence of the magnitude of the anomalous “Mg on the Al/Mg ratio can be interpreted as resulting from the decay of the radioactive nuclide “Al. Another extinct radioactive system is based on the decay of ^°Td (with a half-life of 6.5 x 10* a) to ’°^Ag. Kelly & Wasserburg (1978) proved the existence of excess ^^Ag in iron meteorites with *°^Ag excesses of up to 20%. The excess ^°^Ag correlates with Pd to form an "isochron" with a calculated A of - 10^ years between the last injection of nucleosynthetic material and the melting and differentiation of small planetary bodies. A ^^Ba enrichment, which was observed in an Allende inclusion by McCulloch & Wasserbur g (1978b), is due to the decay of ^“Cs (which has a half-life of 3.3x10*). 11 Journal of the Royal Society of Western Australia, 77 (1), March 1994 f exrinct radionuclides may be present in A number of oth Bosnian and I investigated the ^eteoritic matena ^ (with a half lifeof 10*) in iron meteorites, decayofto excess ’“Te. This provides a but could g ^ pg Laeter & Rosman 1984). lower limit Of lu y u . Hv of extinct radionuclides therefore provides the The study Of ex material, containing a full following thetic products (including some range « uudeo- y^^ condensed to form a radionuclides tha estimated molecular^clou ^ chronometer). Rapidly evolving stars a "list minute" injection of fresh nucleosynthetic provided a las condensing cloud r"‘“whS the Solar System was formed within a time pTrkid of a few million years. Nucleocosmochronology is the use of the relative Nucleocosmoc determine the time abundances o r nuclides. Schramm & scales for "^ ^^"Yshowed that a nucleosynthetic formalism Wasserburg ( , ^g-Iived radionuclides, and that this canbedeve oped for ^ elements, independent of the can give production model adopted. This mean age iTl'owSS to" the period T over which nucleosynthesis has taken place. The estimate of the duration of nucleosynthesis is a challenging scientiHc problem. If we can calculate this period T, then we can estimate the age of the galaxy T by Lding T to the age of the Solar System and the formation age of the Solar System A, such that Tc=r + A +T^ This estimate is then a lower limit for the age of the universe. I have already mentioned the work on the '^^Lu/’^^Hf cosmochronometer by Malcolm McCulloch, Kevin Rosman and myself in the mid-1970s, (McCulloch et at. 1976). 1 was anxious to calibrate this chronometer because I thought it offered the best chance of measuring the duration of nucleosynthesis and hence the age of the Universe. Although the uoLu/‘^‘Hf ratio was known from mass soectrometric determinations in our laboratory, and the neutron capture values and half life of '^‘Lu had been measured, the nuclear systemaHcs of ’^'Lu were not well understood in the mid 1970's. When '^‘Lu is formed by the capture of '”Lu by a neutron, it can exist in two isomeric states - the ground state which decays to '^"Hf with a half life of3 57 X lO'^’a, and an excited state which has a half life of only 3.68 hours (Figure 8). In the mid 1970's it was believed that the two states were completely independent and thus all that was required was to measure the branching ratio B, that is find out what fraction of^^Lu existed in the ground state. 1 persuaded some nuclear physicists at Lucas Heights Atomic Energy Establishment in NSW to help me to measure the branching ratio in their nuclear accelerator, which could simulate the Figure 8. Decay scheme for neutron capture on ^^^Lu, showing the two possible radioactive decay modes of ’^^Lu to conditions that existed in red giant stars. We obtained a value of B = 21% (Allen et ai 1981). 1 remember, as if it was yesterday, making the calculations to estimate the age of nucleosynthesis from the data, and tlie shock of finding a negative age. This implied that the Big Bang had not yet occurred! There was obviously a small problem somewhere. So for the next 12 months or so, we laboriously repeated the branching ratio experiment, necessitating several trips to Lucas Heights. However, the value remained essentially unchanged. Furthermore our value was in conflict with a value of B from a much more prestigious group in Karlsruhe in Germ any (Beer 1984), so thatwe were understandably nervous about our conclusion, which was that the ^^^’Hf chronometer couldn't keep time. The correct answer finally emerged when the energy levels of ’^*’Lu were remeasured. It turned out that the energy separation between the ground state and excited state of *^^Lu was much smaller than first reported in the literature, and at the temperatures which existed in red giant stars (lO'^ K), the two states could overlap and hence the '^'*Lu ground state could leak away via the ’^^Lu excited state to ^^*’Hf. This problem is somewhat akin to the U-He leakage method. '^*Lu was a cosmothermometer not a cosmochronometer (De Laeter et at. 1988). Sufficient to say that this was a terribly disappointing result. 1 had spent almost ten years of research investigating the age of the Universe only to find that Nature had managed to frustrate us. The one pleasing fact that has emerged is that our conclusions have now been accepted by the international scientific community, including the Karlsruhe group (Lesko et ai 1991). An independent estimate of the age of the Universe, calculated from the Hubble recession of the Galaxies, is approximately 19 x 10*^ (Sandage & Tamman 1982). Other astrophysical evidence on globular dusters, and estimates of the age of nucleosynthesis by the U-Pb (Thielemann et ai 1983)and Re-Os (Yokoietai 1983)cosmochronometers, give tentative support to this value, but it must be emphasised that the age is an approximate value with large uncertainties. Thus cosmochronological studies have provided a reasonable time scale for the early history of theSolar System, including the age of formation of the meteorites. Some success has also been achieved in deriving a mean age of galactic nucleosynthesis, but the age of the galaxy has not yet been accurately determined from nucleocosmochronological studies due to the constraints associated with each of the long-lived chronometers used to determine the duration of nucleosynthesis. 12 Journal of the Royal Society of Western Australia, 77 (1), March 1994 Conclusions I have endeavoured to give you a cursory glimpse of our endeavours to delve into that most intriguing Question of Time. As Arthur Holmes' has said: 'Tt is perhaps a little indelicate to ask our Mother Earth her age, but Science acknowledges no shame, and from time to time has boldly attempted to wrest from her a secret which is proverbially well guarded ". My journey in time commenced in 1968 when AlecTrendall and Bill Compston came to visit me in theMass Spectrometry Laboratory at WAIT 25 years ago, almost to the day, and persuaded me to assist in that indelicate question of asking Mother Earth her age. Looking back down that 25 year time period (which is something I probably wouldn't have done if it was not for this Lecture) my over-riding impression is how much still remains to be achieved. Although we have developed beautiful "high-tech" ma.ss spectrometers and many measurements have been made, some of the objectives we set out to achieve haven't been accomplished. But I look to the future with confidence. SHRIMP is a mass spectrometer which is ideally suited to unravelling the problems of mineral exploration, and we have an excellent team of physicists and geologists from UWA, GSWA, and Curtin together with overseas collaborators, who can exploit its potential to the full. My most enduring memories of my chronological odyssey have been the people 1 have worked with, the friendships made, the teamwork forged, the localities visited. Maybe one day we will learn all the answers, but in the meantime perhaps we should listen to Tolly Cobbold's injunction "Time, Gentlemen, Time", and leave some of the questions to another time and perhaps to another Royal Society Medallist to answer. Acknowledgements: The outstanding contribution of Dr Peter Jeffery in the establishment of geochronology in Australia, and the work of Prof Bill Compston who has recently received an FRS for his development of SHRIMP and other initiatives, together with the long .standing interest of Drs Alec Trendall, Will Libby, Ian Fletcher and others, needs to be recognised. I wou Id like to acknowledge my family, research colleagues and .students for their contribution and support to the work described in this review. I would also wish to thank Curtin University of Technology and the Australian Research Council for supporting the mass spectrometry laboratory over a period of twenty five years. Ms Tiffanie Carr prepared this manuscript with her usual skill and efficiency. References Aldrich L T & Nier A O 1948 Argon 40 in potassium minerals. Physical Review 74: 876-877. Allen B J, Lowcnlhal G C & De Lacter J R 1981 S-process branch at *^'’Lu. Journal of Physics G: Nuclear Physics 7:1271-1284. Aston F W1929 The mass spectrum of uranium lead and the atomic weight of proactinium. Nature 123:313. Beer H, Walter G, Macklin R L & Patchett P J 1984 Neutron capture cross sections and solar abundances of '*^'*‘Dy,‘'"Yb, and i76,i77j-jf s-process analysis of the radionuclide ’’'’Lu. Physical Review C. 30; 464-478. Bickle M J, Bettenay L F, Barley M E, Chapman H J, Groves D I, Campbell IH &De LaelerJ R1983 A 3500 Ma Plutonic and Volcanic Calc-Alkaline Province in the Archaean East Pilbara Block. Contributions of Mineralogy and Petrology 84: 25-35. Chapman H J, Bickle M J, De Laeter J R, Bettenay L F, Groves D I, Andersen L S, Binns R A & Gorton M1981 Rb-Sr geochronology of granitic rocks from tlie Dicmals area, Central Yilgarn Block. Special Publication No. 7 Geological Society of Australia, 173-186. Clarke W B, De Laeter J R, Schwarez H P & Shane K J 1970 ^A1 /^*Mg dating of feldspar in meteorites. Journal of Geophysical Research 75:448- 462. Compston W, De Laeter J R & Vernon M j 1972 Strontium isotope geochemistry of Apollo 15 basalts. In: The Apollo ISLunarSamples. (Chamberlain, J.W. & Watkins) The Lunar Science Institute, 347-351. Compston W & Jeffery P M1959 Anomalous "common strontium" in granite. Nature, 184:1792-1793. Compston W & Pidgeon R T1986Jack Hills, evidence of more very old detrital zircon.s in Western Aastralia. Nature 321:766-769. Cotton L A1926 Age of certain radium-bearing rocks in Australia. American Journal of Sciences 12:42-45. De Laeter] R, Allen BJ, Lowenthal G C & Boldeman J W1988 Constraints on the “’’'’Lu cosmochronometer. Journal of Astrophysics and Astronomy 9: 7-15. De Laeter J R, Fletcher IR, Rosman K J R, Williams I R, Gee, R D & Libby W G 1981 Early Archaean gneisses from the Yilgarn Block, Western Australia. Nature 292; 322-324. De Laeter J R, Fletcher IR, Bickle M J, Myers J S, Libby W G & Williams IR1985 Rb-Sr, Sm-Nd, Pb-Pb geochronology of ancient gneisses from Mt Narryer, Western Australia. Australian Journal of Earth Sciences 32: 349-358. De Laeter J R & Libby W G 1993 Early Palaeozoic biotite Rb-Sr dates in the Yilgarn Block, near Harvey, Western Australia. Australian Journal of Earth Sciences 40; 445-453. De Laeter J R, Libby W G & Trendall A F 1981 The Older Precambrian Geochronology of Western Australia. Special Publication of the Geological Society of Australia 7; 145-157. De Laeter J R & Rosman K J 1984 A possible ‘^n chronometer for the early solar system. Meteoritics 19: 217. De Laeter J R & Trendall A F 1970 The age of the Copper Hills porphyry. Western Australia Geological Survey Annual Report 1969: 54-59. De Laeter J R & Trendall A F 1979 The contribution of geochronology to precambrian studies in Western Australia. Journal of the Royal Society of Western Australia 62:21-31. De Laeter J R, Vernon M J & Compston W1973 Revision of lunar Rb-Sr ages. Geochimica et Cosmochimica Acta 27: 700-702. Fletcher IF & Rosman K J R1980 The effect of trace gases on neodymium ion formation. International Journal of Mass Spectrometry and Ion Physics 36:253-257. Froude D O, Ireland T R, Kinny P D, Williams IS, Compston W, Williams IR. & Myers J S1983 Ion Microprobe identification of 4100 to 4200 Ma- old terrestrial zircons. Nature 304; 616-618. Gee R D, Baxter J L, WildeS A & Williams IR1981 Crustal development in the Archaean Yilgarn Block, Western Australia. Geological Society of Australia, Special Publication 7:43-56. Gray C&CompstonW 1974 Excess“^Mgin theAllendc meteorite. Nature251: 495-497. Greenhalgh D & Jeffery P M1959 A contribution to the Precambrian chronology of Australia. Geochimica et Cosmochimica Acta 16:39-57. Holmes A & Lawson R W 1927 Factors involved in the calculation of the ages of radioactive minerals. American Journal of Science 13: 327-344. Holmes A 1927 The age of the Earth. Bonn's Six penny Library No 102, London, Ernest Benn Ltd. Jeffery P M 1976 Stable Isotope Abundance Studies in Western Australia. Australian Physicist 13: 26-28. Jeffery P M, Compston W, Greenhalgh D & De Lacter J R1955 On the Carbon- 13 Abundance of Limestones and Coals. Geochimica et Cosmochimica Acta 7:255-286. Jeffery P M & Reynolds J H1961. Origin of the excess '^’Xe in stone meteorites. Journal of Geophysical Research 66; 3582-3584. Jenkin J G 1985 Frederick Soddy's 1904 Visit to Australia and the Subsequent Soddy-Bragg Correspondence: Isolation from Without and Within. Historical Records of Au.stralian Science 6:153-169. Kelly W R & Wasserburg G J1978 Evidence for the existence of "^Pd in the early solar system. Geophysical Research Letters 5:1079-1082. 13 Journal of the Royal Society of Western Australia, 77 (1), March 1994 ^ HP D O Ireland T R & Compston W1988 Early p D WilUams IS, Fro^e ^ ^ ( 55^5 ana anorthosites at Mount Archaeanz^n«S;^^^X Research 38:328-341. Narryer, Single-zircon dating by stepwise Pb FidgeortRT&LippoH ) ^ of detrital zircons evapo^^^^C^t W^ern A^tralia. Earth and Planetary Science from The Letters 91. vVasserburg G J1977 Aluminium-26 in the early LeeT,Papan«^=^;°tt^^^^^^^^^ , -^^Pr R M Sur B & Beausang C B 1991 ‘«Lu : An Lesko K T, ,' 0 ^" chronometer. Physical Review C 44:2850-2864. unrelia es- nKJR1976Theisotopiccomposition McCulloch M T, De ^ of lutetium in meteorites and terrestrial and elemental abundance Earth and Planetary r"’„^c«Le«ert28:308-322. Av.re G I 1978a Sm-Nd and Rb-Sr chronology of P I 497 gb Barium and neodymium isotopic McCulloch M T & ^““^^“^^demeteorite. AstrophysicalJournal220:L15- anomalies in in T19 ■ meter for routine isotope abundance ^ ° 'ture^erRevOscienHficInstrumentsll: 2 ^ measurem hey B F 1941 The isotopic constitution of Nier A measurement of geological time. Physical Review 66 : 112-116. ^ r 1956 Age of meteorites and the earth. Geochimica et Letters 4; 8-10. Rutherford E 1906 Radioactive Transformations. Yale University Press, New Haven. Rutherford E1929 Origin of actinium and the age of the earth. Nature 123* 313 - 314. Sandagc A & Tammann G A1982 Steps towards the Hubble constant with the global value. Astrophysical Journal 256:339-345. Schramm D N & VVasserburg G J 1970 Nucleochronologies and the mean ago of the elements. Astrophysical journal 162:57-69. Simpson E S 1910 Pilbarite, a new mineral from the Pilbara Goldfields : Verbatim report of paper read to National History and Science Society of Western Australia. We.st Australian, 17 August 1910. Tera F, Papanastassiou D A & Wasserburg G J 1974 Isotopic evidence for a terminal lunar cataclysm. Earth and Planetary Sciences Letters 22-1- 21 . Thielemann F K, Metzinger J & Klapdor H V 1983 Bela delayed fission and neutron emission: Consequences for the astrophysical r-process and the age of the galaxy. Zeilschrift fur Physik A 309; 301-317. Wasserburg G J 1987 Isotopic abundances: Inferences on solar system and planetary evolution. Earth and Planetary Sciences Letters 86-129- 173. Wasserburg G J, Papanastassiou D A, Tera F & Huneke ] C 1977 The accumulation and bulk composition of the Moon. Philosophical Transactions of the Royal Society, London A 285:7-22. Wilson A F, Compston W, Jeffery P M & Riley G H1960 Radioactive Ages from the Precambrian rocks in Australia. Journal of the Geological Society of Australia 6:179-196. Yokoi K, Takahashi K & Arnould M 1983 The ‘"^Re-"*’’C>s chronology and chemical evolution of the galaxy. Astronomy and Astrophysics 117* 65-82. 14 Journal of the Royal Society of Western Australia, 77: 15-22, 1994 The Impact of Prolonged Flooding on the Vegetation of Coomalbidgup Swamp, Western Australia. R H Froend^ & P G van der MoezeP ^Surface Water Branch, Water Authority of Western Australia, PO Box 100 Leederville, WA 6007 ^Alan Tingay and Associates, 35 Labouchere Rd, South Perth, WA 6151. Manuscript received ]uly 1993; accepted February 1994 Abstract A preliminary study of the impacts of prolonged flooding on the vegetation of Coomalbidgup Swamp, an ephemeral wetland near Esperance, Western Australia, identified changes in composition and physiognomy of dryland and wetland vegetation. A high mortality was observed after flooding for ail species typically part of the surrounding dryland flora. Some stands of wetland and dryland vegetation have been partially inundated for up to 6 years, which has resulted in 100% mortality of dryland species. The wetland species, Melaleuca cuHcularis and Eucalyptus occidentalis, showed tolerance towards these conditions, although 45.5% of the individuals that were healthy prior to flooding were either dead or dying after prolonged flooding. There was no apparent relationship between tree vigour and age. Melaleuca cuticidaris and Eucalyptus occidentalis seedlings emerged around the wetland margins within months of the water levels receding. Little recruitment of native dryland species was evident even 2 years after water had receded. However, extensive weed invasion and establishment of E. occidentalis and M. cuHcularis in this zone contributed to a change in composition of the peripheral vegetation. Secondary salinity and the extent of flooding within the catchment is likely to increase, leading to further degradation and change in wetland vegetation. The study provides an insight into the possible effects of altering the flooding regime of ephemeral wetlands. Introduction The species composition of wetland vegetation is influenced primarily by water regime, the key parameters being water depth, flooding frequency and duration. Flooding tolerance varies depending on species, age and quality of floodwaters, however rhizospherc oxygen deprivation is eventually fatal to all species irrespective of their flooding tolerance. Even among wetland species, there are no known cases of any prolonged survival over weeks or months of roots being entirely deprived of oxygen (Crawford 1992). Death of wetland plants, asaconsequenceof prolonged flooding, accelerates wetland degradation because of a reduced or ineffective buffer to nutrient input, reduced evapotranspiration resulting in increased capillary rise of saline water, destabilisation of sediment, and loss of food source and habitat for fauna. Changes in water regime and water quality that lead to the decline of wetland vegetation are usually associated with disturbances within the catchment but are external to the wetland (Froend ei al. 1987; Froend &c McComb 1991). Removal of native dryland vegetation results in increased groundwater recharge, associated increased mobility of salt stored in sub-soils, and increased salt and nutrient concentration in the wetland. There have been few south¬ western Australian studies which haveexamined the process of wetland vegetation degradation due to salinity and waterlogging, and/or nutrient enrichment, although degradation is widespread in the south-west (Halse 1993). Of the Western Australian rural wetlands studied to-date. © Royal Society of Western Australia 1994 e.g. Lake Toolibin (Mattiske 1978; Froend et al. 1987; Halse 1987; Anon. 1987; Bell & Froend 1990) and Lake Towerrinning (Froend & McComb 1991), the development of secondary salinity and waterlogging within the catchment has been well advanced by the time of study. Little is known about the changes that wetland plant communities undergo during the early part of the process of secondary salinisation and waterlogging. Without baseline data collected before or during the early phases of vegetation change, it is difficult to understand the ecological processes involved. At present we rely largely on historical information to determine the sequence of events which lead to environmental change and degradation. Of particular importance is the impact that increased salinity and waterlogging has on the recruitment of wetland plants. Most species are dependent on seed production, germination and seedling establishment for successful recruitment, although sedges, rushes and submerged macrophytes readily reproduce by vegetative means (Froend et al. 1993). If increased salinity and waterlogging results in the death or degradation of mature individuals at lower elevations, then the continued survival of the population is dependent on the successful recruitment of seedlings at higher elevations. The effect of salinity and waterlogging on germination at higher elevations in the species range at a wetland is therefore critical. On the south coast of Western Australia near Esperance, a combination of extensive clearing and above-average rainfall years (1986, 1989) caused increased groundwater recharge and surface water retention to an extent where natural ephemeral water courses and basins have retained 15 Journal of the Royal Society of Western Australia, 77 (1), March 1994 surface water for prolonged periods. One ephemeral wetland, Coomalbidgup Swamp, has reportedly contained surface water since the winter of 1986. During 1989, winter rainfall in the catchment was heavy, causing severe water erosion and flooding in the catchment (Anon. 1990), and increased water levels in the swamp. Mortality of fringing wetland and dryland vegetation has resulted due to prolonged inundation. Coomalbidgup Swamp provides an opportunity to examine changes in wetland vegetation structure and composition as a consequence of unusually high water levels and a prolonged flooding period. As the catchment had been cleared in relatively recent times (since 1964), the study permitted the examination of wetland plant community response at a relatively early phase of altered catchment history (cf. wheatbelt catchments). The aspects of wetland vegetation dynamics that were investigated include the extent of vegetation mortality as a result of the 1986 and 1989 flood events, the success of seedling recruitment, and the colonisation of the swamp banks as water levels receded. Study Area Coomalbidgup Swamp is situated approximately 45 km west of Esperance, Western Australia. A single intermittently- flowing creek drains an area of 97 km^ and empties into the swamp from the north-east (Fig 1). Most of the Coomalbidgup catchment (approx. 95%) was cleared for agriculture during 1964-1972. Smallareasof remnant vegetation exist primarily along water courses and around wetland basins, and the vegetation surrounding Coomalbidgup swamp represents the largest stand of native vegetation in the catchment. Secondary salinity is not currently a major concern within the catchment, however, research by the Western Australian Department of Agriculture (Bob Nulsen, pers. comm.) indicates that subsoil salt stores in the catchment are extensive, and pronounced salinisation will occur in the near future with continued elevation of groundwater levels. Figure 1. Location of sample sites at Coomalbidgup Swamp. Sites 1,2,3 and 4 represent the shore transects, and SI and S2 the seedling plots. The traverse transect is shown by the dotted line between shore transect 1 and 2. The area has a 25 year average (1965-89) annual rainfall oi 480 mm. The annual rainfall during 1989 was 35% above the average, and contained the wettest two, three, and four consecutive months recorded since 1969. Runoff in the catchment of Coomalbidgup Swamp is generated after several months of high rainfall rather than by a single wet month or individual high intensity storms of shorter duration. As the water table rises and saturated areas within the catchment increase, an increased proportion of rainfall is expected to runoff. As a result, there is increasing pressure to implement water management programs on agricultural land within the catchment (Anon. 1990). Diversion of excess water into natural water courses downstream, via a system of contour banks and draias, is one means used to control salinity and waterlogging. Wetlands therefore, are receiving much more water and salt than under natural conditions. Methods Species Distribution and Mortality - Transects Permanent belt transects were established at 4 sites around Coomalbidgup swamp in June 1990 (Fig 1) to represent the different plant communities present. Each transect was 30m X 2m with one end at the water's edge; both ends were marked with a star picket. Changes in elevation along the transects were determined at 2m intervals using a dumpy level. For comparison between swamp and terrestrial vegetation, each transect was extended 50 - 60m into the lake to include inundated trees and shrubs on the swamp bed. Due to the depth of water, the identity of dead understorey species on the deepest part of the transect could not be recorded. Vegetation physiognomy, species presence and mortality were recorded at 2m intervals along the transects during June 1990 and June 1992. Sampling in June 1990 took place sufficiently close to the time of death to determine the identity of most species. Specimens were collected of all species and identified at the Western Australian Herbarium. Species Distribution, Vigour and Mortality - Traverse A permanent transect traversing the swamp was established between transects 1 and 2 in June 1990 (Fig 1). A compass bearing of 145® (from transect 1) was followed by boat and each tree that occurred within Im either side of the boat was tagged, species recorded, trunk diameter at the height of the tag (1 m above 1990 water level) and tree height measured and vigour determined. Water depth was recorded at approximately 20m intervals or at major changes in tree composition/distribution. Understorey species were not recorded due to the water depth. The traverse was re-visited during June 1992 and the vigour of all tagged trees determined. Vigour was estimated using a scale of 1 to 5; 1 = dead for >1 year, no leaves; 2 = many dead/dying branches, few green leaves visible; 3 = visible signs of stress but 50 - 70% green canopy; 4 = visibly healthy with few signs of stress, 70 - 90% green canopy; 5 = healthy, no signs of stress, 90 -100% green canopy. A vigour class of 1.5 was assigned to trees which appeared to have died recently (within the last year), judging by the persistence of dead leaves. Height was determined with an extendable tree measuring pole. Trunk diameter at tag height was measured with a diameter tape. 16 Journal of the Royal Society of Western Australia, 77 (1), March 1994 Seedling Plots Two seedling plots were established in June 1990 to determine the density of seedling recruitment on the banks of the swamp. Both plots (SI and S2; Fig 1) were situated close to the swamp shore and orientated perpendicular to the shoreline. The size of the seedling transects varied according to the density and distribution of seedlings; SI was 15 X 10m and S2 was 10 x 6m. The sites cho.sen for the transects were representative of recruitment observed in open shrubland (SI) and closed woodland (S2). Recruitment only occurred along the flotsam line(s) on the shore. Each plot was divided into Im^ cells within which the density and maximum height of each seedling species was determined. Results Transects The species present and their distribution at each of the transects is shown in Fig 2a - d. Transect 1 vegetation at June 1990 varied from Banksia speciosa low (4-5m) woodland on the upper and lower slopes of the present shore, to Eucalyptus occidentalis tall (12-14m) woodland on the lake bed (Fig 2a). Understorey growth on the shore was dense, the dominant species being jacksonia spiuosa. Apart from a few deaths of /. sphwsa, more likely due toinsect attack or'natural' senescence, there were no apparent deaths due to flooding on the upper slope of the shore. The lower part of the shore and the nearshore shallows displayed a high mortality of species present before flooding, though Amphipogon sp. and Patersonia ocf/de«f/7//s were relatively tolerant, and the introduced species Cirsiuni vulgare and Solatium nigrum invaded the saturated soils. The presence of E. occidentalis seedlings near the June 1990 shoreline, suggests a transition from scattered wetland vegetation on the swamp bed to a littoral community. The presence of dead B. speciosa in the nearshore shallow water indicated that the dryland sandplain vegetation once extended below the June 1990 shoreline. On the lake bed itself £. occidentalis was the only species observed. The June 1992 water level was approximately 2 metres lower than in June 1990. Due to the death of the original trees and shrubs and the moist sediment, the area submerged during 1990 but exposed in 1992 represented ideal ground for weed invasion and recruitment of wetland species. As a result, a significant increase in the density of weed and herbaceous native species was recorded on the exposed sedimentinJune 1992. SpeciessuchasC. vulgare, Hypocliaeris glabra, Sonchus oleraceus and Eragrostis curvula were prevalent on the damp exposed shoreline. There was also extensive recruitmentofE. occidentalis andMelaleucacuticularisseedlings between 1990 and 1992 as the water level receded. The shore vegetation of Transect 2 (Fig 2b) was typified by a mid-high (0.5-lm) shrubland with emergent £. occidentalis and M. cuticidaris trees. Although there was no structurally dominant species in the understorey, common species were Grevillea nudiflora, Leucopogon spp. and Hibbertia hyperkoides. The slope of the shore at Transect 2 was shallower than that of Transect 1 indicating wetter conditions over a greater area of the transect. This is supported by the presence of M. cuticidaris on the mid to lower slopes of theshore. Vegetation on the lake bed consisted of M. CHfindrtr/s mid-high woodland, with dead Beaufortia sp.. Acacia alata and Hakea laurina closer lo the shore. Shrubs common on the upper slopes of the shore {e.g. Micromyrtus elobata and G. nudiflora) showed a high degree of mortality when present on the lower flooded areas. Live plants of the lower shore and near shore areas were typically those tolerant of waterlogged conditions, such as £. occidentalis and M. cuticularis; the introduced species S. nigrum also invaded the saturated sediment. The pattern of tree deaths on the lake bed was similar to those of Transect 1. From these remnants of dead vegetation in the nearshore region, it is evident that the dryland vegetation extended 20- 30m further downslopeofthejune 1990 water level. Transect 2 in June 1992, like Transect 1, displayed extensive weed (C. vulgare, S. oleraceus, Rumex acetosella) establishment in the area that was submerged in 1990. Both £. occidentalis and M. cuticularis seedlings had also established over most of this area. The vegetation of Transect 3 (Fig 2c) was similar in structure and dominant species composition to Transect 1. A B. speciosa low (3-6m) open woodland covered the shore and a £. occiden talis tall woodland covered the lake bed. Dominant shore species were Leptospermum erubescens and Melaleuca thymoides. There was high mortality amongst all the species in waterlogged and flooded parts of the transect except seedlings of £. occidentalis and M. cuticularis, and the introduced species C. vulgare and S. nigrum which probably established soon after water levels began to recede. The sandplain vegetation had extended 20-30m further downslope of the June 1990 water level. The mortality amongst the £. occidai talis on the lake bed was similar to the other transects. Transect 3 in June 1992, like all other transects, had extensive weed (C. vulgare, 5. oleraceus, E. curvula) and native herb (Senecio quadridentatus, Epilobium billardierianum, Dampiera sericantha, Muehlenbeckia adpressa, Juncus pallidus, Velleia fn'/ien;/s)establishment in thearea that was submerged in 1990. Both £. occidentalis and M. cuticularis seedlings had also established over this area. Vegetation on Transect 4 (Fig 2d) was similar in structure toTransect2, with the understorey being mid-high shrubland covered by scattered emergent E. occidentalis and M. cuticularis. However, the dominant shrubs of the understorey {Hakea corymbosa, Pbymatocarpus maxivellii and A. sulcata) were different to those of Transect 2. Emergent £. tetragonaand theshrubs Acaciasubcaerulea,Hibbertiaacerosa and Melaleuca striata were only found at the uppermost, and driest, part of the transect. The pattern of shrub and tree deaths was also similar to Transect 2. Transect 4 in June 1992 had weed (C. indgare) and native herb (S. quadridentatus, E. billardierianum, D. sericantha, V. trinenus, £. monostachya) establishment in the area that was submerged in 1990. Both E. occidentalis and M. cuticidaris seedlings also established over most of this area. At all the transects, small seedlings of E. occidentalis and / or M. cuticidaris were found within 2-6m of the shore in June 1990 and 1-26 m of the shore in June 1992. These seedlings were of varying age and had germinated between winter 1989 and autumn 1992. Generally the age of the seedlings increased with increasing elevation. Traverse The lake is a flat-bottomed basin with gently sloping sides and had a maximum depth of approximately 5m during the 1990 sampling period. 17 Journal of the Royal Society of Western Australia, 77 (1), March 1994 5 ”3) ^ ' S S □3 m H U UJ c/5 Z 06 H UJ D Z w U w o ■u 03 O) TJ -o C 03 cu > n C 73 73 C (H 4 -) J= .SP 'o OJ CD c 03 o -C I/) bO c c D 2h ‘0 0/ Cu cn 73 C 03 C/5 E 03 bo .2 2 o es a> u 3 bC T o 73 (Z > I □ T (N O' CJ > 18 Journal of the Royal Society of Western Australia, 77 (1), March 1994 ■ cs ffl m ' 8) 0 c ® 2 J. e'S 2 .2 S P .jj ^ ^ O 3 VI SS-Sow®* mS S ^ c c c .c ^ « a ■ £ 2® ®iS5S.-SS'o. c V) m^^EoS® -.1:1:11 l-g I I mSaSS&l'ol s •S c o T n u a • VN b 19 Journal of the Royal Society of Western Australia, 77 (1), March 1994 X Wp 1 shows the vigour classes of the trees sampled and . an trunk diameter of each class. A total of 173 and 18 M. cuticularis plants were recorded ‘’^'Ihe llOOm transect. It was noted that tree deaths fp'lfddentaUs) along the transect could be classified into two ^ ins those that had recently died (s 1 year ago), and those b had been dead for a much longer period of time. The f‘ r eroup consisted of 107 (61.8%) trees which perished in fire in December 1976 (neighbouring landowners, ^ omm.). This was evident in the lack of bark and minor ches weathered trunk and absence of dead attached I ^"ps Even though local residents report that the lake has ^Tdried since the 1986 flood event, post-1976 tree deaths rP verv recent (s 1 year before sampling). This implies that Ip inundation between 1986 and 1989 was not suffiaent to se tree death within that period, and that the recent ‘^^nrtality observed on the lake bed is a response to the "’mulative effect of the significantly higher water levels ^ ftir the 1989 flood event. Mortality of the d ry land vegetation ^ also likely to be due to the 1989 flood, as there is no idence to suggest this vegetation was inundated after the 1986 flood All recent deaths and poor vigour of the lake bed frees were assumed to be due to prolonged inundation since 1 Q8f, Of the remaining live trees most were found to be m oLr to fair health (vigour class 3 & 4). Injune 1990, trees that had died or were dying since flooding (vigour classes 1.5 and 2) comprised 21.2% of trees alive prior to flooding (Table 1). The lune 1992 monitoring of the trees showed that the nroportion of dead and dying trees increased to 45.5% of those trees alive prior to flooding. Most live trees remained in the poor to fair health categories, however, there was an increase in the number of dying trees (from 5 to 9) and poor health trees (15 to 24). There was a decrease in the number of fair health trees (from 31 to 12) and healthy trees (from 6 to 0), and the number of live trees (classes 2, 3, 4 and 5) decreased from 55 (83.3%) to 45 (68.2%). Apart from an area of small diameter trees in the last 250m of the traverse (SE end), there was no apparent pattern of distribution of vigour and diameter classes, and height across the transect. Figure 3 shows the trunk diameter classes versus vigour categories for the trees on the traverse. Most trees sampled were of small to medium diameter, 0-16 cm. However, the numbers in these classes were made up significantly by trees that perished in the 1976 fire (Dead category=vigour class 1; Fig 3). This ind icales that at the time of the fire a large proportion of the population died, and the survivors have matured since, increasing the range in diameter classes. Trees that had recently died or were dying as a result of prolonged flooding (Recently Dead category = vigour class 1.5+2), were mostly medium to large in d iameter (up to 26-28 cm) but with a significant number of small diameter trees. Live category trees (vigour class 3+4 + 5) had the widest range of diameter classes from 2 cm to 36 cm, with most less than 22 cm. Figure 3. Proportion of trees in each trunk diameter class for different vigour categories. Vigour category Dead = vigour class 1; Live and Recently Dead = vigour class 1.5+2+3+4+5; Live = vigour class 2+3+4+5; Recently dead = 1.5+2. See methods for definition of vigour class. Seedling Plots Seedlings were found amongst the flotsam (dead leaves, capsules, seeds) that collected in a series of lines along the shore of the swamp. It is suggested that the trees on the lake bed may have flowered and dropped seed on the water Table 1 Number (N) and percentage of total trees in each vigour class, and trunk diameter (mean, standard error and range) of trees sampled along the ti averse (see Fig 1) for 1990 and 1992 1990 1992 1990 TRUNK DIAMETER (cm) VIGOUR CLASS N % total N % total MEAN S.E. RANGE all classes 1 (Dead) 1.5 (Recently Dead) 2 (Dying-very Poor Health) 3 (Poor Health) 4 (Fair Health) 5 (Healthy) 173 107* 9 5 15 31 6 61.8 5.2 2.9 8.7 17.9 3.5 173 125 3 9 24 12 0 72.3 1.7 5.2 13.9 6.9 0 10.43 7.93 16.22 6.68 13.19 16.55 14.92 0.54 0.45 3.25 1.50 1.90 1.37 5.87 1-36 1-26.8 1-26 3.1-13 5.4-28.5 5.9-34 4.8-36 Number of trees alive prior to flooding 66 Number of Recently Dead+Dying Trees 14 30 (% of trees alive prior to flooding) (21.2) (45.5) *trees died during 1976 fire prior to flooding 20 Journal of the Royal Society of Western Australia, 77 (1), March 1994 surface each year since the 1989 flood. Measurements of seedling density in June 1990 show variation in seedling numbers in relation to the distance from the water's edge, with peak density at the flotsam lines (Fig 4). At the time of sampling, seedling height on the upper flotsam line (Plot 1) was up to 30 cm, and less than 5 cm on the lower flotsam line would have germinated in Autumn 1990 (just prior to the June 1990 sample date). Only a small proportion of the seedlings observed were M. cuticularis (Plot 1 only). This is probably due to the relative scarcity of this species at the swamp combined with the timing of seed release. The higher Upper Flotsam Lower Flotsam Line Line Towards Waters Edge Lower Flotsam Line Distance (m) Towards Waters Edge — Figure 4. Seedling density along seedling plots 1 and 2. Values are mean ± standard error. • = Eucalyptus occiden tails, O = Melaleuca cuticularis seedling density in Plot 1 may be due to the very open tree and shrub canopy and presence of larger areas of bare ground. At Plot 2, where seedling density was lower, the tree and shrub canopy was dense, therefore restricting light, and the ground was covered with a deep litter layer. Although tlie seedling plots were not monitored in 1992, it was noted that further flotsam lines developed at lower elevations as the water level receded. This resulted in up to 4 lines (1989- 92) of seedling recruitment decreasing in age from high elevations to lower elevations. However these lines were not pronounced where dense vegetation/debris broke up the flotsam. Although not measured, a significant increase in M. cuticularis seedlings on the swamp shore was noted in June 1992. Discussion The results of this study document the detrimental effect of abnormally high lake levels and prolonged flooding. In all four transects, a high mortality was observed for all species typical of the surrounding dryland sandplain flora. The duration and depth of flooding that resulted in plant mortality is difficult to determine. Local farmers recall that the swamp "filled" during heavy rains of 1986 and has not dried, or had substatial lowering of water levels, since. If this is the case, then plants at lower elevations on the shore transects would havebeen partially inundated or submerged for up to 4 years by 1990 and 6 years by 1992. However, the evidence suggests that plant mortality at higher elevations on the shore transects occurred s 1 year before June 1990. Those plants at lower elevations on the shore transects {i.e. totally submerged in 1990) may have died relatively soon after inundation in late 1986 or 1987. The water depths where dead plants were found in June 1990 varied from waterlogged (up to 50 cm upslope of the June 1990 level) to 3m depth. As a consequence of the deaths of dryland species, the width of the remaining peripheral dryland vegetation at some parts of the swamp was reduced by one half. Such vegetation is vital to the conservation of wetlands because it acts as a buffer to disturbance and runoff, and is often the only intact native vegetation and habitat for fauna in agricultural areas. More than 45% of the trees on the lake bed that were alive at the time of the 1989 flooding were either dead or dying by 1992 as a result of prolonged flooding. Mortality of the lake bed trees is expected to increase if the lake does not dry out regularly (every 1-2 years). On all transects, M. cuticularis and £. occidentalis were the only perennial species that displayed flooding tolerance. Both species occur naturally in areas subject to inundation, but the duration of flooding under normal conditions would be much shorter (a few months of the year) than experienced under the elevated water regime of Coomalbidgup Swamp. Depending on future changes in water regime, all or part of the peripheral sandplain vegetation is expected to regenerate gradually. However, from the differences observed between the two sample dates, species composition and diversity of the regenerated vegetation are expected to differ from the pre-flooding condition. Species such as B. speciosa require a fire to trigger seed release, and unless this occurs, recruitment is unlikely. Open disturbed areas. 21 Journal of the Royal Society of Western Australia, 77 (1), March 1994 J u of vegetation due to flooding, were rapidly caused by de growth may colonised by ^ recruitment and establishment of native severely res ^ 1988 ). in some areas speaes (Hf occidentalis germinated amongst dead where seedl g ^ survival of the seedlings may r "'’LTn » Svatd waler regime. Ae »..d by the June depend on a ^iingsofE-ocddcnffl/isandM. cuhculam 1992 obse^ati jj^e water level receded. With continued and subsequent mortality of the trees TthTswamp bed, the distribution of these species may b^ome limited to the Uttoral zone. Ai oresent, Coomalbidgup Swamp contains relatively , u^i-icb water (54 ppt). Given that water levels are "f ■’ S .hi' ruLI *a. wiih lower w.ler levels, ™ win bc'groater than 4 ppt. Upon drying, salt in the f^r^Mil of the swamp bed may reach a concentration in adversely affect the surviving vegetation, which wi continued groundwater rises and fncreasedm^off from the surrounding catchment the 4 bill for secondary salinisation is significant. Subsoil LTt^ore levels in the Sobidge Creek area are thought to be h ih fBob Nulsen, pers. comm.), adding to the future threat of ^ Hon Although £. occidentalis is relatively tolerant of SS and waterlogging (Van der Moezel el «l. 1991), hwhefsalinltles are expected to I,aveadetnmental effect on ti^fvegetation of the lake bed and periphery. The adverse pffeL of higher water levels (without periods of drying) and Potential increases insalinity on wetland vegetation in south- ^ Australia has been documented (Froend ef al. 1987; SThmid 1990; Froend 4. McComb 1991). If the disturbance in the water and salt balance of the catchment goes unchecked, the result is a gradual degradation of the low-lving wetland areas. It is clear that at Coomalbidgup Swainp, the death of a significant proportion of the surrounding mature sandplain vegetation occurred during 1986-1992 because of abnormally high water levels and prolonged flooding. Judging from the age of the trees that died during the recent flooding, similar water levels (and duration of flooding) have not occurred within 15-20 years before 1986. Under pristine catchment conditions, flooding events of this magnitude would be rare (Anon 1990). As studies elsewhere have indicated (Froend & McComb 1991), regular flushing (outflow) of Coomalbidgup Swamp during winter and spring would decrease its total salt load. To reduce mean water depth, ensure regular drying, and increase through-flow, the drainage of the lake may be improved by lowering the elevation of the outflow channel. However, the detrimental effects of increased discharge on other wetlands downstream would need to be determined. Other remedial measures, aimed at the cause rather than the symptom, may include better water management practices on agricultural land such as improved retention of water on farms coupled with greater water use through transpiration. Management plans for wetlands of conservation value in the Coomalbidgup area should consider the effects of increased surface runoff, elevated groundwater, and associated transport of nutrients and salt, and means of reducing them. Acknowledgements: Funding for this project was provided by the Water Resources Planning Branch of the Water Authority of Western Australia and by the Department of Conservation and Land Management. We arc grateful to Viv Reid for making available his information on CoobidgeCreek catchment and initial talks with Viv gave us the impetus for the study. We are indebted to Bob Piggott and all the landholders in the Coobidge Creek catchment for access to wetlands and information regarding rainfall patterns, water levels and the history of the catchment. References Anon. 1987 The status and future of Lake Toolibin as a Wildlife Reserve. A report prepared by the Northern Arthur River Wetlands Rehabilitation Committee. Report No. WS 2, Water Authority of Western Australia, Perth. Anon. 1990 Coobidge Creek Catchment Group Drainage. Gutteridge Haskins and Davey Pty Ltd. Unpublished report to the Agriculture Department of Western Australia. Bell D T & Froend R H 1990 Mortality and growth of tree species under stress at Lake Toolibin in the Western Australian Whoatbelt. Journal of the Royal Society of Western Australia 72:63-66. Crawford R M M 19920xygen availability as an ecological limit to distribution. Advances in Ecological Research 23:93-185. Froend R H & McComb A J1991 An account of the decline of LakeTowerrinning, a whcalbelt wetland. Journal of the Royal Society of Western Australia 73:123-128. Froend R H, Farrell R C C, Wilkins C F, Wilson C C, & McComb A J I993 Wetlands of the Swan Coastal Plain, Vol 4. The Effect of Altered Water Regime on Welland Plants. Water Authority of Western Australia and the Environmental Protection Authority, Perth, 178 pp. Froend RH, HeddleEM, BellDT & McComb A J1987 Effectsof salinity and waterlogging on the vegetation of Lake Toolibin, Western Australia. Australian Journal of Ecology 12; 281-294. HalseS A1987 Probableeffectof increasing salinity on the waterbirds of Lake Toolibin. Technical Report No. 15, Department of Conservation and Land Management, Perth, Western Australia. Halse S A 1993 Vegetation of depth-gauged wetlands in nature reserves of south-west Western Australia. Technical Report No. 30, Department of Conservation and Land Management, Perth, Western Australia. Hobbs RJ1988 Effectsof annual plants on the establishment of shrub .seedlings on a nature reserve in the West Australian wheatbelt. Annual Revegetation Newsletter 1:33-34. Hobbs R J & Atkins L 1988 Effect of disturbance and nutrient addition on native and introduced annuals in plant communities in the Western Australian wheatbelt. Australian Journal of Ecology 13: 171-179. Matti.skeEM1978 Vegetation studies of Lake Toolibin and surroundings. In: Northern Arthur River Wetlands Rehabilitation Committee - Progress Report. Appendix D. Unpublished report to the Minister of Fisheries and Wildlife. Van der Moezel P G, Pearce-Pinto G V N & Bell D T1991 Screening for salt and waterlogging tolerance in Eucalyptus and Melaleuca species. Forest Ecology Management 40: 27-37. 22 Journal of the Royal Society of Western Australia, 77: 23-32,1994 Rottnest Island artifacts and palaeosols in the context of Greater Swan Region prehistory C E Dortch^ Patrick A Hesp^ ^ Anthropology Department, Western Australian Museum, Francis Street, Perth, WA 6000 ^ Geography Department, National University of Singapore, Singapore 0511 Manuscript received October 1993; accepted February 1994 Abstract The prehistoric record of Rottnest island 19 km offshore the Swan Region, Western Australia, consists solely of three stone artifacts. Two are Eocene fossiliferous chert flakes probably deriving from palaeosols in the Tamala Limestone cliffs at Fish Hook Bay and Little Armstrong Bay. A third is a calcrete flake from a siliceous dune blow-out near Fish Hook Bay. A feldspar pebble in situ in a palaeosol intercalated between aeolian calcarenite units at City of York Bay is probably a manuport. The age of the Little Armstrong Bay and City of York Bay palaeosols is estimated to be 15,000 to 50,000 years old. Similarities in pedology and in stratigraphic position suggest that these two palaeosols belong to a single palaeosol unit extending along Rottnest Island's northern shore, a possibility that could give scope to further prehistoric investigations on the island. Prehistoric remains could also be in situ in palaeosols and sandy sediments infilling solution pipes and other Tamala Limestone features on the island's littoral and submerged offshore. Of possible archaeological significance are two charcoal concentrations and a pit- like feature in situ in the lowermost palaeosol unit in the aeolian calcarenite cliffs at Fish Hook Bay. Charcoal from one concentration is radiocarbon dated ca. 18,600 yr b.p. A Turbo shell sample from a storm beach deposit emplaced in a wave-cut notch 3.4 m above sea level in the Fish Hook Bay palaeosol is radiocarbon dated ca. 5700 yr b.p. The former date is probably erroneous, and the palaeosol is estimated to be 40,000 to 80,000 years old. Factors that could account for the dearth of prehistoric evidence on Rottnest Island are: (1) the island's position near the seaward edge of the emergent continental shelf, which was probably less suitable for human occupation than the shelf's more inland parts; (2) site destruction on the island littoral during the Early to Middle Holocene period of rising and high sea level, when the island was 40% larger in area than now; (3) poor surface visibility on the present- day island. The island's minimal prehistoric record is evidence that it was not occupied extensively prior to its formation; its distance offshore implies that it was not visited by Late Holocene Aboriginal sea voyagers. Palaeoenvironmental and archaeological site distributional data from the emergent continental shelf are used in the appraisal of pre-transgression terrestrial environments and prehistoric occupation in the Rottnest locality and elsewhere in the Greater Swan Region. The Rottnest sites appear to be some of the oldest in the region, and suggest the archaeological potential of Tamala Limestone palaeosols. Introduction Rottnest Island, 19 km offshore the Swan Region in south¬ western Western Australia, was separated from the mainland by glacio-eustatic sea level rise about 6500 years ago (Fig 1; Playford 1983, 1988). The 1900 ha limestone island's prehistoric record consists solely of three stone flakes, two collected in 1984 and the third in 1992, and two pebbles (Table 1). These artifacts are the result of many, mostly unsystematic, searches by numerous archaeologists and other Quaternary investigators since the 1960's. Two of these flakes, and another more problematical specimen, a feldspar pebble also collected in 1992, suggest the archaeological potential of the palaeosol horizons intercalated with the aeolian calcarenite units comprising the greater part of the Tamala Limestone (formerly Coastal Limestone), which is © Royal Society of Western Australia 1994 the constituent rock of Rottnest Island, and one of the main Quaternary units in the Perth Basin (Playford 1983; 1988; Playford et al. 1976). The two 1992 finds and the palaeosols at their find spots are described below, along with another recently recorded palaeosol that may be of archaeological significance. We also examine some of the factors that could account for Rottnest Island's extraordinarily sparse prehistoric remains, assess the archaeological potential of the island's littoral and offshore environs, and review palaeoenvironmental data that give insight into the suitability of the Rottnest area for human occupation prior to post-glacial marine transgression. Fundamental to this assessment of Rottnest island prehistory is the premise that the submerged continental shelf west of the lower reaches of the Swan estuary and environs (Fig 1) is an integral part of the "Greater Swan Region" - an informal term referring to the exposed shelf 23 Journal of the Royal Society of Western Australia, 77 (1), March 1994 Table 1 Summary of 1984 and 1992 Rottnest Island prehistoric stone artifacts and other finds Specimen Find spot WA Museum Accession No. Aboriginal Sites Dept Reg. No. 1984 Eocene chert flake Fish Hook Bay B5612 S02099 Calcrete flake Fish Hook Bay, East B3123 S02099 1992 Feldspar pebble City of York Bay B7712 S02276 Eocene chert flake Little Armstrong Bay B7713 S02275 Quartzite pebble Little Armstrong Bay B7746 S02275 Figure 1. Rottnest island and the Greater Swan Region, Western Australia, showing the 10,50 and 100 m bathymetric depth contours. Inset A: Rottnest Island showing salt lakes. Island size ca. 6500 yr b.p. is based on 5 m depth contour (stippled line). The zone between the 5 and 10 m contours (solid line) indicates approximate land area just before island formation (after Department of Marine and Harbours, WA 1988). Inset B; Rottnest Island at maximum inundation during the Middle Holocene. Modern shoreline and salt lakes indicated by stippled lines (after Playford 1988). Insets C and D: Locality maps. area accessible to hunter-gatherer populations during the glacio-eustatic low sea levels that persisted through the late Pleistocene and Early Holocene (Chappell & Shackleton 1986,Thom & Chappell 1975). This region extends westward from the Darling Fault scarp through the Swan riverine / estuarine drainage basin to the steepening edge of the continental shelf 10 km west of Rottnest Island, demarcated by the 50 and 100 m depth contours (Fig 1). Archaeological assessment of pre-transgression human occupation on this and other parts of the submerged shelf of f the Indicin Ocean coast of southwestern Australia is enhanced by the record of numerous Eocene fossiliferous chert artifact assemblages distributed throughout the emergent Perth Basin and Leeuwin Block (Fig ID; Glover 1984), Interpreted as deriving from a "concealed western provenance", i.e. quarry-factories centred on chert outcrops totally or mostly submerged by glacio-eustatic sea level rise (Glover & Lee 1984), these onshore chert assemblages comprise "the most extensive material record in Australia directly relating to human activities [minimally chert quarrying and knapping] on the now-submerged continental shelves" (Dortch 1991). Survey aims and methods Rottnest Island consists mostly of aeolian calcarenite, that is generally covered by heath- vegetated dunes (Hesp et al. 1983, Playford 1983, 1988); the salt lakes cover approximately 10 % of the island's area (Fig 1 A). Survey for prehistoric sites on the island, mainly carried out by one of us (CED) for the WA Museum, has been largely confined to eroded areas on ornearTamala Limestone coastal cliffs and headlands. These features are important in the survey because intercalated with the aeolian calcarenite units forming these cliffs are calcareous palaeosols which could yield prehistoric occupation material, as was suggested by the provenance and surface condition of an Eocene fossiliferous chert flake found on the island in 1984 (see below). The primary survey aim in 1992 was to test the hypothesis that Rottnest Island palaeosols, occurring within the Tamala Limestone and dating to the Late Pleistocene, contained stone artifacts or other prehistoric remains. The survey has closely covered an estimated 50% of the palaeosols exposed on the summits and in the faces of the 24 i Journal of the Royal Society of Western Australia, 77 (1), March 1994 island's coastal cliffs and headlands, and perhaps 30% of the eroded areas of aeolian calcarenite and palaeosol remnants in the island's interior, particularly those near the salt lakes (Fig 1 A). Survey of the smaller areas of siliceous dune blow¬ outs, that are residual from the weathering of the aeolian calcarenite (Playford 1988:22), has not been extensive, with at most 20 % of these features having been searched. Very little survey has been carried out in dense heath, in the extant patches of native low forest, or in the much larger areas of tree plantation. Underwater survey of Tamala Limestone features submerged offshore has been done around Parker Point, in Fish Hook Bay and in Little Armstrong Bay (Fig IB). The archaeological survey has been concerned solely vvith locating and recording individual small finds or features relating to prehistoric occupation prior to Rottnest's formation because, as discussed below, it is improbable that the island was later reached by mid-late Holocene prehistoric voyagers from the mainland. Survey has not been orientated toward terrestrial habitats on the island that would seem likely to have invited prehistoric occupation. This is because the present-day island is a scant and heavily eroded remnant of a vast area of emergent continental shelf that was exploited by prehistoric populations in ways that are conjectural, and because the locality's pre-transgression terrestrial habitats probably differed significantly from those existing at the time of European discovery (Backhouse 1993, Storr et al. 1959). Rottnest prehistoric finds and their provenances The 1984 finds One of the two stone artifacts found on Rottnest in 1984 is an Eocene fo.ssiIiferous chert flake (Fig 2:1) collected from "limestone rubble exposed by the deflation of one or more dune soils" (Dortch & Morse 1984) on an eroding Tamala Limestone cliff-top overlooking Fish Hook Bay (Fig 1B). It is significant that this flake's surfaces (ventral/dorsal faces and butt) are semi-lustrous and only slightly weathered, showing that it has not been subjected long to open-air conditions. The second 1984 find, a deeply weathered calcrete flake (Fig 2:2; Dortch 1991), is from a siliceous dune blow-out on the limestone cliff summit about 300 m east of this bay. The 1992 finds The smaller end of the feldspar pebble from City of York Bay (Fig 2:3) has a fracture surface along a cleavage plane that was broken after the stone was rounded; no archaeological significance is attributed to this fracture. The nearest known sources of feldspar pebbles deriving from the Yilgarn Craton are mainland alluvial gravels in the vicinity of the Darling Fault, 50 km east of Rottnest Island (Fig ID ). This feldspar pebble (weight 12.41 g, maximum length 33 mm) is in the size range of emu crop stones (pers. comm.: R Johnstone, Department of Terrestrial Vertebrates, WA Museum). However, the considerable distance between Rottnest Island and the nearest known feldspar sources, and this pebble's relatively large size suggest to us that it is more likely to be a manuport (i.e, an object transported through human agency) than a crop stone. Figure 2. Prehistoric finds from Rottnest Island. 1984 specimens: 1 Eocene fossiliferous chert flake from Fish Hook Bay; 2 calcrete flake from 300 m East of Fish Hook Bay. 1992 specimens: 3 feldspar pebble from City of York Bay; 4 Eocene fossiliferous chert flake from Little Armstrong Bay. The Eocene fossiliferous chert flake from Little Armstrong Bay (Fig 2:4) weighs 6.98 g, and has a maximum length of 44 mm. It has a diffuse bulb of percussion and a linear butt, features characteristic of bipolar percussion. The two prominent flake scars covering the dorsal face have been produced by blows against its distal end, probably by means of counter-percussion, as suggested by the very small flake scar facets and crushing that have removed the proximal ends of the prominent flake scars. The flake has a prominent notch midway along its left edge' produced deliberately by three or more blows. The flake's surface is uniformly matt, finely porous, and light-hued, whitish-buff, similar to the colour of the sandy palaeosol from which it probably derives (see below). Like the 1984 chert flake (Fig 2:1), the 1992 flake " ... appears to have been buried in a stable deposit for most of the time since it was discarded" (Dortch & Morse 1984). A weathered quartzite pebble (Table 1: B7746) was collected in situ in the palaeosol at Little Armstrong Bay 1 m east of the chert flake's find spot. This piece is conceivably a manuport, although its very small size (weight 1.26 g; maximum dimension 16 mm) suggests that it was naturally transported; for example, it could be a bird crop stone. The City of York Bay and Little Armstrong Bay find sites The two 1992 prehistoric finds are from aeolian calcarenite cliffs at City of York Bay, and at Little Armstrong Bay, 1100 m further north-eastward along the northern shore of the island (Fig 1, Table 1). Each cliff face features a prominent palaeosol intercalated between aeolian calcarenite units. The feldspar pebble was found in situ in the City of York Bay palaeosol (Fig 3a). The chert flake from Little Armstrong Bay, however, was not in situ, but found lying on a 10 cm-thick deposit of fine carbonate sand covering a calcarenite ledge at the foot of the prominent palaeosol exposed in a 1 to 1.3 m- 25 Journal of the Royal Society of Western Australia, 77 (1), March 1994 . .u- rUff face (Fig 3b). This sand closely high section in this almost certainly is resembles that in P section, as is the case with the chert or the aeolian calcarenite unit forming a rrS overhang above .he palaeoaol secdon. S.ev.ng of Figure 3. Tamala Limestone cliffs on showing cross-bedded aeoline calcarenite palaeosols. TOP ritv of York Bay the position of the feldspar pebble (Fig 2.3) Sued r* is in live shaded are. henea.h ,hc calcarende overhang; BOTTOM Little Armstrong Bay, 15 cm scale in centre of palaeosol section. approximately 100 litres of the loose sand on which the flake was lying yielded no archaeological finds, but did reveal small calcareous nodules, rhizotubules and land snail shells (Austrosuccinea sp.), as found in the palaeosol. The palaeosols atboth these find spots display moderately deep (ca. 0.6 m), light-grey A horizons with diffuse contacts to a lower C horizon; no B horizon is present. The A and C horizons in the Little Armstrong Bay palaeosol are formed in fine carbonate sand, with very small fraction of fine-medium quartz grains dispersed throughout. The City of York Bay palaeosol has an A horizon that is 60% fine carbonate sand and 40% fine-medium quartz sand; its C horizon consists of equal amounts of fine-medium carbonatesand and medium- coarse, well-rounded quartz grains. In both palaeosols, small carbonate nodules and rhizotubules (maximum dimension - 1 cm) are present in frequencies estimated at 100 per 1 m^. Land snails {Austrosucchiea sp.) are common in the upper two thirds of the A horizon of each palaeosol. The pale cream-white C horizon in each palaeosol appears to be parent material lacking laminae or bedding planes. Observations of presumed Early to Middle Holocene calcareous soils on Rottnest indicate that the C horizon develops via chemical breakdown of the laminated limestone as the A horizon forms. The slightly cemented C horizons have the same colour, texture, and grain size distributions as the underlying, laminated aeolian calcarenite. Limited organic development or leaching takes place in the A horizon presumably due to low rainfall {ca. 715 mm p.a.), high exposure, semi-arid vegetation with limited turnover, a Mediterranean climate and limited soil fauna. The palaeosols at theCity of York Bay and Little Armstrong Bay find spots are thus pedogenetically very similar, and their stratigraphic positions within each cliff face, and heights above sea level are also much the same (Table 2; Fig 3a, b). Closely resembling these two soil horizons are ones intercalated with calcarenite units at Charlotte Point and at the east end of Catherine Beach, 300-400 m east of the City of York Bay find spot. The Little Armstrong Bay palaeosol extends eastward along the shore about 300 m, and several palaeosol horizons exposed in coastal cliffs several hundred metres further to the east may be part of this same unit. While we stress that the palaeosols cannot be traced laterally very far, the City of York Bay and Little Armstrong Bay palaeosols appear to belong to a single soil unit extending more than 1500 m along the northern shore of the island. Other exposed palaeosols in the cliffed scries of headlands near the western end of the island, 3 to 5 km from City of York Bay, may also belong to this putative soil unit. This is a question of some consequence, since the widespread distribution on the island of a palaeosol having potential for at least rare finds would greatly enhance the chances for discoveries of prehistoric remains. Estimated age of The City of York Bay and Little Armstrong Bay palaeosols The City of York Bay and Little Armstrong Bay palaeosols are overlain by thick aeolian calcarenite units. These are truncated and cliffed lee slope and slipface beds, the foreset beds having been removed by wave action. Even if these aeolian units were formed as steep, climbing dunes, it is highly unlikely that they were formed at modern or mid- Holocene sea levels. This is because the dune crests are also absent, and if one projects the crests and windward (foreset) slopes seaward, the toe of the windward slopes would extend below sea level. The dune units were thus formed 26 Journal of the Royal Society of Western Australia, 77 (1), March 1994 during Pleistocene lower sea levels when the glacial beach was some distance north and west of Rottnest. The age range for the underlying palaeosols is therefore speculated to be ca. 15,000 to 50,000 years old (Chappell & Shackleton 1986; Thom & Chappell 1975). This estimate is further supported by the fact that the overlying aeolian dune units are lithified, whereas younger dunes mantling the Tamala Limestone are poorly lithified or unconsolidated. Many of these younger dunes have steep seaward faces, and appear to have formed during or after the latter stages of post-glacial marine transgression. carbonate sand, and has a very deep (2.3 m) A horizon varying in colour from dark to light brown (mean 7.5 YR 5 / 3: brown). The soil contains Leptopius pupal cases, rare carbonate nodules, and in its uppermost 1 metre AustrosHCcinea shells. The A horizon has a diffuse boundary with a 1.2 m thick lower 'unit' (tentatively designated as a C horizon) showing intense rhizotubule development. In the middle of the A horizon are two charcoal concentrations separated by a pit-like structure (Fig 4a). The charcoal concentrations are approximately 70 cm from the Table 2 Physical description of palaeosol horizons at City of York Bay, Little Armstrong Bay and Fish Hook Bay. Elevation is a.s.l. (top of a horizon). City of York Bay Little Fish Hook Bay Munsell colour A horizon 10YR7/2 10YR7/2 7.5 YR5/3 C horizon 10YR8/1 5 Y8/1 5Y8/1 grain lithology A horizon carbona te / quartz carbonate carbonate C horizon carbonate / quartz carbonate - grain size A horizon fine-medium sand fine sand fine sand C horizon fine-medium-coarse fine - organic content nil nil scattered or concentrated charcoal fragments A horizon thickness 0.8-1.2m 0.6 m 2.3 m elevation 7m 9 m 4 m The Fish Hook Bay Site Most of the palaeosol sections exposed in cliff faces along the northern shore of Rottnest Island have been searched for archaeological material, as have many palaeosol sections on the island's southern shore. Apart from the finds listed in Table 1, the only other possible prehistoric remains recorded so far are in a palaeosol section at the base of the 20 to 25 m - high calcarenite cliffs in the eastern corner of Fish Hook Bay (Table 2). No other exposure of this palaeosol, which is at the base of a series of three aeolian calcarenite units intercalated with two other palaeosols, has been identified at this bay or elsewhere on the island. This lowermost Fish Hook Bay palaeosol is exposed beneath a calcarenite overhang in an eroded section 3.5 m thick; the unit is concealed from view by an 8 m - high mass of huge limestone blocks and scree material fallen from the face of the overlying cliff. A wave-cut notch is formed in the palaeosol at ca. 3.4 m above sea level, and its base is being eroded by wave or tidal action to form an active sea cave of unknown length. The south-facing palaeosol is exposed in a 10 m long section. It consists of very slightly cemented fine top of the A horizon, and about 70 cm apart. Each consists of very friable charcoal fragments 2-15 mm in maximum dimension. The northernmost, best-defined concentration is a horizontal band 65-70 cm long, and about 1 cm thick; the charcoal fragments forming this band arc not contiguous but are separated by sand fill, showing that the charcoal and the sand were deposited at the same time (Fig 4b). This charcoal band shows no signs of having been burnt in situ, cis it has no associated white ash, or fire-crazed and scorched sand grains characteristic of hearths and other fire zones in primary position. The other charcoal concentration is more widely redistributed, and consists of several isolated charcoal fragments forming a horizontally oriented cluster about 30 cm long and 15 cm high. Minute charcoal fragments are present in the palaeosol section above the two charcoal concentrations but not below. Whether the charcoal in these concentrations derives from bush fires, or from hearths or other fires associated with human activities is open to question. The pit-like structure (Fig 4a) is 50 cm wide at its top, and measures 45 cm from top to bottom. The feature is defined by its fill of darkbrown sand, and by a thin carbonate encrustation 27 Journal of the Royal Society of Western Australia, 77 (1), March 1994 1 Fish Hook Bay palaeosol; TOP northernmost f- oncentration and pit-like structure; BOTTOM to^?p of charcoal concentration in 4a. Inboth photographs is a 15 cm scale. on its right side and a line of very small calcarenitc nodules on the other. It could be a burnt tree root or possibly an Aboriginal fire pit. The two charcoal concentrations and the top of this structure are much the same height, suggesting that all three are contemporaneous. Thepalaeosolhasbeennotchedbywaveactionsubsequent to its formaHon and to the deposition of the overlying 20-25 m-high aeolian calcarcnite cliff. A storm beach deposit, 2.3 m long and 30 cm thick, and comprising small (5-8 cm) to large (30 cm) calcarcnite cobbles is emplaced in the palaeosol where the wave-cut notch broadens into a bench at the southern end of the section, approximately 3.4 m above sea level (Fig 5). A few marine mollusc (Turbo) shells are in situ among the imbricated calcarcnite cobbles. The beach deposit ispartlyburied by collapseoftheoverlyingcalcarcnite units, which formed a wide overhang above the storm beach and palaeosol, judging by the massive amounts and size of the calcarenite blocks accumulated in front. Figure 5. Fish Hook Bay wave-cut notch and storm beach deposit, elevation ca. 3.4 m above sea level. Feature consists of imbricated calcarcnite cobbles emplaced in palaeosol. At upper left of cobble deposit is 15 cm scale. The arrow indicates the position of the dated Turbo shell. Age estimation of the Fish Hook Bay palaeosol Judging by its stratigraphic situation, the Fish Hook Bay palaeosol is potentially one of the oldest of these features on Rottnest. However, a radiocarbon sample of charcoal from the northernmost charcoal concentration (Fig4a) gave an age of 18,660±250 yr b.p. (SUA 3030), which we consider to be erroneously young presumably as a result of contamination by younger carbonate. Although the sample was pre-treated to remove younger carbonate by boiling in dilute hyd rochloric acid, because of its small size it was only lightly pre-treated by soaking in alkali-prophosphate to remove humic acids possibly present. Radiocarbon age of the Fish Hook Bay storm beach deposit Our interpretation of the wave-cut notch and storm beach deposit shown in Fig 5 as one of the many Rottnest shoreline features created by Middle Holocene high sea levels relative to the island's littoral and salt lake shorelines (Play ford 1983, 1988, see below) is supported by a radiocarbon date of 5,730 ±60 yr B.P. (SUA 3037) for a whole Tiu/'o shell from the storm beach deposit. A1-2 m storm surge occurring when sea level was ca. 2 m above present level can easily account for the storm beach deposit's 3.4 m elevation. Archaeological potential of shoreline and offshore zones Theoldest knownTamalaLimestoneon Rottnest island is exposed just at and below mean sea level at Fairbridge Bluff in Salmon Bay (Fig 1). Here, Tamala Limestone underlies the partly-emergent Rottnest limestone coral-reef unit (Playford 1988), dated by uranium- thorium technique to 132,000 ±5000 yr b.p. (Szabo 1979). This date suggests that the oldest Tamala Limestone shoreline features on the island, or submerged at shallow depths offshore, could pre-date the earliest human presence in this region. However, exposed at many places on Rottnest Island's mainly rocky shores are cemented soil horizons and sandy sediments that have infiltrated the cavities of solution pipes and other older 28 Journal of the Royal Society of Western Australia, 77 (1), March 1994 limestone structures. At Little Armstrong Bay, for example, part of a palaeosol is exposed in an intertidal erosion pool. This palaeosol, which has in it rhizotubules and Leptopiits weevil pupal cases, is intercalated between wave-eroded units of aeolian calcarenite. The lowermost palaeosol unit exposed at City of York Bay is approximately 1 m above sea level. These shoreline palaeosol units and sandy infillings are potentially places where isolated stone artifacts or other relevant finds could be located, both above and within the2.4 m high inundation zone dating to the Middle Holocene transgression (see below). Archaeological investigation of Rottnest Island rocky shorelines, including test excavation of uncoasolidated sand filling the hollows of two large solution pipes just above sea level at the south-west end of Porpoise Bay, has not revealed any prehistoric material or terrestrial fossils (Figs IB, 6). The partly excavated solution pipe illustrated rests on a 30-cm- thick reddish indurated palaeosol with rhizotubules, whose base is ca. 30 cm above sea level. The sediment excavated from these hvo pipes is a fine, iron-stained (reddish-yellow: 7.5YR 6/8) quartz sand similar to that of the Spearwood Dune System, which is one of the major constituents of the mainland Tamala Limestone (McArthur & Bettenay 1974), though not recorded on Rottnest Island. The presence on Rottnest of these sands is archaeologically significant, since some late Pleistocene to Middle Holocene assemblages of Eocene chert artifacts in the emergent Perth Basin are associated with Spearwood dune.s,c^^^ in the Pinnacles Desert 170 km north of Perth (Fig ID; McNamara 1983), at Minim Cove near the mouth of the Swan Estuary (Fig 1; Clarke & Dortch 1977) and at Dunsborough (Ferguson 1982), as well as with dune soils on the western parts of the Leeuwin Block (Fig ID) closely resembling and occupying the same sequential position as those of the Spearwood Dune System, e.g. at Quininup Brook (Ferguson 1981), Ellen Brook (Bindon & Dortch 1982), Devil's lair (Dortch 1984) and Arumvale (Dortch & McArthur 1985). Figure 6. Partly excavated Tamala Limestone solution pipe, south-western end of Porpoise Bay, Rottnest Island, WA. Playford (1988) notes that sub-aerial features, including aeolian calcarenite, palaeosols and limestone solution pipes, extend below sea level at many localities around Rottnest Island, probably reaching depthr of 70 m or more. For example, there are a dozen solution pipes, 1.5 to 2.0 m high and 0.4 to 0.7 m wide, rising from the sea floor at depths of 3-5 m offshore Parker Point and Little Salmon Bay (Fig IB). These Rottnest solution pipes closely resemble the one in Fig 7, photographed among a large number of submerged solution pipes surrounding a massive calcrete horizon in water 3 m deep near Little Island, two km offshore the mainland, and 25 km north-east of Rottnest Island (Fig 1). Although it is improbable that these features have much potential for archaeological survey, they are still part of a pre-transgression landscape traversed by human groups. Figure 7. Submerged solution pipe inan outcrop of Tamala Limestone, near Little Island, offshore the Swan Region, W.A. (Photographed by Clay Bryce, WA Museum, Perth). Discussion Three factors are probably significant in accounting for Rottnest Island's extremely sparse prehistoric record. 1. The paucity of prehistoric finds may largely reflect late Pleistocene/Early Holocene occupation patterns in the Greater Swan Region. One of us (Dortch 1991) has proposed that the pre-transgression site distribution on the emergent, western half of the continental shelf was similar to that recorded in the present coastal plain (approximating the Perth metropolitan region). Here, Aboriginal occupation during the latter half of the Holocene, and probably during earlier millenia, was concentrated in the coastal plain's eastern, inland half, where abundant surface water and diverse resources are available (Hallam 1987). As suggested then, the outer, western part of the emergent shelf was not much used by prehistoric groups, and occupation instead was concentrated around wetlands and lakes in the shelf's inner parts, a likely example being the terminal Pleistocene - Early Holocene lagoon delineated by the 10m contour in the central part of Cockburn Sound (Fig 1; Churchill 1959, Dept of Marine and Harbours 1988, Searle and Seminiuk 1985). 2. Middle Holocene high sea level (Thom and Chappell 1975), including sea level rise that reduced the newly formed island's area by more than a third following its 29 of the Royal Society of Western Australia, 77 (1), March 1994 Journal unn and continued to rise, thereby time of initial ‘ f^rn^ed island's low-lying areas, inundating the newly extremely probably helps accoun ^ e sparse the Rottnest locality shortly before human occupation ot have been concentrated itsbecomingan islami _^y ^he vicimt.es at low elevation along ^eem to have of the island's •'„ by new evidence from Barker been f''^®l’''^Rf^'Lussedbelow,oraround thesaltlakes. Swamp (Fig brackish or fresh-water at that which perhaps also w time. . „ of the sea floor between Rottnest Island and the ma. J . Jjal isthmus connecting the 1988) imply that ^e pos^‘|^^ ^^^^hed a level 5-10 m two was severed ^ ^^pjh contour line below present rnea ^ ^ contour lines in Fig lA). in Fig. 1, the sea at that level means that at Island formation wt ^ ^go, Rottnest was forty its initial separation ca. 650J (Churchill 1959, per cent larger m ^^rbonr, 1988). However, Department parentlywithinafewcentur.es continuingsealeve ^^^P^ reaching a inundated the no y ^ level relaflve to peak ca. J 2.4 m higher the island is iggg, 1988). At that time, most than at prese shoreline was inundated, and of the present-day . .^333 marine embayment sheltered frona P Rottnest The evidence for these radiocarbon-dated ^'olluscarshell deposits located in modem quarry sites molluscan snt associated with wave-cut ‘"1 ^^^nd notches on the island's salt lake shorelines platforms supplemented by the no archaeological features are associated with any of U 1 shoreline features and deposits, their 1 ™v htlp )o“plai,. the dear.!, ol Rottnest fs'Sd prchioric sites. PUyford (1988) has suggested that the Middle Holocene high sea levels relative to Rottnest island may be more the result of localised s • m Iban of clacio-eustatic sea level rise, conhniung ir;;,tJ:. p-- its causes, inundation on this scale, preceded by the loss of movo than a third of the newly formed island s area, has a destructive potential that cannot be ignored when assessing Rottnest Island prehistory. Poor surface visibility must in part account for Rottnest Island's sparse prehistoric record, with most of the island s Tamala Limestone (including calcareous soil horizons and quartz residual dunes) covered by thick semb or late Holocene dunes that restrict archaeological survey on the island to cliffs and other eroded Tamala Limestone features - particularly palaeosols, and to dune blow-outs and road cuttings. However, it is uncertain whether widespread exposure of subsurface features and ground surfaces on the island would reveal extensive or numerous prehistoric sites. Prehistoric voyages to Rottnest Island? Prehistoric Aboriginal voyages to Rottnest Island can be virtually discounted. For several Australian coastal regions there are ethnohistoric data relating to the seaworthiness of various kinds of Aboriginal watercraft. These data (Jones 1977) show that Rottnest is much too far offshore to have been within feasible voyaging range from the mainland. This is assuming that there was ever incentive to undertake a 19 km voyage to an offshore island in south-western Australia, a region for which there are plentiful ethnohistoric data for estuarine shoreline fishing (e.^. Moore 1978 [1884]), though where there is no indication of Aboriginal watercraft of any kind (Dortch & Morse 1984), and nothing to suggest prehistorievoyagestoany offshore island, includingGarden Island separated from the mainland by a 2 km-wide strait (Fig 1; Dortch & Morse 1984). However, in considering possible prehistoric human visits to Rottnest following its formation, it is necessary to discuss the putative shell midden reported on the island by Hughes et al. (1978), which is a bed of marine mollusc shells (predominately Turbo, with a few limpets) on a low ledge near Parker Point (Fig IB). This particular shell bed is considered to have accumulated through Pacific Gulls dropping shells in order to break them (Playford 1988), as observed on Rottnest Island by Teichert & Serventy (1947) and by Storr (1965). One of us has suggested that the shell bed is a storm beach deposit (Dortch 1991). We concur that the Parker Point shell bed probably is natural ly accumulated, by one or perhaps by both of the processes noted above. Moreover, the low elevation of the deposit (1 to 3 m above sea level) places it within the range of the earlier noted Middle Holocene high sea levels that significantly eroded the island littoral. Therefore, this intact shell bed post-dates these high sea levels, and if it is a midden would presumably result from a human presence on the island following its formation. However, the weathered appearance of the shells in the upper part of the Parker Point deposit suggests to Hughes et al. (1978)thatitisprehistoric rather than having "accumulated as a result of of shell gathering by Aboriginal convicts confined to the island late last century." If the Parker Point deposit is a shell midden, it would be surprising that a deposit of this size, at least 100 times greater in mass than any of the 10 extremely small shell midden deposits recorded along south-western Australia's coastline - extending 1800 km from latitude 29° South to longitude 123° East (Dortch et al. 1984), would bo found on a small, arid and relatively distant offshore island that otherwise has yielded absolutely minimal prehistoric remains of any kind (Table 1; cf. above comments on watercraft). The regional evidence for Aboriginal mollusc exploitation "is decidedly sparse and sometimes equivocal;" (Dortch cf a/. 1984), though this comment does not apply to the Parker Point shell deposit, which is prolific and almost certainly natural. (An archaeological examination of another Rottnest Island shell deposit resembling a midden has been made by Bindon et al (1978). Palaeoenvironmenlal considerations During the last glacial maximum, with the sea at its lowest levels (Chappell & Shackleton 1986, Thom and Chappell 1975), Rottnest can reasonably be asumed to have been a 30 Journal of the Royal Society of Western Australia, 77 (1)^ March 1994 waterless seriesofresourcc-poor limestone ridges and dunes. At that time, the only parts of Rottnest that may have been attractive for human groups are the present-day salt lakes (Fig 1B), which are probably karst structures (Playford 1988). Prior to sea level rise, these presumed caves or dolines could have featured rainwater pools. Although the terrain on Rottnest and other parts of the outer shelf may have been uninviting during the glacial maximum, by the Early to early Middle Holocene, the shelf s still emergent parts would seem to have been suitable for human occupation, in part as a result of coastal plain water table rise consequent to glacio-eustatic sea level rise. These favourable conditions are suggested by radiocarbon dated biotic data from the mainland and from Rottnest Island. A pollen sequence for Loch McNess near the coast 50 km north of Perth (FiglD) shows little change in local vegetation (mainly Eucal\/ptns woodland interspersed with swamp dominated by sedge communities) from 9000 yr b.p. until the present (Newsome & Pickett 1993). A corresponding record is provided by a radiocarbon dated sequence of peat, pollen, aquatic molluscs and ostracods (Backhouse 1993) from the lower part of a core at Barker Swamp, Rottnest Island, which shows that about 7,200 years ago and continuing for several centuries thereafter, this 1 ha swamp was an open freshwater lake surrounded by "sedges and Callitris [native pine] low forest, witharestrictedjarrah /tuartlEucalyptusmarginata/E. gomphocejjhala] woodland present nearby" . The Barker Swamp record is to some extent supported by previous, much less well documented evidence consisting of the remains of a grass-tree {Xauthorrhoea sp.) radiocarbon dated ca. 7,000 yr b.p, and collected from a bore at an unknown locality on Rottnest Island (Churchill 1960). Xauthorrhoea is an understorey plant in jarrah and tuart woodland (Beard 1981), and remains of the.se species are also dated 7,000 yr b.p. in the Barker Swamp core (Backhouse 1993). Barker Swamp is is less than 1 m above sea level; its core record suggests that 7000 years ago, some or all of the island's other seven swamps were fresh-water bodies, as may have been the present-day salt lakes. This core record is evidence that the Rottnest area offered congenial conditions for human occupation for at least several centuries prior to its separation from the mainland. The use of palynological and other palaeoenvironmental data, as well as submerged shelf topography and archaeological site distributions (Dortch 1991), in assessing former terrestrial environments and occupation patterns, as outlined here for the Greater Swan Region, has even greater applicability in those regions where the presence of numerous or extensive Pleistocene/Early Holocene archaeologicalsites on offshore islands or on mainland coasts leaves little doubt that the adjacent submerged shelf was an integral part of the landscape exploited by pre-transgression hunter-gatherer populations. In southern Australia (Fig 1C), these regions include Shark Bay (Bowdler 1990), the Archipelago of the Recherche (Dortch & Morse 1984), Kangaroo Island and neighbouring mainland peninsulas (Lamport 1981), and Bass Strait (c.,^. Jones 1977). Significance of the Rottnest prehistoric finds The Rottnest Island prehistoric finds (Table 1) are all problematical, the two chert flakes because they were not ^ound in situ, and the feldspar pebble because of its less than absolute cultural association. The calcrete flake (Fig 2:2) shows the archaeological potential of the Rottnest residual siliceous dunes; it is relevant that 117 limestone (mostly calcrete) artifacts were recovered from the late Pleistocene deposit at Devil's Lair cave (Fig ID; Dortch 1984). The feldspar pebble (Fig 2: 3) and the two Eocene fossiliferous chert flakes (Fig 2:1, 4) clearly pre-date Rottnest Island's formation during the early Middle Holocene. The two flakes' presence on Rottnest island, like the Uvo Eocene fossiliferous chert flakes from Garden Island and the many hundreds of others from the Archipelagoof the Recherche (Fig 1C; Dortch & Morse 1984), is in keeping with their presumed age and derivation from chert outcrops on the emergent shelf. These rare prehistoric finds to date imply that the pre¬ transgression land mass broadly synonymous with Rottnest Island was probably never occupied intensively. If such occupation did take place, the archaeological remains have been destroyed by Middle Holocene sea level rise and continuing marine conditions, or are buried beneath marine/ lacustrine sediments, dune sands or some of the island's aeolian calcarenite units. The potential of palaeosols within Tamala Limestone sequences on Rottnest Island for yielding rare archaeological finds is now apparent, and this potential may extend to palaeosol units in the Tamala Limestone throughout the Perth Basin and Leeuwin Block. The presence of what may be asinglepalaeosol unit extendingfor several km alongRottnest Island's northern shore offers perhaps the best opportunities for further prehistoric investigations on the island. The cultural finds described here and the estimated age of the palaeosols at the Little Armstrong Bay and City of York Bay tind sites are suggestive of an age for prehistoric occupation insouth-westem Australia equalling or exceeding thatshown for Devil's Lair - ca. 33,000 yr b.p. (Dortch 1984) and the Upper Swan site - ca. 38,000 yr b.p. (Pearce & Barbetti 1981). Further investigations on Rottnest should clarify the ages of some of these palaeosols, and verify the occurrence of prehistoric material within them. Acknoivledgements: We wish to thank Uie Rottnest Island Authority for sponsoring the radiocarbon dating, and for accommodation and support. Our thanks to Mike Barbetti and Gillian Taylor of The N W G. Macintosh Centre For Quaternary Dating (University of Sydney) for advice and assistance in problems relating to the radiocarbon age of the Fish Hook Bay palaeosol and associated features. Thanks also to Ms L K Lee for cartography, to Joe Dortch for assisting in the June 1992 Rottnest archaeological survey, and to WA Museum staff members for the following help. Clay Bryce, Molluscs Department, advised and took the underw^atcr photograph in Fig 7, which was printed by Dougla.s Elford, Photography Department. Ken McNamara, Alex BevanandGcorge Kendrick, Departmentof Earth and Planetary Sciences advised, ns did Mance Lofgren of Anthropology Department. Thanks to Peter Bindon of the last named department, and Richard Gould, Anthropology Department, Brown University for advice and encouragement, and to an anonymous reviewer for detailed and constructive criticism. References Backhouse] 1993 Holocene vegetationand climate record from BarkerSwamp, Rottnest Island, Western Australia. Journal of the Royal Society of Western Australia 76:53- 61. Beard J S1981 Swan 1:100 000 Vegetation Series Explanatory Notes to Sheet 7. University of Western Australia, Perth. Bindon P & Dortch C E 1982 Dating problems at the Ellen Brook site. Australian Archaeology 14:13-17. Bindon P, Dortch C & Kendrick G 1978 A 2500 year old pseudo shell midden on Long Reach Bay, Rottnest island, WA. Australian Archaeology 8; 162-171. 31 Journal of the Royal Society of Western Australia, 77 (1), March 1994 BowdlerS 1990The Silver Dollar site. Shark Bay: an interim report. Australian Aboriginal Studies 2:60-63. Chappell,]&ShaclM’-;--t-*.Ai#;' ■-•!■' :•* '! •• - ■ ■ “ , '■-*:' ■■^■'; J^t'^ „• r*a '*4 •• tpwifi - ^ ^ ^ t ■ Ti \mC^ ^ ^ ' 3r -. V ■> . 4v ■*’.n>'*,“' ^:'l, *'*■•»-.*■* ^ ■■ 4^-*'-‘ ‘*'Xt / J i * • . • ' . :'»'5^^' ' ■ ' * 1 - ^ ■ , r - , f Vi' ' 'V • ,i .^' .* ,. -iS-.p. J '/I' f *' ■ ?*’ ■' /• ^ .► - i . . 1^ fc t* • { I . '*' 4 * V -" JM ■ . • >>>w I- Vi?’- -**- ‘ ' ' « ] •* '. Vs.^-,-V ’'t'--■ ‘ J *, ‘^■^m - ■ ' ” ’.V- t, **, fVi;S,4 ' . * '‘- I* .. * - .-i ? -r *• -14 ’ \‘i - - . .f Vf'- ■;i^ -'v? ■ ' ^ »«. .' r--i. . , •*», ..:.A ,.,>i*'. -. • •■ a«i^^. ■ •3^'.■■-it-,■>; ’ • ^ : : *■ '.,. T-;-. Journal of the Royal Society of Western Australia^ 77: 37-43,1994 Convergent Evolution in the Dentitions of Grazing Macropodine Marsupials and the Grass-Eating Cercopithecine Primate Theropithecus gelada N G Jablonski Department of Anatomy and Human Biology, and Centre for Human Biology, The University of Western Australia, Nedlands, Western Australia 6009 Manuscript received January 1994; accepted May 1994 Abstract This study examines the dentitions of two sets of non-ungulate grazers in relation to their diets. These sets were the grazing macropodine marsupials (including the genera Macropus, Peradorcas, Onychogalea and Lagorchestes) and the large, grass-eating cercopithecine primate Theropithecus gelada. The two sets are highly selective grazers whose diets consist mainly of green grass parts. Both were found to share four unusual dental adaptations for trituration of grasses and extension of the active life of the cheek teeth despite the severe abrasive effects of their diets: 1) bilophodont molars, which when worn present distinctive patterns of longitudinal and transverse enamel cutting ridges; 2) arrangement of the mandibular cheek teeth in an upwardly convex curve (reversed Curve of Spee) so that occlusion between these teeth and their maxillary counterparts is concentrated in a small anterior area of the tooth row instead of along its entire length; 3) accelerated anteriad movement of cheek teeth propelled by the anterior force of occlusion that permits relatively unworn, posterior elements of the cheek tooth battery to be moved anteriorly into areas of intensified function; and 4) enlargement of the transeptal interdental fibres of the periodontal ligament to maintain close contact between the elements of the cheek tooth row and thus convert the row into a single functional unit. This convergent evolution of grazing macropodine marsupials and a grass-eating primate is particularly interesting because the adaptations for grazing are considered highly derived in both the macropodine and cercopithecine lineages, and that the pathway of evolutionary change followed similar courses in both groups. Adaptations for grass-eating are thought to have arisen in ancestral types of both lineages that subsisted on diets of softer, less resistant vegetation and that possessed dentitions that lacked specializations for the reduction of large quantities of abrasive, high-fibre vegetation and the preservation of relatively unworn tooth substance throughout adult life. Introduction Herbivores face considerable problems imposed by the ^^gh rate of wear on their teeth caused by mastication of large 9^antities of tough, fibrous and frequently abrasive vegeta- These problems are particularly serious in grazers, ^hich consume large quantities of grasses that are rich in ^trasives. Grasses contain endogenous abrasive silicates (phytoliths) tliat deter attack from herbivores, and poten- ^*slly damaging exogenous abrasive materials such as sand ^nd grit from soil. Many mammalian species have evolved ^^rious combinations of dental modifications to enlarge the ^^rface area of the postcanine teeth, to prolong the life of ^hese teeth, or both that compensate for the rapid rate of ^^ntal wear caused by these abrasive materials. Enlarging total surface area of the cheek tooth permits better Subdivision of plant material into small parts, although this *^^9uires increased muscular force. This increased subdivi- facilitates either direct digestion of starch or indirect %estion of cellulose through microbial symbionts. A simi- effect is gained by increasing the total length of the tooth's ^Uarnel cutting edges. Society of Western Australia 1994 The strategies that have evolved in various mammalian lineages for the exploitation of grasses as a primary source of food are many and varied, with each grazing lineage pos¬ sessing its own unique combination of features for harvest¬ ing, chewing and digesting grass parts. Because grass is low in simple sugars, low in protein and high in fibre, successful grazers must strike a balance between the metabolic energy they expend in processing grass and the energy they receive in return. The complex carbohydrates present in grasses carmot be directly metabolized by mammals, and so most grass-eating mammals have evolved modifications of their digestive tracts for fermentative breakdown of these carbo¬ hydrates into usable sugars and volatile (short-chain) fatty acids. The two basic styles of gut fermentation that have evolved in mammals, foregut fermentation and hindgut fermentation, differ in their efficiencies and have been incor¬ porated in two rather divergent feeding strategies in the animals that use them. Foregut fermenters {e.g. artiodactyls, kangaroos, and colobine monkeys) extract high metabolic yields from moderate amounts of relatively high-quality forage, while hindgut fermenters (c.^. perissodactyls) rely on bulk feeding of lower-quality forage to compensate for a reduced nutrient extraction rate (Janis 1976). Accordingly, 37 Journal of the Royal Society of Western Australia, 77 (2), June 1994 the teeth of foregut fermenters tend to be somewhat simpler in design than those of hindgut fermenters because they are required to process less vegetation of generally higher qual¬ ity {i.e. lower in fibre and abrasives, and higher in water content). The mechanical properties of foods differ, with some requiring more compression or crushing and others (such as grasses) requiring more cutting or shearing. An optimal crush/shear ratio exists for breaking up each type of food (Osborn & Lumsden 1978), and the masticatory apparatus of every species has evolved, in part, to generate the crush/ shear ratio best suited to its preferred diet. As discussed by Osborn (1993), this ratio depends on the direction of tooth movement during mastication, the orientation of the occlu¬ sal surfaces of the teeth and the direction of the bite force. Feeding on grasses is usually related to the evolution of several specific dental modifications that are used in various combinations. The most common of these modifications include an extreme increase in the surface area of the teeth, the evolution of complex enamel patterns on the occlusal surfaces, an increase in the height of the crowns of the molars, and the conversion of the molars into rootless, con¬ tinuously growing structures. Of these, the first two act more to increase the surface area and food-processing effi¬ ciency of the tooth while the last two act to prolong the life of the tooth. The high-crowned (hypsodont) molars of certain perissodactyls, artiodactyls and rodents, for example, are usually associated with an abrasive diet and are commonly assumed to have evolved in response to the consumption of abrasives associated with grasses (Stirton 1947; Janis & Fortelius 1988). Manatees (genus Trichechiis), consumers of sea grasses, have evolved yet another strategy - that of continuous posterior eruption of an indefinite number of new molars in response to the wear and anterior shedding of the anterior molars (Domning & Hayek 1984). Because the dentition is only one part of the grazer's digestive armament, it breaks down grasses in a manner that is complementary to the type of fermentation that will occur in the gut. On the basis of the rather limited number of major types of molar and jaw design we see in grazing mammals, however, there would appear to be design constraints that are operating on the evolution of the.se structures in specific lineages. Some of these constraints no doubt derive from the mechanical properties of the dental hard tissues them.selves and the need to minimize the potential for crack propagation in brittle enamel (Koenigswaldetfi/. 1987). In each grazing lineage, there is a unique compromise between the size of individual cheek teeth, the design of the occlusal surfaces of the teeth, the total number of cheek teeth an animal can develop in its life, and the number of teeth in occlusion at any one time. The anatomical compromise that evolves in any grazing lineage is teeth that must be powerful enough to physically break down grasses yet durable enough to last through the entirety of the animals' reproductive lives. In¬ creasing the amount of tooth material (and therefore the area and total length of cutting edges) is one solution for increas- ing grass comminution and reducing wear. However, the cost of larger teeth to the animal, in the development of the teeth themselves, in larger masticatory muscles and in dental supporting tissues, is great. An enlarged masticatory appa¬ ratus is energetically costly to construct and maintain, and in some lineages the costs appear to outweigh the benefits. Among mammals, neither herbivorous marsupials nor primates exhibit the extreme dental modifications for coping with high rates of dental wear seen in lineages with long evolutionary histories of grazing {e.g. perissodactyls or artiodactyls). Yet, among both the Metatheria and the Pri¬ mates there are species that rely exclusively on diets of grasses. These species display unusual dental modifications that increase the efficiency of the postcanine teeth in the comminution of grasses and that prolong the life of the postcanine dental battery, but that differ from those of other more widely known grazers (Jablonski 1981). In this study, the dentitions of the grazing Australian macropodine mar¬ supials comprising the genera Macropus, Peradorcas, Onychogalea and Lagorchestes and the large cercopithecine monkeys known as geladas or gelada baboons {Theropifhecus gelada) of the highlands of central Ethiopia arc examined. These groups have been compared because of the remark¬ able similarities observed in their postcanine dentitions. Similarities in diet and dental morphology between the grazing macropodines and the geladas are explored in de¬ tail. The evolutionary histories of these animals are then compared to see if the apparent convergence in dental struc¬ ture and function between the groups can be related to similarities in selective pressures acting on similar ancestral morphologies in their respective lineages., Dental Structure and Function in Relation to Diet Grazing Macropodines Among the extant Macropodidae, species of Macropus, Peradorcas, Onychogalea and Lagorchestes have been classified as grazers because of their reliance on abrasive, high-fibre vegetation often in the form of grasses. Although .some species-specific and regional variations in dietary prefer¬ ences have been reported among these species, grasses domi¬ nate their diets Qarman 1984). For instance, the diet of Macropiis gigauteus in southwestern Queensland is supple¬ mented by dicotyledons throughout the year (Griffiths & Barker 1966, Griffiths et ai 1974), while that of M. rufiis includesdrought-resistantshrubsin dry times (Jarmanl984). Grazing macropodines are highly selective feeders, which can use their narrow muzzles and incisal arcades to pick individual grass blades. They show a preference for green, readily digestible material and avoid, where possible, highly fibrous and dry plant parts (Jarman 1984). The structure and function of the teeth in grazing macropodines was described in detail by Sanson (1978, 1980), who defined a series of five characters associated witK a diet of abrasive, high-fibre plants. These were: 1) bilophodont molars with strong links (longitudinal ridges) between the transverse loph(id)s; 2) a convexly curved lowet- tooth row (describing a reversed Curve of Spee) that meets the upper row at a tangent so that only the anterior portionn of the tooth row are in occlusion; 3) occlusion that consists o f an initial forward motion of the lower molars followed by marked lateral movement in which the developed links ar^ in contact with the opposing lophs; 4) molar progression, by which relatively unworn lower molars are advanced to anterior positions in the jaw while worn molars are sheej^ anteriorly; and 5) reduced or vestigial permanent premolar^. 38 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Figure 1. Comparison of the dentitions of a grazing macropodinc (A; Macropus robustus; W.A. Museum M5339) and a grass- eatingpnmate(B;TheropithecusgeladaHKlJ/UWA0242)\n lateral view. In the line drawings of the same specimens, the curves represent the reversed Curves of Spee and the arrows indicate the regions of the dentitions at the most superior points on the curves where the forces of mastication are most strongly concentrated. Scale bars in all views indicate one centimeter. These basic anatomical characteristics of the dentition of grazing macropodines are illustrated in Fig 1A and Fig 2A. The Curve of Spee is a term used to describe the upwardly concave profileof the human mandibular dentition as viewed from the side. A reversed Curve of Spec is an upwardly convex curve of the lower dentition. As described by Sanson (1980), mastication in grazing macropodines is ideally suited to the comminution of tough, fibrous materials. Processing of plant materials is accom¬ plished mostly by cutting on the sharp enamel edges of the molar teeth. Although the sharpest cutting edges are on unworn molars, further cutting edges are exposed as the enamel on the surface of the crown wears and the underlying dentine is exposed and excavated (Fig 2A). The efficiency of the molar increases up to a certain point, after which the molar is so worn that it is no longer effective as a cutting tool. Because the focus of cutting action is at the point of contact between two enamel cutting surfaces, the wearing down of those surfaces will result in diffusion of occlusal pressure over a larger area and less effective comminution (Sanson 1980). As their functional life approaches an end, molars move into nonfunctional positions in the anterior portion of the jaws before being shed. Through the process of molar progression, a molar tooth erupting in the mandible moves anteriorly and dorsally into the occlusal plane. The molar continues to move anteriorly and dorsally as it becomes worn and, at the height of its functional life, occupies a position on the most superior point on the reversed Curve of Spee (Figure 1). As the tooth wears further, molar progres¬ sion moves it anteriorly and ventrally out of the occlusal plane until it is shed (Sanson 1980). This process begins with shedding of the premolars and their position in the jaw being taken by the first molar, with the other erupted molars following. Molar progression continues with the wearing down and shedding of the first, second and subsequent molars in succession (Tyndale-Biscoe 1973). The process of molar progression appears to be driven by the direction of the force of occlusion. Anteriad mov'ement of the mandibu¬ lar dentition relative to the maxillary dentition during mas¬ tication produces anteriorly directed forces that, in turn, are transmitted to the dentition as a whole by the transeptal interdental fibres of the periodontal ligament (Sanson & Miller 1979). These fibres act as links in a chain to maintain tooth-to-tooth contact. Normally, in kangaroos, only four molars erupt in each jaw so that very old individuals may have only one worn molar tooth left in each jaw. Thus, the 39 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Figure 2. Cheek teeth of a grazing macropodine {Macropus robustus) and a grass-eating primate enamel occlusal view (specimens as in Fig 1). Line drawings of the same specimen indicate the transverse and long ridges that shear against their counterparts in the upper dentition to comminute grass parts during masnca in all views indicate one centimeter. life span of these animals is dictated in large part by the durability of their teeth. Only in the little rock wallaby, Peradorcas concinna, are new molars continuously produced in the back of the jaws, apparently in response to a diet of grasses and ferns that are rich in abrasive silica (Sanson et al. 1985). Theropithecus gelada Among Primates, Theropithecus gelada is unique in its almost exclusive reliance on a diet of grasses. Geladas prefer grass blades to herbs or shrub vegetation when the former is available (see Iwamoto 1993a for a review). In the localities where the feeding behaviour of geladas has been observed throughout the year, grasses were found to constitute 90% of the diet in the wet season, but around 60% at the height of the dry season. Although green grass blades constitute most of the grass parts eaten, grass seeds are intensively exploited by geladas when they are available. Geladas are highly selec¬ tive feeders but, unlike grazing macropodines which crop grass with their incisors, they harvest young grass blades between the highly opposable first and second digits of the manus before transferring them to the mouth. The same technique is used for the harvesting of rhizomes, which form an important part of the diet in the dry season (Iwamoto 1993a). Because geladas do not clean the rhizomes before ingestion, this increases their consumption of exogenous abrasives from the soil. Feeding is generally a continuous operation in which the animal masticates the previous hand¬ ful of grass parts while gathering the next (Dunbar 1977). The time geladas spend feeding each day is longer in the dry season than in the wet season because the animals must spend longer to find green blades and succulent stem bases of grasses (Iwamoto 1993a, b). The ability of geladas to harvest vegetation quickly and with great precision appears to have been an ancient attribute of the Theropithecus lineage. as judged by the fact that the digital proportions of were presaged in a Pliocene member of the lineage (Jablonski 1986). In T. gelada, the demands of mastication of an abrasive, high-fibre diet are met with a series of structural modifica tions of the teeth and jaws that make possible particularly efficient trituration of large quantities of grass parts QoHy 1972, Jablonski 1981,1993a). In the dentidon, these modifi¬ cations comprise five major features: 1 ) bilophodont molars with columnar cusps that, when worn, present a pattern of complexly curved enamel cutting ridges; 2 ) alignment of the mandibular cheek teeth in a reversed Curve of Spec so that, for much of the life of the individual, contact between the upper and lower tooth cheek tooth rows is concentrated anteriorly; 3 ) occlusion that is characterized by a short anteriad movement followed by a marked lateral component that permits the transversely and longitudinally oriented enamel cutting edges in the lower cheek tooth row to move across those of the upper row; 4) accelerated mesial drift of the molar teeth promoted by heavy interproximal wear and delayed eruption of the lower third molars, by which rela¬ tively unworn molars are moved to more functional, ante¬ rior positions in the jaws; and 5) possession of enlarged mesial shelves and distal accessory cuspules on the first and second permanent molars, and a large and consistently well- formed hypoconulid on the lower third molars in order to increase the functional length of the cheek tooth rows 1972; Jablonski 1981,1993a; Swindler & Beynon 1993). The basic anatomical characteristics of the dentition of T. gelada are illustrated in Fig IB and Fig 2B. The molars of the gelada display considerably more occlusal relief at most wear stages than do those of any other cercopithecines, but they probably approach optimum op¬ erational efficiency in a relatively unworn state when the 40 Journal of the Royal Society of Western Australia, 77 (2), June 1994 greatest number of exposed enamel ridges is av'^ailable for separation of plant material (Jablonski 1981; Meikle 1977). In unworn molar teeth, the cusp pairs of the mesial and distal loph(id)s are partly separated by deeply waisted buccal and lingual enamel folds. When worn, this arrangement pro¬ duces a series of incomplete transverse ridges of harder enamel alternating with softer dentine (Fig 2B). Because of the height of the cusps of the molars and the depth of the clefts and basins, the molars of T. gelada retain occlusal features for a long time (Meikle 1977). Further, the long buccal surfaces of the molar teeth permit the long anatomical crown to function until the root supporting system of the teeth is compromised (Swindler & Beynon 1993). In T. gelada, the highest occlusal pressures can be gener¬ ated at an anterior position in the jaws, at the apex of the reversed Curve of Spec Qablonski 1981; see Fig 1B). During the early part of adult life, this is where the upper and lower first permanent molars come into contact, but as these mo¬ lars wear and move anteriorly in the jaws their positions are gradually occupied by the anterior loph(id)s of the second permanent molars, followed by the posterior loph(ids). The curved orientation of the lower tooth row, the action of mesial drift and the delayed eruption of the elongated lower third molar tooth help to ensure that relatively unworn tooth substance is retained until late in lifeOahlonski 1981). In one mandibular specimen (HKU/UWA 0242), the fourth premolar showed severe wear of the interproximal enamel and a worn occlusal surface, the first molar was heavily worn on the occlusal surface while the second molar was only slightly so, and the large third molar was barely worn and the root apices were still open (Miller & Jablonski, unpub¬ lished observations). One of the effects of this process is the production of a steep wear gradient in the molars oiT. gelada, characterized by considerably heavier wear in the anterior than the posterior molars at any given time Qablonski 1981). The mesial drift that occurs in the gelada is not the same as the molar progression in the grazing macropodines, in which the continual movement of molars is accommodated by sequential loss of anterior teeth. InT. gelada, aWihemoiars are, generally, retained throughout life and are not shed, despite very heavy occlusal and interproximal wear. Near loss of anterior cheek teeth by shedding has only been observed in one aged female specimen, which interestingly, also possessed supernumerary upper fourth molars. In 7. gelada, several features of the alveolar bone and periodontal ligament indicate that the cheek teeth are functioning and moving as a unit. These include the presence of buttressing bone around the mandibular cheek teeth and enlargement of the trans-septal interdental fibres of the periodontal liga¬ ment (Miller & Jablonski, unpublished observations). Ante¬ rior movement of the cheek teeth in the gelada appears to be driven, at least in part, by the anterior force of occlusion, but marked anterior movement (and shedding of obsolescent teeth anteriorly) seems to be prevented by the stabilizing effect of the large upper canine and lower sectorial premolar (Miller & Jablonski, unpublished observations). Evolution of Dental Form and Function The similarities in dental morphology shared between grazing macropodines and geladas are striking, despite major differences between the two groups in overall cranial shape. Both groups possess bilophodont molars which, while different in the cusp and crest disposition in the unworn state, appear remarkably similar when moderately worn. The sequence of dentine and enamel ridge exposures produced by interproximal (interdental) wear and wear on the occlusal surfaces is virtually identical in both groups, although it appears that the more bunodont macropodine molars approach functional obsolescence earlier than those of the gelada. The similarities between the groups in the conformation and movement of the cheek teeth in the jaws are perhaps even more remarkable than those of the teeth themselves, and distinguish them from their respective non-grazing relatives. In both groups, the reversed Curve of Spec acts to concentrate the muscular force exerted by the muscles of mastication in a relatively small, anteriorly located area of the tooth row. Molar progression in the macropodines and mesial drift in the geladas provide mechanisms by which animals of both groups are provided with a continuous supply of relatively unworn and highly efficient triturating surfaces to this focal point of occlusal pressure. The geladas' adaptation is the more conservative because heavily worn molars are retained in the jaws, resulting in an ev^entual flattening of the occlusal profile in old age and apparent dispersion of the force produced by the muscles of mastica¬ tion over a larger area of the molar row. Anteriad movement of molars continues throughout life in the gelada, but the continuous, slot-machine-like replacement of molars at the back of the jaws does not occur as it does in grazing macropodines or, as it does in the even more extreme form, in the manatee and Peradorcas. In this regard the geladas can be viewed as lying at the conservative end of a spectrum of dental and gnathic adaptations common in non-ungulate grazing mammals to extend the active life of the cheek teeth despite the severe abrasive effects of diet. Discussion Morphologists arc quick to identify “adaptive features" in the animals they study, but it is difficult to prove or establish the adaptedness of these features without resorting to circular argumentation. Reference to conditions in an “outgroup" is a method that reduces circularity and, in the caseof the animals under coasideration here, the adaptedness of the dental morphology in one group is supported by that discovered in the other distantly related group. As Davis stated, “the convergent appearance of similar conditions in more or less remotely related organisms under similar or identical environmental conditions is the most readily avail¬ able proof of adaptive value" (1949, p. 80). The identification of a similar suite of dental characteris¬ tics in two phyletically distant lineages argues strongly in favour of the occurrence of convergent evolution of dental form and function in response to closely comparable envi¬ ronmental stimuli in the two lineages. Further, this demon¬ stration of convergence argues strongly for the adaptive advantage of the unusual suite of features of the dentitions of the two lineages in relation to graminivory. Sanson (1978) has argued that macropodines belonging to the “derived grazing grade" can be contrasted to those of the “ancestral browsing grade". Those macropodines he 12966—2 Journal of the Royal Society of Western Australia, 77 (2), June 1994 assigned to the browsing grade {e.g. Wallabia, Dorcopsis, Dendrolagus and Setonix) subsist on diets of relatively non¬ abrasive vegetation of low fibre content, such as the leaves of dicotyledonous plants, and occasional fruit and flowers. Their dentitions lack the conspicuous adaptations for the processing of tough vegetation and the preservation of tooth substance, and instead are characterized by bilophodont molars with weak longitudinal ridges, a flat cheek tooth row in which the permanent premolar and the four molars in the upper and lower jaws meet each other along a flat occlusal plane, and modest amounts of mesial drift (Sanson 1978). The derived grazing grade of macropodines evolved from the ancestral browsing grade, Sanson (1978) reasoned, as a response to climatic and ecological events beginning in the terminal Miocene that spurred the decline of mesic vegeta¬ tion in the central regions of Australia and the rise of xeric grasslands. The specialized nature of the dentition of Tlieropithecus has been recognized for many years and, as in the case of the grazing macropodine dentition, is a relatively recent innova¬ tion that can ultimately be traced tolateTertiaryclimaticand ecological change. Although some aspects of the emergence of the genus are still not fully understood (see Jablonski 1993b), the T/icrop/f/iccHsdentition clearly representsa highly derived condition for the tribe comprising the largest cercopithecines, the Papionini. The dentitions of other papionins such as macaques (Macaca) and common baboons {Papio) more closely resemble the primitive condition for the tribe in the absence of conspicuous adaptations for exclusive graminivory. Compared with the molars of Theropifhecus, their molars possess more bunodont crowns with shallow notches and clefts and are arrayed in a flat tooth row. Moderate levels of molar wear in these animals result in the exposure of a simpler pattern of exposed enamel and dentine on the occlusal surfaces and the production of a more even wear gradient along the cheek tooth row. The emergence and early diversification of Theropitheciis in the Pliocene in East Africa was clearly linked to the evolution of a feeding apparatus specialized for the eating of grasses (Jablonski 1981, 1993a). This made possible the invasion of the more open, grassland environments that were evolving in East Africa in the late Miocene. The specializations for grazing in Tlieropithecus included those of the hand — that permitted early theropiths to harvest the vegetation of the emergent grassland environments — and those of the masticatory apparatus (Jablonski 1986, 1993a). Theropithecus thus occupies a position in what could be called the derived grazing grade for primates. The dental specializations in both the grazing macropodines and the geladas represent apparent compro¬ mises between the demands of diet and constraints of dental and gnathic design in the two lineages. In both, the total length of the enamel shearing crests and cutting edges of the molars has been increased relative to that seen in their browsing counterparts (Benefit & McCrossin 1990) and, at least in the case of the geladas, it can be argued that these features could not have been increased further within the constraintsof a bilophodont tooth. In addition, both lineages — with their reversed Curves of Spec and mechanisms of molar progression or mesial drift — have evolved similar, simple mechanisms for conserving relatively unworn occlu¬ sal surfaces long into adult life while maximizing the occlu¬ sal pressures that can be exerted between opposing molar teeth. The reversed Curve of Spee concentrates occlusal pressures in a small area of the cheek tooth row and thus helps the animals make the most of molars that, by the standards of grazing ungulates, are relatively simple. It would thus appear that the constellation of dental features shared by grazing macropodines and geladas constitutes an alternative "blueprint" for facing the adaptive challenges of graminivory. This suggestion is supported by the fact that the bilophodont lower molars of manatees, which are con¬ sumers of sea grasses, are also arranged in a reversed Curve of Spee and also are shed and replaced through molar progression. The many parallels between the diets and dental anato¬ mies of the grazing macropodines and the geladas still leave us with the question of why the dentitions of these distant lineages evolved in such remarkably similar fashions. Both groups represent ungulate-like grazers in non-ungulate or¬ ders and so have evolved their dental specializations in lineages that lack the more established, long-term anatomi¬ cal commitments to grazing seen in perissodactyl and artiodactyl lineages. The basic pattern of bilophodonty seen in the molars of both groups under consideration is itself thought to be a modification that evolved to facilitate the processing of vegetation. Bilophodont molars were well established within the Macropodidae and Cercopithecidae by Middle Miocene times or, possibly, earlier (Benefit 1987; Hume et ai 1989) and thus must be considered as ancient elements of their anatomies. The fact that the evolution of grazing adaptations in the dentitions of both lineages took very similar courses would tend to suggest that the basic template of bilophodonty in both lineages was the product of a rigid developmental pattern that acted to limit or constrain architectural possibilities within the dentitions of both groups Qanis & Fortelius 1988). Such constraints may have seriously limited the potential anatomical responses to selective pres¬ sures in both lineages. As Janis & Fortelius (1988) noted, the evolution of hypsodonty — an increase in the height of the cusps of the teeth — is not a response to the pressure of increased wear that appears to be available to animals with bilophondont dentitions. The relatively few modifications ofthemolars thatappeartohavebeen possible in the animals under consideration were augmented by modifications of the supporting tissues to accentuate the food processing capabilities of the teeth while preserving relatively unworn tooth substance far into adult life. Selective feeding behav¬ iour emerges as critical to the grazing adaptations of both groups, especially T. gelada, because of the lack of potential for generation of new tooth material once the adult ration has been exhausted. In herbivorous mammals, the methods for mechanical breakdown of vegetation must complement those involved in chemical digestion, and both methods arc clearly related to the metabolic rate of the animal under consideration. It is thus appropriate to consider how these parameters compare between the grazing macropodines and Theropithecus. As marsupials, grazing macropodines have lower metabolic rates than most placental mammals (McNab 1980). Their method of foregut fermentation permits them to break down the complex carbohydrates in cell walls, and their bilophodont molars serve the important function of chewing the food very finely in order to increase the surface area for microbial attack. Empirical determinations of basal metabolic rate in 42 Journal of the Royal Society of Western Australia, 77 (2), June 1994 geladas have not been performed, but evidence from the study of comparative brain volumes in cercopithecids sug¬ gests that their basal metabolic rate is low relative to closely related taxa of comparable body size (Martin 1993). Geladas are able to digest more than 50% of the crude fibre in their diet and appear to possess a hindgut microbial fauna for the fermentation of fibre (Iwamoto 1993b). Because of the lim¬ ited capacity of geladas to digest the components of plant cell walls, their bilophodont molars must work to expose the contents of the cells by rupturing, as well as to finely commi¬ nute cell walls in preparation for microbial attack in the hindgut. Geladas are similar to ungulate hindgut fermenters in the ability to physically divide fibrous vegetation in the oral cavity, but they are less efficient at extracting protein (Dunbar & Bose 1991). In the dry season, when the nutri¬ tional valueofgrassesdeclinesbecauseof desiccation, geladas increase their grass intake, but supplement their diet with large volumes of the succulent herb TrifoUujii (Iwamoto 1993b). That geladas do not engage in the same degree of bulk feeding on grasses in the dry season as do ungulate hindgut fermenters reflects their limited ability to extract protein from this source. When estimated metabolic requirements are taken into account, grazing macropodines and geladas appear to have similar compromises in their modes of physical and chemi¬ cal breakdown of vegetation. The macropodines can satisfy their relatively low energy requirements and protein needs through the relatively slow and thorough chemical digestion of very finely divided grasses in the forcgut. Geladas satisfy their higherenergy requirements and protein needs by high- volume intake and fine physical reduction of consistently high-quality vegetation. This is followed by relatively fast "bulk processing" in the gut that involves only partial fer¬ mentation of ingesta in the hindgut. Acktjowledgtnents: I am grateful to Dr. John Long for providing access to collections of macropodine skeletal materials at the Western Australian M useum. This paper benefi ted greatly from discussions with George Chaplin, and from the comments and corrections of three anonymous reviewers and the Honorary Editor. I also wish to thank Martin Thompson for rendering the line drawings. References Benefit B R 1987 The molar morphology, natural history and phylogenetic position of the Middle Miocene monkey Victoriapithecus. PhD disserta¬ tion, New York University. Benefit B R & McCrossin M L 1990 Diet, species diversity and distribution of African fossil baboons. Kroeber Anthropological Society Papers. Nos 71 - 72, 77-93. Davis DD 1949 Comparative anatomy and the evolution of vertebrates. In: Genetics, Paleontolog>' and Evolution (ed G L Jep.sen, G G Simpson & E Mayr) Princeton University Press, Princeton, 64-89. Domning D P & Hayek L-A C 1984 Horizontal tooth replacement in the Amazonian manatee {Trichechus inun^^uis). Mammalia 48:105-127. DunbarRIM 1977 Feeding ecology of geladababooas: a preliminary report. In: PrimateEcologyrStudiesofFeedingRangingin Lemurs, Monkeysand Apes (ed T H Clufton-Brock) Academic Press, London, 251-273. Dunbar RI M & Bose U 1991 Adaptation to grass-eating in gelada baboons. Primates 32:1-7. Griffiths M & Barker R 1966 The plants eaten by sheep and kangaroos grazing together in a paddock in .south-western Queensland. Wildlife Research 11:145-167. Griffiths M, Barker R & MacLean L (1974) Further observations on the plants eaten by kangaroos and sheep grazing together in a paddock in south¬ western Queensland. Australian Wildlife Research 1:27-43. HumeID,jarmanPJ,RenfrecMB&Tcmple-SmithPD 1989 Macropodidae. In: Fauna of Australia. Mammalia (ed D W Walton & B J Richardson) Canberra; Australian Government Publishing Service. Vol IB, 679-715. Iwamoto T 1993a Tl\ceco\o^ of Theropithecus gelada. In: Tlieropithecus: The Rise and Fall of a Primate Genus (ed N G Jablonski) Cambridge University Press, Cambridge, 441-452. Iwamoto T 1993b Food digestion and energetic conditions in Theropithecus gcladu. In: Theropithecus: The Rise and Fall of a Primate Genus (ed N G Jablonski) Cambridge University Press, Cambridge, 453-463. Jablonski N G 1981 Functional analysis of the masticatory apparatus of Theropithecus gelada (Primates: Cercopitheddae). PhD dissertation. Uni¬ versity of Washington. Jablonski NG 1986 Thehand ofTheropithecusbrumpti. In: Primate Evolution (ed P Lee & J Else). Proceedings of the 10th Congress of the International Primalological Sodety, Vol. 1. Cambridge University Press, Cambridge, 173-182. Jablonski N G 1993a Evolution of the masticatory apparatus in Theropithecus. In: Theropithecus: The Rise and Fall of a Primate Genus (edNG Jablonski) Cambridge University Press, Cambridge, 299-329. Jablonski NG 1993b The phylogeny of T/icropif/iccus. In: Theropithecus: The Rise and Fall of a Primate Genus (ed N G Jablonski) Cambridge University Press, Cambridge, 209-224. Janis C1976 The evolutionary strategy of the Equidae, and the origin of rumen and cecal digestion. Evolution 30:757-774. Janis C M & Fortelius M 1988 On the means whereby mammals achieve increased functional durability of their dentitions, with spedal reference to limiting factors. Biological Reviews 63:197-230. Jarman PJ 1984 The dietary ecology of macropod marsupials. Proceedings of the Nutrition Sodety of Australia 9:82-87. Jolly C J 1972 The classification and natural history of Theropithecus (Simopithecus) (Andrews, 1916), baboons of the African Plio-PIeistocene. Bulletin of the British Museum (Natural History), Geology 22:1-123. Koenigswald W, Rensberger J M & Pretzshner H U 1987 Changes in the tooth enamel of early Paleoccne mammals allowing increased diet diversity. Nature 328:150-152. Martin RD 1993 Allometric aspects of skull morphology in T/icmpif/iccws. In: Theropithecus: The Rise and Fall of a Primate Genus (cd N G Jablonski) Cambridge University Press, Cambridge, 273-298. McNab B K 1980 Food habits, energetics and the population biology of mammals. American Naturalist 116:106-124. Meikle W E 1977 Molar wear stages in Theropithecus gelada. Kroeber Anthropological Society Papers. No 50, 21-25. Osborn JW 1993 Orientation of the masseter muscle and the Curve of Spee in relation to crushing forces on the molar teeth of primates. American Journal of Physical Anthropology 92:99-106. Osborn J W & Lumsden A S 1978 An alternative to "Thegosis" and a re¬ examination of the ways in which mammalian molars work. Neues Jahrbuch fur Geologisches und Palaontologisches Abhandlung 156:371- 392. Sanson G D 1978 The evolution and significance of mastication in the Macropodidae. Australian Mammalogy 2:23-28. Sanson G D 1980 The morphology and occlusion of the molariform cheek teeth in some Macropodinae (Macropodidae: Marsupialia). Australian Journal of Zoology 28:341-365. Sanson G D & Miller W A 1979 Mechanism of molar progression in macropods. Anatomical Record 193:674 (abstract). SansonGD, NelsonJE&FellP 1985 Ecology ofPcradorcflscoHcm;w in Arnhem Land in a wet and dry season. In: Ecology of the Wet-Dry Tropics (ed M E Ridpath & L K Corbett). Ecological Sodety of Australia, pp 65-72. StirtonRA 1947 Observations on evolutionary rates in hypsodonty. Evolu¬ tion 1:32-41. Swindler D R & Beynon A D 1993 The development and microstructure of the dentitionofT7jrrop/7/iccMS. In: Theropithecus: TheRiseandFallofaPrimate Genus (ed N G Jablonski) Cambridge University Press, Cambridge, 351- 381. Tyndale-Biscoe H 1973 Life of Marsupials. American Elsevier Publishing Company, New York. 43 - w • y, ^ ^ • H|--'J.>>' -' *>;rZ .;': -^« ‘** ^ ^ . — ^"‘-■’-'-v - ( :■■ /■<• '■■'-'' .''. ,%-i r«iVi/:iv.-•>■¥■ ?-■ .y-r r- :^ '■ _/ .,, * ___ n,v . ii> ^ •^' -i . i. ..^ -.^.•■^15.- __ * -i-- *' r 1 ■ ,...r ■ -s ^ Hij.'vrt. vri fii ii^ ■; / V- ^ '■•• V__ ■»«*yr»' - --.«.. . -V-J__ H ' ' ‘.> :‘ ’ ^ ll.*^ ' '• •» in £^’ t. u« K««‘ —;'■ ' V-' =■ .• ■» ".. . .-fi’j’.i-?'' ■ ~ a S.‘. 'X-- ^ - r. . . : . ■» »i ' ’ vrA^, Mv<»5*' ■., ■■ ' ' u* « «p/. -.'i^* '.* 1^ *y . ■■ r -J.* ,ai ‘i ' !••»* '■ 'f'V V' * ■• ■ •* ? *»'i' ■*■•< ^-wy:,V ' Sft V- -s; -.'Jy / ■ . i4:/'>.- ' . J ■ Vni,» 4 »“i . ' ^^v - A .K^ - ' 5* ^ >-.‘-;'yj .4.r-¥v fii'^’'"-i^»4^. ".s'A'ija »; ^ u** ; 5 . T' ' ._;>'^;:-'jB^^«¥''^ r^-:. \ ^**<5 ^ ^ .i '..f ■' ’ » »• 'r* f r» .. . ;s« Journal of the Royal Society of Western Australia, 77: 45-49,1994 Re-examination of the Murchison Downs meteorite: A fragment of the Dalgaranga mesosiderite? A W R Bevan' & B J Griffin^ ' Department of Earth and Planetary Sciences, Western Australian Museum Francis Steet, Perth, WA 6000 ^Centre for Microscopy and Microanalysis, University of Western Australia, Nedlands WA 6009 Manuscript received March 1994; accepted June 1994 Abstract The Murchison Downs mesosiderite was reportedly recovered in 1925 from a locality ca. 200 km to the NE of the crater- forming Dalgaranga mesosiderite in Western Australia. A comparison of data from the literature on the chemistry and mineralogy of Murchison Downs and Dalgaranga, and a re-investigation of the metallography and mineralogy of Murchison Downs and Dalgaranga, suggests strongly that the two meteorites belong to the same fall. Murchison Downs may be one of the few examples of a meteorite transported by Aborigines and, pending further work, should be paired with the Dalgaranga meteorite. Introduction Meteorites have been recovered throughout Western Australia over the last century. Currently, specimens from 141 distinct meteorites have been documented from the State, representing more than 50% of all meteorites known from Australia (Sevan 1992). Conventionally, meteorites take the name of the geographical locality where they fell or were found. 'Paired' meteorites are those suggested, because of geographical propinquity and classification, to belong to a single fall (Hey 1966). However, when two or more mete¬ orites were found at different times and allocated different names, but were subsequently proved conclusively to be from the same fall or find, then they are said to be 'synony¬ mous' and the name of the meteorite first recovered usually takes precedence. Conversely, meteorites thought to be from the same fall are sometimes found on further examination to be distinct. For these reasons, the number of distinct mete¬ orites known from Western Australia has fluctuated in the past without necessarily any addition of new material. The Murchison Downs meteorite (Western Australian Museum registration number WAM12586), a small metallic slug weighing 33.5 grams (Figure l),was found in 1925 and described briefly by Simpson (1927), who noted that an etched surface of the meteorite displayed a Widmanstatten pattern and classified the meteorite as an iron with a fine octahedral (Of) structure (e.^. see Graham et al. 1985). Re¬ cently, Wasson et al. (1989) have provided a modem analysis of the Murchison Downs meteorite showing that it is a metallic nodule from a mesosiderite. In this paper, metallographic, mineralogical and chemical data are pre- © Royal Society of Western Australia 1994 sented suggesting that the Murchison Downs meteorite is probably a transported fragment of the Dalgaranga mesosiderite. Figure 1. The Murchison Downs meteorite (Western Australian Museum number 12586). Recovery and historical details Few details of the discovery of the Murchison Downs meteorite are known. Neither the exact locality, nor the name of the finder was recorded. In the Annual Report of the Geological Survey of Western Australia for 1925, Gibb Maitland records the find-site as Murchison Downs Station 45 Journal of the Royal Society of Western Australia, 77 (2), June 1994 in stmpson (1927) records the co-ordinates of the find-site as approximately 26® 40'S, 119° O'E which correspond to a site close to'North Cattle Well', situated cfl. 15 km north of the homestead on 'Murchison Downs' Station and approxi¬ mately 200 km to the north-east of the Dalgaranga crater. Since Simpson's (1927) observations were made, no detailed metallographic description of the meteorite has been pub¬ lished. However, McCall and de Laeler (1965) noted that the Murchison Downs meteorite "was similar to the small twisted irons commonly found near meteorite craters." In contrast, the discovery of the Dalgaranga crater and recovery of its associated meteorites are well documented. This small crater, measuring 25 m in diameter, was discov¬ ered in 1923 by G E P Wellard at co-ordinates 27° 43'S, 117° 15'E on Dalgaranga Station, north of Yalgoo (Simpson 1938). Wellard is reported to have recovered a large number of meteorite fragments from the vicinity of the crater but the repository of this material is unknown. Simpson (1938) described a metallic fragment of the Dalgaranga meteorite weighing 42 grams, which he classified as a medium (Om) octahedrite. In 1959, and again in 1960, H HNiningerand GI Huss of the American Meteorite Laboratory visited the Dalgaranga crater and from the surrounding plain collected 207 speci¬ mens with an aggregate weight of 1098 grams. In addirion. the Ky arra district of the Murchison Division and notes the -or' of the specimen (GSWA 1/3894) as "Richardson". they recovered 280 specimens, weighing approximately 9.1 kg, of deeply weathered material buried beneath the crater floor (Nininger & Huss 1960). Most of the specimens from the plain around the crater weighed individually less than 5 grams, and the largest weighed 57 grams. The material comprises both metallic and achondritic stony portions and the classification of the meteorite as a mesosiderite by Nininger & Huss (1960) was later confirmed by McCall (1965). Metallographic, mineralogical and chemical details A metallographic examination (this work) of an etched section (1.5 x 1.5 cm) of the Murchison Downs meteorite (WAM 12586—formerly Geological Survey of Western Australia Collection 1/3894) reveals that this meteorite consists predominantly of plates of a-Fe,Ni (kamacite), y- Fe,Ni (taenite), Y'-Fe,Ni (tetratacnite) and a+YFe,Ni (plessite) in octahedral arrangement. On two perpendicular sections, the bandwidths of plates of kamacite (excluding those in plessite) vary from 0,3-0.5 mm with a meanof0.45±0.1 mm. This kamacite bandwidth lies within the 'fine octahedrite' group of the modem structural classification of iron mete¬ orites (Buchwald 1975). Troilite and schreibersite occur as irregular inclusions, and several small (mm-sized) inclu¬ sions of non-metallic minerals composed of low-Ca orthopyroxene (Fs 3 ^), anorthitic plagioclase feldspar (An^^), Table 1 Electron microprobe analyses of non-metallic minerals in the Murchison Downs and Dalgaranga meteorites. Analysts: B J Griffin and G D Pooley. - indicates not detected Murchison Downs Dalgaranga orthopyroxene plagioclase chromite chromite range* SiO^ 52.7 46.2 - - TiOj 0.32 - 1.20 1.32 0.18-1.32 Aip, 0.69 34.3 10.8 13.5 13.0-14.3 CrA - - 54.6 52.0 51.3-52.41 VA - - 0.54 0.52 0.47-0.62 FeO'> 21.1 - 28.6 28.2 27.0-28.2 MnO 0.83 - • 1.74 1.60 1.50-1.88 MgO 22.6 - 2.70 3.54 3.42-4.19 CaO 1.80 18.4 - - Na^O - 0.93 - - K,0 - 0.00 - - Totals 100.04 99.83 100.18 100.68 Molecular % FS34 ^^91.6 ulvospinel 3.1 3.3 £"62.4 AW. spinel 21.9 26.8 WO3, 0 b chromite 74.3 69.1 magnetite 0.7 0.7 100Cr/(Cr+Al) 77.2 72.0 100Fe/(Fe+Mg) 85.6 81.7 ^range based on six analyses. ^All Fe reported as FeO. 46 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Al-rich chromite and a silica polymorph occur throughout the section. These inclusions are swathed with bands of kamacite up to 1.0 mm thick. Electron microprobe analyses of the non-metallic minerals in the Murchison Downs me¬ teorite are given in Table 1. The Widmanstatten pattern displayed by the Murchison Downs meteorite is not continuous; in addition to swathing kamacite around silicate inclusions, in the section exam¬ ined a thick (1.00 mm) band of swathing kamacite partly bounds the external surface of the meteorite. Locally, the Widmanstatten structure displays moderate to severe me¬ chanical deformation and kamacite and taenite plates are bent and kneaded. Kamacite is shock-hardened, displaying the cross-hatched e-kamacite structure and abundant Neumann bands. Narrow zones of shear deformation occur in the metallic micro-structure of the meteorite along which fine scale (<1 pm) recrystallization has taken place. Under crossed polars, troilite di.splays abundant shock-twins and, where inclusions are traversed by shear zones, troilite has been recrystallized. Table 2 Summary of the mineralogy of the Murchison Downs and Dalgaranga mesosiderites (mineral compositions determined by electron microprobe analyser unless otherwise stated - indicates not recorded) Murchison Downs Dalgaranga Wasson et al. (1989) This work Hassanzadeh et al (1990) McCall (1965) Nehru et al (1980) This work silicates olivine - - - Fajj orthopyroxene Fs„Eiv,WOj FSj^En^^WOj, present Fs,,^ - plagioclasc An,„Ab,, anorthite AnTjAbjg - silica polymorph present present - ■ present metallic minerals kamacite (mm bandwidth) (0.5) (0.45±0.1) Fe,3,Ni5,Co„53 (0.3±0.1) - present taenite present present present - present tetrataenite present present ^*^48.3^'51.6^®0 09 - present other minerals troilite present present present present present chromite present 'Jsp,,Spj,,Chr3,3Mt„, present (Al-rich) present Usp„Sp3„Chr„,Mt„ schreibersite present present present - present cohenite * present - - present * determined by X-ray diffraction; ^ determined optically. The Dalgaranga meteorite has been described by Nininger & Huss (1960) and McCall (1965). The material consists of disrupted fragments of mesosiderite; individual specimens ranging from nodules formed almost entirely of metal, through mixtures of metal and silicate, to essentially basaltic achondritic material. Nininger & Huss (1960) noted that those fragments composed mainly of metal are polycrystalline and display Widmanstatten patterns rang¬ ing from coarsest (Ogg) to finest (Off) octahedrite. The structures of many of these metallic slugs show extensive gross mechanical deformation and localised thermal altera¬ tion of the type often encountered in crater-forming iron meteorites. Other fragments show few signs of the effects of impact shock-metamorphism (Nininger & Huss 1960). Hassanzadeh et al. (1990) have described a metallic slug from the Dalgaranga meteorite and note the presence of clumps of silicates that comprise low-Ca pyroxene, anorthite, accessory troilite, schreibersite and an Al-rich chromite. Published mineralogical data for both Murchison Downs and Dalgaranga meteorites (and those determined in this work) are summarised in Table 2. Simpson (1938), Wasson Table 3 Summary of published analyses of metal in the Murchison Downs and Dalgaranga mesosiderites Ni (%) Co (%) Ga f^g/g Ge l^g/g Ir l*g/g Cr lig/g Cu l^g/g As |4g/g Sb ng/g W iig/g Re ng/g Ft Fg/g Au Fg/g Murchison Downs' 9.16 0.49 13.6 56.1 4.98 221 144 12.5 310 1110 600 8.9 1.32 Dalgaranga ^ 10.27 0.48 12.7 - 4.99 12 172 11.9 260 990 600 8.0 1.37 Dalgaranga ^ 8.8 - 15.5 56.0 4.2 - - - - - - - - Dalgaranga '* 8.63 - - - - - - - - - - - - ^ Wasson et al. (1989); ^ Hassanzadeh et al. (1990); ^ Wasson et al. (1974); ^ Simpson (1938) 47 Journal of the Royal Society of Western Australia, 77 (2), June 1994 et al (1974) and Hassanzadeh etal. (1990) have analysed the metal in the Dalgaranga meteorite, and their data, compared with those of Wasson et al (1989) for the Murchison Downs meteorite, are shown in Table 3 Discussion Excluding olivine, the silicate mineralogy of both the Dalgaranga and Murchison Downs meteorites consists essentially of low-Ca orthopyroxene and anorthite. Acces¬ sory minerals in both meteorites include troilite, schreibersite and an Al-rich chromite. In the Murchison Downs meteor¬ ite, Wasson et al. (1989) reported a fine-grained silica polymorph, probably tridymite, which is confirmed in this work. In the Dalgaranga meteorite, olivine with the range of compositions (Nehru et al. 1980), generally occurs as nodules and phenocrysts in the stony portions of the meteorite (McCall 1965). Olivine has not been found in the Murchison Downs meteorite. However, in mesosiderites, olivine is only rarely associated with metallic nodules. Additionally, tridymite has yet to be reported from the Dalgaranga meteorite, although Hassanzadeh et al. (1990) note that it is a sub-group'A' mesosiderite that are known to be tridymite-rich (Hewins 1984). Nevertheless, the min¬ eral compositions of those silicates in the Dalgaranga and Murchison Downs meteorites that have been analysed by modern methods are generally very similar (Table 1). The composition of the pyroxene (Fs 3 ^) reported in this work for the Murchison Downs meteorite is identical to that reported for the Dalgaranga meteorite by McCall (1965). While this differs from the pyroxene composition (Fs,^) reported for the Murchison Downs meteorite by Hassanzadeh et al. (1990), Powell (1971) has shown that pyroxene compositions within individual mesosiderites can be highly variable, though rarely falling outside the range Fs.,^^,. The compositions of grains of plagioclase can also be variable ranging from An^^^, to An^^. McCall (1965) reports an optical determination of plagioclase in the Dalgaranga meteorite with the composi¬ tion An_^Ab 2 g (bytownite) that is very different from ana¬ lysed plagioclase reported for the Murchison Downs mete¬ orite (Wasson efrt/. 1989; see Table 2), but is also different from the 'anorthite' reported for the Dalgaranga meteorite by Hassandazch et al. (1990). In contrast to silicates, Powell (1971) and Bunch &Keil (1971) noted that there is generally little compositional vari¬ ability in chromite for a given mesosiderite. Major oxides in chromite rarely vary by more than 13% of the amounts present, with TiO, being the most variable component. Vari¬ ations in the composition of chromite (notably Al and Mg) between the Murchison Downs and Dalgaranga meteorites are greater than one would normally expect from a single meteorite. However, chromite compositions within frag¬ ments of the Dalgaranga meteorite are similarly variable. One grain of chromite associated with a serpentine-group mineral in a weathered portion of the Dalgaranga meteorite showed extremeTiOj depletion and MgO and Al^O^ enrich¬ ment that may be attributed to alteration during severe terrestrial weathering. Simpson (1927) noted one small grain of cohenite in the Murchison Downs meteorite which is confirmed in this work. Minor amounts of cohenite also occur in the Dalgaranga meteorite. The reported bulk Ni contents of metal in the Dalgaranga meteorite (Table 3) vary from 8.63 %wt (Simpson 1938) to 10.27 %wt (Hassanzadeh et al. 1990) consistent with the observed structural heterogeneity of the metallic portions of the meteorite. With the exception of Cr, there is a very close correspondencebetween the major, minor and trace element contents of metal in the Dalgaranga and Murchison Downs meteorites (Table 3).TheCr content of metal in theMurchison Downs meteorite (221 pg/g) reported by Wasson efrt/. (1989) is 18.5 times greater than that reported by Hassanzadeh et al. (1990) for the Dalgaranga meteorite (12 pg/ g). This is out¬ side the usual variation in replicate analyses of the same meteorite and could indicate that the two meteorites are distinct. However, it is possible that the high Cr content of the Murchison Downs meteorite reported by Was.son et al. (1989) is due to the presence of microscopic inclu.sions of chromite in the small sample of metal (off WAM 12586) that was analysed 0 T Wasson, pers. comtu.). On the basis of cluster analysis for a number of elements, notably Ni and Au, in metallic nodules from twelve mesosiderites, Hassanzadeh et al. (1990) recognised various sub-groups of mesosiderites. Eight closely related clusters are recognised, and three moderately related clusters are designated low-AuNi, high-AuNi and intermediate-AuNi. Significantly, the Murchison Downs and Dalgaranga mete¬ orites, along with the South Australian mesosiderite, Pinnaroo, form one closely related cluster belonging to the high-AuNi sub-group of mesosiderites (Hassanzadeh et al. 1990). Structurally, the thick band of swathing kamacite that bounds a portion of the exterior surface of the Murchison Downs meteorite is typical of the heterogenous nucleation of this mineral encountered in the metallic nodules from mesosiderites, and indicates that the metal originally formed in contact with either silicates, or some other non-metallic phase. The deformed and locally heat-altered nature of the metallographic structure of Murchison Downs is character¬ istic of meteorites that have been involved in a crater¬ forming impact. The style and extent of thermo-mechanical alteration displayed by the Murchison Downs meteorite is identical to that in many of the metallic slugs of the Dalgaranga meteorite. To dale, no young meteorite impact crater other than the Dalgaranga crater has been identified in the immediate vicinity of Murchison Downs Station or elsewhere in the Murchison Division of Western Australia. Summary and Conclusions The data presented in this paper confirm the re-classifica- tion by Wasson et al. (1989) of the Murchison Downs meteor¬ ite as a mesosiderite but are insufficient to prove, conclu¬ sively, that it is a fragment of the Dalgaranga mesosiderite. Notwithstanding, there is little evidence to suggest that they arc from different falls. The distance between the find-sites of Dalgaranga and Murchison Downs {c«. 200 km) would normally preclude pairing. Mesosideritesare extremely rare and account for less than 1% of all known meteorites. Re¬ markably, two other mesosiderites (Mount Padbury and Pennyweight) have also been found in the same general area as the Dalgaranga and Murchison Downs meteorites. The suggestion by Mason & Jarosewich (1973) that the Mount 48 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Padbury and Dalgaranga meteorites may be fragments of the same meteorite has been shown by Wasson et al (1974) to be highly unlikely. Also, Hassandazeh et al. (1990) show that the Pennyweight and Murchison Downs mesosiderites belong to the low- and high-AuNi sub-groups, respectively, and are probably distinct. However, the discovery in the same general area of two mesosiderites (Dalgaranga and Murchison Downs) that have been involved in crater-form¬ ing impact militates against their being from separate falls. There are a number of explanations as to how the Murchison Downs fragment could have became displaced from the vicinity of the Dalgaranga crater. One possibility is that the Murchison Downs fragment became detached dur¬ ing atmospheric passage of the impacting projectile. How¬ ever, if this was the case, then the meteorite would not show evidence of damage due to large scale, explosive impact. Moreover, the small size of the Dalgaranga crater (25 m in diameter) makes it very unlikely that the fragment could have been thrown 200 km by the impact event. The most likely explanation is that the fragment was transported by human agency. This could have occurred at any time prior to the discovery of the Murchison Downs meteorite in 1925. The age of the Dalgaranga crater is variably reported from 3000 years (Shoemaker & Shoemaker 1988) to around 25,000 years (Nininger & Huss 1960; Grieve 1991). These ages lie well within the accepted time of Aboriginal occupation of Australia (40,000 years) and it is possible that the impact of the Dalgaranga meteorite was witnessed by Aborigines. Although Aborigines were not known to have collected or utilized meteoritic iron, it is po.ssible that the Murchison Downs meteorite is one of the few examples that have been transported by Aborigines. Pending further work, it is sug¬ gested that Murchison Downs be paired with Dalgaranga. Acknozvledgements: The authors thank Greg Pooley for his assistance with electron microprobe analysis, and for commenting on an earlier version of the manuscript. The authors also thank J T Wasson and anonymous reviewer for their suggested improvements to the manuscript. References Sevan A W R1992 Australian meteorites. Records of the Australian Museum, Supplement 15:1-27. Buchwald V F 1975 Handbook of Iron Meteorites. University of California Press, Los Angeles. Bunch, T E & Keil K1971 Chromite and ilmenite innon-chondritic meteorites. American Mineralogist 56:146-157. Graham A L, Bevan A W R & Hutchison R1985 Catalogue of Meteorites (with special reference to those represented in the collection of the British Museum (Natural History]). edition. British Museum [Natural History] and University of Arizona Press. Grieve R A F 1991 Terrestrial impact: The record in the rocks. Meteoritics 26:175-194. Hassanzadeh J, Rubin A E & Wasson J T 1990 Compositions of large metal nodules in mesosiderites: Links to iron meteorite group IIIAB and the origin of mesosiderite subgroups. Geochimica et Cosmochimica Acta 54:3197-3208. Hewins R H1984The case for a melt matrix in plagioclase-POIK mesosiderites. Proceedings of the 15'^ Lunar Planetary Science Conference C289-C297. Hey M H 1966 Catalogue of Meteorites (with special reference to those represented in the collection of the British Museum [Natural History]). 3"* edition. British Museum [Natural History). McCall G J H1965 New material from, and reconstruction of, the Dalgaranga meteorite and crater. Western Australia. Mineralogical Magazine 37:476- 487. McCall G J H & de Laeter J R1965 Catalogue of Western Australian Meteorite Collections. Western Australian Museum Special Publication number 3. Mason B & Jarosewich E1973The Barea, Dyarrl Island, and Emery mesosiderites and a review of the mesosiderites. Mineralogical Magazine 39:204-215. Nehru C E, Zucker S M, Harlow G E & Prinz M 1980 Olivines and olivine coronas in mesosiderites. Geochimica et Cosmochimica Acta 44:1103- 1118. Nininger H H & Huss G I 1960 The unique meteorite crater at Dalgaranga, Western Australia. Mineralogical Magazine 32:619. Powell B N1971 Petrology and chemistry of mesosiderites-II. Silicate textures and compositions and metal-silicate relationships. Geochimica et Cosmochimica Acta 35:5-34. Shoemaker E M & Shoemaker C S1988 Impact structures of Australia. Lunar and Planetary Science Conference XIX, 1079-1080 (abstract). Simpson E S 1927 Contributions to the mineralogy of Western Australia. Journal of the Royal Society of Western Australia 13:37-48. Simpson E S 1938 Some new and little-known meteorites found in Western Australia. Mineralogical Magazine 25:157-171. Wasson j T, Schaudy R, Bild R W & Chou C L 1974 Mesosiderites -I. Compositions of their metallic portions and possible relationship to other metal-rich meteorite groups. Geochimica et Cosmochimica Acta 38:135- 149. Wasson J T, Xin wei Ouyang, Jianmin Wang & Jerdc E1989 Chemical classifi¬ cation of iron meteorites:XI. Multi-element studies of 38 new irons and the high abundance of ungrouped irons from Antarctica. Geochimica et Cosmochimica Acta 53:735-744. 49 ■■ .' •■' . r ;• ••■■. > .- ' i il- ^ '■ „ > ■» ■•• ■ ■> ■■ ' ■-'•f'r; -^^•.-'t ?>■ •',V J ,. ■ i- * ^ ' I?.’***' ■'*^ 'k,*^*#v’' •** ' »* * Ml ' *» • • >j ;V’ >- •" i,_i^jr^:),;-i.ij;^ -3Hm .' - ■,* ‘ !..-«,-^jpi»»,'*»K';fc,'*v •f-7jm-^^^ t ' ,,,• .;^-r''*‘*<- ■ •:;»^’««r:r v‘? '•■'v®\ -'.V'UV* *4> V , »■*. ,'.«'*^-olfliel ' ':^*!S ■ -■- • . ^ f J . •-■.“’•vr, , -.Y^ 5-- % :'**•*: . - .f- i'f ,^1* “ ’ I’ ■« . t'”*? ‘•‘•-i*'. .-■•**>»■! 1 ■% vsl *4"' ' *- V ■^' v^.*-V %♦’■• ■'.. • A •• ■•*'•■ . S ■•♦* . - ■ * ■ fa - t.' ••AsT'.’^ *• ' , ‘ >-.i - - ''■a. .rj. • t fS i" ** . . >- . , .Vt.^ I - .,4^' * f -M *1 - t*' ■» .-' -,.-^ /-'.Gifts ' •i*-^V,«■' ■ -y ,N - ■■ '.^ ' ^ ■< ‘ . ttJ ,»■'- » V N' > > -V- e;.» ,fi>> . . -d&m ... V. •■> r ^ I *» *" * ~ iT ‘* ' ■ '^ Journal of the Royal Society of Western Australia, 77: 51-63,1994 Invertebrate community structure related to physico-chemical parameters of permanent lakes of the south coast of Western Australia D H D Edward, P Gazey & P M Davies Aquatic Research Laboratory, Department of Zoology, The University of Western Australia, Nedlands WA 6009 Manuscript received December 1993; accepted June 1994 Abstract The aquatic invertebrate fauna and a range of physical and chemical parameters were recorded in twenty- three permanent lakes within 20 km of the coast between Cape Naturaliste and Albany, Western Australia. Invertebrates were collected by qualitative sweeps, benthic cores and plankton trawls to sample all major habitats. A total of 209 invertebrate taxa were recorded, representing a rich faunal diversity. Multivariate analyses showed that invertebrate community structure was most closely associated with salinity and nutrient status of the lakes; however, all salinities were below the limnologically accepted 3 g 1’ upper limit for freshwater (Bayly & Williams 1973). Human activities are most likely responsible for the elevated nutrient levels recorded in some of the lakes. Introduction Lentic wetlands in the south-west of Australia, particu¬ larly lakes on the Swan Coastal Plain associated with urban development, have recently been the focus of much research. Wetlands on the Swan Coastal Plain today represent only about 30% of wetlands present prior to European settlement (Halse 1989) and many are now eutrophic due to urbaniza¬ tion and agriculture (Davis & Rolls 1987; Balia & Davis 1993; Davis etal. 1993). In contrast, the permanent lakes along the south coast are less disturbed and have received littleattention. Limnological studies on south coast lentic wetlands have mainly been restricted to temporary systems (Bayly 1982, 1992a; Christensen 1982; Pusey & Edward 1990a,b) although two permanent pools were included in the latter study. Recently, a survey of aquatic invertebrates in three lakes in the Two Peoples Bay area (Storey et al 1993) outlined a highly diverse invertebrate fauna, with possible biogeographic links to South-western Tasmania. Establishing a database of physi¬ cal, chemical and biological parameters in south-coast 'vetlands is essential for developing management proce- t^ures to deal with the impact of increased human activity, ^his study elucidates relationships between environmental parameters and invertebrate community structure for the lakes surveyed, and provides a database to assess the conser¬ vation value of the lakes to enable formulation of future Management programmes. Materials and Methods Study sites Twenty-three permanent lakes in the south of Western ^^stralia were studied, all located on Vacant Crown Land 'vithin 20 km of the coast between Cape Naturaliste and ® Royal Society of Western Australia 1994 Albany (Fig 1). Several of the lakes were not officially named and are therefore referred to as follows; the lake near the junction of Charley and Dunes Roads, Pemberton is Charley Lake; the lake north east of Windy Harbour is Windy Har¬ bour Lake; for the group of lakes near Boat Harbour Road, Denmark, the eastern lake is Boat Harbour Lake 1, the western lake is Boat Harbour Lake 3 and the northern lake is Boat Harbour Lake 4. Sampling regime Fifteen lakes were sampled in 1991 during winter (25 June-2 July) and six of these, located across the geographical range of the survey, were re-sampled in spring (4-13 No¬ vember), together with six additional lakes. Charley Lake and Lake Williams were sampled in early summer (18 & 19 December) because of forest quarantine restrictions; these were considered spring samples because typical hot/dry summer conditions had not commenced. Environmental parameters The physical and chemical parameters measured in the lakes, and the methods used are summarised in Table 1. Surface area of the lakes was estimated from enlarged pho¬ tocopies of 1:50000 maps, using a Delta-T^’^^ Area Meter. Temperature, dissolved oxygen and pH were measured from surface waters at each lake between 1030 and 1430 h. Depth was recorded along at least two transects for those lakes accessible to a boat. At these lakes, vertical profiles of dissolved oxygen and temperature were recorded to deter¬ mine the extent of any stratification. The ratio of surface to bottom readings for temperature and dissolved oxygen was used as a measure of stratification in subsequent statistical analyses, where a ratio of one indicated no stratification. Undisturbed water samples were taken for analyses for colour, turbidity, anions and cations. Salinity was calculated as the total concentration of the major cations (Mg, Ca, K and Na) and chloride in solution. These ions account for approxi¬ mately 95% of the total soluble salts (TSS) for drainage basins 51 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Figure 1. Map of locations of the lakes. 1. QuininupLake 2. Lake Davies 3. Lake Quitjup 4. Lake Jasper 5. Lake Wilson 6. Lake Smith 7. Charley Lake 8. Ycagarup Lake 9. Neanup Lake 10. South Yeagarup Lake 11. DoggerupLake 12. Lake Samuel 13. Lake Florence 14. Windy Harbour Lake 15. LakeMaringup 16. CHvingup Lake 17. Boat Harbour Lake 1 18. Boat Harbour Lake 3 19. Boat Harbour Lake 4 20. Lake 12046 21. Lake Williams 22. LakeSaide 23. Lake Powell 3345’ 48" S 115 00' 06" E 3413‘ 20" S 115 01’ 58" E 34 22’ 58" S 11535’ 40" E 34 24’ 40" S 115 40' 59" E 34 25'39" S 115 43' 00" E 34 25’45" S 115 43’ 26" E 34 30' 21" S 115 49’ 28" E 34 32’43" S 115 52'14" E 34 32‘ 47" S 115 51’32" E 34 33' OS'S 115 5r58’'E 34 43’03" S 116 03' 36" E 34 43'55"S 116 03' 26" E 34 44’ 03” S 116 05' 57" E 34 50’ 00" S 116 02’ 25" E 34 50’ 00" S 11611’55" E 3459' 56" S 117 03' 57" E 35 or or S 117 05' 59" E 35 Ol’OVS 11705'17" E 3500’ 53" S 11705’ 46" E 34 59’ 56" S 11713' 23" E 35 0V00"S 11716’03"E 35 02' 34" S 11728' 24" E 35 01'14’’S 117 44'16" E on the south coast of Western Australia (Loh et ai 1983). Water samples for total nitrogen and phosphorus determinations were filtered in the field through a 0.22 pm Millipore™ filter using a 50 ml syringe. Samples for chloro¬ phyll (a) determinations were taken by filtering a measured volume (approx. 11) of water through a Whatman'^^ GF/C filter, and the retained cells, containing chlorophyll, were stabilized with a few drops of a saturated magnesium car¬ bonate solution. The filter was folded, blotted between gauze swabs and placed in a plastic bag on ice, out of the light, before being stored frozen in the laboratory. Chloro¬ phyll (a) was then measured using the method described in Strickland & Parsons (1968). At each lake, a 10 cm deep core¬ sample of the benthic material was collected in a vertical¬ sided vial with 13.2 cm^ lid opening. In the laboratory, the percent organic content was determined after drying at 40 °C and ashing in a muffle furnace at 450 °C for approximately eight hours. Invertebrates The methodologies for collecting fauna were designed to sample the major aquatic habitats and maximise the number of species recorded from each lake. Six random replicate benthic samples were taken with a 72 cm^ core sampler to 10 cm depth from each lake. Samples were immediately pre¬ served in 5% formalin. In the laboratory, the organic fraction was separated from the sediment by water elutriation and washed through a 250 pm sieve. The fauna was removed from the organic fraction using a dissecting microscope. All individuals were identified to the lowest taxon possible, usually species, either by the use of keys or by matching specimens to a voucher collection at the Aquatic Research Laboratory, Department of Zoology, The University of West¬ ern Australia. The Calanoida, Ostracoda and Cladocera Table 1 Environmental parameters, methods of measurement and units of precision. Parameter Method/Apparatus Precision Acronym Lake surface area Estimated from 1:50000 map SA Temperature Mercury thermometer & Yeo-Kal 602 Hamon 0.5'‘C Temp salinity / temperature bridge 0.5"C Dissolved oxygen Nester portable meter 0.1 mg L* DO pH Kane-May KM 7001 portable pH meter 0.1 pH unit pH Depth Graduated line 0.05 m Depth Colour UV-VIS spectrophotometer 5 APHA units Col Turbidity Nephelometric method 0.1 NTU Turb Anions & cations Atomic absorption spectrophotometer 0.4 mg 1 ‘ chemical symbols Salinity Sum of anion & cation concentrations Sal Total nitrogen Auto analyser 0.01 mg 1-1 N Total phosphorus Auto analyser o o 3 P Chlorophyll (a) Spectrophotometer 0.01 pg V Chloro Benthic organic matter Gravimetric method 0.01% BOM 52 Journal of the Royal Society of Western Australia, 77 (2), June 1994 were forwarded to specialist taxonomists for identification. We include within Calamoecia tasmanica (Smith) $.1. the two forms C. tasmanica and C. tasmanica subatteruiata described from Western Australia (Bayly 1992b). An estimate of abun¬ dance for each species was made according to the following categories; < 50 (rare), 50-500 (common) and > 500 (abun¬ dant) individuals. A standard Freshwater Biological Association (FBA) D- net with 110 ^im mesh was used to collect two-minute qualitative sweep samples from the substrate and amongst the macrophytes from the littoral margin of each lake. Sam¬ ples were preserved in 5% formalin. In the laboratory, the fauna was removed from the organic matter in the samples, identified and categorised as for the benthic samples. Plankton was sampled with a 110 |xm mesh net, attached to a standard FBA D-net frame, held just below the water surface. Two samples were taken from each lake, using a standardised 50 m trawl for biomass determination and a shorter trawl for identification of the plankton species. Each sample was preserved in 5% formalin. Biomass was deter¬ mined after drying the samples to constant weight at 40 The taxa were identified and abundance categories esti¬ mated as for the benthic samples. Data analyses Multivariate analyses are useful techniques to outline patterns in complex biological phenomena (Gauch 1982) and these patterns are generally correlated to underlying envi¬ ronmental gradients (Wright et al. 1984; Furse et al. 1984; Moss et al. 1987). Principal components analysis (PCA) was used to ordinate the pattern of co-occurrence of the physical, chemical and morphological parameters measured for the lakes. This technique was used to reduce the complex data¬ set to a few, underlying 'factors' and to eliminate redundan¬ cies inherent in the data (Noruss 1986). This analysis was performed using the SPSS/PC^ advanced statistics proce¬ dure 'FACTOR'. The total species information from each lake was classi¬ fied by a polythetic divisive multivariate technique (Two- Way INdicator SPecies ANalysis; TWINSPAN, Hill 1979a). Subsequent groupings formed by TWINSPAN were corre¬ lated to environmental parameters by Multiple Discrimi¬ nant Analysis (MDA; Noruss 1986) using the SPSS/PC^ version DSCRIMINANT. This analysis was performed be¬ tween each TWINSPAN division. DEtrended CORrespond- ence ANAlysis (DECORANA, Hill 1979b) was used to ordi¬ nate the lakes on the basis of invertebrate community struc¬ ture. DECORANA orders samples along an axis of similar¬ ity, where the lakes closest together on each axis have a more similar invertebrate community structure than those further apart. Results Environmental parameters The 23 lakes studied ranged in surface area from 0.13 ha (Quininup Lake) to over 400 ha (Lake Jasper), with the majority of the lakes less than 100 ha in surface area (Table 2). The deepest lakes were Yeagarup and Jasper, both over 10 m (Table 2). Most of the lakes with larger surface area were shallow, ranging between about 0.8 m and about 1.3 m for Lake Powell and Owingup Swamp respectively. Water colour was highly variable between lakes, ranging from < 5 APHA units at Quininup Lake to 740 APHA units at Lake Williams. The lakes were visually classified in this study as clear, brown and black corresponding to three APHA unit categories where clear is < 100, brown 100-300 and black > 300 (Table 3). The pH ranged from 4.4 at Lake Florence to 8.6 at Lake Davies (Table 2). The values for pH were highly negatively correlated with colour (correlation coefficient r = -0.83, n = 29, p < 0.001). The low pH values probably reflected high humic / tannic content of darker waters (Table 3). The major¬ ity of these dark-water lakes lie on acid peat flats, a source of humic material. Lake colour was also correlated with water temperature (r = 0.51, n = 29, p < 0.01), probably due to dark water absorbing more solar radiation than clear water. Water temperatures between lakes ranged from 9.2-15°C in winter to 12.3-24°C in spring (Table 2). Temperature stratification (Table 2), with colder hypolimnetic water, which is the normal condition for deep water lakes, was recorded in Charley and Yeagarup Lakes in spring. In contrast, lakes Maringup and Jasper were stratified in winter, with warmer hypolimnetic water, which was interpreted as heating of the bottom sediments and hypolimnetic water by solar radiation in these dear-water lakes. Salinity categories for the lakes were based on the classi¬ fication for potable surface water by the Water Authority of Western Australia (1989a). The majority of the lakes were fresh; however Boat Harbour Lake 3, Lake Saide and Lake Powell were classified as marginal and Lake Davies and Owingup Swamp as brackish (Table 3). On the basis of nutrient status (Wetzel 1975) Windy Harbour, Williams, Saide and Powell, were classified as eutrophic (Table 3). The other lakes were either oligo- mesotrophic or meso-eutrophic. Lakes sampled in winter usually had higher levels of phosphorus, probably from run¬ off of winter rainfall, elevating them into the meso-eutrophic category. The typical ratio of total nitrogen to total phospho¬ rus (N:P) within the tissue of aquatic algae and macrophytes is 7:1 (Wetzel 1975). In Quininup Lake, Lake Smith, Neanup Swamp, and Boat Harbour Lake 3, theN:P ratios in the water were greater than 70.T and therefore phosphorus was more likely the limiting nutrient. In contrast, for Lake Williams and, in winter. Lake Powell, the ratio was less than 4:1, indicating that nitrogen was the more probable limiting nutrient. Turbidity measured in nephelometric turbidity units (NTU) ranged widely (Table 3), showing no obvious associa¬ tion with individual lakes. Turbidity and chlorophyll (a) were correlated (r = 0.61, n = 29, p < 0.01); high turbidity probably reflected the abundance of algal cells. Values for chlorophyll (a) ranged from 0.19 pg I ' in Yeagarup in late spring to 14.41 pg in Lake Saide in winter (Table 3). Lower values for chlorophyll (a) were consistently recorded during spring. Based on the classification of Wetzel (1975) using chlorophyll (a) levels, the trophic status of most of the lakes was classified as oligotrophic and only lakes Powell, Saide, 12046, Boat Harbour 1, Boat Harbour 4, Windy Harbour and Owingup Swamp were classified as meso-eutrophic. Benthic organic matter ranged from 0.16% in Lake Wilson to 87% in Boat Harbour Lake 1 (Table 3). The wide range reflects both differences in the nature of the catchments and 53 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Table 2 Environmental information collected for each lake/occasion. W = winter, S = spring, ** = not recorded. Lake Season SA (km^) Max. depth (m) Temp surface rc) Temp ratio' DO surface (rngV) DO ratio' pH Ugi’: Na ) (mg 1') K (mg T') Ca (mg T') Mg (mg T‘) Cl (mg 1') Cation Dominance Quininup Lake S 0.0013 - 18.0 1 1 8.30 131.73 4.3 48.50 17.01 200.65 Na >Ca > Mg > K Lake Davies W 0.0116 4.3 13.1 1 11.1 1.12 8.64 474.74 11.7 30.46 84.56 814.64 Na > Mg >Ca > K Lake Davies S 0.0116 4.5 17.0 1 11.4 1 8.60 424.86 9.8 32.46 78.25 720.70 Na > Mg >Ca > K Lake Quitjup w 0.7261 1.4 12.2 1 12.4 1 7.53 70.81 1.6 1.60 6.80 113.79 Na > Mg >Ca - K Lake Jasper w 4.3751 10.1 10.7 0.88 12.5 2.78 7.52 67.59 2.1 8.42 6.32 107.41 Na > Ca >Mg> K Lake Wilson w 0.1731 1.7 12.0 1 12.5 1.19 5.55 51.73 1.2 1.20 ,4.62 84.02 Na > Mg >Ca = K Lake Smith w 0.0450 1.6 13.2 1 12.4 1 4.70 45.29 0.6 <0.40 3.89 72.32 Na > Mg > K > Ca Lake Smith s 0.0450 1.6 16.0 1 11.1 1 4.50 31.27 0.7 <0.40 2.67 48.92 Na>Mg>K>Ca Charley Lake s 0.0260 6.4 24.0 1.33 7.8 13.00 6.50 39.31 0.8 1.30 3.30 72.32 Na > Mg >Ca > K Yeagarup Lake w 0.1697 10.8 11.5 1 13.8 1 7.05 34.49 0.9 6.01 3.65 54.59 Na > Ca >Mg > K Yeagarup Lake s 0.1697 10.1 15.2 1.27 12.8 1 6.80 33.11 1.0 5.21 3.40 57.78 Na > Ca >Mg > K Neanup Swamp s 0.0833 •• 16.0 1 ** 1 6.40 34.71 1.1 8.02 2.92 54.24 Na > Ca >Mg > K South Yeagarup Lake s 0.0610 *• 12.3 1 ** 1 6.70 26.21 1.0 43.69 3.65 41.12 Ca>Na >Mg>K Doggerup Lake Doggerup Lake Lake Samuel w 0.0831 ** 10.8 1 12.3 1 it* 34.26 1.1 <0.40 3.40 52.82 Na > Mg > K> Ca s 0.0831 2.5 18.0 1 9.8 1 5.20 36.09 0.8 < 0.40 2.92 58.49 Na > Mg > K > Ca s 0.0667 1.1 17.5 1 11.2 1 4.70 36.32 0.7 <0.40 3.40 62.75 Na > Mg > K > Ca Lake Florence s 0.1044 1.5 18.0 1 10.4 1 4.40 35.40 0.9 <0.40 2.92 51.76 Na > Mg > K> Ca Windy Harbour Lake w 0.0186 0.35 9.2 1 13.7 1 6.47 82.07 2.0 4.41 9.96 124.43 Na > Mg > Ca > K Lake Maringup w 1.3602 4.6 9.7 0.8 12.0 1 7.58 64.14 1.9 26.05 7.29 102.81 Na > Ca > Mg > K Lake Maringup s 1.3602 4.8 15.8 1 11.5 1 7.90 54.49 1.7 22.04 6.08 86.85 Na > Ca > Mg > K Owingup Swamp s 1.7886 1.3 19.5 1 11.2 1 7.40 340.48 1.5 17.64 58.08 717.15 Na > Mg > Ca > K Boat Harbour Lake 1 w 0.2507 0.6 13.5 1 12.5 1 8.00 150.12 4.1 29.26 18.71 245.31 Na > Ca > Mg > K Boat Harbour Lake 3 w 0.4235 0.8 14.5 1 12.2 2.09 8.10 241.40 6.2 48.10 27.70 407.32 Na > Ca > Mg > K Boat Harbour Lake 4 w 0.1097 12.5 1 12.4 1 8.05 149.66 5.6 20.44 23.09 246.02 Na > Mg > Ca > K Lake 12046 w 0.1014 4.9 10.3 1 12.9 1 7.70 182.31 16.9 22.04 16.77 250.63 Na > Ca > K > Mg Lake Williams s 0.0211 2.9 21.5 1 8.0 1.60 5.80 139.78 46.9 7.30 13.30 235.74 . Na>K>Mg>Ca Lake Saide w 0.4125 1.1 13.0 1 12.7 1 8.40 150.35 6.6 99.80 22.60 245.31 Na>Ca>Mg>K Lake Powell w 1.3973 0.7 15.0 1 11.8 1 8.10 291.51 10.1 29.66 28.92 492.05 Na > Ca > Mg > K Lake Powell s 1.3973 0.9 19.0 1 11.0 1 7.10 211.05 6.2 25.25 22.36 363.72 Na>Ca>Mg>K ^Ratio of surface to bottom readings. Table 3 Environmental information collected for each lake/occasion. W = winter, S = spring, * = below detection. Lake Season Salinity (mg/l) Salinity Colour Colour Turb Chloro N P Trophic status^ BOM Category' (APHA classification^ (NTU) (mgl') (mgl-') (%) units) (figl') Quininup Lake S 403 fresh <5 clear Lake Davies W 1417 brackish 10 clear Lake Davies s 126 brackish 10 clear Lake Quitjup w 195 fresh 110 brown Lake Jasper w 192 fresh 15 clear Lake Wilson w 143 fresh 170 brown Lake Smith w 122 fresh 380 black Lake Smith S 84 fresh 530 black Charley Lake s 117 fresh 310 black Yeagarup Lake w 100 fresh 180 brown Yeagarup Lake s 91 fresh 260 brown Neanup Swamp s 101 fresh 240 brown South Yeagarup Lake s 116 fresh 100 brown Doggerup Lake w 92 fresh 300 black Doggerup Lake s 99 fresh 330 black Lake Samuel s 104 fresh 470 black Lake Florence s 101 fresh 630 black Windy Harbour Lake w 223 fresh 340 black Lake Maringup w 202 fresh 20 clear Lake Maringup 5 171 fresh 55 clear Owingup Swamp s 113 bracki,sh 220 brown Boat Harbour Lake 1 w 448 fresh 70 clear Boat Harbour Lake 3 w 731 marginal 20 clear Boat Harbour Lake 4 w 445 fresh 65 clear Lake 12046 w 489 fresh no brown Lake Williams s 443 fresh 740 black Lake Saide w 525 marginal 80 clear Lake Powell w 852 marginal 220 brown Lake Powell s 629 marginal 190 brown 0.3 0.58 5.80 <0.01 oligo-mesotrophic 2.47 0.4 0.64 1.10 0.01 meso-eutrophic 1.92 0.4 1.53 0.83 <0.01 oligo-mesotrophic 1.29 0.9 0.42 0.56 0.01 meso-eutrophic 0.73 0.8 2.56 0.69 0.01 meso-eutrophic 1.16 0.7 1.07 0.46 0.02 meso-eutrophic 0.16 0.4 0.43 0.70 0.01 meso-eutrophic 1.24 0.9 0‘ 0.70 < 0.01 oligo-mesotrophic 3.44 0.4 0.97 0.64 0.01 meso-eutrophic 11.25 3.0 0.41 0.55 0.01 meso-eutrophic 3.62 0.7 0.19 0.55 <0.01 oligo-mesotrophic 2.24 0.8 0.80 0.73 <0.01 oligo-mesotrophic 0.18 2.6 0* 0.43 <0.01 oligo-mesotrophic 44.82 0.4 0.53 0.52 0.01 meso-eutrophic 17.31 0.4 0.20 0.55 0.01 meso-eutrophic 0.43 1.1 1.78 0.79 0.01 meso-eutrophic 66.04 0.7 0.39 0.89 0.01 meso-eutrophic 1.42 0.9 4.84 1.00 0.05 eutrophic 72.20 0.3 0.62 0.61 0.01 meso-eutrophic 7.40 0.3 0.40 0.48 <0.01 oligo-mesotrophic 17.10 0.6 3.59 0.69 0.01 meso-eutrophic 0.61 0.4 3.24 0.92 0.01 meso-eutrophic 87.10 0.5 2.45 1.40 0.01 meso-eutrophic 43.82 0.6 4.82 0.83 0.01 meso-eutrophic 44.68 1.7 3.72 0.82 0.02 meso-eutrophic 46.42 0.3 1.07 1.70 0.43 eutrophic 0.37 1.3 14.41 0.90 0.04 eutrophic 1.22 12.0 12.71 1.50 0.47 eutrophic 0.87 3.1 3.29 0.97 0.12 eutrophic 1.60 ’Based on the Water Authority of Western Australia (1989a) classification for surface water; fresh = 500 mg I ’ TSS, marginal = 500-1000 mg I-’ TSS, brackish = 1000-5000 mg 1’ TSS. ^Classification according to colour measurements in APHA units; < 100 = clear, 100-300 = brown, > 300 = black. ^Based on Wetzel (1975); oligo-mesotrophic = 5-10 pg 1’ P & 250-600 pg 1’ N, meso-eutrophic = 10-30 pg 1’ P & 300-1100 pg 1’ N, eutrophic = 30-100 pg 1’ P & 500-15000 pg 1' N. 54 Journal of the Royal Society of Western Australia, 77 (2), June 1994 the patchy distribution of organic material within each lake as shown for the six re-sampled lakes (Table 3). Factor analysis of the environmental data using principal components (PCA) indicated seven major factors (Table 4). The factors are numbered in decreasing amount of variation explained and therefore considered of decreasing 'impor¬ tance'. Factor 1 showed the co-occurrence of depth with stratification and low nitrogen levels. Factor 2 showed co¬ occurrence of all cations and anions measured, except potas¬ sium, and salinity, illustrating that in the lakes with rela¬ tively elevated salinities, no single cation was responsible. Additionally, the lakes with elevated salinity had low col¬ our. Factor 3 was a gradient generally of parameters that were different seasonally; temperature, dissolved oxygen and colour. Factor 4 showed the association of elevated phosphorus with chlorophyll and high turbidity (Table 4). Factor 5 showed that lakes with high pH, calcium, nitrogen and chlorophyll also had low colour levels. Factor 6 showed a gradient of nutrient levels (nitrogen, phosphorus and potassium) and colour. Factor 7 was a gradient of the association of lake size with low levels of benthic organic matter. A total of 85.9% variation in environmental param¬ eters was explained by these seven factors. The identifica¬ tion of these seven underlying factors greatly simplified the large data array, collapsing the information into co-occur¬ ring parameters. In subsequent analyses identifying gradi¬ ents in the ordinations of invertebrate community structure, factors, rather than individual variables, were correlated with axes scores. Invertebrates A total of 209 taxa belonging to 6 phyla were recorded from the lakes, and the occurrence of the taxa in the lakes is shown in Appendix 1. The number of invertebrate species collected from each lake ranged from 18 in Lake Davies to 55 and 56 species in lakes Charley and South Yeagarup, respec¬ tively. Stepwise multiple regression analysis, with hierar¬ chical inclusion of the number of species recorded in each lake against factor scores (Table 4), showed that Factors 2,3 and 5 were significantly correlated (p < 0.05, F-value = 5.7, Table 4 Factor scores calculated for the environmental variables and the percentage of variation explained by each Factor. A principal components analysis of the physico-chemical conditions associated with each site shows the pattern of co¬ occurrence of variables. The parameters highlighted in bold were considered 'significant', i.e. loading on any axis at > +0.30 or <-0.30 (Child 1970). Parameter FI F2 F3 F4 F5 F6 F7 SA 0.163 0.007 -0.166 0.237 0.163 -0.247 0.685 Season -0.134 -0.054 0.916 -0.153 -0.024 -0.026 0.034 Depth 0.5.S9 -0.056 -0.154 -0.302 -0.014 0.245 0.598 Temp (surface) 0.106 0.100 0.908 -0.004 -0.003 0.201 0.057 Temp (bottom) 0.900 0.133 0.261 0.060 -0.032 0.103 0.151 Temp (ratio) 0.945 0.072 0.035 0.029 0.035 0.062 0.150 DO (surface) 0.806 0.117 -0.425 0.045 -0.268 -0.027 0.027 DO (bottom) 0,871 0.130 -0.183 0.094 0.058 -0.090 -0.120 DO (ratio) 0.877 0.014 0.122 0.012 0.079 -0.018 0.039 pH 0.310 0.225 0.161 0.097 0.786 -0.049 -0.026 Col 0.204 -0.352 0.469 0.123 -0.584 0.336 -0.196 Turb -0.007 0.069 -0.045 0.880 0.024 0.017 0.154 Na 0.103 0.963 -0.022 0.117 0.155 0.119 0.027 K 0.100 0.253 0.105 0.172 -0.056 0.824 -0.097 Ca -0.201 0.335 -0.170 0.213 0.702 0.072 -0.090 Mg 0.072 0.981 0.005 -0.037 0.117 0.025 0.017 Cl 0.110 0.970 0.026 0.116 0.125 0.073 0.049 Sal 0.093 0.965 0.002 0.121 0.169 0.107 0.030 Chloro 0.126 0.184 -0.249 0.751 0.340 0.041 -0.098 N -0.426 0.063 0.131 -0.101 0.475 0.553 0.137 P 0.098 0.086 0.138 0.735 -0.126 0,593 0.082 BOM 0.032 -0.112 -0.293 -0.046 0.145 -0.071 -0.735 % variation explained 25.9 19.4 13.1 9.8 6.9 6.1 4.7 55 Journal of the Royal Society of Western Australia, 77 (2), June 1994 cumulative variation explained r^ = 0.41). This indicated the importance of salinity, where lakes with lower salinities had higher species richness (Fig 2), and to a lesser extent, season, pH and nitrogen levels. 60 • 50 - % • • 40 • • • • • j 30 • • • • 20 ” 10 1 0 i_I_I 500 1000 1500 Salinity (mg I"'') separated Powell, Quininup, Davies, Owingup, Boat Har¬ bour 3 and Saide from the other lakes (division 1 and 2; Fig 3). This division was attributed primarily to salinity, based on the results of multiple discriminant analyses (Table 5). One TWINSPAN level two division separated lakes Saide and Powell from Quininup, Davies, Owingup and Boat Harbour 3 (division 3 and 4; Fig 3) and was associated with Table 5 Discriminant analyses, using factor scores from PCA, on TWINSPAN (presence/absence) groupings. Discriminant analysiswas performed at each TWINSPAN division, termi¬ nating at level two of the classification analysis. The table illustrates percent correct classification, the most important t and Wilk's Lambda. All significance levels shown were p < 0.05. The values in brackets indicate the direction of the correlation between factor scores and TWINSPAN group¬ ings. Figure 2. Regression analysis of total species richness vs salinity. The regression was significant F^j = 7.63, p = 0.01, r = -0.47, n = 29. Zooplankton biomass was variable between lakes, rang¬ ing from zero in Doggerup Lake in winter to 678.5 mg in Lake Powell in winter (Appendix 1). The highest biomass values were recorded from lakes 12046 and Powell and consisted mainly of Daphnia carmata King. TWINSPAN analysis of the total species data-set was taken to two levels and indicator species for each division are shown in Fig 3. The first and most important division Groups % correct Variables Interpretation Lambda 1/2 100 F2 high salinity 0.46 (-) F5 high pH / nitrogen 0.32 (-) F4 high phosphorus 0.27 (-) F7 large size 0.23 (+) 3/4 100 F4 high phosphorus 0.36 (-) F2 high salinity 0.28 (+) FI stratified 0.22 (+) 5/6 95 F5 high pH/nitrogen 0.63 (-) FI stratified 0.37 (+) F3 high temperature 0.26 (+) F2 high salinity/low colour 0.23 (-) 22 W 2 W.S 23 W,S 16 S 18 W Calamoecia tasmanica 3 W 9 S 15 W.S 4 W 10 S 17 w 5 w 11 w,s 19 w 6 w,s 12 s 20 w 7 s 13 s 21 s 8 w.s 14 w Aphroteniella filicomis Hydracarina sp. R Gomphodella aff. maia 7Limnophyes sp. V31 Calamoecia attenuate _Y 5 6 3 W 11 W,S 4 W 15 W.S 5 W 12 S 9 S 17 W 6 W.S 13 S 10 S 19 W 7 S 21 S 14 W 20 W 8 W.S Aphroteniella filicomis Figure 3. Dendrogram showing TWINSPAN classification of the lakes using the total species list. Indicator species are shown at the bottom of the boxes. Lake numbers as in Figure 1. W = winter, S = spring. 56 Journal of the Royal Society of Western Australia, 77 (2), June 1994 phosphorus levels (Table 5). The second level two division of other lakes (division 5 and 6; Fig 3) was associated with pH and nitrogen levels (Table 5). Ordination by DECORANA on the total species data-set is presented in Figure 4. Along axis one, the obvious sepa¬ ration is of lakes Powell and Saide from the other lakes. Axis one was a gradient of pH, salinity and phosphorus concen¬ tration. These variables explained 32,27 and 16% of the total respectively (Table 6). 300 9s 200 1s CNJ .52 100 -100 6.12.13s 10s 6w« 11w /Os 15\^/i6s ^ 2w 17w .ISw •4w 20w 23w 23s 18w 22w I_I_I_i-1 -100 0 100 Axis 1 200 300 acidic water alkaline water low salinity high salinity low nutrients high nutrients - - conservation status low Figure 4. Axis 1 by axis 2 plot of DECORANA scores using total species for each lake, w = winter, s = spring. Lake numbers as in Figure 1. Table 6 Results of stepwise multiple regression analysis, with hier¬ archical inclusion of DECORANA axes scores against envi¬ ronmental parameters. The cumulative variation explained (r^) by each variable and F-values are presented. All factors vvhere the F-values were significant at p < 0.05 are shown in the table. Dependent variable Factor Interpretation r2 F-value Axis 1 scores F5 pH/nitrogen 0.32 12.6 F2 salinity 0.59 18.8 F4 phosphorus 0.76 25.8 Axis 2 scores FI depth/stratification 0.47 24.1 F3 season 0.69 28.4 Discussion The permanent lakes of the south coast of Western Aus¬ tralia contained a highly diverse invertebrate fauna, total¬ ling 209 taxa. However, the total taxa count would have been higher if the Nematoda and Annelida could have been identified to species. An interesting feature of the fauna was the number of species that previously had been recorded in lotic habitats in Western Australia. Five of the dragonfly species; Austrogomphus lateralis (Selys), Heinigomphus armiger (Tillyard), Lathrocordulia metallxca (Tillyard), Hesperocordulia berthoudi (Tillyard) and Synthemis cyanitincta (Tillyard) are considered to be stream species (Watson 1962). Within the Chironomidae, 14 species; Paramerina levidensis (Skuse), IZavreliniyia sp. V20, Ablabesrnyia sp. V37, Orthocladiinae sp. V43, Cricotopus anmdiventris (Skuse), ILimnophyes sp. V31, Thienemanniella sp. V19, Dicrotendipes sp. V47, Nilothauma sp., Riethia sp. V4, Riethia sp. V5, Harnischia sp. VCDIO, Steinpellina ?austraUensis Freeman and Aphrofeniellafilicornis Brundin have been recorded only from upland streams of the jarrah and karri forests (Edward 1986; Storey & Edward 1989). The Trichoptera, except Leptoceridae sp. F, sp. H and sp. I, the Amphipoda Perthia aciititelson Straskraba and the Dytiscidae beetle Sterxiopriscus browni Sharp have been recorded in upland streams of the northern jarrah forest (Aquatic Research Laboratory 1988; Bunn ei al. 1986; Storey etal 1990). The provisionally identified Glacidorbis sp., from South Yeagarup Lake, is of interest because the only previous representative from Western Australia, Glacidorbis occidentalis Bunn & Stoddart, is highly associated with intermittently flowing streams (Bunn et al 1989). The genus is known from both lotic and lentic waters in eastern Australia (Ponder 1986). The species richness of the fauna was high, in both a regional and State context, to the extent that many of the lakes could be considered environments with high conserva¬ tion significance. Species richness of invertebrates provides a useful comparison for different aquatic systems, with more diverse systems being considered 'healthier' than less di¬ verse ones (Magurran 1988). Table 7 shows a comparison of species richness for invertebrates from different biogeographic regions of Western Australia. Species rich¬ ness was high for lakes from both the Swan Coastal Plain and the south coast regions. The permanent lakes of the Swan Coastal Plain were characterised by nutrient enrichment to an extent where many were eutrophic, and the lakes were sampled on more than one occasion (Davis et al. 1993), which would have substantially increased the total species list. Sampling in summer/autumn would be likely to increase the total number of species for the south coast lakes. The major differences in species composition were that only about 30% of the identified species were shared by lakes in the two regions, with high numbers of species of Coleoptera in the Swan Coastal Plain lakes and high numbers of species of Chironomidae in the south coast lakes. Multivariate classification of the lakes on the basis of invertebrate community structure and subsequent correla¬ tion with environmental parameters illustrated a tight cou¬ pling. Multiple discriminant analyses using TWINSPAN groupings produced an approximately 98% correct classifi¬ cation, suggesting that a model can be derived where by the type of invertebrate community can be predicted with a very hi gh degree of accuracy, given a set of environmental param¬ eters collected during winter/spring. Any departure from this predictive success could indicate some level of distur¬ bance. The most important environmental parameters, deter¬ mined by multiple discriminant analyses to be significantly associated with invertebrate community structure, were Journal of the Royal Society of Western Australia, 77 (2), June 1994 Table 7 nf invertebrates recorded from a range of methodologically similar studies of lentic systems in Western Species richness oi Australia. System South coast lakes Two Peoples / (south coast) Robe River pools (Pilbara) Swan Coastal Plain lal Swan Coastal Plain lal Swan Coastal Plain lal TamworthLake (Swan Coastal Plain) Collie wetlands Catchment Lakes Seasonal sampling Species richness Reference mostly undisturbed 23 limited 209 this paper mostly undisturbed 3 yes 123 • 1 Storey et al. 1993 mostly undisturbed 10 limited 80 Streamtec 1991a urban 5 yes 87 Davis & Rolls 1987 urban 6 yes 176 Balia & Davis 1993 urban 41 yes 253 Davis et al. 1993 rural 1 no 48 Streamtec 1992 semi-disturbed rural 2 no 31 Streamtec 1991b disturbed, revegetated 1 no 44 Cale & Edward 1990 , I r.( cnlinitv pH and nutrient status. All the lakes had IfnLs well below 3 g 1’ considered the upper limit for ^ wkal freshwater (Bayly & Williams 1973). However, *e higher saliniHes, with a maximum of 1626 mg V at Lake naviSin winter, appearedsufficienttocausealocahsedloss S some, presumably less-tolerant, species. Deferences m *e geolories and origins of the lakes, particularly recent past connections to the sea, however, also need to be consid- ered. Two of the indicator species determined by the TWINSPAN analysis were associated with dark coloured, acidic water. Aphroteniellafilicornis, which is usually found in the lower order streams throughout Australia, has been recorded in perched acid, dune lakes on Fraser Island, Queensland (Cranston & Edward 1992). In this study, it only occurred in the acid dark-water lakes. The most commonly collected species of Ostracoda was GomphodeUa aff. maia Do Dekker which was mainly associated with the black-water acidic lakes. The Trichoptera Oecetis sp., was an important indicator species for the separation of lakes Quininup, Davies, Owingup, Boat Harbour 3, Saide and Powell from the other lakes in the TWINSPAN analysis. This genus is common in eutrophic lakes of the Swan Coastal Plain (Balia & Davis 1993 ) and in lower rivers of the northern jarrah forest (Storey etal. 1990). Lakes Powell, Saide, Owingup and BoatHarbour 3 had both elevated salinity and nutrients and contained fauna more typical of the eutrophic lakes of the Swan Coastal Plain including the species Daplitua carinata, Triplectides australisNavas,Sarsq/pridopsi$aculeata(Cosiz\Austrochiltonia subteiiuis (Sayce), Palaemonetes australis Dakin, C/iiro«omMS occidentalis Skuse and Polypedilum mibifer (Skusc) (Balia & Davis 1993; Pinder et al 1991). Candonocypris novaezelandiae (Baird) is often associated with eutrophic water (De Deckker 1981) and was only recorded from lakes 12046 and Powell. The high zooplankton biomass in Lakes 12046 and Powell consisted mainly of Daphnia carinata and this species was also the dominant species of Cladocera in the eutrophic lakes on the Swan Coastal Plain (Davis & Rolls 1987; Davis et al. 1993). Multivariate analyses showed that the structure of the invertebrate fauna of lakes Powell and Saide was highly associated with nutrient enrichment. Past and present land practices within the catchments of these lakes may have degraded water quality and subsequently adversely affected composition of the invertebrate fauna. The absence of Calamoecia from lakes Saide and Powell may reflect human activity in these eutrophic lakes. However, the geological history of the lakes should also be considered, as the only record of the euryhaline estuarine species Gladioferens imparipes Thompson was from Lake Powell, indicating a recent past connection with the sea. Lake Saide is associated with farming activities, particularly, in the past, with potato production. Lake Powell receives large nutrient inputs via Five and Seven Mile creeks, of secondary treated sewage from the Timewell Road Treatment Plant, Albany (Water Authority of Western Australia 1989b). Human activity should therefore be considered as a possible cause of el¬ evated nutrient status for all of the eutrophic lakes. Growns et al. (1993), using multivariate analysis of inver¬ tebrate data from 33 wetlands on the Swan Coastal Plain, found that the majority of wetlands could be grouped on the basis of nutrient status and colour, and concluded that low nutrient levels and highly coloured waters were the prob¬ able state of wetlands prior to European settlement. If this conclusion is correct, then many of the dark-water lakes of the south coast would have a high conservation status. Analysis of the total species dataset by DECORANA (Fig 4) showed a gradient of putative conservation status of the 58 Journal of the Royal Society of Western Australia, 77 (2), June 1994 lakes, from 'high' quality, having low scores on axis one {e.g. lakes Doggerup, Smith, Wilson, Quitjup, Williams and Maringup), through 'intermediate' quality lakes (those in the centre of the axis and including Boat Harbour Lake 3, Lake Davies, Owingup Swamp and Lake 12046) to 'low' quality, having high axis one scores {e.g. lakes Powell and Saide). This nominal conservation status of the lakes, based on invertebrate community structure, was reiterated on the important underlying gradients. pH, salinity and phospho¬ rus concentration respectively are the most important envi¬ ronmental parameters describing the above pattern. Future management programmes should include the routine moni¬ toring of these parameters. Acknowledgtnents: Thesludy was funded by ihe Western Australian Depart¬ ment of Conservation and Land Management and the Australian National Parks and Wildlife Service, jim Lane, the Project Officer (CALM), is especially thanked for his guidance and cooperation. The following are appreciated for their specifictaxonomiccxpertise; Ian Bayly (Copepoda), RussShiel (Cladocera) and Stuart Halse (Ostracoda). References AquaticResearch Laboratory 1988 Biological monitoring in the catchmentsof the northern jarrah forest and Swan Coastal Plain. Unpublished report ARL013 to the Water Authority of Western Australia, Perth. Balia S A & Davis J A 1993 Wetlands of the Swan Coastal Plain. Vol 5. Managing Perth's wetlands to conserve the aquatic fauna. Water Author¬ ity of Western Australia, Perth. Bayly IA E1982 Invertebrate fauna and ecology of temporary poolson granite outcrops in southern Western Australia. Australian Journal of Marine and Freshwater Research 33:599-606. Bayly 1 A E 1992a The micro-Crustacea and physico-chemical features of temporar)' ponds near Norlhcliffe, Western Australia. Journal of the Royal Society of Western Australia 75:99-106. Bayly I A E 1992b The non-marine Centropagidae (Copepoda: Calanoida) of the world. SPB Academic Publishing, The Hague. BayIyIAE&WiIliamsWD1973 Inland Waters and their Ecology. Longman, Australia. BunnSE, Edward DH& LoneraganN R1986 Spatial and temporal variation in the invertebrate fauna of streams of the northern jarrah forest, Western Australia: community structure. Freshwater Biology 16:67-91. Burtn S E, Davies P M & Edward D H 1989 The association of Glacidorbis ecc/dc«ffl/is Bunn (StStoddart 1983 (Gastropoda: Gladdorbidae) with inter¬ mittently-flowing, forest streams in south-western Australia. Journal of the Malacological Society of Australia. 10:25-34. Cale D J & Edward D H D 1990 Tlie influence of aquatic macrophytes on invertebrate communities in Swamphen Lake, Capel, Western Australia. Technical Report No. 10. RGC WetlandsCentre, Capel, Western Australia. Child D 1970 The Essentials of Factor Analysis. Holt, Rinehart & Winston, London. Christensen P1982 The distribution of Lqndogalaxiassalantandroides and other small freshwater fishes in the lower south-west of Western Australia. Journal of the Royal Society of Western Australia 65:131-141. Cranston PS& Edward D H D1992 A systematic reappraisal of the Australian Aphrotcniinac (Diptera: Chironomidae) with dating from vicariance biogeography. Systematic Entomology 17:41-54. Davis J A & Rolls SW 1987 A baseline biological monitoring programme for the urban wetlands of the Swan Coastal Plain, Western Australia. Bulletin 265. Environmental Protection Authority, Perth. Davis J A, Rosich R S, Bradley J S, Crowns J E, Schmidt L G & Cheat F 1993 Wetlands of the Swan Coastal Plain. Vol 6. Wetland classification on the ba.sis of water quality and invertebrate community data. Water Authority of Western Australia, Perth. De Dcckker P 1981 Taxonomy and ecological notes on some ostracods from Australian inland waters. Transactions of the Royal Society of South Au.stralia 105:91-138. Edward D H D 1986 Chironomidae (Diptera) of Australia. In: Limnology in Australia (eds P De Deckker & W D Williams) CSIRO, Melbourne, 159- 173. Purse MT, Moss D, Wright J F & Armitage P D1984 The influence of seasonal and taxonomic factors on the ordination and classification of running- water sites in Great Britain and on the prediction of their macro-inverte¬ brate communities. Freshwater Biology 14:257-280. Gauch Jr. H G1982 Multivariate Analysis in Community Ecology. Cambridge University Pres.s, Cambridge. Crowns J E, Da vis J A, Cheal F, Schmidt L & Rosich R1993 Multivariate pattern analysis of wetland invertebrate communities and environmental vari¬ ables in Western Australia. Australian Journal of Ecology 17:275-288. Halse S A 1989 Wetlands of the Swan Coastal Plain—past and present. In: Proceedings of the Swan Coastal Plain Groundwater Management Con¬ ference (ed G Lowe) Western Australian Water Resources Council, Perth, 105-112. Hill M O 1979a TWINSPAN—A FORTRAN Program for Arranging Multivariate Data in an Ordered Two-way Table by Classification of the Individuals and Attributes. Ecology and Systematics, Cornell University, Ithaca. Hill M 01979b DECORANA—A FORTRAN Program for Detrended Corre¬ spondence Analysis and Reciprocal Averaging. Ecology and Systematics, Cornell University, Ithaca. Loh IC, Ventriss H B & Collins P D K1983 Water quality—Its significance in Western Australia. In: Water Resources Quality in Western Australia— Proceedings of the Water Research Foundation of Australia Seminar Perth, 2-10. Magurran A E 1988 Ecological Diversity and its Measurement. Princeton University Press, Princeton. Moss D, Furse M T, Wright J F & Armitage P D 1987 The prediction of the macro-invertebrate fauna of unpolluted running-water sites in Great Britain using environmental data. Freshwater Biology 17:41-52. Noru§sMJ 1986 SPSS/PC’Advanced Statistics. SPSS Inc., Chicago. Finder A M, Trayler K M & Davis J A 1991 Chironomid control in Perth wetlands. Final report and recommendations. Unpublished report, Murdoch University, Murdoch. Ponder W F 1986 Glacidorbidae (Glacidorbacea: Ba.sommatophora), a new family and .superfamily of operculate freshwater gastropods. Zoological Journal of the Linnean Society 87:53-83. Pusey B J & Edward D H D 1990a Structure of fish assemblages in waters of the southern acid peat flats. South-western Australia. Australian Journal of Marine and Freshwater Research 41:721-734. Pusey B J & Edward D H D 1990b Limnology of the southern acid peat flats. South-western Australia. Journal of the Royal Society of Western Aus¬ tralia 73:29-46. Storey A W & Edward D H D 1989 Longitudinal variation in community structure of Chironomidae (Diptera) in two south-western Australian riversystems. Acta Biologica Debrecen, Oecologica Hungarica 3:315-328. Storey A W, Bunn S E, Davies P M & Edward D H 1990 Classification of the invertebrate fauna of two river systems in southwestern Australia in relation to physical and chemical parameters. Regulated Rivers: Research & Management 5:217-232, Storey A W, Halse S A & Shiel R J1993 Aquatic invertebrate fauna of the Two Peoples Bay area, .southwestern Australia. Journal of the Royal Society of Western Australia 76:25-32. Streamtec 1991a Aquatic Fauna and Water Chemistry; Robe River. Unpub¬ lished report ST 136. Robe River Iron Pty Ltd, Perth. Streamtec 1991b Proposed Collie Power Station: Wetland Survey. Unpub¬ lished report ST 134. Kinhill Engineers, Perth. Streamtec 1992 Baseline Aquatic Survey of the Biology of Tamworth Lake, Baldivis. Unpublished report. Wood & Grieve Engineers Pty Ltd, Perth. Strickland G D H & Parsons T R 1968 A practical handbook of seawater analysis. Fisheries Research Board of Canada Bulletin 167. Water Authority of Western Australia 1989a Stream salinity and its reclama¬ tion in south-west Western Australia. Water Resources Directorate, Surface Water Branch Report WS 52. Water Authority of Western Aus¬ tralia, Perth, Water Authority of Western Australia 1989b Albany wastewater treatment and disposal: A plan for the future. Unpublished report. Water Authority of Western Australia, Perth. Watson J A L1962 The dragonflies of south-western Australia. Handbook No. 7. Western Australian Naturali.st's Club, Perth. Wetzel R G 1975 Limnology. Saunders, Philadelphia. Wright J F, Moss D, Armitage P D & Fur.se M T 1984 A preliminary classification of running water sites in Great Britain based on macro¬ invertebrate species and the pred Iction of community type using environ¬ mental data. Freshwater Biology 14:221-256. 59 Journal of the Royal Society of Western Australia, 77 (2), June 1994 Appendix 1. Species richness and list of taxa identified from the benthic, sweep and zooplankton samples and zooplankton biomass (dry weight) in a standard 50 m trawl. W = winter, S = spring. * = open water inaccessable. Numbers represent abundance categories for each species in any of the samples; 1 = < 50,2 = 50-500,3 == > 500 individuals. An 'A' after the taxa = adult — Coleoptera only. Lake numbers as in Figure 1. IS 2W 2S 3W 4W 5W 6W 6S 7S 8W 8S 9S lOSllW IIS 12S 13S 14W 15W15S 16S 17W 18W19W 20W 21S 22W 23W23S Spedes richness (total taxa) 28 18 18 35 22 28 43 38 55 33 25 45 56 51 50 33 43 28 38 . 41 45 32 34 39 29 37 25 31 33 Zooplankton biomass (mg) 30.6 81.9 47.8 143.9 57.6 86.6 18 72.8 20.2 83.1 11.9 313 * 0 24.2 67.5 18.8 7.4 117 34.5 24.3 43.4 39.4 : 35.1 : 764.3 68.9 16.5 679 587 CNIDARIA Hydra sp. 2 1 1 1 PLATYHELMINTHES TEMNOCEPHALOIDEA Temnocqjhala sp. 1 1 TURBELLARIA Dugesiidae sp. 1 1 NEMATODA Nematoda spp. 2 1113 2 11111 2 2 2 1 1 1 2 2 1 1 1 1 1 13 12 1 MOLLUSCA BIVALVIA Bivalva sp. 1 1 Bivalva sp. 2 1 Westralunio carteri Iredale 1 1 1 GASTROPODA IClacidorhis sp. 1 Ferrissia petterdi (Johnston) 1 1 1 1 2 1 1 Gastropoda sp. 1 3 Gastropoda sp. 2 1 1 ?Physastra 1 1 1 1 1 1 1 2 ANNELIDA OLIGOCHAETA Oligochaeta spp. 3 2 1 1 3 1 1 1 2 2 1 1 1 1 1 1 2 1 2 1 2 1 2 2 2 112 1 HIRUDINEA Richardsonianidae sp. 1 1 1 2 1 1 Glossiphoniidae sp 2 1 ARTHROPODA arachnida Hydracarina sp. C 1 Hydracarina sp. D 1 Hydracarina sp. E 1 1 1 Hydracarina sp. F 1111 1 1 1 1 1 1 1 1 Hydracarina sp. G 1 1 1 1 1 1 1 1 1 1 1 1 Hydracarina sp. I 1 1 1 1 1 1 1 Hydracarina sp. J 1 1 1 1 Hydracarina sp. L 1 Hydracarina sp. 0 1 Hydracarina sp. Q 1 1 Hydracarina sp. R 1111 1 1 1 1 1 1 1 1 1 1 Hydracarina sp. S 1 1 1 1 Hydracarina sp. T 1 11111 1 Hydracarina sp. U 1 Hydracarina sp. V 1 CRUSTACEA Cladocera Biapertura nr. setigera Brehm 1 1 1 1 1 1 1 1 Bosmina meridionalis Sars 3 1 3 Chydorus sp. Leach 1 2 111* 1 1 1 1 1 1 1 Neothrix cf. armata Gurney 1 1 2 1 1 Daphnia carinata King 1 3 3 3 Alonella sp. 11 11 1 1 1 1 1 Cladocera undescribed genu.s V13 1 1 1 1 Cladocera ?genus V16 1 Cladocera ?genus V17 1 Cladocera ?genus V18 1 Cladocera ?genus V15 2 Craptoleberis testudinaria Sars 2 1 1 Camptocercus cf. australis Sars 1 1 1 Ostracoda Newnhamia fenestra King 1 2 Cypretta baylyi McKwtzie 1 2 1 Newnhamia sp. 295 1 Sarsnpridopsis aculeata (Costa) 3 1 1 1 1 2 2 limnocythere mowbrayensis Chapman 1 1 Candonopsis tenuis (Brady) 11 11 1 1 2 1 1 1 1 Gomphodella aff. maia De Dcckker 1 1112 1 2 1 2 1 3 1 1 Cyprididae undescribed genus 1 2 1 1 1 1 1 Alboa worooa De Deckker 1 2 1 1 2 Ilyodromus sp. 255 1 1 1 1 1 Paralimnocythere sp. 262 1 1 Candonocypris novaezelandiae (Baird) 1 2 1 60 Journal of the Royal Society of Western Australia^ 77 (2), June 1994 IS 2W 2S 3W 4W 5W 6W 6S 7S 8W 8S 9S lOSllW llS 12S 13S 14W 15W15S 16S 17W18W19W20W 21S 22W23W23S Species richness (total taxa) 28 18 18 35 22 28 43 38 55 33 25 45 56 51 50 33 43 28 38 41 45 32 34 39 29 37 25 31 33 Zooplankton biomass (mg) 30.6 81.9 47.8 143.9 57.6 86.6 18 72.820.2 83.1 11.9 31.3 * 0 24.2 67.5 18.8 7.4 117 34.5 243 43.4 39.4 35.1 764.3 68.9 16.5 679 587 Copepoda Cyclopoida Cyclopoida spp. Harpactacoida Harpactacoida spp. Calanoida Calanoida sp. Calamoecia tasmanica (Smith) s.l. Calatnoecia aitenuata Fairbridge Hemiboeckella searli Sars Hemiboeckella andersonae Bayly Cladioferens imparipes Thompson Isopoda Phreatoicidea Phreatoicidea sp. A Amphipoda Amphipoda sp. A Amphipoda sp. B Gammaridae Perlhia branchialis (Nicholls) Perthia acutitelson Straskraba Ceinidae Ausirochiltonia subtenuis (Sayce) Decapoda Parastacidae Cherax sp. (immature) Cherax (juinquecarinatus (Gray) Cherax lenuimanus (Smith) Cherax destructor Clark Palaemonidae Palaemonetes australis Dakin INSECTA Ephemcroptera Leptophlebiidae Leptophlebiidae sp. Neboissophlehia occidentalis Dean Bibulmena kadjina Dean Baetidae Cloeon sp. Caenidae Tasmanocoenis tillyardi (Lestage) Odonata Odonata sp. Aeshnidae Aeshna brevistyla (Rambur) Libeliulidae Orthetrum caledonicum (Brauer) Austrothemis ni^rescens (Martin) CorduHidae Lathrocordulia metallica (Tillyard) Hesperocordulia berthoudi (Tillyard) Hemicordulia australie (Rambur) Procordulia affinis (Selys) Synthemidae Synthemis cyanitincta (Tillyard) Gomphidae Austrogomphus lateralis (Selys) Hemigomphus armiger (Tillyard) Austrogomphus collaris Hagen Ustidae Austrolestes annulosus (Selys) Mcgapodagriidae Austroagrion cyane (Selys) Coenagriidae Cocnagriidae sp. Hemiplera Corixidae Micronecta sp. Agraptocorixa sp. A Sigara sp. Agraptocorixa sp. B Veliidae Veliidae sp. 2 Notonectidae Notonectidae sp. Anisops sp. ^otonecta sp. 111111 1111 12 1 111 1 2 11 1 3 3 2 3 3 3 1 3 3 3 3 2 1 1 1 3 13 2 11 2 1 1 3 3 3 3 1 1 1 1 1 111 11 111 3 3 2 1 1 1 2 2 1 1 1 1 3 3 3 2 1 1 1 2 1 1 1 1 1 11 11111 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 2 1 1 1 1 1 11112 1 11 2 1 3 3 3 3 3 3 1 1 12 11 2 12 12 1 1 1 1 1 1 3 2 2 2 11112 1 1 1 3 1 2 2 3 3 1 1 3 1 1 1 1 2 1 2 3 2 3 1 1 1 1 2 3 3 1 2 1 1 1 1 1 1 1 1 1 1 1112 1 1 1 1 1 1 1 2 61 Journal of the Royal Society of Western Australia, 77 (2), June 1994 species richness (total taxa) Zooplankton bi omass (mg) Nepidae Ranatra sp. Diptera Chironomidae Paramerina levidensis (Skuse) Prochdius paludicola Skuse procladius villosimanus Kieffer ?Zairrelimyia sp. V20 Macropelopia dalyupetisis Freeman Macropclopio sp. VSCL50 Ablabesmyia sp. V37 Coelopynia pridnosa Freeman Orthocladiinae sp. VSCL3 Orthocladiinae sp. VSCl.7 Orthocladiinae sp. VSCL20 Orthocladiinae sp. VSCL36 Orthocladiinae sp. V5CL38 Orthocladiinae sp. VSCL43 Orthocladiinae sp. V43 Cricolopus amuliventris (Skuse) Corynonma sp. Limnophyespullulus (Skase) ILimnophyes sp. V31 ILimnophyes sp. VSCL45 Thienemanniella sp. V19 SHctocladius sp. VSCL24 Dicrotendipes sp. V47 Dicrolendipes Iconjunctus Walker Kiefferulus martini Freeman Ki^ulus inlertinctus Skuse Cryptochironomus griseidorsum Kieffer Nilothauma sp. 7Stenochiron(mius sp. Stenochironomus sp. VSCL14 Stenochironomus sp. VSCL31 Riethia sp. V5 Rielhia sp. V4 Polypedilum sp. VSCL8 Polypedilum sp. VSCL16 Polypedilum sp. VSCL 33 Polypedilum nubifer (Skuse) Chironomus occidenlalis Sku.se Chironomus aff. alternans Walker Cladopelma curtivalva Kieffer HdrHisc/iifl sp. VG310 Chironomini ?genus VSCL27 Chironomini ?genus VSCL34 Chironomini ?genas VSCL35 Tanytarsini ?genus Tanytarsus sp VSCL5 Tanytarsus sp. VSCL9 Tanytarsus sp. VSCL18 Cladotanytarsus sp. VSCLIO Stempellina laustraliensis Freeman Aphroteniellafilicomis Brundin Ceratopogonidae Ccratopogonidae sp. B Ceratopogonidae sp. F Ceratopogonidae sp. G Ceratopogonidae sp. K Ceratopogonidae sp. L Ceratopogonidae sp. N Ceratopogonidae sp. O Ceratopogonidae sp. P Tipulidae Limoniinae sp. A Limoniinae sp. B Limoniinae sp. F Empididae Empididae sp. A Empididae sp. B Tabanidae Tabanidae sp. A Dolichopodidae Dolichopodidae sp. A Culicidae Culicidae sp. IS 2W 25 3W 4W.5W 6W 6S 7S 8W 8S 9S lOSllW llS 125 135 14W 15W15S 16S 17W 18W19W 20W 21S 22W 23W23S 28 18 18 35 22 28 43 38 55 33 25 45 56 51 50 33 43 28 38 41 45 32 34 39 29 37 25 31 33 30.6 81.9 47.8 143.9 57.6 86.6 18 72.820.2 83.1 11.9 31.3 * 0 24.2 67.5 18.8 7.4 117 34.5 24.3 43.4 39.4 35.1 764.3 68.9 16.5 679 587 1 1 1 1 1 1 1 1 1 1 11111 1 1 1 1 1 2 11111 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1111 1 1 1 1 1 1 1 1 1 1111 11111 1 1 1 1 1 1 1 1 1 1 1 1 11111 1 1 111 1 11 1 1 1 1 1 1 1 112 1 1 1 2 111111 1 1 2 1 1 1 1 1 1 2 2 2 111 11111 1 1 1 1 1 1 1 11111 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1111 1 1 11111 1 1 1 2 1 1 111 111 1 1121 1 1 1 1 1 1 1 1 1 1 1 1 1111 2 2 62 Journal of the Royal Society of Western Australia, 77 (2), June 1994 15 2W 2S 3W 4W 5W 6W 65 7S 8W 8S 9S IPS 11 W 11 S 12S 13S 14W 15W15S 16S 17W 18W19W 20W 21S 22W 23W23S Species richness (total taxa) 28 18 18 35 22 28 43 38 55 33 25 45 56 51 50 33 43 28 38 41 45 32 34 39 29 37 25 31 33 Zooplankton biomass (mg) 30.6 81.9 47,8 143.9 57.6 86.6 18 72.8 20.2 83.1 11.9 31.3 ‘ 0 24.2 67.5 18.8 7.4 117 34.5 24.3 43.4 39.4 35.1 764.3 68.9 16.5 679 587 Ephydridae Ephydridae sp. A Trichoptera Ecnomidae Ecnomidae sp. Ecnomina ?trulla Neboiss ?Eawmus sp. 1 Ecnomus pansus Neboiss Ecnomus turgidus Neboiss Leptoceridae Leptoceridae spp. 1 Ledrides parilis Neboiss Triplectides sp. B Notoperata lenax Neboiss Notalina sp. A Oecetis sp. 1 Triplectides australis Navas Leptoceridae sp. C 1 Leptoceridae sp. F Leptoceridae sp. H Leptoceridae sp. I Hydroptilidae Hydroptilidae sp. Acritoptila globosa Wells Hellyethira sp.B 1 Atriplectididae Atriplectides ?dubius Mosley Polycentropodidae Polycentropodidae sp. Plectronemia cximia Neboiss Coleoptera Dytiscidae Stermypriscus Ibrowni Sharp Sternopriscus sp. A liodessus sp. Megaporus sp. Homeodytes scutellaris (Germar) 1 Homeodytes sp. Sternopriscus broumi Sharp A 12 11 Stermypriscus marginata Watts A 1 1 BJiantus sp. A liodessus itwrnalus (Sharp) A Mecterosoma daneini (Babington) A Antiporus femuralis (Boheman) A Megaporus liowitti (Clark) A Megaporus solidus (Sharp) A 12 Homeodytes scutellaris (Germar) A Lancetes lanceolatus Clark A Liodessus dispar (Germar) A 1 Megaporus sp. A Hclodidae Helodidae Cyrinidae Macrogyrus sp. A Hydrophitidae Hydrophilidae sp. 6 Hydrophitidae sp. 1 A 2 Hydrophilidae sp. 2 A Hydrophilidae sp. 3 A Hydrophilidae sp. 9 A Hydrophilidae sp. II A Carabidae ?Carabidae sp. A 1 1 1 1 1 1 1 1 1 1 1 1111 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 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 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 1 1 1 1 1 11 111 1 1 1 1 1 1 1 1 1 63 Biosystematics of Australian mygalomorph spiders: Description of a new species of Aname and its aerial tube (Araneae: Nemesiidae) Barbara York Main Department of Zoology, The University of Western Australia, Nedlands WA 6009 Manuscript received April 1994; accepted June 1994 Abstract A new species of the Australian mygalomorph genus Aname is described. The species has an unusual burrow structure which consists of a soil-plastered tube that extends into the foliage of a supporting shrub. It is postulated that the aerial tube is an anti-flooding adaptation and that the spider has secondarily become a forager amongst foliage. Such a tube and behaviour have not been recorded for any other Australian mygalomorph spider. Introduction This paper is the first in a planned series in which new species of Aname will be described and their respective biology d iscussed. The endemic nemesiid genus A name Koch is the most widely distributed of all Australian mygalomorph genera. It occurs all over the continent in diverse habitats (Main 1972; Raven 1981,1984a, b, 1985), in Tasmania (Hogg 1902; Raven 1984b) and on several offshore islands (Main 1976,1982). Main (1985) recognised 14 nominal species and indicated that there are many undescribed species. Raven (1981) synonymised Chenistonia with Aname, a decision not concurred with by Main (1985). Raven later (1984a, b, 1985) attributed 17 new species from northern and eastern Aus¬ tralia and one from Tasmania to Aname. Of these. Main (1985) listed two in Chenistonia. Main (1993) gave a summary account of anti-flooding burrow specialisations of some mygalomorph species. In¬ cluded amongst the specialisations, were listed those of nests of some species which characteristically construct nests on the trunks of trees, and which have no contact with the ground, e.g. of certain theraphosids, Sason (Barychelidae), Conothele (Ctenizidae), Chenistoniavillosa Raibow and Pulleine (Nemesiidae) and Moggridgea tingleMam (Migidae). In addi¬ tion, it is well known that some species of the notorious Australian funnel web spiders, Hadronychc (of which some species have been cited as Atrax species) similarly construct silk tubes on the bark and trunks of trees (Gray 1978,1981, 1984,1988; McKeown 1962; Mascord 1970; Main 1964,1976), sometimes as high as 15m above ground (Gray 1981). Main (1993) also described the aerial tube of an un¬ named Aname species. The nest of this species is unique in that, while other arboreal mygalomorphs have their tubes associated with the trunk or branches of trees, the Aname species has its tube extending into foliage. Main postulated that this aerial lube, which is supported by spinifex (Triodia) tussocks or chenopod shrubs, primarily prevents flooding of the burrow during sudden deluges, and that secondarily it has exposed the spider to a new foraging situation (for a mygalomorph) i.e. within the foliage of plants. While descriptions of numerous new species oi Aname are in progress, this species has been singled out here for de¬ scription because of its noteworthy burrow structure and distinctive foraging behaviour. Systematics Abbreviations BYM, Barbara York Main collection (housed in the Zoology Department, University of Western Australia). SAM, South Australian Museum. WAM, Western Australian Museum. SA, South Australia. WA, Western Australia. ALE, anterior lateral eyes; AME, anterior median eyes; PLE, posterior lateral eyes; PME, posterior median eyes, v, ven¬ tral; d, dorsal; p, prolateral; r, retrolateral (in reference to position of leg spines). Aname turrigera sp. nov. (Figures 1 A - M, 2 A - B, 3; Table 1) Holotype: Female, 24 km W of Balladonia Roadhouse, EyreHighway,WA(32°15' S123°2' E), collected by B Y Main, 20 November 1986 (BYM 1986/174, WAM 92/2629). Paratype: Female (internal genitalia figured), data as for holotype (BYM 1986/173, WAM 92/2630). Other paratypes: (all collected by B Y Main unless other¬ wise stated). Western Australia: 3 females, same data as holotype (BYM 1986/175,176,177);2females, same locality as holotype, 24 May 1986 (BYM 1986/157,158); female with brood, data as for preceding (BYM 1986/159); female, Mallura, A R Main, 21 August 1960 (BYM 1960/19). South Australia: 2 females, 18 km west of Lock, 8 May 1986 (BYM 1986/29, with brood young; BYM 1986/60 (SAM N1994396) internal genitalia dissected); 3 females and one © Royal Society of Western Australia 1993 65 17554—1 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Figure 1. Attflmefumgerfl, female (holotype). A, Dorsal view of spider, carapace and abdomen. B, Abdomen, ventral. C Eves. D Sternal area. E, Maxillary cuspules, F, Pseudorastellum. G, Palp tarsi and metatarsi, r = right, 1 = left. H, Left leg I, ventral, tarsus, metatarsus, tibia. I, J, K, Female internal genitalia; I, (BYM 1986/173) W Balladonia, J, (BYM 1986/110) | Yalata, K, (BYM 1986/60) W of Lock. L, Spinifex tussock with aerial tubes. M, Spider in foraging position at entrance ot tube (adapted from photograph by D Hirst). Scale bars: A, B, D, E, G, H = 1.0 mm; C = 0.5 mm, F, not to scale, I - K — 0.5 mrn/ 1 L, = 10.0 cm; M = 1.0cm. J Journal of the Royal Society of Western Australia, 77 (3), September 1994 cast skin (separate burrow) same locality as preceding, 18 Mayl986(BYM1986/98,99,101,102); female, 29.6 km north of Minnipa, 16 May 1981 (BYM 1981/4); 2 females, Yalata (Yalata Swamp, Head of the Bight), 20 May 1986 (BYM 1986/ 109; 110 internal genitalia dissected); female, Ardrossan, 16 May 1986 (BYM 1986/67) (tentative identity). Description is of holotype except where otherwise stated. Male unknown. Female . Colour generally tan brown, abdomen dorsally with dark bands on pale yellowish background, venter pale. Carapace glabrous, with light covering of fine hairs; length 4.6 mm, width 3.5, caput width 2.4. Eye tubercle 0.4 long, 1.0 wide, pronounced anterior mound. Eye diameters: ALE 0.25, AME 0.15, PLE 0.2, PME 0.1; AME apart 0.2, ALE apart 0.65. Fovea procurved. Chelicerae with long heavy apical teeth (pseudorastellum). Teeth on groove of paturon irregular sizes, prolateral 10 (right), 7 (left), without basal retrolateral teeth. Sternum length 2.5, width 2.2; sigilla, posterior pair ovoid, submarginal. Labium length 0.4, indented anteriorly. Maxillae with about 35 long, thin pin-like cuspules. Legs. Scopula complete on tarsi of palp, legs I and 11 (without dividing line of bristles), and metatarsi I; on metatarsi II complete on retrolateral aspect and apical half only of prolateral aspect; absent on legs III and IV which have dense bristles on tarsi. Paired tarsal claws I and II with uniform comb of at least six or seven teeth respectively on each claw; Ill with fewer teeth, inner rows with fewer teeth and distal to outer rows; IV, teeth present only on "outer" edges of daws, prolateral claw with 3 prolateral teeth, retrolateral daw with 3 or 4 basal retrolateral teeth. Spines absent on all leg tarsi. Palp tarsus v 5 (right), v 4 (left - arranged as 1 apical and 3 in proximal line; this is the "typical" arrangement for most specimens - see under Variation and Fig 1 G ). Leg I, metatarsus v 2-2-1-3 (apical), tibia v 5. II, metatarsus v 2-1-3, tibia V (fine bristles) 1 -1-1. Ill, metatarsus v 5, d 1-2, rd 4, tibia V (bristles) 1-1-2, d 2, p 2, r 2, patella p3 + dense thorn-like "spinules", r 2. IV, metatarsus v 9, r3, tibia v (fine bristles) 1- 1-2, r 2, Abdomen with marked dorsal pattern of bars. Spinner- ets with terminal segment of posterior lateral pair ovoid. Internal genitalia (Paratype WAM 92/2630, BYM 1986/173) bilobed, main lobe bottle shaped with broad "open" base, narrow neck and terminally bulbous; lateral lobe arising from main lobe below constriction of "neck" (Fig II). Variation Table 1 Leg dimensions of holotype female of Aname turrigera. ^8 formula = length of leg divided by carapace length. Tibial index = width of patella X 100 divided by length of tibia + patella (Petrunkevitch 1942). Leg formula; 4/3.02,1/2.2,2/2.0,3/1.73 Femur Pat lih Mi la Total 3.2 2.0 2.0 1.8 1.3 10.3 2.7 1.8 1.8 1.7 1.2 9.2 2.2 1.4 1.2 1.9 1.3 8.0 3.1 2.1 3.4 3.0 1.3 13.9 2.2 1.5 1.4 - 1.5 6.6 JJJjdth of patella I at knee = 0.6. Tibial index = 15.0. ^*dth of patella IV at knee = 0.7. Tibial index = 12.7. Chelicerae , usually 6 or 7 prolateral teeth on groove of specimens with carapace length upwards of 3.5 mm (may be asymmetrical on individual specimens); 3 to 5 basal retrolateral teeth usually present (absent on holotype). Juve¬ nile specimens (brood of BYM 86 /159) with prolateral teeth. Labium of paratype (BYM 86 /173) with a single long pointed cuspule. Some specimens (e.g. BYM 86/ 60, from Lock, Eyre Peninsula, S A) with short thorn-like spines on promargin of coxae I (absent on holotype). Spines on palp tarsus . While the typical number is like that of holotype, 1 distal and three in proximal row ( left tarsus Fig 1 G), the number varies (unrelated to size or locality) as 1, 2, occasionally 3 distal spines, and 3 or rarely 4 proximal (17 sampled). Carapace length/width of females ranges from 3.5 / 2.8 {i.e. from smallest with apparent adult internal genitalia), 3.6/2.7 (female with brood young) to 4.7/3.5 (largest specimen collected). The internal genitalia of specimens from widely distant localities have a remarkably similar configuration (see Figs 11, J, K) e.g. Balladonia (paratype BYM 1986/173) carapace L/W 4.6/ 3.4); Yalata (BYM 1986/110) carapace L/W 4.2/2.3), Lock (BYM 1986/60, carapace L/W 3.5/2.8) Etymology The name turrigera, from the Latin (turris-gero) meaning "tower-bearing", in reference to the tower-like aerial tube. Affinities The species has some morphological features similar to Kwonkan Main and Yilgarnia Main, namely the configuration of the internal genitalia (see Main 1983 Figs 23 - 26,1986 Fig. 1 h, i), a strong "pseudorastellum" and the pin-like maxillary cuspules characteristic of at least some Kwonkan species. It is assumed that the male will have a relatively short and broadly based embolus and a cluster of retrolateral short spines on the palp tibia like that of these two genera. Some undescribed species oiAnaine also share these male features. However no such specimens can be unequivocally attrib¬ uted to Aname turrigera as none have been collected from the precise habitats of turrigera females. Nevertheless, the undescribed species and Aname turrigera, with further data may warrant erection of an inclusive new genus. On a cladistic analysis, Kwonkan and Yilgarnia would be placed as apomorphic derivatives from such a hypothetical genus. Natural history Geographic Distribution and Habitat The distribution of Aname turrigera ranges from west of Balladonia to Eyre Peninsula, in patchily distributed fa¬ voured habitats. Although it has been collected fromMallura north of the Nullarbor Plain and in coastal, lagoon-like flats of the Yalata Swamp at the Head of the Bight (southeast of the Nullarbor), it has not been observed on the Plain proper or on coastal areas immediately to the south, such as the Roe Plain. The species Aname turrigera occurs in semi arid habitats but in situations which are subjected to periodic sheet flooding or inundation, such as Melaleuca summer-dry swamps (e.g. Eyre Peninsula), mallee/spinifex associations (Eucalyptus / Triodia pungens ) sites which experience flooding especially during summer thunderstorms (e.g. west of Balladonia and Eyre Peninsula), flood-prone mulga (Acacia aneura ) swales 67 Journal of the Royal Society of Western Australia, 77 (3), September 1994 in the sand dune country north of the Nullarbor Plain (e.g. Mallura) and saline flats dominated by chenopod shrubs {e.g. coastal “lagoons" of Yalata Swamp at the Head of the Bight). Burrow structure and foraging behaviour Main (1993) recorded the burrow and characteristic aerial tube extending from it of Aname tiirrigera . The vertical burrows are shallow (up to about 15cm deep) and lined with a thin, stocking-like silk tube which is loosely attached to the burrow wall. The tube extends from the burrow as a soil- encased turret into the supporting scaffold of a shrub (e.g. chenopod or glass wort) or spinifex (Triodia ) grass tussock (Fig IL, 2 A). Turrets may be up to about 25 cm high and about 2 or 3 cm in diameter although the lumen is less than a centimetre. Occasionally shorter, and free standing, tubes have been observed (Fig 2B ). The first turrets I observed were mistaken for termite flight towers. Turrets supported by foliage often have thick, irregularly "lumpy" walls, which suggests that the nests are deepened after rain when the soil is wet and gobs of spoil are dumped from the mouth of the tube from where some soil slides down the outside wall as a paste, while at the same time the turret is gradually height¬ ened. As the soil dries out it would harden. Although bur¬ rows may be periodically flooded, the shrub-supported tubes (into which spiders can "escape") apparently persist and are heightened seasonally. However, free standing tubes probably collapse during heavy rain, as they never attain the height of those supported in foliage. As a consequence of the specialised burrow, which seems primarily associated with flood avoidance, the spiders have secondarily been exposed to a "new" foraging opportunity i.e. to catch prey amongst foliage (Main 1993). Presumably Aname turrigera catches prey from within reach of the tube entrance.David Hirst (pers, comm.)has observed and photo¬ graphed a spider sitting with anterior legs splayed in typical "sit and wait" foraging pose in a turret entrance (see Fig IM). It is also possible, as with some Aname species which forage at ground level, that Aname turrigera may sometimes "chase" prey (within the canopy) from an ambushing stance. Al¬ though a number of arboreal Australian mygalomorphs make tubes on tree trunks or against the boles of trees or stems of shrubs (see Main 1993), no other species have been observed to forage within a shrub canopy. Life history The reproductive phenology of Aname turrigera is not completely known. Two females with brood young have been collected. One from near Lock, S A on 8 May 1986 (BYM 1986/29) and the other on 24 May 1986, 24 km west of Balladonia Roadhouse (BYM 1986/159). Tlie mean carapace length of 10 specimens from a sample of brood young collected from the burrow of the latter was 1.72 mm. It is not known when males wander, but by comparison with other species which release young from the burrow in autumn/ winter, it is assumed that male wandering and hence mating would likewise be in association with autumn/winter rainy periods. Entrances of turrets are sometimes sealed with soil plugs during dry summer weather. Figure 2. Aerial tubes of Aname turrigera. A, Tube in spinifex tussock. Lock, S A. B, Free standing tubes at base of mulga tree, Mallura, W A. Scale bars = 5.0 cm. Figure 3. Map of south eastern Western Australia and Eyre Peninsula, South Australia, showing known distribution of Aname turrigera (excluding Ardrossan in Yorke Peninsula). 68 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Acknowledgements: I am grateful for facilities provided by the Zoology Department, University of Westeni Australia. A R Main collected the Mallura specimens and took the photograph of nests at the same site. The Yalata Aboriginal Community Idndly granted permission to travel through and collect spiders in the Yalata Aboriginal Reserve. 1 very much appreciated information which David Hirst (South Australian Museum) made available from his field observations. References Gray M R 1978 A. Systematics and distribution. [In Venoms of Dipluridae. Chapter 7, Gray M R & Sutherland S K]. In: Arthropod Venoms (ed S Bettini). Handbook of Experimental Pharmacology 48:121-126. Gray M1981 Getting to know funnel webs. Australian Natural History20:265- 270. Gray M R 1984 A guide to funnel web spider identification. The Medical Journal of Australia 141:837 - 840. Gray M R1988 Aspects of the systema tics of the Australian funnel web spiders (Araneae: Hexathelidae: Atracinae) based upon morphological and electrophoretic data. In: Australian Arachnology (eds. A D Austin & N W Heather) Australian Entomological Society, Miscellaneous Publication No. 5, Brisbane 113-125. Hogg H R1902 On some additions to the Australian spiders of the suborder Mygalomorphae. Proceedings of the Zoological Society of London 1902(2); 121 -142. Main B Y 1964 Australian Spiders. Jacaranda, Brisbane. Main B Y1972The mygalomorph spider ^enus Stanwellia Rainbow & Pulleine (Dipluridae) and its relationship to Aname Koch and certain other diplurine genera. Journal of the Royal Society of Western Australia 56 :100 -114. Main B Y1976 Spiders. Collins, Sydney. Main B Y1982 Notes on the revised taxonomic position of the Black Wishbone spider Dekana diversicolor Hogg (Mygalomorphae: Dipluridae). Journal of the Royal Society of Western Australia 65 :25 - 29. Main B Y 1983 Further studies on the systematics of Australian Diplurinae (Chelicerala: Mygalomorphae: Dipluridae): Two new genera from south western Australia. Journal of Natural History 17:923 - 949. Main B Y1985 Mygalomorphae. In: Zoological Catalogue of Australia, Vol. 3 (edDW Walton) Australian Government Publishing Service, Canberra 1-48. Main B Y 1986 Further studies on the systematics of Australian Diplurinae (Araneae: Mygalomorphae: Dipluridae): A new genus from south-west¬ ern Australia. Records of the Western Australian Mu.seum 12:395-402. Main B Y1993 From flood avoidance to foraging: adaptive shifts in trapdoor spider behaviour. Memoirs of the Queensland Museum 33 :599 - 606. Mascord R 1970 Australian Spiders in Colour. A H & A W Reed, Sydney. McKeown K 1962 Australian Spiders. Angus & Robertson, Sydney. Petrunkevitch A1942 A study of amber spiders. Transactions of the Connecti¬ cut Academy of Arts and Sciences 34:119 - 464. Raven R J1981A review of the Australian genera of the mygalomorph spider subfamily Diplurinae (Dipluridae: Chelicerata). Australian Journal of Zoology 29:321-363. Raven R J 1984a A new diplurid genus from eastern Australia and a related Aname species (Diplurinae; Dipluridae: Araneae). Australian Journal of Zoology, Supplementary Series 96:1 - 51. Raven R J 1984b A revision of the Aname maculata species group (Dipluridae, Araneae) with notes on biogeography. Journal of Arachnology 12:177 - 193. Raven R J 1985 A revision of the Aname pallida species-group in northern Australia (Anaminae: Nemesiidae: Araneae). Australian Journal of Zool¬ ogy 33: 377 - 409. 69 ■ r ; ^•‘ .t ■«' ■ ? '■ '^ 1-- p •. L . in *•, 4 In >. V. M •■►.,. • . m , V J “v - «:|J|i; ur-4.:■ ^ik ' >< vm4i^[,S^difs. -r' V ‘'i .i ‘K . ■ •* ''' >*•#€?• ■*• 1i^ yl:^* • v^ -■ ,,- -^A'" **' >■■’* '■ ■*^'''' '" *‘t^ ■’ ''■ ' ^ t •- 4 ■• fc t '* . ' ••-*•. t-iv-., >\%i;^, ;r4flp*iN>.'.V 'iu wi«i- '^■■^•-^,r-kJ‘>.‘?* »»->♦•’V I rrfl-^ %'? *■•■ j'v"' irf ^ •«•' ,^*4i' 'i)i^l^.,#^«! -iiiiV* ‘■'»‘*'^1pr- wstWek '''-^ ■i^*'-i ^ *^- ■ :»*k »■ ■* ■■ ■■•' .’J»Jhii 4;->M^‘5*>if.^-'*5'• ' ■ ■ '^^:-. ,.. • V ^,?^ ', 51 -f.?!; -^'•- ;r' ■?4■^^^. '-i" '. *. ' I • ■^■'>Nl'i^^yVs8<9i-* ••TWfi- s «,.* g. <|»i iW ^ 4Nb|^ . y > |ip . i 4i , ->.».■'».-ry .jU-»^p^»|||te.v:. . - ^^v^^^-,, J- ,•■ t ••i* \'ip-r*. " '' ■ ■ T - '*• A •,’>?«' . "S?' p* ' 4 Journal of the Royal Society of Western Australia, 77; 71-79, 1994 Wet heathlands of the southern Swan Coastal Plain, Western Australia: A phytosociological study R S Smith^'^ and P G Ladd^ ^ 23 King Road, Bunbury, Western Australia 6230 ^ School of Biological and Environmental Sciences Murdoch University, Murdoch, Western Australia 6150 Manuscript received March 1994; accepted June 1994 Abstract Wet heathlands, formed by sclerophyllous nanophanerophytes and graminoids growing on seasonally waterlogged soil, were surveyed on the southern Swan Coastal Plain and classified using two-way indicator species analysis (TWINSP AN). Three main wet heathland groups were identified, these being the Pericii/ymmfl - Hypocalymma, Pericalymma - Regelia and Melaleuca - Cassytha alliances respectively. The Pericalymma - Hypocalymma alliance is found north of the Capel River in depressions in the Bassendean Dune system on colluvial sands which have impermeable organic, clay or ferruginous horizons at about a metre. The Pericalymma - Regelia alliance is found south of the Capel River primarily on colluvial wet sands, and shallow sands and loams, in the Abba soil association. Variations in vegetation composition within this alliance are associated with the depth to the impermeable layer (usually lateritic) and the winter height of the perched water table. The Melaleuca - Cassytha alliance was found north of the Capel River mainly on heavy clays formed from Quartemary alluvium within the Serpentine River Association. These heathlands were species poor compared to Australian dry heathlands, with the number of species per 100 m^ ranging from 9 to 26. The ecology of these heathlands is discussed in the light of the limited information available. Their existence as fragmented islands within a primarily agricultural landscape makes them vulnerable to changes in ecosystem processes. Introduction Wet heathlands, which are plant formations dominated by evergreen sclerophyllous nanophanerophytes and graminoids growing on seasonally waterlogged soil (Groves & Specht 1965; Specht 1981), are widespread but of very limited area in the south west of Western Australia. Dry heathland, now generally termed kwongan, has been the subject of much scientific research in Western Australia over the last ten years (e.g. Brown & Hopkins 1983; Pate & Beard 1984; Bell «& Loneragan 1985; Brown 1989), but wet heathland remains largely undescribed in regard to its phytosociology and ecology. Undoubtedly this is because wet heathland in Western Australia is much less extensive and less species rich than dry heathland. Dry heathland or kwongan, most of which occurs on sandplain soils in the low to moderate (300-600 mm) rainfall areas of the South West Botanical Province, has been esti¬ mated to cover 118,000 km^ or 27% of the province (Beard 1984). No comparable estimate of the area of wet heathland, which is virtually confined to the high rainfall (> 800 mm) areas of the south-west, is available. Because of its limited and scattered occurrence, wet heathland has usually been mapped in conjunction with sedgeland and "swamp vegeta- Hon" (Smith 1973; Beard 1981). The dry heathland of the sandplains is extremely species rich at small (< 1 ha) sample sizes with an average of 60 species per 100 m^ in the central wheatbelt (Brown 1989) and 70 per 100 m^ in the lateritic uplands of the Mt Lesueur area (Hopkins & Griffin 1984). There are few published reports of species richness in Western Australian wet heathland. Wardell-Johnson et al. (1989) found an average of 21 species per 314 m^ plot in a heathland community on damp shallow sandy sites near Walpole on the south-western coast. How¬ ever, this study did not include herbaceous perennials and some annuals. Much of the vegetation which originally included wet heathland communities has been cleared for agriculture and the area of this vegetation type left in Western Australia would probably cover no more than 500 km^ with the most extensive areas being within the Warren Botanical Sub¬ district (Smith 1972). The wet heathland of the southern Swan Coastal Plain (SCP), which is the focus of this study, is primarily restricted to small conservation reserves and road verges surrounded by cleared farmland. The main objective of this study was to characterize the wet heathland communities of the southern SCP in terms of their characteristic species and to relate these communities to soil and other environmental factors. Identification of the community types provides a framework to plan the manage¬ ment and conservation of these communities, which despite their relative paucity of species provide a habitat for many of the rare and endangered plant taxa of the high rainfall areas of the South West Botanical Province (Keighery & Robinson 1990). © Royal Society of Western Australia 1994 71 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Regional Setting Geology, Geomorphology and Soils All study sites were on the southern part of the Swan Coastal Plain (SCP), Western Australia, between latitude 33*^ 00’ S and 33” 45' S (Fig 1). The SCP extends from the Darling and Whicher Scarps to the Indian Ocean and to about 60 m above sea level. The plain is underlain by the Phanerozoic sediments of the Perth Basin and several landform units lying parallel to the coastline and closely related to the geology can be identified (Wilde & Walker 1982). At the foot of the Darling Scarp is a zone comprised of coalesced colluvial fans and the remnants of a strandline deposit. This zone merges into a 10 km-wide alluvial plain (often called the Pinjarra Plain), which has been lateritized and then exten¬ sively stripped to form soils which are predominately meadow podzolic in nature (Mulcahy 1973). The meadow podzols consist of a sandy surface overlying a poorly struc¬ tured clay of low permeability developed on a lateritic pallid zone. Along the streams which cross the plain, younger deposits in the form of terraces are incised into the meadow podzols and in places alluvial fans overly them. These younger soils are red and yellow podzolics and undifferentiated soils. Figure 1. Wet heathland vegetation survey areas on the southern Swan Coastal Plain. 1. Riverdale Nature Reserve, 2. Reserve No. 20331,3. Wellard Nature Reserve, 4. Guthrie Forest Block, 5. Byrd Swamp Nature Reserve, 6. Hay Park Recreation Reserve, 7. Reserve No. 1167,8. Ruabon Nature Reserve, 9*Ruabon (railway reserve), 10. Williams Road (railway reserve), 11. Tutunup (railway reserve), 12. Yoongarillup Reserve, 13. Fish Road Reserve, 14. Ambergate Nature Reserve. Further west, at least three generations of dune soils, overlying the fluviatile deposits, are evident. At the present coastline is the youngest dune system, the Quindalup Sys¬ tem of McArthur & Bettenay (1960), which consists of highly calcareous shell sand deposited during the Holocene. Inland of this system lies a belt of slightly podzolized yellow sand overlying limestone called the Spearwood System (McArthur & Bettenay 1960). The Spearwood sands are almost entirely quartz although there are localized deposits of heavy min¬ eral sands. Lying between the Spearwood System and the alluvial plain is the Bassendean System (McArthur & Bettenay 1960), composed of the oldest aeolian deposits, which have now lost their dune morphology. It is of lower relief than the younger dunes nearer the coast and the soils are highly leached and podzolized white quartz sands with B horizons of accumulated iron and organic matter. The depressions between these low dunes are filled with swamp and lacustrine deposits which may be peat, peaty sand, sand or clay (Wilde & Walker 1982). Details of the soil associations of the three dune systems and the alluvial plain for the southern SCP are given by McArthur & Bettenay (1956,1958), Pym & Poutsma (1957), McArthur (1958) and Tille & Lantzke (1990). Climate The climate of the area is "mediterranean" with cool wet winters and warm to hot, dry, summers (Gentilli 1972), Proximity to the coast provides a moderating influence and frosts are infrequent, with mean temperatures of the coldest and hottest months being 11-13”C and 25-29” C. Mean annual rainfall ranges from about 850 to 1000 mm through the study area with the higher levels occurring near the scarps. The pasture growing season is seven months, from April to October inclusive, and during this period the excess of rainfall over potential evapotranspi ration of pastures is 560 mm at Harvey (Pym & Poutsma 1957) and 380 mm at Bunbury (McArthur & Bettenay 1956). This excess of water is available for run-off, storage in the sub-soil or penetration to the underlying aquifers. The fiat topography of much of the coastal plain ensures that there is little run-off and where the subsoils are relatively impermeable much of this excess is probably stored in the subsoil and therefore is available for growth of deep rooted native species beyond the period of pasture growth (McArthur & Bettenay 1956). Methods Data collection The fourteen reserve and roadside sites containing areas of wet heathland were visited, usually at least twice, in spring 1992 and 1993. At each site between three and eight 10 m X 10 m quadrats were sampled. Within each quadrat, all vascular plant species were either identified in the field or given a code number and collected for later identification. Thenomenclature of Green (1985) and Marchantef a/. (1987) was used. Each species was given a cover value between 1 and 5 on a modified Braun-Blanquet scale: 1, < 5%; 2,5-25%; 3,25-50%; 4,50-75%; 5,75-100%. At least one soil pit was dug at each site down to 1 metre, or to a limiting horizon. Soil texture of each horizon was recorded and samples were taken within each horizon for determination of the Munsell colour. Published soil surveys were also used to determine the soil type at each site and note was taken at the time of survey of the depth of watertable at each site. 72 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Data analysis Vegetation data were classified by two-way indicator species analysis using the polythetic divisive computer pro¬ gram TWINSPAN (Hill 1979) and the Braun-Blanquet cover value for each species. Only the 116 species which were found in at least two quadrats were used in the analysis. In general, classification was terminated at the third level of division to give three groups (alliances) and six subgroups (associations) of the 47 quadrats; a further sub-group was recognized at the fourth level of division. Groupings of less than four quadrats at the fourth level of division were not recognized. The taxonomic composition of the spedes groups was compared using the Sorenson coeffident of community (Gauch & Whittaker 1972). Detrended Correspondence Analysis (DCA; Hill & Gauch 1980) was used to ordinate the 47 quadrats so their relationship with identified edaphic variables could be examined. Results Vegetation and soils The TWINSPAN classification produced three main groups of wet heathland species, which are termed alliances in this paper and seven sub-groups or associations (Appendix I). Two of the groups, the Pericalymma- Hypocalymma alliance, with two associations, and the Pericalytmm-Regelia alliance, with three associations, contain the shrub Pericalymma ellipticum as one of the characteristic species. The other main wet heathland group, the Melaleuca- Cassytha alliance which is divided into two associations, does not contain P. ellipticum but is usually dominated by shrubs of Melaleuca spp., the most characteristic of which is M. incana. The climber Cassytha glabella is also a characteristic component of these heathlands. In general the Pericalymma- HypocaIy 7 iima alliance is found on deep sands (humus podzols), with organic depositional layers at 1 to 1.5 metres, north of the Capel River and the Pericalymma-Regelia alliance is found on a range of soils from deep sand to shallow loams, usually with a laterite hardpan at less than a metre, south of the Capel River. The Mclaleuca-Cassytha alliance is found on deep clays or sand over clay north of the Capel River. The typical soils for each of the associations are summarized in Table 1. Similarity of alliances, associations and quadrats The Sorensen coefficients of community of the various alliances and associations are given in Table 2. This coeffi¬ cient, which has a maximum value of 200 when the two samples being compared have all their species in common. Table 1 Soil descriptions for wet heathland communities of the southern Swan Coastal Plain. Association Soil Description p£ricalymma-Hypocalumrtia alliance P- elUpticum-H. angustifolium-Hibbertia naginata [Al] Deep (> 1 m) light grey’ sand sometimes with an organic deposition hardpan at 1 to 1.5 m. High organic matter content in Al horizon [Uc 2.33p. Top soiP-C/N ratio: moderate-high (23-27), P (total): very low (20-30 ppm), pH: strongly acid (5.0-5.5)^. P- ellipticiim-Evattdra \nuciflora~Hypo\aena exsulca [A2] Deep (> 1 m) light grey to greyish brown sand sometimes with an organic deposition or lateritehardpanatl tol.5m[Uc2.31]. High organic matter content in Al horizon. Topsoil- C/ N ratio: moderate-high (23-27), P (total): very low (20-30 ppm), pH; strongly acid (4.9-5.5). ^Slicalumma-Regelia alliance P- ellipticum-R. ciliata-Leptocarpus [Bl] Pale brown, brown or grey sand to sandy loam over laterite hardpan at 0.25-1 m [Uc 2.21]. Top soil-C/N ratio; low-moderate (18-21), P (total): very low-low (40-70 ppm), pH: moder alely acid (5.4-6.1). ellipticum-Chamelaucium ^^ycei-Crevillea diversifolia [B2] Dark reddish brown loam to greyish brown or brown sandy loam over laterite hardpan at 0.1 to0.5m [Urn 5.21/Uc5.11]. Top soil-C/N ratio: low (16-18), P (total) very low to high (40- 400 ppm), pH: moderately acid (5.4-5.9). P- ^llipticum-Kunzea recurva-Daviesia P^^issii [63] Deep (> 1 m) light grey fine sand to grey brown sandy loam over laterite caprock at 1 to 1.5 m [Uc 2.21]. High organic matter content in Al horizon. Top soil-C/N ratio moderate to high (23-28), P (total) very low (40-50 ppm), pH moderately acid (5.5-5.9). ^^^^Okidca-Cassutba alliance ^^^iminea-Isolepsis jwdosa-C. glabella Deep (> 1 m) grey ish brown to light yellowish brown heavy clay [Ug 5.141. Top soil: C / N ratio very low-low (9-17), P (total) low-moderate (40-120 ppm), pH strongly-moderately acid (4.8- 6.2). ^^cana-M. hamulos£-K. recurva [C2] rr -- Deep (> 1 m) grey brown sand overlying very pale brown sandy clay [Uc 1.21]. Top soil: moderately acid (5.5-6.0), may have high salt content. !^>1 colour according to Munsell colour charts. ^ Northcote (1975) soil classification. ^ Soil chemical data from Pym & Poutsma (1957), ^^Arthur & Bettenay (1956,1958), McArthur (1991) and Smith 1994. 73 j umal of the Royal Society of Western Australia, 77 (3), September 1994 in composition of the species ,ri 7 es the similarity ^ heathland alli- fnces have the , j) the similarity between alli- Sland and h-*^teIativelylow^^^^^^ ancesbasedonshared^ea ^ 22. This indica es "efficientsofcomm^ty^^^ qnite low species a high degree ^.^^f^^nces, the highest coeffiaent is 49 richness, ^j^^n 32. and it is generally of the Pericalymtna-Hypocalymma alliance. The third axis (not shown) divided the P. ellipticum-R. ciliata-L canus and P. ellipticurn-K. recurva~D. preissii association quadrats at one end (high values) from the P. ellipticum-C. roycei-G. diversifolia association (low values) with the Pericalymma-Regelia alli¬ ance quadrats occupying intermediate values. Unfortunately, because soil data were not recorded from each quadrat, it is not possible to more precisely relate values on the DC A axes to edaphic variables. Table 2 Total species nun^ and Sorenson coefficients of community of the wet heathland alliances and associations. Alliance A Pericalymma-Hypocalymma (56 spp.) B Pericalymma-Regelia (.no spp^ CMelaleuca-Cassyiha (39spp.)_ B 21.6 C 16.8 17.6 —---— A1 A2 B1 B2 B3 Cl C2 Association 49.2 21.7 12.3 18.9 13.6 3.9 A1 (33 spp-) 21.0 7.9 23.2 14.8 4.3 A2 (28 spp.) _ 29.2 30.6 12.9 7.0 B1 (96 spp.) 18.0 8.1 6.1 B2 (48 spp.) _ 14.9 10.2 B3 (41 spp.) - 31.8 Cl (26 spp.) C2 (18 spp.) r c nf the first four axes of the DCA were The eigenvalues of the h 0.74, .Uiancs and reinforced .he quadrats ccifi ration (Fig 2). It also confirmed the outcome of the cla ^ p,ricflymma-Hypocalymma and greater y g^j-h other than each has Pericalymina-Reg alliance The values along the „i.h .he VP- ■■ r'MSdsUO The hlghes. values on axislcorrespondwi.h heathlands tes. me g Se^eTwthslightlylowervalues associated with thedeep reserve , 6 values on both axis 1 and rx^aTaStediththeshallowsandyloams^^ S the Tutunup-Ruabon area. An increase m value on axis 2 appears to coLspond with an increase m soil depth and fncrLsing sandiness of the soil, with the highest values occurring at reserve T1167 which has deep grey sands, but which alL has a higher cover of sedges than the other sites Axl«1 Figure 2. Detrended Correspondence Analysis ordination (axes 1 & 2) of 47 wet heathland quadrats from the southern Swan Coastal Plain showing the three alliances derived from the TWINSPAN classification, m Pericalymma-Hypocalymma aWiance, • Pericalymma- Regelia alliance, ^ Melaleuca-Cassytha alliance. The Pericalymma-Hypocalymma alliance. The P. ellipticurn-Hypocalymma angustifolium-Hibbertia vaginata as¬ sociation (association A1 in Appendix I, Table 1) is found on the deep acid grey sands series and related Swamp types) of the Bassendean Association (Pym & Poutsma 1957) in the northern part of the study area. These soils have an organic B horizon, which may be concreted to form a hardpan, at 1 to 1.5 metres. It is probably the most extensive of the wet heathland formations on the southern SCP being found in State Forest on the east side of the McLarty and Myalup Plantations and in nature reserves in the Kemerton area. The wetlands in which this heathland association are found occur as regularly spaced depressions, or interdune swales, and are classed as sumplands (seasonally inundated basins) of the Riverdale suite by Semeniuk (1988). Recharge of the sumplands is by precipitation or groundwater rise. Structur¬ ally, the associationis mid-dense to closed heath land (Walker 1983) dominated by shrubs up to about 1 m with varying amounts of graminoids up to abou 125% of ground cover and in some places emergent Xanthorrhoea preissii, Melaleuca preissiajia or Nwyfsm floribunda. The heathlands grade into Eucalyptus niarginata~E. calophylla open forest on low sandy rises, M. rhaphiophytla and M. preissiana low forest in wetter areas, and E. gomphocephala-E. marginata forest on the sands of the Karrakatta Association to the west. The P. elliplicum-Evatidra pauciflora-H. exsulca association (A2) is found on similar soils and topographical positions to the association described above, and they have many species in common. The main difference is the greater dominance of P. ellipiicum and the larger proportion of the Cyperaceae and Restionaceae in the P. ellipticum-E. pauciflora-H. exsulca asso¬ ciation which may indicate that the sites are somewhat wetter. This association is found in State Forest west of 74 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Harvey, and on several small reserves in the same area and also south of Bunbury. In the Harvey area, the soils are of the Joel series and related types of the Bassendean Association, and in the Bunbury-Capel area they are the Swamp series VI of the Southern River Association (McArthur and Bettenay 1956). The presence of organic or ferruginous hardpans at about 1 metre as found in some of these soils may account for the increased proportion of graminoids characteristic of this association. Structurally, the association is a mid-dense to closed heathland, generally below 1.2 metres tall, with occa¬ sional emergent Kunzea ericifolia or M. preissiana. It may grade into a low forest of M. preissiam or a sedgeland dominated by £. paiiciflora in welter areas and into £. margimta-E. calophylla forest or K. ericifolia tall shrubland or Banksia attenuata woodland on interswale dunes. The Pericalymtna-Regelia alliance. The P. ellipticurn- Regelia ciliata-Lcptocar}nts canus association (Bl) is the most widespread group within this alliance, being found on the Wet Sand soil type of the Abba Wet Ironstone Flats land unit (Tille & Lantzke 1990). These acid grey sands are similar to the Joel Sand of the Bassendean Association and have a similar origin. They may grade into the Bog Iron Ore Sand of Tille and Lantzke (1990) when the laterite hardpan comes to within less than 50 cm of the surface. The heathlands of this association occupy a similar position topographically to the Joel Sand of the Bassendean Association being found in shallow depressions and swales in areas of sandplain and low dune fields. The association is restricted to road verges and three small (<200 ha) nature reserves between Tutunup and Ambergate south of Busselton. The structure of the mature formation is mid-dense to closed heathland which varies in average height from 0.6 to 1,2 m but may be up to 1.6-1.8 m (tall heathland) in some areas. The species richness of the shorter heathland, particularly that at the Ruabon Nature Reserve and part of the Fish Road Nature Reserve, is only about half of that of the taller heathland. In other areas, with deeper soils, species characteristic of the association such as Melaleuca uncimta, M. hanwlosa, Hakea varia, R. ciliata may grow to more than 2 metres high. At Tutunup the association merges quite sharply with £. calophylla forest, which has an almost completely different suite of species in the understorey, and is situated on the low rises interspersed with the heathlands. The soils on these rises are deep (1.5-2.0 m) Abba and Busselton sands (Tille & Lantzke 1990). The P. ellipticum-Chamelaucium roycei-Grevillea diversifolia association (B2) is perhaps the most restricted unit in the Pericalymma-Regelia alliance. It is confined to shallow loams over laterite hardpan along a road and rail reserve in the Tutunup area southeast of Busselton and covers less than 10 hectares. The soils are the Bog Iron Ore Loams of the Abba Wet Ironstone Flats land unit (Tille & Lantzke 1990) with less than 30 cm of brown loam overlying a massive laterite hardpan. This association merges with the P. ellipticum-R. ciliata-L canus association which occurs where the laterite is deeper than about 30 cm and the loam gives way to loamy sand and sand. It is wetter than the latter association and sometimes water stands several centimetres deep above the soil surface over the winter months. The drainage along the verge has been modified by road and railway construction, the latter taking place over 120 years ago. The declared endangered species C. roycei is characteristic of this associa¬ tion. Several undescribed taxaincludingspedesofLexocflri/fl, Restio and Dryandra are also found. Structurally the associa¬ tion is closed heathland to tall closed heathland (Walker 1983) with emergent Viminariaderiudata,Xanthorrhoeapreissii, and various Melaleuca species up to 3 m high. The P. ellipticum-Kunzea recurva-Daviesia preissii associa¬ tion (B3) differs from the other two associations within the alliance mainly in regard to the taxa that it does not have. Regelia ciliata is replaced by a similar myrtaceaous shrub Kunzea recurva and it does not have the rich cover of species in the genera Lepyrodia, Leptocarpus, Restio and Loxocarya found in the other associations. It also has a restricted distri¬ bution, being found at Ambergate and Yoongarillup on small nature reserves. At Ambergate the soils are of the Wet Sand or Busselton type (Tille & Lantzke 1990) with laterite hardpan at about 1 m, at Yoongarillup the soils are brown sandy loams over sandy clay loams (Mixed Alluvial Soils; Tille & Lantzke 1990). The P. ellipticum^K. recurva-D. preissii association is also a mid-dense to closed heathland with occasional emergent X. preissii, grading into £. marginata- E. calophylla forest on the slight rises surrounding the shal¬ low depressions in which the heathland is found. The Melaleuca-CassythaaWiance.The Melaleuca viminea- Isolepsis nodosa-Cassytha glabella association (Cl) is found in the northern part of the study area on small remnants of uncleared heavy kaolinitic clays of the Serpentine River soil association (Pym & Poutsma 1957). The very low permeabil¬ ity of these clays and the level topography restricts drainage and water lies on the surface for several months over winter. They occur near the boundary between the Bassendean Dunes and the alluvial Pinjarra Plain and the soils were formed by the build-up of alluvium from streams terminat¬ ing in the Bassendean Dunes (Semeniuk 1988). It forms a closed heathland at the Wellard Reserve and a tall heathland (1.5-2.5 m) at Byrd Swamp Reserve, with up to half of the ground cover being provided by sedges and restiads {Isolepsis, Lepyrodia, Leidocarpus, Gahnia) in some areas. Climbers, pri¬ marily C. glabella, are also prominent. On the low sandhills surrounding the heathland association occurs or Kunzea ericifolia low forest or £. marginata-E. calophylla forest with an understorey of Banksia species. In wetter areas M. rhaphiophylla low forest with an understorey including M. lateritia, Asfartea fascicularis and Lepidosperma longitudinale may be found. Further south, near Bunbury, the Melaleuca incana-M. hamulosa-Kunzea recurva association occurs on deep grey sand over sandy clay. This soil is similar to the Stirling VII Sand and Swamp Series IV of McArthur & Bettenay (1956). Although it is also alluvial in origin, it is much more perme¬ able than the clays of the M. vitninea-L nodosa-C. glabella association, however during winter the watertable may rise above the ground surface. This association is represented by only one site, within the city of Bunbury, although it prob¬ ably occurs in other areas south of Bunbury near the junction of the Bassendean and Spearwood dune systems. It is similar structurally to the tall heathland variant of the M. vitninea-L nodosa-C. glabella association found at Byrd Swamp but with the sedge Gahnia trifida and the native grass Sporobolus virginicus forming the graminoid component rather than Isolepsis and Lepyrodia species. On better drained soil to the east of the heathland occurs a low woodland of Melaleuca preissiana. Journal of the Royal Society of Western Australia, 77 (3), September 1994 Discussion Comparisons with other heathlands: Species richness The Pericalj/ftnna-Regelia and Pericalymrna-Hypocalymma wet heathland alliances of the southern Swan Coastal Plain have a comparable species richness to similar communities in New South Wales and Queensland. At the 100 m’ scale, wet heathland in Royal National Park and Ku-ring-gai Chase near Sydney averaged 23.9 and 18.0 species respectively (Specht & Specht 1989) compared to 15.8 (up to 21 /100 m^) for the Pericalymma-Hypocalymma alliance and 18.4 (up to 26/ 100 m^) for the Pericalymma-Regelia alliance. Mowever the southern SCP alliances are considerably less species rich than other coastal heathlands on sand in Western Australia. At Nomalup National Park, wet heathlands growing on similar soils had 34 species per 50 m^ (George et al 1979) and at Scott River a low open heathland on deep grey sand had 40 species per 100 m^ (Specht & Specht 1989). The low species richness of the Melaleuca-Cassytha alliance and of some of the wetter areas of the other two southern SCP communities (9- 12 species /100 m^) is similar to that of the species poor heathlands of northern Europe (Gimingham et al. 1979). Comparisons with other heathlands: Ftoristics There are few published reports of the floristics of other wet heathlands in Western Australia, and none of these has attempted a classification of this community type. Wardell- Johnson et al. (1989) describe two wet heathland communi¬ ties growing on humus podzols over deep sands at Nomalup near the south coast of Western Australia, called the Beaufortia sparsa plain and Dasypogon brorneliifdlius heathland commu¬ nities. These communities share Adenanthos obovatus, Lysinema ciliatum, Melaleuca thyinoides, and D. bromeliifolitis with the Pericalymma-Hypocalymma alliance. Havel (1968) describes a low M. preissiana woodland growing within the Bassendean Dune system near Perth. Apart from M. preissiana, this community hasP. ellipticiwi, A. oboimtus, D. brotneliifolius and H. angtistifoliuw in common with the Pericalymma- Hypocalymma alliance. Keighery & Trudgeon (1992) describe several wet heathland communities growing on the SCP Table 3 Numbers and percentage (in brackets) of species in each of the major plant families in the southern Swan Coastal Plain wet heathlands. Only those species which were positively identified have been included. Family Pericalymma-Hypocalymma Pericalymma-Regelia Melaleuca-Cassytha Myrtaceae 8 (15%) 11 (12%) 11 (31%) Proteaceae 3 (6%) 15 (21%) 3 (8%) Leguminosae 7(14%) 8 (8%) 5 (14%) Restionaceae 6 (12%) 11 (12%) 3 (8%) Cyperaceae 3 (6%) 3 (3%) 4 (11%) Liliaceae 4 (8%) 5 (5%) 1 (3%) Epacridaceae 3 (6%) 1 (1%) - Stylideaceae 1 (2%) 5 (5%) - Dilleneaceae 2 (4%) 4 (4%) - Droseraceae 2 (4%) 3 (3%) 1 (3%) Orchidaceae 2 (4%) - 3 (8%) near Perth, which are dominated by P. ellipticum and which, in common with the Pericalymma-Regelia alliance, include P. ellipticum, Hakeasulcata, Xanthonhoeapreissii, Stirlhigia latifolia, Calothamnus lateralis and the sedges Mcsomelaena tetragona and Hypolaena exsidca. Keighery & Robinson (1990) mention wet heathlands growing on red clays and loams over laterite hardpan on the Scott Plains east of Augusta, which share a number of rare taxa with the Pericalymma-Regelia alliance. The Melaleuca-Cassytha wet heathland alliance has taxo¬ nomic affinities with low woodlands and low closed forests dominated by M. rhaphiophylla and M. ciiticularis around swamps and on seasonally wet areas of the Yoongarillup Plain landforrn south of Perth (Trudgeon 1991; R Smith pers. obs.). The remnant wet heathlands growing on the northern part of the SCP and on the Scott Plains are apparently species rich, and the bulk of the taxa occurring in them are different from those in the communities which were the subject of this study. Tliere is a clear need for a formal classification of these communities because they are threatened by urban develop¬ ment and mining (Keighery & Robinson 1990; Gibson & Keighery 1992). A comparison of the most important plant families within each of the southern SCP wet heathland alliances (Table 3) shows a very high proportion of Myrtaceae in the Melaleuca- Cassytha alliance. The percentage of Restionaceae in the Pericalymma-Hypocalymma and Pericalymma-Regelia alliances is about double that of the 25 mostly dry heathland sites surveyed in the south west of Western Australia by George et al. (1979). Except for the Pericalymma-Regelia alliance, the proportion of Proteaceae in the SCP heathlands is lower than most Western Australian dry heathlands (George et al. 1979; Brown & Hopkins 1983) where it is generally above 15%. As with most Australian heathlands, there was a low propor¬ tion of species of the Epacridaceae in the flora of the southern SCP heathlands and where they did occur they had low foliage cover values. This is in contrast to the wet heathlands of Europe and South Africa where the closely related family Ericaceae is often a dominant component of the community (Gimingham et al. 1979; Cowling & Holmes 1992). 76 Journal of the Royal Society of W estern Australia, 77 ( 3 ), September 1994 Ecological processes Only some general comments can be made on ecological processes within the wet heathlands of the SCP as there has been little ecological research in these communities. Clearly the soil moisture regime is critical in determining the extent of the wet heathland formation and in modifying plant species composition within the formation. The Perkalymma- HypocaI\/mma and Pericali/WTua-Regelia alliances are similar to the groundwater heaths described for eastern Australia by Groves & Specht (1965) and Siddiqi et ai (1972) which are subjected to extremes in soil moisture availability. In the rainy season the watertable is close to the surface for several months, while during the dry season the soil above the impermeable layer of clay, peat, or laterite hardpan under¬ lain by clay, drys out. As plant roots are generally not able to penetrate this layer, water deficits become severe in sum¬ mer. The rapid drying out of these soils was illustrated by a study of watertable levels along a transect from heathland on shallow loam over laterite hardpan to woodland on deep sands at Tutunup (Smith 1994). In late August and early September, the watertablc was 3 cm to 5 cm above the surface in the heathland and 5 cm to 14 cm below the surface in the woodland. By late October the watertable in the heathland had withdrawn to within or below the laterite (50 cm below the surface) and in the woodland the watertable was greater than 85 cm below the surface. As well as the stress caused by the drying out of the relatively shallow soils in summer, high watertables in win¬ ter cause additional stress to the heathland species. Waterlogging by perched watertables causes soil air, which is essential to root respiration, to be expelled from the soil (Specht 1981). The duration of waterlogging is important in determing the species composition of wet heathlands (Webster 1962; Siddiqi et al 1972). The seedlings of many sclerophylious trees are very sensitive to prolonged waterlogging (> 2-3 months) and thus fail to establish in adjacent heathland soils whereas the seedlings of heathland plants generally have morphological or physiological adap¬ tations which alleviate the anaerobic conditions (Specht 1981). Few nitrifying or nitrogen-fixing bacteria can exist under the poorly aerated conditions of waterlogged soils (Woolhouse 1981) and therefore these soils may be low in available N. Inhibition of the nitrification process (Groves 1981) is indicated by the high C/N ratios (23-27) of the podzols of the Perkalymma-Hypocalymma alliance (Table 2). Leaching of the soils in wet heathlands may also lead to a reduction in some cations, notably Ca"\ Mg^' and K\ Re¬ search into the ecology of European wet heathlands has shown that nutrient-poor, waterlogged sites act as refugia for slow growing plants which are not able to compete well on better aerated, fertile soils (Berendse & Aerts 1984). A comparison of soil nutrients in the Bog Iron Ore and Wet Sand soils under wet heathland at Tutunup with those in the Abba Sand under Eucalyptus marginata~E. calophylla forest adjacent to them shows the heathland soils to have similar levels of total N and total P, higher extractable K and Na but substantially lower extractable Ca and Al (Smith 1994). Some of the Bog Iron Ore loams have quite high levels of total P (up to 400 ppm) but this nutrient is probably strongly bound by iron oxides and extractable P is low (McArthur 1991). The low total P concentrations of most of the wet heathland soils of the southern SCP (Table 2) are typical of Australian heathlands (Groves 1981; Keith & Myerscough 1993). Disturbance in the wet heathlands Most of the wet heathlands surveyed in this study occur on reserves of less than 200 ha or on narrow road reserves, and are surrounded by cleared agricultural land or pine plantations. Fragmentation of these areas of natural vegeta¬ tion may have major effects on various ecological processes, such as the water and nutrient cycles, and especially in the smaller fragments, the radiation balance and wind regimes (Hobbs 1993). Clearing of the surrounding native vegetation may lead to a rise in the watertable and consequent salt accumulation near the surface, as has occurred on the lower SCP since at least the 1950's (McArthur & Bettenay 1956). Fertilizer drift from adjacent farmland may disrupt the natu¬ ral nutrient cycle and encourage the invasion of exotic plants (Smith 1990). Too frequent fire is another stress which may lead to the deterioration of natural vegetation in ecosystem fragments. Although most of the sites surveyed for this study had apparently not been burnt for 8-10 years, much of the heathland along the Tutunup-Ruabon railway was recently burnt, and was burnt for fuel reduction every few years in the past (F N^gus personal cojumunkation). The vegetation of the wet heathlands recovers quite rapidly after fire due to the high proportion of resprouters (R Smith personal observation). However the post fire environment with its abundant light and increased nutrient levels in the ash bed provides an ideal seed bed for invasive weed species which may out-compete slower growing native species (Baird 1977). The combina¬ tion of recurrent fire and rising salinity levels may place extra stress on the recovery of species which normally require establishment of seedlings after fire (Baird 1984). This survey provides a systematic analysis of a commu¬ nity which is likely to be one of the rarer components of the Western Australian vegetation. If biodiversity is to be an important criterion for conservation of this vegetation type in this State, then the high species turnover (beta diversity) of the heathlands needs to be taken into account. Several of the sites surveyed for this study are not in conservation reserves, in particular the highly diverse heathlands of the Tutunup area. To preserve what is left of the diversity of this ecosystem, all of the sites included in this study need to be in conservation reserves where they can be more easily man¬ aged. More detailed work needs to be completed on the northern part of the coastal plain and along the south coast to provide a complete picture of the wet heathlands of Western Australia. References Baird A M 1977 Regeneration after fire in King's park. Western Australia. Journal of the Royal Society of Western Australia 60:1-22. Baird A M 1984 Observations on regeneration after fire in the Yule Brook Reserve near Perth, Western Australia. Journal of the Royal Society of Western Australia 66:147-162. Beard J S 1981 Vegetation Survey of Western Australia, 1:1 000 000 Series. Explanatory notes to Sheet 7. University of Western Australia Press, Nedlands. 77 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Beard J S1984 Biogeography of the kwongan. In: Kwongan, Plant Life of the Sandplain (ed J S Pate & J S Beard) University of Western Australia Press, Nedlands, 1-26. Bell D T & Loneragan W A1985 The relationship of fire and soil type to floristic patterns within heathland vegetation near Badgingarra, Western Aus¬ tralia. Journal of the Royal Society of Western Australia 67:98-108. Berendse F & Aerls R1984 Competition between Erica tetralix L and Molinia caerulea (L.) Moench as affected by the availability of nutrients. Acta Oecologia/Oecologia Planla 5:3-14. Brown J M 1989 Regional variation in kwongan in the central wheatbelt of Western Australia. Australian Journal of Ecology 14:345-355. Brown J M & Hopkins A J M1983 The kwongan (sderophyllous shrublands) of Tutanning Nature Reserve, Western Australia. Australian Journal of Ecology 8:63-73. Cowling R M & Holmes P M 1992 Flora and vegetation. In: The Ecology of Fynbos. Nutrients, Fire and Diversity (ed R M Cowling) Oxford Univer¬ sity Press, Oxford, 23-61. Gauch H G & Whittaker R H 1972 Comparison of ordination techniques. Ecology 53:868-875. Gentilli J 1972 Australian Climate Patterns. Nelson Academic Press, Mel¬ bourne. George A S, Hopkins A J M & Marchant N G1979 The heathlands of Western Australia In: Ecosystems of the World Vol 9A, Heathlands and Related Shrublands (ed R L Specht) Elsevier Scientific, Amsterdam, 211-230. Gibson N & Keighery G1992 Plant communities of the northern Swan Coastal Plain-with special reference to uncommon and potentially rare plant communities. In: Bushland in our Backyard. Proceedings of a workshop held by the Wildflower Society of Western Australia. Gimingham C H, Chapman S B & Webb N R1979 European heathlands. In: Ecosystems of the World; Heathlands and Related Shrublands Vol 9A (ed R L Specht) Elsevier Scientific, Amsterdam, 365-409. Green J W1985 Census of the vascular plants of Western Australia. Western Australian Herbarium, Department of Agriculture, Perth. Groves R H1981 Heathland soils and their fertility status. In: Ecosystems of the World; Heathlands and related shrublands Vol 9B (ed R L Specht) Elsevier Scientific, Amsterdam, 143-150. Groves R H & Specht R L1965 Growth of heath vegetation; I. Annual growth curves of two heath ecosystems in Australia. Australian Journal of Botany 13:261-80. Havel J J 1968 The potential of the northern Swan Coastal Plain for Pm«s pinaster Ait plantations. Bulletin No. 76, Forests Department of Western Australia, Perth. Hill MO 1979 TWINSPAN-A FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and attributes. Ecology and Systematics, Cornell University, Ithaca, New York. Hill M O & Gauch H G1980 Detrended correspondence analysis: an improved ordination technique. Vegetatio 42:47-58. Hobbs R J1993 Effects of landscape fragmentation on ecosystem processes in the Western Australian wheatbelt. Biological Conservation 64:193-201. Hopkins A J M and Griffin E A1984 Floristic patterns. In: Kwongan, Plant Life of the Sandplain (eds J S Pate & J S Beard) University of Western Australia Press, Nedlands, 69-^. Keighery G and Robinson C 1990 A survey of declared rare flora and other plants in need of special protection on the Scott Plains. Unpublished report to the Australian National Parks and Wildlife Service, Endangered Species Programme. Keighery B J & Trudgeon M E1992 Remnant vegetation on the alluvial soils of the eastern side of the Swan Coastal Plain. Department of Conservation and Land Management, Western Australia; unpublished report to the Australian Heritage Commission. Keith D A & Myerscough P J1993 Floristics and soil relations of upland swamp vegetation near Sydney. Australian Journal of Ecology 18:325-344. Marchant N G, Wheeler J R, Rye B L, Bennett E M, Lander N S & McFarlane T D 1987 Flora of the Perth Region. Parts 1 and 2. Western Australian Herbarium, Department of Agriculture, Perth. McArthur W M 1958 Further investigatiorvs of the soils of the Harvey and Waroona areas, WA. CSIRO, Division of Soils, Division Report 4/58. McArthur W M 1991 Reference soils of south-western Australia, (eds D Johnston & L J Snell) Western Australia Department of Agriculture, Perth. McArthur W M & Bettenay E 1956 The soils and irrigation potential of the Capel-Boyanup area. Western Australia. CSll^O, Division of Soils, Soils and Land Use Scries No. 16. McArthur W M & Bettenay E 1958 The soils of the Busselton area. Western Australia. CSIRO, Division of Soils, Divisional Report 3/58. McArthur W M & Bettenay E1960 The development and distribution of the soils of the Swan Coastal Plain, Western Australia. Australian CSIRO, Soil Publication 16. Mulcahy M J1973 Landforms and soils of southwestern Australia. Journal of the Royal Society of Western Australia 56:16-22. Northcote K H, Hubble G D, Isbell F F, Thompson C H & Bettenay E1975 A description of Australian Soils. CSIRO, Australia. Pate J S & Beard j S1984 Kwongan, Plant Life of the Sandplain. University of Western Australia Press, Nedlands. Pym L W & Poutsma T 1957 Soils and land use in the Harvey area. Western Australia. CSIRO, Division of Soils, Soils and Land Use Series No. 20. Semeniuk C A 1988 Consanguineous wetlands and their distribution in the Darling System, southwestern Australia. Journal of the Royal Society of Western Australia 70:69-87. Siddiqi M Y, Carolin R C & Anderson D J1972 Studies in the ecology of coastal heath in New South Wales. 1. Vegetation structure. Proceedings of the I innean Society of New South Wales 97:211-224. Smith F G 1972 Vegetation Survey of Western Australia, 1:250,000 series. Pemberton and Irwin Inlet. Western Australian Department of Agricul¬ ture, Perth. Smith F G 1973 Vegetation Survey of Western Australia, 1:250,000 series. Busselton and Augusta. Western Australian Department of Agriculture, Perth. Smith R S1990 Mineral nutrient influxes and additions and their effects in a Banksia woodland ecosystem in south western Australia. Unpublished Honours Thesis, Murdoch University, Western Australia. Smith R S 1994 The ecology of two rare Chamelaucium species from south Western Australia. MSc Thesis, Murdoch University, Western Australia. Specht R L1981 The water relations of heathlands: Seasonal waterlogging. In: Ecosystems of the World: Heathlands and related shrublands (ed R L Specht) Elsevier Scientific, Amsterdam, 99-121. Specht R L & Specht A 1989 Spedcs richness of sderophyll (heathy) plant communities in Australia - the influence of overstorey cover. Australian Journal of Botany 37:337-50. Tille P & Lantzke N 1990 Busselton-Margaret River-Augusta land capability study; methodology and results (2 volumes). Department of Agriculture, Western Australia, Technical Report 109. Trudgen M 1991 A flora and vegetation survey of the coast of the city of Mandurah. Department of Planning and Urban Development, Western Australia. Walker J 1983 Description and classification of Australian non-rainforest vegetation. In: Survey Methods for Nature Conservation (eds K Myers C R Margules & I Musto CSIRO) Australia, 18-35. Wardcll-Johnson G, Inions G & Annels A 1989 A floristic classification of the Walpole-Nomalup National Park, Western Australia. Forest Ecology and Management 28:259-279. Webster J R 1962 The composition of wet-heath vegetation in relation to aeration of the ground-water and soil. Journal of Ecology 50:619-637. Wilde S A & Walker IW1982 Collie, Western Australia, 1:250,000 Geological Sheet-Explanatory Notes. Geological Survey of Western Australia, De¬ partment of Mines, Perth. Woolhouse H W1981 Soil aridity, aluminium toxicity and related problems in the nutrient environment of heathlands In: Ecosystems of the World: Heathlands and Related Shrublands Vol 9B (cd R L Specht) Elsevier Scientific, Amsterdam, 215-224. 78 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Appendix 1. Two-way table produced by TWINS?AN classification of 47 southern Swan Coastal Plain wet heathland quadrats. Underlined species names are the characteristic species of the alliances and associations. Values shown within the table are the Braun-Blanquet cover values for the species. Only species occurring in more than two quadrats are included. 27 29 33 34 35 46 47 28 45 39 40 41 Quadrats 8 7 10 12 13 14 11 15 18 17 19 20 16 21 22 1 2 3 4 5 6 9 23 36 37 38 24 30 31 32 25 26 42 43 44 Alliance Association (see Table 1) Pericalymma-Hypocalymma A1 A2 Pericalymma-Regelia B1 B2 B3 Melaleuca-Cassytha Cl C2 Conostylis serrulate Phylidrella pygmaea ‘ParentucelUa viscosa Acacia stenoplera Chamelauciurn rovcei Crfivillea diversifdlk Lqjtocarpus aristatus Loxocarya a((. flexuosa 'Briza maxima Hakea ceratophylla Hakea sulcata Isopogou formosus Rfg^lia ciliata Ltptocarvus canus Leptocarpus elegans Melaleuca uncinata Conostylis aculeate Xanthorrhoea gracilis Alhcasuaritw humilus Calothamnus lateralis Dryandra nivea Creuitlea brachystylis Helichrysum cotula Loxocarya fascicularis Stirlingia latijblia Hibbertia rhadinopoda Mesomelaena telragona Synaphea reticulata Tricorytte eliator Dajtinia preissU Boronia dichotoma Acacia pulchella rilipticum Xanthorrhoea prrissii Adenanthos meisneri Dampiera linearis Acacia flagelliformis Evatidra pauc\flora Acacia semitrullata Adenanthos obovatus Dasypogon bromeliifdlius Euchilopsis linearis H ypocalymma angust{htium Kunzca ericifolia Lysinema cilialum Hibbertia vagiiiata ‘Hypochaeris glabra Lyginia barbate iepidosperma angustatum *Ursinia anthemoides Melaleuca preissiana Drosera macrantha Hakea oaria Sphaerolobium medium Acacia saligna Kunzea recurve T'hysanotus patersonii Lepyrodia glauca Hakea marginata lepyrodia muirii holepsis fiodosus glabella Melaleuca Gtthnia trifida Sporobolus virginicus Melaleuc a hamulosa 1 1 1 1 1 1 1 1 1111 1 1 2 1 1 2 12 112 11 1 1 2 1 1 1 1 2 12 2 4 3 3122323341 1 2 2 2 2 1 2 2 2 2 12 1 11 1 111 1111 1 2 2 1 1 1 2 1 1 1 1 1 1 1 2 2 2 2 2 3 2 2 2 3 1 1 1 2 1 2 1 3 3 1 1 1 1 1112 112 1 1 1 2 1 1 2 1 1 1 1 1 2 2 111 1 1 1 1 1 1 1 11111 12 12 3 1111 4 3 1 1 3 2 1 2 2 1 1 1 111 1 1 1 1112 1 1 11111 1 1 1 1 4 4 3 3 4 1 1 1 1 1 2 1 1 1 1 1 1 11 3 3 12 1 1 1 2 111 2 2 2 2 2 1111 2 2 12 1 2 2 11 2 1 1 2 1 11111 12 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1111 1 1 1 1 1 2 2 2 1 2 1 1 1 1 1 142443444 43 112 2 2 112 2 11 2 1 1 2 4 4 4 4 1 1 1 1 1 1 1 4 4 4 4 3 1 1 1 1 1 1 1 4 1 1 1 1 1 2 1 1 1 1 3 2 1 1 1 1 1 1 1 1111 1 1 1 1 1 1 1 1 1 2 1 2 2 2 3 1 11111 1 1 1 1 1 1 1 3 2 1 1 1 3 1 4 1 1 1 1111 1 1 1 1 1 1 1 4 3 1 1 1 1 1 1 2 2 4 1 1 1 1 2 1 2 3 2 1 1 2 1 2 2 *Exotic speacs 79 ~»i ’ » * V ' .. ' L r. ' , • . . ' ' ■' —• <• ;.;» «iv •' * .^'||f. ■w^«V*-0|ii^'‘A La; >. tv. ' • ’ ■ Vi , .-‘'i ■r. VI •'5 i> % r ♦ .• ’ *-i»^'?. ■- ‘'l- 6*'4>iplii 4? 't *. I ,- •. -^' ' ' '9 -"H "-* - •d- .<;■•• *• -•^, V '.^ ' Ui., B'** > -ik '-■'■-'i^ "(f . i ,< #’,'.,;srL' ' . I-';*-it':'*- *.*5s . >IS vi;^' S : 'i. ■* *■■'. ■ ?-3' '“f'"’ * i:G]l ■ ■ m - •V % 'tv r---!-*;' "*•' i’ - ■ 1 .> *• ,• '.' - • ■•^iT ':• ' •.^%. i,. .'/*■_ /'.’"^'fe'-'' * ‘V:^ I" < aj ■% ’ *:.v* T .A -» ^ -• » tft-'•- '^y^; • -'ri «* =• 'A I, ’ i V. •. ■« . ' %A-f f . • ?- 1»V-^''-* v^# • . " 'f‘m- t ' A i# ■ A t f f . 1 V r‘'-> W-5?%1% ■ ’ s. * 'I. ,.K 4* V*'" .. ^ :..V ■- * -p/% • ' ' ' i- % • . . *- I* •_ M ^ -1^'' r“i»i5^ ^ ' ^- ’<'*•'«■-■ . * "*■ ^ •«*' ’ :.^N ■“ tf ' . *f - . •- ; - . ' *1 • r ,. V '■ . ~- -t. f , ■» »-■- StTui. ’^-•\.. .‘1 .r-i'-s k * • s I «V U Mi- U/-- i “' •. ‘ i'-’ 1*■' ‘'' . *? '• '^ a I *■ \ * - ■rftSii —■ b r i Y .'j . :/ ^ ^1 ■ • rt# k ^ ‘ V ’. yi .. --.iibMi^tifc " ■ •av.' ; - ,> * V- E-" >' .t ■ c. •.,' -J • ■%-'t- ■ a n ^ ' wky.i V M ' ■ *•• I «»* m' • ■: f ^ '-4k ^ * \ ■ .- t(i*isjY<^ * '"*• • •<% ■ bit h 'V i Journal of the Royal Society of Western Australia, 77: 81-85,1994 Seed dispersal of Hibbertia hypericoides (Dilleniaceae) by ants A Schatral^ S G Kailis^ & J E D Fox^ ^School of Environmental Biology, and ^School of Pharmacy, Curtin University of Technology, GPO Box U 1987, Perth 6001, Western Australia. Manuscript received April 1994; accepted September 1994 Abstract The "Yellow Buttercup", Hibbertia hypericoides occurs abundantly in association with several different vegetational types in the south west of Western Australia. In remnant urban bushland at Perth, Western Austral ia, seeds are collected by iv/oMelophorus species, Rhytidoponera violacea (Forel) and Iridomyrmex discors (ForeI).Inbanksialowwood]andatCooljarIoo,aspeciesofthe/r/do»iyrme.rrw/on/^crgroup and Rhytidoponera violacea (Forel) collect seeds of H. hypericoides. One of the Melophorus species discriminates between potentially viable seeds, which are enclosed by a pulpy aril and non-viable seeds which show only a thin, short aril. Large numbers of potentially viable seeds are collected; the aril is consumed and the seeds discarded in the refuse heaps of the colony. The aril is rich in lipids, and the most common fatty acid in both seed and aril is oleic acid. Preliminary laboratory experiments indicate that seed germination is similar in sand collected around ant nests and sand collected elsewhere in the same habitat. INTRODUCTION Many Australian native plants produce seeds with fleshy and highly nutritious appendages, called food bodies or elaiosomes (Semandcr 1906; Berg 1975). In numerous plants these are the arils, structures growing from some part of the ovule, or funicle, after fertilization (Corner 1976; Takhtajan 1991). Ants, attracted to the elaiosome, may carry the whole diaspore (seed plus elaiosome) to their nest and consume the food body but leave the seed intact. Such seed dispersal is called myrmecochory. In contrast, seed harvesting (granivorous) ants destroy the whole diaspore (Berg 1975; Major & Lamont 1985; Andersen 1990; Majer 1990; Hughes & Westoby 1992a). Myrmecochory is particularly common in the dry sclerophyll vegetation of Australia (Berg 1975; Davidson & Morton 1981a,b; Rice & Westoby 1981) and South Africa (Milewski & Bond 1982). Plants are assumed to benefit from ®ced dispersal by ants since seeds may be carried to more advantageous germination sites (Handel 1976; Davidson & Morton 1981a,b; but see Drake 1981; Hughes & Westoby 1992a). Seeds that have been handled by ants may also germinate better (Horvitz 1981). Furthermore, seed disper- ^al by ants may allow colonisation of more distant habitats, ^nd dispersed seedlings may escape competition from par- plants and siblings (Berg 1975). Finally, seeds inside the nest may be protected from fire (Berg 1975) and predators (Heithaus 1981). The ants, in return, benefit from their ^yrmecochorous behaviour since the elaiosome is rich in ^^pidsand vitamins (Bresinsky 1963; Beattie 1985;Oostermeijer 1989; Hughes & Westoby 1992b). Viable seeds of Hibbertia (Dilleniaceae) are partly or com¬ pletely enclosed by a pulpy aril. Ants are important dispersal Agents of seeds of H.vestitata, H.obtusifolia and H.ovata (Drake Royal Society of Western Australia 1994 1981; Berg 1975; Hughes & Westoby 1992b) in Eastern Aus¬ tralia, and it has been shown for Hibbertia ovata that the aril functions as an elaiosome (Hughes & Westoby 1992b). The only observations on seed dispersal for Western Australian Hibbertia species are for H. cuneiformis. Seeds of this species are enclosed by a pulpy, bright orange aril and are collected by birds (N Marchant, quoted in Stebbins & Hoogland 1976). Hibbertia hypericoides, the "Yellow Buttercup" is wide¬ spread in low banksia woodlands from Augusta to Northhampton, Western Australia (Wheeler 1987), and of¬ ten dominates the understorey of a site. Mature seeds are either brown, red or black. Brown and red seeds have a short transparent aril, but black seeds are almost fully enclosed by a white, pulpy aril. Black seeds contain firm, white endosperm, but brown and red seeds are characterized by shrivelled or no endosperm. Hence only black seeds are potentially viable. Initial observations indicate that ants collect the seeds of H. hypericoides soon after they have dropped to the ground. Here we describe seed collecting ants and the amount and type of seeds taken, and examine the following hypoth¬ eses: (1) ants collect potentially viable seeds more often than non-viable seeds, (2) the aril is rich in lipids, (3) ants are attracted to the aril, and (4) germination of seeds as well as survival and growth of seedlings is higher in ant nests than elsewhere in the same habitat. METHODS Study sites and voucher deposition Field study sites were in Kings Park and Wireless Hill Park, both within the Perth metropolitan area. Western Australia (n5‘'50' E, 31‘’56’ S), and Cooljarloo, ca. 200 km north of Perth near Cataby (115°35' E, 32‘’0' S). The natural bushland of Kings Park and Wireless Hill Park is character- 81 Journal of the Royal Society of Western Australia, 77 (3), September 1994 ized by low banksia-sheoak woodland {Banksia attenuata, B. menziesii, Allocasuarimfraseriam, A. huinilis) with jarrah (Ew- califptus marginata) and marri (£. calophylla) interspersed. Hibbertia hypericoides dominates the understorey, with indi¬ vidual plants on average between 3 and 5 m apart. Other common understorey plants are Macrozatnia riedlei, Xanthorrhoea preissii, SHrlingia latifolia, Hypocalymma robustum and Jacksonia sternbergiana. The naturalbushland at Cataby is open low banksia-sheoak woodland interspersed with vari¬ ous Melaleuca species, Conospermum stoechadis and Hypocalymma robustum. Plants with ripe fruits were visited in Kings Park, Wire¬ less Hill Park and at Cooljarloo in December 1990. Seed collecting ants were recorded and sampled for subsequent identification. Voucher specimens are held in a reference collection at the School of Biology, Curtin University or in the Australian National Insect Collection. When species names were unavailable they are either identified with Curtin Uni¬ versity code numbers QUM) or, if voucher specimens are deposited there, with Australian National Insect Collection codes (ANIC). Seed yield During December 1990 a total of 40 nests of the most significant seed collecting ant species were visited in Kings Park and Wireless Hill Park. At Cooljarloo 16 nests were visited. The numbers and types of H. hypericoides seeds in the refuse heaps were counted, then the upper 30 cm of the soil layer of each nest was removed in order to determine the number and type of seeds inside the nest (Shea et al. 1979 found no seeds buried deeper than 30 cm by ants of Melophortis sp. 1 (ANIC) in the southern jarrah forest). The distance between each ant nest and the nearest H. hypericoides plant was recorded. Lipid content of seed and aril Since the nutritional value of seed and aril is mainly due to lipids, a chemical analysis was carried out in order deter¬ mine the lipid content of seed and aril in H. hypericoides. Diaspores (black seeds with pulpy aril) were collected in December 1990 in the Perth metropolitan area, and the aril was removed. Seventy five mg of both arils and seeds were analysed for lipids. Samples were homogenised with 2 ml of isopropanol; 3 ml of hexane was then added to each sample, and the mixtures allowed to stand for 12 h at room temperature. The supernatants were removed and the residues re-extracted using the same procedure. Supernatants were pooled for each sample and evaporated either under vacuum or with nitrogen. The residues, which contained the 'Tipid fraction", were weighed to establish the yield. Lipid fractions (10-50 mg) dissolved in toluene were subjected to a methylation process by adding 4 ml of 1% sulphuric acid in methanol and allowing to stand for 12 h at 50°C. Five ml of 5% aequeous sodium chloride solution was added to each sample, followed by 5 ml of hexane. The mixtures were thoroughfully shaken and allowed to sepa¬ rate. The supernatant from each mixture was removed and the residue re-extracted with an additional 5 ml of hexane. Supernatants for each sample were pooled and then passed through a column of anhydrous sodium sulphate. These methylated lipid fractions in hexane were used for gas- chromatograph / mass-spectrograph analyses. The methylated lipid fractions, which contained methyl¬ ated fatty acids (FAMES) were analysed using an Hewlett Packard GC-MS (Model 5971). Chromatography was per¬ formed with an Econocap** Carbowax 20M column (ID 0.32 mm and film thickness 0.25 p) using helium as the carrier gas. The initial column temperature was 50''C for 5 min, followed by an increasing gradient of 5"C min * to a temperature of 240‘’C, with the latter temperature being held for at least another 20 min. Run times were approximately one hour. FAMES were identified by using known retention times of standards, and mass spectra. Attractiveness of the aril A choice experiment was conducted to determine if the aril is the main attractant to ants of Melophorus sp.l. Fifteen of their nests were randomly selected in Kings Park in December 1990. Three seeds, one black and shiny seed with a pulpy aril, one brown seed with a short, brittle aril, and one black seed without an aril (the aril had been removed with forceps), were placed 50 cm away from each nest. Ant behaviour towards these seeds was observed in terms of the time ants spent attending the seeds, whether the aril was consumed and whether the seeds were carried into the nest. All observations were made from 10 am to 1 pm, when temperatures were above 25'’C. Effect of soil on germination The following experiment was performed in order to test whether the soil inside ant nests is beneficial to the germina¬ tion of H. hypericoides seeds. Freshly collected (in 1990) black seeds were air dried for 3 weeks at room temperature and then planted in sand trays. Sand was collected from the natural habitat in Kings Park. Sand was taken randomly (from any open area which was not close to an ant nest), from underneath H. hyf^ericoides bushes, and from ant nests of Melophorus sp. JDM 358. Sterilized coarse river sand was used as a fourth treatment. Between 3 and 6 sand trays were prepared with each of the four soil types, twenty five seeds placed in each. All trays were prepared at the end of Decem¬ ber 1990 and left unwalered in the glass house of the School of Environmental Biology until the 20*'" of March 1991, when watering commenced. Surface soil temperatures from De¬ cember to February ranged from 30 to 45®C during the day. Tlie number of germinating seeds was counted every second day until the 30**' of September 1991, when watering ceased. Watering of the trays re-commenced in March 1992, and the number of germinating seeds was counted regularly until the beginning of September 1992. Survival of seedlings in the field In Kings Park and Wireless Hill Park, plants and ant nests were scrutinised for seedlings in December 1991. Seedlings were tagged with flagging tape, re-visited in June 1992, and the number of surviving seedlings was recorded. RESULTS Ants as seed collectors Ant species observed collecting seeds of H. hypericoides in the study sites belong to the genera Melophorus (Family Formicinae), Rhytidoponera (Fam. Ponerinae) and Iridomyrmex 82 Journal of the Royal Society of Western Australia, 77 (3), September 1994 (Fam. Dolichoderinae). Melophoriis sp JDM 358 was the main collector of H. hypericoides seed in I^gs Park and Wireless Hill Park in 1990 (Table 1). Other ants taking seeds of H. hypericoides are Melophorus sp. 1 (ANIC), Iridomyrmex discors (Forel) and Rhytidoponera violacea (Forel). At Cooljarloo, Rhytidoponera violacea (Forel) and a species oithe Iridomymiex rufoniger group removed seeds of H. hypericoides (Table 1). Attractiveness of the aril The aril of potentially viable seeds of H. hypericoides functions as an elaiosome. Melophorus sp. JDM 358 is at¬ tracted to the aril of potentially viable seeds. Ants either take black seeds which are enclosed by an aril to the nest imme¬ diately or consume the aril first and then take the seeds back to the nest. Seeds without arils attached are less frequently Table 1 Frequency with which different ant species collected seed of H. hypericoides as a percentage of the total number of observa¬ tions (N) in three study sites in December 1990. Dashes: species was not observed in study site. Site N Melophorus sp. JDM 358 Melophorus sp. 1 (ANIC) R. violacea I. discors 7. rufoniger group Kings Park 63 70 20 5 5 - Wireless Hill 32 100 - - - - Cooljarloo 16 - - 70 - 30 Nests of Melophorus sp. JDM 358 are abundant in Kings Park and Wireless Hill Park and are on average 1.24 m (n = 24) and 1.5 m (n = 36) respectively from the nearest Hibberfia plant. At Cooljarloo, nests of Rhytidoponera violacea are on average 1.2 m from the nearest H. hypericoides plant (n = 16). In the Perth metropolitan area, Melophorus sp.l (ANIC) and Melophorus sp. JDM 358 start building (or re-open) nests from the previous season in September. They collect large num¬ bers of Hibbertia petals which are carried inside the nest and subsequently discarded outside the nest on refuse heaps or middens, hours or days later. Ants begin to collect seeds at the end of October and continue until December when the last seeds mature. Seed yield Seed numbers were determined in nests of Melophorus sp.l (ANIC) and Melophorus sp. JDM 358 at Kings Park, in nests of an unidentified species of the Iridomyrmex rufiger group and in nests of Rhytidoponera violacea (Forel) at Cooljarloo, as well as in nests of Melophorus sp. JDM 358 at Wireless Hill Park. Ants mainly collected potentially viable seeds. Large numbers of black, potentially viable seeds lay discarded in colony middens at all study sites. The numbers of black seeds found around and inside ant nests ranged from 4 (Kings Park) to 49 (Cooljarloo) in 1990. (Kings Park: number of nests = 26, average number of seeds (mean ± SE) n = 4.3 ± 1.2; Wireless Hill Park: 14 nests, n = 9.9 ± 6.9; Cooljarloo: 16 nests, n = 48.6 ± 8.4) Of all seeds found, none had the aril still attached. In all three study sites, between 2 and 5% of the seeds were brown. Although plants produce many more brown seed s than black seeds (see Schatral & Fox in press), brown seeds were collected only infrequently. Lipid content of seed and aril The most abundant fatty acids, in botli seed and aril are oleic acid, linoleic add, palmitic add and stearic add (Table 2). Qualitative differences in the lipid content between seed and aril were not found. One saturated acid, lauric acid (C12:0), the unsaturated fatty adds C16:l(n-7), C16:l (palmitic add) and C18:l (oleic add) are relatively more abundant in the food body. C18:2 (linoleic acid) is more common in the Seed than in the aril. Table 2 The abundance of fatty acids in seed and aril of H. hypericoides (black seeds only). Abundance is expressed as total % and as a ratio to oleic add. Fatty acid Name SEED ARIL % Ratio % Ratio 12:0 Lauric low* - 0.4 0.8 14.0 Myristic 1.3 3.5 0.7 1.4 16:0 Palmitic 19.8 53.5 26.8 45.9 16:1 2.1 5.8 2 4.3 16:1 (n-7) Palmitic 0.3 0.9 4.6 9.8 17.0 low* - 0.31 0.7 18:0 Stearic 8.4 22.7 6.2 13.3 18:1 (n-9) Oleic 37 100 46.7 100 18:1 0.6 1.7 0 0 18:2 (n-6) Linoleic 28.4 77 11.0 23.6 18:3 (n-3) Linolenic 0.6 1.5 0.3 0.7 20:0 Arachioic 0.5 1.5 0.5 1.2 ’ The amount was too small to be measured accurately. collected than black seeds with arils. Thus, 80 % of the black seeds with arils were taken back inside nests but only 53.3 and 46.7 % of the brown and black seeds without arils, respectively, were taken. Brown seeds as well as black seeds without arils are attended for significantly shorter time periods (analysis of variance, F^^ = 5.53, p<0.05) than seeds with arils (time attended: n = 15; black seeds with aril, 40.4 ± 5.8 min; black seeds without aril, 13.9 ± 2.6; brown seeds, 15 ± 2.5 min). After the seeds had been introduced, ants took only a few minutes to detect them. Seeds were usually attended by 2-12 minor morphs simultaneously (ants of the genus Melophorus show continuous polymorphism, J Majer, pers. comm.). However, in most cases a major morph helped to carry the seeds inside the nest. The time that elapsed until the ants carried the seeds back to the nest was not signifi¬ cantly different (analysis of variance, F^. = 0.014, NS) for black seeds enclosed by an aril and brown and black seeds without arils respectively (time taken to carry seed to nest: black seeds with aril, n = 11, 34.1 ± 6.3 min; black seeds 83 Journal of the Royal Society of Western Australia, 77 (3), September 1994 without aril, n = 7, 34.7 ± 6.5 min; brown seeds, n = 6,35.7 ± 5.7 min). Ant dispersal and its effect on seed germination Seed germination is similar in sand collected around ant nests and sand collected elsewhere in the same habitat. During the first season, seeds germinated significantly slower (analysis of variance, = 4.28, p<0.05) in sterile river sand than in any of the three natural soil treatments (Table 3). However, the number of germinating seeds was not differ¬ ent between soil treatments after half of the experimental time period had elapsed and at the end of the experiment (analysis of variance on arcsine transformed data, after 50 % of time; F, = 1.1, NS; at the end of the experiment; = 0.11, NS). During the second season, seeds started to germinate significantly later in river sand than in any of the natural soil treatments (Fj „ = 14,4, p<0.05). After half of the experimen¬ tal time had elapsed, none of the seeds had germinated in river sand but a small percentage of the seeds in the natural soil treatments had done so (Table 3). The number of germi¬ nating seeds was not significantly different between the natural soil treatments (F^ g =17.11, p<0.05). At the end of the experiment, the final germination percentage was similar for river sand and the three natural soil treatments (F3 g = 0.009, Germination of seeds of H. hypericoides (mean ± SE, %) from 4 soils over two seasons; First day of germination (mean ± SE), germination (mean ± SE, %) after half of the experimental time period had elapsed and at the end of the experiment, n = number of replicates (= sand trays, 25 seeds per tray). During the second season, fewer replicates were available since no germination was recorded for several trays. First Season First day of Germination % 1991 Germination After half of At the end of the experiment experiment Soil Type n random 5 59 ± 6.2 10.0 ± 4.0 11.2 ±4.3 bushes 3 76 ± 8.5 5.3 ±2.7 9.3 ± 5.8 ant nests 5 54 ± 9.3 6.0 ± 2.0 6.4 ± 4.9 river sand 6 108 ±15.0 4.3 ± 4.3 9.3 ± 4.9 Second Season First day of Germination % 1992 Germination After half of At the end of the experiment experiment Soil Type n random 4 00 +1 4.5 ±1.5 5.5 ±1.5 bushes 3 65 ± 2.9 4.5 ± 0.9 5.0 ± 0.6 ant nests 4 74 ± 4.1 3.0 ± 0.6 4.5 ± 0.9 river sand 4 123 ±10.6 0 6.0 ± 2.3 Survival of seedlings in the field Seedlings were found close to, or on, seven of 40 (17.5 %) ant nests at Wireless Hill Park and Kings Park in December 1991. The number of seedlings ranged from 2 to 15 (mean ± SE, 5.4 ± 1.67). Seedlings were also observed underneath parent plants, but the average number of germinating seeds was lower; of 60 plants examined, seedlings were detected underneath 19, with the average number of seedlings of 1.3 ± 0.13. By June 1992, none of the seedlings had survived, regardless of its germination site. DISCUSSION Mclophoriis sp.l (ANIC) is one of the most significant seed takers in the northern jarrah forest (Major 1982). These ants collect elaiosomes although they sometimes consume entire seeds, depending on seed size (Majer 1982). However, the species is omnivorous, with arthropod fragments and seeds found scattered around the nests. Other Melophorus species are considered granivores (Davidson & Morton 1981a). De¬ tails of the ecology of Melophorus sp. JDM 358, the most significant collector of H. hypericoides seeds in Kings Park, are unknown. At Cooljarloo, Rhytidopouera violacea and Iridojnynuex rufotiigcr sp. collect seeds of H. hypericoides. In contrast, species of the genus Iridomyrmex have never been recorded eating seeds. Seeds of H. hypericoides drop passively to the ground as soon as they are ripe (dispersal mechanism is of the Viola odorata type; Semander 1906; Berg 1975; Drake 1981). Once dropped from the plant onto the ground, the seeds are removed rapidly. Similarly, ants in the dry sclerophyll forest on North Stradbroke Island in Queensland remove virtually all elaiosome bearing seeds within two days of seed fall (Drake 1981). The high removal rate of H. hyjjericoides seeds suggests that only few seeds remain underneath the parent plant, and therefore most seedlings would escape competi¬ tion with the parent plant and possible predators under¬ neath the plant. However, the accumulation of seeds around ant nests may cause strong competition with siblings (see below). It is not known whether seeds of H. hypericoides, once collected by elaiosomc-consuming ants, are subsequently removed by seed-feeding animals (see Hughes & Westoby 1992b for a discussion of this problem). Since Melophorus sp. JDM 358 removes black seeds of H. hypericoides still enclosed by an elaiosome in preference to brown seeds and black seeds without the elaiosome, they must be able to discriminate between these seed types. Ants appear to choose between different seeds on the basis of size and/or the chemical compositon of the elaiosome. It has been found that the presence of elaiosomes increases the removal rates by Rliytidoponera metallica (Hughes & Westoby 1992b) and other ants (Majer 1982; Oostermeijer 1989; Drake 1981; Hughes & Westoby 1992b). Moreover, ants can distin¬ guish fertile Eucalyptus regtians seeds from chaff, and a reducing, sugary substance is regarded as the attractant (Ashton 1979). The chemical analysis demonstrates that both the aril and seed of H. hypericoides are rich in the same fatty acids. The elaiosomes of many plants arc rich in lipids, but the compo¬ sition of the lipids varies between species. Oleic acid is the major fatty acid present in the elaiosomes of Hepatica americam, Viola odorata and four other ant-dispersed North American herbaceous plants (Marshall et al. 1979, Skidmore & Heithaus 1988; Kusmenoglu et al. 1989). Oleic acid is also abundant in both elaiosome and seed of H. hypericoides. Oleic acid and especially 1,2 diolein, a diglyceride which is de¬ rived from oleic acid, are the main attractants for seed 84 Journal of the Royal Society of Western Australia, 77 (3), September 1994 collecting ants (Marshall et al. 1979; Skidmore & Heithaus 1988; Brew et aL 1989; Kustnenoglu et al. 1989; Hughes & Westoby 1992b). In contrast, Brcsinsky (1963) suggested that the diglyceride ricinolic acid attracts ants to the elaiosomes of Viola odorata. Further studies are necessary to determine which chemical substance in the elaiosome of H. hypericoides induces seed carrying behaviour in ants. Germination of H. hypericoides seeds in sand collected from ant nests is not enlianced compared with germination attained from sand collected underneath bushes, randomly collected sand and river sand. However, the conditions inside an ant nest will be more complex than the experimen¬ tal conditions. Moisture, temperature, aeration and the pres¬ ence of microorganisms may affect the germination of seeds in nature. Since seeds germinated more slowly in sterile river sand than in the natural soil treatments, river sand may lack nutrients and/or certain micro-organisms that promote the successful germination of seeds in natural soil. The death rate of seedlings observed in the field may be increased as a result of competition on the ant mounds (Beattie & Lyons 1975; Shea1979), although this hypoth¬ esis does not explain the equally high death rate for seedlings that grow underneath parent plants. Competition between siblings would be mainly for nutrients and water. The nutri¬ ent content of the soil has not been analysed during the present study on H. hypericoides. Majer (1982) has found no differences in total nitrogen or available phosphorus con¬ tents between soils from Melophorus sp. 1 (ANIC) nests and soils from control areas 1 m away from each nest. In contrast, nutrients are concentrated locally on the mounds of some other ants (Davidson & Morton 1981b). Drake (1981) is scepticalofthehypothesisthatantnestsarebeneficial germi¬ nation sites. It is not only that seed dumps may cause high competition between seedlings, but the nests of some myrmecochorous ant species do not appear to offer suitable germination conditions. Nests of R. mctalUca, for instance, are sometimes located in tree trunks and germination is highly unlikely because of the restricted depth of soil avail¬ able. Whether the removal of the aril benefits seed germination is unknown for H. hypericoides but it promotes germination in H. cuneifonnis (unpublished data). Simi larly, Horvitz (1981) found that the germination of the tropical perennial herb Calathea microcephala was enhanced by the removal of the elaiosome. AcknowledgtnctUs: This study was supported by the Mineral and Energy Research Institute of Western Australia (MERIWA). We are very grateful to B Heterick, T Postle, J Majer and S Shattuck for identification of the ants. J Majer also kindly read an earlier draft of this manuscript and we would like to thank him for his critical comments. References Andersen A N 1990 Seed harvesting ant pests in Australia. In: Applied Myrmecology (eds R K Van der Meer, K Jaffe & A Cedeno) Westview Press, Boulder, 34-39. Ashton D 1979 Seed harvesting by ants in forests of Eucalyptus regnans F. Muell. in central Victoria. Australian Journal of Ecology 4:265-2^. Beattie A J 1985 llie Evolutionary Ecology of Ant-Plant Mutualisms. Cam¬ bridge University Press, Cambridge. Beattie A] & Lyons N1975 Seed dispersal in Viola (Violaceae): adaptations and strategies. American Journal of Botany 62:714-722. Berg R Y1975 Myrmecochorous plants in Australia and their dispersal by ants. Australian Journal of Botany 23:475-508. Bresinsky A 1963 Bau, Enlwicklungsgeschichte und Inhaltsstoffe der Elaiosomen. Studien zur myrmekochoren Verbreitung von Samen und Fruchten. Bibliotheca Botanica 126:1-54. Brew C R, O'Dowd D J & Rae I D 1989. Seed dispersal by ants; behaviour- releasing compounds in elaiosomes. Oecologia 80:490-497. Corner E J H 1976 Tlie seeds of Dicotyledons. (2 volumes). Cambridge University Press, Cambridge. Davidson D W & Morton S R 1981a Myrmecochory in some plants (Family Chenopodiaceae) of the Australian arid zone. Oecologia 50:357-366. Davidson DW & Morion SR 1981bCompetition for dispersal in ant-dispersed plants. Sdence213:1259-1261. Drake W E1981 Ant-seed interactions in dry sclerophyll forest on Stradbroke Island, Queensland. Australian Journal of Botany 29:293-309. Handel S N1976 Dispersal ecology of Carex peduuculata (Cyperaceae), a new northern American myrmecochore. American Journal of Botany 63:1071- 1079. Heithaus E R 1981 Seed predation by rodents on three ant-dispersed plants. Ecology 62:136-145. Horvitz C C 1981 Analysis of how ant behaviours affect germination in a tropical myrmecochore Calathea microcephala (P. & E.) Koernicke (Marantaceae): microsite selection and aril removal by neotropical ants Odontomachus, Pachycondylaand Solenopsis (Formicidae). Oecologia 51:47- 52. Hughes L & Westoby M 1992a Fate of seeds adapted for dispersal by ants in Australian sclerophyll vegetation. Ecology 73:1285-1299. Hughes L & Westoby M 1992b. Effect of diaspore characteristics on removal of seeds adapted for dispersal by ants. Ecology 73:1300-1312. Kusmenoglu S, Rockwood L & Gretz M 1989 Fatty acids and diacylglycerols from elaiosomes of some ant dispersed seeds. Phytochemistry 28:2601- 2602. Majer J D 1982. Ant-plant interactions in the Darling Botanical District of Western Australia. In: Ant-plant interactions in Australia (ed R C Buckley) Dr W Junk Publishers, The Hague, 45-61. Majer J D 1990. The role of ants in Australian land reclamation seeding operations. In: Applied Myrmecology (eds R K Van der Meer, K Jaffe & A Cedeno). Westview Press, Boulder, 544-554. Majer ] D & Lamont B B 1985 Removal of seeds of Grevillca pteridofolia (Proteaccae) by ants. Australian Journal of Botany 33:611-618. Marshall D L, Beattie A J & Bollenbacher W E1979 Evidence for diglycerides as attractants in an ant-seed interaction. Journal of Chemical Ecology 5:335-343. Milewski A V & Bond W J 1982. Convergence of myrmecochory in mediterranean Australia and South Africa. In: Ant-plant interactions in Australia (ed R C Buckley) Dr W Junk publishers. The Hague, Nether¬ lands, 89-98. Oostermejer J G B1989 Myrmecochory in Polygala imlgaris L., Luzula campestris and Viola curtisii Forster in a Dutch dune area. Oecologia 78:302-311. Rice B L & Westoby M 1981 Myrmecochory in sclerophyll vegetation of the West Head, New South Wales, Australian Journal of Ecology 6:291-298. Schatral A & Fox J E D in press. Viability of seeds in the genus Hibbertia. Seed Science and Technology. Sernander R 1906 Entwurf einer Monographic der europaischen Mynnekochoren. K Svens.Vetenskasakad. 41:1-40. Shea S R, McCormick ] & Portlock C C1979. The effect of fires on regeneration of legume species in the southern Jarrah (Eucalyptus marginata) forest in Western Australia. Australian Journal of Ecology 4:195-205. Skidmore A B & Heithaus E R 1988 Lipid cues for seed-carrying by ants in Hepatica americaiia. Journal of Chemical Ecology 14:2185-2196. Stebbins G L & Hoogland R D 1976 Species diversity, ecology and evolution in a primitive angiosporm genus: Hibbertia (Dilleniaceac). Plant Systemat- ics and Evolution 125:139-154. Takhtajan A1991 EvoIutionaryTrendsinFlowering Plants. ColumbiaUniver- sity Press, New York. Wheeler J R 1987 Family 38: Dilleniaceae. In: Flora of the Perth Region. Vol. 1 (eds N G Marchant, J R Wheeler, B L Rye, E Bennett, N S Lander, & T D Macfarlanc). Western Australian Herbarium, Department of Agriculture, 119-133. 85 -y.} * ,■ • i v>« ■ *’ -^ ■ ' ^;-?,? .',n.^ ^l.f .!|t ' > *i iV ^ f «: -i JPi *..■«■ ' /■r.*?3''V' ’ J J* '• ’ -- ^••41* ‘ * •►'yfc - ' '-■* A^9 ‘r,- I V. • ’• O'. 1^1^ r, ^il.* 4v‘ ^ » • '^ ' V • ’'V . ja*} * - :.- vy'#'INWJ6^’ '*■ ; ' *\ ’ ^ ?K**^ ■ * __ t •€.% «• »' ‘i*' -■« « c 1- -'■■' n ' . J I « ,-4 ■*. .. - # JVvy ■-hn .»i - Bfc';t*s>'U I..*-#*- . : ■• 1 . v;=;-> ■ «*■. '••■ .‘-i Sit . f . * : "i *♦-'6^- r.- ♦ • • ^ • n .■ *, ^.•' • • >v - «• r 1 ■'-•i^'. ■^*' ‘ • .'Jl ^ - ' N* '»• 1-" ■ .-^ •«%. V ‘^^•, V ' r<’ , • .* <4^ t * I 'ft “#: :^. > , ' ' g% %• V. V ’' r « s. ^w* ' '’akV.'- •j 'll s . • ■*.»'^ ' • t** Journal of the Royal Society of Western Australia, 77: 87-95,1994 Holdings in the Library of The Royal Society of Western Australia [Books, current Journals and Maps] BOOKS Abramovich, I I & V V Gruza 1972 Fatsialno-formatsionnyi analiz magmaticheskikh kompleksov. Leningrad: Nedra. 240 p: maps. Akademiia NaukSSSR 1968GeoIogicheskii putevoditel po kanalu im. Moskvy i Volgo-Baltiiskomu vodnomu puti. Leningrad: Nauka. 211 p: ill. Alekseev, A K 1963 Paleogenovaja fauna molluskov sevemogo Priaralja. Erevan: Izdatelstvo Akademii Nauk Armjanskoi 5SR. 176 p: plates. Alexander, RMcN 1968 Animal mechanics. Seattle: University of Washington Press, xi, 346 p. Allan Hancock Foundation for Scientific Researcli 1955 Essays in the natural sciences in honor of Captain Allan Hancock. Los Angeles; University of Southern California, xii, 345 p; ill, maps. Allied Geographical Section. Southwest Pacific Area 1944 An annotated bibliography of the Southwest Pacific and adjacent areas. 3 v. Almukhanbetov D 1972 Elektromagnitnoye zondirovaniye v Kazakhstane. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 211 p: ill. AtzJ W1971 Dean bibliographyoffishcs 1968. New York; American Museum of Natural History. 512 p. Australia Resources Information and Development Branch. 1962 Index to Australian resources maps of 1940-59. Canberra: Dept of National Devel¬ opment. 241 p: 4 folded maps. Australia Dept Home Affairs and Environment. 1982 National conservation strategy for Australia: Living resource conservation for sustainable devel¬ opment. Proceedings of the National Seminar, Canberra, 30 November • 3 December 1981. Canberra: AGPS. 1 v. Australia Dept Home Affairs and Environment. 1982 National conservation strategy for Australia; Living resource conservation for sustainable devel¬ opment. Conference draft, February 1983. Canberra; AGPS. 1 v: maps. Australia Dept Home Affairs and Environment. 1982 National conservation strategy for Australia: Living resource conservation forsustainable devel¬ opment. Towards a national conservation strategy: A discussion paper. Canberra; AGPS. 71 p. Australia Dept of Home Affairs and Environment. 1983 National conserva¬ tion strategy for Australia; Living resource conservation for sustainable development. Proposed by a conference held in Canberra in June 1983. Canberra: AGPS. 14 p. Australia Senate Standing Committee on Science, Technology and the Envi¬ ronment. Parliament. 1983 Preservation of the Abbott's Booby on Christ¬ mas Island. Canberra; AGPS. Australian Academy of Science 1968 National parks and reserves in Australia. Canberra: Australian Academy of Science. Hi, 45 p. Australian Environment Council; Australia. 1982 Australian achievements in environmental protection and nature conservation. 1972-1982. Canberra: Australian Environment Council and Council of Nature Conservation Ministers, in, 34 p: ill. Australian Nature Conservation Agency; Australian National Parks and Wildlife Service 1993 Christmas Island National Park : plan of manage¬ ment. Canberra: Australian Nature Conservation Agency, vii, 71 p: maps. Babbington C1874 Manual of British botany, containing the flowering plants and ferns, arranged according to the natural orders. 7*’ od. London: van Voorst. Ixiii, 473 p. Bailey F M1890 A synopsis of the Queensland flora. 3^** supplement containing both the Phaenogamous and Cryptogamous plants. Brisbane: Govt Printer. 135 p: ill. Bailey F M 1893 A companion for the Queensland student of plant life. Brisbane: Department of Agriculture. 108 p. Bailey F M1897-1913 Contributions to the flora of Queensland and New Guinea, 1897-1913. Brisbane: Queensland Dept of Agriculture. 66 extracts from the Queensland Agricultural Journal, bound in 1 volume. Bailey F M 1899-1902 The Queensland flora. Brisbane; Govt Printer. 6 parts and index. Bailey L H1901 The principles of fruit-growing. 4'^ ed. New York: Macmillan. xvii, 516 p: ill. Baker R T 1913 Cabinet timbers of Australia. Sydney: Govt Printer. 186 p: chiefly col ill. Baker R T & H G Smith 1910 A research on the pines of Australia. Sydney: Govt Printer, xiv, 458 p: ill, maps. © Royal Society of Western Australia 1994 Barnes L C 1992 Opal: South Australia's gem stone. Rev. ed. Parkside, SA: Dept of Mines and Energy, Geological Survey. 176 p: ill. Bastian L V 1982 Minerals and their relationships in the Leeuwin Block, Leeuwin-Naturaliste National Park. Perth: Govt Chemical Laboratories. 24 p: ill. Bekmukhametov A E 1970 Formirovanie skarnovo-rudnykh zon magnetilovykh meslorozhdeniiyuzhnogoTurgaya. Alma Ata: Akademiya Nauk Kazakhskoi SSR. 206 p; ill. Bentham G & Baron F von Mueller 1863-1878 Flora Australiensis. London: Lovell. 7 v. Biological Diversity Advisory Committee (Australia) 1992[?] A national strat¬ egy for the conservation of Au.stralia's biological diversity: draft for public comment. Canberra; Dept of the Arts, Sport, the Environment and Terri¬ tories. 38 p. Black J G 1885 Chemistry for the gold fields: including lectures on the non- mctallic elements, metallurgy, and the testing and assaying of metals, metallic ores, and other minerals, by the test-tube, the blow-pipe, and the crudble. Dunedin: Horsburgh. x, 569 p. Blackall W E & B J Grieve 1981 How to know Western Australian wildflowers: A key to the flora of temperate regions of Western Australia. Part III. Nedlands: University of Western Australia Press. Ixxviii, 460-595 p: 12 p plates: ill. Bliss D & D M Skinner 1963 Tissue respiration in invertebrates. New York: American Museum of Natural History, x, 139 p: tables. Bok I 1 1965 Agronomicheskie rudy (osnovy ikh geologii i poiskovo- otsenochnyc priznaki). 2nd ed. Alma Ata: Akademiya Nauk KazakhskoiSSR, Kazakhskii Politekhnicheskii Institul. 308 p: ill. Bok I I 1965 Geologiya i resursy agrophimicheskogo syrya Kasakhstana. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 215 p. Booker F W 1961 Studies in Permian sedimentation in the Sydney Basin. Sydney: NSW Mines Dept. 53 p, 5 p plates. British Museum (Natural History) 1910 Guido to the Crustacea, Arachnida, Onychophora and Myriopoda exhibited in the Department of Zoology, British Museum (Natural History). London: Trustees of the British Mu¬ seum. 133 p: ill. Brown G M & J A Crutchfield 1982 Economics of ocean resources, a research agenda: Proceedings of a national workshop. Sponsored by Office of Ocean Resources Coordination and Assessment, National Oceanic and Atmospheric Administration, Orcas Island, Washington, September 13- 16,1981. Seattle; University of Washington Press, xii, 242 p. Brown R 1866-1867 The miscellaneous botanical works of Robert Brown. London: Tlic Ray Society. 3 v. {inii, 612; viii, 786 p.). Brown R1874 A manual of botany, anatomical and physiological. Edinburgh: William Blackwood, xviii, 614 p. Bublichenko N L1976 LowerCarboniferous brachiopods of altai ore deposits. Alma-Ata; Nauka. 211 p; ill. Buchanan j 1880 Manual of the indigenous grasses of New Zealand. New Zealand Colonial Museum and Geological Survey Department. Welling¬ ton; NZ Govt Printer, xiv, 156 p, 58 p plates: ill. Burn R S1901 Outlines of modem farming. Vol. 4. The dairy, pigs, poultry. 8* ed. London; Crosby Lockwood. 211,16,48 p. Campbell N & R MSSmelliel983 The Royal Society of Edinburgh (1783-1983): T^e first two hundred years. Edinburgh: Royal Society of Edinburgh, xvi, 186 p: ill, ports (some col) Campbell W D 1899 Aboriginal carvings of Port Jackson and Broken Bay measured and described by W. D. Campbell. Sydney: NSW Govt Printer. V, 73 p, 29 p plates: map. Carter G S1967Structure and habit in vertebrate evolution. Seattle; University of Washington Press, xhf, 520 p: ill. Chadwick A C & S L Sutton 1984 Tropica! rain-forest: The Leeds symposium. Leeds: Leeds Philosopliical and Literary Society, xvi, 335 p: ill. ChampeJ L (assisted by W RWedelcffl/.) 1949 Proceedings of the Fifth Plains Conference for archaeology. Nebraska: University of Nebraska, x, 136 p, 12 p plates: ill. Chang Wen-Tang, Chen Pei-chi & Shen Yan-Bin 1976 Fossil Conchostraca of China. Peking; Science Press. 325 p: ill. Chao N L, W Kirby-Smith 1985 Proceedings of the International Symposium on Utilization of Coastal Ecosystems: Planning, pollution and productiv¬ ity, 21-27 Nov 19 Rio Grande. Rio Grande, Brazil: Editora da Furg. ill, maps. Cihar J 1987 Obojzivelnici a Plazi: Katalog k cxpozici zoologickeho oddeleni Narodniho muzea v Praze. Prague: Narodni Muzeum. 132 p: ill. 87 of the Royal Society of Western Australia, 77 (3), September 1994 Journal 1080 Evolutionary biology of the New World o =,lore IndiaJanuary 1 .^^,oman from marine invertebrates in fcFE 1910 Minnesota m - ^'‘^Tota Presa. 169 p: iH 1912 Minnesota trees and shrubs. Cobbold T S 1879 j^i^-cctozoa. London; Churchill, .ri, 508 p: ill. ticluding some account f h oiltivation and plantation Coghlan H L & ] ^ ^^^b^Lockwood. z, 128p. 10 p plates: ill. ^machinery. . ^Jfii,,xondon: Routlc-dgo. 179p; col platc-s. ColemanWS1860Britishbuttern Collander R & V IWessa ° ;J^,„,,.Helsinld:Sod withanappendixon , ^ ^ and Geological Survey Dept 1873 Catalogue of the land ^"‘Xllu'^ca of New Zealand xwi Enterprise. Colonial Sugar xuiV, 500 p; ill. Sydney: Angus and (Canberra) 1955 Bibliography of the flora Commonwealth Nation^.^^^^ Commonwealth National Library. and fauna of New o 26 p. Akademie der Wissenschaften der DDR 1973 Copernicus-Komitee an 500.Berlin: Nicolaus Copemcu . 103 p: ill- Akademie- Verlag fo Southern Great Barrier Reef. Part 1. Cribb A B 1983 Marmc ^^^^^banCora^ Rhodophyta. Brisb. • ^ North America; A bibliography. Crispens C G1960 P- Seattle: Uraversity . forest mensuration as a basis for the Cromer DAN 1961 ^PP^, ^Canberra: Forestry and Timber Bureau, vi. management or i uiu^ off New South Wales. i an i. oic ing C°^215 P: 'IP resultsof the fishingexperimenLscarried DannevigHC1911-1933.^ g 1909-1914. Sydney: Dept Trade and out by the h- 1. =»• Dar^" 0868 urvarij^ion of animals and plants under domestication. 1. ‘^“"fl" 89 oSestLtureanddistributionofcoralreefs:Alsogeolos^^^^^ Darwin C1890On the ^ j ts of South America visited Dar^ttlS scientific papers, volume 2.Tida^ Cambridge: Cambridge University Pre^s. m 516 p^ De Laeter J R 1993 A question of time. Bentley, WA: Curtin University of Technology. H sound cassette (ca. 60 mm.) + programme) DeGraff R M 1980 Forest habitat for birds of the Northeast. Wa,shmgton: U S Government Printing Office. Ill, 598 p: ill, maps. Devillers P 1989 Atlas des oiseaux nicheurs de Belgique. Bruxelles: Inshtut Royal des Sciences Naturellesde Belgique. 394 p: maps. Deyl M 1946 Study of the genus Sesleria. Prague: Ceskoslovenska botanicka spolecnost. 256 p: ill, maps. . c u n Diels L 1922 Beitrage zur Kenntnis der Vegetation und Flora der Seychellen. Jenna: Gustav Fischer. 60 p: ill-, map. Diels L & E Pritzel 1905 Fragmenta phytographiae Australiae occidentalis: Beitrage zur kenntnis der pflanzen Westaustraliens ihrer verbre.tung und ihrer lebens-verhaltnisse. Leipzig: Wilhelm Engelmann. p56^62: ill. Dixey F A practical handbook of water supply. London; Murby. xxviii, 571 p. ill, maps. Dodd A P1940 The biological campaign against prickly-pear. Brisbane: Govt Printer, ii, 177 p: ill, map. Domin K 1913 Beitrage zur flora und pflanzengeographie Australiens. 1. Lieferung. Stuttgart: E Schweizerbart. 238 p, 8 p plates: ill. Drew F 1875 The Jummoo and Kashmir territories: A geographical account. London: Stanford, xv, 568 p: ill, col maps, ports. Dumont d'Urville JSC 1834 - 1835 Voyage pittoresque autour du monde. Paris: L Tenre. 2 v. (576; 584): ill, maps, ports. Dumont d'Urville J & J A Boisduval 1832-1835 Voyage de decouverles de I'Astrolabe execute par ordre du Roi pendant les annees 1826-1827-1828- 1929 sous le commandement de M. J. Dumont d'Urville. Faune entomologique de I'Ocean Pacifique avec I'illustration des insectes nouveaux recueillis pendant le voyage. Paris: Tastu. iv, 716 p. Dumont d 'Urville J, J R C Quoy & j Gaimard 1830-1834 Voyage de decouvertes de I'Astrolabe execute par ordre du Roi, pendant les annees 1826-1827- 1828-1829 sous le commandement de M. j. Dumont d'Urville. Zoologie. Paris: Tastu. 6 v. Dzhurkashev T N1972 Antropogenovaia istoria Balkhash-Alakolskoi vpadiny. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 126 p: maps. Elazari-Volcani 1 1938 Planned mixed farming. Rehovol, Palestine: Agricul¬ tural Research Station. xt>, 154 p: map. Environmental Protection Authority, WA 1987 A guide to the Environmental Protection Act 1986. Perth: Environmental Protection Authority. 17 p. Eriksson N 1978 Kungl Vetenskaps-och Vitterhets-Samhallet i Goteborg 1778-1874. Goteborg: The Royal Society. 416 p: ill. Eriksson N 1985 Kungl. Vetenskaps-och Vitterhets-Samhallet i Goteborg 1875-1953. Goteborg: Rundquists Boktryckcri. 364 p ; ill Erofeyev B N 1976 Metodicheskoye posobiye po metallogenii. Moscow: Nedra. 271 p. Esenov Sh E 1973 Tektonika i neftegazonosnost solianokupolnykh oblastey S.S.S.R. Alma-Ata; Akademiya Nauk Kazakhskoi SSR. 263 p: maps. Esenov Sh E1974 Problemy metallogenii i rudogeneza. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 356 p. Etheridge R 1916 The cylindro-conical and comule stone implements of Western New South Wales and their significance — 2. The Warrigal or "dingo" introduced or indigenous? Sydney: Govt Printer. 54 p, 12 p plates. Etheridge R 1918 The dendroglyplis, or 'carved trees' of New South Wales. Sydney: Govt Printer, vii, 104 p 39 p plates: map. Ewart A J 1909 The weeds, poison plants, and naturalized aliens of Victoria. Melbourne; Govt Printer. 110 p, [33] p plates: col ill Ewart A J 1930 Flora of Victoria. Melbourne: University Printer. 1257 p; ill Ewart A J, O B Davies, J H Maiden, A A Hamilton & E Chcel 1917 The flora of the Northern Territory. Melbourne: Govt Printer, ru’i, 387 p 27, p plates; ill, map. Fain E E1971 Spektrograficheskoe oprcdelenic reniya v rudakh i mineralakh. Alma Ata: Akademiya Nauk Kazakhskoi SSR. 67 p: ill Federation of Australian University Staff Associatioas 1979 Report on re¬ search in universities. Melbourne: Federation of Australian University Staff Associations. 51 p. Fielding J W 1927|?1 Australasian ticks. Melbourne: Division of Tropical Hygiene of the Commonwealth Department of Health. 114 p: ill FiguierL 1873(?1 Reptiles and birds: A popularaccountof their various orders with a description of the habits and economy of the most interesting. London: Cassell uni, 624 p: ill Fitzgerald R D 1877 Australian orchids. Sydney: Govt Printer. Plates [with description]. Fitzpatrick K R1979 The Armidale area. Sydney: Dept of Mineral Resources and Development, x, 102 p; ill, maps. Fletcher] J 1893 Macleay memorial volume. Sydney: The Linnean Society of New South Wales. Ii, 308 p, 42 p plates: ill (some col), port. Flinders M1814 A voyage to Terra Australia... in the years 1801,1802 and 1803, in His Majesty's Ship the Investigator. London: G Nicol 2 v. Forbes H O 1903 The natural history of Sokotra and Abd-el-Kuri, being the report upon the results of the conjoint expedition to these islands in 1898- 9, by Mr. W. R. Ogilvie-Grant, of the British Museum and Dr. 110. Forbes, of the Liverpool Museums. Liverpool: Henry Young, xhii, 598 p: ill, maps. Francis W D1929 Australian rain-forest trees: Excluding the species confined to the tropics. Brisbane: Govt Printer. x(, 347 p; ill, map. Fryer G1993 Tlie freshwater Crustacea of Yorkshire: A faunistic and ecological survey. Yorkshire Naturalists' Union: Leeds Philosophical and Literary Society. 312 p: ill maps. Gardner C A 1931 Enumeratio plantarum Australiae occidentalis: A system¬ atic census of the plants occuring in Western Australia. Perth; WA Govt Printer. it>, 150 p. Garnet J R 1965 The vegetation of Wyperfeld National Park (north-west Victoria); A survey of its vegetation and plant communities, together with a check-list of the vascular flora as at December 1964. Melbourne: Field Naturalists Club of Victoria in conjunction with members of the Commit¬ tee of Management of Wyperfeld National Park. 95 p: ill map, plates. Geier P W... [cf al. j 1973 Insects: studies in population managument. Canberra: Ecological Society of Australia, vii, 295 p. Geikie A 1882 Geological sketches at home and abroad. London: Macmillan. X, 382 p: ill, maps. 88 Journal of the Royal Society of Western Australia, 77 (3), June 1994 Geikie A 1882 Text-book of geology. London: Macmillan, xi, 971 p: ill. Geisler W1930 Australien und Ozeanien. 3”* rev. Leipzig: Bibliographisches Institut. xi, 424 p: ill, map. Giles G M 1900 A handbook of the gnats or mosquitoes giving the anatomy and life history of the Culddae. London: John Bale, viii, 374 p: ill. Goldschmidt R B 1960 In and out of the ivory tower: the autobiography of Richard B. Goldschmidt. Seattle; Univ Washington Press, xiii, 352 p. Graetz R D & K M W Howes 1979 Studies of the Australian arid zone. 4. Chenopod shrublands: Proceedings of a symposium hold by the Rangelands Research Unit (Riverina Laboratories Deruliquin, N.S.W. 1975). Melbourne: CSIRO. v, 196 p: ill, maps. Gray J E1985 A report on the collection of native plants in Australian botanic gardens and arboreta. Canberra: Royal Australian Institute of Parks and Recreation, xi, 69 p: ill. Griffin J J 1873 Scientific handicraft: A descriptive, illustrated and priced catalogue of apparatus suitable for the performance of elememtary ex¬ periments in physics. Volume 1. Mechanics, hydrostatics, hydrodynamics and pneumatics. London: J J Griffin, xv, 186 p: ill. Guberlet M L 1956 Seaweeds at ebb tide. Seattle: University of Washington Press, xvi, 182 p: ill. Guilfoyle W R 1883 Catalogue of plants under cultivation in the Melbourne Botanic Gardens, alphabetically arranged. Melbourne: Govt Printer, xii, 200 p: ill. Gvozdetsky N A 1959 Speleology and studies in karst: Proceedings of the Conference on Speleology and Karst Study, December 17-18,1958. Mos¬ cow: Moscow Society of Natural History. Geographical Section. 200 p: ill, maps. Hall A R 1969 The Cambridge Philosophical Society: A history, 1819-1969. Cambridge: Cambridge Philosophical Society. 114 p: ill, ports. Hammond-Tooke W D1994 Creed and confession in South African ancestor religion. Cape Town: South African Museum, lip. Hansen H L & W Sorensen 1904 On two orders of Arachnida: Opiliones, especially the suborder Cyphophthalmi, and Ricinulci, namely the family Cryptostemmatoida. Cambridge: Cambridge University Press, xi, 182 p. Hansen V ... [et al] 1939 Catalogus coleopterorum Daniae et Fennoscandiae. Helsinki: Finska Velenskaps Societeten. 129 p: map. Harlog M, I B J Sollas, S J Hickson & E W MacBride 1906 The Cambridge Natural History. Vol. 1. London: Macmillan. 671 p: ill. Hedge IC & J M Lamond 1970 Index of collectors in the Edinburgh Herbarium. Edinburgh; HMSO. v, 147 p. Hegi G1906 Illustricrte Flora von Mittel-Europa. Munchen: J F Lehmann. 2 v. Hcndley Q B 1973 Early man. Cape Town: South African Museum. 26 p: ill. Henfrcy A & M T Masters 1878 An elementary course of botany, structural, physiological, and systematic. 3^ ed. London: van Voorst, xv, 741 p; ill. Heran 11989 Zvirata a fotographie. Prague: Narodni Muzeum. 94 p: ill. Hill C C1958 Spring flowers of the lower Columbia Valley. Seattle: University of Washington Press, xi, 164 p: ill. Hitchcock C L & A Cronquist 1973 Flora of the Pacific Northwest. Seattle: University of Washington Press, xix, 730 p: ill. Hitchcock C L ... [el al] 1955-1961 Vascular plants of the Pacific Northwest. Seattle: University of Washington Press. Parts 3, 4, and 5 only. Hoesch W1955 Die Vogelwelt Sudwestafrikas. Windhoek: Meinert. 300 p: ill. Hooker Sir J D1864-1867 Handbook of the New Zealand flora: A systematic description of the native plants of New Zealand and the Chatham, Kermadec's, Lord Auckland's, Campbell's, and Macquarie's Islands. London: Reeve. 2 v. Hooker Sir J D 1867 Handbook of the New Zealand flora: A systematic description of the native plants of New Zealand and the Chatham, Kermadec's, Lord Auckland's, Campbell's, and Macquarie's Islands. London: Reeve. 2 pts. in 1 v. (786 p). Hooker Sir] D1878 The student's flora of the British Islands. 2"'’ ed. London: Macmillan, xx, 539 p. Hooker, Sir W J 1813 Journal of a tour in Iceland, in the summer of 1809. Volume 1.2'"^ ed. London: Longman, evi, 369 p: ill, map. Hooker, Sir W J1837-1841 leones plantarum or figures with brief descriptive characters and remarks of new or rare plants selected from the author's herbarium. London: Longman. 4 v: ill. Hooker Sir W J 1842 The British flora. Volume 1. The Phaenogamous or flowering plants and the ferns. 5'*'ed. London: Longman, xxxviii, 464 p: ill. Hooker, Sir W J1844 Species filicum: Being descriptions of all known ferns. London: William Pamplin. 3 parts (192 p, 56 p plates); ill. Hooker, Sir W J & J G Baker 1868 Synopsis filicum, or a synopsis of all known ferns including the Osmindaceac, Schizaeaceae, Marattiaceae, and Ophioglossaceae. London: Robert Hardwicke. 482 p, 7 p plates: ill. Homy R & F Basil 1970 Type specimens of fossils in the National Museum Prague. Volume 1. Trilobita. Prague; National Museum. Museum of Natural History. 354 p: ill. Howard G E 1950 Lichens of the Slate of Wa.shington. Seattle: University of Washington Press, ix, 191 p, 11 p plates: ill, map. Hughes M R & A Chadwick 1989 Progress in avian osmoregulation. Leeds: Leeds Philosophical and Literary Society, xv, 346 p: ill, port. Hull E 1873 The coal-fields of Great Britain; Their history, structure and resources with notices of the coal-fields of other parts of the world. 3"* ed. rev. and enl. London: Stanford, xxiii, 499 p: ill, maps. Hurst E 1942 The poison plants of New South Wales. Sydney: New South Wales Poison Plants Committee, xiv, 498 p: ill. Hutchinson H N 1890 The autobiography of the earth: A popular account of geological history. London: Stanford, xiv, 290 p: ill. Hutton F W1872 Catalogue of the Edunodermata of New Zealand. Welling¬ ton, NZ: Colonial Museum and Geological Survey Dept 17 p. Hutton F W1873 Catalogue of the marine Mollusca of New Zealand. Welling¬ ton, NZ: Colonial Museum and Geological Survey Dept, xx, 116 p. Hutton F W 1873 Catalogue of the tertiary Mollusca and Echinodermata of New Zealand. Wellington, NZ: Colonial Museum and Geological Survey Department, xvi, 48 p. Hutton F W 1880 Manual of the New Zealand Mollusca. Wellington, NZ: Colonial Museum and Geological Survey Dept, xvi, 224 p. Hutton F W & J Hector 1872 Fishes of New Zealand: Catalogue with diagnosis of the species. Wellington, NZ; Colonial Museum and Geological Survey. XV, 133, a, 12 p plates: ill. Huxley T H1877 American addresses with a lecture on the study of biology. London: Macmillan. 164 p; ill. Huxley T H 1877 Lay sermons, addresses and reviews. 6* ed. London: Macmillan, xi, 344 p. Ingwersen F1976 Vegetation of the Jervis Bay territory. Canberra: AGPS. vi, 80 p: ill, map. Inshin P V 1972 O mekhanizmakh differentsiatsii magmy. Alma Ata: Akademiya Nauk Kazakhskoi SSR. 247 p: ill. Institute of Agriculture and Natural History (Tel-Aviv, Palestine) 1926 First report covering a period of five years 1921-1926. Tel-Aviv: Institute of Agriculture and Natural History. 103 p: ill, port. International Aphidological Symposium 1985 Evolution and biosystematics of aphids: Proceedings of the International Aphidological Symposium at Jablonna, 5-11 April, 1981. Warsaw: Polish Academy of Sciences. 510 p: ill. International Botanical Congress 1961 Recent advances in botany: From lectures and symposia presented to the IX International Botanical Con¬ gress, Montreal, 1959. Volume 2. Toronto: University of Toronto Press, xiv, 951-1766 p: ill. Issac L A 1949 Better Douglas fir forests from better seed. Seattle: University of Washington Press, x, 64 p: ill, map. Jackson L d'A 1880 Aid to survey-practice: For reference in surveying, levelling, and setting-out; and in route-surveys of travellers by land and sea. London: Lockwood, xx, 350 p: ill. Jackson S W1907 Egg collecting and bird life of Australia: Catalogue and data of the "Jacksonian zoological collection". Sydney: White. 171 p: ill. James M J 1991 Galapagos marine invertebrates: Taxonomy, biogeography, and evolution in Darwin's islands. New York: Plenum Press, xiv, 474 p: ill, maps. Jewett S G ... [et al] 1953 Birds of Washington State. Seattle; University of Washington Press, xxxiii, 767 p, 12 p plates; ill (some col), maps. Johns R K 1976 History and role of government geological surveys in Aus¬ tralia. Adelaide: Govt Printer. Ill p: ill, ports. Johns R K1986 Cornish mining heritage. Adelaide; Department of Mines and Energy, South Australia. 50 p: ill, map. Johns R K 1988 Mineral resources of the Adelaide geosyncline. Adelaide: South Australian Department of Mines and Energy. 99 p: col ill, maps. Johnson R W 1976 Volcanism in Australasia: A collection of papers in honour of the late GAM Taylor. Amsterdam; Elsevier, xvi, 405 p: ill, maps, port. Johnson R W 1993 Volcanic eruptions and atmospheric change. Canberra: Australian Geological Sur\'ey Organisation, Dept Primary Industries and Energy. 36 p: ill (.some col), map. Johnston C D ... [ef al] 1983 Water movement through preferred paths in lateritic profiles of the Darling Plateau, Western Australia. Melbourne: CSIRO. 34 p: ill (some col). Jones R & J Allen 1985 Archaeological research in Kakadu National Park. Canberra: Australian National Parks and Wildlife Service, xviii, 317 p: ill, maps. Jones T R 1868 The animal creation: A popular introduction to zoology. London: Society for Promoting Christian Knowledge, xi, 607 p: ill. Jukes J B1876 The school manual of geology. 3"* ed. rev. and enl. Edinburgh: Adam & Charles Black, xvi, 402 p: ill. Kaljo D L 1977 Fades and fauna of the Baltic Silurian. Tallinn; Academy of Sciences of the Estonian SSR. 286 p: ill. Kayupov A K... [et al] 1975 Geology and metallogeny of the anticlinorium of Zhaman-Sarysuisk. Alma-Ata; Akademiya Nauk Kazakhskoi SSR. 222 p: ill. 89 Journal of the Royal Society of Western Australia, 77 (3), September 1994 of iron and manganese ores of western Kayupova Ka^khslan.^lma-Ata; Akademiya Nauk Kazakhskoi SSR. 232 p: isopods of southern Africa, Cape Town: KensleyBl^^ African Museum. 173 p: ill- Trustees of trie so Permo-Triassic: The fortunes of the King G M 1990 L«fe®™ ^^ tiios.CapeTown: South African Museum in dicynodont Royaf Society of South Africa. 17 p: port, oollaborarion ^ lepidop«rist's guide. London: van Voorst. 122 p. Knaggs H G W9^ balearica. Etude phytogeographique sur les lies Knoche H ' mcr- (nstilut de Botanique. 4 v. The Noonkanbah story. Dunedin: University of Otago Press. KoligE iw/ ‘o E & K Heider 1895-1900 Text-book of the embryology of mverte- Ata: Akademiya of tbe Pacific Northwest: An illustrated Kozloff E N 1976 Plan_ Western Oregon, Washington and British SumbiStl^: uldvemity of Washington Press, ix. 264 p, 48 p plates: ‘TS'ceologv of the north-east of Asia. Moscow: Nedra 4 v. Krasny LI formatsii yugo-zapadnogo altaya i ikh Kuzebnyi^ V ®^yg^^|n,a.Ata: Akademiya Nauk Kazakhskoi SSR. 342 p: ill, folded maps. Handbook of structural timber design. 3"' ed. UnglandsI&AJ 1 Melbourne. , n van & W M D van Lecuwen 1926 The Zoocedidia of NeSInds Lst Indies. Batavia: Drukkerij de Unie, 601 p, 7 leaves of plates: ill. ^^orld of man and the world of spirit: An ^^'^nt^retrtion of the Linton rock painHngs. Cape Town; South African Museuin. 16 P'treasury of botany: A popular dictionary of the Lindley J & T M« New rev. ed. London: Longmans. 2 v: ill. UndTefj“^ e’’a’’iq3TFiftv years of museum work: Autobiography, unpublished S Swhy f Frrfync A. Ne» V.,t A™ Museum of Natural History. 1,81 p: ill, ports. .u t M H & 1M Lindsay 1985 Stratigraphy, palaeontology, malacology: honour of Dr Nell Ludbrook. Adelaide: Govt Printer, tii-i, 387 p: Ludbrook N H1984 Quaternary molluscs of South AustraUa. Adelaide; Govt Printer 327 p: iH (some col). Lunney D 1992 Zoology in court. Mosman, NSW: Royal Zoological Soaety of New South Wales, u, 90 p: ill port. Lvovskoe Geologichcskoe Obshchestvo 1967 Voprosy geologn Karpat. Lvov: Lvovskii Universitet. 188 p: Ul maps. j • x/i 1 «i Dt* I'influence de remetique sur 1 homme et les ammaux. ^^Coim lu a rpmlr7classe de 1’ Institut de France, le 23 aout 1813. Mai^rj hTe Betche 1916 A census of New South Wales plants. Sydney: Govt Printer, xx, 216 p. , . , ^ - i ^ Maiden J H1889 The useful plants of Australia including Tasmania. London: Trubner. xii, 696 p. Maiden J H 1904-1925 The forest flora of New South Wales. Sydney: Govt Printer. 77 parts in 8 v. Maiden J H 1907 Sir Joseph Banks ; the "father of Australia". Sydney; Govt Printer. x.rii>, 244 p; ill, maps, ports. , ^ , . , _ Mansfield L L1961 Handbook on quarrymg. Adelaide: South Australia Mines Dept. 185 p; ill. , Marie R & P Lesne 1917 Catalogue des Coleopteres de la region Malgache, Paris; Imprimerie Nationale. iv, 180 p. Markham N L & H Basden 1974 The mineral deposits of New South Wales. Sydney: Geological Survey of New South Wales. 682 p: ill. Martens E von 1873 Critical list of Mollusca of New Zealand contained in European Collections. Wellington, NZ: Colonial Museum and Geological Survey Dept, v, 51 p. Martin A W 1961 Comparative physiology of carbohydrate metabolism in heterothermic animals. Seattle: University of Washington Press. 144 p, 2 p plates: ill. Matsuy V M1973 Pozdnii Kainozoi kazakhstanskogo Priirtyshya. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 143 p. McAlpine D1895 Systematic arrangement of Australian fungi, together with host-index and list of works on the subject. Melbourne; Govt Printer, vi, 236 p. McCallum F1926 An epidemiological report on an outbreak of small-pox on ship-board. Melbourne: H J Green, Government Printer. 29 p. McCarthy F D1938 Australian Aboriginal decorative art. Sydney: Australian Museum. 48 p: ill (some col). McCoy F 1878-1889 Prodromus of the zoology of Victoria; Or, figures and descriptions of the living species of classes of the Victorian indigenous animals. Melbourne: Govt Printer. 2 v (375; 200 p plates); ill (some col). MecklenburgC von 1974 Saccus vasculosus in some teleosts. Lund: University of Lund. 35 p. Michaelsen J W1888-1936 Collected papers: Oligochaeta. 5 v. Michaelsen J W 1889-1934 Collected papers: Miscellaneous. 2 v. Mill J S1865 Principles of political economy with some of their applications to social philosophy. London: Longmaas. xx, 591 p. Millar A 1978 Orchids of Papua New Guinea: An introduction. Seattle: University of Washington Press. 101 p: col ill. Moore C & E Betche 1893 Handbook of the flora of New South Wales: A description of the flowering plants and ferns indigenous to New South Wales. Sydney: Govt Printer, xxxix, 582 p. Morrissy N 1969 Manon of Western Australia {Cherax tenuimanus). Perth: Dept of Fisheries and Fauna. 18 p: ill, map. Morton H1982 The whale's wake. Dunedin: University of Otago Press. 396 p: ill, plates. Mueller F Baron von 1869-1881 Fragmenta phytographiae Australiae. Mel¬ bourne: Govt Printer. 4 v. Mueller F Baron von 1874 Obser\'ations on new vegetable fossils of the auriferous drifts. Melbourne: Govt Printer. 31 p: ill. Mueller F Baron von 1879-1883 Eucalyptographia: A descriptive atlas of the eucaly pts of Australia and the adjoining islands. Melbourne: Govt Printer. 9 decades. Mueller F Baron von 1882 General information respecting the present condi¬ tion of the forests and timber industry of the southern part of the colony, with some remarks and suggestions on future conservation and manage¬ ment of the timber areas, from various authorities. With a reprint of the regulations and laws in force for the renting or leasing of timbered lands, together with a report on the forest resources of the colony. Perth: Govt Printer. 27 p, 30 p, 20 p plates: ill, folded map. Mueller F Baron von 1883 Observations on new vegetable fossils of the auriferous drifts. Melbourne: Govt Printer. 23 p: ill. Mueller F Baron von 1885-1888 Key to the system of Victorian plants. Mel¬ bourne: Govt Printer. 2 v. Mueller F Baron von 1886 Description and illustration of the myoporinous plants of Australia. Part. 11. Lithograms. Melbourne; Govt Printer. 74 p plates. Mueller F Baron von 1887-1888 Iconography of Australian species of acacia and cognate genera. Melbourne: Govt Printer. 13 decades in 1 v: chiefly ill. Mueller F Baron von 1888 Select extra-tropical plants readily available for industrial culture or naturalisation with indications of their native coun¬ tries and some of their uses. 7* ed., rev. and enl. Melbourne: Govt Printer, ix, 517 p. Mueller F Baron von 1889 Iconography of Australian salsolaceous plants. Melbourne: Govt Printer. 1 decade: chiefly ill. Mueller F Baron von 1889 Second systematic census of Australian plants, with chronologic, literary and geographic annotations. Part 1 Vasculares. Mel¬ bourne: Govt Printer. 244. Mueller F Baron von 1891 Select extra-tropical plants readily eligible for industrial culture or naturalisation, with indications of their native coun¬ tries and some of their uses. 8'^ ed. Melbourne: Govt Printer, viii, 594 p. Mueller F Baron von 1892 Iconography of candolleaceous plants. Melbourne: Govt Printer. First decade, 10 p plates. Mueller F Baron von 1895 Select extra-tropical plants readily eligible for indu-strial culture or naturalisation, with indications of their native coun¬ tries and some of their uses. 9'^ ed, rev and enl. Melbourne: Govt Printer, xi, 654 p. Mueller F Baron von & M Fraser 1879 Report on the forest resources of Western Australia (being appendix to "General information respecting the present condition of the forest and timber trade..." by Malcolm Fraser). London; L Reeve. 30 p, 20 p plates: ill. Mukhamedzhanov S M ... \etal.] 1965Podzemnye vody khrebta Tarbagatai i ego ravninnykh predgorii. Alma Ata: Akademiya Nauk Kazakliskoi SSK Institut Geologichesldkh Nauk im K1. Satpaeva. 148 p: ill, maps. Murray RAF 1887 Geology and physical geography (of Victoria). Melbourne: Govt Printer, in, 179 p: ill, map. Nachtrieb H F, E E Hemingway & J P Moore 1912 The leeches of Minnesota. Minneapolis, Minn: Geological and Natural History Survey of Minnesota. vi, 150 p: ill. Nansen F1900-1905 The Norwegian North Polar expedition 1893-1896: Scien¬ tific results. Christiana: Fridtjof Nansen Fund. 6 v: ill. Narvail G E ... (et fll.) 1974 Mednoyc orudneniye Mugodzhar. Alma-Ata: Nauka. 176 p. Nason H B & C F Chandler 1874 Elderhorst's manual of qualitative blow-pipe analysis and determinative mineralogy. 4*^ ed. rev. and enl. London: T Edward 2^11. xr, 310 p; ill. 90 Journal of the Royal Society of Western Australia, 77 (3), September 1994 National Museum of Victoria; National Gallery of Victoria 1955 Exhibition of works of primitive art of the National Museum of Victoria, the ethno¬ graphical collection of the University of Melbourne, and private collec¬ tions : Held on the occasion of the thirty-first meetingof the Australian and New Zealand Association for the Advancement of ^ence, Melbourne 17- 24 August, 1955. Melbourne; National Gallery of Victoria. 19 p: ill. Nevins A 1941 This is England today. New York: Scribner, x, 164 p. Nicholson H A1879 A manual of palaeontology for the use of students, with a general introduction on the principles of palaeontology. 2*^ ed. rev. and enl. Edinburgh; Blackwood. 2 v (511:531). Nikitin 1F 19730rdovikKazakhstana. Chast2. Paleogeografiya, paleotektonika. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 99 p; ill. Osborne G D 1950 The Structural evolution of the Hunter-Manning-Myall Province, New South Wales. Sydney; Royal Society of New South Wales. 80 p: ill, folded maps. Palanca-Soler A 1987 Aspcctos faunisticos y ecologicos de lepidopteros altoaragonesas. Antonio Palanca-Soler. Madrid: Consejo Superior de Invcsligadones Cientificas. 317 p: ill. Patalakha EI & T V Giorgobiani 1975 Struktumyi analiz lineynoy skladchatosti na primere khrebta Karatau, kaledonskii Isykl. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 195 p: ill, folded maps. Patalakha E1... |d al] 1974 Geneticheskiye tipy geosinknalnoy skladchatosti (Kazakhstan). Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 208 p. Patin S A 1982 Pollution and biological resources of the oceans. London: Bulterworth. xi, 287 p: ill. Patten T 1986 Scotland and Petrochemicals: Proceedings of the symposium held in Edinburgh, 4 and 5 June 1985. Edinburgh; Royal Society of Edinburgh. 121 p: ill, maps. Paxton Sir J & S Hereman 1868 Paxton's botanical dictionary comprising the names, history, and culture of all plants known in Britain including all the new plants up to the present year revised and corrected by Samuel Hereman. New ed. London: Bradbury, xii, 623 p. Pedrocchi-Renault C 1987 Fauna omitica del Alto Aragon Occidental. Ma¬ drid: Consejo Superior de Investigaciones Cientificas. 210 p, plates. Perkins G H 1908 Report of the state geologist on the mineral industries and geology of certain areas of Vermont. Concord, NH: Rumford. Xi, 302 p: ill, maps. Peron F, L de Frcydnet, C A Lesueur & N-M Petit 1807-1816 Voyage de decourvertes aux terres Australes execute par ordre de sa majesle I'empereur et roi sur les corvettes le Geographe, Ic Naturaliste, et la goelette la Casuarina, pendant les annees 1800,1801,1802,1803, et 1804. Paris: Imprimerie Imperiale. 3 v. Pert M 1963 Pinus radiata: A bibliography to 1963. Canberra: Forestry and Timber Bureau, ix, 145 p. Pescott R T M 1954 Collections of a century: The history of the first hundred years of the National Museum of Victoria. Melbourne: National Museum of Victoria, xiii, 186 p; ill, port. Pettibone M H 1953 Some scale-bearing Polychactes of Puget Sound and adjacent waters. Seattle: University of Washington. 89 p, 40 p plates. Playford P E 1988 Guidebook to the geology of Rottnest Island. Perth; Geological Society of Australia (WA Division) and Geological Survey of Western Australia. 67 p: ill, maps. Pocock, R 1.1928 Guide to the Arachnida, millipedes and centipedes exhibited in the Department of Zoology, British Museum (Natural History). Lon¬ don: The Trustees of the British Museum (Natural History). 56 p: ill. Popov V I 1963 Rukovod.stvo po opredeleniyu osadochnykh fatsialnykh komplcksov i mctodika...kartirovania. Leningrad: Gostoptekhizdat. 713 p: ill, maps. Privat-Deschanel P & M Zimmermann 1930Geographie Universelle. Volume 10. Paris: Armand Colin. 368 p; ill. Raath M A1984 Dinosaurs and diatremes; The life and work of Sidney Henry Haughlon, 1888-1982. Cape Town: South African Museum. 10 p. Radchenko M11967Kamennougolnaia flora yugo-vostochnogo Kazakhstana. Alma-Ata: Akademia Nauk Kazakhskoi SSR. 75 p: ill. Ravikovitch S 1960 Soils of Israel. Rehovot, Israel: Hebrew University of Jerusalem, xix, 89 p, plates: ill. Raychaudhuri S P 1977 A manual of virus diseases of tropical plants. Delhi: Macmillan, xix, 299 p; ill. Reed D W 1987Spirit of enterprise: The 1987 Rolex awards. Wokingham, Berk: Van Nostrand Reinhold, xx, 460 p: ill (some col). Reed D W 1990 Spirit of enterprise; The 1990 Rolex awards. Bern: Buri International, xxii, 489 p: col ill, col ports. Reed D W1993 Spirit of enterprise: The 1993 Rolex awards. Bern: Buri. xxii, 489 p: col ill, col ports. Rehn JAG 1953 The grasshoppers and locusts (Acridoidea) of Australia. Volume 2. Family Acrididae (subfamily Pyrgomorphinae). Melbourne: CSIRO. 270 p, 32 p plates: ill. Renwick A & S R Ross 1980 The Australian geological delegation to China. Canberra: AGPS. xvi, 130 p: ill. Reynolds B 1984 Material culture: A system of communication. Cape Town: South African Museum. 13 p: port. Roberts R W1989 Land conservation in Australia: A 200 year stocktake. West Pennant Hills, NSW: Soil and Water Conservation Association of Aus¬ tralia. 32 p: col ill. Rohacek J & S A Marshall 1985 The genus Trachyopella Duda (Diptera, Sphaeroceridae) of the Holarctic Region. Torine: Musoe Regionale di Science Nalurali. 109 p: ill. Romanes G J 1882 The scientific evidence of organic evolution. London: Macmillan, vi, 88 p. Rotay A P1975 The main features of carboniferous stratigraphy of the USSR. Leningrad: Nedra. 334 p: maps. RSPCA Australia 1987 Incidence of cruelty to wallabies in commercial and non-commercial operations in Tasmania: Report to the Australian Na¬ tional Parks and Wildlife Service by RSPCA Australia. Fyshwick, ACT: RSPCA Australia. m,143 p: ill. Rutherford W1880 A text book of physiology. Edinburgh: Black, vi, 160 p: ill. Ryman N & F Utter 1987 Population genetics and fishery management. Seattle: University of Washington Press, xvii, 420 p. Savillc-Kent W 1893 The Great Barrier Reef of Australia: Its products and potentialities. London: W H Allen, xvii, 387 p: ill (some col), map. Scheibner E1974 Explanatory notes on the tectonic map of New South Wales. Sydney: Geological Survey of New South Wales, xi, 283 p: ill. Scherzer K 1861-1863 Narrative of the circumnavigation of the globe by the Austrian frigate Novara ... in the years 1857, 1858 & 1859. London: Saunders. 3 v. Schnabel C & H Louis 1898 Handbook of metallurgy. London: MacMillan. 2 v ( xvi, 876; xii», 732). Schwarz F1961 You can trust the communists (... to do exactly as they say!). Englewood Cliffs, NJ: Prentice-Hall. 187 p. Schweitzer F R1975 Die steentydperk; The stone age. Cape Town: Trustees of the South African Museum. 20p: ill. Scott R H1890 Elementary meteorology. 5* ed. London; Kcgan Paul, Trench, Trubner. xiv, 410 p: ill. Scott W B 1903-1915 Reports of the Princeton University expeditions to Patagonia, 1869-1899. Princeton, N J: Princeton University Press. 8 v. Sedgwick A 1897-1905 A student's text-book of zoology. London: Swan ^nncnschein. 2 v: ill. Selby J 1987 South Australia's mining heritage. Adelaide: Department of Mines and Energy South Australia, jointly with the Australasian Institute of Mining and Metallurgy. 203 p, ill, maps, ports. Semon R1899 In the Australian bush and on the coast of the Coral Sea, being the experiences and observations of a naturalist in Australia, New Guinea and the Molucca.s. London: Macmillan, xv, 552 p: ill, maps. Seyfullin S Sh & N N Nuralin 1964 Geologo-struklurnyie usloviya formirovaniya mestorozhdeniya Dzhezkazgan. Alma-Ata: Akademiya Nauk. 77 p. Sharpe G & W Sharpe 1954101 wildflowers of Olympic National Park. Seattle: University of Washington Press. 40 p: ill. ShatalovET1972MctaIlogenicheskii analiz rudokontroliruyushchikhfaktorov v rudnykh rayonakh. Moscow: Nedra. 296 p. Shaw E M & P Davidson 1973 Die suid-Nguni; The southern Nguni. Cape Town: Trustees of the South African Museum. 28 p: ill, map. Shaw E M 1971 Die Boesman; The Bushmen. Cape Town: South African Museum. 17 p: ill. Shaw E M 1972 Die Hottentotl; The Hottentots. Cape Town: South African Museum. 20 p: ill. Shcheglov A D 1968 Metallogeny of the regions of autonomous aclivization. Leningrad: Nedra. 180 p. Simmonds N W1976 Evolution of crop plants. London: Longman, xii, 339 p. Smyth R B 1878 The Aborigines of Victoria: With notes relating to the habits of the natives of other parts of Australia and Tasmania, [compiled from various sources for the Government of Victoria]. Melbourne: Govt Printer. 2 V {Ixxii, 483 p; w, 456 p); ill, folded maps. Sollas W J1905-1906 The rocks of Cape Colville Peninsula, Auckland, New Zealand. Wellington: Govt Printer. 2 v: ill, map. Sousa A F G e 1960 Dendrologia de Mocambique. V. Distrito de Manica e Sofala. Lourenco Marques: Impresna Nadonal de Mocambique. 217 p: ill, map. Sousa A F G e & A De P Bartolomeu 1951 Dendrology of Mozambique. 1. Some commercial timbers. Lourenco Marques]?]: Imprensa Nacional de Mo¬ cambique. 247 p: ill (some col). Spencer B1915 Guide to the Australian ethnological collection exhibited in the National Museum of Victoria. 2"^ ed. Melbourne: National Museum. 128 p: ill. 91 f the Royal Society of Western Australia, 77 (3), September 1994 Journal . ',-^>1 rollectionexhibitedinthe SpencerB1922Guideto*e^A^^^^^^^ NaUonal Museum of National ■ .hps Wien, Fuhrer durch die Victoria. 142 p: Naturhislonsches SpitzenbcrgerF 198J Wi^emapplications. London: Schausammiung- theory, source, Sprague ] T 1875 Electna Alexander Teixeira de Stewart ^,^/®^Jied to clear water. 7* e Missouri Botanical particularly apph nnraofMisso^*^-^*^® Steyermark J A1940 flowers 15*^ ed. London: Garden. nL 582 p: lU- vegetables and flowers. Sutton and Sons 1913 The cu 1973-1975. Simpkin.454p:m. ^ bibliography. Suppl^-ent for o'f'^'l^ncw o7lhe USSR. l^P^Pj^‘^^^^fi<.ExploringExpei^^ TateR18'f2SdentificresuUso j^^^5j„lia.236p, ppa . . Adelaide, SA: Royal i«,icdovaruya dzhezkazganskikh swit TazhibayevaPT1964Lifolo^*^j^^g^^^ vsviazisproblemoigei’^t. . . ggR. 277 p; db'"‘“ps. Ata: Akademiya Nauk Kaza jg^^jbAustraliangeology. Adelaide: Teesdale-SmithEN1959Bibhograp y Govt Printer. 240 p: -"^P' ^^^e conservaUon outside rescues: A Thackway R & F Stevenson 19^ ^ promoting nature ^ summary of government ^gr Australian National Parks lands outside parks and and Wildlife Service, u, 62 p. Freshwater Science list of "■TSir. SSS. N.« - tute. 135 p: maps. . Water Sciences list of publications 1954- ll,ompson R M C 1^89 Divis'on ^ Oceanographic 1986. Supplement No k 1989. vve ^ Institute, Division of Saences, , The non-metals. London: Thorpe T E1874 A manual of inorgara Collins. 399 p: ill. «,e Myxophyceae of North America Tilden J1910 Minnesota ^'gf ■ , America, Greenland, Bermuda, and adjacent regions University of Minnesota Press. the West Indies and Hawaii. Minneapolis, u 328 p, 20 p plates: ilk half-caste problem in South Australia. Tindale N B 1940-19H . codety South Australian Branch. Adelaide: R°y“'bb 1976 Natural history of the Adelaide Twidale C R, M ] Tyler & B P . Australia. 189 p: ill, maps. *d,Uld, Rd,.l cl«8dd "I ".."OS'.pW. U.S.S.R. State Geological ^omm He 9 Mu.seum. Leningrad: Nedra. paleontological collechons kept m CNIGK MU P' , ,.^1- RJfklev WA: Perth Observatory. Utting M 1989 Cooke's Perth Observatory. Bickley, w 22 p: ill, ports^ Mamiferos Vivientes y Extinguidos de las '''TntiUru SranTcuba: Ins.ituto de Zoologia de la Academia de United States and such forei^londs Washington: Govt IiLa 1^Ica11kcv A* AW , woe --kinds as have been introduced... with an pT;n“th"e:;:iiSr imposition of grasses. Washington: Govt “%92elticheskaiageologiia:RefemHvn^^ l::LtLnovnoiliteraturypo.l-Mathemat.caI geology:^ f basic literature to 1968). Leningrad: BAN. 2^ p. :er J B & D M Stoddart 1993 Walk to the west. Hobart: Artemi-s on behalf f the Royal Society of Tasmania. iJcmit, 76 p: ill (some col), folded map, 'orts. arth's surface. London: Macmillan. 2 v; ill, maps, ice A R1886The Malay Arcliipelago: The land of the orang-utan and bird f paradise, a narrative of travel, with studies of man and nature. London: lacmillan. xvi, 653 p: ill, maps. on W & W Bean 1890 Orchids; Their culture and management, with escriptions of all the kinds in general cultivation. London. Gill, ix, 554 p. 1 (some col). Weaver C E1931 Paleontology of the Jurassic and Cretaceous of west central Argentina. Seattle: University of Washington Press, xv, 594 p: col ill, maps. Webb L 1968 Nature protection in Europe. Sydney: Wildlife Preservation Society of Australia. 50 p. Wells M J1962 Brain and behaviour in cephalopods. Stanford, Calif: Stanford University Press. 171 p: ill. Welsch U & V Storch 1976 Comparative animal cytology and histology. Seattle: University of Washington Press, xiv, 343 p: ill. Western Australia. Dept of Mines 1966 Mineral resources of Western Aus¬ tralia. Perth: Govt Printer. 84 p: ill (some col), maps. Western Australia. Metropolitan Region Planning Authority 1970 Report 70; The Perth region, a ten year review, Metropolitan Region Planning Authority, Western Australia. Perth: Govt Printer. 28 p: ill. Western Australian Greenhouse Coordination Council 1989 Addressing the greenhouse effect: A discussion paper. Perth: Environmental Protection Authority. 21 p. Wheeler G C & J N Wheeler 1986 The ants of Nevada. Los Angeles: Natural History Mu.seum of Los Angeles County. Pii, 138 p: ill (some col), maps. White, E. G. 1972 Directory of New Zealand entomology. Nelson, NZ; Ento¬ mological Society of New Zealand. 56 p. Whiting J W & R E Relph 1962The occurrence of opal at Lightening Ridge and Grawin, with geological notes on County Finch. Sydney: Dept Mines. 7-21: map. Willis J H 1941 Victorian fungi: A key and descriptive notes to 120 different toadstools, family Agaricaceae, with remarks on several other families of the higher fungi. Melbourne: Field Naturalists' Club of Victoria. 72 p, 16 p plates; ill (some col). Wood J G & T Wood [18..?] The field naturalist's handbook. London; Cassell, Pettcr, Galpin. 167 p. Woodward S P, R Tate, A N Waterhouse & J W Lowry 1868 A manual of the Mollusca: A treatise of recent and fossil shells. 2'^ ed. London: Virtue, xiv, 542 p, 23 p plates, 86 p appendix: ill, folded map. Wright W P 1923[?| Cassell's dictionary of practical gardening: An illustrated encyclopaedia of practical horticulture for all classes, special edition. London: Waverlcy Book Co. 4 v. (480 p): ill (some col). Yocom C F 1951 Waterfowl and their food plants in Washington. Seattle: University of Washington Press, xvi, 272 p : ill. Zayas F de 1974 Enlomofauna Cubana; Topicos entomologicos a nivel medio para uso didactico. Tomo 111. Subclase Polyneoptcra. La Habana: Institute Cubano del Libro. 130 p: ill. Zhevago V S 1972 Geotermiya i termalnyie vody Kazakhstana. Alma-Ata: Akademiya Nauk Kazakhskoi SSR. 254 p: ill, maps. JOURNALS (Current) Abhandlungen Herausgegeben Vom Naturwissenschaftlichen Verein Zu Bremen Abstracta Botanica Acta Biologica Cracoviensia. Series: Botanica Acta Biologica Cracoviensia. Series: Zoologia Acta Biologica Hungarica Acta Botanica Fennica Acta Musei Macedonid Sdentiarum Naturalium [= Izdanija Prirodonaucen Muzej Na Makedonija.) Acta Omithologica Acta Physica Hungarica Acta RegiaeSodetalis Scientiarum Et Litterarum Gothoburgensis: Geophysica Acta Regiae Sodetatis Sdentiarum Et Litterarum Gothoburgensis: Botanica Acta Regiae Sodetatis Sdentiarum Et Litterarum Gothoburgensis: Zoologica Acta Sodetatis Zoologicae Bohcmoslovacae Acta Zoologica Fennica AGSO Bulletin AGSO Journal Of Australian Geology & Geophysics AGSO Yearbook American Museum Novitales Anales De La Sodedad Cientifica Argentina Annalen Des Naturhistorischen Museums In Wien Annales Botanid Fennid Annals HLstorico-Nalurales Musei Nationalis Hungarici Annales Zoologid Armales Zoologid Fennid Annals Of Carnegie Museum Annals Of The Cape Provincial Museums (Human Sdences) Annals Of The Cape Provincial Museums (Natural History) Annals Of The Natal Museum Annals Of The South African Museum Annuaire - Academic Royale De Belgique 92 Journal of the Royal Society of Western Australia, 77 (3), September 1994 Annual Report Of The Australian Nature Conservation Agency Annual Report Of The Chief Directorate Marine Development, Department Of Environment Affairs Annual Report Of The CSIRO Annual Report Of The Department Of Mines And Energy, South Australia Annual Report Of The Environmental Protection Authority, Western Aus¬ tralia Annual Report Of The Institute Of Geoscience, The University Of Tsukuba Annual Report Of The International North Pacific Fisheries Commission Annual Review Of Ecology And Systematics Anthropological Records Anthropology News; Proceedings Of The Anthropological Society Of West¬ ern Australia Anuario Da Sociedade Brotcriana Aracnologia Arsbok - Kungliga Vetenskaps- Och Vitterhets-Samhallet I Goteborg Arsbok. Serie A - Societas Scientiarum Fennica AtDella Societa Italiana Di Scienze Naturali Atlantica Austrobaileya Beaufortia Belgian Journal Of Botany: Bulletin De La Sodete Royale De Botanique De Belgique Biological Bulletin Biological Notes Of The Ohio Biological Survey Biological Reviews Biologiske Skrifter Biology And Envirorunenl: Proceedings Of The Royal Irish Academy, Sect B Biullelen Moskovskogo Obshchestva Ispytatelei Prirody. Otdel Biologicheskii [* Bulletin Of The Biological Section, Moscow Sodety For Natural Re¬ search) Biulleten Moskovskogo Obshchestva Ispytatelei Prirody. Otdel Geologicheskii Bulletin Of Tlie Geological Section, Moscow Society For Natural Research) Boletim Da Sociedade Broteriana Boletim De Fisiologia Animal Boletim De Zoologia Boletim Ig-Dsp Publicacao Especial. Institute De Geociencias, Universidade De Sao Paulo Boletim Ig-Usp Serie Cientifica. Institute De Geociendas, Universidade De Sao Paulo Boletim Paranaense De Geodencias Bolcdn De La Sodedad Zoologica Del Uruguay Bolleti De La Societal D'historia Natural De Les Balears Bollettino Del Museo Civico Di Storia Naturale Di Verona Bothalia Bulletin De L'Institut Royal Des Sciences Naturclles De Belgique: Biologic Bulletin De L'Institut Royal Des Sdenccs Naturelles De Belgique: Entomologie Bulletin Do L'Institut Royal Des Sdences Naturelles De Belgique: Sciences De La Terre Bulletin De La Classe Des Sdences Bulletin De La Sociele D'Histoire Naturelle De Toulouse Bulletin De La Sodete Sdentifique De Bretagne Bulletin Du Musee D'Histoire Naturelle De Marseille Bulletin Du Museum National D'Histoire Naturelle. Miscellanea Bulletin Du Museum National D'Histoire Naturelle. Section C; Sdences De La Terre Bulletin Du Museum National D'Histoire Naturelle: Botanique Bulletin Of Applied Botany, Genetics And Plant Breeding (Leningrad) [= Trudy Po Prikladnoi Botanike, Genetike I Selektsii] Bulletin Of The American Museum Of Natural History Bulletin Of The Auckland Institute And Mu.seum Bulletin Of The Biological Section, Moscow Sodety For Natural Research Bulletin Of The Carnegie Museum Of Natural History Bulletin Of Tlie Florida Museum Of Natural History, Biological Sciences Bulletin Of The Geological Institutions Of The University Of Uppsala Bulletin Of iTie Geological Section, Moscow Sodety For Natural Research Bulletin Of The Geological Survey Of Canada Bulletin Of The Geological Survey Of New South Wales Bulletin Of The Geological Survey Of South Australia Bulletin Of The Geological Survey Of Western Australia Bulletin Of The International North Padfic Fisheries Commission Bulletin Of The Museum Of Comparative Zoology Bulletin Of The Natural History Museum. Geology Series Bulletin Of The Ohio Biological Survey Bulletin Of The Osaka Museum Of Natural History Bulletin Of The Royal Sodety Of New South Wales Bulletin Of The Toyama Sdence Museum Bulletin Of The University Of Nebraska State Museum Bulletin Zoologisch Museum Caribbean Journal Of Sdence Ceylon Journal Of Medical Sdence Ceylon Journal Of Sdence: Biological Sdences Cimbebasia Cimbebasia. Memoir Circular Of The Geological Survey, Indiana Coastal Chart Series - New Zealand Oceanographic Institute Collected Reprints - Institut Fur Meereskunde An Der Universitat Kiel Comptes Rendus De L'Academic Bulgare Des Sdences [= Doklady Bulgarska Akademija Na Naukite) Comunicadones Antropologicas Del Museo De Historia Natural De Monte¬ video Comunicacioncs Botanicas Del Museo De Historia Natural De Montevideo Comunicadones Paleontologicas Del Museo De Historia Natural De Monte¬ video Comunicadones Paleontologicas Del Museo Paleontologico Municipal (Colonia-Uruguay) Comunicaciones Zoologicas Del Museo De Historia Natural De Montevideo Contributions From The Institute Of Geology And Paleontology, Tohoku University Contributions In Sdence Contributions To Geology, University Of Wyoming Contributions To Geology, University Of Wyoming Spedal Paper Cunninghamia Directory Of The Geological Survey, Indiana Doklady Bulgarska Akademija Na Naukite Durban Museum Novitates Economic Botany Edinburgh journal Of Botany Entomologische Abhandlungen Environmental Department Research Bulletin - ALCOA, Australia Environmental Department Research Note - ALCOA, Australia Fauna Polski Fieldiana: Botany Fieldiana: Geology Fisheries Research Bulletin Fisheries Research Report Of The Fisheries Department Of Western Australia Folia Biologica Folia Musei Rerum Naturalium Bohemiae Ocddentalis: Botanica Folia Musei Rerum Naturalium Bohemiae Ocddentalis: Geologica Folia Musei Rerum Naturalium Bohemiae Occidentalis: Zoologica Fragmenta Balcanica Fragmenla Faunistica Geochronique Geodendas Revista Da Universidade De Aveiro Geologica Balcanica Geological Atlas 1:50,000 Series (Tasmania) Geological Survey Bulletin (Tasmania) Geological Survey Bulletin, Indiana Geological Survey Explanatory Report : Geological Atlas 1:50,000 Series (Tasmania) Geological Survey Of New South Wales; Memoir Geology Geologicheskaia Literatura Sssr [* Geological Literature USSR] Geologic En Mijnbouw Geologiia I Geofizika |= Geology And Geophysics) Geologisches Jahrbuch. Rcihe B: Regionalc Geologic Ausland Geology And Geophysics Glasra Contributions From The National Botanic Gardens, Glasnevin Historical Records Of Australian Sdence History Of Australian Sdence Newsletter Hombill Informative Circular Of The Ohio Biological Survey Investigational Report Of The Division Of Sea Fisheries, South Africa 93 Journal of the Royal Society of Western Australia, 77 (3), Septen\ber 1994 V South Wales .eneMcCunochFou.a^ti.« Island Chart Senes ^ g. Makedoruja. Izdanija PrirodoMUce Kazakhskoi Ssr: Seriia Geologicheskaia [ = Qazaq Izvestiia Akademii SsrGhylyml n^r Wissenschaften In Gottingen ,ahtbu<*DerA^"sOfTheRoy^ ^°“"''mf?he Adelaide Botanic Gardens ^ SSe Bombay Natural Histon' Socety o ^he Iowa Academy Of Science ’°“"’ 1 onhe Namibia Scientific Society ’““""I Royal Society Of Western Austraha ToralolTheTokyo university Of Fishenes Kapala Cruise Report l^o^dusNewsletterOfTheSWA/NamibiaBudClub Mauritius institute Bulletin Melhorame^ Geological Survey Of Western Australia Memoir Of ® Geologique De L'Universite De Louvain rrr:D:utcieteRoyaleDeBotaniqueOe Belgique Memoirs Of The Museum Of Victona Memoirs Of The Queensland Museum "‘^"’“^T^.Stat^Pro Fauna Et Flora Fennnica rrldedadl^OenciasNatura^^^ “elaJournalOfllieUniverti^OfGuam Mineral Geological Survey Of Western Australia tJneralogicheskii Sbomik [= Mineralogical Collechon] Mines And Energy Review, South Australia Miscelanea Zoologica (Cuba) Miscellanea Zo^ogm Zealand Oceanographic Institute Si”'™ PublloBm Mil.. N.W Z«.l.nrfW*V®-/^*- V ^ C-_ “(> " ■ -A’icf.T ■>;? iSBa' fc V3r''?A^-, • . • • ‘■3riJfit.->.-.V»i»\-.>'. .i-. “-*W?!*ft,^-»t>-:^A.''9^‘." i'njitpai.y-J*»v:- -" t»%<#>i;fiwi>fy»i»*^cb^?^ p- . • -■ ■• “-^ >■.'-.>?*^- rr'‘ ^>'^r-y y ■ AS!^* ■•- -uSi - ' > W» ^ AiteUfK '‘’•6%V ?*>♦ .iT^ to ^ ^ M i /*• .*5r« ©f-VT^S^?'^ 1 ^ i . - ■-■•, ■ r u_> •» »s* 4. - ’ '.. ' 11 TTllfl ■ -r - r^? m I L-.p .• r- -Si 'vj, -■ • • • * ^ ■ . r’i'./V .'•^ .-.a *jrc- « '.; '**■ '•■► ■' ) * * * • - <-■. ,-- ; -•. - ■'■■ '’■V ’ !tr« ■• i *•. ■■■ « .. f*v' ’ *■ • r ’ .4^^' • £• if* .. n ' < -*^- * Ei- ^ ■' r-i»i n ■■ ■*' '« * •■■■•,< v‘ ^ r ' ■ ■ -ii: T "*' •■ »*«» - '^V * *:«’ ^iv V.- ‘ ■.T. > f' • r/". .*/; ■*!9»V«' ' • ._•••-■'<» _ V? S!y ; •'^' X: '' 4*> ^ *^-5 *, Y -‘' ‘ v\3^^/^^.3%■•■'' ‘ '.,■ V > ? v: %;Y ^ c.'ryi.v-.^-, '-'' ' ■ '■** ■^■' »• ''■ *' -■ ■ ■ ' i< J t:r> - ->:-■ ‘■-^' r .. , -- . >—V-- • ■=• -- ,4-' ■ ' :■■ r ;■. - ' -a* -• : • -y> -j I - <• • ,^-- .-. ** , •r-y\ :•< ■' .i ■^'-' ^'* ‘'-j * ^. \* ■ •"-• E V.^ , .* _ ; rf 'j'.V -» '. ■' .♦ * ••. - ra 'v I .l.' • " - • V- •: L. ')• I ••%.ir, - . T j- r :*v --ijs-'" ■>--T/._'.■ ^:r' ../ :-, -v.,^' -'■■ "’ ' ...a'' Journal of the Royal Society of Western Australia Volume 77 Part 4 December 1994 esented by: The Royal Society of Western Australia The Ecological Society of Australia ^ ' •^^nsored by: MUSEUM OF VICTORIA PLANT DISEASES IN ECOSYSTEMS: threats and impacts in south-western Australia Locality map of main towns and features mentioned in the volume. t N 250 km V 0^' ' \ C CD CD CJ o c: •2 c: IB Eneabba Cervantes ■ \ ^ Moore River \ ■ Gingin \ B Bullsbrook -JtKarragullen Garden /s/andT ■tardup Southwestern Australia Point D' ^ntrecasteaux Karragullen - . Cardup ^ p Popanyinntng B Dwellingup ^ , I ^ ^ East ,. / Mount lunsborough aKojonup Barren^etou^ i_ Augusta STIRUNG RANGE ^-f^gerald Riv6T ♦ ^Bremer Bay 'in \ KamballuDB j^C^pe Riche Cheyne Beach ‘kvo f^ples Bay Jbany Yalgorup Cape Naturalists, iDunsborough Augusta Cape Leeuwm^\ Kamballupi West Cape Howe yraiiup^ j| alpol^^A'C 'w 'est^Albar Esperance :ape Cape Arid Le Grand 150 km Wills 94 Additional copies of this issue can be purchased for $30 + postage by contacting The Journals Manager, Royal Society of Western Australia, c/o W.A. Museum, Francis Street, Perth, WA 6000. ^ Front Cover artwork: honey possum {Tarsipes rostratus) on a nodding Banksia bloom {Banksia nutans) from south coast heathlands threatened by dieback; photograph provided by Steve Hopper, Kings Park and Botanic Gardens, Perth. Plant Diseases in Ecosystems: Threats and impacts in south-western Australia Journal of The Royal Society of Western Australia Volume 77 Part 4, December 1994 Proceedings of a Symposium of the Royal Society of Western Australia and The Ecological Society of Australia, held at Murdoch University, Perth, Western Australia, 16 April 1994 Edited by P C Withers^ W A Cowling^ & R T Wills^ *The Royal Society of Western Australia c/o W. A. Museum, Francis Street, Perth WA 6000 ^The Ecological Society of Australia PO Box 1564, Canberra ACT 2601 HONORARY EDITOR'S PREFACE This symposium, on Plant Diseases in Ecosystems: Threats and Impacts in South-Western Australia, was jointly organised by the Royal Society of Western Australia and the Ecological Society of Australia, and was sponsored by the Australian Nature Conservation Agency, Alcoa Ltd of Australia, and the Gordon Reid Foundation for Conservation. Its purpose was to bring together a multidisciplinary group of researchers studying various aspectsof the ecological, economic and social impact of plant diseases, including Phytophthora, Armillaria and canker, in the natural ecosystems of south-western Australia. This volume presents a summary of the symposium, session summaries, and papers reflecting the presentations of invited speakers in the sessions. These papers have been peer-reviewed, and the editors of this volume are grateful to the following reviewers for their efforts in ensuring that a high scientific standard has been maintained, and for their prompt cooperation; M Brown (Forestry Commission of Tasmania, Hobart) E Davison (CALM, Manjimup) B Dell (Curtin University, Perth) D M Cahill (Australian National University, Canberra) W A Cowling (W A Department of Agriculture, Perth) S Hopper (Kings Park and Botanic Gardens, Perth) S James (University of Western Australia, Perth) K Old (CSIRO, Canberra) J Pate (University of Western Australia, Perth) R Shivas (W A Department of Agriculture, Perth) ] A Simpson (State Forests of NSW, West Pennant Hills) K Sivasithamparam (University of Western Australia, Perth) M Sweetingham (WA Department of Agriculture, Perth) G Weste (University of Melbourne, Parkville) R T Wills (WA Herbarium, CALM, Perth) R Wooller (Murdoch University, Perth) Abstracts of the invited papers and contributed posters, as well as a list of symposium registrants, are presented elsewhere, in The Handbook of the Symposium on Plant Diseases in Ecosystems: Threats and Impacts in South-Western Australia, which is available from the Royal Society of Western Australia for $A10 including postage. P C Withers Honorary Editor Royal Society of Western Australia Foreword Phytoplitlwra citimwwwi is regarded as one of the most devastating pathogens in natural ecosystems yet recorded. Across southern Australia, it has had a major impact on a wide range of plant species from small shrubs to large eucalypt trees, the most notable of which are the majestic jarrah trees {Eucalyptus margimta) of sou th-westem Australia. Little is known about when or how the fungus established itself in south-western Australia, but few doubt the fact that it was introduced by European settlers. Considerable effort has been expended in researching the biology of P. ciuriamomi and the disease it causes, but are we any closer in answering key questions? Why is the disease spreading so rapidly today in reserves and parks on the south coast, such as the Stirling Ranges? What is the impact of radical ecological changes, such as clear-felling or prescribed burning of forests, on dieback and other diseases? What environmental conditions are conducive to dieback development? Many such questions will be explored in this Volume, sometimes challenging conventional views about ecosystem management for disease control. P. cinnamomt is not the only pathogen causing widespread destruction of natural ecosystems in south¬ western Australia. Several canker fungi have been reported causing severe dieback in coastal shrublands and Banksia woodlands, sometimes following the tracks left by a single vehicle several years earlier. What changes are occurring in these ecosystems that may be enhancing disease development? How can society resolve the conflict that occurs between those that want access to these ecosystems for economic or recreational reasons, and the need to control spread of disease? The Royal Society of Western Australia is in a unique position, as a multi-disciplinary scientific society, to bring together scientists from a wide range of disciplines to discuss issues of importance in a Symposium setting. However, plant diseases in south-western Australia havesuch a wide-reaching impact on ecosystems, including plant, microbe and animal life, that the Plant Diseases in Ecosystems Symposium would not have been possible without the full backing of the Ecological Society of Australia. The only major research done on the impact of plant disease on mammal populations is in eastern Australia. The Ecological Society of Australia brought key scientists to the Symposium from across Australia, and helped to involve the Australian Nature Conservation Agency (ANCA), which funds dieback research at the national level. ANCA fully supported the Symposium with the opening address from its Director, and in generous sponsorship of this publication. This document is a valuable summary of current and past research into plant diseases in ecosystems that is relevant to the whole of southern Australia. It is one of the few (if not the only) compilations on the subject, and is of course a major reason for holding the Symposium. The Symposium allowed a wide range of views to be aired, and not all views are supported by data. In many cases, data is simply not available. This has lead to conflict in the past among researchers, and such conflict can only be resolved by identifying the problems and setting about solving them. This Symposium has, we hope, allowed many participants to define the important issues, or to see old issues from new perspectives. One such issue is the effect of environment on development of dieback in jarrah. It is clearly very important to identify the key environmental parameters for disease development so that management strategies may be aimed at reducing disease impact. Private industry is keenly aware of its responsibilities in its management of native ecosystems, and industry representatives presented papers at the Symposium. We are grateful for support received from industry for the Symposium, and we thank Alcoa of Australia Ltd for their generous sponsorship of this publication. As organizers of the Symposium, we have attempted to provide a fair and balanced treatment of the subject by inviting a wide range of participants from all institutions that are actively involved in research into plant diseases in ecosystems, with an emphasis on south-western Australia. One resounding conclusion from the Symposium was that more research is needed into the effect of diseases on ecosystems, and that management decisions must be based on sound knowledge from research. We believe that a measure of success of the Symposium will be closer interaction among scientists and administrators in the future to achieve this result. This document should help both groups to plan and execute good research. We would like to thank all those who volunteered their services to help run the Symposium. We thank the Gordon Reid Foundation for financial assistance with running costs and to allow us to publish a Handbook of Abstracts on the day of the Symposium. Last, but not least, we thank our families who endured another bout of late nights and long days. W A Cowling President, 1993/1994 Royal Society of Western Australia and R T Wills Honorary Bulletin Editor Ecological Society of Australia 97 ‘ .-6. i. I Mnhl i i y^ ~~ii^"'~fffrinr^P^n^*'"*^***^ gi ri.--' i - r-** igfett. •f v-^,‘ -I',. ' ■ - •*-r:;*^J:'V * :■ .»v-_. . . . - - ' ‘-.-V.:**-.^,^ ?■>•!./ -4 ■ •, "^v.* ^ ^ ;‘V :vvr' v.. - - :* -rv^crr* ■‘ .; - - >y; ‘•- V-ivvi<>J/.%''.;*,v-ii f-* « 4 •? :'..*• ’^Vj(»**k iV- ' . '. ■-■-.^ ,V^S-- ^' ^ l: ;■■> ^ - * f- - % ^ •’*‘ . ■' "* -I -..^/..-‘e^'A » ' Jt- ^VU’' i *''5 r-: A S'Tx^^frfi^ * - ’ ' ■ " ‘ . r ’ • ’■ *''^.',- ' ’ ;-v 5fdrjU*^-‘ Journal of the Royal Society of Western Australia, 77: 99-100,1994 Plant Diseases in Ecosystems: threats and impacts in south-western Australia Symposium Summary S H James Department of Botany, The University of Western Australia, Nedlands WA 6009 This symposium reflected the levels of awareness, com¬ petence and action focussed upon the problems of plant diseases in Western Australian ecosystems. The problem of Phytophthora chinamorni, the major pathogen in our plant communities, has been likened to a massive invasion^. While it is unlikely that the invader will ever be routed, this symposium may help in formulating appropriate strategic and tactical responses to impede its progress and minimise its destructive effects. Peter Bridgewater* outlined the Commonwealth Govern¬ ment's legislative interest in and financial support of dieback research. He emphasised that disease in plant ecosystems was a problem affecting a large proportion of the Australian continent and that the diseases were permanently altering ecosystem.s, eliminating many keystone species and leaving behind a considerably less diverse community of resistant survivors. In the south-west land division of Western Aus¬ tralia some 1500 to 2000 of the approximately 9000 extant species were susceptible to P. citmamomi, the major pathogen causing dieback disease, but a number of additional patho¬ gens were involved. The reason why such a large proportion of the native flora was susceptible to dieback diseases was a question still not properly answered. The point was made that research and management practices should beware of concentrating upon the curing of symptoms rather than attacking the disease. This puts in sharp focus the possibility that plant diseases in ecosystems may well be a symptom of a more deeply seated ecological malaise. The basic malaise is obviously human involvement in ecosystem dynamics, both direct and indirect, and the minimisation and rectifica¬ tion of the damage resulting from plant diseases in natural ecosystems must clearly involve land management proce¬ dures. Frank Podger^ likens the advance of P. cinmmomi in our plant communities to an invasion and considers it essentially unstoppable so that it will ultimately infect and affect all the available population systems within the continent. He at¬ tributes the entry and the primary spread of the disease directly and unequivocally to the activity of man, but is optimistic about our ability to slow its progress. Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 Bryan Shearer^ outlined the taxonomic constitution, dis¬ tribution and impact of four classes of disease producing fungal pathogens in native plant communities in south¬ western Australia. As well as Phytophthora, of which P. cirmamomi is the most destructive, Armillaria leutobuhalina is a very important rootrot fungus which possibly has a wider host range than Phytophthora. There is a variety of stem canker fungi and rusts. Anuiltaria, the stem cankers and rusts are probably endemic fungi normally with restricted im¬ pacts, but are they becoming increasingly important as hith¬ erto undefined ecological balances become perturbed. Ex¬ panded taxonomic research and an inventory of disease incidence is required to document the importance of these pathogens, as well as that due to Phytophthora. Elaine Davison^ provided evidence strongly suggesting that dieback symptoms and death may be induced in jarrah as a consequence of waterlogging, even without P. cintiatriomi infection. However, the waterlogged condi¬ tions may promote pathogen activity, and secondarily result in increased levels of infection. This is indeed a situation where symptoms and disease may be easily confused. Ray Wills and Greg Keighery-^ reported that 38% of 460 plant taxa examined are susceptible to P. cinnamomi, while 59% of 436 taxa were sensitive to canker fungi. Some 86% of Proteaceae species are susceptible to these pathogens, and after infection the percentage cover of dominantProteaceous species may be reduced by 95%. The species removed are often keystone species, providing food, cover and nesting resources for associated animals^ and altering the light re¬ gimes necessary to support erstwhile associated plants. Removal of the dominant species by P. cinnamomi and canker disease leaves a simpler, less diverse community dominated by rushes and sedges. While these remnants are resistance to P. cinnamomi and canker diseases, they are susceptible to a suite of smuts and rootrots^ and the subsequent perform¬ ance of these residual populations is conjectural. While the effect of plant diseases in the loss of diversity and structure of complex plant communities is quite strik¬ ing, it may not be recognised as such by an untrained observer. Nick Malajczuk and Martin Pearce® emphasised that the changes ind uced in the soil microflora and microfauna may be equally or even more dramatic. He pointed out that the taxonomic complexity of the soil microorganisms is orders of magnitude greater than that of the above ground 99 Journal of the Royal Society of Western Australia, 77 (4), December 1994 vascular plants, and that some 50% of the energy captured by plant photosynthesis finds its way into maintaining the rhizosphere. We are only beginning to chart the dynamics of rhizosphere ecology. Stuart Crombie outlined the difficulties in assessing the effects of plant disease on timber production in local forests'*. The Western Australian forests currently produce some 2 x lO'^m^timber per annum, from which thestatereceivesabout $100 million. In addition, the forests have value in water catchment, tourism and education. The effects of timber harvesting on the spread of disease, however, is pronounced, with the major outbreaks of P. cinnatuomi being closely associated with logging activities, especially the establish¬ ment of road netu'orks throughout the forests*. Much of the expense associated with timber harvesting is presently con¬ cerned with the implementation of disease minimisation. Ian Colquhoun and Anthony Petersen outlined thedieback control and environmental restoration measures adopted by Alcoa of Australia and RGC Mineral Sands Ltd". Dieback control is a source of additional costs in both industries, but probably represent less than 0.5% of the value of the indus¬ tries per year. Industries dependent upon the products of natural plant communities and the threats posed by plant diseases were outlined by Chris Robinson and Ray Wills'”. Minor indus¬ tries include honey production {$2 million per annum) and cut flowers ($18 million perannum). Both these industries are severely affected by plant diseases, but both may ultimately become based on cultivated native plant farms. The tourism industry in Western Australia is a major source of wealth, ($3 billion perannum), but only an indeterminate fraction of this industry canbe attributed to the attraction of our native plant communities and it is not clear that any changes wrought by plant diseases have diminished that attraction. Kelly Gillen and Anna Napier'^ reviewed the manage¬ ment of access for recreational activities in the control of dieback diseases in native plant communities. Public access is clearly critical in the spread of dieback. The severe impact of P. cinnamomi on almost 75% of the Stirling Range National Park demonstrates the effect of facilitating recreational use in spreading disease. Management of access requires con¬ siderable resources in planning and providing appropriate surfaces and drainage patterns, the implementation of sea-, sonal or permanent closure, and the provision of washdown facilities. Giles Hardy et al. reviewed a wide variety of techniques now available for the detection and control of plant patho¬ gens in native plant communities^^ and Jen McComb et al. showed that the breeding and cloning of native plants resist¬ ant to some pathogens is possible’''. However, the costs of implementing these "solutions", the large number of imper¬ illed species and the extensive area of infection necessitates a system of prioritisation. Some targeted species and some targeted communities may be protectable at substantial costs. Greg Keighery et al. showed that almost every human activity in and near native plant communities is likely to create an ecological imbalance and to promote the spread of disease'^. Ecosystem changes wrought by the diseases especially in association with land clearing and a new set of grazers, predators, pollinators and weed competitors is permanent. Joanna Young suggested that disease spread and the threat to our flora and its associated fauna may be reduced by land management practices which take basic ecological principles into account'^ It is, however, arguable that many of the pertinent ecological principles are not yet known or are inimical to our society's requirements. References [from Handbook of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia 1994. Eds RT Wills & W A Cowling. Royal Society of Western Australia and Ecological Society of Australia, Perth.] ‘ Opening Address: P Bridgewater ^ History of Research: F D Podger ^ Major plant pathogens in ecosystems: B L Shearer Role of environment in dieback of jarrah: E Davison ^ Impact of plant disease on plant ecology: R T Wills & G J Keighery ^ Smut and root rots on native rushes and sedges- K A Websdane, 1 M Sieler, K Sivasithamparam & K W Dixon ^ Impact of plant disease on animal ecology: B A Wilson G Newell, W S Laidlaw & G Friend ^ Impact of plant disea.se on microbial ecology: N Malaczuk & M H Pearce ^ Disease and forest production in WA: D S Crombie Threats of plant disease to flora-based industries- CJ Robinson &RT Wills " The impact of plant disease on mining: I J Colquhoun & A E Petersen Management of access: K Gillen & A Napier ''' Control options of plant pathogens in native plant com munities in south-western Australia: G E St J Hardy, P A O'Brien & B L Shearer Future ecosystems - use of genetic resistance; J A McComb, M Stukely & 1 Bennett Future ecosystems - ecological balance: G J Keighery, N Gibson & D J Coates Future ecosystems - effects of plant disease on Society: J T Young 100 Journal of the Royal Society of Western Australia, 77:101, 1994 Session 1: Biology KM Old CSIRO Division of Forestry, PO Box 4008, Queen Victoria Terrace, Canberra ACT 2600 The first paper scheduled for this session was a personal account of the impact of European settlement on native vegetation communities as seen from the perspective of a plant pathologist, and focussing on Phytophthora-induced dieback. The paper was presented by Dr Frank Podger and was based on his career-long involvement in the study of diseases of forests and native plant communities, including the first association of P. dtwamomi with jarrah dieback. He pointed out that Phytophthora-induced diseases of native vegetation is a national problem, with regional circum¬ stances of climate, soil and vegetation differences demand¬ ing differing approaches to management. Accidental intro¬ duction of P. cuumwomi into native vegetation has caused major changes in plant communities and despite the en¬ hanced level of awareness of the consequences of disease, there are major difficulties in achieving containment. He pointed out the dangers of over-emphasis on reductionist research which may provide information that cannot be translated to practical solutions for reducing the spread and impact of Phytophthora spp. and advocated a holistic ap¬ proach to disease management. Although it was appropriate for a paper on Phytophthora to lead this session, the second speaker. Dr Bryan Shearer, gave a much broader picture of the range of pathogens that currently affect the health of native plant communities in south western Australia. He pointed out the need for sys¬ tematic disease surveys of such communities, and the need to maintain adequate mycological expertise within govern¬ ment agencies and the universities. Without such skills, the recognition and containment or control of epidemics will not even be attempted. A database, which is being compiled for a wide range of families and their diseases, is a first essential step in an assessment of the significance of the main groups of diseases including root rots, stem cankers and leaf dis¬ eases. The final paper in the session was a discussion by Dr Elaine Davison of the role of environment in the dieback of jarrah, especially the effects of waterlogging on the physiol¬ ogy and anatomy of tree roots, and root infection by P. cin?iamotm. Although the epidemic disease, which is so dam¬ aging on the understorey and ground cover in the jarrah forest, Banksia woodland and healhlands, appears to have a somewhat simple etiology, the impact of the fungus on large jarrah trees is more ciy^ptic. Of the several hypotheses to explain local and rapid death of large trees in the jarrah forest over the last two decades, none have proved to apply un¬ equivocally to all drcumstances. This partly reflects the inherent difficulty of diagnosis of dieback diseases in large trees. However, controlled glasshouse studies coupled with field investigation suggest that periodic waterlogging may contribute to death by direct effects on root conductivity (through induction of tyloses) and predisposition of roots to infection by P. cwtiatuotui. The three papers emphasised the need for a multidisciplinary approach to the management of dieback diseases of native vegetation. Systematic gathering of infor¬ mation and rigorous study of the etiology of diseases, host pathogen interactions and of environmental effects are needed for the development of effective management pre- sciiptions. Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 101 B « « k,'-m T 'Jj.' It .A-^V'W'W'T’ ■{f‘ *»' *1 it, ^ •;-^ ,_ PT» > - •* • '■ C^- : ;,^..3 ■-. . , . .■ P^?>\ , - /f 8 « • . «- . »‘^-**J«.: ISSS?*i''^'-^'’-’* '?53Pf«''T^ ‘ *’ ' “ < -■- jar- ; ,. ■ “- ’ ■ V ■• S- ^ . »,» ■■-' ■’.'i-'iSt •V 1 .;,. ‘SraL.,,v ^ , ft-.'. ►i^ -r >* ■_? wh ►,^ ..- ■■ . «#*' -’'''■*^ t » f-v‘ ^ " ' '” i,V. if ■' '■ 'fe»' ' ^ ' ' 5.j»L.^££««' ■' ■ '-"' .'. - V 7 .V ^ V.'■^■' r Ji^» o’ *' iii '•' S^fa^,T?rr*. .v* ^ *s- V' j r.. f m^'■ • ’ _ ' 3^11 • '- ^ [' ■':' ■’■J*?^,* '9*^* ■■ ' A* ?rt* ’*■>- - ’t'. srjj;-n V v^: l‘4i^ /r*'- ^ * . ,t' - t y^.,x. r- • I ^4' t . In i 4 ' r* ■ V wcstern Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 erals Sands were discussed and despite the very diverse range of site conditions (e.g. Darling Ranges and Coastal Plain) there was common recognition of the impact of the species of Phytophtliora on the operations. In these latter presentations, it was stressed that the impact was more than just financial, although this aspect was still evident from the presentations. Both companies recognise the potential threat of the plant diseases affecting the adjacent natural vegeta¬ tion, the ability of establishing key plant species in rehabili¬ tation sites, and achieving high species richness in these sites. The potential impact of all plant diseases on these ecosys¬ tems is very difficult to determine as there is lack of taxo¬ nomic and basic research on the range of pathogens present in the environment. In fact, the effort of some of the commer¬ cial or licensed industries such as forestry and mining have assisted in defining and determining the impacts of the plant diseases in the south-western part of Australia. On the other hand, as Gillen and Napier discussed in their paper, the less controlled or un-licensed activities such as recreation and tourism in some ways pose greater threats to the ecosystems. The wider responsibilities and costs then return to the man¬ agement authorities, the government departments and ulti¬ mately the Australian community to determine equable rules to minimize the spread and intensification of the plant diseases. The latter is achieved primarily through manage¬ ment plans and operational activities (such as closing roads and quarantine measures). The costs associated with these less controlled activities is very difficult to determine as many cannot be determined without a substantial improve¬ ment in the understanding of the plant diseases, the ecosys¬ tems, the different site conditions and the diversity of activi¬ ties concerned. ^ *H^ - .^4j: % .- -« r. t\ ^ ■:' %.' ‘ ‘.i-Xr ' - -' . -L -•...•<:^* -;., *- J'r •'~-t. 1 ' V-S’.*^’ 'sfjf_ M * •'■f r A' F * •> • #J|f V'l I ■|^r ■' 2*- V^:" < ,V, --5^, <» ■• ■’•i* N .*■ ?i- ' 't' ■-'•?•* ' “' 'i « .4 ^ 5«!fS(6;->«r «'■••.■::»*, . w'v '■ > J-= -r - - r»if .*.£ ^ . •, ^r., •; .t S'* '•>•*/ ^ V1 'V ■‘' - ■ ti:< •^■.y.s < *4;4 J >' > ■'-t ' r ’.' ■ fc. ^ ' * •• • , » ' ' •''* is ' ■ •*' *•« isvi . • -.*3*4^ j£. >- 1: ^^- ‘ ^t>dj C y*-ti^vc Vl^ 1>3' ii 4^, \ : •i *■' ^ < ,.x ?■'"■:>-’•■ f ^ -.■ ^'ii». ;. 4» ^ .'. : - _i!sb»6i^%a.|64tes^^ i'-jf ■• ‘^ “s«i*4«^’^-“> ''“ -4t‘ " '■■“■ ■ ft ^_^,'V:- A'?,.;] - • *■ — r * '' ' .-^ j4i':,i4 ^^•p*e^3fSS« .- .'■'. /’sf ry'hk'* — '■ • •t': r '■• ■: K%‘- ■mf*2— ‘ j ^ ‘’ifc? ■ r*-' » t r ,*■ jr* ^ I .. ! -V -f. ■»<. ^ Vi. A S’ »S ‘ I'V i '*. ' -■ V i ■•Vsii ■; /, f. ivx\h* . *', ■ .‘"vr (-'siivv'i ■ “ ■ ifi .c,-'tA--\li fr-f. j - - i •* ♦v • '>’ *• ' ^ -- ’. - 'r ' ' :•_ > V ' '^' * -V'^ A r--* et vV , TB^^,- ■'■■•’;• , ' “» ■»^' esi V* y#-^ V**,' •". r * a.ii ,1 r' «, ■ . ;i ■■*'» ■^'F-U' ■»'---''yji' - -■■■-%' '■^ »*)■ --|> , * *. f >' •'* • ^:'V'■■ - »* VI ;#.r; fit'. ■* rr^/. •i}r; « "iAs '<* ■ ■- iSpfe,- d ' sr:.;.- ■- i?. ■ ■’V,-;:-':^. - Nit ■ > ^ i %n V.J •’ ■ \ . LUff* Journal of the Royal Society of Western Australia, 77:107^ 1994 Session 4: The Future S H James Department of Botany, University of Western Australia, Nedlands WA 6009 This session juxtaposed optimism with a sombre realism, if not pessimism. The optimism derives from the promise and achievements of modem technological capabilities and focussed research. The darker view is generated by a global, but rational, view of the magnitude and permeance of the changes in plant communities wrought by dieback diseases and the robustness of the social and economic imperatives which are promoting the spread of those diseases. Giles Hardy et al presented a broad overview of the current and possible options for the control of plant patho¬ gens in native plant communities. They emphasised the great difference between the vastly diverse natural ecosys¬ tems and the monocultures of commercial plant production which provide essentially all our experience in plant disease control, the importance of ecological imbalance in the gen¬ eration of disease, and the fact that almost every human activity in our native ecosystems promotes imbalance. While phosphorous acid treatment is dramatically effective in mo¬ bilising defence responses in certain species otherwise sus¬ ceptible to P. cinmmomi, the general levels of phytotoxicity of this and other chemical controlling measures, and their effects on micro-organisms antagonistic to the pathogen, or otherwise beneficial, are not known. Untargeted chemical application, as per aerial spraying or soil drenching, is poten¬ tially capable of inducing deleterious ecosystemic imbal¬ ance. Infecting host plants with avirulent or hypovirulent pathogens may increase their resistance to virulent patho¬ gens without any significant perturbation of the ecosystem. It should be possible to construct appropriate hypo-virulent pathogens using ds-RNA elements. Breeding, or selecting, resistant strains of host plants is a demonstrated possibility and thereislittledoubtthatgenetically-engineered, resistant hosts will be constructed in the near future. However, the scope for using targeted chemical and biological protectants, and selected or engineered stock is limited because of the cost, the area of bushland infected, and the number of susceptible species involved. Hardy et al also described the sophisticated techniques now available for detecting and accurately identifying pathogens including immunological and DNA based procedures. The paper also reviewed the management procedures which are currently practised, in¬ cluding hazard rating and risk assessment, hygiene practice and quarantine, and outlined the variable expression of P. cinnamoml and other pathogens, with varying ecological conditions. Jen McComb et al. reported a series of studies which show convincingly that resistance of jarrah to P. cinnamomi is genetically determined, that their selected resistant stocks are indeed resistant, and that the selected susceptible stocks are indeed susceptible to P. cinnatnotni when grown in reha¬ bilitated bauxite mine pits. They posed and provided an¬ swers to several relevant questions: re-establishment of jarrah in graveyard sites seems possible, but has not yet been accomplished; the technique of selection for resistance and micro propagation could be adapted to any species, but the cost and work involved makes it an option only for selected priority species; the resistant and susceptible stocks avail¬ able may provide the pedigreed stocks necessary for the detection of molecular markers for P. cinmmomi resistance in jarrah; appropriate techniques for introducing genetically engineered resistance into native species in natural ecosys¬ tems are presently not in hand and might be subject to public resistance on ethical grounds. Greg Keighery et al pointed out that the rates of change in Western Australian ecosystems had been dramatically in¬ creased since European settlement. Plant disease is but one, albeit extremely important, determinant of ecosystem change, but when associated with the synergism of broadacre land clearing, weeds and arrays of animals utilizing those plants, the change becomes permanent destruction. Rare species which are disease-susceptible and associated with remnant vegetation are fatally threatened. Their remnant ecosystems are often targeted for activities which promote the spread of disease, such as waste disposal and power-line construction. There is likely to be no source for post disease recolonization of remnants in broadacre cleared agricultural areas. A continuing remnantization of native plant communities will promote increasing levels of extinction. Disease within larger continuous areas of native vegetation will promote extinction of susceptible species and domination by fewer resistant species. Stemming this tide of extinction is perhaps the greatest challenge to land management. Joanna Young emphasised the rapidity and magnitude of change in our plant communities associated with repeated disturbance and human exploitation. Disease may strongly exacerbate these changes, and our experience of disease control in agricultural plant production is not appropriate for disease control in complex natural ecosystems. Young is of the opinion that management must be more strongly aligned with the ecological principles which underly the relative stability of the natural ecosystem rather than being driven by human economic considerations. The natural plant communities and ecosystems of the south-west of Western Australia are diminishing, and will be less available to fewer people in future generations. Our loss can be limited only be management procedures specifically targeted at preservation and protection from disease. Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 107 r-^ :f" > "■ . '■ igi^ .•A*» < • - ‘w-'V*" ». ■''* ••> 5*iVv»;'’r'*' */***:»- nV'- .- if : .=■-. '.- • •■ ^ -‘rr .f f N*.*,*4i ■»-'4 V i? ,'^"'.iw *'.: ^ *.'7*'- £“'4V' .r. jir: ■ 5 •!:» •■ ■ t , r - ^.' f ^ • w'i - ■'" >’ -r^ ■ 4- ■ ~ ‘ ' / ?;*AAWC *“ ^V*'^ ■/ % -5 ■I*, ft -- rrtf, Ht *:. O'r.t ! r* t » ■' ‘' *t-:'-^f^ J 'f-^lf^*.*'*^" »■'. V-i !-!>> f'V*'';*^r‘ it»-v . • ■ ■ . - .P* *V . .'iT,. irr>.*'U4 ^'Y«r Wri'j ,. ^ ^%'aJ -^'il 1*:^ :-■;: tv 7 - ■■ i*.»t^W r^:'*;; tAvf'.v-i V'i. '.'n^^t/'- «* v.*:- -f ^3^52* "t'-; '.(<*M.'^ ■*• * i^^t:i>Kr< . ^ Vtl , - %^ji r'^l • -; *-5^.« •.?; ' ^tUti" 4' JSr*'/ ' IN-. ’* . ■ - . • ■* ■ w; ;< ^.. 'i ■ ' '.f'-'i ■ * '•- fA>*M 1^ ‘ ^-» 'N^, “ ■Sf^'-A. ' - x^j''.-r»-,,Y-'.<^'- * «- i. sWlWB' *“' *v» -- -*- ■-»“■ ■'‘ fljif|^':»^'t»-»i.^_-,-‘,VV -■«, - " ja^ V V , ?<•-•■■ ^.« iK#. ft 1' -- v^ *,. "'*i*>5'ft ft"; t1*.i ' ''-* •^iS' ?'-ft 'V ■ • ' /^ ti^n- ,4**-'. ■■.••■ .» 1^ fco- .. ■ _• #•-. -rt&i . hltiV^'ya .-.■ ''■« \‘ - j| i iii i»#s^-ii l li t » i'>^i i a >.- t . » ir ‘ ^ - • -0^>Cnp^,-VJ!>» ^ i •/VlE^-V v-^ -ir; crtr ■(-»>'^ y!^'f>:< .. . .> v: v^^vtfiiit*5j^Sr** 4^ .y .53^ ■ *' Atar'iK^ ,>..-^!k..*i.’■J'J.-'t * *.' ^ \ P* L-v \5-r-\^^ ^:.'V . ■t <•-« ^ A ■:.H- *t ■ i?a '•^"<44^ < 1 .• ‘ %•>%!#•,V <4irnn iari*- : r* ».■. Journal of the Royal Society of Western Australia, 77:109-111,1994 Ecosystem pathogens : A view from the centre (east) P B Bridgewater & B Edgar Australian Nature Conservation Agency, GPO Box 636, Canberra ACT 2601 Introduction In 1846 to 1851 over a million Irish died, and a million more emigrated, starting a trend which kept the population of Ireland small and the semi-natural ecosystems of Ireland more or less intact, or as intact as a landscape depleted by the British Navy and the Industrial Revolution could be. These deaths and population movements were the result of the Irish Potato Famine, caused by the impact of Phytophthora infestans on the Irish potato crop, the introduced potato having become the staple of Irish peasant diet. This syn¬ drome, of an introduced food staple being subject to massive infestation and destruction, is not unknown and may well increase in the next decades as global diversity decreases. What is interesting, in the context of today's discussions, is that this fungal attack, causing dysfunction to the main Irish agroecosystem, actually saved the Irish landscape, and made it what it is today—one of the best preserved of the European landscapes. The other feature of interest is the candidate of destruction, which is a species of one of the key genera of concern to remaining species-rich Australian eco¬ systems— Phytophthora. Of course, much has now been con¬ fused by the misleading use of the term "dieback", being a general phenomenon of arborescent shrubs and trees wher¬ ever there are ecosystem stresses. Our focus is on one major, insidious, and relatively unrecognised threat, dieback and die-off caused by fungal attack. Although you will traverse a broader coverage during your discussions today, I want to particularly focus on Phytophthora cinnamomi, because it is listed as one of five key threatening processes under the 1992 Federal Endangered Species Act. Major problem syndromes and areas In the twentieth century, Phytophthora has had a major affect on industries such as nurseries, horticulture, cut flowers, field crops and pastures. In Australia, Phytophthora was first recognised as a major problem for flora and fauna with the onset of forest dieback in the jarrah forests of Western Australia and in the mixed eucalypt forests of Eastern Australia. It has only been more recently, however, that Phytophthora has been recognised as having a major impact on Australian biodiversity. In particular P. cinnamomi has been identified as causing major ecosystem disruption in the species rich communities of southern Western Australia. In addition cool temperate rainforests, heathlands and the understorey of dry sclerophyll woodlands are affected in Tasmania, Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 Victoria and New South Wales. In Northern Australia, Phytophthora is reported from tropical rainforest, low sclerophyll shrub woodlands, mangroves and heathland. The scale of infestation in South Australia may be only now becoming apparent, with areas such as Kangaroo Island reporting infections. This insidious pathogen is altering the ecosystems of Australia on a mammoth scale, sometimes in a subtle way, but also dramatically as occurs here in Western Australia. The floristic composition of vegetation communities on a landscape scale are being permanently altered and this is having subsequent impacts on fauna. For example, Banksia coccinea, a species highly susceptible to infection by Phytophthora, may be a keystone species because of the reliance of birds and small marsupials on its flowers for food. Relationship with areas of high biodiversity The effects of Phytophthora are most dramatic in areas of high species diversity. In the south-west of Western Aus¬ tralia, for example, it is estimated that 1500 to 2000 species of the estimated 9000 species of vascular plants may be suscep¬ tible to infection. Many of these species are highly endemic and have been taken to the edge of extinction. The pathogen is particularly devastating on species from the families of Proteaceae, Papilionaceae, Mimosaceae and Epacridaceae. These families comprise the bulk of species from which the wildflower export and tourism industry relies. For the Proteaceae family, 85% of species found in the Stirling Range National Park have been rated as susceptible to Phytophthora (Wills 1993). Is it possible that areas of high species diversity are more susceptible to Phytophthora, or is it simply that it is in these areas that the problem is more apparent because they dem¬ onstrate the most stark examples of alterations in commu¬ nity structure? Whatever the answer, the ability of the pathogen to radically reduce the complexity of a community to a relatively fewer number of tolerant species is having a profound affect on the biodiversity of significant regions of Australia. Many of the species threatened by fungal syndromes in this State, and indeed in the Eastern States, are either poorly known or not yet subject to taxonomic description. It is axiomatic that there are many species becoming greatly diminished, or extinct, without ever being recognised. Now that clearing of land has become recognised as a major problem for biodiversity conservation, it is ironic that the remaining uncleared areas are being threatened by these syndromes. But is it really serious? Is there not so much redundancy (Walker 1992) in these species-rich systems that 109 Journal of the Royal Society of Western Australia, 77 (4), December 1994 the loss of a few, albeit colourful, species will not really be of concern? I pose this as the kind of rhetorical question that will be asked of us by senior bureaucrats and politicians. Considerations for management at a national level There are important considerations for a national ap¬ proach to this pandemic pathogen. These include: • continuing to develop strategies to control the problem in the short to medium term through effective hygiene, the development of control methods such as the use of phosphonate, and improved prediction and mapping of occurrence of incidence; • investigating longer term control options that may be¬ come available; • improving communication between agencies and institu¬ tions; • conducting research so that advancements and strategies can be rapidly applied; • developing a national approach to identifying gaps in knowledge and targeting research towards filling those gaps; • co-ordinating research that no single agency has the resources or expertise to conduct on its own; • ensuring that limited resources are well targeted and that the effort is complimentary rather than involving dupli¬ cation; • investigating novel approaches to control and manage¬ ment of Phytophthora and liaising with international ex¬ perts and agencies. Linkage of the problem The problem of Phytophthora has common threads of economic loss and ecosystem breakdown across Australia. However, the relative impact in the various regions of Aus¬ tralia differs. For example, where a complex understorey exists, with associated dependant fauna, ecosystem damage comes from loss of floristic and structural richness leading in turn to loss of faunistic richness. Where death of the overstorey is the most apparent symptom, there can be a resultant change in understorey structure from reduced competition. Climate, geology, soils and the resultant veg¬ etation communities are important factors in determining the relative impact of Phytophthora from the north of Aus¬ tralia to Tasmania and the south-west of Western Australia. The number of researchers involved with Phytophthora in natural and agroecosystems around Australia is about 150, which includes research staff, technicians and students. Approximately 75% of the researchers work in Western Australia, Queensland and Victoria. Many horticultural crops, and even subterranean clover, are threatened by Phytophthora species. But six species, especially P. cinnawomi and P. megasperma, are of particular threat to the south coast of Western Australia. In economic terms (wildflowers), the problem could be a loss of upwards of $50 million. But, it is the incalculable damage to the species rich ecosystems which is of prime concern. How do you measure the change from one of the worlds most species-rich shrublands to a bland covering of grasses, sedges and restioids? In National Parks and State Forests, quarantine of dis¬ eased forest is the main strategy for control in both Western Australia and Victoria. Vehicle check-points and washdown, as well as logging practices which minimise disease spread are other strategies employed. There appears, however, no management practice which is totally successful at eradicat¬ ing disease originating from infestation by Phytophthora. Given that, research priorities should attempt to focus on Integrated Pest Management (IPM), and yet there seem to be demands from the community for biocontrol or resistance for breeding as the leading research efforts (Cahill 1993). Recent work in Western Australia particularly focuses on the use of phosphorous acid to protect individual trees, and small areas of ecosystems of high conservation value. Other groups have suggested that bioconlrol may be feasible, or that hypovirulence could be induced into some populations Despite these potential developments, the clear message from the current dire situation is that there is no one cure, and that IPM is the only sensible way to go. Role of the Australian Government The role of the Australian Government is to promote a national approach and facilitate research and management action. This has been achieved to dale through funding and promoting communication between the various parties re¬ sponsible for research and management. The Australian Government also has the role of discharg¬ ing its legislative responsibilities under the Emiangered Spe¬ cies Protection Act 1992 (ESP Act). This new and important legislation tackles endangered species conservation in a number of innovative ways. For example, it provides for the recognition and protection of ecological communities. It also tackles threatening processes, which operate across a range of habitats and affect many species, whether threatened or not. The Act recognises Phytophthora as a key threatening process and establishes the premise that a nationally coordi¬ nated plan would be of major benefit in tackling the problem. Such a plan must be in place by 1999. The Commonwealth has significant funding that can be brought to bear on Phytophthora, such as from the Endan¬ gered Species Program. Important work being funded at present includes the development of disease control by the application of phosphonate, the development of an inexpen¬ sive and simple diagnostic test for the presence of Phytophthora, the use of GIS for mapping and predicting distribution and severity of disease, and the identification of susceptible taxa and long-term storage of germplasm. The Commonwealth can also be of assistance by provid¬ ing research expertise of a high quality to provide direction and support to agencies directly responsible for tackling Phytophthora. There are now a number of successful exam¬ ples of Co-operative Research Centres (CRC) in Australia, where a collaborative and co-operative approach has been employed to tackle important issues such as Phytophthora. Given that Integrated Pest Management appears the only 1 Journal of the Royal Society of Western Australia, 77 (4), December 1994 sensible way to tackle the problem, perhaps development of a CRC is a practical and effective way to proceed into an uncertain future. Symposia like this one today are a start on the road to attaining greater certainty. One final point, which may seem heretical, needs to be posed. Bridgewater &c Ivanovici (1992) discuss the phenom¬ enon of "constraint syndromes" on the natural ecosystem, using as exemplars Drupella and Acanthaster in the marine environment. Are the various fungal syndromes to be dis¬ cussed today in a similar category? If so, simply solving the problems caused by the fungal species effecting the con¬ straint may not be the whole answer. It would be rather like curing the symptom, rather than the disease. To cure the disease, we suspect a greater understanding is needed of human influences at work in the landscape at large, for broadly-based human-induced change may just be the pri¬ mary driving forces for the more obvious features such as dieback / dieoff. Just as IPM may be the answer to solving the fungal problem, we may need to combine it with Integrated Landscape Management, to give full effect to our palliative measures. References Bridgewater P B & Ivanovici A 1992 Achieving a representative system of marine and estuarine protected areas for Australia. In: Marine and estua¬ rine protected areas in Australia (ed A Ivanovici) ACIUCN, Canberra, 23- 29. Cahill D1993 Review ofP/iyfop/if/iora diseases in Australia. RIRDC Research Paper Series. 93/4,78pp. Walker B1992 Biodiversity and ecological redundancy. Conservation Biology 6:18-23. Wills R1993 The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18:145-159. ^ ^ t ?"' s * V' • *»*»r li*‘" - V •v.,« .---r-', - • "i . fV^'*-"- ^ ■ ■■■'''■Wt:^''' /.Mf. ■,-.1 :ir'v ^ . ■ ■ V ' •TO ■■ -■■ - ' ■“ '<-’ r » - . ' « • - V*>. !^: 4 |ffl|!|.V.>r;-f,^“ _'■' ' '* |§j|i|!i ^ ^'' * ' ,,j.r^ V ' ^’' » * .^ 3 jf *■» ■ * . ^ f € ■ 'V ,- -, . - A*S-'' 9 i 31 , .. * r ‘ . » Vi- .. < 'M-* ' • ■' . I •. •• . . ■'- - ■'< - ■«■ _'■■• - kZ’wl^'lr'. „, ' • * j *1 ^ - ■>' ‘-V ■ ■' *'''»^ • *■^v -y‘-f-.2?-jj..'-«v^ ' « . - *' . . s ' ^*‘A V '7 > .fcrf' fV- ■:“** y- rfH ■■’ *3 « r * •» r - _■. iir'V*- ■■ .i .* ,V <'. -J* r4.*. Journal of the Royal Society of Western Australia, 77: 113-122,1994 The major plant pathogens occurring in native ecosystems of south-western Australia B L Shearer Science and Information Division, Department of Conservation and Land Management, 50 Hayman Road, Como WA 6152 Abstract Objective assessment of the relative importance of pathogens on conservation and production values in native plant communities of south-western Australia is impeded by the lack of systematic disease surveys. The occurrence of diseases and pathogens on Western Australian native plants was compiled from published information, other reports and personal databases. Pathogens were databased according to name, host name and family, disease group and Botanical Province, giving a total of 936 entries that did not include reports of pathogens on hosts in nurseries. Ninety-one per cent of the pathogen reports were from the South-West Botanical Province and 2% from each of the Eremaean and Northern Botanical Provinces. Bacterial diseases, galls, downy and black mildews, ergot and leaf moulds were infrequently reported on native plants. Pathogens were infrequently reported on species within the families: Aizoaceae, Amaranthaceae, Amaryllidaceae, Annonaceae, Anthericaceae, Apocynaceae, Arecaceae, Asphodelaceae, Cupressaceae, Cyperaceae, Dennstaedtiaceae, Geraniaceae, Juncaceae, Lamiaceae, Linaceae, Loganiaceae, Olacaceae, Onagraceae, Phormiaceae, Pittosporaceae, Podocarpaceae, Polygonaceae, Portulacaceae, Rubiaceae, Solanaceae, Stylidiaceae, Tremandraceae, Verbenaceae and Zamiaceae. Pylhiacious root rots, rusts, Armillaria root rots, stem cankers, and leaf spots and blights were frequently reported on native plants. Families most affected by disease were: Proteaceae, Myrtaceae, Mimosaceae, Papilionaceae, Haemodoraceae, Goodeniaceae, Epacridaceae, Poaceae and Chenopodiaceae. Families mostly affected by rusts were least affected by the root rots, stem cankers and leaf spots and blights. The biology, distribution and disease expression of Phytophthora cinnamomi, rust fungi, Arwillnria lufeobiibalina and Cryptodiaporthe canker of Proteaceae in native plant communities are described. Conservation of plant taxa requires a much better inventory, than is available at present, of the incidence and status of the various plant pathogens that occur in native communities of south¬ western Australia. Prediction of the likely long-term effects of pathogens on native plant communities requires a much better understanding of their life cycles and biology in the south-western Australian environment. Introduction The lack of systematic disease surveys in native plant communities of south-western Australia impedes objective assessment of the relative importance of the impacts of pathogens on conservation and production values (Shearer & Hill 1989; Shearer 1992a). There have been no coordinated regional surveys of disease occurrence in native communi¬ ties of south-western Australia, similar to the regular assess¬ ment of disease and pest conditions in Canadian forests (Forestry Canada 1993). This is somewhat surprising consid¬ ering the exceptional species richness and high degree of endemicity of the flora of south-western Australia. At least 7000 species of described native vascular plants occur within the state (Green 1985), of which over 3000 are endemic to the area (Keighery 1992). Knowledge of the diseases of native plant taxa is important for maintenance of long-term conser¬ vation and production values, especially in the case of rare 9nd endangered taxa. Symposium on Plant Diseases in Ecosystems: l^hreats and impacts in south-western Australia. held on April 16, 1994, at Murdoch University, by the Royal Society of 'Western Australia and the Ecological Society of Australia. ® Royal Society of Western Australia 1994 Current knowledge of the occurrence, biology and im¬ pact of pathogens in Western Australia has mainly accumu¬ lated from research initiated in response to potential disease threats to wood production, such as early research on wood rots and jarrah dieback (Shearer 1992b), or from opportunis¬ tic individual curiosity and observation. Since the mid 1980's, there has been a growing appreciation of the threat of disease to conservation values (Shearer 1992b). This paper assesses the relative occurrence of pathogens on native plant species from published information and personal databases and then describes the biology, distribu¬ tion and disease expression of the major pathogens affecting Western Australian native plant communities. Diseases and pathogens of Western Australian native plants Occurrence of diseases and pathogens on Western Aus¬ tralia native plants was compiled from published informa¬ tion and reports (Brandis et aL 1984; Brittain 1989; Hill 1990; Shearer 1992a; Shivas 1989; Wills 1993) and from survey and isolation databases of my own and those of S Bellgard, F Bunny and the Plant Health Service, Department Conser¬ vation and Land Management. Pathogens were databased 113 i according to name, host name and family, disease group ( .g. Pythiacious root rots, Armillaria root rots, rust, mildew, etc) and Botanical Province. Nomenclature of plant taxa follows that of Green (1985). The database consisted of a total of 936 entries, and did not include reports of pathogens on hosts m nurseries. Because of the lack of comprehensive surveys, the current information is incomplete and the database repre¬ sents an indication of the occurrence of a pathogen on a particular host species rather than the intrinsic frequency of its occurrence. The number of occurrences within families can give some measure of the pathogens relative impor¬ tance. The analysis will, however, favour pathogens and plant taxa most studied. The analysis will also favour patho¬ gens with wide host range, and disfavour those with high impact but narrow host range. Ninety-one per cent of the pathogen reports were from the South-West Botanical Province and 2% from «ch of the Eremaean and Northern Botanical Provinces. Th>s reflects the greater concentration of research activity that has occurred in the South-West Botanical Province. Disease groups occurring in three or more families and families within which there was three or more occurrences o a pathogen are shown in Table 1. Bacterial diseases, galls, downy and black mildews, ergot, and leaf moulds were infrequently reported on native plants. Pathogens were in¬ frequently reported on species within the families: Aizoaceae, Amaranthaceae, Amaryllidaceae, Annonaceae, Anthericaceae, Apocynaceae, Arecaceae, Asphodelaceae, Cupressaceae, Cyperaceae, Dennstaedtiaceae, Geraniaceae, Juncaceae, Lamiaceae, Linaceae, Loganiaceae, Olacaceae, Onagraceae, Phormiaceae, Pittosporaceae, Podocarpaceae, Polygonaceae, Portulacaceae, Rubiaceae, Solanaceae, Styiidiaceae, Tremandraceae, Verbenaceae and Zamiaceae. This list of families with infrequent disease does not simply reflect families with few species, as only 25% of the families have 5 or less species and just over a third have more than 40 species (Green 1985). The list may represent families that are relatively disease free, but plant taxa may also be included because of limited investigation of disease occurrence. Pythiacious root rots, rusts, Armillaria root rots, stem cankers, and leaf spots and blights were frequently reported on native plants (Table 1). Families most affected by disease were: Proteaceae, Myrtaceae, Mimosaceae, Papilionaceae, Haemodoraceae, Goodeniaceae, Epacridaceae, Poaceae and Table 1 r within disease groups and families for which there was three or more records; less Frequency of occurrence of pathog included are listed in the text. Totals are for all entries m the database and do freauentlv occurring groups and families not mcluaea are iisieu u i u .c q y ^ not necessarily match the row or column totals. Family Pythiacious Rusts Armillari root rots root rot Proteaceae 136 3 39 Myrtaceae 55 25 Mimosaceae 5 73 13 Papilionaceae 29 7 18 Haemodoraceae 5 27 1 Goodeniaceae 2 37 1 Epacridaceae 24 14 Poaceae 14 Chenopodiaceae 13 Dilleniaceae 12 4 Asteraceae 11 1 Rutaceae 7 2 2 Colchicaceae 10 Orchid aceae 9 Xanthorrhoeaceae 6 2 Casuarinaceae 4 3 Iridaceae 6 1 Rhamnaceae 1 3 2 Apiaceae 1 2 1 Dasypogonaceae 4 1 Euphorbiaceae 2 3 Restionaceae 2 Santalaceae 2 Sterculiaceae 3 1 1 Thymelaeaceae 2 3 Ranunculaceae 2 1 Myoporaceae 1 1 Zygophyllaceae 3 Total for all families 310 232 147 Disease group Smuts Total for all disease Stem cankers Leaf spots and blights Wood rots Powdery Crown mildews rot White rust groups 53 33 27 17 7 6 6 1 1 5 1 22 17 258 161 98 61 42 41 39 35 22 17 15 11 10 10 9 8 8 6 5 5 5 5 5 5 5 4 3 3 91 79 25 23 12 4 936 114 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Chenopodiaceae. Frequency of ocairrence of Pythiacious root rots, Armillaria root rots, stem cankers and leaf spots and blights were similar between families, although there was a more frequent occurrence of Armillaria root rot in the Mimosaceae than for the other disease groups. In contrast, frequency of occurrence of rusts between families was the inverse of that for the previously mentioned disease groups. Families mostly affected by rusts were least affected by the root rots, stem cankers and leaf spots and blights (Table 1). This is especially the case with the Colchicaceae and Orchidaceae which seem to be little affected by diseases other than rust, although this may also be due to limited research on the diseases of these plant taxa. Within the most frequently occurring Pythiacious root rots (Table 1), Phytophthora chinamomi Rands accounted for 54% of reports, P. megasperrna Drechsler for 21% and P. citricola Sawada for 13%. Within the rusts, 53% were Pucciniasp. and 32% Uromycladium tepperianioii (Sacc) McAlpine gall rust of Acacia species. Only one confirmed Armillaria species, A. liiteobubalina Watling & Kile, is known to cause Armillaria root rot in Western Australia (Kile et al. 1983). Of the 17 stem canker pathogens recorded, Botrposphaeria sp. was the most frequent (45% of reports) followed by Zythiostroma sp. (14%) and Cryptodiaporthe sp. (11%). Of the 46 leaf spot and blight pathogens recorded, three species were the most frequently recorded, with each only 6% of reports. Biology of major pathogens Knowledge of the life cycle and biology of P. cinnamomi in native communities of south-western Australia is derived mainly from research conducted in the northern Eucalyptus marginata Donn ex Smith forest (Dell & Malajczuk 1989; Shearer & Tippett 1989), but relatively little specific informa¬ tion is known of the biology of P. cinnamomi in non-forest communities. In addition, little specific information is known of the factors affecting spore production, infection and host susceptibility to infection for Phytophthora species other than P. cinnamomi, rusts, A. liiteobubalina and stem cankers in the south-western Australian environment. Phytophthora root rots Phytophthora spp. are introduced soil-borne pathogens belonging to the class Oomycota, a relative primitive group of fungi having a number of morphological, physiological and biochemical characteristics found in certain protozoa and bacteria and an ancestral affiliation with heterokont algae (Barr 1983). In evolutionary development, Phytophthora belongs to a transitional group between entirely aquatic and completely terrestrial fungi. This is reflected in their complex life cycles dependent on moist conditions for survival, sporulation, dispersal and infection, and in the initiation of Various adaptation strategies to cope with the fluctuating soil environment. Phytophthora cinnamomi is a major pathogen in the alter¬ nating temperature and moisture mediterranean climate of south-western Australia, despite the fact it is an introduced, nioisture-dependent microorganism. This has occurred be¬ cause movement of infected soil by human activity has spread the pathogen throughout the region (see below). In addition, the soils and topography in conjunction with the hydrological cycle and susceptible plant communities have provided niches within the soil profile whereby P. cinnamomi can survive adverse conditions, and be spread in water or by root-to-root contact to infect the roots of a wide range of hosts. The interactions that have created the diversity of microenvironments and conditions favourable for sporulation, survival, dispersal and infection are detailed in Shearer & Tippett (1989) and can only briefly be described here. Phytophthora cinnamomi takes advantage of favourable warm and moist soil conditions in autumn and spring, and presence of susceptible tissue, by rapidly producing various spore types in an expanding phase of population growth. During unfavourable conditions of low soil moisture, ab¬ sence of susceptible tissue, and high microbial activity, the fungal hyphae are lysed and disintegrate, releasing resistant spores specialised for survival. Vegetative reproduction is by sporangia that release infectious motile zoospores in water. This is the main way the Phytophthora species repro¬ duce and infect plants. Spherical, sedentary chlamydospores may also be vegetati vely produced, but their role in infection and survival in south-western Australia is poorly under¬ stood. Under certain conditions, sexual reproduction by thick-walled oospores occurs. Oospore production by P. cinnamomi is probably infrequent in south-western Australia as two mating types are required for spore induction but only one mating type predominates in the region. In com¬ parison, P. citricola and P. megasperrna readily produce oospores from the one mating type. Reproduction by oospores is probably an important survival mechanism for P. citricola and P. megasperrna as the thick walled spores are more resistant to drying than are zoospores. Once Phytophthora species have entered the roots of sus¬ ceptible hosts, primary symptoms of infection are evident as advancing fronts of necrosis (lesions) in the inner bark of roots and stems. Lesions are most evident in fleshy primary roots as a root rot. The fungi kill their hosts by destroying the roots and girdling the base of the stem, depriving the plant of access to nutrients and water. Host plant species occur mainly in the Proteaceae, Myrtaceae, Papilionaceae, Epacridaceae and Dilleniaceae (Table 1). Rusts Although rusts are the second most frequent pathogens on native plant taxa in south-western Australia (Table 1), research on their biology in the region is limited to only three studies (Goodwin 1963; Verhoogt & Sivasithamparam 1985; Nichol 1986). Rust fungi are of the order Uredinales of the class Basidiomycota and are destructive pathogens to many agriculture and forest crops. In contrast to Phytophthora, rust pathogens on native plants are probably endemic, they complete their life cycles on the above ground plant parts and they are mainly dispersed as air borne spores. Also, unlike root rots and stem cankers which can live and repro¬ duce on dead tissue, the rust fungi are obligatory parasitic, requiring living hosts for normal development. The life cycles of rusts are more complex than those found in any other group of fungi, and typically consist of four or five reproductive stages in a regular sequence. Details of the stages can be found elsewhere (Agrios 1978) and are briefly described as follows. Pycniospores and receptive hyphae are Journal of the Royal Society of Western Australia, 77 (4), December 1994 produced in pycnia. Pycniospores serve as spermatia and are transferred to other pycnia by insects and fuse to form binucleatehyphae. Aeciospores formed from the binucleate hyphae are wind-dispersed to infect hosts other than the one on which they are produced. Uredospores are produced from binucleate mycelium from a germinating aeciospore or a uredospore. Uredospores are generally the main repeating stage of rusts and can withstand adverse conditions of long- range dispersal from plant to plant by wind. Sexual repro¬ duction is by teleutospores, which are not dispersed but germinate to produce basidiospores. The basidiospores are temperature and moisture sensitive, and dispersed by wind over short distances. Within this life cycle pattern, long- cycled rusts produce at least one type of binucleate spore in addition to the teleutospore, while for short-cycled rusts the teleutospore is the only binucleate spore produced. The life cycle may be completed on the one host (autoecious) or on two distinct hosts (heteroecious). In Western Australia, ga\\rust(Urom}/cladium tepperiamim) is a short-cycled autoecious rust producing pycniospores, teleutospores and basidiospores mainly on Acacia species (Goodwin 1963). Some rust taxa on orchids are long-cycled as they produce aeciospores, uredospores and teleutospores (Nichol et al. 1988). It is not known whether the life cycle of rusts on plant taxa other than the Mimosaceae (Table 1) are autoecious or heteroecious. Host plant species occur mainly in the Mimosaceae, Goodeniaceae, Haemodoraceae, Poaceae, Chenopodiaceae, Asteraceae, Orchidaceae and Colchicaceae (Table 1). Uromydadium tepperiauum infection stimulates the Acacia host to form galls and/or'witches brooms' (Goodwin 1963). Infection of the growing point results in a witches broom caused by reduction of the growing axis and a proliferation of lateral buds. Galls may be globose or elongated and can form on different plant organs, although formation on a particular plant part is consistent within an Acacia sp. (Good¬ win 1963). Gall formation on inflorescences reduces fertilisa¬ tion and fruit development. Uredospore production by leaf rusts rupture the leaf epidermis, reducing photosynthetic and transpiration processes. In Orchidaceae, leaves infected with rust senesce earlier than healthy leaves and rusted plants produce fewer flowers than healthy plants (Nichol 1986). Armillaria root rot Research on A. luteobubalina in south-western Australia has mainly concentrated on the impact of the pathogen in forested areas (Pearce et al. 1986; Shearer & Tippett 1988; Pearce & Malajczuk 1990a), and the potential for biocontrol with wood decay fungi (Pearce & Malajczuk 1990b). The impact of the pathogen on shrubland and heathland com¬ munities has only recently been recognised (Shearer et al. 1994). Specific details are lacking on the mechanisms of infection and host colonisation of A. luteobuhalina in Western Australia and factors affecting host susceptibility to infec¬ tion (Shearer 1992a). Details of the life cycle of Arinillaria species are reviewed in Shaw & Kile (1991). Armillaria luteobubalifia is an indigenous species of mush- room-producing primary pathogen of the order Agaricales, class Basidiomycota. Infection from A. luteobubalina occurs from aerial dispersed basidiospores or through mycelial transfer at root contacts. Growth through the soil by rhizomorphs is not an important mechanism of spread in south-western Australia (Pearce et al. 1986; Shearer & Tippett 1988) as the seasonal pattern of temperature and moisture associated with the mediterranean climate of the region, is not conducive for rhizomorph growth (Pearce & Malajczuk 1990c). Basidiospores, formed by sexual recombination of gametes, are shed in autumn-winter from annual fruiting bodies that develop on decayed roots and stems of dead and living trees. Fruiting bodies of A. luteobubalina are mainly produced in June and July (Pearce et al. 1986; Shearer & Tippett 1988). How basidiospores infect woody tissue is poorly understood and is probably an infrequent event (Kile 1983). The distribution of infection points and aerial dis¬ persed sexually produced basidiospores results in a discon¬ tinuous, discrete distribution of infections of different geno¬ types. The number and distribution of different genotypes can provide an e.stimate of the frequency of infection from basidiospores (Kile 1983), but no analysis of this type has been done for south-western Australia. The pathogen spreads within disease centres by mycelial growth through roots. In susceptible E. loandoo Blakely, the mean rate of disease extension over a 8 year period was 2.04 ± 1.05 m yr’ (Shearer unpuh. obs.). This is comparable to mean maximum rates of 0.7-1.6 m yr^ found by Kile (1983) for Victorian forest.-New infections are established by contact between roots and stems, and dead roots and stumps increase the inoculum level. In mixed eucalypt forests in the highlands of west- central Victoria, the pathogen can survive in stumps for up to 30 years (Kile 1981). Armillaria luteobubalina establishes in the bark and causes columns of decay within roots and stems of host species. The pathogen spreads tangentially in the inner bark of suscepti¬ ble hosts, often resulting in girdling of the stem collar and host death (Pearce et ai 1986; Shearer & Tippett 1988). Host plant species occur mainly in the Proteaceae, Myrtaceae, Papilionaceae, Epacridaceae and Mimosaceae (Table 1). Stem cankers The contribution of canker fungi to stem and branch death in south-western Australia has largely been ignored (Davison & Tay 1983; Shearer 1992a). Mortality and decline of marri and red flowering gum were associated with stem cankers in the mid 1930's (Smith 1970). Davison & Tay (1983) identified a number of pathogenic fungi associated with stem and branch cankers of forest trees in south-western Australia. In 1989, a species of Diplodina (sexual stage Cryptodiaportlie) was found killing Banksia coccinea R. Brown on the south coast of the state (Shearer & Fairman 1991)- Interpretation of the cause of stem cankering can be compli¬ cated as some fungi are frequently isolated from cankers but they are secondary invaders of the diseased tissue, Cytospora eucalypticola van der Westhuizen is an example of a fre¬ quently isolated fungus that pathogenicity tests have shown to be a nonagressive facultative parasite (Davison & Tay 1983; Shearer et al. 1987). The origins of stem canker fungi in south-western Aus¬ tralia are uncertain. Botryosphaeria ribis Gossenb. & Dugg. is possibly an introduced pathogen (Davison & Tay 1983) and it is widely distributed on a diverse range of hosts in the tropical and temperate regions of the world. The Cryptodiaporthe pathogen of B. coccinea is possibly endemic as 116 Journal of the Royal Society of Western Australia, 77 (4), December 1994 it is a new species (Bathgate et al. 1994) and has a very limited host range within the Proteaceae (see below). How the canker-causing fungi complete their life cycles in south-western Australia requires further research. This is complicated by uncertainties in the identity of the various spore stages of canker fungi on native plants in this state. For example, the asexual stage of Endothia isolated from Myrtaceae in Western Australia has been identified as En. gyrosa (Schw. Fr.) Fr. by isozyme analysis against voucher specimens (Davison & Coates 1991), Even though the sexual ascospore stage occurs in eastern Australia (Walker et al 1985), it has yet to be recorded in Western Australia. Canker fungi kill the aerial parts of plants. This is in comparison to disease caused by Phytophthora and Armillaria that kill plants from the roots up. Hosts affected by canker fungi occur mainly in the Proteaceae and Myrtaceae (Table 1). The fungi sporulate in dead bark and are dispersed as sexually produced ascospores in wind currents or asexually produced pycnidiospores in rain splash. The mode of entry of germinating spores is either direct or gained through lenticels or wounds from branch stubs, broken branches and insect damage. Phloem and sapwood invasion results in sunken cracked areas on the stem that may expose the xylem and exude kino. Cankers thus formed can be annual, peren¬ nial or diffuse. In annual cankers, lesion development is contained by host defense mechanisms wi thin the first year's invasion. Botryosphaeria ribis generally forms annual cankers unless stress factors affect the host-pathogen interaction, as described in the next section. Perennial cankers denoted by concentric rings are formed when invasion by the pathogen is walled off, but the pathogen survives on dead tissue to re¬ invade healthy tissue in the following years. Large Target'- like cankers occuronE.cfl/op/iy//flLindIey and E.gomphocephala DC, but the causal pathogen has yet to be determined. Diffuse cankers occur when lesions rapidly progress along the stem, resulting in gradual decline from death of twigs and lateral branches to rapid death of leaders in a few years. Diffuse canker development by Cryptodiaporthe sp. leads to death of infected B. coccmea, and destruction of diseased stands in a few years (Shearer & Fairman 1991; see below). The effect of death of canker-infected stems and branches on leaf area and host plant functioning has not been deter¬ mined. Disease caused by canker fungi can be aggravated by transient stress factors (Schoeneweiss 1975). Trees planted outside the normal range may experience environmental stress with an associated decline in resistance to infection by canker organisms (Shearer et al 1987). Stress from two days of above 40 °C and high winds in February 1991, was associated with rapid extension of Bo. ribis lesions in stems of B. speciosa R Brown near Hopetoun. The stand was severely debilitated by the infection and trees died. Twelve months later, many of the surviving B. speciosa trees had again contained the Bo. ribis lesions and formed new epicormics below the wallcd-off lesion margin. Distribution Phytophthora cinnamomi Phytophthora cinnamomi is the most common and destruc¬ tive of the Phytophthora species found in native communities of the south-west. It occurs in the area bounded by Eneabba north of Perth, east of Dryandra near Popanyinning, and Cape Arid east of Esperance on the south coast (Fig 1). Figure 1. Distribution of Phytophthora cinnamomi disease centres in south-western Australia, compiled mainly from assessment plots and mapping, and supplemented by isolation records. Greatest incidence of P. cinnamomi occurs in the northern and southern E. marginata forest (Fig 1). This is partly due to environment and partly to historical factors related to hu¬ man activity (Shearer & Tippett 1989; Shearer 1992a). The pathogen frequently occurs on the acidic leached sands of the Bassendean Dune System of the Swan Coastal Plain, Gavin Sands of the Leeuwin-Naturaliste Ridge, laterite soils and winter wet flats of the d'Entrecasteaux and Walpole- Nornalup National Parks and the Keystone and Gardner geomorphic units in areas on the south coast such as West Cape Howe and Two Peoples Bay. Incidence is high in the sandy deposits of the Stirling Range National Park (the rectangle of occurrences north of Albany, Fig 1). Infections fringe the Fitzgerald River National Park east of Bremer Bay, but a 6 km long infection occurs within the park. A transect from the coast, inland between 31.5° and 33.5° S, shows that P, cinnamomi disease centres are absent from coastal dunes, but increase in frequency in the Bassendean Dune System and Pinjarra Plain to the west of the Darling Scarp (Fig 2). Frequency of occurrence is greatest in the northern £. marginata forest on the western edge of the Darling Scarp, decreasing rapidly to the drier eastern edge of the E. marginata forest (Fig 2). Rusts Rusts are widely distributed on native plant taxa through- out the south-west (Fig 3). This is especially so for 17. tepperianum, which occurs relatively frequently on Acacia spp. in coastal areas and in the eastern wheatbelt and gold¬ fields. Uromycladium tejiperianum isprobab\y the most widely distributed pathogen in native communities in south-west¬ ern Australia (compare Fig 3 with Figs 1 and 4-6). However, because of the limited research on rusts of native plants of Western Australia, many more surveys are needed for a more accurate picture of the distribution of rusts in native communities of the state. Armillaria luteobubalina Armillaria luteobubalina disease centres mainly occur in coastal dune vegetation and forested areas (Fig 4). In vegeta- 117 Journal of the Royal Society of Western Australia, 77 (4), December 1994 I . -- - 116 116.5 117 Longitude (class midpoint) Figure 2. Occurrence of disease centres of Armillaria luteobiibalina and Phytophthora cinnamomi in a transect between 31.5° and 33.5° S and from the coast (115.6° E) inland to 117.4° E. The plot is percent¬ age of occurrence of disease centres in longitude classes of one tenth of a degree. Figure 3. Distribution of rusts on native plants in south-western Australia, compiled from reports in Shivas (1989) and Nichol(1986). Rust taxa were: Puccinia for Haemodoraceae, Uromycladium for Mimosaceae, Aecidium and Puccinia for Asteraceae and Colchicaceae, and Puccinia and Uromyces for Goodeniaceae and Orchidaceae. tion on the non-podsol sands of the coastal dunes, A. luteobubalim occurs as far north as Cervantes and around the coast to Cape Arid (Fig 4). The pathogen also occurs in E. gomphocephala forest and Banksia woodland of the Spearwood Dune System and equivalents, just inland from coastal dunes but rarely occurs in communities on the acid sands of the Bassendean Dune System. The pathogen fre¬ quently occurs in the northern and southern £. marginata forest, the £. diversicolor F. Muell. forest in the south, and in £. wandoo forest to the east. Figure 4. Distribution of Armillaria luteobiibalina disease centres in south-western Australia, compiled from isolation records, assess¬ ment plots and mapping. In comparison to P. cinnamomi, A. luteobubalina occurs on the coast and rarely occurs in the Bassendean Dune system to the west of the Darling scarp (Fig 2). Distribution within the northern forest tends to be more skewed to the east, than for P. cinnamomi, and there is a greater frequency of occur¬ rence in the E. wandoo forest east of the £. ttiarginata forest (Fig 2). Stem canker pathogens Various canker pathogens, mainly on Myrtaceae and Proteaceae, are widely distributed throughout the south¬ western Australian region (Fig 5). The distribution map is incomplete, however, as there has been inadequate sam¬ pling in the eastern wheatbelt and goldfields. Figure 5. Distribution of stem canker fungi on Proteaceae and Myrtaceae in south-western Australia, compiled from isolation records. The recently discovered Cryptodiaporthe canker of Proteaceae has an interesting discontinuous distribution (Fig 6). On the south coast, the pathogen infects B. coccinea throughout its geographic range (Fig 6). However for B. grandis Willd. and Dryandra sessilis (Knight) Domin, Cryptodiaporthe canker only occurs within a small portion of the geographic range of these two hosts. The pathogen has Journal of the Royal Society of Western Australia, 77 (4), December 1994 fiot been found in an area between the south coast and west ^:oast (Fig 6), even though the area has been sampled (Fig 5). On the west coast, Cryptodiaporthe is an aggressive canker of p. sessilis north of Perth and on B. grandis south of Perth (Fig ^). Curiously, it has been infrequently isolated from these {WO hosts in other areas, even though these species occur and pave been sampled throughout the south-west. Possible pauses of this distribution are currently under investigation. j^igure 6. Distribution of Cryptodiaporthe stem canker of Proteaceae \n south-western Australia, compiled from isolation records. Communities affected fhytophthora cinnatnomi Death of the susceptible understorey species of the proteaceae, Myrtaceae, Papilionaceae and Epacridaceae (Ta- ple 1) is the first indication that P. cinnamomi has spread into a new area. On sites favourable to disease development, a line of dead and dying understorey marks the 'infection front' at the boundary of infested and uninfested areas (Fig 7). Disease impact is more subtle on less favourable, free draining sites, and there is often no clear demarcation be¬ tween infested and uninfested areas. Figure 7. High impact of Phytophthora cinnamomi in Banksia wood¬ land on the Bassendean Dune System of the Swan Coastal Plain. Most of the overstorey of Banksia attenuata R Brown, B. ilicifolia R Brown and B. menziesii R Brown has died in the infested area. Death of dominant key overstorey and understorey species result in reduction of vegetation biomass in the diseased area and the disease front being delineated by a sharp boundary of dying plants. Shearer (1990) described the impact of P. cinnamomi ac¬ cording to a grouping of vegetation systems of Beard (1981). Impact of P. cinnamomi tends to be lowest in coastal commu¬ nities on coastal limestone and forest communities on rela¬ tively fertile red earths associated with major valleys. Impact of P. cinnamomi is also low in inland woodlands and shrublands. However, low disease expression in inland ar¬ eas is probably due to low rainfall unfavourable for patho¬ gen survival and sporulation, rather than a lack of suscepti¬ ble vegetation or soil profile characteristics favourable for pathogen development (Shearer 1990). This is illustrated by the recently observed infection of rare and endangered B. cuneata A S George, located east of Dryandra on the western edge of the wheatbelt (Fig 1). Impact of P. cinnamomi is highest in the E. marginata forest understorey on laterites and Banksia woodlands associated with leached sands and laterites of the Northern and South¬ ern Sandplains and the Swan Coastal Plain. Within these vulnerable communities, the impact of P. cinnamomi results more in changes in community structure and function than in total number of species. For example, infestation of Bfln/csffl woodland on the Swan Coastal Plain resulted in an average of 7 fewer species in infested than non-infested woodland (Shearer & Dillon 1994). However, the loss of these species often resulted from the almost complete death of the domi¬ nant susceptible overstorey and understorey vegetation with a substantial reduction of the vegetation biomass in the diseased area (Shearer & Dillon 1994; Fig 7). Thus in commu¬ nities dominated by rare and endangered plant taxa, such as B. brownii Baxter ex R Brown in the Albany region, infesta¬ tion is resulting in elimination of the threatened taxa. Keighery (1992) lists6 species (2 Andersonia spp., B. brozvnii, 2 Dryandra spp. and Lambertia orbifolia C Gardner) that are currently threatened with extinction from P. cintjamo?7ii infestation. All of these species of Proteaceae occur in the Southern Sandplains, in areas of high impact of P. cintiamotni. Within areas where P. cinnamomi has caused significant damage to susceptible communities, such as on the Swan Coastal Plain, the £. marginata forest and a number of reserves and national parks on the south coast such as Stirling Range National Park, Cape Arid National Park and Two Peoples Bay Nature Reserve, the main challenge is the development of suitable management strategies for communities irreversibly changed by impact of the pathogen. Rusts There is little information available on the impact of rusts in native plant communities of south-western Australia. The impact of rust on native communities of the state cannot accurately be assessed from the current information. Severe infection of Li. tepperianum ultimately results in death of the host (Goodwin 1963) and the pathogen has been used in biological control oi Acaciasaligna (Labill) H L Wendl, a weed in South Africa (Morris 1991). In Orchidaceae, rust infection reduced thecapacity of T/iWymifra crinita Lindley to produce flowers (Nichol 1986). Thus, rust found on rare and endan¬ gered T. tnactnillanii F Muell would need to be considered in conservation plans, as seed production may be reduced by infection (Nichol 1986). Amiillaria hiteobubalina The impact of A. luteobitbalina disease centres can be expressed as: 1) an expanding patch of dead and dying hosts; 119 Journal of the Royal Society of Western Australia, 77 (4), December 1994 2) dead hosts occurring frequently, but at random, in patches; 3) dead hosts occurring infrequently, but individually, or at random in patches; and 4) small patches of dead and dying hosts occurring in young stands, but the patches of mortality fail to expand as the stand ages. The first and second impact type mainly occur in coastal dune vegetation (Fig 8) and E. wandoo forest (Fig 9). The disease centres can be quite large, averaging 1.7 ± 0.2 ha (range 0.02 - 6.5 ha) for coastal dune vegetation (Shearer et al. 1994) and 1.2 ± 0.3 ha (range 0.01 - 8 ha) for £. wandoo forest (Shearer iin}mb. ofcs.). Most of the susceptible hosts are killed within the disease centres of coastal dune and £. wandoo communities, leaving open de¬ nuded areas which encourage severe wind erosion of coastal dunes (Fig 8A). In coastal dunes, geographically restricted Callitris prcisii Miq (Fig 8B) and rare and endangered B, brownii and B. occidentalis R Brownformosa Hopper are threatened by infection. In the £. ivandoo forest of the Stirling Range Na¬ tional Park, A. luteobubalina infestation is killing Choretrum glomeratiim R Brown, the only food plant for the larvae of the rare brown azure butterfly (Wills & Kinnear 1993). The third and fourth impact type mainly occur in E. diversicolor, E. gomphocqiiiala and E. marginata forests. Stem canker pathogens Stem canker pathogens are having considerable impact in communities dominated by Proteaceae and Myrtaceae in south-western Australia. Cri/ptodiaporthe stem canker is caus¬ ing high mortality of B. coccinea (Fig 10) throughout the Banksia's geographic range on the south coast (Fig 6). In one monitored site, plant death increased from 40% to 98% in 2.7 years. The pathogen is also causing severe branch and stem cankering of D. sessilis north of Perth and B. grandis, south of Perth. On the south coast, a Zpthiostroina sp. causes stem cankers of B. baxteri and Bo. ribis infection has debilitated stands of B. speciosa in association with climatic stress. In eucalypt forest communities, stem canker fungi are associated with crown decline, stem cankering and mortal¬ ity of E.ficifolia F Muell, £. calopliylla and £. gomphocephala. In each case, the causal pathogen has yet to be identified, although En. gyrosa and Bo. ribis have been isolated from dying £. gomphocephala (Shearer unpub. obs.). Canker fungi have been associated with the complex of factors causing crown decline in £. wandoo (Albone 1989). Cankers are also having an impact on myrtacious dominated communities, other than forest. For example, a Phomopsis sp. was isolated from dying branches of Calothamnus quadrifidus R Brown showing severe canopy decline throughout the northern sandplain in 1993 (Shearer unpiib. obs.}. Conclusions Functional diversity and dynamic balance in native eco¬ systems result in explosive epidemics of disease being un¬ common and limited in space and time (Zadoks & Schein 1979). Why then are such explosiveepidemics of Phylophthora species, A. luteobubalina and Cryptodiaporthe canker of Proteaceae currently occurring in plant communitiesof south¬ western Australia? Phytophthora species are human intro¬ duced pathogens to native plant ecosystems of south-west¬ ern Australia and their impact is related to the intensity of human activity, occurrence of sub-surface soil moisture and temperature conditions that favour survival, multiplication Figure 8. High impact of Armillaria httcolmbalina in coastal dune vegetation. A, Death of dune vegetation in an infested area has resulted in denuded areas subject to wind erosion in Yalgorup National Park; B, Most Callitris prcisii have died in a disease centre on Carden Island. Figure 9. High mortality of hosts in an Armillaria luteohubalim disease centre in Eucalyptus wandoo forest near Kojonup. and spread of the pathogen and large numbers of .susceptible key plant taxa that have not co-evo!vcd with the pathogens. Research has elucidated many of these interactions for P. cinnamojui (Shearer & Tippett 1989), but a greater under¬ standing is required for other Phytophthora species such as P. citricola and P. megasperma. In contrast to Phytophthora, the situation for A. luteobubalina and Ciyptodiaporthe canker of Proteaceae is different, as they arc probably native patho- 120 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Figure 10. A stand of Batiksia cocdnea at Cheyne Beach, east of Albany, killed by Cryptodiaporthe canker. Mortality within this stand increased from 40% in 1989 to 98% in 1992. gens and presumably have co-evoIved with the existing vegetation communities. Current knowledge is inadequate to determine whether the prevailing impacts observed relate to a periodic change in disease intensity, or whether they represent more permanent long-term changes. Conservation of plant taxa requires a much better inven¬ tory, than is available at present, of the incidence and status of the various pathogens that occur in native communities of south-western Australia. As noted in this paper, the record¬ ing of most pathogen occurrences on native plants in this state is the result of opportunistic re.search, and comprehen¬ sive surveys have yet to be attempted. Pathological research of native plant taxa has tended to be dominated by P. cinnamofni to the exclusion of other pathogens. Compre¬ hensive surveys would ensure objective assessment of the importance of hitherto ignored pathogens or pathogen/ community interactions. This is illustrated by the recent recognition of the high impact of A. luteobiibalimi in coastal communities (Shearer et al. 1994) and Cri/ptodiaporthe canker of Proteaceae (Shearer & Fairman 1991; Bathgate ct al. 1994). Presumably, these pathogens have been impacting on the respective communities well before their recent recognition. Biogeographical surveys of fauna and flora in communities or National Parks need to include a census of fungi occurring within the areas. Uncertainties in the taxonomy of fungi in this state com¬ plicate inventory of the occurrence and importance of patho¬ gens on native plants (Shearer 1992a). A number of patho¬ gens are undescribed species. This is further complicated by the occurrence of biological species within species com¬ plexes, such as may be occurring in P. megasperma (Bcllgard ct al. 1994). Fungal taxonomic studies are fundamental to assessment of the relative importance of pathogens. Prediction of the likely long-term effects of pathogens on native plant communities requires a much better under¬ standing of their life cycles and biology in the south-western Australian environment than is available at present. By their impact, pathogens are undeniably affecting the evolution of plant communities of the state. However there is only a conceptual understanding of the selection pressures patho¬ gens are placing on community composition and function¬ ing, and in turn, the selection pressures environment and community composition are placing on the pathogens. In¬ formation on the biology and ecology of pathogens in native communities is needed to determine whether current im¬ pacts of endemic pathogens are short term perturbations or part of long term cycles in pathogen-community-environ¬ ment interactions. Such information is also essential to the determination of the likely consequences of disease, and the application of appropriate control strategies. Achwwlcdgwents: My thanks to interpreters of the Forest Management Branch and South Coa.sl Region for locations of P. dunamomi and A. luteobubaliria disease centres, and Colin Crane and John Dodd for help in preparing the disease centre co-ordinates for mapping. References Agrios G N 1978 Plant Pathology. Academic Press, London. Albone P1989 Wandoo dieback. In: Insect and Rural Tree Decline: In Search of Solutions (ed B Schur) Denmark Environment Centre, Denmark, West¬ ern Australia, 17-18. Barr D J S 1983 The zoosporic grouping of plant pathogens. Entity or non¬ entity? In: Zoosporic Plant Pathogens. A Modern Perspective (ed S T Buczacki) Academic Press, London, 43-83. Bathgate], Shearer B & Roh! L1994 Cryptodiaporthesp: a new canker pathogen threatening Bauksia cocemea. In: Handbook of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Au.stralia, Perth, 19. Beard J S1981 Vegetation Survey of Western Australia. The Vegetation of the Swan Area. University of Western Australia Press, Nedlands. Bellgard S E,Shearer B L, Crane C E &Smith B j 1994 Morphological variability exhibited by Pbytophthora tnegaiperma retrieved from diseased areas of Western Australian bushland. In: Handbook of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia (eds R T Wills & W A Coxviing) Royal Society of Western Au.stralia and the Ecological Society of Australia, Perth, 20. Brandis A, Hill T, Keighery G & Tippett J 1984 Dieback in the Cape Arid National Park and other areas of concern. Department Conservation and Land Management, Como, unpubli.shed report. Brittain T 1989 A survey of dieback disease due to Phytophtlwra dnnamomi in Two Peoples Bay Nature Reserve. Department Con.servation and Land Management, Como, unpublished report. Davison E M & Coates D J 1991 Identification of Cn/phouectria cubensis and Endothingyrosa from cucalypts in Western Australia using isozyme analy¬ sis. Australasian Plant Pathology 20:157-160. Davison E M & Tay F C S 1983 Twig, branch and upper trunk cankers of Eucalyptus margiiiata. Plant Disease 67:1285-1287. Dell B & Mala/czuk N 1989 Jarrah dieback - a disease caused by Phytophthora cifinawomi. In: The Jarrah Forest. A Complex Mediterranean Ecosystem (eds B Dell, J J Havel & N Malajczuk) Kluwer Academic Publishers, Dordrecht, 67-87. Forestry Canada 1993 Forest Insect and Disease Conditions in Canada 1990. (Compiled by B H Moody). Forest Insect and Disease Survey, Ontario. Goodwin J1963 A study of the Acacia rust Uromycladium tepperiammi (Sacc.) McAlpine. University of Western Australia, Perth, Honours Thesis. Green J W 1985 Census of the Vascular Plants of Western Australia, 2"‘* Ed. We.stern Australian Herbarium, Perth. Hill T C j 1990 Dieback disease and other Phytophthora species in the Northern Kwongan. In: Nature Coaservation, Landscape and Recreation Values of the Lesueur Area (eds A A Burbidge, S D Hopper & S van Leeuwen) Bulletin 424, Environment Protection Authority, Perth, 89-97. Keighery G1992 The impact of Phytophthora species on rare plants. In: Dieback - What is the Future? (Compiled by M J Freeman, R Hart & M Ryall) The Northern Sandplains Dieback Working Part)', Perth, 29-36. Kile G A 1981 Armillaria lutcohubalina: a primary cause of decline and death of trees in mixed species eucalypt forests in central Victoria. Australian Forest Researdi 11:63-77. Kile G A 1983 Identification of genotypes and the clonal development of Armillaria luteohubaliua Watling & Kile in eucalypt forests. Australian Journal of Botany 31:657-671. 121 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Kile G A, Walling R, Malajczuk N & Shearer B L1983 Occurrence of Annillaria luteobubalina Walling and Kile in Weslem Australia. Australasian Plant Pathology 12:18-20. Morris M J1991 The use of plant pathogens for biological weed control in South Africa. Agriculture, Ecosystems and Environment 37:239-255. Nichol AMI 986 Rust diseases of native orchids of the south-west of Western Australia. University of Western Australia, Perth, Honours Thesis. Nichol A, Sivasithamparam K & Dixon K W1988 Rust infections of Western Australian orchids. Lindleyana 3:1-8. PearceMH&MalajczukN1990aStumpcoloni2ationby Armillarialuteobubalitta and other wood decay fungi in an age series of cut-over stumps in karri {Eucalyptusdiversicolor)Te^TOw\h forests in south-western Australia. New Phytologist 115:129-138. Pearce M H & Malajczuk N 1990b Inoculation of Eucalyptus diversicolor thinning stumps with wood decay fungi for control of Armillaria luteobubalina. Mycological Research 94:32-37. Pearce M H & Malajczuk N 1990c Factors affecting growth of Armillaria luteobubalina rhizomorphs in soil. Mycological Research 94:38-48. Pearce M H, Malajczuk N & Kile G A 1986 The occurrence and effects of Armillaria luteobubalina in the karri {Eucalyptus diversicolor F. Muell.) forests of Western Australia. Australian Forest Research 16:243-259. Schoeneweiss D F 1975 Predispositioa stress, and plant disease. Annual Review of Phytopathology 13:193-211. Shaw C G & Kile G A 1991 Armillaria root disease. Agriculture Handbook Number 691. Forest Service, United States Department of Agriculture, Washington, D.C. Shearer B L1990 Dieback of native plant communities caused by Phyfophthora species - A major factor affecting land use in south-western Australia. Land and Water Research News No. 5:15-26. Shearer B L 1992a The ecological implications of disease in the southern forest of south-w'estem Australia. In: Research on the Impact of Forest Manage¬ ment in South-West Weslem Australia (ed M Lewis) CALM Occasional Paper Number 2 /92. Department Conser\'ation and Land Management, Como, 99-113. Shearer B L 1992b Plant disease research in the Department Conservation and Land Management. In: Dieback • What is the Future? (ed M J Freeman, R Hart & M Ryall) The Northern Sandplains Dieback Working Party, Perth, 7-13. Shearer B L & Dillon M 1994 Impact of Phytophthora cinnamomi infestation in Banksia woodlands on the Swan Coastal Plain south of Perth. In: Hand¬ book of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 36. Shearer B L & Fairman R C1991 Aerial canker fungi threaten Banksia coccinea. Abstract 85/C16. Proceedings of the Conservation Biology in Australia and Oceania Conference, University of Queensland. Shearer B L & Hill T C1989 Diseases of Banksia woodlands on the Bassendean and Spearwood Dune Systems. Journal of the Royal Society of Western Australia 71:113-114. Shearer B L & Tippett] T1988 Distribution and impact oi Armillaria luteobubalina in the Eucalyptus marginata forest of south-western Australia. Australian Journal of Botany 36:433-445. Shearer B L & Tippett J T1989 Jarrah dieback: The dynamics and management of Phytophthora cinnamomi in the jarrah (Eucalyptus marginata) forest of south-western Australia. Department of Conservation and Land Manage¬ ment, Research Bulletin 3. Shearer B L, Tippett J T & Bartle J R 1987 Botryosphaeria ribis infection associated with death of Eucalyptus radiata in species selection trials. Plant Disease 71:140-145. Shearer B L, Fairman R, Crane C & Grant M 1994 Impact of Armillaria luteobubalina infestation in coastal dune communities of south-western Australia. In: Handbook of the Symposium on Plant Diseases in Ecosys¬ tems: threats and impacts in south-western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 35. Shivas R G1989 Fungal and bacterial diseases of plants in Weslem Australia. Journal of the Royal Society of Western Australia 72:1-62. Smith W P C1970 Stem canker disease of red flowering gums. The Journal of Agriculture of Western Australia 11:33-39. Verhoogt M M & Sivasithamparam K 1985 Rust disease of Kangaroo Paws Anigozanthos Labill and Macropidia Drumm ex. Harve. in Western Aus¬ tralia. Australasian Plant Pathology 14:40-42. Walker J, Old K M & Murray D I L 1985 Endothia gyrosa on Eucalyptus in Australia with notes on some other species of Endothia and Cryphonectria. Mycotaxon 23:353-370. Wills R T1993 The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18:145-159. Wills R & Kirmear J 1993 Threats to the Stirling Range. In: Mountains of Mystery. A Natural History of the Stirling Range (eds C Thomson, G Hall & G Friend). Department of Conservation and Land Management, Como, 135-141. Zadoks J C & Schein R D1979 Epidemiology and Plant Disease Management. Oxford University Press, New York. 122 Journal of the Royal Society of Western Australia, 77: 123-126,1994 Role of environment in dieback of jarrah: Effects of waterlogging on jarrah and Phytophthora cinnamomi, and infection of jarrah by P. cinnamomi. E M Davison Department of Conservation and Land Management, PO Box 104, Como WA 6152 Present address: Department of Conservation and Land Management, Brain St, Manjimup WA 6258 Abstract The association of jarrah deaths with poorly drained dieback sites, following exceptionally heavy rainfall, indicates that hypoxic and anoxic soil conditions, which develop when soil is saturated, may be important in determining why jarrah dies. Experimental work has shown that when jarrah seedlings are waterlogged, xylem vessels in the tap root embolise and are sealed off with tyloses, thus root conductivity is suddenly reduced. Growth and development of Phytophthora cinnamomi are also affected because sporulation and vegetative growth are reduced under anaerobic conditions. When jarrah seedlings are inoculated with zoospores, more lesions form on roots in waterlogged soil than on roots maintained in moist soil. Predictions from these experiments are that in the jarrah forest, soil saturation may directly affect jarrah roots by firstly reducing the number of functional vessels in the sapwood; secondly it will decrease sporulation of P, cinnamomibut thirdly will increase the probability of root infection. Less efficient roots and increased infection will affect tree growth and survival. Introduction Although 14.2 percent of the jarrah forest is infested by Phytophthora cinnamomi (Davison & Shearer 1989), jarrah does not invariably die on all infested (dieback) sites. Groups of trees die suddenly and spectacularly weeks or months after exceptionally heavy rainfall. Groups of jarrah deaths in the late 1940's followed the exceptionally wet winters of 1945-1948 (Harding 1949; Waring 1950), deaths in the late 1950's followed exceptionally heavy summer and winter rainfall in 1955 (Podger pers. comm.), deaths in the mid 1960's followed the exceptionally wet winters of 1963-1964 (Podger 1968) and deaths in 1982 followed exceptionally heavy rain¬ fall in January 1982 (Shea et al. 1982). These mass collapse sites, i.c. dieback sites where jarrah trees of all sizes and ages die suddenly, are in water gaining situations or are on soils with impeded drainage (Podger et al. 1965; Shea et al. 1982; Wallace & Hatch 1953; Waring 1950), not on high quality sites with deep, well drained soil profiles (Waring 1950). This association ofjarrah deaths with poorly drained sites after periods of excessive rainfall im¬ plies that mass collapse occurs on sites in which the soil profile is saturated or partly saturated in the occasional wet years. Thus soil saturation (waterlogging) is an important environmental factor which is associated with jarrah deaths, and which has the potential to affect jarrah, P. cinnamomi and root infection. Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 Symptoms in jarrah and their physiological basis When jarrah trees die on mass collapse sites, the whole crown turns brown within a few days (Podger et al. 1965; Shea et al. 1982). These symptoms indicate severe water deficiency which will result if any of the following occur, either singly or in combination: (i) excessive transpiration (e.g. following extremely hot weather), (ii) reduced water uptake by fine roots (c.^. because the soil has dried out or a large proportion of fine roots havebeen damaged by pests or pathogens), (iii) reduced conduction of water between fine roots and foliage {e.g. caused by mechanical damage or rotting of the sapwood). Observations of individual jarrah trees prior to crown death shows that there is a rapid decrease in stem girth for several weeks or months before foliage dies (e.g. Fig 1; Davison & Tay unpublished data). New leaves may still be produced in the crown even though the stem shows symp¬ toms of gradually drying out. The tree shows symptoms of undergoing severe water deficiency which occurs in the stem before the crown, implying that there is either reduced water uptake by fine roots, or there is reduced water move¬ ment between roots and foliage. Water moves between the roots and foliage in sapwood, an outer annulus of lighter coloured wood (xylem). Micro¬ scopic examination of sapwood shows that it is largely composed of small, thick-walled cells which give wood its strength and rigidity, and large xylem vessels through which water moves. Xylem vessels are capillary tubes which, in jarrah, are up to 0.4 mm in diameter, and approximately 50 cm long (Davison & Tay ujtpublished data). In sapwood the majority of vessels conduct water, but a few may be non- 123 Journal of the Royal Society of Western Australia, 77 (4). December 1994 trace ............ tree 22, uninfested area; ^ young leaves present in the crown; ^ crown death. conducting, as indicated by ingrowths, tyloses, from adja¬ cent cells. Tyloses only form in xylem vessels which are gas- filled, not water-filled (Zimmerman 1983). They are one mechanism by which a plant seals off damaged xylem, but it is not known what triggers their formation. Once tyloses form in a xylem vessel, that vessel will never conduct water, so that if there are large numbers of tylosed vessels in sapwood the conductivity will be reduced. Effects of waterlogging on jarrah, P. cinnamomi and root infection Experimental work When soil is saturated with water, it becomes anaerobic because oxygen in the soil solution is rapidly consumed by micro-organisms and roots (Drew 1992). As oxygen diffuses Kl* times more slowly through water than through air, it is used more quickly than it is replaced. Saturated soil will become hypoxic and anoxic more rapidly in summer than in winter, because the solubility of oxygen decreases with increasing temperature, and because respiration increases with increasing temperature. Laboratory measurements using jarrah forest soil show that, when saturated with water, it will become anoxic within 2 days at 20° C, and within 4 to 5 days at 16° C (Davison Tay 1991). Anaerobic respiration is much less efficient than aerobic respiration so that, quite apart from any other physical and chemical changes in the soil, this sudden development of hypoxic and anoxic conditions is biologically very damaging. Podger (1967) showed that jarrah was more sensitive to waterlogging than other forest eucalypts. Further work (Davison & Tay 1985) has shown that when jarrah seedlings are waterlogged under controlled conditions, xylem vessels in the tap root embolise and are sealed off with tyloses, so that roots become less efficient. This happens quickly; the proportion of tylosed vessels is correlated with the duration of waterlogging, and after 14 days at 20° C half of the vessels are blocked. Many woody plants close their stomata when their roots are waterlogged, but jarrah seedlings continue to transpire. Thus, the rate at which seedlings wilt and die depends on both the duration of waterlogging and the transpiration rate of the plants (Davison & Tay 1985). Waterlogging also affects P. cinnaviomi, a fungus which requires matric potentials close to zero for the production of sporangia, discharge and dispersal of zoospores (Gisi et al 1980; Shea ct al 1983; Kinal et al 1993). Sporulation, however, either does not occur or occurs very slowly under anaerobic conditions (Davison & Tay 1986). Zoospore germination is not affected by aeration, but germ tube growth is correlated with oxygen concentration (Davison & Tay 1986). Infection of jarrah roots by zoospores of P. cittnamomi is greater in saturated soil than in soil at field capacity (Davison & Tay 1987). This is because more lesions are formed as a result of increased mobility of zoospores in flooded soil and increased attraction of zoospores to anaerobically respiring roots (Allen & Newhook 1973). Infection does not increase the proportion of occluded vessels in tap roots of jarrah seedlings (Davison & Tay 1987). When large stems and roots are wound inoculated with P, cinmmovtl it preferentially colonises inner bark (Tippett et al 1983; Davison et al 1994). Although wound responses result in occlusion of xylem vessels adjacent to phloem lesions, this is of limited extent (Tippett & Hill 1984). Field observations Predictions for the field from the experimental work reviewed above are: (i) if forest soil is waterlogged suddenly as a result of exceptionally heavy rainfall, the soil solution will rap¬ idly become hypoxic and anoxic; (ii) these conditions will not kill large jarrah roots, but will result in xylem vessels cavitating and becoming oc¬ cluded with tyloses, so that these roots are le.ss efficient at conducting water; (iii) after the soil has drained, new functional xylem vessels will be formed by the root cambium. Thus, over time, the tylosed vessels in the sapwood will be replaced by newly formed vessels; (iv) if there are seasonally high watertables in soil, large jarrah roots will be restricted to well aerated, surface horizons, so that jarrah trees on such sites will have shallow root systems; (v) when the soil temperature is above 15°C and the matric potential is close to zero, P. djwatnomi sporangia will be formed on root lesions; (vi) sporangia will not be formed below a watertable be¬ cause aeration is inadequate, but will be formed in the moist soil above; (vii) zoospores released from sporangia will move pas¬ sively in percolating water through the profile into saturated soil where they will be attracted to, and infect, anaerobically respiring roots; A 124 Journal of the Royal Society of Western Australia, 77 (4), December 1994 (viii) root infection will be more frequent in saturated soil than in moist soil; (ix) the main tissue invaded will be the phloem; (x) reduced hydraulic conductivity of root xylem will re¬ duce the movement of water from the soil to the canopy, while increased infection of root phloem will reduce the movement of photosynthate and hormones from the crown to the roots; (xi) both reduced hydraulic conductivity and increased infection will adversely affect tree growth. Some of these predictions can be compared with field data. Field measurements of perched watertables at Dawn Creek (Nanga Block) in June after 50 mm rain in the previous 4 days showed that the oxygen concentration of the soil solution was 49 percent of water saturated with air (Davison & Tay 1991). Thus, perched watertables rapidly become hypoxic. When jarrah dies on such sites, one would expect to find roots with large numbers of tylosed vessels in the sapwood, and extensively infected roots. Past investigations of jarrah deaths have included both anatomical and pathological studies (Table 1). Large numbers of tyloses were noted by Harding (1949), Stahl & Greaves (1959), Dell & Wallace (1981) and Davison (1993). P. cinnamomi lesions have not been found consistently, although failure by Harding (1949), Stahl & Greaves (1959), Podger (1968,1972), Dell & Wallace (1981) and Sheareref al. (1981) to find lesions on vertical roots might be a result of incomplete root excavations (Table 1). Table 1 Observations on tyloses and Phytophthora cinnamomi in jarrah trees. Investigator(s) Anatomical study Pathological investigations Tyloses noted Consistent and extensive lesions Phytophthora isolated Harding (1949) Yes Yes No No Stahl & Greaves (1959) Yes Yes No NA Podger (1968,1972) NR NR NR Yes Dell& Wallace (1981) Yes Yes Yes Yes Shearer, Shea & Fairman (1981) NR NR Yes Yes Shea, Shearer & Tippett (1982) NR NR Yes Yes Davison (1993) Yes Yes No No NR, not reported; NA, not attempted The most recent investigation of dying jarrah trees in a mass collapse site included the assessment of surface and sinker roots for both infection and tylosed sapwood (Davison 1993). No P. cinnamomi lesions were found. The mean pro¬ portion and standard deviation (calculated from arcsine- Jarrah and understorey response to the introduction of Phytophthora cinnamomi on a well drained site Jarrah + mid- and understorey Rain 3 T baa* - B . , Evapo- ^ transpiration > 'j Rapid drainage ' P. cinnamomi 1 1 Rain Sporulation Rootir 4_1 transpuation t ifecrion A A ■ Rapid drainage' Death of some mid- and understorey Jarrah + reduced competition from mid- and understorey .IfflKk:- I ' • ' .-__ ^ Reduced evapo- 1 T T transpiration Rain Sporulation Root infection ■ ■ V . jI^ j I Slower dramage ..■ii—ii.iiii.ii |» I .i/iVtIljlJlWIWiiTfc' Root regeneration 7 Jarrah grows faster Jarrah and understorey response to the introduction of Phytophthora cinnamomi on an impeded drainage site Jarrah + mid- and understorey . "warn s. . iii Rain transpiration ' ^ Slt»w drainage ' transpuation fechon , 'M’ Slow drainage P. cinnamomi i Rain^'" Sporulation Root in \ZL ^ _-J 1 Death of some mid- and understorey I = Jarrah + reduced competition from mid- and understorey Sporulation Root infection ► i transpiration Exceptional ^ .1 ^ rainfall ^oot conduction , Slow drainage through roots 4% waterlogging Reduced conduction through roots I Jarrah death l^igure 2. Hypothesised responses of jarrah and understorey to the introduction of P. cinnamomi on a well drained and an impeded drainage site. 125 Journal of the Royal Society of Western Australia, 77 (4), December 1994 transformed data) of tylosed vessels in 18 sinker roots was 74.8 per cent (standard deviation 44.2-96.3 per cent), and the mean proportion of tylosed vessels in 25 surface roots was 24.6 per cent (standard deviation 0.8-66.1 per cent). The proportion of tylosed vessel in the sapwood of surface roots from live trees from two other sites was 12.6 per cent (stand¬ ard deviation 1.3-31.3 per cent, n=201), calculated from arc¬ sine transformed data (Davison & Tay iiupublished data). Another prediction from experimental work is that there would be more lesions on roots of jarrah trees growing on poorly drained sites than on well drained sites. In excava¬ tions of apparently healthy trees growing on infested sites, Shearer & Tippett (1989) recovered P. cintiamomi from only 4.7 per cent of large roots. Similarly Davison & Tay (unpub¬ lished data) only found P. dmiat7tomi lesions on 3.4 per cent of 44 large roots from two sites which differed in soil drainage. In this latter study, lesions were too infrequent for statistical analysis. Quite apart from its ability to infect jarrah, P. cintiamomi kills many mid-and understorey plant species (Podger 1968). By reducing vegetation density, it will reduce both intercep¬ tion of rainfall and evapotranspiration from the site (Green¬ wood et al. 1985). These changes will have a major effect on site hydrolog)’, so that dieback sites will be wetter than adjacent uninfested areas. If site topography and/or soil profile characteristics result in poor drainage, this will be exacerbated by a removal of vegetation. Thus there will be an increase in both the incidence and duration of waterlogging on such sites in the occasional wet years (Fig 2). Conclusions In any investigation in plant pathology it is important to know as much about the host as about the pathogen. With studies of P. cinnamomi this may be conceptually difficult because this fungus has such a wide host range that it is natural to concentrate on the pathogen rather than on the many species affected, which in turn can affect the whole ecosystem. Phytophthoracinnamomidoesnoi]ust\nfec\.plants, it also has a dramatic effect on site hydrology by reducing vegetation density. It is also important to consider how associated environmental factors affect the known physi¬ ological limitations of both host and pathogen because this may provide insights into how to predict when and where deaths will occur, and how to reduce their incidence. Acknoivledgments: I thank I Colquhoun, G Hardy and other colleagues whose comments have improved this manuscript. References Allen R N & Newhook F J 1973 Chemotaxis of zoospores of Phytophtbora cinnamomi to ethanol in capillaries of soil pore dimensions. Transactions of the British Mycological Society 61:287-302. Davison E M1993 Preliminary report on the mass collapse site. Admiral Road, Gordon Block, Jarrahdale District. Unpublished manuscript, CALM Library. Davison E M & Shearer B L1989 Phytophthora spp. in indigenous forests in Australia. New Zealand Journal of Forestry Science 19:277-289. Davison E M & Tay F C S 1985 The effect of waterlogging on seedlings of Eucalyptus marginata. New Phylologist 101:743-753. Davison E M & Tay F C S 1986 The effect of aeration on colony diameter, sporangium production and zoospore germination of Phytophtlwra cinnamomi. New Phytologisl 103:735-744. Davison E M & Tay F C S 1987 The effect of waterlogging on infection of Eucalyptus marginata seedlings by Phytophthora cinnamomi. New Phy tologist 105:585-594. Davison E M & Tay F C S1991 Measurement of oxygen concentration in sub¬ surface flows. CALM RPP 27/91 unpublished report. Davison E M, Stukely M J C, Crane C E & Tay F C S1994 Invasion of phloem and xylem of woody stems and roots of Eucalyfitus marginata and Pinus radiata by Phytophthora cinnamomi. Phytopathology 84:335-340. Dell B & Wallace IM1981 Recovery of Phytophthora cinnamomi from naturally infected jarrah roots. Australasian Plant Pathology 10:1-2. Drew M C1992 Soil aeration and plant root metabolism. Soil Science 154:259- 268. GisiU,ZentmyerGA&KlureLJ 1980 ProductionofsporangiabyP/ii/top/»f/»or<7 cinnamomi and P. palmivora in soils at different matric potentials. Phytopathology 70:301-306. Greenwood E A N, Klein L, Beresford G D, Watson G D & Wright K D 1985 Evaporation from the understorey in the jarrah (Euca/yplws marginata Don ex sm.) forest, south-western Australia. Journal of Hydrology 80:337-349. Harding J H1949 Pathogenic aspects of dieback in the jarrah forest of Western Australia. Australian Forestry Conference. Pages from Forests Depart¬ ment file391/49. Unpublished manuscript, CALM Library. Kinal J, Shearer B L & Fairman RG 1993 Dispersal of Phytophthora cinnamomi through la teritic soil by laterally flowing subsurface water. Plant Disease 77:1085-1090. Podger F D 1967 Research project W.A. 4 - The cause of jarrah dieback. Progress report number 3 - Waterlogging as a possible cause. Unpub¬ lished manuscript, CALM Library. Podger F D 1968 Aetiology of jarrah dieback, a disease of dry sclerophyll Eucalyptus marginata 5m forests in Western Australia. University of Mel¬ bourne, M. Sc. Thesis. Podger F D 1972 Phytophthora cinnamomi, a cause of lethal disease in indig¬ enous plant communities in Western Australia. Phytopathology 62:972-981. Podger F D, Docpel R F & Zentmyer G A 1965 Association of Phytophthora cinnamomi with a disease of Eucalyptus marginata forest in Western Aus¬ tralia. Plant Disease Reporter 49:943-947. Shea S R, Shearer B &:Tippett J1982 Recovery of Phytopihthora cinnamomi Rands from vertical roots of jarrah (£«c10 years) sites, and immediate impact of canker on live cover compared with total cover, (numbers of species in plots given in brackets) Cover for Grevillea omitted due to small sample size. Number Susceptible Number assessed % difference in cover Dieback All taxa 177 460 29 (191) Proteaceae 101 110 72 (47) Banksia 29 29 93 (8) Dryandra 15 15 79 (10) Grevillea 2 6 - Hakea 16 20 52 (11) Isopogon 12 12 72 (5) Petrophile 9 9 69 (6) Canker Al! taxa 273 436 14 (230) Proteaceae 120 139 25 (60) Banksia 27 29 17 (12) Dryandra 15 17 33 (9) Grevillea 10 11 26 (3) Hakea 29 32 31 (16) Isopogon 9 10 28 (4) Petrophile 13 13 19 (6) Impact of canker fungi In recent years, a new fungal threat has emerged. Several aerially-dispersed, canker-causing fungi have been found in a taxonomically diverse group of native plants from many plant communities in south-western Australia. These in¬ clude a number of taxa classified as vulnerable or endan¬ gered (Murray ct al. 1994, Shearer 1994). The cankers, includ¬ ing species of Botn/ospliaeria and Diplodina, have caused extensive damage to large stands of vegetation in south- coastal areasof Western Australia (Shearer &Fairman 1991a, Wills 1991, Bathgate et al. 1994, Khangura ct al 1994, Murray et al 1994, Shearer 1994), particularly since February 1991 (Wills 1991). It appears likely that unusual weather condi¬ tions, comprising 6 months of serious rainfall deficiency up until May that year, and a heat-wave lasting four days and reaching 47°C, contributed to the rapid growth of the can¬ kers observed since that time in native plant communities. Extensive plant death has also been recorded as a result of periods of drought (Hnatiuk & Hopkins 1980), but the symp¬ toms associated with the drought deaths (A ] M Hopkins, pers. comm.) are distinct from those of canker impact. Surveys on the Southern Sandplain (Wills 1991) and in the south-west (Murray et al 1994) reveal that 59% of species assessed from a range of families were affected by canker fungi (Table 1). About 86% of Proteaceae were damaged and frequently killed by canker fungi (Table 2), although the level of damage sustained by different species was extremely variable. However, some species suffered severe impact with large stands being destroyed e.g. Bauksia coccinea and B. baxteri. These species are restricted to south coastal areas of Western Australia and are both highly susceptible to dam¬ age by canker fungi. As a result, the commercial picking of infloresences from wild populations of these species has now been banned (see Wills & Robinson 1994). Impact of Armillaria luteobiibalina Armillaria luteobubalina has a broad host range and is widespread in jarrah, karri, tuart and wandoo forest as well as woodlands and shrublands throughout the south-west of Australia (Shearer 1994). For example, in coastal dune shrubland communities 307 plant species were recorded in the sites assessed, and 112 of these were hosts to A. luteobubalina. (Shearer ct al. 1994). Susceptible species were mainly from the Proteaceae, Myrtaceae, Epacridaceae, Papilionaceae and Mimosaceae; the species killed included the geographically restricted Callitris preissii (itself forming a rare community), and rare and endangered Bauksia brozvnii and B. occidentalis formosa (Shearer cf al. 1994). Ecological consequences Changes in community structure following infestation are inevitable because virtually all species susceptible to P. cinnamomi and the majority of species susceptible to canker fungi and A. luteobubalina are woody perennials, while many field resistant species are herbaceous perenni¬ als. While even highly susceptible species are generally not eradicated with the initial invasion of the fungi, the abun¬ dance of susceptible species can be greatly reduced. At sites with a longer exposure to P. cinnamomi^ susceptible species may eventually be eliminated (Wills 1993); the same out¬ come may result from the activity of the other two diseases. Evidence for some regeneration of susceptible species at long-infected sites has been found (Weste & Ashton 1994), but it appears unlikely that susceptible species would return to previous levels of abundance, and changes in isolated remnants lacking extant seed sources will probably be irre¬ versible (see Keighery et at. 1994) without intervention. 128 Journal of the Royal Society of Western Australia, 77 (4), December 1994 In the case of P. cintiamoini, the fungus causes not only the decline in species richness of susceptible species at a site, but also a change in plant community structure and biomass as field resistant species, especially herbaceous perennials, be¬ come more abundant. These changes may translate into indirect losses in community productivity due to changes in plant biomass and degrade the capacity of infested sites to support dependant biota. Changes in habitat due to the alteration of community structure and composition may impact plants not affected by the disease. For example, most species of the Stylidiaceae appear to be field resistant to P. cinnamomi. However, one species {Sh/lidium scandens) is common in healthy sites in the Stirling Range National Park but absent in adjacent areas with a high disease impact. It seems likely that this species, which grows in dense understorey below stands of Eucalyp¬ tus marginataf may disappear as the structure of the stand is opened up through the loss of shading canopy and the death of susceptible understorey species. Other species with spe¬ cific habitat requirements may suffer similar indirect effects of the pathogen or may benefit from such changes, For example, introduced annual species invade more readily after disturbances (Hobbs & Atkins 1988), and introduced annuals have been shown to increase in abundance follow¬ ing the removal of the canopy of dense native plant commu¬ nities (Hobbs & Atkins 1991). This may be particularly important given that most annual species may be field resistant to P. cinnamomi (Podger& Brown 1989; Wills 1993), a factor that may enhance the invasibility of sites infested with the pathogen. Furthermore, if annuals become abun¬ dant, regeneration by native perennials is likely to be se¬ verely inhibited (Hobbs & Atkins 1991). Changes in the availability of resources and in habitat due to the alteration of community structure and composition may affect associated groups of animals e.g. pollinators, grazers (Wills 1993, Wills & Kinnear 1993, Laidlaw & Wilson 1994, Newell 1994, Wilsonc/ al. 1994) and soil biota (Malajezuk 1979, Malajezuk & Pearce 1994). Table 3 Summary of 494 out of 1655 Declared Rare Flora and Priority taxa ranked for their susceptibility to dieback and/or canker. Susceptible Total Rated % Total priority Priority Species Dieback 307 494 62^ 1655 Proteaceae 205 213 96 213 Myrtaceae 1 91 1 334 Epacridaceae X. 68 _b 68 Papilionaceae 24 26 92 92 Priority Species Canker 322 353 9P 1655 Proteaceae 213 213 100 213 Myrtaceae 8 12 75 334 Epacridaceae (68)'’ 68 68 Papilionaceae 25 26 96 92 "Biased sample due to ratability ^ Impact variable - all species currently rated as susceptible The disturbance of pollinators can have serious implica¬ tions for plant communities. For example, almost all wind- pollinated .species appear to be unaffected by diseases caused by P. cinnamomi or canker fungi. In contrast, the majority of vertebrate-pollinated species assessed are affected by these pathogens, in part a reflection of the prominence of verte¬ brate pollination in the Proteaceae. Pollinators reliant on susceptible plant species as key nectar sources (e.^. Banksia) may become rare or locally extinct in old-infested areas. Reduction in the population size of pollinators could affect the stability and viability of breeding populations and may result in the local extinction of the animal species at a site. This may be compounded by physical changes to the habitat, especially in cases where monocotyledons such as sedges colonize the spaces created by the loss of susceptible plants. In addition, it is possible that a reduction in the number of pollinators could affect the reproductive success of surviv¬ ing plants, further contributing to a decline in community structure and ecosystem viability. Threatened taxa While emphasis in the above discussion has been on the keystone species, there are rare and vulnerable species that are also at risk (Table 3). Notably, when all rare and vulner¬ able species from the south coast considered most at risk from dieback and/or canker are ranked, all but one of the species considered under greatest threat from plant diseases are proteaceous (Table 4). For example, the only known population of Dryandra sp. (Kamballup) is infected with canker (D L Murray & R T Wills unpublishcdX and all known populations of fjroztvin are infected with P. cinnamomi. Needless to say, other species from other families are also at considerable risk (c.^>. .see Lemson 1994). Furthermore, these diseases threaten the survival of animals, as exemplified by the impact of Armillaria that kills broom bush {Choretrum glomeratiim), the only food plant of the larvae of a rare species of butterfly, Ogyris otames (Wills & Kinnear 1993). Table 4 Taxa from the south coast considered most at risk from dieback and / or canker. Species Conservation Status^ Susceptibility Dieback Canker Lambertia orbifoUa E High High Andcrsoiiia sp.^ E High Moderate Banksia Imnvnii E High Moderate Banksia verticillata E High Moderate Adenanthos linearis 2 High Moderate Banksia occidentalis suhsp. formosa 2 High Moderate Dryandra sp.*’ E High High Isopogon uncinatus E High High Lambertia echinata subsp. echinata E High High Lambertia fairallii E High High Isopogon alcicornis 2 High High ^ Two Peoples Bay (GJK 8229) ^ Kamballup (M Pieroni 20.9.88) " E=endangered; 2=CALM priority 2 taxon; few poorly known populations on conservation lands. Journal of the Royal Society of Western Australia, 77 (4), December 1994 Disease amelioration Currently, the most practical management technique for the control of P. cinnamovii in native plant communities is foliar application of the fungicide phosphonate, ''phospho¬ rous acid" (Shearer & Fairman 1991b, Komorek et al. 1994, Hardy et ai 1994). Field trials in various areas in the south¬ west on plant communities already infested with P. cinnamonii have shown that one application of phosphonate gives excellent control of the disease over several years (see review article by Hardy et al. 1994). While canker fungi are not a major problem in south¬ western Australia, they are distributed throughout this re¬ gion and have the potential to cause very serious damage. Fire is the most practical management tool for the regenera¬ tion of native plant communities after infestation by canker. Acknowledgements: We thank G Friend, T Start, S Hopper and C Wills for their comments on earlier drafts of this paper. References Bathgate J, Shearer B L & Rohl L 1994 Cryptodiaportlie sp: a new canker pathogen threatening Banksia coccinea. In: Handbook of the Symposium on Plant Diseases in Ecosystems: Threats and Impacts in South-Western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 19. Diels L 1906DiePlarLzenwelt vonWest-AustraliensudlichdesWenderkreises. Vegn Erde 7, Leipzig. Froend R H 1987 Investigations into species richness patterns in the Northern Sandplain region of Western Australia. University of Western Australia, PhD Thesis. Gardner C A 1944 The vegetation of Western Au.stralia with particular reference to the climate and soils. Journal of the Royal Society of Western Australia 28:11-87. George A S, Hopkins A J M & Marchant N G 1979 The heathlands of Western Australia. In: Ecosystems of the World 9A (ed R L Specht) Elsevier Scientific Publ Co, Amsterdam, 211-230. Griffin E A, Hopkirts A ] M & Hnatiuk R J 1983 Regional variation in mediterranean-type vegetation at Eneabba, south Western Australia. Vegetatio 52:103-27. Griffin E A, Hopper S D & Hopkins A] M 1990 Flora. In: Nature Conservation, Landscape and Recreation Values of the Lesueur Area (eds A A Burbidge, S D Hopper & S van Leeuwen) Bulletin 424, Environmental Protection Authority, Perth, 39-69. Hardy G E St J, O'Brien P A & Shearer B L 1994 Control options of plant pathogens in native plant communities in south-western Australia. Jour¬ nal of the Royal Society of Western Australia 77:169-177. Hill T C J 1990 Dieback disease and Phytophthora species in the Northern Kwongan. In: Nature Coaservation, Landscape and Recreation Values- of the Lesueur Area (eds A A Burbidge, S D Hopper & S van Leeuwen) Bulletin 424, Environmental Protection Authority, Perth, 89-97. HnatiukRJ & Hopkins AJ M 1980 Western Au.stralian species-rich kwongan (sclerophyllous shrubland) affected by drought. Australian Journal of Botany 28:573-585. Hobbs R J & Atkins L 1988 Effects of disturbance and nutrient addition on native and introduced annuals in plant communities in the Western Australian wheatbelt. Australian Journal of Ecology 13:171-179. Hobbs R J & Atkins L 1991 Interaction between annual and perennial vegetation components in the Western Australian wheatbelt. Journal of Vegetation Science 2:643-654. Hooker, J D 1859 On the Flora of Australia, its Origins, Affinities and Distribution. Lovell Reeve, London. Hopkins A J M, Keighery G J & Marchant N G 1983 Species-rich uplands of south-western Australia. Proceedings of the Ecological Society of Aus¬ tralia 12:15-26. Hopkins, A J M & Griffin E A 1984 Floristic patterns. In: Kwongan: Plant Life o t 6Sandplain(edsJSPate&JSBeard)UniversityofWesternAustralia Press, Nedlands, 69-83. Hopper S D 1979 Biogeographicul aspects of speciation in the southwest Australian flora. Annual Review of Ecology and Systematics 10; 399-422. Keighery G J, Coates D J & Gibson N 1994 Future ecosystems • ecological balance. Journal of the Royal Society of Western Australia 77:179-182. Kennedy J & Weste G 1986 Vegetation changes associated with invasion by Phytophthora cinuamomi on monitored sites in the Grampians, Western Victoria. Australian Journal of Botany 34:251-279. Khangura R, Hardy G E St J & Wills R T 1994 Fungi associated with cankers in Western Australian plant communities. In: Handbook of the Sympo¬ sium on Plant Diseases in Ecosystems: Threats and Impacts in South- Western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 26. Komorek B, Shearer B L, Smith B & Fairman R 1994 Phosphonate offers a practical method for the control of Phytophlhora cinnamomi in native plant communities. In: Handbook of the Symposium on Plant Diseases in Ecosystems: Threats and Impacts in South-Western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society' of Australia. Perth, 28. LaidlawWS&Wilson BA 1994Small mammal abundancesinheathland with dieback disease in south-eastern Australia. In; Handbook of the Sympo¬ sium on Plant Diseases in Ecosystems: Threats and Impacts in South- Western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 29. Lamont B B, Hopkins A J M «/% - s _ - * V !> -w'i ■i/^ .? t - L..i ' • »■•JWtAj. f'.-itt4^H||P|P||tt - ik-^' ' -■■ W t.\..^ ' ': " '■ ■ .7 -VI pm ► 4ir, ..sjwer i»y> f ^ >f ’ .‘Ss^^Sf*, t • ^rc ^ * -a)' ^ j-* • ■* r 1 vi- > • <' iv < •' ■ - ^ ?S. ! * V-^ V ■ ,*-/'’i ;i--^ eK-Xl ***^ * • f ^ --W ' ^ I* * • . /« -r'-t' "fc ifc 4 j, ^ r ■ •*->■■ ’ * H V, --■*• . *"y ■ ' ■*’" , -'■’.* F,''^ - K " «V - • ■' S' •». ^ _;,i V ^ B V'*"'-': T *" ■ %*' 'ft WAcSi«»W- %ftv-^tk!t^IMH: ' -*' * .’■a.--.^ ,j|^ «... tMkf. V:Je' ^ ^ _ V*'.s^. 4P’ 2'ifr*^^!?'-.-i'-''. e>s..,,«1l'0tt ** * -tr^Awv^Air '- C*wi • f' . i^f •« ^ -1’^. ^ «* ik>* ‘ji' .' 4 ;-- * ^ * V ♦ U ^■' Vv <1*. ' ?>(*■ , f^, i _ - -c •- ^i. *-.. I ', f » i ■»'* t » s 4 C ^ ■> * ., * Vi' ^’3 \< * s * .«\ * 4# r' •V’'- ' • < • *■ -A -.. vT^- ■*« -' ' R rr • fc •9 - -t' *•• I .y-' » ' r- •;< •rA' , iC-. S m ■ " ^ .»S . A* *■-**■ . /* a < • #^ • ■ ■ .aS|t' w., ^ * Ia*-'' •'* »rf ttv.'>fir ir ^ ' H; ;; .ft,.. .^^&. • ^ .f - ' y v^yafe -#". ^ ' , 4 .V • ^ ‘ S i V* i Sc^fe^ r mi^*~ ■• ■ " ■ r ■* • -■ .- '-i = eV.Ki-^ ■V» - ; . '.' Ob ^,j'ar’i^ -', ' ‘ - ’• < . *• V ■■ - 1 1 K. ’■•■•••■ - W** » i«. -T * , W* ': » . *, \ A?| •tV^‘ • ' •v . , . *. 4-^ «„ , » ',' . -4, , 4. t V 'A-'-i.] f. I K if:--. -SlEr. - ♦ * Vt-5*l ■■' . V •!»--■: »- V 9f' K.J# •:- »s.4d55S?H^ ^ tS»“ - ' g <5. »'*• »• “ ■.-' i^r- ■ 5V* “ -Ei * rc ;■■ * ' r liipt# ^ >P t * 4^' V/ aifc^ V ■•■<*. K ■Bdk-.lb'. ' ■' Journal of the Royal Society of Western Australia, 77:133-137,1994 Smut and root rots on native rushes (Restionaceae) and sedges (Cyperaceae) K A Websdane’, I M Sieler’, K Sivasithamparam’ & K W Dixon^ ^Soil Science and Plant Nutrition, The University of Western Australia, Nedlands WA 6009 ^Kings Park and Botanic Garden, West Perth WA 6005 Abstract Native rushes (Restionaceae) and sedges (Cyperaceae) are widespread and common components of the vegetation of south-west Western Australia. In phytopathological terms many taxa within these families are susceptible to indigenous smut diseases, which affect culms and reproductive organs, yet are apparently resistant to habitat invasions by Phytophthora species. Smut disease has been found in a quarter of the 113 species of Restionaceae, representing 11 out of the 19 genera for the State, and in 17 species and 9 genera of Cyperaceae. For many species, smut diseases result in total loss of viable seed production. Culm smut has been recorded in only one species, and in this case infection results in retardation in growth and development of culms and reduced seed output. Conversely, introduced Phytophthora species have limited impact on survival and reproduction of native Restionaceae. Controlled inoculation of P. cinnamomi shows that disease symptoms are confined to localised regions on the roots, with a limited degree of cross infectivity between roots. Infected roots of test species produce abundant, healthy lateral roots above the lesions and symptoms in shoots are not apparent. Theuseof Restionaceaeasbarrierplantings for containing infection tosites within a habitat or protecting sites of rare taxa which are susceptible to P. cinnamomi is presently being investigated. An understanding of the diseases of rushes and sedges is important for the management of these plants in natural communities and in horticultural enterprises. Introduction Until recently, the effect of disease on Restionaceae has been limited to taxonomic description of species of smut (Ustilaginales) on host plants (McAlpine 1910) and the in¬ crease in rush and sedge abundance in wild sites affected by Phytophthora (Wills 1993). Many smuts have been described on the closely related Cyperaceae and the impact of P. cinnamomi on sedges has been studied in eastern Australia (Weste 1986; Phillips & Weste 1984; Cahill et al. 1989). Southern-hemisphere rushes and sedges are widespread and common components of the flora of south-west Western Australia. They inhabit a wide range of the northern sand plains, coastal wetlands and to a lesser extent the jarrah forest. Most species are long-lived perennials, with culms initiated from an underground rhizome in late autumn to early winter. Most Restionaceae are dioecious, with a few monoecious and hermaphroditic species, while sedges are monoecious. All species of both families flower from early winter to early summer, are wind pollinated, with most species producing very few viable seeds which are short lived and germinate in response to soil disturbance (Meney & Dixon 1988, Meney et a!. 1994). Taxa can be classified into two broad groups, seeders and resprouters, depending on their reproductive and vegetative biology, and their re¬ sponse to disturbances, such as fire. Their life strategies occur along a continuum from fire sensitive seeders, recruit¬ ing from seed after fire, to resprouters which regenerate Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 from heat resistant or deeply buried perennating buds lo¬ cated on the rhizome (Pate et al. 1990). An understanding of the impact of Phytophthora and smuts on native rushes and sedges is critical, as many species are harvested extensively from natural populations for use in the floriculture industry. Many .species are also important in rehabilitation of mining areas due to the sand-binding nature of their roots (Pate & Dixon 1994) and dominance in pre-mined vegetation (K Meney pers. comm.). Impact of smut on native rushes and sedges McAlpine (1910) described smuts on nine sedge species and two rush species from Eastern Australia. He wrote of the Western Australian smuts, "Only those species are known which attack cultivated crops and those occurring on the native flora are yet to be discovered". In the last seven years, 43 rush and sedge species in 20 genera have been recorded as smut hosts in Western Australia (Table 1). These smuts are as widespread as their hosts and represent the most debili¬ tating pathogens of rushes and sedges. Thus, smut impacts are likely to play a significant role in rehabilitation and conservation of these families in the future. Taxonomy The taxonomy of smuts is based on spore morphology, mode of spore germination and specialisation at host genus level. There are about 43 new species among these smuts and their taxonomy is currently under investigation. With the exception of one smut, all sporulate in inflorescences and have affinities to Tolyposporium. One species, T. restionum, has recently been described on a native rush Alexgeorgea 133 Journal of the Royal Society of Western Australia, 77 (4), December 1994 nitens (Nees) L Johnson & B Briggs (Websdane et al 1994). The exception to the inflorescence smuts is the culm smut Ustilago lyginiae (Websdane et al 1993) on Lyginia barbata (Labill) R Br. This is the first record of a culm smut on Restionaceae. Table 1 Recorded hosts of smuts within Restionaceae and Cyperacea in Western Australia Restionaceae Cyperaceae Genus Species Genus Species ^Alexgeorgea A. nitens A. subterranea *Carex ^Caustis C. fascicularis C. petandra ^Anarthria ^Desmodadus A. laevis D. biformis D. elongaius D. flexuosus Kyathochaeta ^Evandra C. avenacea C. clandestina E. aristata ^Harperia H. confertospicatus Hsolepsis Isolepsis sp. ^Hypolaena ^Lepidobolus H.fastigiata H. macrolepala L chaetocephalus L deserti ^Lepidosperma L. angustatum L. effusum L. gladiatum L. gracile L Longitudinale ^Leptocarpus L. preissianus L aristatus 1. ceramophilus L elegans L. scflnosis ^Mesomalaena ^Shoenus M. gracileps M. pseudostygia M. stygia S. laevigatus Schoenus sp. ^Lepyrodia L. macra ^Tricostularia T. neesii ^■^Lyginia ^Pseudoloxocarya ^Restio L. barbata P. magna R. chaunocoleus R. isomorphus R. leptocarpoides R. microcodon R. sinuosa R. sphacelata R. stenandra Totals 26 17 entity of smuts associated with host taxa. jlyposporium alexgeorgii. entatively identified as Fflrysifl species. i? entatively identified as an Anthracoidea species by R Shivas nidentified mally tufted inflorescences (Fig 1-3) become swollen when smutted (Fig 1 -4) and, according to Nees, resembled inflores¬ cences of female Restio species. Nees called the host species R. jiiteiis and mistook the smut for a rust which he named Uredo restionum. Johnson & Briggs (1896) resolved this misidentification and clarified the taxonomic identity of the host. The identity of the pathogen has also been clarified recently as T. restionum (Websdane et al 1994). In certain other host species examined, the only evidence of smut infection was a slight swelling of the host inflores¬ cence and in many sedges smutted spikelets cannot be distinguished from those which are healthy. The culm smut recorded on Lyginia barbata is easily recog¬ nised as a raised, brown, crust-like structure, or peridium, derived from host epidermal cells, which encloses the pow¬ dery spore mass (Fig 1-6). These spores have high viability, germinate readily and survive for at least one year and possibly longer. Spores of inflorescence smuts on the other hand appear to have either a long period of dormancy or a short period of viability and spores of most of these smuts have not been successfully germinated. The development of the culm smut sorus occurs during culm growth and well before inflorescence initiation. The sorus prevents further development of culms eventually causing dieback of culm apices. Thus all infected culms are sterile (Fig 1-5; Websdane et al 1993). Disease impact Smuts often fully replace the reproductive structures of host plants and render the plants sterile. In many seeder species, smut infection causes total loss of reproductive capacity in infected plants. Field observations on the resprouter species A. nitens indicate that smut infection is systemic via the ramifying rhizome system. This was shown by tracking rhizomes from smutted parent plants to under¬ ground buds and enclosing developing buds in spore-proof glassine bags. At inflorescence maturity all new plantlets originating from smut infected parent plants were also in¬ fected, indicating systemacy (K Websdane, datij). Disease incidence ymptoms The symptoms of inflorescence smuts are not always bvious/as the spore mass or sorus is general^ “j le inflorescence. In Restionaceae, smut infections s massive sporulation by the fungus in the ovaries (Fig W nd anther sLs of inflorescences. Infection results in partial 3 total loss of seed productive capacity In the monoecious ?yperaceae, the sorus surrounds developing anthers and tamens rendering them sterile. Transvestism is an interesHng abnormality observed in everal dioecious rush species affected by smuts. For exam- ,le, when smutted the normally ™ '’"•'S' r‘‘7 etr. in male plan.s iXT raid" Nee, IIMh,. The nor- Smuts have been found on hosts from isolated, undis¬ turbed habitats. For example, a population of Lepidobolus deserti Gilg was found to have less than 1% smutting at the Queen Victoria Springs on the edge of the Great Victorian Desert. It appears however that disturbance exacerbates the disease incidence as 20 to 50% smutting has been recorded in populations experiencing frequent fires, mining activities or road works (K Websdane, unpublished data). Implications of smut disease for conservation and restora¬ tion Smut infection is likely to be a problem in post-mining situations where successful restoration of species of rushes and sedges is dependent on creating self-sustaining populations of plants producing seed and healthy seedlings. For example, smutted plants of Lepidobolus chaetocephalus Nees, a seeder species, experience up to 100% loss of seed production in sites adjacent to mining areas in northern 134 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Figure 1. 1. Cross-section of female R.micrPCorff)«spikelet showing the production of smut sori in ovaries, preventing seed production. S=sorus 0=ovary wall (bar=lmm). 2. Smutted spikelets (SS; left) and uninfected spikelets (US; right) of a male R. microcodou plant. Smutted male spikelets appear identical to healthy female spikelets (bar-1 cm). 3. Uninfected spikelets (US) of a male 4. jiiVcms plant (bar=0.5cm). 4. Smutted spikelets (SS) of a male A. niteiis plant thought by Nees (1846) to resemble female spikelets of Restio species (bar=0.5cm). 5. Whole plant of L barbata infected by U. lygiuiae (culm smut) showing sori (arrows) (bar=5cm). 6. Close-up of culm smut sori along a culm of L barbata. P-peridium. S=scale leaf, (bar^lcm). 7. Roots of Loxocarya maf^na root rot symptoms. Healthy lateral roots (arrow) are initiated above water soaked lesions (double arrow) (bar=2cm). 8. The difference in root mass, at the base of the pot, between control and inoculated plants of Loxocarya mayna. Kwongan regions. This has serious implications for the survival of this species in rehabilitation programs. This host relies on current season seed production for seedling recruit¬ ment after fire, rather than soil seed banks (Meney 1993). Thus, a high incidenceof smut infection and fire can result in death of parent plants and dramatically reduced seedling recruitment. These factors, in addition to high levels of seedling mortality recorded for this species (Meney ef al. 1994), may result in localised extinction of this species (K Websdane, unpublished data). Of the 17 rush and sedge species harvested from the wild for the cut-flower industry, half have been recorded as hosts for smuts. In a number of these species infection of inflores¬ cences is diagnosed only after careful examination. This means that smutted inflorescences could unknowingly be harvested along with those which are healthy. The disease is therefore likely to have an impact on the sustainability of bush harvesting and on product quality particularly for export markets. It is critical for pickers to be aware of the disease in order to recognise and avoid smutted plants and populations and prevent the export of diseased material around the world. 135 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Impact of Phytophthora cinnamomi on native rushes and sedges Since the invasion of P. cimiaftwtni into the sclerophyll forests of Australia there has been a marked reduction in the diversity of understorey species present in infested sites (Weste 1986, Wills 1993). Members of the Epacridaceae and Proteaceae which are highly susceptible to the pathogen have been replaced by monocotyledons, especially members of the Restionaceae and Cyperaceae. Within the Stirling Range National Park a greater number of Restionaceae spe¬ cies were recorded in old-infested sites than healthy sites and their percentage cover increased significantly from four percent in healthy sites to ten percent in old-infested sites, with a similar increase in the abundance and frequency of Cyperaceae in long term, diseased sites (Wills 1993). The dominance of these species in diseased areas has been attrib¬ uted to their field resistance to Phytophthora root rots. Previous research showed P. cinnamomi to infect the roots of three species of sedges from the Brisbane Ranges in Victoria (Phillips & Weste 1984, Cahill et al. 1989), however only field observations have been made on the field resistant nature of the Restionaceae. The resistance of tw'O species of Restionaceae to increasing levels of P. cinnamomi inoculum was studied under glasshouse conditions. Clonally propa¬ gated plants of Loxocarya magna Meney & Dixon (ined) and Restio amblycoleus F Muell were inoculated with P. cinnamom i. Examination of the roots eight weeks later showed the development of water-soaked lesions immediately behind the root tip but new lateral roots formed above these lesions and limited cross infection was apparent (Fig 1-7; Sieler et al 1993). Inoculated plants of L magna produced a greater number of thick, depth-seeking roots and fewer bifurcated laterals than control plants (Fig 1-8). There were no symp¬ toms in the above-ground portion of the plants except for some yellowing of juvenile leaves and culm tip die-back in some treatments. Production of new reproductive culms was high for all inoculated plants, with a greater number of new culms being produced at high levels of inoculum. The ability of these plants to produce new roots and reproductive culms once infected has important implications for the reha¬ bilitation of affected sites and the management and recovery of rare species of rushes and sedges. At least 11 species of rare and endangered or priority rushes and sedges are located in areas infested with p cinnamomi. The potential benefits of field resistance in rushes and sedges is considerable including reinforcement and reintroduction into disease affected areas. In addition, species such as R. ustulatus (F Muell ex Ewart & Sharman) L which is both a priority species and recorded as being harvested, may be used in the reclaimation of sites or high risk horticultural sites as an alternative to bush picking. The use of rushes and sedges to act as biological barriers to reduce the impact and rate of spread of P. ciwiamomi is currently being investigated. Phillips & Weste (1989) showed zoospores to be produced from infected roots of the sedge Cahnia radula but the potential of infected roots to act as a future source of inoculum is not known. In a study of the barrier potential of Restionaceae to P. cinnamomi, adult plants of R. amblycoleus were placed into the middle of pots divided into three portions with Mira cloth (Calbiochem. Corp, La Jolla, USA). Bfln/rsw seedlings were planted on either side of therushes and one set of plants inoculated with P. cinnamo77u. In comparison to control pots, banksias in pots containing rushes showed a mark reduction in disease presence and delayed movement of inoculum. Passage of P. cinnamomi through the soil occupied by roots of a rush has been slow, and studies are continuing to determine the effectiveness of the Restionaceae as a biological barrier to the movement of the pathogen through soil. Conclusions Although there is a great diversity of flora in Western Australia, phytopathological studies have concentrated on those components which are most visibly affected by patho¬ gens. In contrast to the timber and highly floriferous species, only limited work on disease impact and interactions has been conducted on the Cyperaceae and Restionaceae. Their importance in floriculture and rehabilitation of mine sites, Phytophthora affected areas or high risk horticultural sites indicates a more economic role for these families in the future and the need for on going research. Acknowledgments: We would like to thankK Meney iorinvaluable assistance in identification and advice on Restionaceae and Cyperaceae. Work on the impact of smut on the Restionaceae and Cyperaceae was sponsored by Alcoa Australia, RGC Mineral Sands, Tiwest Joint Venture and the Minerals and Energy Research Institute of WA. Western Australian Wildflower Producers Association and The Horticultural Institute of Australia supported research on the impact of Phytophthora on commercially significant species. References Cahill D, Legge N, Grant B & Weste G1989 Cellular and histological changes induced by Phytophthora cinnamomi in a group of plant species ranging from fully susceptible to full resistant. Phytopathology 79:417-424. Johnson L A S & Briggs B G 1986 Alexgeorgea nitens, a new combination in Restionaceae. SC. Telopea 2:781-782. McAlpine D 1910 The Smuts of Australia. Department of Agriculture, Victoria. Meney K A 1993 Functional aspects of the growth, development and repro¬ duction of southern hemisphere rushes and sedges (Restionaceae). Uni¬ versity of Western Australia. Ph D Thesis. Meney K A & Dixon K W 1988 Phenology, reproductive biology and seed development in four rush and sedge species from Western Australia. Australian Journal of Botany 36:711-726 Meney K A, NeilssenG M & Dixon KW1994Seed bank patterns in Restionaceae and Epacridaceae after wildfire in kwongan in southwestern Australia. Journal of Vegetation Science 5:5-12. Nees von Esenbeck C G D1846 Restio nitens Nees. In: Plantae Preissianae sive Enumeratio plantarum quas in Aastralasia occidentali et meridionali- occidentali annis 1838-1841 collegit Ludovicus Preiss ed (Lehmann). Vol II, Hamburg, 59-60. Pale J S & Dixon K W 1994 Convergence and divergence in the South-West Australian flora in adaplalionsof roots to limited availability of water and nutrients, fire and heal stress. Proceedings of the Symposium on the Ecology and Conservation of Western Australian Biota. Australian Sys¬ tematic Botany Society, National Herbarium Sydney. In press. Pate J S, Froend R H, Bowen B J, Hansen A & Kuo J1990 Seedling growth and storage characteristics of seeder and resproutcr species of Mediterranean- type ecosystems of south-western Australia. Annals of Botany 65:585-601 • Phillips D & Weste G 1984 Field resistance In three native monocotyledon species that colonise indigenous sclerophyll forest after invasion by Phytophthora cinnamomi. Australian Journal of Botany 32:339-52. Sieler I M, Sivasithamparam K & Dixon K W 1993 Impact of Phytophthora cinnamomi on the roots of 'Resistant' Cyperaceae and Restionaceae. 6*^’ International Congress of Plant Patliology, National Research Council, Ontario, Canada. Abstract. Websdane K A, Sivasithamparam K, Dixon K W & Pate J S1993 Biology and description of Ush'/Jifo lyginiae sp. nov., a new culm smut on Lyginia barbarta (Restionaceae) in Western Australia. Mycotaxon 48:483-489. 136 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Websdane K A, Sivasilhamparam K, Dixon K W & Meney K A 1994 Tolyposporium restionum comb. nov. on Alexgeorgea species (Rcslionaceae) in Western Australia. Mycotaxon 51:471-477. Weste G 1986 Vegetation changes associated with invasion by Phylophlhora cinnamomi of defined plots In the Brisbane ranges, Victoria, 1975-1985. Australian Journal of Botany 34:633-644. Wills R T1993 The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18:145-159. 137 'r .’-i-T, -mi ■■ m . r" fe*^, U'-^'yV %. ^ ^ ^ -H ■c’t • . -V J-V • tc. • V « S' s ff ■ > ' . 3? n.V ■• 1- -«4i^^-;v. ; ■, •' V •. ■ wrf.. »;^ •- .^.- . :-1U'a^ ^'4 r «s > !t ' r. ' ts'm »'• n . K- C . • t. ■ ,,r • / V i‘ir. '‘^1^ »*- * S ' ’ ■ ' St fete ' tsMM*'-''’ ^ " .» ,/*! '-’ ,,iUi V";*- ~ *■• fvv>.>**''ipi ^ r~'' ••>--- ■^■- 5. ••• '■ i A -ai -a ■ *v^^i; « ;a.A%:, '- _ . ' i !-■ ,gf. '^- ^'fi' 'J»r,| >v- •• • ^>?.;.«^-i > • ■'s^t'* rl'L^ ' B'1*- ^--'''’ v.- t * “ '• ■ S*'^' 4y ■•" _ ’-rR.y *• * 1 ' ' .-. *, ' . V ■*‘- - - • ■' ^ .■- . or .-•* ■ ' -r .'•rai<»-3 , ^ ; # * •’ ♦ “ J?- -'A* •- - !■ r ■ ' **. M - V .-*-■ j. , -' % " ■ r. > . * ■ ' • ;> ♦ i ■•••■■,^^- '■ A V ■ 'V* V* ' ,_ -- ■• -^.-1 4 , . . * W''« ' * . -. • -* .-I- ' - ^ Lf , V‘ ^ ■' Wy--: ■ ^ - , f -1 ,. ’V. ^i». " • .* S - .•» • »•■» ' r* ,-.j 3u . i.,- -■-|*‘si.- V 1^ Vr: ■yi- * ■» C , ’‘.J • « •--^ ■-- - - - -:■ -« - .^v. ; - 1^*' a V ^ MJ . ► -lii .-,• r V- ’ ^ -.i-i^'J'/ -,^1. Journal of the Royal Society of Western Australia, 77: 139-143,1994 Impact of plant diseases on faunal communities B A Wilson’, G Newell’, W S Laidlaw’ & G Friend^ ^School of Biological & Chemical Science, Deakin University, Geelong VIC 3217 ^Wildlife Research Centre, Department of Conservation and Land Management, PO Box 51, Wanneroo WA 6065 Abstract Plant pathogens can have a major effect on vegetation floristics and structure. Effects include loss of plant species, decline in vegetation cover, and an increase in bare ground and the abundance of resistant plant species. These changes would be predicted to affect the faunal communities inhabiting infected habitats, but there have been few studies which examine the relationship between faunal abundance and composition and plant pathogens. Tliis paper considers the potential affects of plant pathogens on faunal communities and reviews recent work on the effects of Phytophthorn cirmamomi (cinnamon fungus) on mammals and inverte¬ brates. Analysis of a number of disturbance factors in hcathland and woodlands of south-eastern Australia have identified P. cinnamomi infection as being associated with low species richness, and low abundance of small mammals. Studies on populations of Antechiuus stuarfii (brown antechinus) in woodlands found that there were lower capture rates, and habitat utilisation was altered. The major contributing factor was alterations to vegetation structure, rather than food availability. In heathlands, species such as Rattus lutrcolus (swamp rat), Rattusfuscipes (bush rat) and Antechinus stuartii were found to be less abundant in areas infected with P. cinnammomi. Introduction The habitat and microhabitat preferences exhibited by animal species are determined by a number of basic require¬ ments such as the provision of adequate food resources, the presence of cover (protection from predators, suitable microclimate) and access to breeding, basking or roosting sites. Vegetation attributes make a major contribution to these habitatcomponents. For example, understorey vegeta¬ tion and litter provide important refuges and breeding areas for invertebrates. Omnivorous or herbivorous terrestrial lizards are dependent on vegetation as foraging sites, while some species require elevated sites for thermoregulation. Birds require vegetation for nest sites, some using tree cavi¬ ties, others shrub or dense ground vegetation, and most species also depend on vegetation for food such as nectar, seeds, fruit or invertebrates. Small terrestrial mammals de¬ pend on vegetation for cover, food and protection. Tree bark, litter and woody debris, harbouring a diverse range of invertebrates, also provide foraging sites for insectivores. Studies have provided evidence that small mammals exhibit microhabitat selection based on vegetation charac¬ teristics (Braithwaite & Gullan 1978, Braithwaite et al. 1978, Cockburn 1978,1981, Fox & Fox 1981). Analy.ses of factors found that structural attributes of vegetation were impor¬ tant indicators for habitat preferences for some species (Barnett 1978, Stewart 1979, Newsome & Catling 1979), while for others both floristic and structural characteristics were significant (Braithwaite & Gullan 1978, Braithwaite et ai 1978, Fox & Fox 1981). For example, a number of pseudomyine rodent species exhibit preferences for diverse Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16,1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 floristic communities, which are likely to be related to di¬ etary requirements (Braithwaite & Gullan 1978, Cockburn 1978, 1981, Fox & Fox 1981, 1984). Species such as the omnivorous bush rat (Rattus fuscipes) respond mainly to structural factors while the herbivorous swamp rat (Rattus lutreolus) has a requirement for a sedge food resource (Braithwaite et al 1978, Barnett et ai 1978). Plant pathogens can have a major effect on vegetation communities. They may decrease the fitness of individual plants, alter the size and genetic structure of individual populations and thus the structure and diversity of whole plant communities (Burdon 1991). There may also be a range of effects from pathogens depending on the intensity of the pathogen and the conditions present at the time of infection. The consequences for animal habitats will also vary accord¬ ing to the intensity of pathogen effects and the percentage of resistant plants present. In some cases, marked changes to habitat conditions may occur if the pathogen affects all the vegetation present, or if there is total mortality of susceptible species. Indeed, some pathogens lead to almost complete loss of plants present, from the understorey to the canopy. In other situations only understorey vegetation may be af¬ fected. Thus, animals that rely on different habitat compo¬ nents would be affected differentially. For example, arboreal marsupials (e.g. folivores, nectarivores) and many bird spe¬ cies are likely to suffer as the consequence of canopy dam¬ age. Terrestrial mammals and invertebrates would be seri¬ ously affected if all understorey vegetation is eliminated. Furthermore, some animal species are reliant on particular plant species for their food. If those plant species are highly susceptible to a pathogen, then this will have important consequences for animal food resources. In this paper we examine evidence and recent research on the impact of a particular plant pathogen, cinnamon fungus (Phytopihthora cinnattwmi) on fauna. We assess the affects of P. cinnamomi on vegetation together with consequential 139 Journal of the Royal Society of Western Australia, 77 (4), December 1994 effects on animal populations, communities, habitats and diet. The role of P. cinnamorni as a disturbance regime in native communities is examined, and the long-term conse¬ quences of infestation and resilience of these communities is assessed. Effects of P. cinnamorni on vegetation and animal habitats The fungal pathogen P. cinnamorni causes extensive "dieback" of Australian native vegetation. It is widespread and occurs in forests ranging from the jarrah forests of Western Australia to the stringybark and silvertop ash for¬ ests of Victoria, and the tropical rainforests of Queensland. It also occurs in woodlands and heathland communities c.^. the Grampians and Brisbane Ranges, in Victoria and in Western Australia (Newhook & Podge Ashton 1970, Weste & Taylor 1971, Weste 1974, Wills 1993). The pathogen has a wide host range although its pathogeLity To different hosts varies, and its growth and distribution are influ ygjj^ility (Weste & ^ /lyrics 1987, varies from site to site. Studies of m ec ed shown that P.cimwmomi has a major e e , . jj j. and up to 60% of the plant speaes presen f . natedLrinfection(Westel974,Dawsonri^ & Weste 1986, Wills 1993). In^^td L ^opuTa^^^^^ seedling regeneration, specie ^a^son et al. 1985, Kennedy density of species (Weste 19 , j^crease in the frequency & Weste 1986), but there is often ^^cre and cover of field resistant mon ^ j severely 1985, Weste 1986, Kenney has ^ affected areas, the ® ° g^ge-dominated understorey, from diverse sclerophyll p cinnamorni. In some due to the resistance of denuded of vegetation, situations large areas ha tKennedy & Weste 1986). 0 «e„ resuUing sever. The changes in to affect the fauna P. cinnamorni infection wou P potentially present (Table 1). Decrease marsupials, while simplifi- affect nesting birds and arbor sources of cation of the understorey v 8 mammals, birds and seed, nectar and pollen ava Uoney possum (Tarsipes insects. Indeed, species sue ^ gpedalized diet of pollen rostratus), which rely enhreiy proteaceous spe- and nectar mainly from P affected by P. cies (e.g. Banksias), are likeiy (Friend 1992 . tation and litter could Alterations to , fj^vertebrates. Thus, one may alter habitat conditions ^^r ^ compositionof inver- expect changes in th^abunda^ influence the diets of tebrate communities w gmall dasyurid marsupials, insectivorous animals such a ^^^ample, wererediscov- Dibblers(Pflhm/er/»■"“s"P'""/^^*j; the south coast of Western ered in 1967 at Cheyne Beach Australia (Morcombe l967)^ah j brateswhich inhabited the wild inceridiiu -- inder great threat from Table 1 Predicted effects on fauna due to the presence of P. cifinamomi (cinnamon fungus) in vegetation communities. Effects on vegetation Effects on fauna 1. Loss of susceptible plant species in the understorey, midstorey a) Direct loss of food sources e.g. seeds, nectar, pollen. b) Indirect loss of food sources e.g. invertebrates. 2. Decline in plant species' richness and diversity a) Loss of food for species that prefer floristically rich vegetation. b) Loss of seasonal food availability. 3. Decrease in plant cover, increase in bare ground, erosion. a) Loss of habitat for species dependant on thick ground cover. b) Increased predation risk. c) Changes to microclimate conditions. 4. Decrease in canopy cover a) Loss of food for arboreal species. b) Lo.ss of habitat for arboreal species. 5. Decrease in litter fall. a) Decline in litter invertebrates (dry conditions) b) Decline in invertebrate food sources for insectivores. 6. Post infection increase in frequency of field resistant plant species e.g. sedges. Increase in food for specialist herbivores Species such as the pseudomyine rodents have been shown to be dependent on floristically-diverse understorey (Cockburn 1978, Cockburn et al. 1981, Fox 8z Fox 1984) and are potentially endangered by the simplification of diverse sclerophyll communities resulting from P. cinnamorni infec¬ tion. By contrast, animals with relatively generalized diets that require dense, low vegetation for shelter may prefer habitat that has been infected by P. cinnamorni, if there has been a consequent increase in the cover of field resistant monocotyledons. In severely affected areas, decreases in plant cover could lead to loss of habitat or a decline in the carrying capacity for these animals. Effects of P. cinnamorni on animal populations Although there have been a substantial number of stud¬ ies on the effects of P. cinnamorni on vegetation (see review by Weste & Marks 1987), there has been little work investigating the less direct effects of the pathogen on faunal populations and communities. Two studies have compared invertebrate communities in healthy and infected jarrah forests in Western Australia. Postle ct al. (1986) found that infected jarrah forest had 48% less litterfall, and a standing biomass of leaf litter 8.4% of healthy forest, although these differences were not tested statistically. Numbers of soil and litter invertebrates were generally lower in diseased forest, but there were variations between seasons and between taxa for both the soil and litter components. Nichols & Burrows (1985) recorded lower Journal of the Royal Society of Western Australia, 77 (4), December 1994 numbers of invertebrate species and individuals in a dis¬ eased forest, as well as fewer trees and shrubs and lower mean litter cover values. Variations in abundance of inver¬ tebrate taxa were also observed between an uninfected forrest {i.e. n=l) and an infected (n=l) forest depending upon the habitat requirements ofeach taxon. For example, Dermaptera which require dense litter cover were only recorded in healthy forest, whereas the majority of ant species were found in both diseased and healthy forests, although the total abundance of ants was lower in diseased forest. The low number of sites in these studies makes statistically- teslable conclusions impossible. Reptiles and frogs have been surveyed in both healthy and diseased jarrah forest. Diseased forest supported lower numbers of species and lower abundances than healthy forest (Nichols & Bamford 1985). Some species however, were more abundant in diseased forest minor and Crx/ptoblepitarus pIa;^iocephalus), perhaps reflecting the in¬ creased insolation on elevated surfaces logs) which these species use for basking and foraging (Nichols & Bamford 1985; Wilson & Knowles 1988). Again, low site numbers precluded statistical testing of the data. There is some evidence that there are differences in the avifauna present in diseased and uninfected forests. Nichols & Watkins (1984) described a dieback-affected forest that had low bird species richness and abundance compared with healthy sites. At another dieback site, however, the bird density and species richness were comparable to those in healthy forest. Some species were recorded in higher densi¬ ties in diseased forest, while other species were absent, resulting in different bird species composition in the two forest types. In a study of small mammal communities in heathy woodland and heathlands, Wilson (1990) and Wilson et al. (1990) found the percentage of vegetation modified by P. cinnamomi to be a significant variable in explaining small mammal diversity and density. Further studies by Newell & Wilson (1993) in heathy communities of the Brisbane Ranges National Park, Victoria, found the abundance of AH/cd/hius stiiartii to be lower in P. cintuwwmi affected areas. Vegetation volume was .significantly lower between 0 and 60 cm, in diseased areas and the abundance of A. stuartii was signifi¬ cantly correlated with this variable. The change in structure in diseased areas was predominantly related to the loss of the austral grass-tree (Kanthorrhoeu ntistralis). Whether this loss affected A. stuartii through reduced cover, or altered food availability was unclear. Studies in coastal heathland at Anglesea, in southern Victoria, found that several small mammal species e.*^. R. lutreolus (swamp rat), R. fusdpes (bush rat), and A. stuartii (brown antechinus) were less abundant in diseased heathland than in healthy stands (Laidlaw & Wilson inipuhlisfied data). Mean species richness of small mammal communities was also lower at infected sites. The utilization of habitat by the dnsyurid marsupial A. stuartii in P. cinnamomi infected and non-infected areas have been investigated at several sites in the Brisbane Ranges, Victoria. A. stuartii was found to forage almost exclusively at ground level, and frequently used nest sites located at ground level in large X. australis plants (Newell 1994). The movement and home range of A. stuartii were investigated using trapping and radiotelemetry. Home ranges displayed a high degree of overlap with areas that were uninfected with P. cinnamomi and animals actively selected uninfected habitat, and avoided areas infected with P. cinnamomi (Newell 1994). Individuals occasionally crossed bare, long-term infected areas to enter other uninfected habitat. These re¬ sults suggest that A. stuartii relied heavily on vegetation cover. The effect of P. cinnamomi on the dietary items of A. stuartii was also investigated (Newell 1994). There was no relationship between invertebrate abundance, and the cap¬ ture rate of A. stuartii in infected or unifected areas. The above studies provide evidence that the modification of habitat due to the presence of P. cinnamomi can lead to declines in the overall abundance of fauna. There is evidence that these changes also result in reduction in species rich¬ ness, and/or diversity. Changes in the abundance of indi¬ vidual taxa, may depend on their habitat requirements, with some taxa increasing and others declining. Some of the studies outlined were limited in design, and mainly ad¬ dressed other environmental disturbances such as mining. There is evidence of altered utilization of habitat by A. stuartii due to the presence of P. cinnamomi. Further work needs to be undertaken to establish the relationships be¬ tween the modification of habitat components and changes in faunal communities due to the affects of P. cinnamomi. P. cinnamomi, disturbance and resilience Australian vegetation and fauna communities are well adapted to disturbance factors such as fire. In areas where fire frequency is high, communities often exhibit high de¬ grees of resilience (or more specifically, elasticity; sensu Waltman 1986) following disturbance. This resilience is likely to be a consequence of the evolution of adaptive features over long periods of exposure to disturbance re¬ gimes. The presence of P. cinnamomi may also be considered a component of a disturbance regime. However, there is evidence that P. cinnamomi has recently been introduced to Australia (Weste & Marks 1974,1987) and native plant and animal communities have not been exposed to the regime for a long period of (evolutionary) time. Therefore these com¬ munities are unlikely to have adapted to the presence of P. cinnamomi. Compared with the relatively ephemeral effects of fire, P. cinnamomi may have a severe and long-lasting impact on plant and animal communities. Infection generally results in removal of a wide range of susceptible species and leads to simplification of the understorey. Furthermore, the patho¬ gen is capable of remaining in the soil following infection, so may reinfect vegetation or infect new vegetation. Secondary plant succession after fire normally results in the re-estab- iishment of the original floristic community, following an initial floristic composition model (Noble & Slay ter 1981). There is evidence, however, that susceptible plant species rarely recolonise after P. cinnamomi infection (Weste & Marks 1987). The only evidence of recolonisation is where the austral grass-tree {Xanthorrhoea australis) has been recorded, at least 20 years after infection (Dawson et a!. 1985, Weste 1993). Given large alterations to vegetation species compo¬ sition after P. cinnamomi infestation, it is difficult to predict regeneration of the original vegetation community (Grubb & Hopkins 1986). 141 i Journal of the Royal Society of Western Australia, 77 (4), December 1994 There is often substantial recolonisation by small mam¬ mals after fire (Fox 1982; Fox &: McKay 1981, Newsome et al. 1975, Wilson et al. 1990). Recolonisation rates depend on factors such as the fire regime, the regeneration of vegetation and sources of recolonisers {i.e. unburnt patches). Some species enter the succession early (2-4 years) due to their preference for diverse young vegetation, while others may take up to ten years to return (Fox 1982, Wilson et al. 1990). Successional changes of vegetation communities following P. cintiarnomi infestation are yet to be established. The resultant community is likely to differ substantially from the original vegetation community. It is presently difficult to accurately determine animal succession patterns; however, one would expect a very slow recolonisation process result¬ ing in a different faunal community. Conclusions It is clear from the above review that P. cinnamomi has the potential to severely influence the abundance and composi¬ tion of many faunal commuities. These effects are largely indirect, resulting from changes in plant species richness and composition, and from alterations to the structural compo¬ nents of the habitat. Although data are limited, there is evidence that a broad range of taxa are affected including small mammals, reptiles, birds and invertebrates. The effect on individual taxa will depend on that species' requirements for food and shelter, and its reproductive and foraging strategies. It may be predicted that some generalist species which require relatively open habitats may be favoured, while more specialized species (especially those with re¬ stricted diets which inhabit dense species-rich shrublands) would markedly decline in the dieback-affected areas. Simi¬ lar principles have been shown to apply in assessing the impact of fire on small vertebrates (Friend 1993), and open the way for predictive modelling of the impact of plant diseases on faunal populations. In tandem with these broad approaches to develop a classification of tolerant and sensitive faunal species (as has been done for many plant species; Wilis 1993), there needs to be more detailed studies of movement patterns and resource utilization for the most sensitive species, particularly those which are regarded as rare or endangered. Although de¬ tailed habitat utilization work has been carried out on A. stuartii in relation to infection by P. ciunatitotm (Newell 1994), there has been little such work on other potentially sensitive species like T. rostratus, P. apicalis and the western pygniy possum {Ccrcartetus concinnus). Unique and threatened communities also need to be identified, and measures enforced to assure their long-term protection from plant diseases. Such work is now underway in Western Australia, where a GIS-based decision support system is being developed to monitor and manage P/iyfop/it/iorfl-sensitive taxa and communities, and experi¬ ments are being carried out to evaluate the efficacy of aerial application of phosphonate to control Phytophthora in native plant communities. It is only through application of such multidisciplinary studies that we can begin to understand the processes by which plant diseases influence the structure and composition of plant and animal communities. References Barnett J L, How R A & Humphreys VV F1978 The use of habitat components by small mammals in eastern Australia. Australian Journal of Ecology 3:277-285. Braithwaite R W & Gullan P K 1978 Habitat selection by small mammals in a Victorian heathland. Australian Journal of Ecology 3:109-27. Braithwaite R W, Cockbum A & Lt'e A K1978 Resource partitioning by small mammals in lowland heath communities of south-eastern Australia. Australian journal of Ecology 3:423-445. Burdon j J1991 Fungal pathogens as selective forces in plant populations and communities. Australian journal of Ecology 16:423-32. CiKkbum A1978Thedistribution of PscMdemy.s shortridgeii (Muridae: Rodentia) and its relevance to that of other heathland P.st’Mi/amys. Australian Wildlife Research 5:213-219. Cockbum A1981 Population regulation and dispersion of the smoky mouse, Pseudomyii fumciis I. Dietary determinants of microhabitat preference. Australian journal of Ecology 6:231-254. Cockbum A, Braithwaite R W & Lee A K 1981 The response of the heath rat, Pseudomys shortrid^ei, to pyric succes.sion; A temporally dynamic life- history strategy, journal of Animal Ecology 50:649-66. Dawson P, Weste G &c Ashton D 1985 Regeneration of vegetation in the Brisbane Ranges after fire and infestation by Phytophthora cinnamomi. Australian Journal of Botany 33:15-26. Fox B j 1982 Fire and mammalian .secondary succession in an Australian coastal heath. Ecology 63:1332-41. Fox B j & Fox M D 1981 A comparison of vegetation classifications as descriptors of small mammalhabilatpreference.In: Vegetation classifica¬ tion in the Australian region, (ed A N GilHson & D j Anderson)CSIRO and Australian National University Press, Canberra, 166-180. Fox B j & Fox M D1984 Small-mammal recolonization of open-forest follow¬ ing sand mining. Australian Journal of Ecology 9:241-52. Fox B J & McKay G M 1981 Small mammal responses to pyric successional changes in eucalypt forest. Australian journal of Ecology 3:29-41. Friend G R 1992 Possum in peril. Landscope 7: 22-27. Friend G R1993 Impact of fire on small vertebrates in malice woodlands and heathlands of temperate Australia: a review. Biological Conservation 65: 99-114. Grubb P J & Hopkias A j M1986 Resilience al the level of the plant community. In: Resilience in Mediterranean-type Ecosystems, (ed B Dell, A J M Hopkins & B B Lamont) Dr W Junk, Dordrecht, 21-37. Kennedy j & Weste G 1986 Vegetation changes associated with invasion by Phytophthora cinnamomi on monitored sites in Grampians, Western Victo¬ ria. Australian journal of Botany 34:251-279. Marks G C & Smith 1 W 1991 The Cinnamon Fungus in Victorian forests. Department of Conservation and Environment, Melbourne. Morcombe M K1967 llie rediscovery after 83 years of the Dibbler Antechinus apicalii (Marsupialia, Dasyuridae). Western Australian Naturalist 10: 103-111. Newell G R 1994 The effects of Phytophthora cinnamomi on the habitat utiliza¬ tion of Antechinus stuartii in a Victorian forest. Deakin University, PhD Thesis. Newell G R & Wilson B A 1993 The relationship between cinnamon fungus {Phytophthora cinnamomi) and the abundance of Auft’c/imu.s stuartii (Dasyuridae: Marsupialia) in the Brisbane Ranges, Victoria - a preliminary investigation. Wildlife Research 20: 251-259. Newhook F j & Podger F D 1972 The role of Phytophthora cinnamomi in Australian and New Zealand forests Annual Review of Phytopathology 10: 299-326. Newsome A E & Catling, P C 1979 Habitat preferences of vertebrates inhabilaling heathlands of coastal, montane and alpine regions of south¬ east Australia. In: Ecosystems of the World 9A. Heathlands and related shrublands (ed R I. Specht) Dr W Junk, Dovolrechl, 301-316. Newsome A, Mcllroy K & Calling P 1975 The effects of an extensive wildfire on populations of twenty ground vertebrates in south-east Australia. Proceedings of the Ecological Society of Australia 9:107-123. Nichols, O G & Bamford, M J1985 Reptile and frog utilization of rehabilitated bauxite minesiles and dieback-affected sites in Western Australia's jarrah Eucalyptus marginata forest. Biological Coaservation 34: 227-250. Nichols O G & Burrows R 1985 Recolonization of revegetated bauxite mine sites by predatory invertebrates. Forest Ecology and Management 10: 49-64. 142 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Nichols O G & Watkins D 1984 Bird utilization of rehabilitated bauxite minesites in Western Australia. Biological Conservation 30:109-132. Noble IR & Slatyer R 01981 The use of vital attributes to predict successional changes in plant communities subject to recurrent disturbance. Vegetatio 43: 5-21. Podger F D & Ashton D H1970 Ph\/tophthora cinttamonii in dying vegetation on the Brisbane Ranges, Victoria. Australian Forestry Research 4:33-36. Postle A C, Majer J D & Bell D T 1986 Soil and litter invertebrates and litter decomposition in jarrah {Biical\/ptus margifiata) forest affected by jarrah dieback fungus {Phxftophthora cinnamomi). Pedobiologia 29; 47-69. Stewart A P 1979 Trapping success in relation to trap placement with three species of small mammals, fuscipes, Antecimus swaitisonii and A. stuartii. Australian Wildlife Research 6:165-172. Waltman W E 1986 Resilience; concepts and measures. In: Resilience in Mediterranean-type EcosystemsfedsB Dell A J M Hopkins&B B Lamont) Dr W Junk Publishers, Dordrecht, 5-19. Weste G 1974 Phi/tophthora dtviamomi - the cause of severe disease in certain native communities in Victoria. Australian Journal of Botany 22:1-8. Weste G 1986 Vegetation changes associated with invasion by Phi/tophthora cinnamomi of defined plots in the Brisbane Ranges, Victoria, 1975-1985. Australian Journal of Botany 34: 633-648. Weste G1993 The cinnamon fungus. Is it a threat to Australian native plants? Victorian Naturalist 110; 78-84. Weste G & Marks G C 1974 The distribution of Phytophthora cinnamomi in Victoria. Transcripts of the British Mycological Society 63:559-572. Weste G & Marks G C 1987 The Biology of Phytophthora cinnamomi in Austral¬ ian Forests. Annual Review of Phytopathology 25:207-229. Weste G M & Taylor P 1971 The invasion of native forest by Phytophthora cinnamomi. 1 Brisbane Ranges. Australian Journal of Botany 19:281-294. Wills R T1993 Tlie ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18; 145-59. Wilson B A 1990 The effects of vegetation, fire and other disturbance factors on small mammal ecology and conserv'ation. Deakin University, PhD. Thesis. Wilson B A, Robertson D, Moloney D J, Newell G R & Laidlaw W S1990 Factors affecting small mammal di.stribution and abundance in the eastern Otway Ranges, Victoria. Proceedings of the Ecological Society of Australia 16: 379-396. Wilson S K & Knowles D G 1988 Australia's Reptiles-A Photographic Refer¬ ence to the Terrestrial Reptiles of Australia. Collins Publishers, Australia. 143 . ;;r; . ,v> ^ •< - “^‘--‘i^: -- ♦^TtS^f^-^-Trv?' iT« . . >-35« V. -.-vrr’i -^3 V - 'its ' w ‘ :- -A'^'-« ? ^ •. i •*>«»r - V.-', c^ ?w«t»iaiyy ♦ ■uiSirii- !kv V m , vV- ^-■H; *>;S^*: '■ ^~!-*^* Wi . .'' r.-.Ttv: ’•'.y'-; -x!; '■^‘■l;l<“i•/. - . -'AV.;- l$;j|&it^' ' ' ";"''' *- ■ V • ■ '^*‘Vr^’?, . ^ijiil-J-J|-,* >,I J^.-;,-^-,,^’’, • *’ fl? *-. ' )»li^ ^f'a.$'}i4l- fe :^' . ■•^ *‘a* -> r< -» f .^^^;,^^''-Vvc^^i •"' -V:^' x.ff^ jfe ,:#'■•'* ]? til ;‘ .-1 V:i if J/ .»*'. * •V ^ -»'-r ■■..r‘A*«>if» j»«C,ji'M'jK^^-. V ^ W ’ .irV l\^V# i '■'^ ''^'■ * • "* «■ ' *^.''.» ' I^'S'-'' A -.-: ^ ^ ^ ;« %,.• .__' , *.•'. "^^-ir'*’';^. ,. I r. ' .- : V-.f.^'f •• -i* \.. :'.^, . > V' • ^ JlS ?.T-|P»S . 4 jMlWf_ • ,3. t.- -f'.-,> AI ' -M '-S^' " pw;^, '■ '■ ’■■■ -■■-•<■ :->■' . Y '• '_ ■ ■ ^ /. . '-'3 . ► -.-^ b.fgS)h't . ; - ■ ~X»Cv *i»s'''‘’.-?v^. >N ' '.-';-rta - .vrrw.v . *^1 . ; • r^V\f S-. t -'•a' ."^.■•‘■7 «- i Journal of the Royal Society of Western Australia, 77: 145-149,1994 Disease and forest production in Western Australia with particular reference to the effects of Phytophthora cinnamomi D S Crombie & F J Bunny Department of Conservation and Land Management, Dwellingup WA 6213 Abstract The native forests of Western Australia are valuable for production of timber and water, and for conservation and recreation. Plant diseases affect all of these values. Timber production is reduced and the aesthetic experience of the forest changed when trees are killed or lose vigour. Conservation values are affected when species are eliminated locally or forest structures are changed. The main forest diseases recognised in Western Australia are caused by fungi. They are dieback caused by Phytophthora cinnamo?iiir straw rots {Armillaria luteobttbalina) and karri brown wood (associated with a variety of fungi). The effect of one of these diseases, jarrah dieback, on forest production was examined in a five year dendrometer study of tree growth. Overall diameter increment of jarrah on dieback-affected sites was less than on either control or thinned but dieback-free sites. Growth rates varied greatly both between sites and between trees on the same sites irrespective of disease status. As a consequence, the effects of dieback on tree growth are difficult to separate from site factors (e.^. rainfall, soil, topographical position) and tree factors {e.g. genetic potential, age, vigour and dominance class). Introduction The native forests and woodlands are a major asset of Western Australia. State forests are the source of nearly 2 million cubic metres of timber products annually (Table 1), of approximately half the water used by cities and for irriga¬ tion in the south-west (total storage capacity 910“^ megalitres, Olsen & Skitmore 1991) and provide a varied environment for tourism, recreation and education. Conservation values of the forests are also high as, even in areas subject to logging, disturbance has been much less intense than in neighbouring agricultural or urban areas (Havel 1989). Table 1 Timber production in Western Australia in 1992/93 (from Conservation and Land Management 1993). Category Volume (m^) Eucalyptus margimta (jarrah) 385 819 Eucalyptus calophylla (marri) 45 587 Eucalyptus diversicolor (karri) 195 613 Pinus spp. (mainly Pinus radiata) 149 487 Non-sawlog {e.g. chiplogs, firewood) 1 100 077 Diseases affect forest values adversely by a ffecting growth, hydrological cycles, stand structure and species composi- Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16,1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 tion. The major disease problem of Western Australia's native forests is dieback associated with the soil-borne fun¬ gus Phytophthora ciunarrtomi. The disease is often termed "jarrah dieback" because of its highly visible effects on this dominant forest species. Affected jarrah typically dies after one or more cycles of crown death alternating with periods of partial recovery. Other significant forest diseases identi¬ fied in Western Australia are root rots and basal cankers caused by the fungus Armillaria luteobubalina, and brown wood and rots of karri associated with a number of fungi. occurs throughout the south-west forests (Pearce et al. 1986, Shearer 1992, Shearer & Tippett 1988). Although Armillaria luteobubalma is capable of infect¬ ing a wide range of hosts including jarrah, marri, karri, wandoo and Banksia gratidis, it seldom cau.ses deaths of more than individual trees or small patches of trees. Armillaria luteobubalwa lesions are usually contained by periderms if the host is a vigorous mature tree. However, mortalities can occur when lesions cannot be contained or large numbers of infections are initiated. Trees of highly susceptible species (e.g. Eucalyptuszoandoo)oTwhich are small, whose vigour has been reduced by intense competition or drought, or are close to A. luteobubalina-colomsed stumps supporting high local inoculum populations, are at the greatest risk of damage. Brown wood of karri is a discolouration associated with infection by a number of fungi with Stereum hirsutan and Hymefwchaete sp. being the most common (Davison & Tay 1991). Brown wood does not affect timber strength (Siemon, CALM, Perth, pers. comm.) but it is unsightly and has been identified as a precursor to rots (Davison & Tay 1991). Such rots are unlikely to develop in timber which has been dried to less than 20% moisture content (Bootle 1983). The extent of the brown wood problem in karri is difficult to assess as the discolouration is visible only when stems are sectioned. 145 Journal of the Royal Society of Western Australia, 77 (4), December 1994 However, a limited study of regrowth karri mainly from the Treen Brook area found brown wood in 73% of 270 trees sampled (Davison & Tay 1991). The generality of this finding is untested. The effects of A. htteobubalina and karri brown wood on forest values and forest production are less well understood than are the effects of P. cinnamomi. More extensive reviews of current knowledge of the occurrence and biology of P. cinnamomi, A. luteobubalina, brown wood and other fungal pathogens and saprophytes of the south west forests are given by Hilton et al. (1989) and Shearer (1992). This paper will illustrate some of the difficulties of assess¬ ing the effect of pathogens on forest production by describ¬ ing recent attempts to measure the effect of P. cinnamomi dieback on growth of jarrah. Jarrah dieback Phytophthora cinnamomi infestation of native forest causes changes in stand density and species composition (Davison & Shearer 1989) which affect non-timber values such as water yield (Schofield 1990), biodiversity, honey production and recreation. Pinus radiata on susceptible .sites can also be killed or damaged (Chevis & Stukely 1982, Butcher et al. 1983). P/iyfop/if/iorflc/n«flmom/diebackaffectsanestimatedl4.2% of the jarrah forest (32 000 ha of the 225 000 ha of jarrah forest mapped by aerial photography; H Campbell, cited by Davison & Shearer 1989). The northern and western pa rts of the jarrah forest are affected more than the southern areas (Batini & Hopkins 1972) possibly because of differences in climate and soils which affect fungal behaviour and in the intensity of human activities which spread the fungus (Shearer 1992). Initial prognoses for the survival of the jarrah forest were pessimistic because of early experiences of rapid death of virtually all jarrah trees over large areas. Accordingly, much early work was directed to replacing the native jarrah with Phytophthora resistant exotic species (Bartle & Shea 1978). Better understanding of the factors affecting the spread of P. cinnamomi in the forest and of jarrah's ability to resist P. cinnamomi (see Shearer & Tippett 1989) coupled with changing community attitudes to forest management has resulted in a more optimistic outlook. Management policy now emphasises maintaining as much of the jarrah forest as possible and limiting the spread of P. cinnamomi to uninfected forest by controlling access (Conservation and Land Man¬ agement 1987,1989). Quantifying the effect of dieback on forest production has been difficult. Estimates of the loss of diameter increment of jarrah due to dieback vary from an 87% reduction (Podger 1972), through a smaller 12% loss (Crombie & Tippett 1990) to no effect or even a slight increase (Davison & Tay 1988). It is probable that the outcome depends on the balance be¬ tween the level of damage caused by the pathogen and the benefits of reduced competition as susceptible neighbouring trees and understorey plants are killed. Methods Nine sites comprising a dieback-affected and an adjacent dieback-free area were chosen in the jarrah forest between Perth and Dwellingup. The association of P. cinnamomi with the observed dieback symptoms was confirmed by recovery of Phytophthora from the dieback-affected areas at all sites. A further five dieback-free sites were included to extend the range of site occupation on dieback free sites to encompass that of the dieback sites. Sites selected included elements of the S site type of Havel (1975) as this is representative of the main area available for logging in the northern part of the jarrah forest. The dieback-affected sites and their character¬ istics are listed in Table 2. Four sites (Angle Swamp, Bound¬ ary Road, Canning Dam Road and Karragullen) were those for which initial results have been discussed by Crombie & Tippett (1990). A low intensity fire affected the dieback-affected part only of the Ashendon Road site in the autumn of 1990. Such fires have been associated with increased growth rates (Davison & Tay 1988, Kimber 1978) but the effect in this instance in unknown. Dominant or co-dominant trees of as nearly similar diam¬ eter as could be managed were selected on each of the dieback-affected and dieback-free plots. Selected trees were fitted with stainless steel dendrometer bands (Liming 1957) and increments recorded monthly. Results Mortality Growth of jarrah on dieback sites was monitored for three to five years giving a total of 548 tree-years (number of trees x number of years). Eight trees on three sites died during this time (1.5 % per annum). This compares with previous re- Figure 1. Radial increment of trees on adjacent dieback-free and dieback-affected areas at Pumping Station Road. Standard errors of increments between September 1988 and October 1991 are included (vertical bars) to show variability. There were 11 trees on the dieback- free area (■) and 23 in the dieback-affected area (A). 146 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Table! Characterisations of stands: C = Control, D=Dieback-affected, T=Thinned. Soil type; L^lateritic boulders and lateritic gravel S-L=sandy soil with occasional lateritic rocks, S=sandy soil without lateritic boulders. Soil Type Rainfall (mm) Number of Trees Diameter under bark (mm) Height (m) Stand Density (Stems ha ’) Basal Area (m2 ha-’) Dieback Sites C D C D C D C D C D Angle Swamp L 850 10 10 247 275 14 14 789 280 27 21 Ashendon Road L 1050 12 12 330 323 18 18 1353 1 626 31 29 Boundary Road S-L 1050 11 11 226 251 18 19 865 654 28 22 Canning Dam Road L 1250 10 11 282 265 19 19 1054 318 31 24 Hay Creek S 1100 11 11 326 306 17 15 926 912 34 32 Gravel Pit S 1250 11 11 376 355 19 20 1050 556 38 30 Karragullen L 125U 10 7 252 255 16 16 1455 919 31 24 Pumping Station Road S 1200 12 21 303 248 18 15 934 458 34 20 Sawyers Valley S-L 900 12 11 261 265 17 14 690 269 22 16 Thinned Sites C T C T C T C T C T Banksiadale Road L 1300 20 19 337 349 23 25 988 638 35 15 Jarrahdale Road S-L 1200 12 314 24 1549 44 Mundaring S-L 1100 40 60 251 246 17 17 1 498 1472 29 21 Roma Road L 1250 20 20 295 301 23 22 1722 296 30 13 Torrens Road L 1300 20 20 282 287 21 20 1348 914 36 10 ports of report of 4.2 % per annum deaths (three sites, Podger 1972) and 0.84 % deaths per annum (two sites, Davison & Tay 1988). No trees died on dieback-free sites during this study. Deaths were unevenly distributed between seasons and also between years. Five deaths occurred when evaporative de¬ mand was high in mid to late summer (February to April) and three when it was rising in spring (September or Octo¬ ber). More deaths occurred in 1991 and 1992 than in 1988, 1989 or 1990. Growth Radial increment varied between dieback-affected and dieback-free sites and also between trees within sites, be¬ tween sites and from month to month and year to year. Typical growth patterns are illustrated with data from the Pumping Station Road site in Figure 1. Stem growth occurred when soil water was readily avail¬ able (from the beginning of rains in April or May until January or February) but was slowed by low temperatures during winter Qune to August; see also Abbott & Loneragan 1983). Slower diameter growth or slight shrinkage were associated with water deficits during summer drought (Feb¬ ruary to March). Growth patterns of trees on dieback-affected areas were similar to those of trees on dieback-free sites with the excep¬ tion that summer shrinkage began one to two months earlier and was more pronounced in trees on dieback-affected sites. The effect of dieback on growth was examined by analysis of covariance using the total radial increment occurring be¬ tween October 1988 and October 1991 (Table 3). Data were transformed to square roots to reduce heteroscedascity. Stand basal area (SBA) and site were identified as the major covariates by step-wise regression. Table 3 Analysis of covariance. Tests were conducted by applying the SAS procedure General Linear Models to square root transformed radial increments from October 1988 to October 1991. SBA= Stand Basal Area. Source DF Squares Sum of Square Mean F Value P SBA 1 13.83 13.83 21.12 0.002 Site 13 26.88 2.07 3.16 0.05 Dieback 1 1.93 1.93 2.94 0.12 Site* Dieback 8 5.24 0.65 1.93 0.05 Within plot 411 139.52 0.34 Total 434 201.13 Phytophtlwra chmamomi dieback was less a predictor of tree growth (p = 0.12) than either stand basal area (p = 0.002) or site (p = 0.05). The low level of significance in the results reflects the small sample sizes used (10 to 12 trees per site). A post hoc consideration of the variances in growth suggests that a sample size of 30 trees would have been needed to obtain differences between growth of trees on dieback- affected dud dieback-free sites significant at the 5% level. Comparison of the adjusted mean squares solutions to the model (Table 4) shows that increment on dieback sites 147 Journal of the Royal Society of Western Australia, 77 (4), December 1994 averaged about 80% that on dieback-free sites. However, growth of trees on the dieback-affected areas relative to dieback-free areas varied widely from a minimum of 38% on the Angle Swamp site to a maximum of 140% at Ashendon Road. Table 4 Mean of measured radial increments and least squares adjusted mean increments (as square root of increment) on dieback-free control (C) or dieback-affected (D) sites. Covariates used in the adjustment were stand basal area and site. Site _ Radial Increment (v^mm) Ratio Measured Least squares estimate (D / C) C D C D Angle Swamp 2.1 1.4 2.1 1.3 0.62 Ashendon Road 1.6 1.8 1.6 1.9 1.16 Boundary Road 1.9 2.1 2.0 1.9 0.97 Canning Dam Road 2.6 2.4 2.7 2.2 0.83 Hay Creek 2.2 1.7 2.3 1.8 0.78 Gravel Pit 2.1 2.3 2.3 2.3 1.01 Karragullen 2.7 2.9 2.7 2.7 1.02 Pumping Station Road 2.4 2.1 2.4 2.0 0.83 Sawyers Valley 1.8 2.1 1.8 1.8 1.03 Growth of trees on dieback sites was not affected in a consistent way by outcroppings of lateritic ironstone or sandy soils which might indicate differences in drainage. Thus while growth at two sites with masses of exposed lateritic duricaist (Angle Swamp and Canning Dam Road) was reduced below those of controls, growth was about the same at another site (Karragullen) and was increased sub¬ stantially at a fourth (Ashendon Road). Of the sites with sandy soils, growth was reduced on the dicback areas at two sites (Pumping Station Road and Hay Creek) but was virtu¬ ally unaffected on another two sites (Boundary Road and Gravel Pit). Discussion Synchronisation of tree deaths with times of high or rising evaporative demand and higher death rates in years with unusually high rainfall hasbeen noted before (Davison 1988, Faggcf r?/. 1986, Shea etai 1983, Hamilton, 1951 utipub. nyort cited by Dell & Malajzuk 1989, Tippett et ai 1985). Both observations are consistent with the expected behaviour of P. cinnatnowi which requires water for dissemination and infection and which damages the roots necessary for uptake of water during summer drought (Crombie & Tippett 1990). Overall radial incrementoftreesondieback-affectedsites was about 80% that of similar trees on nearby dieback-free sites, although the estimates are imprecise (see above). The lack of precision is disappointing given that the biotic and environmental factors affecting P. cinnamovii are well known (t\y. Shearer & Tippett 1989). The greatest difficulty encountered was the great vari¬ ability in tree vigour, age, size, spacing, genotype, topogra¬ phy and soil type which occur in native forests (Stoneman et al. 1989). Determination of the effects of disease on growth of jarrah are made more difficult by the very slow growth of jarrah in native forest (diameter increments of 2-3 mm yr*) which require long periods before the effects of disease on growth become evident. Locating suitable control plots was also a problem. Poten¬ tial control sites adjacent to diseased areas are often disease- free because they have some characteristic which has pre¬ vented the pathogen occupying the site. Differences in soil type, amount of deep drainage and sub-surface topography may be very important in determining disease occurrence but are difficult to identify in the field. Even when the factors controlling disease occurrence are known, other factors limit site selection. In particular, the location, shape and size of dieback outbreaks is largely dependent on the interaction of site hydrologic characteris¬ tics and forest operations. Dieback infestations originating from soil dropped from vehicles (Podger 1972) usually spread downhill along drainage lines leaving the tracks from which the infections began as the boundary between dieback- affected and dieback-free areas. Such tracks are often used as convenient boundaries for logging coupes or as firebreaks (as happened at the Ashendon Road site) so that stand structure and history are often not the same on both sides of the track. Areas downstream and possibly to the side of known infections must also be considered to be either in¬ fected or highly likely to become so as spores are distributed by mass flows of soil water (Kinal et al. 1993). Finally, plot access tracks hav'c to be suitable for wet weather use without risking establishing new P. cinnamawi infestations. The time since infection, occurrence of subsequent re¬ infections and the rate and extent at which dieback develops may also be important to the effect of dieback on growth but are seldom known when studying disease in native forests. Similarly, genotypic (Butcher et al. 1984) and environmental factors, including temperature and summer drought(Shearer & Tippett 1989, Tippett et al 1987), are likely to affect the rate of development and severity of disease expression but can¬ not be controlled in most field situations. Conclusion This study has recorded both reductions and increases in growth of jarrah on P. cinnawomi infested sites. The provi¬ sional conclusion is that average diameter increments of trees on d ieback sites are approxima tely 80% those on dieback- free sites when adjusted for differences in site quality and competition. Differences in the effect of P. emmmomi on growth of particular jarrah trees are likely to be related to tree, site and climatic differences. References Abbott I & Loneragan O 1983 Ecology of jarrah {Eucalyptus marginata) in the northern jarrah forest of Western Australia. Department of Conservation and Land Management, Perth, Bulletin Number 1. Bartle ] R & Shea S R 1978 Selection of tree species for rehabilitation of degraded areas in the northern jarrah forest. In: Proceedings: Rehabilita¬ tion of Mined Landsin Western Australia. Western Australian Institute of Technolog)', Bentley, 7-16. Batini F E & Hopkins E R1972 Phytophthora citwantomi Rands -a root pathogen of the jarrah forest. Australian Forestry 36:57-68. Bootle K R1983 Wood in Australia. Types, Propertiesand Uses. McGraw-Hill, Sydney. 148 Journal of the Royal Society of Western Australia, 77 (4), December 1994 ButcherTB, Stukely M J C & Chester G W1983 Genetic variation in resistance of Pintis radiata to Phytophthora cinnamonji. Forest Ecology and Manage¬ ment 8:197-220. Chevis H W & Stukely MJ C1982 Mortalitiesof young established radiata pine associated with Phytophihora spp in the Donnybrook sunkland planta¬ tions in Western Australia. Australian Forestr)' 45:193-200. Conservation and Land Management 1987 Northern Forest Region Regional Management Plan 1987-1997. Department of Conscrv'ation and Land Management, Perth. Conservation and Land Management 1989 Regeneration in forest affected by Phytophthora cinttamomi. Department of Conservation and Land Manage¬ ment, Perth. Silviculture Specification 4/89. Conservation and Land Management 1993. Annual Report 1992. Department of Conservation and Land Management, Perth. Crombie D S & Tippett J T 1990 A comparison of water relations, visual symptoms and changes in stem girth for evaluating impact of Phytophthora dftnanjomi dieback on Eucalyptus imrgwata. Canadian Journal of Forest Research 20: 233-240. Davison E M 1988 The role of waterlogging and Phytophthora cituiamomi in the decline and death of Eucalyptus margimta in Western Australia. Geo- Journal 172:239-244. Davison E M & Shearer B L 1989 Phytophthora spp in indigenous forests in Australia. New Zealand Journal of Forestry lienee 19:277-289. Davison E M & Tay F C S1988 Annual increment of Eucalyptus marghtata trees on sites infested with Phytophthora chwamomi. Australian Journal of Botany 36:101-106. Davison E M & Tay F C S1991 Rot and incipient rot in Eucalyptus dwersicolor. Australasian Plant Pathology. In: Proceedings, 8* Australasian Plant I’athological Society Conference, University of Sydney. Dell B & Malajezuk N 1989Jarrah dieback-A disease caused by Phytophthora cinnamomi. ln:The jarrah forest: A complex mediterranean ecosystem (eds B Dell, J J Havel & N Malajezuk) Kluwer, Dordrecht, 67-87. Fagg P C, Ward B K & Featherston GR1986 Eucalypt dieback associated with Phytophthora cinnamomi following logging wildfire and favourable rain¬ fall. Australian Forestry 49:36-43, Havel J J1975 Site-vegetation mapping in the northern jarrah forest (Darling Range). 1 Definition of site-vegetation types. Forests Department, Perth, Bulletin Number 86. Havel J J1989 Conservation in the northern jarrah forest. In: Tlie jarrah forest: A complex mediterranean ecosystem (eds B Dell, J J Havel & N Malajezuk) Kluwer, Dordrecht, 379-399. Hilton R N, Malajezuk N & Pearce M H 1989 Larger fungi of the jarrah forest: An ecological and taxonomic surv’ey. In: The jarrah forest: A complex mediterranean ecosystem (eds B Dell, J J Havel & N Malajezuk) Kluwer, Dordrecht, 89-109. Kimber P C 1978 Increased girth increment associated with crown scorch in jarrah. Forest Department of Western Australia, Perth, Research Paper Number 37. Kinal J, Shearer B L & Fairman RG 1993 Dispersal of Phytophthora cinnamomi through laterilic soil by laterally flowing subsurface water. Plant Disease 77:1085-1090. Liming F G 1957 Homemade dendrometers. Journal of Forestry 55:575-577. Olsen G & Skitmore E 1991 State of the rivers of the southwest drainage division. Western Australian Water Resources Council, Perth. Pearce M H, Malajezuk N & Kile G A 1986 Tlie occurrence and effects of Armillaria lutcohiibaltna in the karri (Eucalyptus diversicolor F. Muell.) forests of Western Australia. Australian Forest Research 16:243-259. Podger F D1972 Phytophthora cinnamomi a cause of lethal disease in indigenous plant communities in Western Australia. Phytopathology 62:972-981. Schofield N J 1990 Water interactions with land use and climate in south western Australia. Water Authority of Western Australia Perth, Report Number WS 60. Shea S R, Shearer B L, Tippett J T & Deegan P M1983 Distribution, reproduc¬ tion and movement of Phytophthora cinnamomi on sites highly conducive to jarrah dieback in south western Australia. Plant Disease 67:970-973. Shearer B L1992 The ecological implications of disease in the southern forest of south-western Australia. In: Research on the impact of forest manage¬ ment in south-western Western Australia. CALM Occasional Paper No 2 / 92. ShearerBL&TippettJT1988DistributionandimpactofAnnf7/flria/Mfco6wi?fl/mfl in the Eucalyptus marginata forest of south-western Australia. Australian Journal of Botany 36:433-45. Shearer B L & Tippett J T1989 Jarrah dieback: the dynamics and management of Phytoifhthora cinnamomi in the jarrah (Eucalyptus marginata) forest of south-western Australia. Department of Conservation and Land Manage¬ ment, Perth, Research Bulletin Number 3. Stoneman G L, Bradshaw F j & Christensen P1989 Silviculture. In: The jarrah forest: A complex mediterranean ecosystem (eds B Dell, J J Havel & N Malajezuk) Kluwer, Dordrecht, 335-355. Tippett J T, Crombie D S & Hill T C1987 The effect of phloem water relations on the growth of Phytophthora cinnamomi in Eucalyptus marginata. Phytopathology 77:246-250. Tippett J T, Hill T C & Shearer B L1985 Resistance of Eucalyptus spp to invasion by Phytophthora cinnamomi. Australian Journal of Botany 33:409-418. i j' I i i! i i I 1 i Journal of the Royal Society of Western Australia, 77: 151-158,1994 The impact of plant disease on mining IJ Colquhoun’ & A E Petersen^ 'Environmental Department, Alcoa of Australia Limited, PO Box 252, Applecross WA 6153 ^RGC Mineral Sands Limited, PO Box 47, Eneabba WA 6518 Abstract Dieback disease caused by Phytophthora species is the only plant disease having a major impact on the bauxite mining operations of Alcoa of Australia Ltd and the mineral sands mining operations of RGC Mineral Sands Pty Ltd. To mine responsibly in regions of the state where dieback is a concern, both companies have implemented major dieback management programs. The financial costs associated with implementing these programs are substantial. However, estimating the total financial cost of dieback disease management on the mining operations is very difficult because dieback control procedures are integrated with mine planning and other operational procedures. Both companies recognise that these financial costs are part of the overall cost of mining in regions of Western Australia where dieback is present, and conservation of the natural vegetation communities is of high priority. Financial considerations are not the only potential impacts of disease. Plant diseases can also impact on environmental management objectives. Dieback disease has the potential to affect three important environ¬ mental management objectives; protecting the adjacent natural vegetation, establishing key plant species in rehabilitated sites, and achieving high species richness in these sites. Achievement of these objectives is regarded as important to the success of the mining operations of both companies. Alcoa and RGC monitor vegetation adjacent to their mining operations and rehabilitated areas for symptoms of dieback. There has been no Phytophthora-caused plant death found near RGC's operations in areas interpreted previously as uninfested. Monitoring of the forest surrounding a 120 ha bauxite mined area identified two new P.citmamotm infestations; the total area was 0.32 ha. Establishment of key plant species in the rehabilitation {Eucalyptus marghuUa in the jarrah forest and Banksia species in the kwongan) has been successful despite the presence of Phytophthora species in the region. High species richness is achieved in the rehabilitated areas of both operations. Long term monitoring and continued research are still required on many aspects of this disease. By developing dieback control procedures based on scientific knowledge of the pathogen and the disease process, and integrating these procedures with routine mining procedures, we believe both companies have successfully minimised the impact of their mining operations on surrounding and re-established native vegetation communities. Introduction Plant disease has a major financial impact on the mining operations of Alcoa of Australia Limited and RGC Mineral Sands Limited, However, there is more to impact than just financial considerations. Evaluation of the impact of disease on mining also needs to assess the impact of disease on the environmental management objectives of both companies. Achievement of these objectives is regarded as important to the overall success of the mining operations. The only plant disease having a significant impact on Alcoa and RGC is dieback disease caused by Phytophthora species. This disease is a major threat to the native plant communities in the regions of Western Australia where both companies mine. Environmental management objectives have been developed for the mines; their focus is to protect Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16,1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 the natural communities surrounding the mine and to create an ecosystem on the mined areas with vegetation similar to that which occurred before mining. The presence of dieback has the potential to prevent these objectives being achieved. The aim of this review is to discuss both the financial impact of disease on Alcoa and RGC mining operations and the potential impact of plant disease on the environmental objectives of both companies. Background to mining operations Alcoa operates three bauxite mines in the jarrah forest, south east of Perth (Fig 1). Alcoa mines and rehabilitates an average of 450 ha of forest a year. Bauxite ore bodies tend to be on the mid to upper slopes and are of the order of 5-50 ha. The open cut mining procedures remove the hard laterite ''caprock" and the more friable zone below the caprock. The depth of mining is usually between 3 and 5 m. Following mining the area is landscaped, the surface soil is returned and the area is ripped to remove compaction. Finally, the area is seeded with about 50 species of plants indigenous to 151 Journal of the Royal Society of Western Australia, 77 (4), December 1994 I Figure I Location of Alcoa's operations in Western Australia Figure 2 Location of RGC s operations in Western Australia Journal of the Royal Society of Western Australia, 77 (4), December 1994 this region of the northern jarrah forest, jarrah is established as the dominant tree species. Theoverall aim of revegetation is to create a community similar to the one present prior to mining. The mining and rehabilitation procedures are de¬ scribed in more detail by Nichols et al. (1985) and Ward et al. (1993). AH mines are located in the western, high rainfall region of the jarrah forest where the extent of P. chimmomi infestation is considerably greater than the estimate of 14% derived for the entire jarrah forest (Davison & Shearer 1989). P. cittnamonii infestation ranges from 66% of the forest at the Willowdale mine to 33% at the Huntly mine. RGC mines and processes mineral sands to produce ilmenite, rutile and zircon. Its largest operations are near Encabba on the Swan Coastal Plain about 300 km north of Perth and 30 km inland (Fig 2). Currently two mines are operating. One mine is a dredging operation whilst the second uses trucks, scrapers and other conventional dry mining equipment. The ore bodies are located underneath native heathland (kwongan) and agricultural land. The distribution of ore tends to follow long thin “strands" {e.g. 10 km long, 120m wide and 30m deep). Once the ore is mined, the heavy mineral is removed by wet gravity separation using water. The non-mineral slurry is pumped back to the mined-out pits where it is left to dry before it is landscaped, topsoil returned and revegetated. In the species-rich kwongan, the rehabilitation objective is to establish a self sustaining ecosystem similar to the undisturbed heath. Pre¬ mining surveys recorded 429 species from 50 families and 162 genera. Many of the principal genera are susceptible to Phi/tophthora species. Plant diseases Dieback disease caused by Phytophthora species is the only plant disease having a major impact on the mining operations of Alcoa and RGC. The term 'dieback disease' is used here exclusively to describe the diseases of native plants caused by Phytophthora spp, where infection leads to a 'root rot' or 'collar rot', and the eventual death of the plants. The earliest records for this disease in Western Australia came from the jarrah forest where there were unexplained deaths of jarrah trees in the early 1920s. The a.ssocialion between the deaths of jarrah and infection by P. cinnamomi was established in the mid 1960's(Podgercf al. 1965, Podger 1972). Although this disease in the jarrah forest is commonly called 'jarrah dieback', it is known that many other species are susceptible to P. cimiamomi in a range of vegetation communities (Shearer & Tippett 1989, Shearer 1990). Phytophthora cinnawomi is the most destructive of the seven species of Phytophthora found in south-western Aus¬ tralia. It is the only species having a significant impact on the mining operation of Alcoa (Colquhoun 1992) and it is the cause of most Phytophthora related deaths in the northern sandplain around the RGC mine. However, P. dfrtcola, P. ntegasperma var megaspcrtiia, and P. megasperwa var sojae are also known to have caused plant death in the northern sandplain (Hill 1990). All of these Phytophthora species are soil borne and spread readily in soil and water. During mining, huge volumes of soil are transported, large road networks are established and the drainage of areas is altered markedly. All Western Aus¬ tralian mining companies working in Phytophthora-miested areas recognise that mining has the potential to spread the pathogens and thereby increase the area of infestation. The need for dieback control measures is also recognised by the companies. Alcoa and RGC have developed and deployed intensive dieback management programs to minimise the potential to spread the pathogens to native communities and rehabilitated mined areas. The dieback management pro¬ grams are based on an understanding of the pathogen and the disease process. This understanding was gained from scientific studies; many of these were undertaken by local research institutions and funded by West Australian mining companies. There is a range of other plant diseases which are having an impact on mining operations, but the impact is minor in comparison to dieback disease. A range of canker fungi have been isolated from plants growing in rehabilitated bauxite mined areas. The fungi identified include Bofn/osponun/spp, Botryosphaeria ribis, Cytospora eucalypticola and Pestalotioposis sp (Carswell 1993). Only on a few occasions has the presence of these fungi been linked to the death of a plant in the rehabilited areas. Armillaria liiteobubiliua is found frequently in the jarrah forest. Forest sites infested with this pathogen have been mined and the soil used in rehabilitation. Despite regular monitoring of deaths of plants in rehabilitation, and field reconnaissance of most rehabilitated pits, there has been no report of multiple deaths from this pathogen. Alcoa and RGC operate nurseries which propagate native plants for use in the mine rehabilitation program. The control of plant diseases caused by species of Phytophthora, Rhizoctionia, and Pythiuw is critical to successfully producing container grown plants. Both Alcoa and RGC nurseries have been successfully accredited by the Nursery Industry Asso¬ ciation of WA as meeting the required standards of Phytophthora disease control. Two of the most important plant families in the Northern Sandplains are the Restionaceae (rushes) and Cyperaceae (sedges). Previous research had shown that most rush and sedge species in the Northern Sandplains are affected by smut diseases. These diseases may cause up to 100% reduc¬ tion in seed set and reduce seed quality in many species. Smut-affected populations have been shown to have a sig¬ nificantly reduced capacity to recruit seedlings, particularly where the parent plants are killed by fire. Plants of a number of species become chlorotic and senesce following infection by smut diseases. Research work is currently being funded by RGC and conducted by Kings Park Board on controlling smut in rushes and sedges so that improvements can be made in their rehabilitation and the impact of these diseases is lessened. The financial impact of dieback on mining The dieback management programs of Alcoa and RGC are a major cost to the operations. Dieback control measures are present at virtually all stages of mining. Some major costs are easy to determine (Table 1) but the greatest will be the "hidden" costs associated with the day-to-day operations of the mines. Many stages of mining now take longer and require increased use of machinery. To minimise the disrup¬ tions to mining, the control measures have been integrated with the routine mine planning and operational procedures. Due to this integration the total financial impact is very difficult to determine. 153 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Table 1 Major costs associated with Alcoa and RGC dieback management programs ALCOA $ RGC $ Construction of vehicle cleaning facilities 300,000 300,000 Dieback interpretation and mapping 200,000 per year 20,000 per year Mapping site vegetation types 220,000 every 5 years Not applicable Dieback research funding of external projects 210,000 per year 10,000 per year Rather than attempt to determine the cost of the dieback management programs for each company, we have pro¬ vided an overview of the dieback management strategy used by RGC, and a comparison of one stage of bauxite mining with and without dieback control measures. These exam¬ ples will demonstrate the complexity of dieback manage¬ ment and its associated costs. Dieback management strategy - RGC RGC has implemented a dieback management program that addresses hygiene practices, surveys, rehabilitation tech¬ niques, drainage patterns, education, research and regular reviews. Some key components of the strategy are: • mapping the presence of dieback on and around the mining; • advising all contractors, carriers and personnel of their obligations under the dieback management program; • ensuring that all vehicles and equipment entering and leaving the site are clean of soil and plant material, by inspection and authorisation from suitably trained per¬ sonnel; • providing appropriate washdown and inspection facili¬ ties; • constructing new roads and upgrading existing roads to dieback control standards e.g, developing a hard road surface and constructing roads above ground level; • liaising with the Department of Conservation and Land Management, land owners and other authorities to con¬ trol unauthorised access; • • developing procedures for movement of vehicles and equipment between sites and locations; • segregating soil from dieback and non-dieback areas; and • supporting research to increase the understanding of the disease and to develop improved dieback control meas¬ ures. Alcoa employs a similar dieback management strategy. As part of this strategy, Alcoa has developed detailed dieback control measures for each stage of mining. Dieback control measures associated with overburden* removal and storage - Alcoa During this phase of bauxite mining, all of the overbur¬ den is stripped from the orebody (using scrapers) and stored in large stockpiles. Following mining and landscaping of the minepit this overburden is returned to the area. Without dieback control measures, the process is simple; the scraper traverses the orebody in the most efficient manner to remove the overburden and transports it to a stockpile. Thestockpile is usually located at the edge of the orebody, in the most convenient position for the operations (Fig 3a). With the application of the current dieback control meas¬ ures for stripping overburden at the Huntly mine, the oper¬ ating procedures now involve (see Fig 3b): • mapping for the presence of dieback (approx. 10,000 ha for 10 years of mining at one mine); • marking the dieback boundary in the field. This bound¬ ary lends to split the orebody into “dieback" and “dieback- free" areas; • surveying the boundary lines and recording these on a geographical information system (GIS) for use by mining and environmental planners; • stripping the “dieback" and “dieback-free" overburden separately; • ensuring that all vehicles moving across the dieback boundary into the “dieback-free" area are free of infested soil. Vehicle cleaning stations are located on the bounda¬ ries; • storing the “dieback" and “dieback-free" overburden in separate stockpiles; • selecting stockpile locations where the risk of spreading P. dunamomi and the risk of increasing the impact of the pathogen in the infested forest is minimal. This generally results in the “dieback-free" stockpiles being constructed in a previously mined area, where the risk of water draining from the stockpile into the uninfested forest is prevented, and • where necessary, increasing the vertical infiltration of water of the site selected for the stockpile. This usually involves blasting the rock layer on and downslope of the future stockpile area. The impacts of dieback control procedures on the day to day operations of removing and storing overburden are manifest; the procedure takes longer, requires more environ¬ mental monitoring, requires more resources (scrapers, drill rigs from blasting, surveyors, GIS operators, etc.) and re¬ quires more education and training of the operators and planners. This level of intensity of dieback control is achieved at all stages of bauxite mining and rehabilitation. To collate the “hidden" costs associated with dieback management for RGC and Alcoa would be a complex and arduous task, and we do not believe that this is necessary. Both companies recognise that these financial costs are part of the overall cost of mining in the region of Western Aus¬ tralia where P. dnmmomi is present and the conservation of the natural vegetation communities is of high priority. FIow- * Overburden is the part off the soil profile above the ore layer, which remains after the topsoil (top 15 cm of the profile) has been removed. The nutrient level and seed load is low in the overburden so this soil is stockpiled and returned during the rehabilitation phase. 154 Journal of the Royal Society of Western Australia, 77 (4), December 1994 '-"N-X XWsSS . . Ti c c o >, •u o Xi o L- o X 3 (C Si rz B o c o T3 u 3 S u O > o cc c ■q- Cl, o c o a o 3 00 155 Journal of the Royal Society of Western Australia, 77 (4), December 1994 ever, research and development work continually aims to develop more effective and less expensive procedures. Impact of dieback on environmental objectives Dieback has the potential to adversely affect three impor¬ tant objectives for the environmental management of the mining leases; • protect the vegetation surrounding the mined areas, • establish key plant species in rehabilitated mined areas, and • achieve high species richness in rehabilitated mined areas. Protection of the adjacent natural vegetation This environmental management objective is to prevent mining from causing significant adverse effects on natural communities near the mines. The spread of Phytophthora species to these sites has the potential to have a major impact on the health and species composition of these areas. Conse¬ quently, a critical aim of all dieback management programs is to prevent the spread of the pathogen to uninfested, natural communities. Control measures are introduced to minimise the risk of infested soil or water being deposited in these areas. At Alcoa, attempts are being made to quantify the level of spread of P. cinnamomi that can be directly related to mining operations. The procedure involves identifying pits where the adjacent forest was uninfested before mining then re¬ mapping for dieback symptoms after rehabilitation. The increase in the area of infested forest is then calculated. Any new infections found abutting the pit edge would, in most instances, be attributed to mining. This procedure has been applied to an area at the Huntly mine that received the current intensive dieback control measu res. This large mine pit was mined from 1986 to 1990 and rehabilitated in 1991 and 1992. The total area mined was 120 ha and 13.5 km of the pit edge abutted forest interpreted as "dieback free" before mining. The re-interpretation of the pit edge in 1993 identi¬ fied two new infections; the total area of the new infections was 0.32 ha. This is 0.23% of the area cleared for mining. Monitoring of the spread of the pathogen provides infor¬ mation on the impact of mining on the adjacent community; it also provides vital feedback on where and how the patho¬ gen is being spread. This feedback is used to improve the dieback control procedures and their implementation. At RGC the natural vegetation communities around the orebody are inspected regularly. There has been no Phytophthora-caused plant death found near the mining op¬ erations in communities previously interpreted as uninfected. The protection of native communities is an important part of environmental management of most mining opera¬ tions. The measured spread of Pfiytophthora species to these communities is very low but we believe that some long term monitoring is required to ensure that the companies' envi¬ ronmental objectives continue to be achieved. Establishment of key plant species The establishment of particular plant species is regarded as important to the success of mine rehabilitation in the jarrah forest and kwongan. In the jarrah forest the key species is jarrah and in the kwongan the dominant Banksia species are regarded as key species. All of these key species are susceptible to Phytophthora species Alcoa and RGC mine PIiytophthora-infesiQd areas so, irre¬ spective of dieback management programs, Phytophthora species will be present in some of the rehabilitated areas. Will the presence of Phytophthora species prevent the estab¬ lishment of these key species? The objective of re-establishing a forest dominated by jarrah after bauxite mining was not considered in the early years of mining. It was assumed that the presence of P. cifinaniottii would kill the trees. However, in 1978 and 1979 jarrah was established at two mines. In 1986, trials were established at all mines to assess the establishment of jarrah from broadcast seeding. The success of this trial, and high survival of jarrah trees established in earlier revegetation, led to jarrah being re-established as the dominant tree spe¬ cies in most rehabilitated sites after 1987, and all sites after 1991. Phytophthora cmnanioini has been isolated from dead jarrah plants and soil in rehabilitated mined areas but sur¬ vival of jarrah is high in both the infested and uninfested sites (Fig 4). The present target for jarrah establishment is 2000 trees ha ' after 9 months. The eventual stocking density of these sitesis expected to be similar to a fast growing regrowth forest site i.c. 300-500 trees ha '. In 1993 the mean stocking density of jarrah in 32 rehabilitated pits, 9 months after seeding, was 2790 trees ha '. Only two pits had less than 1500 trees ha '. So despite the presence of P. diinafnomi in rehabilitated areas, the early stocking rates are high and the present rate of mortality of jarrah in the rehabilitated areas is low. A co¬ operative research program with CALM, Murdoch Univer¬ sity and Edith Cowan University is identifying and propa¬ gating jarrah plants that are known to have a high resistance to P. dnnaoiooti. Another major research program has been initiated to gain a better understanding of the factors affect¬ ing the survival of jarrah in revegetated bauxite mined areas. These programs should lead to greater certainty on the long term survival of this key species. At Eneabba, Banksia hookcrana, B. atteniiata, B. leptophylla and B. catidoUeatia are propagated in the company nursery and planted into the rehabilitation. Successful rehabilitation has been achieved despite these species being very suscepti¬ ble to Phytophthora. The banksia seedlings are not planted in high risk areas such as drainage lines. This mimics the natural situation where banksias are usually restricted to the dune ridges and slopes and are not found in seasonally wet depressions. The results of the monitoring programs suggest that Phytophthora species are not having a major impact on the establishment of key plant species in the revegetated mineral sand and bauxite mined areas. 156 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Figure 4. Species richness of five rehabilitated bauxite mined areas monitored one, two and five years after mining (Ward, unpublished data). 2 Clematis pubescens “ Cyathochaeta avenaceae D Eucalyptus margmata D Hakea lissocarpha D Hovea chonzemifolia Q Xanthorrhoea preissii I P.citricola Figure 5. Mean proportion of seedling emergence of six species in a revegetated bauxite mined area monitored six weeks after seeding (from Woodman 1993). Emergence presented as least square means of arcsine transformed data with standard error bars. Seedlings C. pubescens and X.preissii were not found. 157 Journal of the Royal Society of Western Australia, 77 (4), December 1994 l«82 1913 liai 1»e» l»8« 1887 1888 1889 1980 IB8I ‘.893 1883 YEAR 0^ REHABllirATtON figure 6. Mean species richness of rehabilitated mineral sands mining areas, one year after establishment and as monitored in 1993. Establishment of high species richness The objective of the present rehabilitation techniques for bauxite-mined areas is to establish a vegetation with a high species richness, similar to that of the adjacent forest. Some species are known to establish well from the seed present in the topsoil, other species establish from seed spread on the pits during rehabilitation. The impact of Phytoplithora spe¬ cies on plant species richness is difficult to assess. Many factors affect plant establishment; plant disease is just one of many. Present levels of species richness in rehabilitated bauxiteminedareasarehigh(Fig5). Initiatives are underway at all mines and Alcoa's nursery to increase species richness. Very few deaths of susceptible species are found during inspections of rehabilitated areas. However, only the death of older plants, of easily observable size, would be found. Phytophthora species may kill the seed or the young (<6 weeks) seedlings of some species. Death at this stage would ^ unnoticed. Preliminary studies of six plant species found that P. cinnamomi had no significant (P=0.05) effect on the emergence of seedlings in two rehabilitated mined areas (Fig 6 ). However, P. dtricola did significantly reduce the emer¬ gence of two species (Woodman 1993). For RGC, the achievement of a high species richness (average of six plants per square metre) is a formal comple¬ tion criterion for their revegetated mined area at Eneabba. Species richness is monitored every year. Results to date indicate that there is a consistent increase in species richness over time and that the revegetated areas are developing a ^hness comparable to the unmined vegetation (Fig 7). Therefore, the impact of Phytophthora, whilst it can be dra¬ matic in isolated infections, does not appear to be limiting the overall success of the rehabilitation program at Eneabba. More research and monitoring is required before the iinpact of Phytophthora species on the species richness of rehabilitated areas can be fully assessed. However, early indications are that high species richness can be obtained despite the presence of Phytophthora species. Conclusion Alcoa and RGC operate large mines in a region of Western Australia where the native vegetation is susceptible to dis¬ ease caused by Phytophthora species. The challenge to both companies has been to develop and implement disease management programs which ensure that mining has mini¬ mal impact on the health of the native communities and the revegetated mined areas, while minimising the financial cost to the companies. The present financial cost of the pro- gra mmes is signi ficanI bu t the resu I ts to date indicate tha 11he disease management measures are successful. However, monitoring of their effectiveness needs to continue. Both companies accept that these costs are part of the essential costs of mining in these regions of the State, but research and development will continue to seek more cost effective dis¬ ease control measures. References Carswell L 1993 Fungi associated with plant deaths in rehabilitated bauxite mines. Murdoch University, Honours Thesis. Colquhoun IJ 1992 Alcoa's Research Direction. In: Dicback - What is tht future? (ed M | Freeman, KHarl & M Ryiill) Northern Sand plains Diebac Working Party, Perth, 15-21. Davison E M & Shearer B L1989 Phiftophthora species in indigenous forests in Australia and Nev\' Zealand. New Zealand Journal of Forest Science 19:277-289. Hill T C J1990 Dieback disease and other Phytophthora species. The Northern Kwongon. In: Nature Conservation, Landscape and Recreation Values o the Lesueur Area. Environmental Protection Authority, Perth, Bulleh'i 424:89-98. Nichols O G, Carbon B A, Colquhoun 1 J, Croton | T & Murray N N J 1^^^ Rehabilitation after bauxite mining in south-western Australia. Lan scape Planning 12:75-92. Podger F D 1972 Phytophthora cinuaniomi, a cause of lethal disease in hidifi cnous plant communities in Western Australia. Phytopathology 62:9/ 981. Podger F D, Doepcl R F & Zentmyer G A 1965 Association of Phytophth^^^^ cinnamomi with a disease of Eucalyptus marginata forest in Western Aus tralia. Plant Disease Reporter 49:943-947. Shearer B L1990 Dicback of native plant communities caused by Phytophth^^^'^ species - A major factor affecting land use in south-western Austr*' Land and Water Research News 5:15-26. Shearer B L & Tippett J T1989 Jarrah dieback: the dynamics and J of Phytophthora cinnamomi in the jarrah {Eucalyptus margiiiata) forests soulh-weslernAu.slralia. Department of Conservation and Land Manage ment, Perth, Research Bulletin Number 3. Ward S C, SlessarG C & Glenister D J 1993 Environmental resource mentpracticesofAlcoa of Australia Limited. In; Australasian Mining^ ^ Metalurgy (ed J T Woodcock & J K Hamilton) Australian Institute u Mining and Metalurgy, Melbourne, 104-108. Woodman G 1993 Damping-off of indigenous jarrah plant Phytopththora chniamomi and PhytOf)hthora dtricola in bauxite pit rehabi tion in the northern jarrah forest. Murdoch University, Honours The- 158 Journal of the Royal Society of Western Australia, 77: 159-162, 1994 Threats to flora-based industries in Western Australia from plant disease R T Wills^ & C J Robinson^ ’Department of Conservation and Land Management, Science & Information Division, WA Herbarium, PO Box 104, Como WA 6152 ^162 Serpentine Rd, Albany WA 6330 Abstract The flora of south-west Western Australia is known internationally for its rich array of species and its uniqueness. This floral wealth attracts many visitors, supports a thriving export trade in cut flowers, and contributes to one of the highest rates of honey production in the world. Many of these plants also have outstanding value or potential for amenity horticulture and floriculture. More recently, the importance of bioresources was highlighted by the discovery of a chemical, derived from a species of Conosperntum, capable of inhibiting Human Immunodeficiency Virus (HIV). The Proteaceae are a key element of south-western Australian ecosystems, and are a key resource of al 1 of these industries. Banksin and Dryandra are of particular importance for their attractive blooms, and are heavily exploited for their inflorescences, foliage and seed, and offer an important source of nectar. The destruction of large stands of these disease-susceptible species by a combination of aerial canker and Phyfophthora diseases, clearing, fire, and other disturbances could cause a significant financial loss to all flora-based industries. Notably, these flora resources are used with little return to the State for their conservation and management. An attempt is made here to quantify the financial impost of such losses, and a case is made for the need to invest in the protection of these bioresources through appropriate management. Introduction The ecological impact of plant pathogens has been well documented (see Shearer 1994 and Wills & Keighery 1994), but the economic implications for flora-based industries are poorly understood. The destruction of large stands of sus¬ ceptible species by various diseases, particularly where other disturbances such as clearing, fire and weed invasion act in concert, could cause a significant financial loss to all indus¬ tries reliant on the native flora. This brief review highlights Some of the potential threats of plant disease to various flora- based industries, and focusses in particular on the family Proteaceae, a key element of south-western Australian eco¬ systems, often a key resource of these industries, and highly susceptible to dieback disease and aerial canker. Wildflower industry The wildflower industry in Western Australia employs about 150 people full-time, and up to a total of 200 people during spring and summer (Anon. 1992). The principal product is dried flowers, but fresh-picked flowers, seed, and resource for craft-based cottage industry (using flowers, nuts, seeds, and dead woods) are also significant. Western Australia dominates the Australian export market, provid¬ ing the majority of overseas sales (Castles 1993). In 1992/ 93, the wildflower industry in Western Australia darned about $17 million. Wildflower exports contributed *nost, earning about $12 million (57% of all Australian wild ^yniposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia. Meld on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. ® Royal Society of Western Australia 1994 flower exports) with about $4 million of this generated from bush-picked flowers. Currently 29% of bush-picked produc¬ tion comes from private land and 71% from Crown Land (Anon. 1994). The Proteaceae are an important resource for the wildflower industry. In 1982, five of the 10 most heavily exploited genera were of the family Proteaceae (Burgman & Hopper 1982). Species of Banksia and Dryandra made up 17% of the stems picked, with two species {B. baxteri and B. grandis) the most heavily exploited for foliage; Banksia spp. were also heavily exploited for seed, making up 16% of all seed harvested by weight and 23% by value. The impacts of plant diseases on the wildflower industry have been greatest in the south of Western Australia, east of Albany. In that area, commercial harvesting is based heavily on 6 . coccinea and B. baxteri. Both of these species have been severely affected by the root rot Phytophthora cinnawonii and more recently aerial canker. From the mid or late 1970's, the B. coccinea harvest was concentrated on unvested reserves at Gull Rock and Cheyne Beach (25 and 80 km east of Albany respectively); picking of B. baxteri was centred on Cheyne Beach. In 1980-81, these species contributed 516,500 and 212,133 flowering stems respectively, to the total of 13,814,000 in Western Australia (Burgman & Hopper 1982) with the majority of these (at least two thirds) picked from Crown land. However, by the late 1980's these areas had been degraded to such an extent, that they were closed to pickers by CALM and other agencies. The creation of illegal vehicle tracks throughout these once dense stands appears to have contributed to the introduction and spread of P. cinnamomi. The disease has spread rapidly through sites killing many susceptible species including 159 Journal of the Royal Society of Western Australia, 77 (4), December 1994 species of Banksia, and has resulted in major structural changes to the vegetation (see Wills & Keighery 1994). As these areas were degraded, wiidflower harvesting activities moved further east, concentrating on unvested reserves near Cape Riche. In assessing B. cocciuca and B. baxteri across their natural ranges from around Albany and east to Esperance, Robinson (1991) found the remaining stands of both species were seriously threatened by aerial canker and PhytophUiora dieback. The study highlighted that areas at Cape Riche, 100 km east of Albany, had numerous large P. cmiuuuomi infestations introduced as a result of the "bush bashing" of illegal tracks for wiidflower picking. The study also noted that there had been a surge in picking activity on private land in the late 1980's as many farmers sought an alternative income by harv-esting B. baxteri when returns from normal rural activities declined dramatically. The combined impact of picking activities and plant disease depleting available stocks led to a ban on the picking of both species of Banksia from Crown land by the Minister for the Environment (B. coednea in September 1991 and B. baxteri in March 1993), and has forced the supply of B. cocdiiea and B. baxteri from Crown land to private land. However, as many remnant stands of Banksia on private lands have also been destroyed by Ptn/tophtliora and aerial canker, there is now a strong move toward cultivation of both species. The volume of trade in B. baxteri was 665,000 stems in 1991, largely through picking from private land (Robinson 1991). Similarly, 60% of 305,000 stems of B. cocc/»c/j produced in 1991 was derived from plantations (Robinson 1991). No¬ tably, a recent study has suggested that remnant bush man¬ aged for sustainable harvest of wildflowers may, in particu¬ lar areas, produce a return per hectare as good as that obtained by traditional forms of cropping (ACIL 1993). In the long term, the cultivation of native flowers to supply the industry' should ensure better quality (particu¬ larly through horticultural selection of preferred forms) and reliable supply. Although Phytoplithora has been recorded within some cultivated stands around Albany for more than five years, its impact has not beenas devastating asin natural bush. The disease can be controlled by application of the fungicide phosphonate ("'phosphorous acid"; Shearer & Fairman 1991, Komorek et at. 1994, Hardy et al. 1994). Aerial canker disease may prove to be a greater problem in plantations. Canker fungi may be spread by cross infec¬ tion from secateur wound sites (J Bathgate, CALM, pers coniw), and it has been shown that the frequency of infection was greater through wounded tissue (Bathgatec/ ai 1994), so it is possible that wiidflower picking may accelerate the spread of aerial canker disease through Banksia stands. For¬ tunately, cross infection can be prevented or controlled by use of appropriate hygiene measures such as the application of fungicides on wounds and equipment. Plant disease has also impacted on the native seed indus¬ try through the destruction of natural stands of B. coccinca by P. dnnamomi. Currently, stocks of seed of this species are very low and demands are not being met by traditional sources (P Luscombe, Nindethana Seeds, pers connn). South Australia is an alternative seed source, but the well estab¬ lished cut flower industry in that state is reluctant to supply Western Australia. B. coednea is now also being established commercially in south-eastern Victoria. Although the trend is toward cloning of superior forms, there will always be a demand for seed to explore the natural variation within the species. Beekeeping Reliance of apiculture on native floral resources is unique to Australia; in Western Australia the honey crop is gained almost entirely from native plant communities. Commercial apiarists in Western Australia practise migratory beekeep¬ ing utilising flowering of various Eucalyptus species in for¬ ests in the south-west of Western Australia in late spring, summer and autumn, and relocate to the sclerophyll shrublands of the Northern Sandplain for the winter flower¬ ing (Wills 1989). In contrast, beekeeping in the rest of the world generally relies on agricultural crops (mainly leg¬ umes) as their major source of pollen and nectar, and there is generally no winter production (Grout 1949, Nye 1980). In 1992, Western Australia produced 2,264 tonnes of honey (Kelly 1993) worth about $2.25 million. While West¬ ern Australia is the second smallest producer, producing only 14% of the Australian honey crop, rates of production are the highest in Australia and among the highest in the world (Wills 1989). Honey bees tend to favour species which are either wide¬ spread and/or locally abundant, although some abundant species may not be visited (van der Moezel et al, 1987, Wills 1989). On the Northern Sandplain, Wills (1989) found that 93% of the total 125 species visited by honey bees were native woody perennials. The same study found that virtually all species of Hakea, GretnUea, and Banksia (Proteaceae) were utilised by honey bees. Two species of Proteaceae, in par¬ ticular Dryandra sessilis and Hakea trifurcata, are considered by beekeepers to be the most important species in the region. Given that P. dnnamomi is principally a pathogen of woody perennial plant species, that most species of Proteaceae are known to be susceptible to dieback disease, and that the Proteaceae are a significant resource for honey bees, then this plant disease poses a very serious threat to apicultural resources, especially in the Northern Sandplain area. Loss of native floral resources through land clearing has been iden¬ tified as one of the most important threats to the continued economic development of the honey industry (Anon. 1983, Anon. 1984, Blyth 1987). Indeed, that part of the honey industry reliant on available native floral resources may have aready reached an upper limit (Anon. 1984); any loss of native vegetation as a result of land clearing, fire, disease and other environmental perturbations will inevitably lead to a decline in apicultural production. Biodiversity industries The Convention on Biological Diversity (established in Rio de Janeiro at the Earth Summit, 1992) included a resolu¬ tion to ensure access to genetic resources for environmen¬ tally sound uses while affirming national sovereignty. The Convention highlights the increasing global awareness of the importance of genetic resources in the production of valuable pharmaceutical, industrial and agricultural prod¬ ucts. 160 Journal of the Royal Society of Western Australia, 77 (4), December 1994 South-western Australia is home to about 9000 species of plants, and as many as 2000 species may be susceptible to pjiytopthora (Wills 1993). The Proteaceae are a key family in \Vestern Australia with 618 species and subspecies, by far t|ie greatest concentration in the world; many of these have outstanding value for amenity horticulture and floriculture (I^amont, Wills & Witkowski, unpiibl ob$.\ and most are susceptible to dieback disease. One species of Proteaceae in the genus Cofwspenmim is of particular significance in the area of therapeutic drugs. Ex¬ tracts from this species have been found to contain chemicals that inhibit Human Immunodeficiency Virus (HIV) itj vitro. lu the spirit of the Biodiversity Treaty, the W A Government has negotiated an agreement with commercial interests which ensures proper protection of the plant and guarantees profits to the State if the chemical proves an effective treatment for HIV. Thediscoveryhasalreadyearned iheStateSl .7million, gild if successfully developed for market could potentially earn royalties in the order of $100 million per annum (Armstrong & Hooper 1994, Armstrong & Abbottpers. cowm.) The Department of Conservation and Land Management is establishing bioprospecting based on four principles: • the WA community receives an equitable share of com¬ mercial benefits derived from use of the State's biological resources; • research and development of WA's biological resources shold involve WA's scientific community; • development of biological resources must be sustainable; and • WA's biological resources must be protected and con¬ served. Tourism Tourists undoubtedly select destinations for a host of reasons, making it difficult to determine the actual contribu¬ tion of floral resources to earnings in the tourist industry. The total 1989/90 income derived from tourism in Western Australia was $2,785 million, estimated using the multiplier effect (Western Australian Tourist Commission 1991). In 1989/90 around 20% of tourist day-trip destinations in¬ volved a visit to a national park/reserve or a scenic drive (Western Australian Tourist Commission 1991). These ac¬ tivities involved .some interaction with and appreciation of the natural features of the surrounding landscape and par¬ ticularly the flora. Further, 1J .5% of 33,246,700 visitor nights in 1991/92 resulted from trips to view wildflowers (Western Australian Tourist Commission 1992). This puts an upper range of the potential annual value of floral resources to the tourist industry of about $280 million to $560 million. The Western Australian Tourist Commission found that 116,000 group trips were specifically undertaken to view wild flowers, with visitor expenditure of about $36 million (A Sands, pers comvi). Consequently, any factor, such as plant disease, which degrades this natural floral resource might be ex¬ pected to have a substantial negative impact on the economy of Western Australia. Or will it? The Albany area receives about 250,000 tourists per year (Western Australian Tourist Commission 1992); the Stirling Range National Park, famous within Australia for its wildflowers, received 209,000 visitors (tourists and locals) in 1992 / 3. Notably, the next most popular visitor destination in the Albany area with 180,000 visitors was The Gap, a natural rock feature above the ocean. A popular destination for visitors to the Stirling Range is Bluff Knoll, which received 36,000 visitors in 1992 /3, yet the flora of Bluff Knoll has been .severely affected by the spread of Phytophtlwra dieback disease from the lower slopes right up to the summit plateau. The jarrah woodland on the lower slopes has been .severely diseased and species such as the Giant Andersonia {A. axilliflora), Drx/nudrn moutaua ms and Banksia oreophila have been killed on the plateau. Banksia brozvnii which was once common is now absent from the summit and upper slopes. The dieback resistant sedge, Lepidosperma sp. has become far more common and domi¬ nant in large areas over the past ten years as dieback has altered the vegetation structure (G j Keighery, CALM, pers comm, Keighery et al. 1993, see also Wills 1993). However it is apparent that most tourists visit only once and do not appreciate the presence of dieback in the landscape. Most are still delighted by the flora and do not notice dieback disease which is so apparent to the trained eye. The 1992Springperiod (September-Decemberinclusive) accounted for 66% of visitors to the Stirling Range; up to 90% of the patrons to Stirling Range Caravan Park come from the Eastern States to see the flowers of the Stirling Range (G Souness, Stirling Range Caravan Park, pers comm). How¬ ever, other high periods of visitation simply coincide with school holidays (Conservation and Land Management 1992) in January (11%) and April (13%). Clearly, the flora is not the only attraction to the Stirling Range, perhaps not even the main attraction. Even so, publicity in eastern Australia or overseas of the impact of dieback on the flora could poten¬ tially have adverse effects on flora-based tourist interests in Western Australia. Downline Effects Decline in activity of any of these flora-based industries will have flow on effects to other service industries that cater for all of the above. Because operations using native floral resources are necessarily regionally-based, they can be quite significant in the economy of small towns, and this may in turn contribute immeasurably losocial values by supporting associated community infrastructures. Conclusions Clearly, the destruction of large stands of these dieback- susceptible plant species by a combination of disease, clear¬ ing, fire, and other disturbances could cause a significant financial loss to all flora-based industries. Protection of bioresources through appropriate management is not only in the interest of conservation, but carries with it financial benefits of the existing resource as well as retention of bioresources of yet unknown value. The wildtlower industry may eventually be able to obtain most of its requirements from plantation; beekeeping might also be able to link into the expansion of plantations for both the wildflower industry and also for agroforestry. But, wildflower-based tourism and biodiversity-based indus¬ tries will necessarily rely on a relatively pristinebushland for i 161 Journal of the Royal Society of Western Australia, 77 (4), December 1994 their continuation, and wildlfower plantations will still re¬ quire wild gene pools to maintain and extend varieties under cultivation. Recommendations Education of industry groups should be a part of licens¬ ing requirements for operation in wild populations of plants. All industries using natural bioresources tend to deny they are, or are ignorant of, contributing any significant ecologi¬ cal damage to the resource that sustains them. Managers and industry both agree that accurate statistical information provides a valuable resource for management; both groups must co-operate to ensure that information obtained as a part of licensing is both accurate and relevant, and that the results of its analysis are put in place. Industry groups should be encouraged to contribute to and assist in the management process. Acknowledgements: We thank S McEvoy (CALM Wildlife Branch) and A Sands (WATC) for access to unpublished data. We also thank K Atkins (CALM Wildlife Branch), H Allison, C Wills, G Friend and T Start (CALM Science «Sc Information Division) and for their comments on earlier drafts of this paper. References ACIL 1993 Making profits from farm bush. Wildlife Branch, CALM, Perth, unpublished report. Anon. 1983 Apiculture Workshop Papers. For the Standing Committee on Agriculture Hawkesbury College. New South Wales Animal Production Division, Department of Agriculture, Richmond, New South Wales. Anon. 1984 Economic analysis of the Australian Honey Industry. Australian Government Publisher, Canberra; Australia Bureau of Agricultural Eco¬ nomics Occ. Pap. Number 88. Anon. 1992 Wildflower industry response to management strategies for the south-west forests of Western Australia. Flower Export Council of Aus¬ tralia, Perth. Anon. 1994 The Australian wildflower industry - a review. Rural Industries Research and Development Corporation, Research Paper Number 94/9. Armstrong J A & Hooper K 1994 Nature's medicine. Landscope 9(4); 11-15. Bathgate j. Shearer B L & Rdhl L 1994 Ciyptodiaporthe sp: a new canker pathogen threatening Banksia coccinea. In: Handbook of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia (ed R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 19. Blyth J D 1987 Beekeeping and Land Management. Proceedings of a work¬ shop held at the Department of Conser\'ation and Land Management, Department of Conservation and Land Management, Perth. Burgman M A & Hopper S D 1982 The Western Australian wildflower industry 1980-81. Department of Fisheries and Wildlife, Perth, Report Number 53. Castle'S I 1993 Foreign Trade, Australian. Australian Bureau of Statistics, Canberra. Catalogue 5436.0. Conservation and Land Management 1992VistatsPolicy&ExtensionBranch, Department of Conservation and Land Management, Perth, unpubished report. Grout R A1949 The beekeeping industry. In; The Hive and the Honey Bee (ed R A Grout) Dadant & Sons, Hamilton, Illinois, 1-10. Hardy G E St), O'Brien P A & Shearer B L 1994 Control options of plant pathogens in native plant communities in south-western Australia. Jour¬ nal of tlie Royal Society of Western Australia 77:169-177. Kelly P C 1993 Livestock and Livestock Products Western Australian Season 1991-92. Australian Bureau of Statistics, Perth. Catalogue 7221.5. Keigher)’ G J, Brown A, Rose A & Thomson C1993 Stirling Range nature walks and drives. In: Mountains of Mystery - A Natural History of the Stirling Range (ed C Thomson, G P Hall & G R Friend) Department of Conserva¬ tion and Land Management, Perth, 155-171. Komorek B, Shearer B L, Smith B & Fairman R 1994 Phosphonatc offers a practical method for the control of Phytophtlwra cinnaniomi in native plant communities. In; Handbook of the Symposium on Plant Diseases in Ecosystems: threats and impacts in south-western Australia (ed R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 28. Nye W P1980 Beekeeping regions in the United States. In: Beekeeping in the United States (ed E C Martin and others) US Department of Agriculture Handbook, Number 335,10-15. Robinson C J1991 Conservation status and economic contribution of Banksia coccinea and Banksia baxteri. Wildlife Branch, Department of Conservation and Land Management, Perth, unpublished report. Shearer B L 1994 The major plant pathogens occurring in ecosystems of south¬ western Australia. Journal of the Royal Society of Western Australia 77; 113-122. Shearer B L & Fairman R 1991 Control of PhytophtUora species in native communities with phosphorus acid. In; Coaservation Biology in Aus¬ tralia and Oceania Programme. Centre for Conservation Biology, Univer¬ sity of Queensland, Queensland, 72. van derMoezel PG, DelfsJ C, Pale JS, Loneragan W A, & Bell DT1987 Pollen selection by Apis tnellifira in shrublands of the Northern Sandplains of Western Australia. Journal of Apicullural Research 26:224-232. Western Australian Tourist Commission 1991 Statistical Profile of Tourism in Western Australia. WA Tourism Commission, Research Division, Perth. Western Australian Tourist Commission 19921991/92Tourism Monitor. WA Tourism Commission, Research Division, Perth. Wills R T 1989 Management of the flora utilised by the European honey bee in kwongan of the Northern Sandplain of Western Australia. University of Western Australia, PhD Thesis. Wills RT 1993Theecological impact of P/iyfopfiffwrflci/inflmomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18; 145-159. Wills R T & Keighery G J 1994 Ecological impact of plant disease on plant communities. Journal of the Royal Society of Western Australia 77:127- 131. 162 Journal of the Royal Society of Western Australia, 77:163-168,1994 Management of access K Gillen’ & A Napier^ ^ Department of Conservation and Land Management, Serpentine Rd, Albany WA 6330 ^Main Roads Department, PO Box 6202, East Perth WA 6004 Abstract The management of access is critical in minimising the spread of Phyfophthora. Recreational activities on CALM-managed lands present much greater risks than those posed by commercial operations which are licence or permit based. This is because the latter are more strictly supervised, and non-compliance with licence conditions can have implications for ongoing activities. In our management of access we attempt to balance the competing demands for requirements of access versus the need to protect areas from introduction oiPhytophthora. With our present state of knowledge it is necessary to control access very strictly on some high value areas. This means total exclusion or permit based entry to some areas. It appears that in general the issue of management control of Phytophthora is still viewed by the public and local government as a State government problem which is mostly too hard for others to address. Main Roads (Western Australia) have procedures which review the cost/benefits of incorporating management in their programmes and make decisions on implementation based on risk and final cost. Main Roads will also assist Shires in developing dieback management techniques in road construction and maintenance. The management of access in relation to Phytophthora requires significant resources in planning such as the cost of management procedures and costs incurred by industry to meet standards imposed. Management options that can be implemented to minimise introduction or spread of the fungus on 2WD and 4WD gravel roads are: • improvement in the surface formation and drainage, • demarcate existing disease areas associated with roads, • manage maintenance according to hygiene standards, • use seasonal or permanent closure as a means of protecting areas, and • establish clean down stations at entrances to national parks and other areas of high conservation value. For foot access on managed paths: • limit activities with the potential to spread the fungus into dieback-free areas, • use techniques such as surface hardening or boardwalks to reduce spread of infected soil, • careful selection of alignment, • implement seasonal closure, • close permanently if high values are at risk, • use strategically-located boot cleaning stations, and • provide information. Introduction Other papers in this issue have provided a wide ranging overview of plant diseases in WA and the extent to which they are influencing our environment. The management of plant disease affecting native vegetation in WA has focused primarily on Phytophthora and in particular P. cinnamomi. This focus is due to the widespread distribution and the high level of impact the fungus is having on native vegetation throughout the South West. This paper is confined to the management of access in relation to plant disease, caused by Phytophthora spp. Access is recognised as being one of, if not the, crucial factor in the artificial spread of the fungus in the south-west Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 of Western Australia and it continues to be a critical and often contentious issue when considering the management objec¬ tives for land set aside for conservation. The issue is conten¬ tious because the management of access is really about the management of people and trying to accommodate their needs/ wants against land management objectives for an area. The definition of access according to the Oxford Dic¬ tionary is: "approach; (to) right or means of approaching or reaching". To many (West) Australians, access particularly on Crown lands has been considered an inalienable right. Where access didn't exist it was created. The age of the 4WD vehicle has provided equivalent motorised opportunity to match that previously enjoyed by horsemen. In many re¬ spects we are talking about a cultural ethos with which many landowners, local authorities and management agencies have had to come to terms in the last 20 years or so. Management of access can be looked at in terms of 1. lands managed by CALM (including State forest), and 2. lands other than those set aside strictly for the purpose of 163 Journal of the Royal Society of Western Australia, 77 (4), December 1994 conservation, i.e. the management of roadsides as well as other reserves and vacant Crown land set aside for a different purpose and managed by authorities/bodies other than CALM. Conservation lands Phytophthora/dieback is currently considered to repre¬ sent the greatest single threat to the conservation values in the south west of the State in the short term. In the longer term, other factors such as the influence of climatic change may be important (but difficult) to manage! As a result of this, CALM and non-govemment conservation groups are concerned that management should be directed toward preventing the introduction of the fungus into areas not currently infested, or minimising its spread where it is already present or carrying out some control measures to protect species or communities from the fungus. This sentiment is strongly presented in the CALM Policy (Conservation and Land Management 1991a) which states as it's objectives: • to minimise the introduction, spread or intensification of the plant diseases caused by Phytophthora species through¬ out the State, with particular emphasis on the south-west, • to monitor for Phytophthora activity in the remainder of the State, including tropical areas, • to undertake and support research into the disease and its control, and • to encourage the West Australian community to share our concern over the problem, and its management. These key objectives, however, must also recognise that one of CALM'S three primary programmes is recreation (Conservation and Land Management 1993a). Provision of access is obviously a significant component of this pro¬ gramme which can often be in conflict with other objectives. In addition to this there are also requirements to accommo¬ date access for a range of government approved activities, such as • timber production in State forests, • mining exploration and mining, • commercial operations apiarists, wildflower picking, • research activities, • fire protection needs, i.e. firebreaks, etc, and • construction and maintenance of roads and powerlines. The development of access carries with it other implica¬ tions for the management of disease. These include the source of road making materials, the dieback status of these materials, and drainage control. To deal objectively with the issues that arise in making decisions about differing needs for access, there is a require¬ ment for effective and practical guidelines. These are gener¬ ally provided by the Department of Conservation and Land Management's dieback policy, Regional Management Plans, Area Management Plans and Dieback Protection Plans, and are dealt with more specifically by separate procedures for miningand commercial activities such as timber production, bee keeping, wildflower picking and research activities. To be able to make decisions about access there is a need to have good information about where the disease is, the risk of introducing the fungus during an operation; chances of it surviving if it is brought in; the impact it may have. Some of these questions resolve themselves into the concept of 'haz¬ ard' which is defined as the final impact of Phytophthora on a site if introduced. Obviously in areas where the hazard and the risk of introduction are low the options to manage are somewhat broader compared to where the hazard is, sav high to very high. For the following discussion we propose to direct our comments to the area within the defined Phytophthora zone in the south-west, where we are dealing with a high to very high hazard. Recreation access The most complex and trying aspect of access manage¬ ment is that related to recreation pressures. This is because management strategies need the co-operation of a large number of mostly unsupervised visitors to be effective. Recreational access can generally be considered in terms of 2WD access, 4WD access, foot access, and other access, i.e. horses, cycling. 2WD roads Sealed roads, once in place, do not present a high (but there is some) risk to vehicle traffic spreading the fungus. If the area which is traversed by the road is dieback-free then management of roadside operations should be conducted with careful attention to maintenance practices. If the align¬ ment was infected prior to sealing then, intensification is likely due to roadside runoff and drainage. Unsealed roads can present a considerable concern for management since roads external to a reserve may be man¬ aged entirely differently to roads within a reserve. There¬ fore, with no guarantee of hygiene on roads outside of a conservation reserve, there is a considerable risk that the fungus will be either introduced or spread by vehicles mov¬ ing into the reserve. In parks. Gazetted roads not under the control of CALM present similar concerns. In the case where a gravel spur road leaves a sealed road, the options presented also apply. If the alignment is not infected initially then risk of infection would be low, particularly if construction was undertaken under strictly controlled conditions. The management options on existing roads are: 1. have 2WD roads in good condition and well drained, 2. identify disease areas and manage the road to limit the risk of picking up infected soil, fc. with use of culverts; raising of road, crowning of road, 3. conduct maintenance operations in dry soil with atten¬ tion to demarcating disease areas within catchment boundaries according to hygiene standards. (Conserva¬ tion and Land Management 1986), 4. impose seasonal access restrictions (such management options can be difficult to implement because of incon¬ sistency in weather conditions from year to year.), 5. impose closures based on conditions which present a high risk This can be difficult to implement in remote locations because of distance and unpredictable response of people who have travelled a considerable distance to get there, such closures must also consider people who 164 Journal of the Royal Society of Western Australia, 77 (4), December 1994 may already be at a site served by a road that is proposed to be closed, and (y. establish clean down stations at boundaries where roads are subject to different management regimes. Such options have been considered at places like the Fitzgerald River National Park, however the costs and practicalities of such facilities have discounted their use to date. In situations where: # there is no complementary management on existing roads that are serving a reserve, and # where dieback hazard is high to very high, and 0 dieback is in the general area, then there is a high degree of inevitability that the fungus vvill be transported into an area. This is because manage- tnent capability in keeping Phi/tophthora out of an area de¬ pends on the success of all phases of operations over a long ti me including the impact of changing personnel on continu¬ ity of work standards. This is therefore a system in which the j-isk of a breakdown of procedures is high. 4WD access If the fungus is known from nearby and conditions are suitable, then 4WD access is a particular concern because the inature of these roads usually means that drainage is a problem and therefore the risk of moving infected soil can be fiigh at particular times throughout the year. The options for management are: 1 . permanent closure and rehabilitation - if this is the best way of meeting management objectives, i.e. values are very high. 2 . continue use - if the alignment is already exhibiting the impact of the disease it may be possible to continue use if this does not place substantially more vegetation at risk. Additionally, if there is concern/risk in taking infected soil away from the road then control over the timing of access may be appropriate. 3 . seasonal closure is an alternative which can deal with the situation where access conditions and risk change dra¬ matically with change of season. 4. ''opportunistic'' closure which is event linked. This op¬ tion is good only if the area can be easily serviced by management, i.e. accessible immediately after the event which is likely to cause concern for access, and therefore enforceable. 5. Upgrading of 4WD standard roads often leads to 2WD standard. Foot access Unmanaged foot access can lead quite rapidly to erosion on slopes greater than 3%, (Land.s, Park and Management 1987). In addition management of foot access can be an importantissuewhereintroductionorspreadofP/n/fo/?//f/mm is a concern. There is a considerable amount of circumstan¬ tial and substantiated (from sampling) evidence to demon¬ strate the spread of PhytoplitJiora by foot traffic, particularly along the south coast. Once introduced to areas high in the profile, the potential for extensive damage is quite signifi¬ cant, as is ably demonstrated by the situation in the Stirling Range National Park. Management of foot access has been based on: 1. mapping of disease occurrence, 2. identification of areas apparently not affected, 3. limiting current activities which have the potential to spread the fungus into areas identified as apparently dieback free, 4. using simple techniques to reduce the risk of taking infected soil upslope (or further along a path) i.e. — improve the path surface, e.g. use of stone to harden surface; boardwalks to avoid high risk areas — use clean down stations, 5. identifying the best location for the alignment, i.e. one side of a ridge, 6. monitoring dieback status, 7. implementing seasonal closure if the risk warrants it, and 8. permanent closure of paths if high values are at risk. The key to reducing the risk in high hazard environments where it is necessary (or preferable) to provide a path is to • provide a good alignment, • provide a good walking surface, • ensure adequate drainage, and • engineer to minimise boggy / wet patches. Information and public feedback Management action must be supported by information to the users of lands to gain their support, understanding and co-operation in relation to managing access. Signs alone don't do the job and there aren't enough people on the ground to educate and enforce different measures. Actions undertaken to control access in conservation lands for the management of Phytophthora are not always popular when first introduced. This is because they inevi¬ tably conflict with existing/ traditional activities. Despite the volume of written material published and that presented through the television and radio media, the depth of under¬ standing of the public concerning the issue is very shallow. This is quite understandable when considering the difficulty in coming to grips with an invisible fungus that in most cases can only be recognised from where it's been {i.e. dead plants), the fact that impacts can vary from dramatic in the short term to incremental over a long time (and most don't see them anyway), and that for most people the issue doesn't directly affect them to any great degree. It really falls into the cat¬ egory of an "SEP" (Someone Else's Problem). This difficulty in coming to terms with what the presence of the disease may mean to plant communities and environment as a whole in even the short to medium term {i.e. 1-5 years) is clearly evident in reviewing the public response to draft manage¬ ment plans for conservation areas. Generally, public perception and response with regard to Phytophthora, its impacts and the measures taken to control its spread in areas managed for conservation seems to be: • the issue exists (some disagree) and • sure, we should do something about it, and • management actions are basically OK as long as they don't affect what I want to do! Access is always a contentious issue in the planning process, especially where a history of existing use is in evidence. An extreme view on the South Coast is that CALM 165 Journal of the Royal Society of Western Australia, 77 (4), December 1994 invented dieback so it could interfere with people's enjoy¬ ment of National Parks. There is also a view expressed by a minority that key conservation areas should be closed to any access. We do not have the answer to the question of "how do we overcome this?" At the moment, the depth of understand¬ ing and concern that individuals have over this issue is closely related to the extent that they are involved in it. If people don't want to know, or aren't interested, then it is unlikely any approach other than enforcement will be effec¬ tive. Education of the younger generation in schools is possibly the best option, but by the time they are old enough to influence the situation the issue is probably going to be all over. It really is a matter of the extent we want to protect and more importantly retain, the diversity and values of our native vegetation. Other activities in conservation lands As stated, the Dieback Policy really sets out how we should go about managing access. In addition, there are specific guidelines for other more commercially oriented activities. —Timber Production. The set of prescriptions in "Timber Harvesting in WA" (Conservation and Land Management 1993a) very clearly sets out the standards for reading and conditions under which access is provided in State Forests. — Parts of State Forest are still included in Disease Risk Areas where access controls are prescribed. These areas were originally gazetted for a period of three years to allow for mapping of dieback disease, during which time access con¬ trols were stringently enforced. This system is now being reviewed. — Apiary activities are addressed by a CALM policy statement (Conservation and Land Management 1992). This includes guidelines on how vehicle access is to be managed and who bears responsibility for costs incurred to ensure ongoing access availability. — Wildflower picking is conducted under a Commercial Purposes licence under the Wildlife Conservation Act which includes specific instructions on access, in particular on land tenures such as State forest. Wildflower picking is not al¬ lowed in Nature Reserves and National Parks. Management of illegal picking has been a major problem in the past and still continues to be an issue of concern. There are considerable risks associated with managing a wildflower industry based primarily on Crown lands de¬ spite a licensing condition specifying the use of existing tracks only. The indiscriminate creation of new tracks and access has lead to the recent removal of B. baxterii and B. coccitiea from the picking list, because of the threat that Phytophthora now presents to these species in all Crown lands. Research Access for research purposes must also conform to the standards that are applied to other land users. This can mean that research proposals have to be amended. These aspects are dealt with in assessing research proposals both for inter¬ nal and external research programmes. In the past there is no doubt that intensive research programmes have contrib¬ uted to disease spread. No group of users is immune from having the capability of being a vector of the fungus. Management activities Management operations and personnel similarly have the potential to spread the fungus and therefore regular training and adherence to procedures is essential. The Hy¬ giene Evaluation Test is a critical tool to ensure the right questions are asked about any proposed operation (Conser¬ vation and Land Management 1993b). Mining The State government's mining policy sets out the proce¬ dures under which all proposals are assessed. These proce¬ dures include referrals to CALM and the Minister for the Environment to consider whether standard conditions are adequate to address environmental concerns. Access (tim¬ ing, method, degree of disturbance) is a critical issue in the assessment of proposals. New access (grid lines, etc) created by legal operations can provide opportunities for unauthor¬ ised access by people who are unaware of the strict condi¬ tions under which such access was developed and utilised. The conditions which are applied to mining and petroleum operations, particularly exploration activities are quite strict and with respect to dieback controls, are from my experience usually well managed. Summary In respect to Conservation land then, there is a significant difference in the management of access for recreation as opposed to the more commercially-based operations. This is because the commercial operations usually involve small number of people who have been longer in the job, are better trained, are involved with localised areas, and are either licence or permit based which carries implications for non- compliance with conditions. They are usually supervised to some extent by CALM staff. The key point about management of access is that ah the various options • cost money to carry our according to the standards set, • need compliance to work, • need management presence/supervision, and • need to be regularly monitored. Much of the access network in conservation lands is managed in the absence of these points. We are not able to provide the money required or the supervision needed to ensure compliance. Management of access on other lands No other agencies currently manage lands with the objec¬ tive of controlling or minimising the introduction or spread of Phytoplitlwra through control of access. Those who do carry out some management include some of the mineral sands mining operations over their lease areas. 166 Journal of the Royal Society of Western Australia, 77 (4), December 1994 To date, few local authorities have been able to develop policies and address dieback issues in their planning of proposed road works despite the fact that a format for such a document has been prepared by CALM and provided to local authorities on request. Despite being approached through the country Shire Councils Association to develop an approach to dieback, it would appear that the issue for local government authorities is just, • too hard, and/or • too expensive, and / or • is perceived to be unnecessary by some. There are a number of difficulties confronting local au¬ thorities in dealing with this issue. They are: • recognition of the disease, • survey and sampling costs, • skills, • operational costs, • administrative hassles. The most active and structured program is that being developed by Main Roads WA (Napier 1992). With over 3000 km of roads within the dieback susceptible areas of the State, Main Roads has a large task with specific problems facing them in the management of roadside areas. However the department has the will and technical expertise to work towards dealing with the issue. Five categories which influ¬ ence the management of roadsides have been identified by MRWA: 1. uncontrolled access, 2. road drainage, 3. on going maintenance, 4. gravel supplies, and 5. dieback mapping. These various factors must be taken into consideration when assessing and planning an operation to see if the balance of ''benefits" from undertaking controls of Dieback are worth the costs and effort. Decision-making flow-charts have been developed to assist in the assessment of proposed works and the selection of relevant Dieback controls. Knowing the extent of the disease over the road network is a primary requirement for their program to proceed and contributes to the manage¬ ment of PIn/toplithora on more than just a local level. How¬ ever, application of dieback management to all possible activities which may spread the fungus is proving extremely difficult, e.g. hygiene procedures for maintenance grading over long sections of road shoulder. This is proving expen¬ sive and often impractical and it is difficult to identify any positive value from the work because of the unknown dieback status and history of most of the roadsides. As with all groups concerned with doing something to ameliorate the threat of Phytophtlwra, there is always a concern that unless more stakeholders are involved and show a similar readi¬ ness to make real efforts then the efforts of an individual group will be greatly jeopardised or worse, be a waste of effort. Case studies Stirling Range National Park A considerable proportion of the Stirling Range National Park has been affected by Phx/tophthoracinnanwwiAiisappar- ent that the combination of soils, rainfall and a diverse susceptible flora has provided a situation very conducive to the survival and activity of the fungus. Most vegetation types are severely affected. The wandoo woodlands, how¬ ever do not exhibit symptoms of the disease due to the lack of susceptible species. Over the last eighteen months considerable effort has been directed at identifying those areas apparently dieback free, particularly in the higher peaks. Current information suggests that few areas of protectable dieback free vegeta¬ tion exist on the higher peaks. There do however appear to be considerable areas, including some of the lower peaks that are apparently die-back free. It seems likely that those areas have remained free of Phi/toplithora because they have offered less of an attraction for bushwalkers and other activi¬ ties that have been conducted in the park. Various options have been considered to protect vulner¬ able areas from introduction of the fungus. The use of boardwalks and clean down stations such as at Mondurup Peak are one means of trying to minimise the risk of intro¬ duction of the fungus. The situation as presented for Stirling Range National Park is now being considered during the Management Planning process. The question of how to deal with access throughout the park is, as usual, complicated by the competing demands of park users. Fitzgerald River National Park The distribution of Phytophthora spp in the Fitzgerald River National Park was presented in the Fitzgerald River National Park Management Plan (Conservation and Land Management 1991b). This map was based on up to date information at the time and showed the distribution of both P. chutamomi and P. Jtiegasperma. The difficulty of recovering P. jnegaspcnua from apparent disease sites and the impact of the 1989/90 fires which affected many previously suspect sites resulted in many areas being identified as "suspect" in theplan, relying on future monitoringto clarify thesituation. Very wet years in 1992 and 1993 initiated widespread symp¬ toms of Phytophthora activity in both eastern and western ends of the park and this has been subsequently confirmed by sampling recoveries of P. wegasperma. It is now apparent that P. megaspertm is present extensively along some road sections in the Fitzgerald River National Park and that these infections are probably quite old. Management operations in the park, particularly since 1986, have been cond ucted under strict hygiene based on the premise that most areas were dieback free. The recent revela¬ tions have highlighted the difficulty of managing road sys¬ tems that were in place prior to vesting. In this situation we have identi fied spread of Phytapbtlwra from infections on old alignments that have been closed to traffic for over a decade. The implications for management is that any work on these road sections which may result in conditions suitable for the fungus (such as drains/culverts) will probably result in 'new' areas of disease expression. The intermittent behav¬ iour of the fungus also makes monitoring of operations much more complicated. The picture in the Fitzgerald River National Park is far from clear and highlights the need for better understanding of P. megasperma in this environment and its potential long term impacts. Journal of the Royal Society of Western Australia, 77 (4), December 1994 References Conservation and Land Management 1986 Dieback Hygiene Manual. Depart¬ ment of Conservation and Land Management, Perth. Conservation and Land Management 1991a Policy Statement Number 3. Phythophtlwra Dieback. Department of Conservation and Land Manage¬ ment, Perth. Conservation and Land Management 1991b Fitzgerald River National Park Management Plan 1991-2001. Department of Conservation and Land Management, Perth, Management Plan Number 15. Conservation and Land Management 1992 Policy Statement Number 41. Beekeeping on Public Land. Department of Conservation and Land Management, Perth Conservation and Land Management 1993a Annual Report. Department of Conservation and Land Management, Perth. Conservation and Land Management 1993b Dieback Hygiene Evaluation: User Guide. Department of Conservation and Land Management, Perth. Lands, Parks and Wildlife 1987 Walking Track Management Manual. Depart¬ ment of Lands, Parks and Wildlife, Tasmania. Napier A1992 Dieback Management on Western Australia's Highways and Main Roads in "Dieback, What Is The Future? Northern Sandplain Dieback Working Party. 168 Journal of the Royal Society of Western Australia, 77:169-177, 1994 Control options of plant pathogens in native plant communities in south-western Australia G E St J Hardy', P A O'Brien' & B L Shearer^ ^School of Biological and Environmental Sciences, Murdoch University, Murdoch WA 6150 ‘Department of Conservation and Land Management, Hayman Road, Como WA 6152 Abstract Control of plant diseases in natural communities can involve management practices such as hazard rating, hygiene measures, quarantine, chemical applications, plant breeding, biological control agents, and molecu¬ lar manipulations involving hosts, pathogens and beneficial microorganisms. This paper will examine traditional, immunological and nucleic acid-based methods for the detection, identification and control of plant pathogens and their application in native plant communities. Introduction The plant diversity of south-western Australia is unique, with some 9,000 plant species, many of which are endemic. As many as 2000 species may be susceptible to Phytophthora species (Wills 1993), and many more are susceptible to other pathogens. Natural ecosystems are important centres of bio¬ diversity and are important assets for tourism, recreation and conservation. The management of plant communities to maintain their intrinsic value includes disease control. Considerable worldwide research has concentrated on the control of plant diseases; however, the majority of this work concerns economically important crop plants. Crops are usually grown in monoculture on a broad scale under well-fertilised, weed-free, and often, irrigated conditions. The plants are usually selected lines, bred for uniformity in germination, growth, crop maturity, and yield; in addition, cultural conditions are usually reasonably homogeneous. This compares with the huge diversity of plant species, uneven distribution of individuals, genetic variability, soil types and micro- and macro-environmental conditions present in a natural ecosystem. There are many ecological differences between agricul¬ tural and natural ecosystems, and those that set them apart embrace the concepts of population size, density, and spatial distribution, genetic variability in host populations, and population continuity or predictability through time (Burdon 1993). In natural ecosystems, plants are adapted to their pathogens; those that are not adapted are replaced by those better adapted. If a plant becomes unusually plentiful, be¬ cause of favourable conditions, its parasites increase with it and reduce the fitness and subsequently the number of susceptible plants, which in turn results in a reduction in the numbers of the parasite. The population of an organism in a given ecosystem is under continuous adaptive selection, through interaction with other organisms, for greater fitness Symposium on Plant Diseases in Ecosystems: Threats and impacts in South*Westcrn Australia. Held on April 16, 1994, at Murdocli University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 (Cook & Baker 1983). It immediately becomes apparent that enormous complexities are involved in applying any control method in a natural, compared to a cultivated, system. Epidemics of plant disease in natural plant communities are relatively commonplace; examples include outbreaks of Albugo Candida and Peronospora parasitica and their effects on the survival of Capsella bursa-pastoris (Alexander & Burdon 1984), and flax rust Melampsora Uni on wild flax Limwt marginak (Burdon & Jarosz 1992). The occurrence of a plant disease epidemic in a natural community indicates that some aspect of the ecosystem is not in balance. A number of essential changes need to occur for disease outbreaks to occur (Cook & Baker 1983): 1) the pathogen is genetically homogeneous, (implies introduction) and highly virulent, in high inoculum density or is not in balance with its antagonists; 2) the abiotic environment is relatively more favourable to the pathogen than to the host and/or the antagonists; 3) the ho.st plant is genetically homogeneous, highly susceptible, and continuously or extensively grown (as in cultivated crops); and 4) the antagonists are absent or in low numbers, lack suitable substrate or the proper environment to function as antagonists, or are inhibited by other microorganisms. Therefore, in cultivated crops plant diseases can be en¬ demic or introduced. In a natural community pathogens are usually introduced or alternatively environmental factors have changed (due to man or nature), that predispose plants to a pathogen. Therefore, it is important to consider in detail the classical disease triangle (host-pathogen-environment), when considering a plant epidemic in a natural ecosystem. In Western Australia, most of the “natural" ecosystems have been changed in some way, through such influences as forestry, wildflower picking, fire management, mining ac¬ tivities, road building, bushwalkers, clearing farmland and hydrological changes due to many of these activities. The.se can all affect the biological equilibrium and result in disease outbreaks. It is often hard if not impossible to associate such changes with a disease outbreak, if the pathogen is proven not to be introduced. Therefore, our information base on our 169 Journal of the Royal Society of Western Australia, 77 (4), December 1994 knowledge of any potential endemic or introduced patho¬ gens needs to be extended. This includes their specific re¬ quirements for survival, means of dissemination, host range, infection processes and pathogenicity. These requirements are further influenced by climatic changes (on pathogen and host), by human activities (management, economic and casual), or by fauna activity (defoliators and borers). It can still be argued that many areas of the biology, ecology and pathology of most if not all pathogens (includ¬ ing P. cinnamomi) are not well understood and, until this is so, the appropriateness of many control strategies such as bio¬ logical and chemical control or burning are questionable. Therefore, the main priority in controlling a pathogen is to have a comprehensive understanding of its biology, ecology, pathology and host range. For example, although many assumptions are made about the importance of chlamydospores and oospores as survival structures, it is still to be conclusively shown for most Phytophtliora species that these survival structures are produced naturally in soils and plants and, if so, how long they survive and how readily they germinate. Considerable research has been done w vitro and in vivo on propagules that have been artificially pro¬ duced on or in rich artificial media and then placed into soils under various temperature and moisture regimes. How these survival structures relate to those that are produced naturally is not known. For example, oospores of P. citricola which were produced axenically in V8 broth cultures had very short survival times compared to those produced in non-sterile soil extracts. Those produced in non-sterile soil extracts survived many months (Hardy, unpublished data). In this paper control options have been divided into five main groups cultural, resistance, chemical, biological and molecular strategies. Cultural control options Cultural control options aim to restrict the spread and reduce the amount of inoculum. In natural ecosystems such as the forests, heathlands and the banksia woodlands of the south-west, one potential control strategy is manipulation of the environment with fire management. Once the biology and ecology of existing pathogens and the ecology of hosts are known, fire could be used to successfully reduce inoculum levels. However, one consequence of using fire may be to provide infection via wounding by other pathogens, particularly aerial ones. Quarantine measures are also effec¬ tive in reducing the spread of soil-borne plant pathogens, particularly for introduced pathogens, such as P. cinnamomi. For aerial pathogens, quarantine is of little value since they can be spread great distances by wind which cannot be controlled by quarantine. Cultural techniques should be explored for the control of plant pathogens in natural ecosys¬ tems, especially for those plants which are endangered. Resistance Induced resistance or immunisation Recently, immunisation (cross protection) against plant disease has been well documented for a range of pathogens in widely diverse plants and plant tissues (Kuc 1990). Plant immunisation has proved successful in expressing, or sensi¬ tising for expression, resistance mechanisms in plants which are considered economically important. Induced resistance essentially involves infecting plants with an avirulent isolate of the pathogen, or a non-related pathogen which induces a host resistance response. This in turn effectively stops the invasion of the pathogen in question. The induction of resistance in tobacco to tobacco mosaic virus with specific chemicals such as salicylic acid and methyl-2,6- dichloroisonicotinic acid has been shown (De Waard cf ai 1993). Such techniques may in the future be useful for the control of certain plant diseases in natural systems, such as the preservation of rare and endangered species. Plant breeding for resistance Breeding for resistance in natural ecosystems where there is a large diversity in plant species is not practical, due to associated costs of developing and introducing resistant plants. The only areas where it would be practical would be to preserve specific attributes of rare and endangered flora and in rehabilitation of severely denuded areas (mass col¬ lapse sites and rehabilitated mines). This topic is covered by McComb et al. (1994). Chemical control Chemicals were used to control plant diseases long before their causal agencies were known. They have been effec¬ tively used in intensive agronomicsituations, but the control of plant pathogens in native plant communities by chemicals is generally not practical due to costs of the chemical(s) and their application. Possibilities of phytotoxicity, resistance, effects on fauna and other beneficial fungi need to be consid¬ ered. It is possible that the use of fungicides can tip the balance in favour of opportunistic pathogens. In such in¬ stances, the use of fungicides at very low rates in conjunction with other factors may be beneficial. Innatural communities, systemic fungicides (which enter the plant, become generally distributed within it, and render the tissues resistant) to attack are the only real chemical option. Most systemic fungicides currently on the market are translocated in tlie apoplast (Manners 1993); therefore appli¬ cation of systemics is likely to result in their being exten¬ sively distributed, via the xylem, throughout the plant. Conversely, application of fungicides to the soil as drenches or seed dressings will not be practical in an extensive and diverse natural environment. However, this should not pre¬ clude their application in certain instances or stop further research on their use. Recently research has shown that the use of neutralised phosphorous acid, which is inexpensive, has low toxicity to plants and animals and has high mobility within plants, will have considerable value in the conservation of rare and endangered plant species (Shearer cf al. 1991). However, on a broad scale, costs of application, differing tolerances of plant species to this chemical, and side effects on inverte¬ brates and other fungi, need further research. It is possible that frequent use might lead to the selection of resistant pathogenic strains to the chemical (Cohen & Coffey 1986, Coffey 1991), although there is no evidence of resistant strains developing during the last decade of use (Guest & Grant 1991). In addition, Phytophtliora species differ in their 170 Journal of the Royal Society of Western Australia, 77 (4), December 1994 tolerance or susceptibility to the chemical, which could give a locally competitive advantage to other Phytophthora spe¬ cies in the area. For example, P. tnegasperma is relatively insensitive to the chemical (Dereks & Buchenauer 1987), and is present in areas where P, cmnawowi also occurs (Bellgard, pers. comm.), such as the Fitzgerald River National Park. Therefore, caution needs to be exercised before phosphorous acid is used extensively in an area. In addition, phosphorous acid does not eradicate the pathogen; it activates the plant defence systems which then stop the spread of the pathogen within tissues. Perhaps the best use of phosphorous acid would be its application for the preservation of rare and endangered plant communities. Additional research is re¬ quired to clearly determine the exact mechanisms involved in disease resistance after the application of phosphorous acid. Many fungicides have different effects on a fungus de¬ pending on the structures it comes into contact with. For example, a number of fungicides sold for the control of Phytophthora species have no effect on endogenously dor¬ mant oospores of P. citricola and P. megasperma, whilst hyphae are killed (Hardy cf ol. wipiiblished data). These oospores could not be induced to germinate in the presence of the fungicides, but would do so once residues of the fungicides had disappeared. It is possible that such effects might occur with other resting structures such as chlamydospores. Therefore, it is necessary to have an under¬ standing of the effects of a particular fungicide on all stages of a fungal life cycle before it can be used with complete confidence. Improved knowledge of fungal biochemistry should al¬ low a more rational choice of fungicides, and open the way for the appropriate resistant genes to be cloned and their role in metabolism and pathogenicity explored (De Waard et ai 1993). Biological control Biological control has been defined as "the reduction of the amount of inoculum or disease producing activity of a pathogen accomplished by or through one or more organ¬ isms other than man" (Cook & Baker 1983). The mechanisms of biological control can be grouped into two categories: I) the use of antagonistic microorganisms, either resident or introduced to reduce the pathogens population level; this includes avirulent or hypovirulent individuals or populations within the pathogenic species itself; and 2) protection of plant surfaces against infection, by genetic manipulation of the plant and the use of specific cultural practices (Cook & Baker 1983). There are three mechanisms by which biological control may operate in the infection court (rhizosphere or phylloplane): 1) parasitism and predation: there is active contact be¬ tween microorganisms which result in the degradation of hyphal walls or mycophagy of whole propagules; 2) amensalism: the biological control agent produces antibiotics or toxic metabolic by-products which inhibit the growth of the pathogen; and 3) competition: involves two or more microorganisms competing for the same limited resource, such as oxygen, space, nutrients and moisture. It is very likely that successful antagonists employ more than one of the above strategies. The major stumbling block for applying a biological control agent in the field is that the environmental niche into which it is being applied is most likely already occupied. Therefore, a niche for the biological agent needs to be estab¬ lished (Powell et al. 1990). Formation of a niche for the bioagent can be accomplished in a number of ways; the biological agent 1) could utilise a substrate not currently used by other microorganisms; 2) may be better adapted physiologically for a particular niche than are the microorganisms currently occupying it; 3) produces an antibiotic which through its activity cre¬ ates a zone of substrate possession around it; 4) the biological agent is applied with or after a physical or chemical treatment that decreases the indigenous microflora; and 5) is added to the environment in such a form that large amounts of new substrate give it a head start over the indigenous microflora. A successful biological control agent is likely to employ more than one of the above conditions. However, in natural ecosystems, as with the application of chemicals, costs of application will be high. There will be associated problems with how the control agents are applied, and in the case of soil-borne plant pathogens, how they can be incorporated into the soil profile where they will be effective. In addition, selection of biological agenl(s) that can survive and function across a broad environmental range (differences in moisture, temperature, pH, fertility, host range, .soil types and micro¬ environments) and plant communities would be difficult. It will be necessary to obtain a good understanding of the host- pathogen-biological control agent-environmental interac¬ tions before the biological control agents can successfully be applied in the field. There is a general misconception that a biological control agent should eradicate a pathogen, but even a balanced equilibrium between a pathogen and its antagonist(s) should be considered beneficial. dsRNA for the control of PJnjtophthora species. Many fungal species have been found to contain double stranded RNA viral genomes in their cytoplasm. These have assumed a great deal of importance in recent years with the discovery that these elements attenuate the virulence of the fungal species which cause chestnut blight (Cryphouectria parasitica) and dutch elm disease (Ophiostoma ulmi). In both of these diseases, dsRNA-containing hypovirulent strains are able to protect the host trees against attack by virulent strains. Protection occurs via transmission of the dsRNA element to the attacking virulent strain converting it to a hypovirulent strain. The dsRNA elements of C. parasitica are the best charac¬ terised and studied. Hypovirulent isolates of C. parasitica contain a number of dsRNA segments (Nuss & Koltin 1990, McDonald (ScFulbright 1991). The number of segments, size and sequence homology may vary between strains. Com¬ parison of dsRNA from an American and a European isolate of C. parasitica revealed that each contained a large dsRNA of 171 Journal of the Royal Society of Western Australia, 77 (4), December 1994 about 12 kilobases in size, and a number of smaller dsRNAs that were derived from the large dsRNA. The number, size, and concentration of the smaller molecules varied with the strain and stage of growth (Nuss & Koltin 1990). Polypeptide coding sequences occupy only a small part of the element. Ophiostoma ulmi isolates contain a specific set of 10 dsRNA segments ranging in size from 0.34 to 3.5 kb which are associated with transmissible hypovirulence. Not all seg¬ ments may be transmitted, and healthy isolates recovered from a diseased isolate were found not to contain segments 4, 7, and 10 (Nuss & Koltin 1990). The discovery of these elements in C. parasitica and O. ulmi, and their association with the hypovirulent state, spurred a search for similar hypovirulent elements in other species of phytopathogenic fungi with the hope that they could be used as biological agents to control disease. dsRNA elements have now been found in both highly and weakly virulent isolates of the wheat pathogen Gaeumamwmyccs graminis var tritici (Nuss & Koltin 1990). However in studies with hypovirulent isolates it was found that virulent isolates could segregate out and that these were free of dsRNA elements, whereas the hypovirulent isolates retained the dsRNA. Isolates of the fungus Helminthosporium inctoriae, which causes oat blight, have been found to contain dsRNA viruses; one is associated with hypovirulence (Nuss & Koltin 1990). Hammarcfa/.(1989)detectedamultisegmented dsRNA element in a hypovirulent isolate of Leiicostoma persooni, and showed that the elimination of these elements restored viru¬ lence. In an elegant series of experiments, Sonnenberg & Van Griensven (1991) showed that La France disease in Agaricus bisporus is due to the transmission of a multisegmented dsRNA from a diseased to a healthy isolate by hyphal anastomosis. DNA markers were used to show that there was no transmission of either nuclei or mitochondria be¬ tween the strains. InspiteoftheevidencethatdsRNA confers hypovirulence to fungal isolates there are instances where they may en¬ hance virulence, or have no effect. Tooley ct al. (1989) studied the distribution of dsRNA elements in isolatesofP/ii/fop/(t/zorfl infesians and found that 36% of Mexican isolates contained dsRNA. However there was no correlation between the presence of dsRNA and virulence. Studies with Rliizoctonia solani have reached different conclusions. Castanho & Butler (1978) isolated three segments of dsRNA from a diseased isolate of R. solani. Healthy isolates recovered from this diseased isolate by hyphal tip culturedid not contain dsRNA. In plate tests it was shown that the diseased isolate could protect plants against the healthy isolate. However, Finkler et ai (1985) compared virulent and hypovirulent isolates of R. solani from Israel, and found that only the virulent isolates contained dsRNA. Transmission of virulence to a hypovirulent strain was found to be associated with trans¬ mission of the dsRNA. More recently Bharalhan & Tvantzis (1990) found dsRNA in all isolates tested from diverse loca¬ tions in the USA and Canada. Isolates from 5 anastomosis groups (AG) representing a wide range of virulence were included in the study. There was a high degree of heteroge¬ neity among dsRN A's from the same isolate, or from isolates within the same AG. This was especially evident in AG4, dsRNA's from these isolates were highly specific for the isolate from which they came indicating a lack of horizontal transmission of genetic elements in this AG. Cross hybridi¬ sation did occuramongdsRN A segments from 3hypovirulent isolates belonging to AG's 2,3, and 5, suggesting that there may be a sequence involved in suppression of virulence. However, the results show that overall there is no correlation between the presence of dsRNA and virulence. Similar con¬ clusions were reached in a study of wheat infecting isolates of R. solarii AG8 in Western Australia (Yang et al. 1994), What are the prospects for the use of dsRNA elements in biocontrol of pathogens such as PIn/fophtlwral Theoretically the use of dsRNA elements for biocontrol of Phi/toplithora is possible. The biocontrol agent would colonise the same microhabitat and is subject to the same influences as the pathogen. Moreover, since the antagonist converts the patho¬ gen to an antagonist, the level of control will not decrease (provided that the dsRNA is not debilitating). Studies on the distribution of dsRNA elements suggest that they are wide¬ spread, and that there should be no trouble finding them in isolates of P. cinuamomi. However, not all of these would be expected to be hypovirulent, and in fact some may be hypervirulent. Hypovirulent elements also have the addi¬ tional disadvantage that they may debilitate the host strain making it less able to withstand competition from other microflora. This would result in the eventual disappearance of the strain from the environment. The results of previous studies suggest that we would be lucky to find a hypovirulent element which could be effectively used as a biocontrol agent. How do we decide which dsRNA elements weshould use? The elements should not severely debilitate the host strain, and should be capable of being transmitted to other strains from the same species despite incompatibility barri¬ ers between the strains. The elements should be stably main¬ tained in different genetic backgrounds. Finally, in deciding which elements are hypovirulent it is important to take into account the level of genetic variation between host strains. Hypovirulent factors are identified by comparison of the pathogenicity of dsRNA containing and dsRNA free iso¬ lates. However, in this regard it is essential that the isolates being compared are characterised by means of DNA finger¬ printing to ensure they are isogenic except for the dsRNA element. The application of DNA fingerprinting techniques has shown that isolates which look and behave the same are often quite different. Finally, if hypovirulence is shown to be an important form of disease control, there are the questions of how will it be introduced into the environment and how long it would take to function? These questions are particu¬ larly pertinent considering the huge and diverse areas of vegetation affected. In the event that we do not identify a suitable dsRNA element for biocontrol, we still have the possibility of using dsRNA elements as delivery vehicles for dominantavirulent genes. The.se genes could beartificially-constructed antisense versions of pathogenesis genes, or naturally-occurring fun¬ gal avirulent genes. Using cloned fungal pathogenesis genes we can construct antisense versions of these genes in the laboratory, and iasert these by gene splicing into dsRNA elements (to create a dsRN A*). T ransformation of the dsRNA* into a strain of the target organism would create a biocontrol agent. Expression of the antisense version of the pathogenesis gene would inhibit expression of the sense gene thereby attenuating virulence. The dsRNA* would be transmitted to pathogenic .strains in the same way as a naturally occurring hypovirulent element, and would convert those virulent strains to hypovirulent strains. 172 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Molecular strategies for disease management Molecular strategies available for disease control are pri¬ marily concerned with crops of agronomic importance. How¬ ever, the use of molecular strategies of disease control in native plant communities will in the medium to long term provide beneficial tools in the area of disease diagnosis. The use of molecular techniques for early diagnosis, genetic manipulation of the pathogen(s), biocontrol agents and the host(s) hold considerable promise. However, increased fund¬ ing into these techniques at the expense of traditional plant pathological strategies and ecological studies must not oc¬ cur. It is imperative that these disciplines occur in conjunc¬ tion with each other. Diagnostics The early and accurate diagnosis of plant diseases using molecular strategies, such as immunological techniques, and nucleic-acid based methods could become an integral managementstrategyofnative plant communities. Manage¬ ment of plant diseases is most effective if control measures can be introduced at an early stage of disease development. Reliance on symptoms is often inadequate in this regard, since symptoms often appear long after disease establish¬ ment. Although biological techniques of disease diagnosis are usually very accurate, they are slow and not amenable to large scale applications. Molecular techniques of diagnosis must be viewed as management tools, to be used in conjunc¬ tion with other diagnostic procedures, knowledge of the host, and an understanding of the ecology of the disease and the biology of the pathogen. For example, a pathogen may be detected in a locality but not cause disease due to one or more of the host, pathogen or environmental factors are not opti¬ mal for the disease to occur. Molecular biology now provides rapid, specific, and sen¬ sitive techniques for detection of some plant pathogens. They will in the future become important early diagnostic tools for the early identification of plant pathogens. Immunoassays in plant pathogen detection Methods such as serological assays for pathogens, par¬ ticularly viruses have been available for many years. Immu¬ nological assays include enzyme linked immunoabsorbent assays (ELISA), immunofluorescent assays, monoclonal and polyclonal antibody assays. Immunoassays have the poten¬ tial to detect and quantify pathogen propagules in soil and other substrates. The role of an immunoassay is to reveal the presence of specific complexes between the antibody and antigen, that are unique to the pathogen. Immunological techniques can aid successful plant protection since they permit the early detection and correct identification of im¬ portant pathogens. As many fungicides are specific only to certain pathogens or groups of pathogens, immunodiagnosis can help in the selection of the most appropriate fungicide treatment (Fox 1993). Immunological techniques can be used to quickly and accurately recognise and identify those patho¬ gens with variable or latent symptoms on the host plant. Nucleic-acid hybridisation based detection of pathogens At present the exploitation of nucleic acids, DNA and RNA, in practical methods for the detection and /or identifi¬ cation of plant pathogens is in its infancy, and it will be a number of years before such methods will be of practical benefit. This is in comparison to the use of molecular meth¬ ods in clinical pathology, or the use of immunological meth¬ ods in phytopathology (Fox 1993). However, the potential advantages of this technology are overwhelming and it is inevitable that their adoption will be widespread. Nucleic- acid hybridisation depends on the high degree of specificity inherent in the pairing of nucleotide base sequences. This specificity allows the technique to be used for diagnostic purposes. The detection of plant pathogens via their nucleic- acids has two major advantages over rival technologies (Fox 1993). Firstly, all viable propagules (virus particles, spores, mycelium, etc.) contain the entire nucleic-acid complement of the organism. The presence of the nucleic-acid sequences is not altered by development or by response to environment or by the host. Antigens, in contrast may only be present at certain points in an organisms life cycle. In addition, the ability of the polymerase chain reaction (PCR) to detect one molecule of a particular sequence, conveys the ability to detect just one viable cell of the pathogen. This ultimate level of sensitivity obviates the need to culture pathogens prior to identification. Secondly, the identity of an organism is the direct result of the expression of its nucleic acids into protein and RNAs. Thus detection of a nucleic-acid sequence is simultaneously a positive identification. The nucleic-acid sequences in pathogens vary in their homology to sequences in other organisms. Thus it is possible to use nucleic-acid based methods for different levels of discrimination. For example, probes could in theory be designed which detect all fungi, or all ascomycetes, or all powdery mildews, or barley mildew or a particular pathotype in which one might be interested (Fox 1993). Despite these positive features, nucleic-acid based meth¬ ods are viewed with suspicion by many plant pathologists. They believe the methods are expensive, complex, slow and involve hazardous chemicals. It is therefore a challenge to plant pathologists to develop cheap, reliable methods suited to routine laboratories and even the end user (forester, farmer etc.). The techniques available include dot-blot as¬ says, non-radioactive labels, restriction fragment length polymorphisms (RFLPs), nucleic-acid probes, cloned probes and synthetic probes. The polymerase chain reaction (PCR) This is a comparatively new method which relies on two specific DNA primers, a thermostable DNA polymerase and temperature cycling to amplify discrete regions of DNA. It is extremely sensitive with the theoretical potential to detect a single target molecule in a complex mixture without using radioactive probes; and it is rapid and versatile (Henson & French 1993). Unlike serology, synthesis of hundreds of different PCR primers generates costs comparable to those of developing only a few monoclonal antibodies. PCR is capa¬ ble of quantifying relative differences as well as absolute amounts of scarce target DNA or RNA sequences. The quantification of plant pathogens in diseased plants is pos¬ sible, since changes of inoculum levels in soil or plants can be monitored by PCR. This can help predict the potential sever¬ ity of the pathogen and assist in control decisions. PCR has considerable potential in epidemiological stud¬ ies. It has the ability to be applied to studying disease resistance and determining at what stage of pathogenesis a 173 Journal of the Royal Society of Western Australia, 77 (4), December 1994 pathogen is inhibited. PCR can be used to estimate the biomass of unculturable microorganisms or obligate biotrophs. Microbial detection methods can be improved by combining PCR with antibody binding, this also gives a better indication of microbial viability. PCR is already being used to advance studies of host-pathogen interactions, such as for Erwinia (Blakemore et al. 1992), Pyrenophora (Reeves & Ball 1991) and Lqytosphaeria maculans (Goodwin & Annis 1991). PCR could be used to construct pathogen genomic or cDN A libraries, or could be used to construct libraries of host or pathogen genes that are differentially expressed during the infection process. As PCR methods for detection of pathogens become available, it will be possible to focus research on studying pathogen populations, biology, ecology, variability and host- pathogen interactions (Henson et al. 1993). An effective diagnostic test must be simple, accurate, rapid and safe to perform, yet sensitive to avoid 'false positives'. Development of resistant plants by genetic engineering In recent years plants resistant to viral diseases have been developed by genetic engineering. This has been achieved by inserting viral genes into the plant genome so that their expression inhibits the normal viral life cycle. This same strategy can be applied to a wide variety of plant viruses. Fungi are much more complex, and use a wider variety of mechanisms to achieve their colonisation of the host plant. Nonetheless we are beginning to identify the mechanisms used by fungi to infect plants. Once we have identified these mechanisms we can engineer the plant to these and thus confer resistance to these plants. One approach is to use fungal inhibitory proteins. This does not depend on knowing the mechanism of infection. Genes for the proteins are inserted into the plant genome where they are expressed. In two separate studies plants resistant to Rhizoctonia solani have been engineered by insert¬ ing the chitinase gene from bean (Broglie & Broglie 1993), and the barley ribosome inhibiting protein gene (Logemann et al 1992) into the plant genome. Both of these mechanisms on their own conferred higher levels of resistance to R. solani. These studies demonstrate the utility of the general ap¬ proach of using antifungal proteins to engineer resistant plants. Many plant species contain antifungal proteins, be they lectins which bind to the fungal cell wall and inhibit growth, or enzymes which degrade the fungal wall. Another approach would be to inhibit the enzymes pro¬ duced by the pathogen and which are necessary for infection (Kotoujansky 1987). Many soft rot pathogens such as Enoinia produce pectic enzymes which breakdown the pectic sub¬ stances producing the typical soft rot symptoms. Most of these enzymes produce oligogalacturonide degradation products which induce a defence response in the plant and thereby limit infection. Highly virulent isolates produce additional compounds which degrade these elicitors and prevent induction of the plant defence responses. These enzymes can be targeted by the use of polygalacturonase inhibiting proteins (PGiP) which have been described in all dicot species (Hoffman & Turner 1984). Potentially, by trans¬ ferring the gene for such an inhibitor to the plant genome, the inhibitor would prevent colonisation by highly virulent strains. A variation on this theme would be to use antibodies, or more correctly plantibodies. A number of studies have now demonstrated that we can now synthesise antibodies in plants (Pluclzltun 1992). This can be achieved by transforma¬ tion with antibody synthesising genes. Antibodies against fungal extracellular enzymes or against the fungal wall components could inhibit infection of the host. Similarly, antibodies against detoxifying enzymes could be used to achieve resistance. The fungus Nectria hacmatococca is a patho¬ gen of peas. The fungus produces the enzyme pisatin demethylase which inactivates the phytoalexin pisatin pro¬ duced by the host (Schafer et al 1989). The production of plantibodies against this enzyme would enable the host to limit colonisation by the fungus. Genetic engineering of resistant plants can be achieved by using genes derived from the pathogen. Incompatible reac¬ tions between the pathogen and the host are determined by avirulent (avr) genes in the pathogen which act with genes in the host to induce defence responses and limit the spread of the pathogen. One novel idea suggested is to take the avr gene and place it in the plant genome by transformation technology (DeWitt 1992). The gene is modified in such a way that any infection would trigger it's expression. This in turn would induce expression of the host defence responses. Thus instead of the pathogen carrying the avr gene, the host would carry it and it could protect against highly virulent isolates. The phenomenon of induced resistance offers great po¬ tential for engineering resistance against a range of fungal pathogens. This is a non-specific resistance induced by infec¬ tion of the plant. For example, the infection of the lower leaves of tobacco with TMV leads to resistance against fungi, viruses, bacteria, and insects. Concomitant with this resist¬ ance a large number of proteins are synthesised within the plant, these include glucanases and chitinases (Garner et al. 1992). Resistance against Phytophtfwra infestans has been achieved by this mechanism. Potentially if we can identify the genes involved in the induced resistance we can develop strategies to turn on these responses rapidly in the event of an infection. The defence responses in the host include activation of lignin and phytoalexin biosynthesis. The level of phytoalexins may be modified to confer resistance to fungal pathogens. It has been found that the activity of the enzyme isoflavone-2- hydroxylase regulates the amount of phyloalexin synthesis in chickpea cell suspensions. Increasing the activity con¬ ferred resistance to Ascochyta pisi (Lamb et al 1992). Greater levels of resistance could be achieved by modification of the gene for increased expression. Plants can be modified to produce new types of phytoalexins, thus conferring resistance against normally pathogenic fungal species. The phytoalexin resveratrol is synthesised by stilbene synthase in a single step from p- coumaroyl CoA and malonyl-CoA (Lamb et al. 1992). Intro¬ duction of the stilbene synthase gene into tobacco caused the synthesis of resveratrol in tobacco. There are now a number of strategies emerging for the genetic engineering of plants resistant to fungal diseases. The impetus for this work is the lack of alternative control measures. In many cases there are simply no natural sources 174 Journal of the Royal Society of Western Australia, 77 (4), December 1994 of resistance, and no effective fungicide treatments for con¬ trol of the pathogen. Integrated control Integrated control will become increasingly important, especially as our understanding of interactioasbetween host plant, pathogen, biological control and environment im¬ proves. However, the range of strategies available for inte¬ grated control in native plant communities are not as diverse as those in broad acre agriculture or horticulture. It is possi¬ ble that an advantage of integrated control is the synergistic effect of combining practices (Coffey 1991). In intensive agriculture, many of the success stories of pathogen control include the incorporation of breeding strategies and chemi¬ cals, cultural and biological applications. However, inte¬ grated control is not a panacea for pest and disease control but an ecological approach to maintaining plant health (Kendrick 1988). Existing control strategies of pathogens in natural plant communities in south-western Australia phytophthora species The main control strategies for Phytophthora species in Western Australia have centered on P. cimmmomi and in¬ clude hazard rating, assessment of risk, hygiene and quaran¬ tine measures, unfavourable to the pathogen but enhancing host resistance (Shearer & Tippett 1989). Hygiene includes planning, training, and exclusion methods such as strategic road placement, washdown facilities between infected and non-infected sites, confining activities such as logging to periodsofleastriskand strategic road management (Shearer & Tippett 1989). Management strategies which are unfa¬ vourable to the pathogen can also be effective. These include the manipulation of understorey by reducing Banksia grandis by fire or stump poisoning which reduces a potential source of inoculum. Fire can also stimulate Acada growth, specifi¬ cally A. pulchella, which has been shown to suppress p. dtimmomi activity. Chemical and biological control can also be used and have been discussed previously. Until it is clearly established whether other Phytophthora species are endemic or not, strategies used for the control of p. dnnamomi may or may not be effective for these other species. Currently, detailed studies are being undertaken on p. dtricola (F Bunny, pers. comm.) and P. megasperma (S Bellgard, pers. comm.). Both the.se species are homothallic and undergo sexual recombination which can add addi¬ tional complexities to control compared to P. cinnamovii vvhich is heterothallic. P. dtricola is most frequently isolated from areas of disturbance such as mine rehabilitation, log landings and drainage lines (F Bunny, pers. comm.). Oospores have always been a.ssumed to be important survival struc¬ tures formed in soil and host plants. However, it is only recently that they have been shown to be produced in nonsterile soils. Oospores are now known to survive for a rninimum of months in field soils which indicates their importance as survival structures. It is now almost certain that P. dtricola is endemic in the jarrah forest (F Bunny, pers. comm.), and this is likely to be true for other areas of the south-west. Therefore, it is extremely important to increase research into the ecology and pathology of P. dnnamomi. The importance of Phytophthora species as damping-off pathogens in natural ecosystems must not be discounted. Recently, it has been shown that P. dtricola can behave as a post-emergent damping-off pathogen in mine rehabilitation (Woodman 1993). Therefore, additional research is required to ascertain whether P. dtricola and other Phytophthora spe¬ cies can cause damping-off, especially in areas of distur¬ bance (drainage lines, road verges, salt effected, waterlogged, fire damaged). It is also important to consider that the ecology and pathology of a pathogen may differ in different environ¬ ments. This has been shown with P. dnnamomi in rehabili¬ tated mines. The mining process changes the soil environ¬ ment substantially. Recently, extensive excavations of 1-6 year old £. marginata growing in rehabilitated mines and exhibiting early symptoms of infection have clearly shown that P. dnnamomi infects the collar and lignotuber region in preference to roots. With time the pathogen moves down into the roots and up the stem. Over 30 jarrah trees have been excavated, and at no time has P. dnnamomi been isolated from roots of trees exhibiting symptoms of early infection (Hardy et at. unpublished data). The eradication of Phytophthora infections on woody plants by the use of chemicals has been shown to be almost impos¬ sible on plants growing in infected soils or container mixes, due to the formation of resistant survival structures such as chlamydospores and oospores (Ribeiro et al. 1991). In addi¬ tion, a range of Phytophthora species, including P. dnnamomi and P. megasperma in plantations of AF/cs procera, Pseudotsuga menziesii, A. magnifica var. shastensis and A. grandis in the north-western United States could not be effectively control¬ led (Ribeiro et al. 1991). In contrast, infection of apple trees with P. cactorum could be eradicated/cured with the sys¬ temic fungicides metalaxyl and fosety 1-Al however, repeated applications were required (Ellis et al. 1986, Orlikowski et al. 1986). Therefore, fungicide application in natural ecosys¬ tems is unlikely to be effective in the long term, even if the costs of the fungicides and their application is not taken into account. Stem and branch cankers The distribution of canker fungi has mainly been ignored in all natural environments of the lower south-west. This is despite the observed increased incidence since the 1970's of eucalypt die-back decline (Kimber 1980). A number of patho¬ genic fungi have been associated with cankers of stem and branches of forest trees in south-western Australia (Davison & Tay 1983), and in heathlands and woodlands (Mu rray et al. unpub. obs.). Recent work on Diplodina (anamorph, Cryptodiaporthe) canker on Banksia coednea suggests that this fungus is endemic (Shearer 1994) The outbreak of this fungus as a major pathogen indicates that some change in the biological balance of the system has changed. Schoeneweiss (1975) has associated disease caused by canker fungi to be aggravated by transient stress factors, such as excessive heat or waterlogging. Murray, Wills & Hardy (unpublished data) examined 1259 cankers on 508 native plants, 49 different genera of fungi were isolated, of which putative pathogens included Botryosphaeria ribis,Diplodia mutila (teleomorph Botryosphacria stevensii), Endothiella and a species of Diplodina. However, 175 Journal of the Royal Society of Western Australia, 77 (4), December 1994 B. ribis and D. jrtutila were isolated respectively, from 53 plant species in 24 genera and 23 species in 13 genera, which indicates their broad host range. In addition, a wide range of fungi were isolated from stem and branch cankers of plants grown on rehabilitated mines (Carswell 1993). Pathogenicity testing proved many of these to be pathogenic. It is likely that conditions in rehabilitated or other severely denuded sites (lack of canopy cover, high or low moisture levels and reduced plant diversity) could predispose plants to canker or other pathogenic fungi. Annillaria htteobubalina. Armillaria luteobubalina is a primary pathogen widely distributed throughout the south-western Australia; it is a native pathogen that infects a wide range of plant species from diverse families (Pearce et ol. 1986; Shearer & Tippett 1988). It has been isolated from wandoo, jarrah, and karri forests as well as throughout the coastal fringe from Cape Arid up to Cervantes (Shearer cf al. 1994). In the jarrah forest, the impact of this pathogen varies between plant community and climatic zone (Shearer & Tippett 1988). Control has been effective in certain instances with other wood rotting fungi such as Coriolus versicolor and Stereian hirsutum (Pearce & Malajczuk 1990, Pearce 1990). These wood decay fungi were shown to significantly reduce the invasion of Eucalifptus diversicolor stumps by A. luteobubalina. However, there are at present fewer options for the control of A. luteobubalina than those available for Phi/topjhthora. Conclusions It is necessary for extensive and detailed surveys to be initiated to increase our knowledge of tlie identity and incidence of pathogens causing disease in the natural plant communities of the south-west of Western Australia. De¬ tailed studiesof pathogen survival, reproduction and spread as well as host infection and susceptibility need to be under¬ taken. Each pathogen and its interactions with each of its hosts must be considered individually. In turn, these need to be related to interacting factors such as environmental changes, insect associations and the influence of human activities. A comprehensive understanding of the pathogens present, their life cycles and how they are influenced by environmental and human interactions will help ensure that our management of these native ecosystems is effective. Consideration must be made to ensure that management practices do not consider a few pathogens to the exclusion of others. References Alexander H M & Burdon) J 1984 The effect of disease induced by Albugo Candida (white rust) and Perotiospora paralitica (downy mildew) on the survival and reproduction of Capsella buna-pastoris (shepherd's purse). Oecologia 64:314-318. Bharathan N & Ta vantzis SMI 990 Assessment of genetic relaledness among double stranded RNA's from isolates of Rhizoctonia solani from diverse geographic origins. Phytopathology 81:411-415. Blakemore E, Reeves] & Ball S 1992 Polymerase chain reaction u.sed in the development of a DNA probe to identify Eru’inia steuwtiif a bacterial pathogen of maize. Seed Science and Technology 20:331-335. Broglie R & Broglie K1993 Production of disease resistant transgenic plants. Current Opinions in Biotechnology 4:148-151. Burdon J J 1993 The structure of pathogen populations in natural plant communities. Annual Reviews of Phytopathology 31:305-323. Burdon J J & Jarosz A M 1992Temporal variation in the racial structure of flax rust (Melampsora Uni) populations growing on natural stands of wild flax (Linum m;ir;^iim/c'): local versus metapopulation dynamics. Plant Pathol¬ ogy 41:165-79. Carswell L 1993 Fungi a.ssociated with plant deaths in rehabilitated bauxite mines. Murdoch University, Honours Thesis. Castanho B, Butler E E & Shepherd R J1978 The assodalion of double strand RNA with RhizocUmia decline. Phytopathology 68:1515-1519. Coffey M D1991 Strategies for the integrated control of soilbome Phytoplit/wra species. In: P/iytophthora (od J A Lucas, R C Shattock, D S Shaw & L R Cooke) Cambridge University Press New York, 411-432. Cohen Y & Coffey M D1986 Systemic fungicides and the control of oomycetes. Annual Review of Phytopathology 24:311-338. Cook R J & Baker K F 1983 The Nature and Practice of Biological Control of Plant Pathogens. American Phytopathological Society, St Paul, Minne¬ sota. Davison E M & Tay F C S 1983 Twig, branch and upper trunk cankers of Eucalyptus marginata. Plant Disease 67:1285-1287. Dereks W & Buchenauer H1987 Comparative studies on the mode of action of aluminium ethyl phosphite in four Phytophthora species. Crop Protection 6:82-89. De Waard M A, Georgopoulos S G, Hollomon D W, Ishii H, Leroux P, Ragsdale N N & Scl\winn F ] 1993 Chemical control of plant disea.ses: Problems and prospects. Annual Review of Phytopatliology 31:403-21. DeWitt J M G1992 The molecular cloning and functions of pathogenicity and avinilence genes of the tomato pathogen Cladosporium fulvum. Proc. I** Gur. Conf. Fungal Genetics, Nottingham University. Ellis M A, Ferree D C & Madden L V1986 Evaluation of metalaxy I and captafol soil drenches, composted hardwood bark soil amendments, and graft union placement on the control of apple collar rot. Plant Disease 70:24-26. Finkler A, Koltin Y, Barash I, Sneh B & D Pozniak 1985 Isolation of a virus from a virulent strain of Rhizoctonia solani Journal of General Vinology 66-1221- 1232. Fox R T V 1993 Principles of Diagnostic Techniques in Plant Pathology. International Mycological Institute, CAB International, Oxon, UK. Garner N, van den Elzen P 6c Cornelissen B J C 1992 The potential for the control of fungal disease in crop plants using gene transfer technology. In: Biotechnology International (ed R Grcenshields) Grosvenor Press Inter¬ national, London, 111-116. Goodwin P & Annis S 1991. Rapid identification of genetic variation and pathotype of Leptospham'a maculans by random amplified polymorphic DNA assay. Applied and Environmental Microbiology 57:2482-2486. Guest D & Grant B 1991 The complex action of phosphonales as antifungal agents. Biological Review 66:159-187. HammarS, Fulbright D W & Adams G.C 1989 Association of double stranded RNA with low virulence in an isolateot Leucostomapersooni. Phytopathology 79:568-572. Henson J M & French R1993 The polymerase chain reaction and plant disease diagnosis. Annual Review of Phytopathathology 31:81-109. Henson J M, Goins T, Grey W, Mathre D E & Elliot M L1993 Use of polymerase chain reaction to delect Caeuniannomyccs graminis DNA in plants in artificially and naturally infested soil. Phytopathology 83:283-287. Hoffman R M & J G Turner 1984 Occurence and specificity of an endopoly galacturonase inhibitor in Pisum sativum. Physiological and Moleculer Plant Pathology 24:49-59. Kendrick J R J B 1988 A viewpoint on integrated pest management. Plant Disease 72:647. Kimber P C 1980 Eucalypt diebacks in the southwest of Western Australia. In: Eucalypt Dieback in Forests and Woodlands (ed K M Old, G A Kile 6c C P Ohmart) C5IRO Division of Forest Research, Canberra. Kotoujansky A 1987 Molecular genetics of pathogenesis by soft rot Erwinias. Annual Review of Phytopathology 25:405-430. Kuc J1990 Immunization for the control of plant disease. In: Biological Control and Soil-borne Plant Pathogens (ed D Hornby) CAB International, Wallingford, 355-373. Lamb C J, Ryals ] A, Ward E R & Dixon R A 1992 Emerging strategies for enhancing crop resistance to microbial pathogens. Biotechnology 10:1537- 1445. 176 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Logemann ], Jach G, Tommerup H, Mundy J & Schell J 1992. Expression of a barley ribosome inactivating protein leads to increased fungal protection in transgenic tobacco plants. Biotechnology 10:305-308. MacDonald W L & Fulbright D W1991 Biological control of chestnut blight. Useandlimitationsof transmissable hypovirulence. Plant Disease 75:656- 661. Manners] G1993 Principles in Plant Pathology. 2^ Ed. Cambridge University Press, Cambridge. McComb J A, Stukely & I Bennett 1994 Future ecosystems—use of genetic resistance. In: Symposium on plant diseases in ecosystems; threats and impacts in south-western Australia (ed P C Withers, WA Cowling & R T Wills) Journal of the Royal Society of Western Australia 77; 179-180. Nuss D L & Kollin Y 1990 Significance of dsRNA genetic elements in plant pathognic fungi. Annual Review of Phytopathology 28:37-58. Orlikowski L B, Leoni-Ebeling M & Schmidle A1986 Efficacy of metalaxyl and fosetyl-A1 in the control of Phytophthora cactorum on apple trees. Zeitschrift fiir Pflanzenkrankheiten und Pflanzenschutz 93:202-209. Pearce M H 1990 In vitro interactions between ArmilUaria luleohubalina and other wood decay fungi. Mycological Research 94:753-761. Pearce M H & Malajczuk N1990 Inoculation of Eucalyptus diversicolor thinning stumps with wood decay fungi for the control of Armillaria luteobubalina. Mycological Research 94:32-37. Pearce M H, Malajczuk N & Kile G A 1986 The occurrence and effects of Armillaria luteobubalina in the karri (Eucalyptus diversicolor F Muell) forests of Western Australia. Australian Journal of Forest Research 16:243-259. Pluclzltun A1991 Antibody engineering: Advances from the use of Escherichia coli expression systems. Biotechnology 9: 545-551. Powell K A, Faull J L & Renwich A 1990 The commercial and regulatory challenge. In: Biological Control and Soil-borne Plant Pathogens (ed D Hornby) CAB International, Wallingford 445-463. Reeves J&Balls 1991 Preliminary results on the identification of Pyrcnop/iorfl species using DNA polymorphisms amplified from arbitrary primers. Plant Varieties and Seeds 4:185-189. Ribeiro O K & Linderman R G 1991 Chemical and biological control of Phytophthora species in woody plants. In: Phytophthora (ed J A Lucas, R C Shattock, D S Shaw & L R Cooke) Cambridge University Press, New York, 399-410. Schafer W, Straney D, CiufetH L, Van Elten H D & Yoder O C1989 One enzyme makes a fungal pathogen but not a saprophyte virulent on a new host plant. Science 246:247-249. Schoeneweiss D F 1975 Predisposition, stress, and plant disease. Annual Review of Phytopathology 13:193-211. Shearer B L1994 The major plant pathogens occurring in ecosystems of south¬ western Australia. In: Proceedings of the symposium on plant diseases in ecosystems; threats and impacts in south-western Australia (ed P C Withers, WACowling&RT Wills) Journal of the Royal Society of Western Australia 77:113-122. ShearerBL&TippettJT1988DistributionandimpactofAnm7/flrwJwfeofJHiJfl/nw in the Eucalyptus marginata forest of South-western Australia. Australian Journal of ^tany 36:433-445, Shearer B L &Tippett J T1989Jarrah dieback: The dynamics and management of Phytophthora cinnamomi in the jarrah (Eucalyptus marginata) forest of south-western Australia. Department of Conservation and Land Manage¬ ment, Perth, Research Bulletin Number 3. Shearer B L, Fairman R G, Crane C & Grant M 1994 Impact of Armillaria luteobubalina infestations on coastal dune communities of south-western Australia In: Handbook of the Symposium on Plant Diseases in Ecosys¬ tems: threats and impacts in south-western Australia (eds R T Wills & W A Cowling) Royal Society of Western Australia & Ecological Society of Australia Perth, 35. ShearerBL, WillsRT&StukelyM 1991 Wiidflower killers. Landscope7(l):30- 34. Sonnenberg A S M & van Griensven L J L D1991 Evidence for transmission of La France disea.se in Agartcus bisporus by dsRNA. In: Genetics and Breed¬ ing of Agaricus (cd L J L D van Grieasven) Pudoc Wageningen, 109-113. Tooley P W, Hewings A D & Falkenslein K F 1989 Detection of double stranded RNA in Phytophthora infestans. Phytopathology 79:470-474. Wills R T1993 The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18:145-159. Woodman G 1993 The role of Phytophthora cinnamomi as a damping-off pathogen in early mine site rehabilitation. Murdoch University, Honours Thesis. Yang H A, Sivasithamparam K, Alemohammad J, Barton J E & O'Brien P A 1993. Association of Rfiizoctonia strains with bare patch disease of wheat in Western Australia. Journal of Plant Pathology (in press). 177 I Journal of the Royal Society of Western Australia, 77:179-180,1994 Future ecosystems — use of genetic resistance J A McComb^ M Stukely^ & IJ Bennett^ ’School of Biological and Environmental Sciences, Murdoch University, Murdoch WA 6150 ^Department of Conservation and Land Management, 50 Hayman Road, Como WA 6152 ^Department of Applied Science, Edith Cowan University, Mt Lawley WA 6050 Abstract Genetic resistance of jarrah to Phytophthora cinnamomi has been identified by glasshouse testing and validated by field trials and analysis of plant/pathogen interactions. Resistant lines of jarrah can be micropropagated and used for revegetation of bauxite mine-sites and have potential for use in replanting dieback graveyard areas in the forest. Experiments are underway to determine the level of resistance of the progeny of selected resistant trees. Several questions are posed in this paper on the use of genetically resistant plants in restoring ecosystems. Introduction We have shown that it is possible to select Eucalyptus marginata (jarrah) resistant to Phytophthora cinnamomi and that this resistance is genetically based. A number of jarrah trees were selected on the basis of their apparent field resist¬ ance or susceptibility to P. cinnamomi, or because they repre¬ sented an ecotype of jarrah. At 12 months of age, half-sib seedlings were screened for their reaction to the pathogen using underbark inoculation or by inoculating the soil. The mean lengths of the lesions, or the percentage of plant deaths, were used to rank families (i.e. progenies from indi¬ vidual open-pollinated mother trees) from most resistant to most susceptible (Stukely & Crane 1994). From the extremes of the range, highly resistant plants from resistant families, and susceptible individuals from susceptible families were chosen and micropropagated (McComb et al. 1990). The clonal plants were planted in dieback-affected sites on rehabilitated bauxite pits and were inoculated with 4 strains of the fungus 1 month after planting. After 5 years in a typical field trial, the resistant plants have shown a low number of deaths and excellent growth, while in some susceptible lines all the plants have died (Fig 1). Laboratory testing of the selected plants has shown that, after infection of root tips with zoospores, the resistant jarrah plants confine the lesion extension; this was also observed in roots of the field resistant species £, calophylla (Cahill al. 1992). Further work is underway to investigate the mecha¬ nism of this resistance and its interaction with environmen¬ tal variables such as temporary waterlogging (Cahill & McComb 1992, Cahill et al. 1993). Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 Clone Figure 1. Mortalities of Eucalyptus marginata clones derived from seedlings resistant (R) and susceptible (S) to Phytophthora cinnamomi, after 5 years of growth in a Phytophthora-infesied bauxite mine-site. Heritability of resistance to P. cinnamomi was estimated from analysis of mortalities after soil inoculation in pots and in the field, and from stem lesion lengths after underbark inoculation of 15 - 50 half-sibs from each of 16 families (Stukely & Crane 1994). Resistance to P. cinnamomi was found to be under strong genetic control as narrow-sense heritability for the families was 0.74 - 0.85 and for individual trees was 0.43. We are studying the inheritance of the resist¬ ance trait further by controlled crosses and will gain infor¬ mation on other characteristics such as flowering times and combining ability for resistance. 179 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Can genetic resistance to P. cinnamomi help restore damaged ecosystems? It is important to investigate the use of resistant clones of jarrah to re-establish the species in areas in which it has died from dieback, that is on graveyard areas. This will be a greater challenge than the establishment of clones in bauxite pits where root competition is absent, drainage is optimised and the soil is friable. We have yet to develop reliable techniques for establishing micropropagated plants in grave¬ yard areas where jarrah plants have to be inserted between existing vegetation, and it is not possible to provide the same level of site preparation as in bauxite pits. In the ideal situation, the trees that are used in graveyard plantings will eventually establish a self-sustaining popula¬ tion of largely resistant trees. This will only be possible if we can identify clonal lines with good combining ability for resistance, and these lines flower at the same time. In restoring a damaged ecosystem we feel it is necessary to utilise, as far as possible, genotypes from the surrounding forest. With this in mind we are now doing further selections from the northern and southern regions of the jarrah forest. However, jarrah is just one species which is killed in natural ecosystems. Replacing the jarrah is a small step towards reversing the floristic impoverishment of the affected areas. Can the method used for selection of resistant jarrah work for other species and other diseases? Theoretically it would be possible to screen other plants with P. cinnamomi or other pathogens and to find resistant individuals in the field. However, on some forest sites jarrah is relatively resistant to P. cinnamomi (Dell & Malajczuk 1989) compared with many other species, in which there may be 100% deaths. There are problems too, not only with the vast number of species that are affected in each ecosystem, but also the length of time required to work out reliable propa¬ gation or micropropagation methods for some species, and to develop appropriate screening methods. Clearly it is a strategy only suitable for priority species. It may be possible to partially restore an ecosystem by use of selected resistant lines of a restricted number of species. This would at least provide more diversity of plant species and animal habitats than are found at present in graveyard areas. Can the new techniques of genetic markers or probes make selection of disease resistant plants faster? In some plant / pathogen interactions it is now possible to identify disease resistant plants by extracting the DNA and probing it for DNA sequences known to indicate resistance. It is not necessary to know the mechanism of the disease resistance and the technique allows screening of plants from disease free areas and of plants whose propagation is diffi¬ cult. This exciting development must be underpinned by ini¬ tial work in each species to identify some resistant and susceptible lines and some study of the heritability of resist¬ ance. However, once appropriate markers are found, screen¬ ing for suitable resistant individuals for seed orchards, veg¬ etative or micropropagation can proceed more rapidly. The technique is likely to work most quickly when there are major genes for disease resistance, but it may also be effective in polygenic systems of resistance such as in jarrah. Can natural resistance be enhanced or replaced by genetically engineered resistance? Genetic engineering offers almost unlimited scope for introducing novel mechanisms for disease resistance into plants. In practice, the problems of working out appropriate techniques for introducing new genetic material into a number of species in a natural ecosystem are immense and conse¬ quently very costly. Added to this are ethical problems. There may be few objections when the introduced gene is from a resistant plant of the same species, but we might expect objections when genes from unrelated organisms are used. Genes introduced by genetic engineering may come from unrelated organisms and in recipient species they form new loci usually dominant in expression. In a cross-pollinating species with a short generation time and under high selec¬ tion pressure, the gene would spread quickly through the population. Are we willing to see this type of genetic engi¬ neering in our natural ecosystems? Is the damage from Phytophthora and other pathogens so great that we would be willing to make the natural ecosystem, to some extent, un¬ natural ? Acknowledgment: We thank Alcoa (Aust) for their continuing support for the work on selection and propagation of dieback resistant jarrah. References Cahill D & McComb J A 1992 A comparison of changes in phenylalanine ammonia-lyase activity, lignin and phenolic synthesis in the roots of Eucalyjitus cahphylla (field resistant) and E. marginata (susceptible) when infected with Phytophthora cinnamomi. Physiological Molecular Plant Pa¬ thology 40:315-332. Cahill D, Bennett I J & McComb ] A 1992 Resistance of micropropagated Eucnlyptus mjirginata to Phytophthora cinnamomi. Plant Disease 76:630-632. Cahill D, Bennett I ) & McComb J A 1993 Mechanisms of resistance to cinnamomi in clonal, micropropagated Eucalyptus marginata. Plant Pathology 42:865-872. Dell B & Malajczuk N 1989 Jarrah dieback - a disease caused by Phytophthora cinnamomi. In: The Jarrah Forest - A Complex Mediterranean Ecosystem (ed B Dell, J Havel & N Malajczuk) Kluwer, Dordrecht, 67-87. McComb J A, Bennett I J, Stukely M & Crane C 1990. Selection and pro¬ pagation of jarrah for dieback resistance. Combined Proceedings of the International Plant Propagators' Society 40: 86-90. Stukely M J C & Crane C E 1994. Genetically based resistance of Eucalyptus marginata to Phytophthora cinnamomi. Phytopathology 84: 650-656. 180 Journal of the Royal Society of Western Australia, 77; 181-184,1994 Future ecosystems — ecological balance (ecological impact of disease causing fungi in south-western Australia) G J Keighery^ D J Coates^ & N Gibson^ ^ Department of Conservation and Land Management, Western Australian Wildlife Research Centre, P.O. Box 51, Wanneroo WA 6065 ^Department of Conservation of Land Management, Western Australian Herbarium, George St, South Perth WA 6151 Abstract Ecosystems are dynamic not static; however, since European settlement the rates of change have been dramatic and rapid. Entirely new types of disturbances have occurred including land clearance, new grazers, new predators, new pollinators, new weeds and new diseases. All of these have impacted on the ecological balance of native plant and animal ecosystems, and their effects are synergistic not unique. South-western Australia has unique species-rich and structurally diverse plant communities, supporting a web of co-evolved fauna species. New plant diseases may lead to simpler ecosystems in terms of structure, diversity and function. Remnants will often be severely impacted, and more prone to weed invasion. Larger areas may become dominated by fewer resistant species. Many susceptible plant species of restricted ranges face the possibility of extinction in the medium term. More widespread susceptible taxa face severe genetic erosion through local extinction of populations in remnants where re-invasion is not possible. Native animal species which rely on the floral diversity for food and shelter may face local extinction, or at least critical reduction in numbers. Perhaps the greatest challenge facing land managers and those attempting to conserve our wild heritage is to understand, detect and lessen deleterious disturbances from completely overwhelming our remaining bushland, and rendering it more vunerable to disease. Introduction All ecosystems are dynamic and change in place and iTomposition over time. Disturbance is a recognised feature of ecosystems, and appears to be vital in maintaining species and community diversity. However, prior to the advent of industrialised European man, this cycle of disturbance and renewal occurred in a setting of continuous bushland. Since settlement in Western Australia and the clearing of land for agriculture, all ecosystems have been rapidly and severely impacted. In the south-west of Western Australia the major change has been the fragmentation of the bush by clearing for other land uses. This has resulted in numerous, often small, isolated bushland remnants many of which appear to have a limited future without careful management. Clearly, fragmentation of ecosystems and its consequences are major challenges facing conservation managers world-wide. Disease is one aspect of natural population control in¬ volved in disturbance and change, but combined with frag¬ mentation it may prove decisive, destructive and unidirec¬ tional. This paper mainly deals with P/ii/fop/ifJiorrtcf»nrtmomi, but it should be noted that several other diseases briefly mentioned here also pose local severe conservation prob¬ lems and deserve closer scrutiny. Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. Held on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 Impact Root cankers: Phytophthora cinnamomi Detailed studies in Victoria (the most climatically-similar area) in dry sclerophyll woodlands, heaths and swamps in the Grampians, Brisbane Ranges and Wilsons Promontory by Dawson et at. (1985), Kennedy & Weste (1986), Weste (1975, 1981 & 1986), Weste & Law (1973), Weste & Taylor (1971) and Weste et al, (1973) have shown that the disease has permanent and severe effects on the structure and floristics of these plant communities. Similarly severe and permanent impacts have been noted in Tasmania (Podger et al, 1990) and Western Australia at Two Peoples Bay (Hart 1983), Cape Arid and Cape Le Grande (Brandis et al 1985), Stirling Range (Wills 1993 ), the jarrah forest (Shearer & Tippett 1989) and the Banksia woodlands of the Swan Coastal Plain (Podger 1972, Shearer & Hill 1989, Shearer 1990). Potentially major impacts can be expected in the heathlands north of Perth (Hill 1990), and in the understorey of many communities of the southern forests (Shearer 1992). The western edge of the wheatbelt and localised populations of proteaceous heaths in suitable catchments in this entire region could also be at risk. In general, the ecological balance which might be achieved in future ecosystems after major disease impact is likely to be reflected in simpler communities in either structure (loss of susceptible dominants) and / or floristics (loss of susceptible species), hence a net local loss of biodiversity. These effects 181 It Journal of the Royal Society of Western Australia, 77 (4), December 1994 are most widespread and severe in the south coastheathlands of Western Australia, from Augusta to Cape Arid. Since this region also contains the highest proportion of vertebrate pollinated flowering plant species in the world (Keighery 1982), many of which are Proteaceae, and susceptible to dieback, secondary effects will also be severe on the diversity of animal communities (Friend 1992) in this area. Bird-pollinated plants often dominate our flowering heathlands in spring, with their large and brilliantly col¬ oured flower; many of these species are highly susceptible to dieback and are replaced by wind-pollinated resistant sedges, a dismal sight to any honeyeater or tourist. Dieback is likely to reduce local populations of these birds. Insects reliant on susceptible species will also face significant reductions in food sources (e.g. the specific bee pollinators, Leioproctus pappus of Conospenmim incunnim, Euryglossa morrisoni of Verticordia nitens and Euryglossa aurea of Verticordia aurea; Houston 1992), but there are few data on the extent of this aspect. The effects on soil organisms, including fungi (see Malajczuk & Pearce 1994) is almost completely unknown, but it must be considerable. The second major impact will be on the hundreds of bushland remnants containing susceptible communities in an arc between Moore River, Kojonup and Esperance. In the studies noted above, the affected areas are within large areas of native vegetation, enabling immigration of resistant taxa, with refuge areas often present for susceptible taxa. In small isolated remnants there is little chance of recruitment from neighbouring remnants, no refugia and local extinction is the major consequence, accompanied by weed invasion. For example, the Swan Coastal Plain Survey has placed over 500 permanent vegetation quadrats in bushland betu'een Gingin and Dunsborough, and has compiled total flora lists for most reserves in this area and on sections of the escarpment. Quadrats at Twin Swamps, Bullsbrook, and Cardup Nature Reserves, shown a major decline in species diversity in communities invaded by Phytophthora cirtuatnomi and a ma¬ jor increase in weed invasion following severe impact by Phytophthora. Average species diversity declined from 69 species to 39 at Cardup or from 49 to 25 at Twin Swamps, and weed species increasing from an average of less than 2% to over 50% of species records at some sites in Twin Swamps. The disease removes susceptible dominants and understorey species, hence opening the community to invasion by resist¬ ant annual and grassy weeds. Flora surveys indicate that local extinctions of many susceptible understorey taxa such as Banksia meisneri var adscendenSf Verticordia nitens and Lambertia rnultiflora var "darlingensis" are likely to occur over the next 5 years at Ruabon, Bullsbrook and Cardup Nature Reserves respectively unless urgent remedial action is taken. Remnants with service corridors through them or with wetlands in or adjacent to the remnant appear to suffer the highest impact, with disease spreading from the wetland or corridor. These remnants are likely to lose their Banksia woodlands entirely, as the dominant Banksia trees become locally extinct. At a species level, the disease has the poten¬ tial to lead to at least the localised extinction of highly susceptible species. A number of recent publications list and publicise the taxa under the greatest threat (Conservation and Land Man¬ agement 1992, Curry 1992, Curry & Kelly 1993, Keighery 1988 a, b, 1991,1992). Eleven taxa are under greatest threat with most or all populations infected (Table 1). Seven of these are considered under threat of extinction in the wild, and the four other taxa are being monitored to determine their status and management options. Table 1 Taxa requiring urgent management action to lessen the impact of Phytophthora A) Taxa under threat of extinction in the wild in the medium term, all populations impacted by disease: Attdcrsonia axilliflora Andersonia sp (Two Peoples Bay) Andersottia sp (Mt Lindesay) Banksia brouniii (Both forms) Dryandra "montam" Isopogon uncinaUis Lambertia echinata B) Taxa with most populations infected by disease, potentially at risk of extinction in the wild: Calytrix breinseta ssp breviseta Conospermum caerulescens ssp "adpressum" Dryandra sp 30 Lambertia orbifolia Current research and management has centred on pro¬ tecting stands of the most threatened species (currently Andersonia species and Banksia broumii)hy phosphonate spray¬ ing and collection of germplasm material (of all taxa) for ex situ long term storage. Ex situ germplasm collections are a last re.sort but in a number of cases may be the only means of preventing total extinction of a species. Information on the effects of disease on many other restricted and potentially threatened species, such as the mountain bells (Danvinia species), Adenanthos dctmoldii, A.pungens, Persoonia rnicrantha, Isopogon uncinatus, Banksia occidentalis ssp, formosus, and Petrophile laterkola, is urgently needed to set priorities and develop strategies for their conservation and management. More widespread susceptible species, such as Banksia coccinea and Lambertia propittqua, are suffering significant genetic erosion through the local extinction of populations and reduction in population size. Other species such as Lambertia orbifolia, Banksia brozonii, Banksia illicifolia and Adenanthos barbigeriis, are in danger of losing genetically and morpho¬ logically distinct races. Again the ecological balance is being shifted to less complex communities, by the loss of regional genetic and specific diversity (geographically restricted taxa and localised population divergence appear to be a feature of plant communities in south-western Australia; Hopper & Coates 1990). Other Phytophthora species Five other species of Phytophthora, namely P. citricola, P. cryptogea, P. dreschleri, P. megasperrna and P. nicotiana have been identified from dead and dying native vegetation. All of these species can have localised severe impact, and prob¬ ably pose a major threat to at least some local populations of susceptible flora (such as the localised, declared rare species, Adenanthos ellipticus on East Mt Barren) and isolated rem¬ nants. 182 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Wood rots Armillaria luteobubalina is the major species of wood rot, with a wide and diverse host range, often from groups resistant to P. cinnamofiiif and in habitats usually at low risk from root rots e.g. Coastal dunes and Spearwood sands (Shearer & Tippett 1989). The impact of disease caused by this fungus on a local scale in Granite heath,Wandoo and Coastal heath can be severe, hence its greatest impact will be on remnants, where it may cause local extinctions and en¬ hanced weed invasion. Localised rare species could also be at risk. Detailed long term studies are required on the effects of this disease, since it is a native species, to understand its role in native plant communities. Stem and branch cankers Currently the major concern is the canker Cryptodiaporthe sp. which affects a broad range of Proteaceae, often causing death of the plant. Banksia coccinea populations are being heavily impacted by this canker, and this species faces local¬ ised extinction in the Albany area. The only known popula¬ tion of Dn/andra sp (Kamballup) is infected with canker (Wills, perscom7n\ and requires monitoring. A series of other cankers can cause dieback and death and loss of overstorey Eucalypt species, including Marri and Red Flowering Gum, Wandoo and Tuart. These diseases could prove severe in remnants, preventing the replacement of spedes after loss of the parental trees. These cankers appear to be a response to various types of disturbance and/or enviromental factors such as drought stress, and their long term impacts are unknown. The future The symposium organisers requested that we speculate, based on current knowledge what the ecological balance of futureecosystemswould be under current trends of diseease spread and impact. Our potential future is major impacts of these diseases in south western Australia, as presented be¬ low. The combined fragmentation and increasing levels of disturbances will inevitably result in a loss of biodiversity. This will result, through the changes in community structure favouring resistant species, in the extinction of local populations of numerous species and the subsequent loss of genetic diversity, the likely extinction of many rare and geographically-restricted susceptible species, secondary spe¬ cies lossof both plantsand animals through lack of pollinators, food plants and shelter, severe genetic erosion of wide¬ spread susceptible taxa, and a loss of scenic values to the public and tourists. Currently the ecological balance seems to be shifting towards simpler, weed-invaded and less visu¬ ally appealing plant and animal communities, that may lack many of their uniquely Western Australian features. There seems little doubt that introduced diseases are hastening this trend. References Brandis A, Hill TCP, Keighery G J & Tippett J T 1985 Protection status and vegetation types in the South Coast Region. Department of Conservation and Land Management, Perth. Unpublished report, 58pp. Conservation and Land Management 1992 A Nature Conservation Strategy for Western Australia: Draft for public comment Department of Conser¬ vation and Land Management, Perth. Curry S1992 Endangered, Prickly Honeysuckle. Landscope 8(2); 40. Curry S & Kelly A 1993 Endangered, the swamp starflower, Calytrix breviseta ssp. breviseta. Landscope 8(4): 27. Dawson P, Weste G & Ashton D H 1985 Regeneration of vegetation in the Brisbane Ranges after fire and infestation with Phytophthora cinnamomi. Australian Journal of Botany 33:15-26. Friend G1992 Possum in peril. Landscope 7(3): 22-27. Hart R1983 Report on dieback due to Phytophtlwra cinnamomi in Two Peoples Bay Nature Reserve. Department of Fisheries and Wildlife, Perth, Unpub¬ lished Report. Hill T C J1990 Dieback disease and other Phytophthora species in the northern Kwongan. In: Nature Conservation, landscape and recreation values of the Lesueur area. Environmental Protection Authority, Perth. Bulletin 424. Hopper S D & Coates D J1990 Conservation of Genetic Resources in Austral¬ ia's Flora and Fauna. In; Australian ecosystems; 200 years of utilization, degredation and reconstruction (ed D A Saunders, A J M Hopkins & R A How) Proceedings of the Ecological Society of Australia 16:567-577. Houston T F1992Three new, monolectic spedes of Euryghssa {Euhemsa) from Western Australia (Hymenoptera: Colletidae). Records of the Western Australian Museum 15:719-728. Keighery G J 1982 Bird pollinated plants in Western Australia and their breeding systems. In: Pollination and Evolution (ed J A Armstrong, J M Powell & A J Richards) Royal Botanic Gardens, Sydney, 77-89. Keighery G J 1988a Dieback briefing paper: Epacridaceae. Department of Conservation and Land Management, Perth. Unpublished report, 53 pp. Keighery G J 1988b Endangered, Browns Banksia, Banksia brownii. Landscope 3 (4): 54. Keighery G J 1991 Endangered, Dieback Prone Plants of the Eastern Stirling Ranges. Landscope 6 (4); 43. Keighery G J1992 The impact of Phytophthora on rare plants. In: Dieback -What Is The Future? The Northern Sandplains Dieback Working Party, Seminar Proceedings,Perth, 29-36. Kennedy J & Weste G 1986 Vegetation changes associated with invasion by Phytophthora cinnamomi on monitored sites in the Grampians, Western Victoria. Australian Journal of Botany 34:251-279. MalajczukN&PearceMH1994ImpactofpIantdiseasesonmicrobial ecology. In; Handbook of the sympsium on plant diseases in ecosystems: threats and impacts in south-western Australia (ed R T Wills & W A Cowling) Royal Society of Western Australia and Ecological Society of Australia, Perth, 10. Podger F D 1972 Phytophthora cinnamomi, a cause of lethal disease in indig¬ enous plant communities in Western Australia. Phytopathology 62.972- 981. PodgerFD,PalzerC&WardIawT1990AguidetotheTasmaniandistribution of Phytophthora cinnamomi and its effects on native vegetation. Tasforests 2:13-20. Shearer B L 1990 Dieback in south-western Australia. Land and Water Resources Newsletter 5:15-26. Shearer B L1992 The ecological implications of disease in the southern forest of south western Australia. Department of Conservation and Land Man¬ agement, Perth. Occasional Paper2/92, 99-113. Shearer B L & Hill T1989 Diseases of Banksia woodlands of the Swan Coastal Plain In: Banksia Woodlands Symposium. Journal of the Royal Society Western Australia 71:113-114. Shearer B L & Tippett J T1989 Jarrah d ieback; The dynamics and management of Phytophthora cinnamomi in the jarrah (Eucalyptus marginata) forest of south-western Australia. Department of Conservation and Land Manage¬ ment, Perth. Research Bulletin 3. Weste G 1975 The distribution of Phytophthora cinnamomi within Wilsons Promontory, Victoria. Australian Journal of Botany 23:67-76. Weste G 1981 Changes in vegetation of shrubby sderophyll woodland asso¬ ciated with invasion by Phytophthora cinnamomi. Australian journal of Botany 29:261-276. Weste G 1986 Vegetation changes associated with invasion by Phytophthora cinnamomi on defined plots in the Brisbane Ranges, Victoria, 1975-1985. Australian Journal of Botany 34:633-648. 183 Journal of the Royal Society of Western Australia, 77 (4), December 1994 Weste G M, & Law C 1973 The invasion of native forest by Phylophthora cinnamomi. Threat to the National Park, Wilsons Promontory. Australian Journal of Botany 21:31-51. Weste G M, & Taylor P 1971 The invasion of native forest by Phytophthora cinnamomi. 1. Brisbane Ranges, Victoria. Australian Journal of Botany 19:281-294. Weste G M, Cooke D & Taylor P 1973 The invasion of native forest by Phytophthora cinnamomi. Regeneration patterns, declining pathogenicity and attempted control, Australian Journal of Botany 21:13-29. Wills R 1993 The ecological impact of Phytophthora cinnamomi in the Stirling Range National Park, Western Australia. Australian Journal of Ecology 18:145-159. 184 Journal of the Royal Society of Western Australia, 77: 185-186,1994 The future — effects of plant disease on society J T Young PO Box 54, Walpole WA 6398 Abstract Just as humans react adversely to too many changes occurring too rapidly, many plant species of the south¬ west of Western Australia will not withstand the rate of change in their environment. There are many ways in which the structure and composition of plant communities are being drastically altered over large areas. Plant diseases may well express as never before. Disease epidemics may be one of the things which force changes in the way the ecosystems of the south-west are managed and utilized. Management should bebased more on basic ecological principles and less on economic considerations. Much greater emphasis should be placed on disease prevention. It must be realised that many systems of disease control developed for horticultural situations are not applicable for broad scale use in natural ecosystems. Once management is more in accordance with basic ecological principles, there will be less conflict in society between those who are fearful for the future health of south-western ecosystems and those who believe "the bush", the forests in particular, will withstand intensive exploitation and repeated disturbance. If disease prevention does not become a prime objective of management, future generations will get fewer and fewer opportunities to appreciate, exploit and enjoy the great diversity, complexity and beauty of the ecosystems of the south-west, which Europeans discovered less than 200 years ago. Effects of plant disease on society Future effects of plant disease on society will be numer¬ ous. There are the obvious things; more regulations, more expensive natural products and decreased supply of prod¬ ucts previously, all taken for granted. One less-obvious effect will be a greater divergence of management tech¬ niques used in natural ecosystems from those used in areas of intensive production such as on farms and in plantations. Plant diseases are difficult and often impossible to control in natural ecosystems; for example the root diseases caused by Phytophthora dnmmomi and Armillaria luteobubalina can be either prevented or controlled in horticultural situations, but both are almost impossible to control in conservation and forest areas where there are numerous infections spread over vast areas. The incidence and severity of plant disease in our south¬ western ecosystems are going to increase and as a conse¬ quence the ways of management will be forced to change. There has to be much more emphasis on disease prevention. Society must be much more honest about the long term objectives of management of both the conservation estate and forests. Economic rather than ecological principles are primarily determining how the ecosystems of the South¬ west are currently being managed. Some plans are initiated from a base of ecological information but then debased by extrapolation over vast areas of widely varying communi¬ ties; for example jarrah logging prescriptions and the treat¬ ment of mixed species Eucalypt forests. This paper is too short to discuss the factors which favour the development of plant diseases in detail. Shearer (1992) Symposium on Plant Diseases in Ecosystems: Threats and impacts in south-western Australia. on April 16, 1994, at Murdoch University, by the Royal Society of Western Australia and the Ecological Society of Australia. © Royal Society of Western Australia 1994 suggested many ways in which human activity could either aggravate or control disease in forest ecosystems. I believe there is plenty of evidence that disease incidence and sever¬ ity will increase as a consequency of human activity. Society seems to be doing much more to encourage plant diseases than to prevent or suppress them. Of particular concern to me is the rate of logging in State Forest. Disease prevention seems to be a low priority. The case for woodchipping and clear-felling is in part based on the ability to regenerate karri forest from seed and seedlings. The ash type eucalypts lend themselves to such management especially when they are in relatively pure stands but there are fewer and fewer stands with predomi¬ nantly karri available for clear-felling. Much of the forest being treated as karri has few stems of karri per hectare; the rest is marri, blackbutt and jarrah. Much of it is essentialy jarrah forest. What is of concern is that karri is being estab¬ lished back on to many areas and soil types where it never grew naturally at high densities. We are imposing a planta¬ tion-type system within natural ecosystems without the knowledge or resources to manage disease. The incidence of brown wood and incipient rot is high in regenerated karri stands (Davison & Tay 1994, Shearer 1992). The stability and benefits of mixed species stands are not understood. Across Australia, often a species of the Eucalyptus sub-genus Monocaplytus and a species of the sub-genus Symphomyrtus seem to form stable a.ssociations (Boland et al. 1984). The recently approved Forest Management Plan 1994 - 2003 (Conserv^ation and Land Management 1994) for this State will mean that areas of jarrah forest, in the order of 130 km^ to 300 km^ will be "clear-felled" every year and require ongoing thinning. Many of the areas so treated will be in the low and intermediate rainfall zones of the forest where the quality of forest is mixed and little research has ever been done. The recent past scale of the operations means that there is already something like 4126 km^ of regenerating jarrah 185 Journal of the Royal Society of Western Australia, 77 (4), December 1994 forest, with less than 20 years of growth (figures collated from Forest Department (WA) and Conservation and Land Management Annual Reports). As stated by the Environmental Protection Authority (1992), the commitment to the type of management de¬ scribed in this paper "would lock the state into long term acceptance of the new forest structure and the new intensive production with insufficient flexibility to adapt manage¬ ment in the light of new information or changes in commu¬ nity expectations". What does happen if, as I predict, dis¬ eases hinder regeneration and growth rates do not meet levels set in computer models? I believe that there are al¬ ready quite inadequate resources to guarantee a healthy crop of trees over all of the areas being logged. Intensive monitoring and disease mangement seems even less likely. Forests are slow to grow and conversely they are slow to die. Consequently, mismanagement may take 20 to 40 yrs to express on a large scale. The danger is that economics rather than ecological principles will drive management of forests for too long and signs of decline will be ignored. Unfortu¬ nately, a single State Department controls most of our natu¬ ral ecosystems and one disadvantage of this state of affairs is that "blanket" prescriptions are imposed over vast areas. When mistakes are made, the consequences will be great. In industries where private enterprise flourishes, many ideas are tried on a small scale all the time. Those that are not any good, fail, without great cost to society. Society must honestly address the question of sustainability and what of the natural ecosystems of the south-west will be left to provide future generations. Multi¬ ple use of areas means increasing pressures on the vegeta¬ tion. The frequency of disturbance, including fires, is ever increasing, presumably resulting in effects such as decreas¬ ing numbers of mature seeding plants and decreasing carbo¬ hydrate reserves within perennials. Fiow do we measure the potential of different plants and communities to reproduce, recover and resist disease after repeated disturbance? We have few measures of vigour or resistance. Slow and gradual decline of species is also difficult to monitor. The decline of species is often masked by the colonization of bare ground by other species. Even with dieback caused by P. ciunamomi few people appreciate the extent and changes in the floristics of plant communities in the south-west because Agonis species and rushes have replaced many of the species killed. Surely one of the things about planning for the future is heeding warnings. Warnings can result from monitoring. There has been far too little monitoring done in vast areas of the south¬ west on dieback or anything else. The effects of disease on regeneration of dominant species should surely be a high priority? What are the effects of P.cinnamomi on jarrah regen¬ eration on different soils? Usually a crisis has tobe well documented before politicans will respond. Plant disease epidemics along with insect plagues will help sway society to call for change. At least the conservation estate should be managed in accordance with basic ecological principles. Compromises, a result of too many conflicting pressures, should be sought warily. 1 believe that, once ecological principles are the basis of vegetation management, there will be less conflict in society between those who believe, "the more you bash it the better it'll be" and those who do have a degree of understanding of the complexity and vulnerability of the ecosystem of the south-west. At least the forest debate could become less polarised. I believe community expectations of sound man¬ agement of our natural ecosystems are going to ever in¬ crease. In the future, the public will be better educated in environmental biology and as more environmental disasters become apparent their assessment of management practices will be more and more rigorous. The current education curricula in schools are stressing an understanding of environmental issues such as dieback and salinity. The general level of appreciation of the flora and fauna is increasing and people will be more sensitive to adverse changes brought about by plant disease. When people go out and plant trees in "greening" programmes and they die, some will ask why? More people will observe and recognise problems as they arise. Society must appreciate the ecosystems on which it impinges, including their complexity and vulnerabity; oth¬ erwise there will not be the will and determination to mini¬ mize the impacts of disease. References Boland D J, Brooker MIH, Chippendale G M, Hall N, Hyland B P M, Johnston R D, Kleinig D A & Turner J D 1984 Forest Trees of Australia. CSIRO, Nelson. Conservation and Land Management 1994 Forest Managmcnt Plan 1994 - 2003. Department of Conservation and Land Management, Perth. Davison E M & Tay F C S 1994 Incipient rot and rot in regrowth karri. In: Handbook of the symposium on plant diseases in ecosystems: threats and impacts in south-western Australia (ed R T Wills & W A Cowling) Royal Society of Western Australia and the Ecological Society of Australia, Perth, 22. Environmental Protection Authority 1992 Proposal to amend the 1987 forest management plans and timber strategy and proposals to meet environ¬ mental conditions on the regional plans and the WACAP ERMP. Depart¬ ment of Conservation and Land Management Report and Recommonda- tioas of the Environmental Protection Authority, Perth. Bulletin 652. Shearer B L1992 The ecological implications of disea.se in the southern forest of south-western Australia. In: Research on the impact of forest manage¬ ment in south-west Western Australia. Department of Conservation and Land Management, Perth. Occasional Paper Number 2/92. 186 I [ r- I i \ '{ I 18158/11/94—2700—U5688 r t lit 1 . • i >i£ I i < ^ *1 M4t'4S