JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 59 PART 1 MAY, 1976 PRICE TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Chief Marshall Sir Wallace Kyle, G.C.B., C.B.E., D.S.O., D.F.C., Governor of Western Australia COUNCIL 1975-1976 B. E. Balme, D.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. A. J. McComb, M.Sc., Ph.D. G. A. Bottomley, B.Sc., Ph.D. G. Perry, B.Sc. (Hons.) M. W. Perry, B.Sc. (Agric.) (Hons.) S. J. Curry, M.A. A. Neumann, B.A. . A. J. McComb, M.Sc., Ph.D. (1975) A. E. Cockbain, B.Sc., Ph.D. (1976) P. E. Playford, B.Sc., Ph.D, J. C. Taylor, B.Sc., Ph.D., A.R.C.S. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. B. Lamont, B.Sc. (Agric.), Ph.D., F.R.H.S. C. E. Dortch, B.S., M. Phil. L. J. Peet, B.Sc., F.G.S. P. R. Atkinson, B.Sc. J. K. Marshall, B.Sc. (Hons.), Ph.D. President Vice Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor 1. — Topical chemistry in Perth’s air Presidential Address, 1975 by G. A. Bottomley^ Delivered 28 July 1975 Introduction Last year Dr. TrendalLs Presidential Address dealt with the banded iron formations in the northwest of this State and their genesis associ- ated with the change in the composition of the Earth’s atmosohere perhaps 2 000 million years ago. I shall be discussing some quite recent atmospheric chemistry again particularly rele- vant to Western Australia and the metropolitan area. At any one average moment over the State of Western Australia there is present 300 million million tons of air (that is over 10,000 times the mass of our known iron ore reserves), carrying water equivalent to at least 10,000 times the annual consumption in the metropolitan area and “traces” of carbon dioxide enough to make a cubical block of wood two miles each edge. It is a big subject. By natural meteorological processes involving prodigious amounts of poten- tially useable solar energy and wind energy the lower atmosphere is generally changing, air arriving from the southwest for example, whilst the current air progresses generally eastward. We humans use it as a source of oxygen reject- ing carbon dioxide to it, plants do the converse by day, and of course it is a rubbish disposal service for factory emissions and traffic ex- hausts. Though neither import duties nor export controls are imposed on its movement by the Federal Government, air is an important and precious commodity. No previous President however has elected to talk about the Atmosphere, indeed, even in the Society’s extensive publications going back to 1914 (and its antecedents to 1897) only a few papers touch the tonic. The air’s inland trans- portation of agriculturally significant chemicals is treated in the Royal Society’s Report on ‘Salinity in Rain in Western Australia’ (Profes- sor Wilsmore Chairman 1928) and in Professor Drover’s ‘Accessions of Sulnhur’ (1960) Would modern replications of the latter be higher? Dr Speck reported in 1953 on ‘Atmospheric Pollen in the City of Perth and Environs’. And four papers indirectly relate to massive combustion in air* Miss Baird’s ‘Regeneration on Garden Island after fire’ (1958), Dr. Hatch’s Burning of the larrah’ (1959), ‘Man the destroyer’, Dr Merrilees’ Presidential Address (1967), and ‘Fire in the jarrah environment’, Dr. Wallace’s Presidential Address (1966). ^Chemistry Department, University of Western Australia Nedland s, Western Australia. 6009. Journal of the Royal Society of Weste Perhaps our oblique interest is because our city has never experienced pollution incidents such as the London pea-soup fogs (now almost eliminated by smokeless zone legislation) in which sulphur dioxide and smoke particles jointly disrupted the health and amenities of European industrial cities. Be that as it may be, no professional chemist need apologise for discussing Air. No precise point can be given for the evolution of modern chemistry (just possibly Dalton’s atomic theory 1807), but my biased choice as a gas experimentalist would be fellow Yorkshireman Joseph Priestley’s characterisation and recog- nition of oxygen as a separate chemical species. Priestley knew about many of the gases I shall refer to later. Oxygen he prepared by the famous burning glass experiment. Solar energy again! Nitrogen he obtained by removing oxygen from air by combustion and by respiration (of mice and men). Nitric oxide he synthesised from the action of many metals on nitric acid. Most importantly for us, he knew that nitric oxide combined with oxygen to give the red water- soluble nitrogen dioxide, indeed this was Priest- ley’s standard quantitative test for the ‘goodness’ of air, viz, its oxygen content. All firmly estab- lished by the ‘pneumatic chemists’ by 1775. The compounds NO and NO 2 (quite distinct from N20-‘laughing-gas’) occur in barely measurable amounts, say a fraction of one part per hundred million in rural, uncontaminated air. Significant amounts are released during combustion of coal, of oil, of natural gas, etc., in furnaces and of petrol in the car engine, therefore both are found above natural levels in city, contaminated atmospheres. Neither is medically harmful at the levels I shall be speak- ing of tonight but they are centrally involved in photochemical smog. Photochemical reactions Controlled photochemical reactions in gases were first demonstrated a century ago by John Tyndall (1868) physicist, master experimentalist in bacteriology (he observed the inhibition of microbial growth by Penicillium fifty years be- fore Fleming), noted science expositor, and a glaciologist who attemoted to conquer the Mat- terhorn just before Whymper’s tragic assault. ^uuraining a trace of amyl nitrite vapour on illumination with sunlight, though intially optically clear, develops a brilliant fog as the energy of the light beam converts the amyl nitrite tp a non-volatile material. The colours rn Australia. Vol. 59, Part 1. May. 1976. 1 of Tyndall’s fogs, pink and green in this repli- cation, failed to attract adequate investigation; similarly coloured monodisperse sols were re- discovered by La Mer (1941) and are explicable with light scattering theories due to Lord Ray- leigh. Many commentators for instance F. W. Went '1966) noted plant physiologist, have suggested that the ethereal atmospheric effects over, for example, Australia’s Blue Mountains might have their origin in extremely fine particulates formed by the natural photochemical process of intense sunlight acting on organic vapours emittsd by trees. Perhaps some of Gruner’s paintings depict the phenomenon. Let me move now to an artificial photochemi- cal experiment on a massive scale. A detailed chemical exposition would be out of place here, I merely remind you of the essential features of photochemical smog formation in air con- taminated with automotive emissions. Energy from sunlight otherwise passing harmlessly to ground is intercepted by the NO 2 molecule which then sets in train chemical degradation of hydrocarbons from uncombusted petrol. The chemical sequence is such that the NO 2 molecule is regenerated and continues its cyclic, catalytic role. In stagnant air the photochemical process results in a marked and characteristic loss of visibility, the formation of eye irritating chemi- cals, and enhanced ozone levels. Well docu- mented social, economic and medical disabilities follow severe and repeated exposure to such conditions. For example, in the Los Angeles basin photochemical smog of man-made origin has severely or at least moderately damaged over 100,000 acres of the San Bernadino National Forest. Restoration of the pre-car era air quality has so far defeated all attempts. Perth air: background information One of the great delights of Perth has been its brilliant atmospheric clarity — a tourist attrac- tion too! In the mid sixties, fresh from eighteen months in Los Angeles, I began to have qualms of chemical conscience, qualms engendered on clear winter mornings by perceptible odours reminiscent of Los Angeles, recognisable by my- self and several others with personal experience of L.A. photochemical smog. (Incidentally, only life-long non-smokers should serve on air- control bodies.) Is there, I mused, the slightest possibility that photochemical processes within the emissions from our expanding road traffic, perhaps compounded by natural terpenes from eucalyptus, might lead at some distant future time to an occasional L.A. type incident? Almost everyone dismissed this as fantasy in ‘the windiest capital city in Australia’. What could I do with very limited facilities which mght partially resolve this question? A remark of Charles Darwin helped: ‘Once a week do a damn fool experiment, they hardly ever work but when they do they’re marvellous’. Chemical analysis for ozone, hydrocarbons, ‘oxidant’, all indeed pose severe chemical and instrumental problems (i.e. financial problems) at the very low con- centrations of interest to air chemists. The determination of NO and NO 2 together, collo- quially NOx, was a practical possibility which might be squeezed in alongside my mainstream research. The principal results are reported with detailed discussion in the Journal of the Royal Socisty (Bottomley and Cattell, 1975). Let me tonight emphasise the salient points, provide some illustrations, and add perhaps a little human interest. Of about two hundred NOx values measured at the University between June 1970 and June 1971 all the concentrations except two are incon- sequential by world city standards. More interest arises when wa note that most high NOx values coincide with impaired atmospheric clarity of a special type. April 29th 1971 is particularly interesting. The NOx value taken 8.00 a.m. to 10.00 a.m., following the morning traffic peak, was the highest recorded in that series, and is followed by a marked fall during the 10.00 a.m. to noon period just as the atmospheric clarity improves. This is simply a chemical verification that car (and other emissions) linger near the ground in stagnant late-autumnal morning air. Unauthorised release of this comparatively high value (23.6 pphm) unleashed in the media an obscuring controversy about Perth’s air. Criticism was made that the University site was quite unrepresentative of the Metropolitan area (though 10 000 of us have to breathe its air) and that Chemistry’s oil-fired boiler was the cause: no critic asked if the boiler was working that morning. Thus Peppermint Grove became the site for further studies of truly suburban air. Long-term studies Photochemical sequelae would be more likely if car emissions remained relatively unmixed throughout a full morning of intense sunshine: I resolved then upon a year long examination, for just such persistence, in the air at home between 10.00 a.m. and noon. The year’s results are really a chemist’s view of the fluctuating meteorology of that year. Examination of the maximum value (10.00 a.m. to noon) obtained in every month locates April, May, June and July to be when inadequate morning mixing occurs. Peppermint Grove experienced some ten inci- dents in a year comparable to the April 29th, 1971, event at the University. If data had been collected 8.00 a.m. to 10.00 a.m. (not two hours later) I believe we would have exceeded the 23.6 pphm figure. The main meteorological cause of the inability of the air to disperse car fumes is the existence of a shallow and intense inversion layer, the air temperature rising with height above ground instead of decreasing as normal. Dispersion of contaminants is delayed until the sun rises high enough to destroy or ‘break-up’ the inversion, when sudden mixing and therefore clearing occurs. The data for 1962 shows how common these inversions can be in Perth (Mackey 1963). The chemical data confirm these shallow in- versions as an extremely important feature of Perth’s meteorology in relation to car and other low-level emissions. Let me offer just one statistical argument wh‘ch directs the blame for the NOx levels Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 2 towards cars. Day by day the values fluctuate remarkably because of Perth’s variable weather, but if you virtually eliminate that factor by averaging the results for every Monday, every Tuesday etc,, then this Table is obtained (over 55 weeks) : Mon. Tues. Weds. Thurs. Frid. Sat. Sun. 1-28 1-22 1-32 1-64 1-79 1.37 1-07 Highest Lowest which surely reflects our car habits. Very simple, very convincing! The scope of these studies was greatly en- larged hereabouts by parallel work at Caporn Street Nedlands, carried out by Dr. Frank Cattell as the Western Mining Corporation Fellow at the University of Western Australia in 1972 and 1973. Clear evidence was produced that Pepper- mint Grove and Nedlands experience somewhat similar effects from traffic emissions and that NO predominates in the ambient NOx mixture. The daily data for both sites has been smoothed to reduce the effects of short term weather fluc- tuations by taking a thirty day moving average: it confirms again that the early winter period is of greatest concern. Furthermore when the NOx on a daily basis is high at Peppermint Grove, it is high also at Nedlands. These highs are not from local sources like Chemistry’s boiler acting in an uncoordinated way, very likely other parts of the Metropolitan area are simil- arly affected. So far I have been concerned with morning traffic: what about the evening exhaust fumes? Mid April 1972 provided some novel and dis- concerting information. The Bureau of Meteor- ology issued an air dispersion alert on April 14th, 1972, for the Friday evening to Saturday noon. The University wind records showed a forty hour calm starting 6.00 p.m. Friday the 14th. Thus alerted we instituted some special studies. We believe the massive late evening peaks to be due to the evening traffic emissions drifting very slowly westward from the city whilst held at high concentration in a very shallow layer by radiation inversions. Perth weekday traffic covers at least a million car-miles between 3.00 p.m. and 8.00 p.m. At the then current Californian standard of 4 grams of NO per mile, this human mobility releases 3 tonnes of NOx. If that is dispersed into a cylinder of air ten kilometres diameter and 100 metres deep (a rough guess for the affected air space) then the average concen- tration is 20 pphm. The higher values in the evening peaks surely provoke speculation about other car emission chemicals. As a rough and ready rule overseas experience is that carbon monoxide levels are 40 times the NOx: if so for our Perth suburbs on these special occasions the United States air quality standards for carbon monoxide (not to be exceeded more than once a year) were violated on at least four evenings. Photochemical events and ozone formation Should the incidents detected in the evenings persist until well into the sunlit morning, per- haps with additional NOx gained from the morn- ing traffic emissions, then photochemical con- s quences would certainly occur. Possibly the April 29th, 1971, morning peak was indeed one such incident. Dr. Cattell made some preliminary assessment of ‘oxidant’ (a photochemical component) by the buffered potassium iodide method and by the ethylene chemiluminescent reaction method which is specific for ozone. A study was made of how N 0 ,N 02 and ozone levels varied during one winter’s day, the time sequence is typical of photochemical reactivity. Ozone concentra- tions on one of three summer days in November- December 1972 (part of a short series studied) reached levels which American authorities would regard as indicative of mild photochemical smog. I mention-caveat-that 1972 was an unusual year weatherwise: ten air dispersion alerts were issued by the Bureau of Meteorology, compared to 2 in 1973 and (I think) one only in 1974. (current practice is different too.) A comfortably convenient viewpoint might well be that the air pollution features I have been talking about are really only of marginal concern in these spec al periods. Of course I can give you no guarantee for the future: our car traffic con- tinues to grow, the weather is fickle — Melbourne recently had an unusual 11 day stagnation period. It is however certain that the basic air- chemistry here gives no built-in immunity, for that you must turn to town planners, controls, politicians, and perhaps the electric car. Ozone levels in Perth were first “measured” almost a century ago. Towards the end of 1875 Sir Malcolm Fraser, then Surveyor General, had established at the Survey Office a meteorological station under his personal supervision and I think at his own expense. The data obtained are reported in the Government Gazette and starting August 18th 1876 include values for the blue colour developed on potassium iodide paper as natural air movement brought the ozone to it, not a method quantitatively validated today. Ozone was then thought to be good for you, you got it at the sunny sea-side and not in indus- trial towns because there the sulphur dioxide obliterated it: Perth scored well. We know some- thing of the instrumentation: Sir James Clark’s Ozone Cage, Dr. Moffat’s papers and ‘scale of tints’, ^ Negretti and Zambra’s superior papers were in use later. Do specimens of these early ozone meters abandoned in 1900 still exist? Use- ful comparative information might yet be ex- tractrd from the historical Perth data? Emissions from vegetation fires Do we have sufficient knowledge of how Perth’s air chemistry is affected by burning-off opera- tions either far removed from the metropolitan area or by major local incidents [see postcript] such as the recurrent fires at Herdsman’s Lake? The logs of Pelsaert, Vlaming and others show that dense smoke patches are not merely modern Journal of the Royal Society of Western Australia, Vol. 59. Part 1 May 1976 3 events. The West Australian reported (Friday, December 11th, 1970) that ‘A smoke haze cover- ing the metropolitan area yesterday afternoon restricted visibility in some places to 1000 yards. Visibility was poor all day and at 5.00 p.m. it was impossible to see across the Swan River from ‘Newspaper House’.’ Earlier that week on a day of similar obscuration my note book (and ‘The West Australian’, Dec. 9th) records that at the edge of the Darling Scarp I could definitely smell the Los Angeles odour. That December week the CSIRO Division of Applied Chemistry (I learned rather surprisingly two years later through the scientific literature) were studying large scale prescribed forest fires in the Manjimup area; the fire on 7/12/70 con- sumed 110,000 tons of fuel, and the smoke from the one on the 9th Dec. could have reached Perth the next day. Aircraft flew through the smoke plumes to measure concentrations of smoke particles and many chemical substances. No chemist expects ozone to form in a fire, directly through combustion alone. No ozone out- side the normal natural air ranges was found in these plume experiments and 50 pphm for NOx was never reached either. In the 1971 re- port’s words ‘. . . it seems most unlikely that any prescribed burn could, of itself, start a photochemical incident’ (Vines et al. 1971). The matter has not ended there. Why “of itself”? Since 1971 additional information has come from new CSIRO forest burn experiments. Ex- cess ozone has now been detected, the peak concentration being several times that of the ambient atmosphere, though d quote) ‘excessive ozone was found only in the top layer in the plume and only when the sun was shining’ (i.e. photochemical) . Let me read the two concluding paragraphs of a CSIRO paper available in January last year (Evans et al. 1974) : “It is indeed fortunate that the excess ozone is confined to a shallow layer at the top of the plume and that the plume from controlled burn- ing rises above 1000 metres, thereby ensuring sufficient dilution of the ozone layer to eliminate any health hazard at ground level. Nevertheless, one should keep in mind that low-level smoke may well constitute a hazard, particularly if it drifts over urban areas and merges with other urban pollutants such as nitrogen oxides, the effects of which cannot yet be predicted. We feel that it is strange that such high ozone concentrations have not been obvious elsewhere in the world. Perhaps the indigenous eucalypt fuels used in the present work yield smokes which are particularly photosensitive.” Local research is impoi*tant. Perhaps my nose was right after all on Decem- ber 8th, 1970! Ozone from massive burns? Ozone from cars? Are the native plants (whose sen- sitivity to ozone we do not know) absolutely safe in Kings Park? The experimentalist you see must first speculate, then measure. Power stations as ozone sources Until recently no one expressed concern about gaseous emissions from oil-fired power stations except over sulphur dioxide. Through steadily improving technology more compact, quickly responding and more sensitive analytical chemi- cal apparatus can now be flown through stack plumes. Just such an experiment was reported by Davis et al. (1974) on the plume gases from a 1000 megawatt generating station near Wash- ington, D.C., which operates on 75% oil 25% coal. In addition to the expected chemistry near the stack, at distances around 70 kilometres down wind through sunlight action on the efflu- ent gases there was a nett production of ozone to levels double ambient values. A surprising finding which will provoke much new experi- mentation and some concern where large power stations are situated to windward of urban areas. The role of a university What should be the University’s role in fur- ther atmospheric studies in Perth? It was Joseph Conrad who authoritatively described Cape Leeuwin as ‘one of the three great Capes in the World’. Far to its west, north and south we have available a vast portion of the water sur- face of the world. Much of Perth’s air comes to us over those oceans with — by world stan- dards — remarkably little air-borne pollution from the few ships and the infrequent aircraft. Perth then is well sited for background studies of numerous naturally occuring chemicals in the uncontaminated atmosphere. Let me cite two examples only. Some splendid studies on gaseous iodine (not sodium iodide, the salt) have re- cently been done at the University of Hawaii (Moyers et al. 1971). A southern hemispherean replication would be a proper task. Secondly, if speaking to you three years ago I would have been urging studies on the rising levels of cai’bon dioxide. Again the deficiency that such studies had only been done in the northern hemisphere (predominantly) is now being corrected by CSIRO’s program of Base Line Atmospheric Carbon Dioxide Monitoring on high altitude commercial flights between Christchurch, New Zealand, through Perth to Mauritius (Pearman & Garratt 1973, 1975). Welcome as this develop- ment is to me, it does point up the question of whether the University can organise, staff and finance environmental programs on more than a very minor scale unless there is a firm com- mitment from senior policy makers, perhaps even a national science policy. Consider — sketchily — the time scale within which an Honours student in one year or a Ph.D candidate in three must face his problem. Lectures, associated study, three or more sem- inars weekly, laboratory teaching, all create a fragmented week within which to cope with the experimental problem and its literature, the apparatus and its problems, and for the Ph.D candidat"^ — getting publishable work. For the regular staff member to deviate from his estab- Lshed research means serious inroads into his private time — a fact University wives will verify. Add that the air is variable daily, weekly, sea- Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 4 sonaly, so that many observations are needed — a problem shared with biologists and farmers. Despite these very real constraints the Univer- sity can act in exploratory roles of value to the Public Health Department and the Environ- mental Protection Department and this I wish to illustrate briefly with the two examples of lead, and “Freon” tracing. Lead emission from cars There is world wide interest in the urban distribution of lead-containing particulates emitted from the exhaust pipes of cars. Cer- tainly humans near highways inhale lead from the air though the public health significance of this is controversial. Mr. L. Boujos as an Honours student in Physical and Inorganic Chemistry obtained local information on lead levels which supplemented the rather meagre public infor- mation available elsewhere in Australia (Bot- tomley &: Boujos 1975). The Causeway across Heirisson Island carries some 75,000 vehicles per day. In the 200 yard length of roadway on the Island roughly two pounds weight of lead is emitted daily through petrol combustion. The finely divided lead salts are brought to the ground to various extents and at various distances depending on the pre- vailing wind and weather conditions, that reaching the ground or plant surfaces is subject to weathering and leaching. Sampling was undertaken to determine the average lead concentration in the top two centi- metres of soil at various lateral distances from the edges of the roadway. The absolute values are somewhat lower than for some other Aus- tralian measurements and are considerably below extreme values for heavily urbanised areas in other countries. Whether simple intercomparison is legitimate or not is a matter itself requiring very close study. The lead is of course not uniformly distri- buted in depth: there is a markedly raised con- centration in the upper few centimetres of soil both near to and remote from the roadway. The immediate surface is richer still in lead, a fact of considerable importance for surface living and feeding animals. The shells of a common snail iPheha visana) at 50 metres from the roadway were five times higher in lead than comparison samples taken at Swan- bourne Beach and at Rottnest Island. These snails are eaten by mice, themselves food for hawks. The levels observed at Heirisson Island cer- tainly do not account for all the lead emitted in past years. Some perhaps is broadcast very widely, some may be carried deep into the soil during the winter rains. Nor do we have any details of the lead run-off from the roadway into the storm drainage system, the riw'r and the ocean. We have kept the recovered lead from Heirisson Island and Dr. Trendall urges an isotopic analysis. An interesting experiment would be to follow the arrival rate of lead at newly created sur- faces: we are doing this on artificial soil at the University, but what about the accumulation in artificial lakes near the Narrows and Hamilton interchanges — both bird sanctuaries in effect. Possible lead insults to the flora and fauna should be examined further if Heirisson Island is to become a zoological reserve (Quarles et al. 1974; National Academy of Sciences 1972). Let me now deflect your interest temporarily to lead contamination in a very distant part of the world, age-dated Greenland glacier ice (Murozumi et al. 1969). There is evidence for regional or hemispheric pollution consequences ot the boom in extraction, processing, and utilisation of lead as the Industrial Revolution encompasses Europe, and further lead deposition is thought to be due to the general adoption of lead antiknock compounds combined with the expansionary phase of car traffic. The Perth metropolitan area doesn’t have con- venient local glaciers, but can we think of a comparable test here? Perhaps dateable sedi- ments in the Swan Estuary? Another possibility is a chemical examination of the annual growth rings of trees growing during the last fifty years alongside busy highways in Perth. Dr. Wycherley has provided me with samples from Kings Park, but what I really covet (I confess) is one of the Norfolk Island pines alongside Stirling High- way at Christ Church, Claremont. Tracing of gas emission By the early seventies a small number of Freon tracing experiments in the Northern Hemisphere had provided information on ground level concentriations of emissions from large scale powerstations as follows. Several kilo- grams of a Freon refrigerant gas are released up the working stack, and are diluted and dis- persed just as the normal gases are under the prevailing weather. On the ground at distances up to 20 kilometres many air samples are taken in plastic bags for later laboratory analysis. This method is practicable only because as little as 10'^^ gram of Freon in 1 cm'^ of air is detectable by gas-liquid chromatography. (On a mass per volume basis, we could detect one needle in 100 large haystacks). This extreme sensitivity is needed to study other industrial emission prob- lems. In Western Australia this system has been applied to determine ground level concentra- tions in the Coogee Air Pollution Study (Anon. 1974). It is worth a few moments delay to trace the origins of this local experiment, the first we believe in Australia, possibly in the southern hemisphere. The University Department of Organic Chemistry has owned a g.l.c. with e.c.d. (to use the jargon) for some years, invariably fully deployed but in very diffei'ent directions. Perceiving the method’s potentiality I wrote in mid 1973 to Dr. O’Brien suggesting that Dr. Cattell (then Western Mining Corporation Fel- low) might develop the local instrumentation appropriately. The Department of Environmen- tal Protection with personnel assistance from the armed services and from school children arranged the logisitics of release, collection, and analysis on a day determined by the Common- Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 5 wealth Bureau ot Meteorology. Remove any one of these links or tl\3 associated financial support and no data can be secured. Cooperation is im- portant. Such information is manifestly important to planning large scale industry within the metro- politan area or further afield and to practical verification of theoretical and simplified models of pollutant dispersion. Diverse additional appli- cations come to mind easily. The emissions of one particular chimney can be distinguished from another emitting the same pollutant. Car emissions could be traced across the city suburbs after release from say a single car crossing the Narrows bridge. The movement of air into and from individual buildings can be followed in time and in amount. Information on persis- tent and troublesome odours can be obtained. My own home is frequently invaded by a smell similar to ‘iodoform’ arriving on gentle S. or S.E. winds in the evening. “Freon” releases from conjectured sources (decomposing lakes vege- tation is one theory) could help disentangle the problem. The extreme sensitivity of the g.l.c. method might permit positive chemical identi- fication of the offending chemical. State Government funds, I am very glad to re- port, have been made available to the Depart- ment of Conservation and the Environment for an additional chromatograph housed at the Uni- versity with priority use on further tracing experiments of these types. I give one more academic example: industrial and domestic use of simple chlorinated hydi'o- carbons had resulted in the Northern Hemis- pherean air being burdened with escape material (Murray et al. 1973; McConnell et al. 1975; Wilkneiss et al. 1975). Comparable measure- ments using the chromatograph on Metropolitan air, for Indian Ocean air and at the Giles Weather Station would make an attractive Honours project and a useful contribution to world environmental information. Funding of relevant university research Recently the Organisation for Economic Co- operation and Development (1974) reviewed scientific and technological activity in Australia. The report states that ‘. . . much more could be achieved if industries, State agencies, and especially CSIRO would engage the Universities more in research work’ and in urging that the Universities need more staff especially to develop their research capacities to fulfil the needs of society, notes that there cannot be a strict pre- rogative for so called academic research work. If the report’s viewpoint becomes accepted Federal Government policy then the Universities must meet those suggestions by the creation of numerous purely research positions from Post Doctoral Fellows through to fulltime and per- manent senior research appointments, which at present within Australia are virtually non- existent and in striking contrast to the position in centres of excellence elsewhere in the world. For almost an hour you have been patiently absorbing my suggestion that atmospheric chemistry is a field to which those recommen- dations might properly be applied at the ter- tiary educational level. Studies in the atmos- pheric environment may however extend from very advanced research levels of physical and mathematical sciences downwards to observa- tions and experience by John Citizen’s children, this is a superbly interdisciplinary subject which should I believe be part of everyone’s education. Whatever your assessment is, please bear in mind that in ‘three score years and ten’ most of us will breathe some five hundred tons of air, and what it contains. Postscript . — Professor R. T. Prider kindly drew my attention at the close of the Address to the fire at Spectacles Swamp, Mandogalup, April 1939, having badly affected Perth residential suburbs with its smoke. The Annual Report of the Mines Department, Western Australia, for 1939 (p. 148) refers to ‘choking fumes . . . plainly and objectionably noticeable even in Perth 20 miles away’. The ‘West Australian’, April 28th 1939, p. 24, describes 1000 acres alight for about three weeks and includes a photograph of soil damage. Two days previously a letter headed ‘Mandogalup Peat’ and signed R.V.R. recalls its author’s prophesy in 1914 when swamp drainage was undertaken. References Anon. (1974).— ‘Coogee Air Pollution Study’ Environ- mental Protection Authority, Perth. Western Australia. Baird, A. M. (1958).— Notes on the regeneration of vegetation of Garden Island after the 1956 fire. Journal of the Royal Society of Western Australia 41: 102-107. Bottomley, G. A., and Boujos, L. P. (1975). — Lead in soil of Heirisson Island. Western Australia. Search 6: 339-390. Bottomley. G. A., and Cattell, F. C. ( 1975) .—Nitrogen oxide levels in suburbs of Perth, Western Australia. Journal of the Royal Society of Western Australia, 58: 65-74. Davis, D. D.. Smith G., and Klauber. G. (1974) .—Trace gas analyses of power plant plumes via aircraft measurement: O 3 , NO^, and SO> chemistry. Science 186: 733-736. Drover. D. P. (1960). — Accessions of sulphur in the rainwater at Perth and Nedlands. Western Australia. Journal of the Royal Society of Western Australia 43: 81-82. Evans, L. F.. King, N. K.. Packham, D. R., and Stephens. E. T. (1974). — Ozone measurements in smoke from forest fires. Environmental Science and Technology 8: 75-76. Hatch, A. B. (1959).— The effect of frequent burning on the jarrah (Eucalyptus marginata) forest soils of Western Australia. Journal of the Royal Society of Western Australia 42: 97- 100 . La Mer, V. K. (1948). — Monodisperse colloids and higher order Tyndall spectra. Journal of Physical and Colloid Chemistry 52: 65-76. Mackey, G. W. (1963).— A survey of the meteorological aspects of “smog” formation in Perth. Bureau of Meteorology Working Paper 58/ 1937 of July 1963. McConnell, G., Ferguson. D. M.. and Pearson. C. R. (1975). — Chlorinated hydrocarbons in the environment. Endeavour 34: 13-18. Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 6 Merrilees, D. (1968). — Man the destroyer: Late Quater- nary changes in the Australian marsupial fauna. Presidential Address 1967. Journal of the Royal Society of Western Australia 51: 1-24. Moyers, J. L., Zoller, W. H., and Duce, R. A. (1971). — Gaseous iodine measurements and their relationships to particulate lead in a pol- luted atmosphere. Journal of Atmospheric Sciences 28: 95-98. Murozumi, M., Chow, T. j., and Patterson, C. (1969). — Chemical concentrations of pollutant lead aerosols, terrestrial dusts, and sea-salts in Greenland and Antarctic snow strata. Geochimica Cosmochimica Acta 33: 1247- 1294. Murray, A. J., and Riley, J. P. (1973). — Occurrence of some chlorinated aliphatic hydrocarbons in the environment. Nature, London 242: 37- 38. National Academy of Sciences, Washington, D.C. (1972). — Biologic effects of atmospheric pollutants. ‘Lead, Airborne Lead in Perspective’. Organisation for Economic Cooperation and Develop- ment (1974). — “OECD Examiners Report on Science and Technology in Australia” . Aus- tralian Government Publishing Service, Canberra. Pearman, G. I., and Garratt, J. R. (1973, 1975).— The CSIRO (Australia) base-line atmospheric carbon dioxide monitoring programme — progress reports. CSIRO. Division of Atmos- pheric Physics, Aspendale, Victoria. No. 2 December 1973, No. 3 February 1975. Quarles, H, D. Ill, Hanawalt, R. B., and Odum, W. E. (1974). — Lead in small mammals, plants and soils at various distances from a highway. Journal of Applied Ecology 11: 937. Speck, N. H. (1953). — Atmospheric Pollen in the City of Perth and environs. Journal of the Royal Society of Western Australia 37: 119-127. Tyndall, J. (1868). — On a new series of chemical reactions produced by light. Proceedings of the Royal Society of London 17: 92-102 and 224-233. Vines, R. G., Gibson, L., Watch, A. B., King, N. K., MacArthur, D. A., Peckham, D. R., and Taylor, R. J. (1971), — On the nature, pro- perties and behaviour of bush-fire smoke. CSIRO. Division of Applied Chemistry Technical Paper No. 1. Wallace, W. R. (1966). — Fire in the jarrah forest en- vironment. Presidential Address 1965. Journal of the Royal Society of Western Australia 49: 33-44. Went, F. W. (1966). — On the nature of Aitken conden- sation nucleii. Tellus 18: 547-556. Wilkneiss, P. E., Swinnerton, J. W., LaMontagne, R. A. and Bressing, D. J. (1975). — Trichlorofluoro- methane in the Troposphere : distribution and increase 1971-1974. Science 187: 832-834. Wilsmore, N. T. M. (Chairman) Salinity in Soils Com- mittee (1928-9). — Salinity of rain in Western Australia. Journal of the Royal Society of Western Australia 15: 22 - 20 . Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 7 2. — The changing ecology of Western Australian wheat production by W. J, R. Boyd\ W. K. Waterhouse^ B. B. SinghS ^ N. A. CampbelP and N. A. Goodchild" Manuscript received 18 March 1975; accepted 15 December 1975. Abstract The ecology of wheat production in Western Australia has changed considerably since the commencement of wheat breeding in 1902. The changes involved are numerous and include technological developments such as the appli- cation of artificial fertilizers, correction for trace element deficiencies, increased use of mechanization, herbicides and insecticides and, the introduction of subterranean clover, which encouraged changes in crop rotational practices. Collectively the changes which have occurred led to greater diversification of farming systems and to the declining popularity of those cul- tivars that were previously adapted. Further changes in crop management practices are considered necessary if the cultivars which are currently dominant are to be replaced. Introduction Agricultural ressarch commenced in Western Australia in 1894 with the establishment of the Bureau, later the Department of Agriculture (Anon. 1967). Since then, under the guidance provided and through the release of more adap- table cultivars, wheat production and the acre- age sown to wh3at have increased considerably (Figure la). In contrast, improvements on yield per hectare have been gradual (Figure lb). In addition to these production statistics, which reflect considerable seasonal variability, changes have occurred in the regional distribution of the area sown to wheat, in the farming systems and rotational practices employed and, in crop management technology; including mechaniza- tion, fertilizer practices and herbicide use. The purpose of this paper is to trace the agro-ecological changes which have occurred in wheat production in Western Australia and to examine their influence on cultivar adapability. Historical (aj Early Estahlishvient Wheat production commenced with the initial settlement (1826) using cultivars of English origin which were, relative to the growing condi- tions prevailing, late maturing (Berthoud, 1903). 1 Agronomy Department and -Biometrics Unit, Institute of Agriculture, University of Western Australia, Nedlands, W.A. 6009. ^ Division of Mathematics and Statistics. CSIRO, Wemb- ley, W.A. 6014. Present addresses: -^The Education Department of Western Australia. Parliament Place, West Perth, W.A. 6005; "’Department of Agriculture, Stock & Fisheries, Lae. Papua New Guinea. For this reason their cultivation was restricted to the south-west coastal areas where reasonable prospects for spring rainfall occur (Berthoud, 1905). The discovery of gold at Coolgardie (1892) and at Kalgoorlie (1894) provided the incentive for an increase in population and for the ex- pansion of inland communication and rural water supplies. These events, in turn, encouraged an increased interest in agriculture which in- tensified following the decline in gold production, in 1903. As a result there was increased clearing of land despite limited agronomic information and the absence of suitable adapted cultivars. Research stations were established at Narrogin (1901), Nabawa (1902), Merredin (1907) and Avondale (1911). In 1905 G. L. Sutton, a con- temporary and disciple of W. Farrer, was ap- pointed Cerealist (later Director of Agriculture) and brought with him, from Wagga (N.S.W.), F:j seed of the cross Gluyas x Bunyip which was to have a major impact on Australian and local wheat production for many years. (b) Breeding and Cultivar Popularity Continuous export of wheat commenced from Western Australia in 1907 largely due to the popularity of the cultivar Gluyas.^ In 1915, from the cross Gluyas x Bunyip Sutton released Nabawa and by 1926 this cultivar occupied nearly 50% of the area sown to wheat; becoming in addition, the most popular cultivar in South Australia (1930-34) and in New South Wales (1930-35). In 1934 the supremacy of Nabawa gave way to Bencubbin; a cultivar derived fi'om back-crossing Nabawa back to its Gluyas parent. (Figure 2). Bencubbin, which flowered earlier than Nabawa, proved even more popular, and with its relative Gluclub, dominated wheat pro- duction in Western Australia until 1945 as well as becoming the leading cultivar in South Aus- tralia and New South Wales throughout the 1940’s. The popularity of Bencubbin (and Gluclub) then declined somewhat in competition with Bungulla, an early flowering line selected from Bencubbin, and together, these various “Gluyas” descendants retained their popularity until the 1953/54 season. (Figure 2). In today’s local terminology the relative flow^ering habits of the Nabawa, Bencubbin and Bungulla would be considered as late, late/mid-season and early, respectively. The original cultivars of English origin would, by comparison, be “very late”. 1 A farmer selection in South Australia from Words’ Prolific: itself a farmer’s selection from the South African variety Du Toit. Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 8 36 32 28 24 20 16 12 8 4 POPULATION (millions) Figure 1. (A) Changes in acreages sown to wheat and pasture and, in sheep population (1900-1972.) — (B) Variation in state average wheat yields (1900-1972). The popularity ot the “Gluyas” cultivars de- creased with the introduction of early/mid- season flowering semi-dwarf cultivars bred in New South Wales and Victoria. These cultivars. Gabo and Insignia, first introduced in the early 1950’s, and their respective descendants. Gam- enya and Heron, steadily increased in popularity until pressure for reasons of quality reduced the popularity of Insignia and Heron. This brief review draws to attention the demise of local cereal breeding, following a period of prolonged success, and a gradual shift toward earlier flowering cultivars of short-stature. This is interpreted as indicating a change in response to changes in the crop management practices under which wheat is grown in Wes- tern Australia and an attempt to account for it is presented below. (c) Changes in Farming Systems and Cultural Practices In the early days of settlement the differential land clearing of soils of high fertility led to sporadic settlement throughout those areas in which spring rainfall was sufficiently reliable to provide for the very late maturing cultivars then Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 9 so 70 - Figure 2— Changes in wheat cultivar popularity (1929-1972). N = Nabawa (Gluyas x Bunyip— released 1915). Be = Bencubbin (Gluyas x Nabawa — released 1929). Gc = Gluclub (Gluyas x Clubhead — released 1922). Bu = Bungulla (Sel'i from Bencubbin — released 1939). Gb = Gabo (Bobin x Gaza x Gular — released 1945). Gm = Gamenya (Gabo-’ x Mentana//Gabo2 x Kenya 117A — released 1960). In - Insignia (Ghurka x Ranee — released 1946). H = Heron {R.D.R. x Insignia-^ — released 1959). F = Falcon (Dundee x Gular2//Bencubbin — released 1960). available. The limited extent of such soils and their exploitive use led to declining production and widespread recognition of the need for fer- tilizers, particularly phosphates (Callaghan & Millington, 1956; Wicken, 1904; Mann, 1905). Such a trend is demonstrated by the declining yield/hectare between 1900 and 1925 (Figure lb). However, the increasing availability of phos- phate fertilizer and the release of more adaptable cultivars, as discussed above, arrested this trend and encouraged a further increase in the area sown to cereals (Figure la). This expansion was suspended temoorarily during the y:ars of the denression and World War II and, again, in 1968-71. due to the imposition of quotas. During the 1920’s and 30’s cereal crops were managed with horse-drawn equipment under a crop-fallow rotation which permitted pre-season weed control and early seeding to which the late and late/mid-season cultivars then available were well adapted. Early flowering cultivars such as Sunset and Noongaar were released but at- tained little popularity. These generalizations represent a consensus of opinion expressed by Thomas and his co-workers (1926 to 1933) who were responsible for much of the experimenta- tion that developed the basis for current cultural practices. The practice of bare fallow under continuous cultivation resulted in deterioration of the physical status of soils and emphasized the need to incorporate a pasture break in the rotation under a ley-farming system (Shier, 1956). At the time the availability of suitably adaoted leguminous pasture cultivars was limited. The eventual release of sub-clover cul- tivars led to an exponential expansion of the pasture acreage (Figure la), a matching in- crease in sheep numbers (Figure la), increased soil fertility and a gradual conversion to crop- pasture rotational practicss. A measure of these modifications on farming systems is provided by regional changes in the proportional utilization of cleared land for cereal crops (Figure 3). The spread of sub -clover pastures and adoption of ley farming coincided with other significant technological events including the Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 10 REGIONAL DISTRIBUTION OF FARMING SYSTEMS SINCE 1929 Figure 3.— Regional distribution of cropped acreage, at intervals from 1929 to 1967. A <10% Cereals, balance in pasture. B >10 <25% Cereals, balance in pasture. C >25 <50% Cereals, balance in pasture. D >50 <75% Cereals, balance in pasture. E >75 Cereals, (shaded) balance in pasture. U Unallocated. recognition of, and correction for, trace element deficiencies (Burvill, 1965), the replacement of horsepower by mechanization and the intro- duction of herbicides and weedicides to control increased weed problems arising from the ab- sence of fallow. The use of artificial fertilizers, principally superphosphate, has increased in conjunction with crop and pasture acreage. This has led to improved soil phosphate levels and further enhancement of soil fertility status, even though, from 1925 to 1960, application per unit area (ranging between 0.113 to 0.125 tonnes/ha has hardly altered. (Figures 4a and b). The change in rotational practices together with the adoption of technological developments have stimulated more diversified farming systems (Figure 3) and, in more recent years, greater intensification within them — as evidenced by increased use of nitrogenous fertilizers (Figure 4). These developments have had num- erous implications in the management of cereal crops; including the need to delay seed-bed preparation so as to conserve pasture for in- creased stock numbers and, a decrease in the use_ of c?real crops for purposes other than grain production i.e. for hay and stubble graz- ing purposes. Delay in seed-bed preparation has contributed to the later seeding of cereal crops and greater weed control problems — disadvan- tages which have been only partly compensated for by increases in soil fertility and increased mechanization because of the need to awaft opening rains and weed germination before cultivation can proceed. Discussion The introduction of leguminous pastures in particular, and the adoption of improved fer- tilizer and weed control technologies, have directly led to diversification of farming systems in Western Australia, and to changes in crop management practices. These changes include conversion from crop-fallow rotational practices favouring pre-season weed control, early seeding and the us3 of tali growing late to late/mid- season cultivars, to crop-pasture rotations characterized by increased soil fertility, in- creased weed problems, and an enforced delay in seedings to permit seed-bed preparation Under these circumstances short-stature culti- vars of early to early/mid-season maturity have gained in popularity. (Figure 5). The sequence of events which has occurred represents a clas- sical case history of changes in cultivar adap- tion because, at no stage has disease or epn- Journal of the Royal Society of Western Australia. Vol. 59. Part 1, May, 1976. 11 YEAR Figure 4.— Changes in fertilizer consumption, particularly superphosphate (A) and in application rate (B), on both crops and pastures siderations of quality been responsible for directing major changes in cultivar use. Un- published data indicate that whilst the currently popular cultivar (Gamenya) is over 20% more productive than its predecessor fBencubbin) under crop-pasture rotations, these cultivars are of equivalent performance if planted early on fallow. In fact Bencubbin was still being recom- mended for early planting as recently as 1965, even though its popularity was low (Figure 2). Despite the current and relative yield advan- tage of Gamenya over Bencubbin district aver- age wheat yields/ha have increased to a more limited extent. This suggests, that the contribu- tion of plant breeding in recent years has been one of developing cultivars more adapted to the changes in cultural practices and farming systems that have taken place. Journal of the Royal Society of Western Australia. Vol. 59. Part 1. May, 1976. 12 YEAR Figure 5. — Changes in the adaptive characteristics of wheat cultivars with changes in cultural practice. The changes which have occurred illustrate the important ecological principle of a dynamic, and delicate, balance between genotype and en- vironment. Modification of the environment affects the productivity and survival of biological organisms within the biosphere. Man’s current concern over his own survival as a consequence of his interference with the en- vironment (e.g. pollution of air and water) stands in sharp contrast to his ancient and continuing need to modify the agricultural environment for the benefit of crops he grows and the livestock he tends. With increasing understanding of plant growth, development and genetics, man has the capacity, through plant breeding, to modify his crop plants to exploit the environmental conditions he can provide for them. The changes in cultural practices dis- cussed in this paper have had the effect of rendering less adaptable and less efficient those cultivars that were once so popular (Figure 2). The continued popularity of Gamenya, despite breeders efforts to improve upon its performance, suggest that additional changes in the crop en- vironment could be a prerequisite to further breeding progress. As the previous changes led to greater diversification of farming systems it is most probable that increased intensification within those systems will now become more urgent. Acknowledgements. — The W.A. State Wheat Industry Research Committee and the W.A. Barley Research Trust Fund provided financial support. References Anon. (1967). — “Western Australian Year Book”. Govt. Printer, Perth, Western Australia. Berthoud, D. (1903).— Notes on State Farm, Hamel. W.A. Jnl. of Agriculture (1st Series) 7: 366-377. Berthoud, D. (1905). — Notes on State Farm, Hamel. W.A. Jnl. of Agriculture (1st Series) 11: 300-403. Burvill, G. H. ( 1965) .—Plant nutrition in Western Australia. Jnl. Dept. Agric. W. Aust. (4th Series) 6 (6): 353. Callaghan, A. R., and Millington, A. J. (1956). — “The Wheat Industry in Australia”. Angus and Robertson, Melb., 486 p. Mann, J. (1905). — Wheat experiments with chemical fertilizers. W.A. Jnl. of Agriculture (1st Series 11: 225-227. Shier, F. L. (1934). — Modified rotation for the wheat belt. W.A. Jnl. of Agriculture (2nd Series) 11: 254-260. Thomas, I., and co-workers. (1926).— Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 3: 202-211. Thomas, I., and co-workers. (1927). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 4: 166-210. Thomas, I., and co-workers. (1928). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 5: 73-87 and 188-256. Thomas, I., and co-workers. (1929). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 6: 127-154; 311-316. Thomas, I., and co-workers. (1930). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 7: 173-194; 270-290. Thomas, I., and co-workers. (1931). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 8: 116-146; 238-259. Thomas, I., and co-workers. (1932). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 9: 56-86; 197-211. Thomas, I., and co-workers. (1933). — Field experiments with wheat. W.A. Jnl. of Agriculture (2nd Series) 10: 17-27; 219-225. Wicken, A. (1904). — Experimental wheat growing. W.A. Jnl. of Agriculture (1st Series) 10: 488-489. 3. — The nutrients and plants of Lake Joondalup, a mildly eutrophic lake experiencing large seasonal changes in volume by R. A. Congdon^ and A. J. McComb^ Manuscript received 17 June 1975; accepted 21 October 1975. Abstract Lake Joondalup is a shallow body of fresh water in a calcareous stable dune system 6 km from the ocean. The fringing and aquatic vegetation is described. Cation ratios resemble sea water apart from relatively high calcium, attributed to leaching from limestone. Seasonal changes in volume greatly affect ionic concen- trations. A bloom of the green alga Dispora coincides with low volume and high ion levels. Total nitrcgen is relatively high compared with phosphorus. The surrounding land is becoming- urbanized, and the data provide a baseline for future reference. Introduction Problems associated with the enrichment of lakes in urban areas, with ‘blooms’ of planktonic algae (some of them toxic to animals), increased bacterial activity, and oxygen depletion at depth, are of worldwide occurrence (e.g. Jackson, 1964). Press reports of the death in late summer of fish and birds in certain lakes of the metropolitan area of Perth indicate that Western Australia is no exception. While it is easy to suggest that the addition of nutrients derived from septic tanks, garbage disposal, and agricultural and lawn fertilisers, are primarily responsible for the eutrophication of these waters, quantitative data concerning nutrient levels and algae are lacKing. The present paper is concirned with the nutrients and plants of Lake Joondalup, which lies in a regmn of rapid urbanization and development. The lake is 32 km north of the centre of Perth and 6 km from the Indian Ocean, and is one of a chain of lakes which r._achEs Yanchep, 20 km further north. The lakes are linear, parallel to the coast, and lie in depres- sions in a Quaternary dune system (McArthur and Bettenay, 1960). They are an important component of the wetlands of the Swan Coastal Plain, which are being reduced in total area by draining and reclamation (Riggert, 1966). The aim of this work was to place on record the seasonal fluctuations in certain nutrients and in the density of planktonic algae to provide a reference aga'nst which future changes in the lake may be assessed. A description of the fringing vegetation is also included to allow comparison with Loch McNess in the Yanchep National Park (McComb and McComb, 1967), and because much of this vegetation, which must relate to the nutrient status of the water, whl be alt:red in the future. I Botany Department, University of Western Australia, Necllands, Western Australia. 6009. Materials and methods Collection of samples Water samples were collected from 6 sites (Figure 1) over a period of one year. Samples for oxygen analyses were collected directly into a sample bottle where the water was shallow, or with a Hals’s water sampler (Welch, 1948) where the water was sufficiently deep to^ allow use of this apparatus. Other samples w:re collected with a plastic bilge pump at a dsnth of 5 cm. All were collected between 1000 and 1600 hr. At each site measurements were m.ade of water temperature and pH (BpH Electromet:r, N. L. Jones, Melbourne). A comparative measure of light transparency was i^btained wi'h a 20 cm- diameter Secchi disc divided into black-and- white quadrants. Samt^les for nretallic cations were stored in clean 250 ml glass bottles, with 1ml of 1:1 (by vol.) nitric acid added to prevent adsorption and biological activity. Other samples were stored at 4** in thoroughly- washed, 5 1 polysthy- Lne jars. Watei' analysis Dissolved oxygen was initially measured in the field with an oxygen meter (Beckman Fieldlab Oxygen Analyser, Beckman Instruments, Fuller- ton, California) and subsequently by the azide mo(Iification of the Winkler Method (Anon. 1955). Conductivity was determined on return to the laboratory with a conductivity meter (E 382, Metrohm Ltd., Herisau, Switzerland) . Ammonia was determined by distillation and nessler.zation (Anon. 1955); organic nitrogen by Kjeldahl digestion (Anon. 1955) followed by titration with an automat'e titrator (W. G. Pye, Cambridge, England) ; inorganic phosphorus as orthophosphate by the molybdate-blue reaction using stannous chloride as the catalyst (Anon. 1955); total phosphorus as orthophosphate after perchlcric ac^d digestion (Robinson, 1941); and chloride by potentiem trie titration using a Clinical Chloride Titrator (4-4415, American Instrument Co., Silver Spring, Maryland). Car- bonate/hicarhonate was determined on one occasion using the double-indicator method (Anon. 1955) . Samples for analysis of metallic cations were given the pretreatment for total metals des- cribed by Parker (1972). Calcium and magne- s um we’-'e then determined by atomic absorption s*~ ctr''r^>^o^ometry (Var'an 1000, Varian Tech- tron Ptv. Ltd., Snringvale, Victoria), and sodium and potassium by flame photometry (EEL Flame Ph-t^m'-t'^r, EEL International Ltd., Bayswater, ip > Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May. 1976. 14 KEY OPCT>l WATER PAPERBARK FOREST SEDGE COMMUNITIES TYPHA COMMUNITIES EUCALYPTUS FOREST ACACIA WOODLAND Figure 1. — Lake Joondalup, showing general features, vegetation, and sampling sites. Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 15 Productivity Samples of 500 ml of water from each site were centrifuged (660xg; 10 min) and the super- natant removed with a suction pump to leave 10 ml of water plus centrifugate, which was transferred to a stopperzd flask and stored in the dark at 4° until phytoplankton counts were made with a Neubauer counting chamber. The problem of obtaining truly representative plank- ton samples for the lake is great, as the large area and shallow water leads to lack of homo- g.neity in phytoplankton populations. For example, although in October counts were gen- erally low at the six s tes, it was clear that a bloom, primarily of Oscillatoria and a colonial alga, was present in parts of the lake not in- cluded by the six sites. Nevertheless, seasonal trends are clear, main specizs were identified, and general variability is reflected in the stan- dard errors of the site means, as given below. Seston productivity was estimated on dupli- cated 500 ml aliquots, centrifugrd as for the phytoplankton counts. Each 10 ml residue was transferred to a 50 ml porcelain crucible, dried at 80° overnight, placed in a des'ccator for 30 min and weighed. It was then ashed in a muff el furnace at 650° for 30 min, cooled in a desiccator and reweighed. Seston productivity was calculated as loss on ignition in grams per litre. Observations and discussion General features The lake (Figure 1) is a shallow, closed body of fresh water 8 km long and 1.2 km wide at its broadest point. The depth varies seasonally but has never exceeded 3.3 m. The water surface is some 18 m above mean sea level, and the area is 6.1 km^ (610 hectares). There is no sur- face outlet for the lake, and it is believed that water passes through the limestone on the western shore, towards the sea. The western bank is relatively steep and little disturbed. To the east the land slopes more gently, and there is a housing settlement. Deep sewerage has be- come available in the area, and at the time of the study about 25% of the 630 houses present had been connected to the system.^ The mean depths of the sampling sites were between 10 and 190 cm during the period of the study, the changes in depth following changes in rainfall (Figure 2). By international stan- dards, therefore. Lake Joondalup is very shallow, and because of this shallow character the changes in depth represent large changes in lake volume. One would therefore expect signi- ficant seasonal changes in concentrations of dissolved ehments because of this factor alone. The water level has been even lower in the past, as indicated by the presence of submerged fence posts in the open water of the southern and eastern regions of the lake; there is evidence ^ The general information included in this paragraph was provided by courtesy of the Shire Clerk, Shire of Wanneroo. that the general water level of wetlands in the coastal plain has increased, presumably as a result of clearing (Evans and Sherlock, 1950; Speck, 1952). Figure 2. — The water level of Lake Joondalup (o), plotted with monthly rainfall data ( • ) for the townsite of Wanneroo, which lies immediately east of the Lake. Vegetation The distribution of main communities is shown in Figure 1, and a transset through the fringing cemmunitiss on the western shore is presented in Figure 3. Open Water and littoral fringes — Shallow waters, less than about 70 cm in summer, are densely populated with the benthic stoneworts, Nitella congesta and Chara baueri, and Pota- rnogeton pectinatus. Najas marina and Myrio- phyllum propinquum are also found in the shallow waters, but are more common in the deeper waters at the northern end of the lake. The floating plants Lemna minor and Azolla filiculoides were collected in the fringing sedge communities, but were never numerous during the period of study. The sedge Baumea articulata (Machaerina articulata, Cladium articulatum) is the dominant macrophyte of the lake, occurring generally in cure stands. It attains a height of 2 to 3 m and, through decomposition, is the main contributor to a ffbrous peat of the lake margins and sedge banks. B. articulata is replaced by Typha in resaocted areas where pasture cr roads encroach on the lake edge. Scirpus validus and Juncus pallidus also occur in restricted areas. Fen vegetation. — This is represented by a few small pockets of Baumea juncea (Machaerina juncea, Cladium juncetim) located on the nor- thern fringes of the two islands, and in scattered areas beneath swamp paperbarks. On Malap Island and on the lake margin to the north Le 2 rdosperma longitudinale occurs in drier areas behind B. juncea. Journal of the Royal Society of Western Australia. Vol. 59, Part 1, May. 1976. 16 WATER LEVEL (ems) 10 METRES 15 20 3 Banksia attenuata 4 Banksia grandis 5 Jacksonia furcellata 6 Eucalyptus marginata 7 Melaleuca raphiophylla 8 Baumea articulata Figure 3. — The vegetation fringing Lake Joondalup. The profile was drawn from a 2 m-wide transect situated at site T, Figure l. Paverbark woodland. — A woodland of swamp paperbark (Melaleuca raphiophylla) borders the fringing sedge communities, and the bases of the trees are inundated during the winter months. The woodland consists of a narrow belt of trees with dense, touching canopies, and often in- cludes Eucalyptus rudis. Baumea juncea and Centella asiatica are common. Seedlings of M. raphiophylla are established in meadows of Baumea articulata and this ‘carr’ is extensive on the north-western side of the lake and to the south-east of Malap Island. Surrounding woodland. — An ecotonal community can be recognized between the paperbark wood- land and the surrounding forest. It consists of a number of species which occur in both com- munities, including Acacia cyclops. Acacia saligna, Jacksonia furcellata, Banksia littoralis and Banksia attenuata. Banksia ilicifolia occurs in the ecotonal region on the north-eastern shore. In two small areas there are patches of Acacia cyclops sufhc’ently distinct to be designated ‘Acacia woodland’ (Figure 1). The surrounding forest has been described by Speck (1952) as tuart forest and jarrah-tuart-ecotone forest, and by Seddon (1972) as tuart-jarrah-marri tall open forest formation; its composition has been well documented by these authors. Much of the forest has been felled to the south for farming, and on the eastern shore is being cleared for housing developments. General. — The fringing vegetation is in general similar to Loch McNess (McComb and McComb. 1967), but there is less snecies diversity at Lake Joondalup, and fen vegetation is not extensive. Scirpus validus is a more prominent member of the sedge communities at Loch McNess, and this may be related to the seasonally more con- stant water level there. In contrast to the sedge communities, the benthic vegetation is more strongly developed at Lake Joondalup, in terms of both species number and plant density. As there is evidence that the level of the lake has increased in relatively recent times, one might expect the swamp and fen successions to be disturbed, and it is quite possible that fen vegetation may have been more extensive in the past. The Melaleuca woodlands are more inundated in winter, and the root systems of the trees may be adversely affected by in- creased waterlogging. Typha appears to be associated with shore disturbance, and further disturbance may Lad to an increase in this species. Temperature and light Water temperatures are closely correlated with mean monthly air temperatures (Figure 4) . Light penetration as measured with the Secchi disc is shown in Table 1 ; the disc was visible to the substratum during February to June, but in August to December, when the lake was rela- tively deep, the limit of Secchi disc transparency was 30 to 130 cm above the substratum. As the point at which the Secchi disc becomes invisible is a useful guide to the depth at which light penetration allows the survival of benthic plants, light may well be a limiting factor for the growth of benthic plants at depth in the lake in August to December. Low light intensities and lower temperatures thus combine to reduce productivity of the benthic plants in winter. As shown below, plankton densities and seston productivity are at their lowest in August to Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May. 1976. 17 December, so that it is depth rather than density of suspended matter which is the prime factor influencing light penetration. ■ I . ■ . I ■ i j ' • ■— JFM4MJJAS0N0 Figure 4. — The temperature of Lake Joondalup {broken line), plotted with atmospheric temperatures at the Upper Swan Research Station: (•, mean monthly maximum; o, mean monthly minimum). Chemical characteristics Oxygen. — The dissolved oxygen concentration varied seasonally between 6 and 9 mg 1 (Table 1). This range is relatively high, since the solubility of oxygen at 760 mm Hg is 8.1 mg/1 at 25° and 10.3 mg/1 at 13° (Hutchinson, 1957‘. (Oxygen becomes critical to aquatic life at 3 mg/ 1; Welch, 1952). These high levels are accounted for by the large surface area of the lake in relation to its volume, and exposure of the sur- face to agitation by wind and rain. The rich growth of benthic plants and presence of phyto- plankton indicates a contribution to the oxygen concentration through photosynthesis during daylight hours. The possibility of oxygen de- pletion at night was not investigated in the present study. Specific conductivity. — This ranged from 1068 to 3114 ^mhos.cm-^ (0.7 — 1.8% Total Dissolved Solids) (Figure 5). At Loch McNess a mean conductivity of 394 (^mhos-cm'^ was reported on one occasion (McComb and McComb, 1967). The concentrations of dissolved salts in Lake Joondalup are comparatively high for a fresh- water lake — for example Juday and Birge (1933) obtained a range of 9 to 124 (^mhos for more than 500 Wisconsin lakes, and Welch (1952) found lakes in Michigan to have a range of 10 to 330. Inland Australian lakes are characteris- tically much more saline (Bayly and Williams, 1973). The high conductivity is a reflection of high concentrations of individual ions (see be- low). The high conductivities of Lake Joondalup and Loch McNess are partially explained by their locations near aeolian limestone deposits. In addition, they are situated within 6 km of the Indian Ocean, and undoubtedly receive ions by wind and rain from that source. Evidence for this is discussed further below, but here we may note that the levels of chloride in the lake are relatively high. Chloride levels and con- ductivity show an expected trend in relation to change in lake volume, (cf. Figures 2 and 5), the increased water of the winter season diluting the dissolved ions. Figure 5. — Specific conductivity ( • ) and chloride concentrations (o) for Lake Joondalup. Acidity . — The lake is alkaline, and pH did not vary greatly with season, the range being 8.4 to 9.2 (Table 1). Constancy can be at least in part attributed to the buffering effect of car- bonate and bicarbonate ions, the levels of which are expected to be high in view of the amounts of calcium and magnesium present ( see below). An analysis of carbonate-bicarbonate carried out in December gave an alkalinity of 180 mg CaCOVl. Leaching of calcareous deposits is presumably the main factor determining the high pH. For example, a South Australian vol- canic lake in contact with limestone gave a pH of 8.2 (Bayly and Williams, 1964), while 19 eastern Australian lakes in silicious dunes had a pH range of 4.3 to 6.0 (Bayly, 1964). Loch McNess has a lower pH than Joondalup, a mean of 7.8 being reported (McComb and McComb, 1967), and this correlates with the lower con- ductivity of that lake. Metallic ions. — Like chloride and conductivity, the concentration of metallic ions decreases with increase in water level (Figure 6). The levels of these ions are compared with certain other lakes in Tables 2 and 3. The first of these tables gives absolute levels, and the second the equiva- lents of each ion expressed as a per- centage of the whole. It is clear that the levels of metallic cations at Joondalup are relatively high for a freshwater lake. The rela- tive proportions of cations resemble those found by Ba.yly (1964) for 19 coastal dune lakes in Queensland and New South Wales, and are also similar to those found in seawater, and in rain- water collected 16 km from the coast (Hutton and Leslie, 1958), except for the higher level of calcium. The proportions in Lake Joondalup Journal of the Royal Society of Western Australia. Vol. 59, Part 1, May, 1976. 18 chloride (mg/L) Table 1 Semonal rhangeH in light penetration, oxggen and pH at Lake Joondalup 3/ertWf are of 6 ftitea: standard errors are in brackets February March April ^ June ' August October December Secchi disc transparency (in) Dissolved oxygen (p.j).m.) I'H ; >0-90' ' 8 -85(0 -.59)=^ > 0 ■ 9()i 8 -60(0 -45)2 9 -22(0 -14) >0-801 ' : 0-12(0-(i0)2 8 -97(0 -21) ; >0-9;>i ' 8-91(0-14) 8-45(0-l]) ' 0-68(0-04) ; ; 8-82(0-51) 8-58(0-23) 0-61(0-02) 8-23(0-16) 8-54(0-24) 0-58(0-04) 8-15(0-62) 8-73(0-53) 1 maximum depth of water at time of •^arnpliiiK 2 determined with oxygen prolie; otiier oxygen data by titi-ation Table 2 Concentrations of ions found in various lakes ])a(a are in mg/l Lake Xa K Ca Mg n Reference Joondalup. W.A 140 to 513 8 to 20 44 to 57 21-5 to 43-5 212 to 655 This paper Coastal dune lakes. N.S.W. and (Queensland .. . 7 - 9 to 26 - 3 0-2 to 1-2 0-2 to 0-8 0 - 7 to 2-8 12-5 to «.3 Bayly 1964 Croispol Loch. Scotland 20 ■ 9 1 -6 27-1 22-2 Spence et al. 1971 German lakes 43-6 to 46 - 3 Hutchinson 1957 Tasmanian inland lakes 0-8 to 82-4 0-2 to 118-6 Williams 1964 Blue Lake, S.A 63 4 36 21 116 Bayly and Williams 1964 Cowan. W.A.* 74 276 387 689 10 435 138 638 Williams 1966 1 Lake Cowan is a salt lake, tlie others are freshwater lakes Table 3 The proportions of cations found in various waters Data are percentages of total equivalents of the ions Source of Water Xa K Ca Mg Reference Lake Joondalup. W.A. 56 - 5 to 77 -8 1 ■ 5 to 2 - 4 9-0 to 20-2 11-4 to 20-9 This ])aper Coastal dune lakes, X'.S.W. and Queensland 78 2 4 16 Bayly 1964 Seawater 77 2 3 18 Bayly 1964 Rainwater, Vic. 731 9 18 Hutton and Leslie 1958 Lake Cowan, W.A 78 0-2 0-8 21 M’illiams 1966 Sedimentary .source 5 8 53 34 Hutchin.son 1957 Loch Croispol. Scotland 22 1 32-7 44-3 Spence et al. 1971 Blue Lake, S.A 451 28 27 Bayly and Williams 1964 * includes sodium and potassium Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 19 are also similar to those found for Lake Cowan, a salt lake, despite the dramatically-different total ion concentrations present (Table 3 cf. Table 2). However, they contrast, (with the pos- sible exception of calcium), with mean values for sedimentary sources (Hutchinson, 1957); and Spence et al. (1971) report that a truly calcareous loch in Scotland, Croispol, had higher milliequivalent proportions of calcium and mag- nesium. At Blue Lake, which has somewhat in- termediate ratios, Bayly and Williams (1964) attribute the proportions to the combined influ- ence of leaching from limestone in the lake’s watershed, and influx of ions from the sea via precipitation. The data from Lake Joondalup, which is situated only 6 km from the ocean, are consistent with the interpretation that pre- cipitation has been the main determining factor in controlling ionic composition, the raised value for calcium being attributable to the leaching of this metal from limestone. Figure 7. — Ammonia nitrogen ( • ) and organic nitrogen (o) in Lake Joondalup. Figure 6. — Seasonal changes in concentrations of metallic cations at Lake Joondalup. Organic nitrogen and ammonia . — Both of these parameters show the same general seasonal trend of decreasing concentration with increas- ing water level (Figure 7), but superimposed on this trend there is an increase in the con- centration of organic nitrogen from June to August. For comparison, Lake Mendota, a eutrophic lake in Wisconsin, had an organic nitrogen concentration ranging from 0.3 to 0.6 mgN/1 (Peterson et al., 1925), values consid- erably lower than those obtained at Lake Joon- dalup, where the range was 1.5 to 3.00 mgN/1. One explanation for the high levels is the re- lease of organic nitrogen by the decomposition of organic material derived from the dense macrophytic vegetation and benthic hydro- phytes. Another possible source of nitrogen is seepage of nutrients from septic tanks. The concentration of ammonia-nitrogen was between 0.05 and 0.48 mg/1. Hutchinson (1957) gives a range of 0.05 to 0.54 mg/1 for the Madison lakes of Wisconsin, and attributes the upper values to derivation from sewage. Phosphorus . — The curves for phosphorus are reminiscent of those for nitrogen. Inorganic phosphorus showed the familiar trend of de- creasing concentration with increasing lake volume, whereas organic phosphorus increased _ when the levels of the lake were increasing S (Figure 8). This increase may be due to in- 1 creased assimilation of inorganic phosphorus by 2 phytoplankton (see below), increased decompo- ^ sition of organic material in the lake margins < and floor, leaching of phosphorus into the lake. I and to the release of phosphorus from the I sediments because of agitation by wind. Figure 8. — Inorganic phosphorus ( • ) and organic phosphorus (o) in Lake Joondalup. Inorganic phosphorus varied between 0.006 and 0.04 mgP/1 which compares with the range of 0.001 to 0.04 given by Pearsall (1930) for 9 lakes of the English Lake District. There phos- phate was the nutrient in relatively smallest quantity, and concentrations were lowest in the shallower lakes. At Joondalup the ratio of nitrogen to phosphorus varied between 70:1 and 140 : ], and the ratio would be higher if nitrate Journal of the Royal Society of Western Australia. Vol. 59, Part 1, May, 1976. 20 had also been determined. As the ratio in plants is about 15:1 to 2:1 (e.g. Gerlaczynska, 1973), one may speculate that of nitrogen and phos- phorus, the lev:l of phosphorus is more likely to be limiting for plant growth. doubt part of this organic material is derived from algal decomposition and bacterial action. Seston productivity (Figure 10) is more closely correlated with organic N than with total plank- ton, again suggesting relatively high levels of organic material in the lake. Phytoplankton Total numbers of phytoplankton are shown in Figure 9, where it can be seen that there was an algal bloom, some 4 x 10^ organisms per 1, in the period April to June. The highest den- Figure 9. — The planktonic algae of Lake Joondalup. The number of individuals of all species is expressed per litre. sity of the bloom was in mid April, coinciding with minimum lake level (Figure 2) and highest concentrations of metallic cations (Figure 6). Temperature was falling (Figure 4). Phosphate phosphorus (Figure 8) peaked a month before th? maximum algal bloom, and an increase in the organic P fractions is associated with the bloom. Not surprisingly, organic N is also rela- tively high (Figure 7). Both organic P and N remain relatively high after the bloom, and no 1 I I I I I ' ■ ■ t- — 1 - ■ . J FMAMJ JASONO Figure 10. — Seston productivity of Lake Joondalup. The distribution of particular species of algae is shown in Figure 11, and it is immediately apparent that the main bloom is due to the presence of Dispora crucigenoides, a green alga. In December there was a minor bloom of a blue- green, Raphidiopsis, but distribution in the lake was very patchy as indicated by the large stan- dard error on the December figure in Figure 9. Occurrence of the blue-green Anabaena spiroides is of interest, as this is known to fix atmospheric nitrogen (Cameron and Fuller, 1960). It appears in the lake in significant num- bers in August, at a time when ammonia levels are low and organic N high, and this suggests that levels of available nitrogen may have be- come limiting to the growth of algae at tha-t time. Anabaena and Anacystis (which is also present in the lake but in low numbers) can produce blooms toxic to animals (e.g. Jackson, 1964). There have been no reports, to our know- ledge, of the death of fish or waterfowl in the lake, as there have been in summer for certain other Perth lakes. There is, however, a possibility of such toxic blooms occurring in Joondalup if further enrichment occurs. General Lake volume is the major factor determining seasonal changes in concentration in dissolved substances. The data are consistent with the view that phosphorus levels may limit algal pro- ductivity in the lake, except for a short period of the year when there is a small increase in Anabaena. Suitable culture experiments could be carried out to examine these possibilities. The observations indicate a need for further moni- toring as urbanisation around the lake pro- pesses, and a need for comparable data on the trophic status of other lakes, including those m natipal parks not affected by urbanisation. The relatively high levels of phosphorus and nitrogen in the water, and the occurrence of algp blooms, justify the provisional designation 01 the lake as ‘mildly eutrophic’. References Anon. Bayly, Bayly, Bayly, (1955).— “Standard Methods for the Examination of Water, Sewerage and Industrial Wastes.” Tenth edition. (American Public Health Association. New York). I. A. E. (1964).— Chemical and biological studies on some acidic lakes of East Australian sandy coastal lowlands. Aust. J. mar Freshw Res. 15: 56-72. I. I. and biological observations on some volcanic lakes m the south-east of South Australia Aust. J. mar. Freshw. Res. 15: 123-32. A. E., and Williams, W. D. Waters and their Ecology.'' tralia. ( 1973). — “Inland Longman, Aus- ^..ameron, K. E., and Fuller. W. H. (I960).— Nitrogen ppion hy some algae of Arizona soils. Proc. Soil Sci. Soc. Amer. 24: 353-6 Journal of the Royal Society of Western Australia, Vol. 59. Part 1 . May, 1976 21 M M O N Figure 11. -Planktonic algae of Lake Joondalup showing the expressed per litre. Note that the scale for Dispora c-ucigenctdes is reduced as comparea Other algae, which is the same in each case. Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 22 Evans, G., and Sherlock, N. (1950). — Butler’s Swamp, Claremont. West. Aust. Nat. Z: 152-60. Gerlaczynska, B. (1973). — Distribution and biomass of macrophytes in lake Dgal Maly. Ekologia Polska 21: 743-752. Hutchinson. G. E. (1957 and 1967). — “A Treatise on Limnology Two volumes. John Wiley and Sons, Inc., New York. Hutton. J. T., and Leslie, T. I. (1958). — Accession of non-nitrogenous ions dissolved in rain- water to soils in Victoria. Aust. J. Agric. Res. 9: 492-507. Jackson, D. J. (ed.) (1964). — “Algae and Man.” (Plenum Press, New York.) Juday, C., and Birge, E. A. (1938). — The transparency, the color and the specific conductance of the lake waters of North-eastern Wis- consin. Trans. Wis. Acad. Sci., Arts, Let. 28: 205-59. McArthur, W. M., and Bettenay, E. (1960). — “The development and distribution of the soils of the Swan Coastal Plain, Western Aus- tralia.” C.S.I.R.O. Australian Soils Publica- tion No. 16. McComb, J. A., and McComb, A. J. (1967). — A prelimin- ary account of the vegetation of Loch McNess, a swamp and fen formation in Western Australia. J. Roy. Soc. W.A. 50: 105-12. Parker, C. R. (1972), — “Water Analysis by Atomic Absorption Spectroscopy.” Varian Techtron Pty. Ltd., Springvale, Victoria. Pearsall, W, H. (1930). — Phytoplankton in the English Lakes. I. The proportions in the waters of some dissolved substances of biological im- portance. J. Ecol. 20: 306-20. Peterson, W. H., Fred, E. B., and Domogalla, B. P. (1925). — The occurrence of amino acids and other organic nitrogen compounds in lake water. J. biol. Chem. 23: 287-95. Riggert, T. L. (1966). — “A Study of the Wetlands of the Swan Coastal Plain.” Dept, of Fisheries and Fauna of Western Australia. Robinson, R, W. (1941). — Perchloric acid oxidation of organic phosphorus in lake waters. Ind. Eng. Chem. (Anal, Ed.) 13: 465-6, Seddon, G. (1972). — “A Sense of Place.” University of Western Australia Press, Western Australia. Speck, N. H. (1952). — “The Ecology of the Metropolitan Sector of the Swan Coastal Plain.” Thesis presented to the University of Western Aus- tralia for the Degree of Master of Science. Spence, D. H. N., Campbell, R. M., and Chrystal, J. (1971). — Spectral intensity in some Scottish freshwater lochs. Freshw. Biol. 1: 321-37. Welch, P. S. (1948). — “Limnological Methods.” McGraw- Hill Book Company, New York. Welch, P. S. (1952). — “Limnology.” Second edition. McGraw-Hill Book Company, New York. Williams, W. D. (1966). — Conductivity and the concen- tration of total dissolved solids in Australian lakes. Aust. J. mar. Freshwat. Res. 17: 169-76. Journal of the Royal Society of Western Australia, Vol. 59. Part 1. May, 1976. 23 4. — Descriptions of three new fishes from Western Australia by Gerald R. Ailen^ Manuscript received 18 November 1975; accepted 15 December 1975 Abstract Three new species of marine fishes are de- scribed from coastal waters of Western Australia. Two of the species, Ellerkeldia rubra and Anthias georgei are serranids and the third, Parapercis biordinis, belongs to the Mugiloididae. The three species are illustrated and keys to the Ellerkeldia and Parapercis from Western Australia are pro- vided. Introduction The most recent comprehznsive listing of Western Australian fishes is that of Whitley (1948). He included approximately 740 species, but this work desperately needs to be updated. In recent years several authors including Mees (1959; 1960a and b; 1961; 1962; 1963; 1964; 1966); Scott (1959); and McKay (1963; 1964; 1966; 1967; 1969; 1970; 1971) have added about 200 additional records. Collections by J. B. Hutchins and the author from 1972 until the present time have contributed nearly 400 more. However, the latter collections remain largely unreported. These will be Included in an an- notated checklist of the fishes of Western Aus- tralia currently in preparation by the author. The present paper includes descriptions of three new species. The majority of specimens involved were located amongst large holdings of un- sorted material during a re-organisation of the Western Australian Museum fish collection in 1975. In addition, three specimens, including the holotype of Ellerkeldia rubra were collected by the author at the Abrolhos Islands. Measurements were made with dial calipers to the nearest 0.1 millimetre (mm). Standard length is abbreviated as SL. The counts and proportions which appear in parentheses under the description section for each species apply to the raratypes when differing from the holo- type except for the single specimen of Para- percis biordinis, in which case the actual milli- metric measurement is given. Type specimens have been deposited at th-^ Australian Museum. Sydney (AM); United States National Museum of Natural History, Washington, D.C. (USNM) ; and the Western Australian Museum, Perth (WAM). The author is grateful to Mr. Jeremy N. Green, Cuiator of Marine Archaeology, for providing ^ Western Australian Museum. Francis Street, Perth, 6000. accommodation during the visit to Beacon Is- land (Abrolhos) in May 1975. Thanks are also due Mr. Pat Baker, who assisted with collections and Mrs. C. Allen who prepared the typescript. Family Serranidae Ellerkeldia rubra n.sp. (Fig. 1; Table 1) Holotype. — WAM P25314-003, 71.0 mm SL, col- lected with multiprong spear off Beacon Island, Wallabi Group, Abrolhos Islands, Western Aus- tralia in 3-4 metres by G. R. Allen on 20 May 1975. Paratypes.—KM 1.18476-001, 59.7 mm SL, col- lected at Abrolhos Islands, Western Australia by A. Robinson, no other collecting data; USNM 214701, 67.5 mm SL, collected with bottom trawl approximately 40 nautical miles west of Bernier Island, Western Australia (25°59’S, 112“27’E) in 71 fathoms by R. George and crew of “Dlaman- tina” on 8 October 1963; WAM P25226-001, 72.6 mm SL, collected with bottom trawl off Cape Inscription, Western Australia in 40 fathoms by Poole brothers aboard “Bluefin” on 9 October 1967; WAM P25311-007, 2 specimens, 29.2 and 49.6 mm SL, collected with rotenone in Goss Passage off Beacon Island, Abrolhos Islands, Western Australia in 30 metres by G. R. Allen on 18 May 1975. Diagnosis. — A species of Ellerkeldia with the following combination of characters: soft dorsal rays 19; soft anal rays 8; tubed lateral-line scales 41-45; colour mostly pale with broad brown stripe on sides from snout to caudal base and series of brown spots and blotches on snout, interorbital, and nape. Description. — Dorsal rays X,19; anal rays III, 8; pectoral rays 15 (16); gill rakers on lower portion of first gill arch 8 + 4 to 5 rudiments: tubed lateral-line scales 44 (41 [11, 42 [2], 43 [21); horizontal scale rows from lateral-line to base of middle dorsal spines 2-3; from lateral- line to anus 17. Body ovate and compressed, the greatest depth 2.7 (2.5 to 2.8), head length 2.4 (2.2 to 2.4), both in standard length. Snout 3.9 (4.2 to 5.0), eye diameter 4.1 (3.8 to 4.0), int~rorbital width 8.4 (8.9 to 11.2), length of maxillary 2.4 (2.1 to 2.2), least depth of caudal peduncle 3.4 (3.2 to 3.8) ! length of caudal peduncle 4.1 (3.8 to 5.2), of Journal of the Royal Society of Western Australia, Vol. 59. Part 1. May, 1976. 24 Figure l.~Ellerkeldia rubra, holotype, 71.0 mm SL, Abrolhos Islands, Western Australia. Table 1 Morphometric Proportions- of Type Specimens of Ellerkeldia Rubra {In thonaandUiH of the sitandard length) Characters ' Holotvpe i Paratypes ; WAM P253U-003 WA51 P25226-001 USNM 214701 AM 1.18476-001 WAM P25311-007 WAM P2531 1-007 Standard len<;th (mm) 71-0 72-6 67-5 59-7 49-6 383 29-2 366 Greatest l)ody depth 369 355 394 360 Head length 415 402 437 412 454 435 Snout length 106 96 89 94 91 103 Eye diameter 101 99 114 104 115 110 Interorbital width 49 41 44 47 40 Length of maxillary 194 168 200 198 ia*? 199 liCast depth of caudal peduncle 123 117 119 127 I'^l 137 Length of caudal peduncle 101 99 84 107 101 Snout to origin of dorsal fin 397 406 406 399 395 387 Snout to origin of anal fin 687 694 726 717 7 ‘^6 Snout to origin of pelvic fin 411 415 433 404 440 394 548 205 •} 1 o Length of dorsal fin base 546 527 527 534 Length of anal fin base 183 178 178 183 173 Length of pectoral fin 293 307 324 305 341 Length of pelvic fin 197 204 233 •^•>6 240 55 137 106 188 89 171 154 Length of 1st dorsal spine 41 55 59 40 54 Length of 4th dorsal spine 120 152 156 134 151 Length of last dorsal spine 97 102 95 101 101 Tallest dorsal ray Length of 1st anal spine 169 99 179 84 190 96 171 69 171 87 Length of 2nd anal spine 152 172 173 173 177 Length of 3rd anal spine 146 139 159 132 141 Longest anal ray 197 204 212 204 •>06 Length of caudal fin 207 241 246 224 242 267 Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May, 1976. 25 pectoral fin 1.4 (1.3 to 1.4), of pelvic fin 2.1 (1.8 to 1.9), of first dorsal spine 10.1 (7.4 to 10 3) of fourth dorsal spine 3.5 (2.8 to 3.2), of last dorsal spine 4.3 (3.9 to 4.6), of longest soft dorsal ray 2.4 (2.2 to 2.6), of first anal spine 4.2 (4.5 to 6.0), of second anal spine 2.7 (2.3 to 2 7) of third anal spine 2.8 (2.8 to 3.2), of longest soft anal ray 2.1 (2.0 to 2.2), of caudal fin 2.0 (1.6 to 1.9), all in the head hngth. Pair of nasal openings on each side of snout, anterior opening with dermal flap; mouth large, oblique, with lower j aw protruding slightly ; supramaxillary present; lateral-line gradually ascending to within 2-3 scales below middle dorsal spines, then gradually descending to middle of side of caudal peduncle, proceeding to base of caudal fin; snout tip, lips, chin, most of dentary, isthmus, and maxillary naked; re- mainder of head and body with finely ctenoid scales; sheath scales covering about basal -J-f of dorsal, anal, caudal, and pectoral fins; pre- orbital entire; preopercle with 3 antrose spines on lower border and about 25 small serrae on posterior edge; subopercle with a few small serrae on postero-ventral border; opercle with 3 spines. Upper and lower jaws with bands of depres- sible villiform teeth, narrowing in width pos- teriorly; in addition, two widely spaced tusk-like canines at front of lower jaw and 1-2 similar teeth about midway back on each dentary, upper jaw with pair of enlarged tusks anteriorly and several enlarged retrcse canines on either side of median diastema; vomer with patch of small conical teeth; palatin:s with small biserial coni- cal teeth; length of maxillary 2.1 (2.1 to 2.4) in the head length. Colour of holotype in ethyl alcohol; head and body generally pale (yellowish); faint brown stripe (about 3-4 scales wide) extending from snout tio (also on upper lip), interrupted by eye, then continuing along upper sides to base of caudal fin, pigmentation most intense just be- hind eye and on caudal peduncle; a pair of brown spots on anterior portion of upper and lower lips, these continuous with brown streak extending across side of snout to front corner of eye; diffuse broken brown stripe on mid- dorsal line from interorbital to occipital region, followed by a pair of short brown streaks, one on each side of mid-dorsal line, then an isolated brown spot on middle of nape and another at base of first dorsal spine; similar brown spot behind upper, posterior corner of each eye; small brown spot on anteriormost extension of isth- mus; fins uniformly pale except basal part of first dorsal spine brown. The paratypes exhibit the same basic pattern except on the 67.5 mm specimen the mid-dorsal stripe is not apparent except as an isolated streak on the post-interorbital. In addition, the broad stripe behind the eye has the appearance of two, large isolated blotches, one immediately above the preopercle opening and the other just above the opercle opening. The 59.7 mm para- type is much darker than the other specimens and the markings on the head are mostly ob- scured except the spots on the isthmus and lower ho. Colour in life . — Head and body generally pinkish-red grading to darker red dorsally; lower portion of head, breast, and abdomen white; prominent red stripe from snout to eye, continued behind eye to upper portion of opercle and two short oblique bands of similar colour and width immediately behind, near upper cor- ner of gill opening; diffuse, dusky brown band running longitudinally on upper sides from upper corner of gill opening to caudal base, more in- tense posteriorly, forming more or less isolated dark brown spot on caudal peduncle; fins pale pink. Remarks . genus Ellerkeldia is a small group of serranid fishes consisting of 5 species confined to the southern Australia-New Zealand region. E. ruhra is separable from the other members of the genus on the basis of the charac- ters given in the following key. Key to the species of Ellerkeldia la. Body with series of 6-7 dark transverse bands (may be faint in preservative) lb. Body without series of 6-7 dark transverse bands, either with horizontal stripe or mottled with irregular broken bands 2a. Soft dorsal rays 17-18; soft anal rays 7; dark bands distinct on both upper and lower half of sides (southern Queensland; New South Wales) 2b. Soft dorsal rays 20 to 21, soft anal rays 8; dark bands, except those on caudal peduncle, either absent or indistinct on lower portion of sides 3a. Greatest body depth 2.8 to 2.9 in standard length; dark bands usually well defined, confined mainly to upper sides (Lord Howe Island; New Zealand) 3b. Greatest body depth 2.4 to 2.5 in standard length; dark bands usually not well defined (at least in preservative), extending well below middle of sides (New South Wales; Western Australia) 4a. Colour mostly pale with faint to prominent longitudinal band on upper sides from snout to base of caudal; soft dorsal rays 19 (Western Australia) 4b. Colour not as in 3a, dark olive- brown above and grey below, mottled with irregular broken transverse bands; soft dorsal rays 20 (southern Queensland; nor- thern New South Wales) In the Abrolhos Islands this species was cccasionally encountered at the bottom of Goss Passage at depths between 25 and 35 metres. It was usually seen resting on the bottom at the entrance of small crevices of the coral reef or near the base of rocky outcrops or sponges. The holotype was the only individual observed outside of the deeper waters of Goss Passage. It was collected from a large patch of staghorn coral (Acropora) in only 3-4 metres depth. This species is named rubra with reference to the red coloration in life. 4 annulata (Gunther) 3 huntii (Hector) maccullochi Whitley rubra n. sp. jamesoni (Ogilby) Journal of the Royal Society of Western 26 Australia. Vol. 59, Part 1, May, 1976. Figure 2 . — Anthias georgei, holotype, 79.5 m:n SL, off Bernier Island, Western Australia. Table 2 Morpltometi'ic Proportions of Tupe Specimens of Anthias georgei {In thodsamlt'is of the standard length) Characters Holotvpe waNi P25205-001 Paratypes WAM 1 WAM l»25205-002 P25205-002 Standard length (mm) 79-5 32-8 29-0 Greatest t>odv deptli 392 390 366 Head length 345 375 331 Snout length . . 69 55 59 Eve diameter 108 125 141 Interorhital width 84 82 75 Length of maxillarv . .. .... .... . . 147 159 159 Least dc])tii of caudal peduncle 143 143 159 Length of caudal peduncle 147 116 131 Snout to origin of dorsal fin Snout to origin of anal fin 350 660 372 680 348 659 Snout to origin of pelvic fin 355 360 355 Length of dorsal fin base 629 595 572 Length of anal fin base 165 174 166 Length of pectoral fin 317 305 297 Length of pelvic fin .... 337 311 300 l.ength of 1st dorsal spine 57 55 66 ]>ength of 3rd dorsal spine ... 167 122 131 Length of last dorsal spine 107 107 117 Tallest dorsal rav 197 183 159 Length of 1st anal spine 57 73 69 Length of 2nd anal spine 136 177 148 Length of 3rd anal spine 131 137 128 Longest soft anal ray 192 220 245 Length of caudal fin 574 • 545 * damaged Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 27 Anthias georgei n.sp. (Fig. 2; Table 2) Holotyve.—VJ AM. P25205-001, 79.5 mm SL. collected with bottom trawl approximately 40 nautical miles west of Bernier Island, Western Australia (24°59’S, 112°27’E) in 71 fathoms by R. George and crew of “Diamantina” on 8 October 1963. Paratyves.—VJ AM P25205-002, 2 specimens, 29.0 and 32.8 mm SL, collected with holotype. Diagnosis . — A species of Anthias with the fol- lowing combination of characters: two and a half to three rows of scales between middle of spinous dorsal fin and lateral-line; tubed lateral- line scales 39-40; gill rakers on first arch 10 or 11 + 22 to 26; third dorsal spine slightly elon- gate; caudal fin deeply forked with prolonged rays (at least in adult males); pectoral rays 18; body d^nth 2.5 to 2.7 in standard length; colour generally pale in preservative, probably reddish in life. Description. — The proportional measurements of the holotype and paratypes are expressed as percentage of the SL in Table 2. Dorsal rays X.16; anal rays 111,7; pectoral rays 18: gill rakers on first arch 10 + 23 (11 -(- 25 or 26); tubed lateral-line scales 39 (39-40); horizontal scale rows from lateral-line to base of middle dorsal spin:s 2-3; from lateral-line to anus 16-17. Body ovate and laterally compressed, the greatest depth 2.5 (2.6 to 2.7), head length 2.9 (2.7 to 3.0), both in standard length. Snout 5.0 (5.6 to 6.8). eye diameter 3.2 (2.3 to 3.0). inter- orbital width 4.1 (4.4 to 4.6), length of maxillary 2.3 (2.1 to 2.4), least depth of caudal peduncle 2.4 (2.1 to 2.6), length of caudal peduncle 2.3 '2.5 to 3.2), of pectoral fin 1.1 d.l to 1.2), of pelvic fin 1.0 (1.1 to 1.2), of first dorsal spine 6.1 (5.1 to 6.8), of third dorsal spine 2.1 (2.5 to 3.1), of last dorsal spin? 3.2 (2.8 to 3.5), of longest soft dorsal ray 1.7 (2.1), of first anal spine 6.1 (4.8 to 5.1), of second anal spine 2.5 '2.1 to 2.2), of third anal spine 2.6 (2.6 to 2.7). of longest soft anal ray 1.8 (1.4 to 1.7), of caudal fin 0.6 (0.6 to 1.1), all in the head length. Pair of nasal openings on each side of snout, anterior opening with dermal flap on posterior edge; mouth large, oblique, with lower jaw pro- truding slightly; lateral-line gradually ascend- ing to within 2-3 scales below middle dorsal spines, then gradually descending to middle of side of caudal peduncle, proceeding to base of caudal fin; area around nostrils, lips, chin, and isthmus naked: remainder of head (including maxillary) and body with finely ctenoid scales: sheath scales covering about basal 4--§ of soft dorsal, anal, caudal, and pectoral fins; rear edge of preopercle serrate, several serrae present on exposed edge of subopercle and interopercle; opercle with three spines, the uppermost blunt and inconspicuous. Upper jaw with inner band of small depres- sible canines and outer row of larger, fixed canines: pair of large tusks on each side at front corner of jaw and a pair of equally large inner teeth on either side of median diastema. An- terior portion of lower jaw with dense patch of small villiform teeth and 1-2 large, laterally flared tusks on each side at front corner; 1-2 large tusks near middle of each dentary with row of small depressible canines on posterior part of jaw. Colour of holotype in ethyl alcohol; head and body generally pale (yellowish-tan) , slightly reddish on upper portion of head and back; faint suggestion of three pale stripes on sides, each about one scale wide, first just below lateral- line, second at level of upper corner of opercle. third at level of lowermost pectoral rays; fins uniformly pal?. The two paratypes, which are juveniles, are uniformly pale with several rows of small brown spots below the spinous dorsal fin. Remarks. — A. georgei appears to be allied to A. conspicuus Heemstra and A. townsendi Boulenger from the western Indian Ocean and Arabian Sea. These species were reviewed by Heem.stra (1973) and are characterised by 2-3 scales between the lateral-line and the base of the middle dorsal spines, 37-41 tubed lateral- line scales, and 15-17 soft dorsal rays. A. georgei differs from A. conspicuus by having the third dorsal spine slightly elongate rather than sub- equal, and the caudal fin deeply forked with prolonged filaments instead of lunate. In addi- tion, there is a significant colour difference. The holotype of A. georgei possesses elongate whitish gonads and is therefore presumed to be a male. The males of A. conspicuus have two wide, dark stripes on the sidss. which join on the caudal fin in contrast to the complete absence of distinguishing marks on A. georgei. A. town- sendi differs by being less deep bodied (2. 9-3.2 vs. 2. 5-2. 7 in SL). and by having a rounded rather than acute anal fin tip, and a lunate caudal. Named in honour of Dr. Ray W. George, Curator of Crustacea at the Western Australian Museum and collector of the only known speci- mens. Family Mugiloididae Parapercis biordinis n.sp. (Fig. 3) Holotype.~WAM P25206-001, 70.0 mm SL, collected with beam trawl southwest of Point Cloates, Western Australia (22°59'S, 113°25’E) in 71 fathoms by C.S.I.R.O. research ship on 31 January 1964. Diagnosis. — A species of Parapercis with the following combination of characters: palatine teeth absent: 6 or 7 teeth in outer row of lower jaw: last dorsal spine connected by membrane to first dorsal ray opposite tip of last dorsal spine: five dorsal spines: about 24 zigzag rows of scales around caudal peduncle; lateral-line scales 54; colour largely pale with 4-5 pairs of brown spots along sides and four spots on caudal fin. Description . — Dorsal rays V,21; anal rays I, 18; pectoral rays 18; pelvic rays 1,5; branched caudal rays 15; gill rakers on first arch 5 -f 12 Journal of the Royal Society of Western Australia, Vol. 59, Part 1, May. 1976. 28 Figure 3 . — Parapercis biordinis, holotype, 70.0 mm SL, off Point Cloates, Western Australia (scales not shown). The pattern of dark spots on the dorsal surface of the head is shown in the drawing on the right. 17; lateral-line scales from upper edge of gill opening to base of caudal fin 54; horizontal scale rows from lateral-line to base of first dorsal spine 3; from lateral-line to anus about 8 fmost are missing on the holotype) ; predorsal scales 7-8. The following measurements (in mm) were recorded for the holotype (only known speci- men) : length of head 20.0 (3.5 in SL); length of snout 5.0 (4.0 in HL) ; diameter of eye 7.5 (2.7 in HL) ; postorbital length of head 8.5 (2.4 in HL) ; interorbital width 2.0 (10.0 in HL) ; snout tip to rear edge of maxillary 7.7 (2.6 in HL); least depth of caudal peduncle 5.7 (3.5 in HL); greatest depth of body 9.5 (7.4 in SL) ; length of fourth dorsal spine 4.9 (4.1 in HL) ; longest pectoral ray 14.2 (1.4 in HL) ; longest pelvic ray 16.4 (1.2 in HL) ; longest caudal ray 13.8 (1.4 in HL); length of dorsal fin base 42.3 <1.7, in SL); length of anal fin base 29.5 (2.4 in SL). Teeth absent on palatines, vomer with 8 rela- tively large conical teeth arranged in a single row; lower jaw with outer row of 6 or possibly 7 (count includes 1 and possibly 2 teeth which are missing) hooked canine teeth on anteiibr- most portion, inside these a band of villiform teeth and a single row of smaller canines on the side of each dentary; upper jaw with about 56 canines in outer row and inside these a dense band of villiform teeth. Scales of body mostly ctenoid, those of pre- opercle region relatively small, embedded, and cycloid; occipital, interorbital, and snout naked. Colour of holotype in ethyl alcohol: head and body mostly yellowish-tan; scales on upper part of body with dusky edges; a series of diffuse brown spots, about pupil size, on sides; the first at level of middle dorsal spines, just below lateral-line, the remainder occurring in 4 pairs at equal intervals below soft dorsal fin (except for first pair, which both lie below lateral-line, the members of each pair are separated by the lateral-line) ; caudal fin with 2 pairs of similar spots, one at base of fin and the other near the centre; pectoral base with faint brown streak; series of dark markings (see Fig. 3) on nape and upper part of opercle; faint brown streak below lower anterior corner of eye connecting suborbital and maxillary; small brown spots on soft dorsal fin as shown in Fig. 3. Remarks. — Cantwell (1964) revised the genus Parapercis and Schultz (1968) published a sup- plemental paper with a key to the Indo-Pacific members of this group and descriptions of four new species. On the basis of coloration and the combination of other characters listed in the diagnosis, P. hiordinis is distinct from the 32 species treated by these authors. They recorded four species from Western Australia: P. allporti (Gunther); P. e7neryana (Richardson); P. haackei (Steindachner ) ; and P. nehulosa (Quoy and Gaimard). In addition, the WAM recently received a specimen (WAM P25342-002) of P. ramsayi, 146 mm SL, which was trawled off Cape Naturaliste, and two specimens (WAM P24582 and P25367-006) of P. cephalopunctata, 56-125 mm SL, collected at North West Cape and the Dampier Archipelago. These represent new records for Western Australia; the latter species, which ranges widely in the Indo-W. Pacific, is also new for Australia. P. ramsayi was pre- viously recorded from New South Wales and South Australia. The species presently known from Western Australia are distinguished in the following key. Key to the species of Parapercis from Western Australia la. Palatine teeth present 2 lb. Palatine teeth absent 3 2a. Dorsal rays V,22; 10 teeth in outer row of lower jaw; last dor- sal fin spine connected by mem- brane to base of first soft dorsal ray; lower portion of sides with- out row of seven spots haackii (Steindachner) 2b. Dorsal spines IV, 24; 8 teeth in outer row of lower jaw; last dorsal fin spine connected by membrane to first soft dorsal ray at about level of tip of last dorsal spine; lower portion of sides with row of seven large spots ramsayi Steindachner 3a. Last dorsal fin spine connected by membrane to base of first soft dorsal ray; horizontal scale rows between lateral-line and first dorsal spine 8-10; soft dorsal rays 22 ; zigzag row of scales around caudal peduncle 35-44 4 3b. Last dorsal fin spine connected by membrane to first soft dorsal ray at about level of tip of last dorsal spine; horizontal scale rows between lateral-line and first dorsal spine 3-5; soft dorsal rays 21; zizag row of scales around caudal peduncle 24-29 5 Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 29 4a. Total gill rakers 17-20; oblique scale rows above lateral-line 70- 77: 3 dark stripes across inter- orbital space; 5 V-shaped dark bars below dorsal fins, but no broad, light, lengthwise streak with dark edge along middle of side, interrupting the dark bars .. emeryana (Richardson) 4b. Total gill rakers 11-17 (usually 13- 14); oblique scale rows above lateral-line 77-87; no dark stripe across interorbital space; dark bars somewhat V-shaped below dorsal fins, interrupted along middle of side by dark-edged, broad, light streak nebulosa (Quoy and Gaimard) 5a. Lower portion of sides with a single row of 10 dark spots, about pupil size, extending horizontally from lower pectoral base to lower caudal peduncle; a dark spot or ocellus on each side of head just above opercle on larger specimens (over about 130 mm SL); vom- erine teeth in 2 rows cephalopunctata (Seale) 5b. Colour pattern not as in 5a; vomerine teeth in 1 or 3 rows . .. 6 6a. Eight vomerine teeth in 3 hori- zontal rows (2 + 4 -f 2); sides with seven dark bars, the an- terior four extending from dorsal fin base to below lateral-line, the posterior three above the lateral- line; no markings on caudal and dorsals fins .... .... .... .... allpovti (Gunther) 6b. Eight vomerine teeth in a single row; sides with 4-5 pairs of spots; 2 pair of similar spots on caudal and several smal spots on basal part of soft dorsal fin .... .... biordinis n.sp. This species is named biordinis (Latin; “double-row”) referring to the double row of spots along the sides. References Cantwell, G. P. (1964). — A revision of the genus Para- percis, family Mugiloididae. Pacific Science 18 (3): 239-280. Heemstra, P. C. (1973. — Antfiias conspicuus sp. nova (Perciformes: Serranidae) from the Indian Ocean, with comments on related species. Copeia, 1973 (2): 200-210. McKay, R. J. (1963). — Second record of the little pine- apple fish ( SoTosichthys ananassa Whitley). West. Austr. Naturalist 8 (7): 171-172. (1964). — Description of a new stonefish of the family Synanceidae from Western Aus- tralia. J. Roy. Soc. W. Austr. 47 (1): 8-12. (1966). — Studies on Western Australian sharks and rays of the families Scyliorhini- dae, Urolophidae, and Torpedinidae. J. Roy. Soc. W. Austr. 49 (3): 65-82. (1967). — Additions to the fish fauna of Wes- tern Australia. West. Austr. Naturalist 10 (4) ; 92-95. (1969). — The genus Tandya in Western Aus- stralia, with a description of a new opistho- gnathid fish, Tandya reticulata sp. nov. J. Roy. Soc. W. Austr., 52 (1): 1-2. (1970). — Additions to the fish fauna of Wes- tern Australia — 5. W. Austr. Fish. Bull., 9 (5) ; 3-24. U971). — Two new genera and five new species of percophidid fishes (Pisces: Per- cophididae) from Western Australia. J. Roy. Soc. W. Austr., 54 (2): 40-46. Mees, G. F. (1959). — Additions to the fish fauna of Wes- tern Australia—!. W. Austr. Fish. Bull., 9 (1) : 5-11. (1960a). — Additions to the fish fauna of Western Australia— 2. W. Austr. Fish. Bull., 9 (2) 13-21. (1960b). — The Uranoscopidae of Western Australia (Pisces, Perciformes). J. Roy. Soc. W. Austr.. 43: 46-58. (1961).— Description of a new fish of the family Galaxiidae from Western Australia. J. Roy. Soc. W. Austr., 44: 33-38. (1962). — Additions to the fish fauna of Western Australia— 3. W. Austr. Fish. Bull, 9 (3) : 23-30. (1963). — The Callionymidae of Western Australia (Pisces). J. Roy. Soc. W. Austr., 46 (3) : 93-99. (1964).— Additions to the fish fauna of Western Australia— 4. W. Austr. Fish. Bull.. 9 (4) : 31-55. (1966). — A new fish of the family Apogonidae from tropical Western Australia. J. Roy. Soc W. Austr., 49 (3) : 1966. Schultz, L. P. (1968). — Four new fishes of the genus Parapercis with notes on other species from the Indo-Pacific area (Family Mugiloididae). Proc. U. S. Nat. Mus., 124 (3636): 1-16. Scott, T. D. 1959.— Notes on Western Australian fishes. no. 1. Trans. Roy. Soc. S. Austr., 82: 73-91. Whitley, G. P. (1948). — A list of the fishes of Western Australia. W. Austr. Fish. Bull. 2: 1-35. Journal of the Royal Society of Western Australia, Vol. 59, Part 1. May, 1976. 30 CORRECTION Geometric microliths from a dated archae- logical deposit near Northcliffe, Western Aus- tralia. By C. E. Dortch, Volume 58, Part 2, Paper 5. The last sentence of the abstract should read: “Analysis of pollen samples taken from the deposit show that Eucalyptus diversicolor, E. calophylla, and E. marginata existed in the locality prior to about 6780 years BP and that all three species were present at times since.” INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor, Royal Society of Western Australia, Western Australian Museum, Perth, Papers are received only from or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All Tables, and captions for Figures, should be on separate sheets. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, includ- ing the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. To avoid unnecessary handling of the original illus- trations, which are usually best prepared between and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. It is the responsibility of authors to adhere to the International Rules of Botanical and Zoological Nomenclature. Palaeontological papers must follow the appropriate rules for zoology or botany, and all new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Com- mittee of the Geological Society of Australia. Thirty reprints are supplied to authors free of charge, up to a maximum of 60 for any one paper. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 59 1976 Part 1 Contents 1. Topical chemistry in Perth’s air. By G. A. Bottomley. 2. The changing ecology of Western Australian wheat production. By W. J. R. Boyd, W. K. Waterhouse, B. B. Singh, N. A. Campbell and N. A. Goodchild. 3. The nutrients and plants of Lake Joondalup, a mildly eutrophic lake experiencing large seasonal changes in volume. By R. A. Congdon and A. J. McComb. 4. Descriptions of three new fishes from Western Australia, By Gerald R. Allen. Correction: Paper by C. E. Dortch, Volume 58, Part 2. Editor: A. E. Cockbain The Royal Society of Western Australia, Western Australian Museum, Perth n3521 /I /76— 625 WILLIAM C. BROWN, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 59 PART 2 SEPTEMBER, 1976 PRICE TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Chief Marshall Sir Wallace Kyle, G.C.B., C.B.E., D.S.O., D.F.C., K.St.J., Governor of Western Australia COUNCIL 1975-1976 President Vice Presidents Past President Joint Hon. Secretaries Hon. Treasurer B. E. Balme, D.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S, A. J. McComb, M.Sc., Ph.D. G. A. Bottomley, B.Sc., Ph.D. G. Perry, B.Sc. (Hons.) M. W. Perry, B.Sc. (Agric.) (Hons.) S. J. Curry, M.A. Hon. Librarian Hon. Editor A. Neumann, B.A. A. J. McComb, M.Sc., Ph.D. (1975) A. E. Cockbain, B.Sc., Ph.D. (1976) P. E. Playford, B.Sc., Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. B. Lamont, B.Sc. (Agric.), PhD., F.R.H.S. C. E. Dortch, B.S., M. Phil. L. J. Peet, B.Sc., F.G.S. P. R. Atkinson, B.Sc. J. K. Marshall, B.Sc. (Hons.), Ph.D. 5. — The petrology and archaeological significance of mylonitic rocks in the Precambrian shield near Perth, Western Australia by J. E. Glover^ Manuscript received 18 November 1975; accepted 15 December 1975 Abstract Mylonitic rock from shear zones in Archaean granitoids on the western margin of the Yilgarn Block, near Perth, Western Australia, has been used extensively for flaked tools by Aborigines. The rock is epidote-rich and flinty, with quartz veinlets, and commonly has a subconchoidal fracture. Flakes are found in the Perth Basin between Eneabba and Mandurah. They are common at eastern sites near the mylonitic zones, but are sparse or absent near the west coast. Introduction The duration of Aboriginal occupation in Aus- tralia exceeds 32 000 years (Barbetti and Allen, 1972), and workers from the Western Australian Museum have provided evidence of occupation in southwestern Australia of at least 25 000 years (Dortch and Merrilees, 1973). Localities strewn with their artifacts are numerous in the south- west (Hallam 1972, Glover 1975). Outcrop in the central Perth Basin (i.e. that part of the basin between Eneabba and Man- durah) is sparse, and most of the rock is lime- stone or sandstone, which is unsuitable for flaking. Artifacts in the area have generally been made from quartzite, quartz, silcrete, granite and dolerite, which have been carried westward from the Yilgarn Block, or from bryo- zoan chert which may have been carried east- ward from quarries now covered by the Indian Ocean. Eastern sites are dominated by quartzite flakes, but bryozoan chert tends to become increasingly important toward the west. One rock represented at many localities, and referred to previously as green non-fossiliferous chert (Hallam 1972; 1974) or veined epidote-bearing chert (Glover 1975) has been an enigma, for both its provenance and petrology have been speculative. Outcrops of the parent rock have now been identified as mylonite and altered mylonite in shear zones on the western part of the Yilgarn Block, and confirm earlier sugges- tions that the flakes were of Precambrian rock derived from the east. This note is concerned with the description of the cherty epidote-bear- ing flakes and their distribution, and with the nomenclature and petrology of their parent rock. The colours and corresponding numerical designations used below refer to the Rock-color chart distributed by the Geological Society of America (Rock-color Chart Committee 1963). 1 Geology Department, University of Western Australia. Nedlands, Western Australia, 6009. The flakes Appearance and typology The flakes of cherty epidote-bearing rock range considerably in size, but many of them have a maximum dimension of about 3 cm. The Aborigines seem to have preferred this kind of rock for making small flat adzes and backed blades (see Hallam 1974, p. 83). The rock generally breaks with subconchoidal fracture, and consists of small disseminated quartz and feldspar grains (porphyroclasts) in an aphanitic groundmass of silica and epidote granules. Many flakes are cut by thin quartz veins and minute faults. Flakes are commonly light greenish grey (5GY8/1), pale greenish yellow (10Y8/2), or pale olive (10Y6/2), and where they have been stained or bleached in surrounding soil or sand, they generally retain a greenish cast because of their epidote content. Distribution in the Perth Basin Cherty epidote-bearing flakes have been found at 31 sites in or just outside the eastern margin of the Perth Basin between Eneabba in the north and Mandurah in the south, a distance of about 300 km (see Fig 1). The sites are com- mercial sandpits or sandy areas in which the artifacts have been concentrated by deflation of sand. Although widespread, the cherty epidote- bearing flakes are generally not abundant at artifact localities in the central Perth Basin: they are absent from 13 sites, and form less than 1% of the flake population at 18 sites (see Table 1). Sites near the west coast are without exception low in these flakes, but higher propor- tions (8-39%) are found near the eastern margin of the basin, in the Gingin-Bullsbrook-Walyunga area. It should be borne in mind that the pro- portion of any one rock type in flakes found on the surface of a wind-deflated area depends on many factors, including the history of human occupation at the site, the depth of erosion, the availability of the particular rock type, the rock’s suitability in the light of typological changes, its durability, and its distance from the souice. The last-mentioned factor seems to have been the most significant here, for the distribution north of the Swan River accords well with a mylonitic source in the Walyunga region, where there are numerous outcrops of mylonite. The mylonitic rocks Outcrop distribution Mylonite and blastomylonite have been recorded by Wilde (1974) in Archaean granitoid of the Walyunga area, and in the Chittering Metamorphic Belt at Mogumber, about 130 km Journal of the Royal Society of Western Australia, Vol. 59. Part 2. September, 1976. 33 Figure 1. — Map of the central Perth Basin, showing location of artifact sites described in Table 1. Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 34 Table 1 Pro'portion of flaken of mylonitic rock at sites in the central Perth Basin. Sites were located on 1 : 250 000 Series It 502 maps, and co-ordinates are based on grid references estimated to the nearest hundredth. Co-ordinates refer to centre points of large sites. Site Xo. Type of site Map Map co-ordinates Flakes counted Percent mylonitic rock 1 Road cutting in .sand, ICooringa area SH50-9 32002760 319 0 2 Sandy soil, X bank Cockleshell Gully SH50-9 30372666 215 0 3 Sand blow-out X bank Hill River SH 50-9 31492481 310 0 4 Sand spoil from dam about 1 km W Dinner Hill SH50-10 357724.52 154 0 5 Sand blow-outs, Pinnacles area SH50-9 31692063 2 107 0 6 Sand near dam, Caro Station SH50-9 34651927 249 0-4 7 Road cutting in sand, S bank Moore River, Regans Ford ... SH50-10 36921652 133 0-7 8 Blow-outs in yellow sand, near mouth of Moore River SH50-14 34851205 886 0-3 9 Sand blow-out 2 km XXW Giiigin Railway Station SH50-14 38801260 406 8*4 10 Sandy area 6 km S Bullsbrook Hast SH50-14 40370758 290 39-0 11 Sandy area Walyunga Xational Park SH50-14 40640743 1 624 10-7 12 Blow-outs in yellow sand. 1 km north Mullaloo Beach .... SH50-14 37190695 127 0 13 Gnangara Sandpit SH50-14 38370671 249 3-0 14 Brambles Sandpit SH50-14 38410669 193 2-6 15 Sandy area, SW margin of L. Gnangara SH.50-14 38550680 684 10 16 Bell Bros Sandpit, Gnangara Road SH50-14 39730675 309 3-3 17 Ready Mix Sandpit, Beechboro SH50-14 39040592 307 4-2 18 Road cut in sand, Beechboro Road SH50-14 39180588 228 0-9 19 Widgee Road Sandpit, Beechboro SH50-14 39380584 291 2-7 20 Sand patch, S side Talbot Way, Woodlands SH50-14 36880540 373 0 21 Sand dune. XW of L. Monger SH50-14 38120517 344 0-3 22 Red sand. E shore of L. Monger SH50-14 38220512 642 0 • 5 23 Sandpit, Maida Vale SH50-14 39880510 347 0-3 24 Exposed sand, airport runway extension SH50-14 39600492 348 1-4 25 Rail cutting in sand near Wittenoom Road, Maida Vale .... SH50-14 39820496 353 0-2 26 Sand blow-out near Bingham Street, Maida Vale SH50-14 39800490 384 0-2 27 Exposed sand, Kewdale SH50-14 39550449 669 0-4 28 Sandpit, Hardey Road, Cioverdale SH50-14 39630455 717 0-1 29 Exposed sand. Xewburn Road, Kewdale SH50-14 39690452 413 0-2 30 Sand blow-out X side Dowd Street. Kewdale SH50-14 39550434 318 0-5 31 Sandpit XE White Street, Orange Grove S 150-2 40140397 306 0 32 Exposed dune, corner High Road, Leach Highway, Riverton S150-2 38700377 378 0 33 Exposed sand. Metcalf Road, Lynwood S150-2 39160377 181 0-5 34 Sand blow-out, corner Riley Road, Xicol Road, Lynwood S150-2 38980364 317 0 35 Snashall Bros Sandpit, Bibra Lake area S150-2 38080308 394 0-2 36 Coopers Sandpit, Canning Vale S150-2 39220303 277 0-4 37 Hot Mix Sandpit, Gosnells S150-2 40110265 361 0-3 38 Ready Mix Sandpit, Forrest Road, Jandakot SI 50-2 38590273 299 0 39 Calsil Sandpit, Forrest Road, Jandakot SI50-2 38780266 118 0 40 MWSS (fe OB Sandpit, Lilian Avenue, Armadale . .. SJ50-2 40110266 359 0-6 41 Fremnells Sandpit, Hopkinson Road, Cardup SI 50-2 39710126 1 589 0 42 Wellard Sandpit, Parmelia S 150-2 38540114 174 1-7 43 Sand blow-outs, Lang’s Farm, Mundijong S150-2 39820092 500 2-5 44 Sand blow-out. 0-5 km S railway bridge, Mundijong S150-2 40040058 374 1-6 45 Sandpit 4*5 km E Stake Hill Bridge, Mandurah ST50-2 38049805 116 0 Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 35 to the north. The rocks are found in a NNE- trending zone. Wilde suggests that the present Darling Fault runs subparallel to an ancient zone of north to north-east shearing that may have been reactivated many times. Mylonite is also known at Cookernup over 100 km south of the area mapped by Wilde, and is likely to crop out in intervening country. Terminology Two terms are commonly used for strongly coherent fine-grained rocks of shear zones, namely mylonite and cataclasite (or cataclas- tite). Modifications of these terms are numerous, but only the most important are considered here. The nomenclature of cataclastically deformed rocks has recently been discussed by a large number of authors, including Christie (I960), Reed (1964), Joplin (1968), Spry (1969), Bell and Etheridge (1973) and Zeck (1974), and a confusing array of names has arisen. Even the style of deformation is now questioned, so that although most authors have assumed a pre- dominance of brittle deformation as implied by the terms mylonite (Gr. myle, mill) and cata- clasite (Gr. klastos, broken in pieces), Bell and Etheridge believe that deformation can be essen- tially ductile. It is commonly held that mylonite is charac- terized by foliation, as indicated by Lapworth (1885) in his original definition, whereas cata- clasite is said to lack foliation (see Christie 1960 Reed 1964, Spry 1969). Christie, in addition, implies that cataclasite is derived from mylonite by ultrabrecciation. On the other hand, Zeck (1974) uses the term cataclastite in a fairly wide sense for fine-grained foliated and massive rocks that have been formed by rupture and show no noteworthy recrystallization. Cataclastic rocks whose rupture has been accompanied or followed by sufficient recrystal- lization or neomineralization to obscure their cataclastic nature have commonly been given names prefixed by hlasto, e.g. blastomylonite, blastocataclasite. In this paper, the term mylonitic rock is used to include all the varieties of strongly coherent, fine-grained rock considered above. Petrography Mylonitic rocks from the bed of the Swan River 0.8 km SSE of Mt Mambup range from mylonite to blastomylonite. They are greenish grey (5GY6/1-5G8/1) to medium bluish grey (5B5/1) with rough to subconchoidal fracture, and some are strongly foliated. Here and there folia are displaced a few millimetres by faults, and are cut by thin quartz veins. The mylonites contain elongate, aligned quartz and feldspar porphyroclasts set in a poorly foliated ground- mass mainly of very finely divided silica and epidote. The quartz porphyroclasts show undu- lose extinction or have been converted into numerous subgrains. Some rocks have a granular groundmass with a mean grainsize of 0.06 mm. evidently due to recrystallization, and are best called blastomylonite. At Walyunga, the mylonitic rock is a flinty, almost aphanitic, pale greenish yellow to pale olive ( 10Y8/2-10Y6/2) rock that is practically massive and breaks with subconchoidal fracture. Porphyroclasts of quartz and subordinate sodic plagioclase occur in a groundmass of epidote granules having a mean grainsize of about 5 ^m, and a little silica. A few porphyroblasts of epi- dote attain a diameter of 100 ^m. The quartz porphyroclasts are strained or finely granulated, whereas the feldspar has survived well except for local displacement of twin lamellae by micro- faults. The porphyroclasts compose only about 5% of the rock and tend to have a common lineation. The rock is cut by veins of micro- crystalline quartz that are generally less than 0.5 mm thick. The veins have many orientations, but there is commonly one set more or less parallel to the orientation of the porphyroclasts, and two other sets at about 60° to that direction. Some veins pinch and swell, others consist of narrow lenses arranged en echelon. This rock is difficult to fit precisely into existing classifica- tions, but is evidently an altered mylonite or cataclasite. Comparison with flakes. The cherty epidote-bearing flakes of the Perth Basin, and of the Walyunga area just outside its eastern margin, are very like the mylonitic rocks in colour, fracture, mineralogy and texture. Thin sections of flakes and mylonitic rock are practically indistinguishable (see Fig. 2 A-F). Discussion The pronounced lithological similarity between the cherty epidote-bearing flakes and mylonitic rocks of the Walyunga area provides a clear indication of their mylonitic origin. As noted earlier (Glover 1975), the textures of some flakes resemble those of metasedimentary and meta- Figure 2. — A. — Mylonitic rock from outcrop at Walyunga. White porphyroclasts are strained to finely granulated quartz, groundmass is finely granular epidote with a little silica. Veins are quartz. University No. 74903. Width of field 4.75 mm. Plane polarized light. B. — Artifact from Bell Bros Sandpit, Gnangara Road, Grid Ref. 39730675 (locality 16). White porphyroclasts are finely granulatsd quartz, groundmass is finely granulated epidote and a little silica. Veins are quartz. University No. 74737. Width of field 4.75 mm. Plane polarized light. C. — Mylonite from outcrop in bed of Swan River, 0.8 km SSE of Mt Mambup. White porphyroclasts are strained and granulated quartz with subordinate sodic plagioclase, groundmass is a fine mixture of silica and epidote. Veins are quartz. University No. 74900. Width of field 4.75 mm. Plane polarized light. D. — Artifact from sand. Kewdale, Grid Ref. 39550449 (Locality 27). White porphyroclasts are strained and granulated quartz, groundmass is a fine mixture of silica and epidote. Veins are quartz. University No. 74617. Width of field 4.75 mm. Plane polarized light. E. — Mylonite from outcrop in bed of Swan River 0.8 km SSE of Mt Mambup. White porphyroclasts are strained and granulated quartz, groundmass is a fine mixture of silica and epidote with a little fine white mica parallel to the foliation. Chlorite is also locally present. University No. 74901. Width of field 4.75 mm. Plane polarized light. P. — Artifact from sandy area, Walyunga, Grid Reference 40640743 (locality 11 ). White porphyroclasts are strained and granulated quartz, groundmass is a fine mixture of silica and epidote with a little fine white mica parallel to the foliation. University No. 74904. Width of field 4.75 mm. Plane polarized light. Journal of the Royal Society of Western Australia. Vol. 59, Part 2, September, 1976. 36 Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 37 volcanic rocks, but it can scarcely be doubted that these textures arise from the cataclasis and neomineralization of Archaean granitoid. The tendency for the mylonitic flakes in areas of the Perth Basin north of the Swan River to increase in frequency toward Gingin and Wal- yunga shows that the sources could have been in mylonitic rocks of the Walyunga area, or in counterparts along the strike of the Darling Fault Zone towards Mogumber. There may have been other sources of mylonitic rock for sites south of the Swan River. Acknowledgements. — Helpful discussions were had with Mrs. S. J. Hallam and Mr. R. H. Pearce of the Depart- ment of Anthropology. University of Western Australia. Dr. R. Marston, Geology Department. University of Western Ausirralia. and Dr. S. A. Wilde of the Geological Survey of Western Australia. References Barbetti, M.. and Allen, H. (1972). — Prehistoric man at Lake Mungo, Australia, by 32 000 years B.P. Nature, Lond., 240 (5375), 46-48. Bell. T. H., and Etheridge, M. A. (1973). — Microstructure of mylonites and their descriptive termin- ology. Lithos, 6, 337-348. Christie. J. M. (1960). — Mylonitic rocks of the Moine Thrust-Zone in the Assynt Region. North- West Scotland. Trans Edinburgh Geol. Soc.. 18. 79-93. Dortch, C. E., and Merrilees. D. (1973). — Human occupa- tion of Devil’s Lair. Western Australia, dur- ing the Pleistocene. Archaeol. Phys. Arithrop. in Oceania, 8. 89-115. Glover, J. E. (1975). — The petrology and probable strati- graphic significance of Aboriginal artifacts from part of southwestern Australia. J. R. Soc. West. Aust. 58, 75-85. Hallam, Sylvia J. (1972). — An archaeological survey of the Perth area. Western Australia: a pro- gress report on art and artefacts, dates and demography. Aust Inst. Aboriginal Studies Newsletter, 3 (5), 11-19. Hallam, Sylvia J. (1974). — Excavations in the Orchestra Shell Cave, Wanneroo, Western Australia. Part 1. Ethnographic and environmental background. Archaeology and Physical Anthropology in Oceania, 9 (1), 66-84. Joplin, G. A. (1968). — A Petrography of Australian Meta- morphic Rocks. Angus & Robertson, Sydney. Lapworth, C. (1885). — The Highland controversy in British geology: its causes, course and conse- quences. Nature, Lond. 32. 558-559. Rsed. J. J. (1964). — Mylonites, cataclasites, and associ- ated rocks along the Alpine Fault. South Island. New Zealand. N.Z. Jl. Geol. Geophys. 7, 645-684. Rock-color Chart Committee (1963). — Rock~color Chart. Geol. Soc. Am., New York. Spry. A. (1969). — Metamorphic Textures. Pergamon, Oxford. Wilde. S. A. (1974). — Explanatory notes on the Archaean rocks of the Perth 1 :250 000 geological sheet. Western Australia. Rec. Geol. Surv. West. Aust. 1974/15 (unpubl.) Zeck. H. P. (1974). — Cataclastites, hemiclastites, holo- clastites, blasto-ditto, ana myloblastites — cataclastic rocks. Am. J. Sci. 274, 1064-1073. Journal of the Royal Society of Western Australia, Vol. 59. Part 2, September, 1976. 38 6. — The history of two coastal lagoons at Augusta, Western Australia by E. P. Hodgkin^ Manuscript received 15 December 1975; accepted 17 February 1976 Abstract The Deadwater and Swan Lake are two small lagoons which form appendages to the estuary of the Blackwood River near its mouth. Neither was a part of the estuary when the first white settlers arrived in 1830; Swan Lake was then a fresh-water lake with a stream to the estuary and the Deadwater is not shown at all on early maps. The Deadwater now has the characteristic form of a coastal lagoon that has developed as the result of diversion of the river mouth parallel to the shore behind a wave-built barrier on a prograding coast. The mouth is known to have migrated some distance eastwards during the 1930s and the river then flowed through what is now the Deadwater. The bar closed for the only time on record in 1945 and was reopened near its original site. This left the Deadwater as a coastal lagoon behind low dunes and the coastal alignment is now almost identical with that of the 1830s. Introduction This enquiry into the origin of the Deadwater and Swan Lake at the mouth of the Blackwood River has arisen out of a study being made for the Environmental Protection Authority of 1 6 Princes Street. Mosman Park, Western Australia. 6012. Western Australia into all aspects of the estuarine ecosystem (Hodgkin, 1976). These two lagoons have been added to the estuary in this century and are now a significant part of it, at least from the biological point of view. It is of considerable interest to understand how this came about, to know whether the events des- cribed were the result of natural processes or are attributable to human interference. Was the formation of the Deadwater a fortuitous event caused by vagaries of short term climatic change or was it caused by human activities? Is it evidence of a prograding coastline with an excess of mobile sand? The following history has been pieced together from a few documents, from old maps arid air photographs, and from the sometimes conflicting accounts given by local inhabitants of what happened 30 to 50 years ago. The catchment and the estuary The Blackwood River is the largest river of the south west, with an estimated average annual discharge of 1 057 x lO'^ m^ more than twice that Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 39 Figure 2. — Tracing from map dated 1832-4-8. This and the subsequent figures have all been reduced to the same scale. Broken lines outline sand banks. of the Swan River with its larger catchment. Floods have resulted in discharge rates in excess of 480 mVsec (Public Works Department, West- ern Australia, 1972). Flow is extremely seasonal with 97.5% of river runoff occurring during the six months June to November (as measured at the nearest gauging station). In consequence the estuary is fresh throughout in winter and brackish to marine in summer, often with extreme stratification of the water body. Tides though small (maximum daily range about 0.6 m) are little damped in the estuary proper and strong tidal currents flow in and out of the two lagoons. The Deadwater is stratified in winter, surface water may be fresh and that below the level of the sill about half sea water salinity. The Dunsborough Fault runs north-south through the estuary (Fig. 1). To the west of this is the Naturaliste-Leeuwin ridge of Pre- cambrian rocks while to the east there are Mesozoic sediments of the Donnybrook Sunkland and Quaternary sands of the Scott Coastal Plain where three old shore lines are recognised parallel to the present coast (Lowry, 1967). The estuary is 42 km long; the greater part is a tidal river 50 to 100 m wide and about 5 m deep which discharges into the small, very shallow. Hardy Inlet with a deep channel to the mouth and bar. The mouth is sheltered from the pre- dominantly westerly winds in winter by the Leeuwin ridge, but is exposed to strong south- easterly winds in summer. 1830 Early maps and paintings by Thomas Turner make it clear that when Captain Stirling and the early settlers came to the Blackwood in May 1830 the mouth of the river was where it is now. The first map, dated 1832-4-8 by Hillman, Turner and Edwards represents an accurate survey of the lower part of the estuary. On this map (Fig. 2) Dukes Head and Point Frederick and their associated sand banks have much the same form they have now. To the east of the mouth the shore line was where it is now. Unfortunately there is no indication of the nature of the shore and it can only be assumed that then as now there were low dunes. The large Swan Lake had its present size and shape and a smaller lake extended about 1 km further east. A small winding stream, Rushy Creek, flowed from the lake to the estuary near its mouth. There was no Deadwater. Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 40 1878 The Admiralty Chart (Fig. 3) only shows the coastal features in detail, as is usual in these charts. Again the mouth and shore line are depicted with essentially their present form. Low dunes are shown in what is now the western end of the Deadwater. A “Townsite” plan dated 1899 appears to show considerable erosion of Point Frederick, but it is difficult to match this plan to other maps and photographs. 1925 A detailed hydrographic survey of the channel part of the estuary (Public Works Department, Western Australia, 23962) again shows the mouth with much the same form as in 1830 (Fig. 4). A low sandy patch extends 150 m east from Dukes Head and the mouth has moved slightly east at the expense of Point Frederick which has retreated about 100 m. The eastern shore line is unbroken; there is nothing to show where Rushy Creek discharged. The survey was made in April and the mouth of the creek was probably blocked with sand. Figure 4.— Public Works Department, Western Australia, chart dated 1925. Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 41 1929 By this date: “the bar has so silted up that getting into or out of it from the Bay is a matter of impossibility” and “The mouth of the Blackwood is fast being blocked by a sand bar a few chains in width . . (Letters on Fisheries Department file 165/21). There is no indica- tion where the mouth was at this time. 1933 Now “the opening to the sea is about a mile east and each year sees it moving further east- ward, the rate of movement has been fairly rapid, the present distance of about a mile being covered in the last twelve years” (letter from Augusta-Margaret River Shire, 29/4/1933). The letter petitions the Minister for Works to make a survey and re-open the old entrance, because of shallowing of the bar and an adverse effect on the fishing industry. 1943 Between 1933 and 1943 the mouth did not in fact move much further east; movement seems to have ceased when the mouth cams to a big dune about 2.5 km from the old mouth. An air photo taken 20/11/1943 (Fig. 5) shows the river flowing in a channel 100 to 200 m wide through what is now the Deadwater to the new mouth. Between this and the sea there is a sand spit 200 to 300 m wide that extends from Dukes Head to the new mouth. It is clear from the correspondence and from the recollection of local residents that between 1925 and 1930 the sand round Dukes Head built out eastward as a broad spit and that Point Frederick retreated, in spite of an attempt to stabilise it with marram grass about 1927 or 1928. By 1930 the spit was sufficiently consolidated for people to drive along it in trucks and fish from them into the estuary. Light planes landed and took off on the spit and the perilously short landing strip was reportedly about opposite the present entrance to Swan Lake from the Dead- water; a photograph dated 1st January 1931 shows a Tiger Moth plane on the spit close to the then mouth. The air photo shows that the spit was wider than the present beach ridge and the shore line about 100 m seaward of its position today, or in 1830. Scattered vegetation grew on the western half but this was probably only pioneer plants such as Arctotheca populi- folia which colonise open sandy shores. Figure 5.— Tracing from air photo of 1943. Stippled areas: bare sand, thought to be the result of attempts to block the creek. Journal of the Royal Society of Western Australia, Vol. 59. Part 2, September. 1976. 42 There is no evidence that the bar ever closed during this period, though it was evidently very shallow. There were continual complaints from fishermen both because the passage was dangerous and because the bar was believed to obstruct the entry of fish. Whatever the con- dition of the bar in summer a considerable volume of water flowed through the new channel and mouth in winter keeping them scoured out. 1945 The events of this year are told in an article in the West Australian newspaper by Noel M. Brazier, dated 14th July 1945 and entitled: Opening the Blackwood, how the river reached the sea again. “By the middle of March this year the water of the Blackwood River was barely flowing over its long sand bed where it entered the sea. The whole landscape of the river mouth had altered from what it was some 54 years ago; the new mouth had been diverted from its then course to about one mile east. “For some years small round sand hills had filled the previous outlet of the river and continued east for nearly a mile . . . “By the end of March the river mouth had silted up; strong winds from the sea had raised the outlet some 3 ft. The river was completely shut off from the sea; fishing was bad. Seeing the danger to the low-lying land adjacent to the river Keith McWhae now permanently residing there, took his instruments out to find the easiest place to let the water out. He decided the best place was close to the natural outlet of over 50 years before. “The people were anxious and appealed to the Government. Two engineers came down and agreed with Mr. McWhae but they said no men could be spared to do the work. The river had now risen some 4 ft. and the camping ground was flooded. “The work must be done — or where would the water rise to? So a team of men started to work at the selected spot — and some of the sand hills were fairly high. Yet they got the water through in a small stream; the trouble did not end there. The sand hills began falling in; ... But three men stuck it out for five days, when the strength of the river suddenly began to break through. They had succeeded. “A huge gap, 100 yards wide and many feet deep, was torn through the cut and in a few hours had lowered the height of the river some 4 ft. The troubles of the mouth were over. The cost in wages was less than £18. The fish from the sea can now reach the river; again the Blackwood will become a good fishing ground, and the friendly por- poise sport around”. Local residents still recall the spectacular out- rush of river water that spread as far as St. Alouarn Island 9 km from the mouth and carried with it fishermen’s nets which had been set in the estuary. The Public Works Department did in fact pre- pare a plan (Public Works Department, Western Australia, 23962 dated 24/5/1945) for a channel 325 m long across the bar. The locality sketch shows “high sand hills” where the mouth is now and the proposed chan- nel is marked about 400 m east of this where the highest point on the bar was about 1.5 m above sea level. This was probably the “selected spot” referred to above. It would be interesting to know the actual date on which the bar was breached. River flow data, as recorded at Nannup 140 km upstream, show that even in May flow had increased little above the low summer levels and it was only in mid-June that the river began to flow strongly. Fortune, tely 1945 was a wet winter with the greatest river flow recorded in the 17 years for which there are records (Public Works Depart- ment, Western Australia, Water Resources Sec- tion, 1972) and the river scoured out for itself a good entrance which has never again closed. 1955 Air photos taken 23/5/1955 and 3/12/1955 (Fig. 6) show a shore alignment which is almost identical with that of 1830; the mouth has returned to its previous position and both Dukes Head and Point Frederick have essentially the form and dimensions shown on the 1830 map and that of the present time. The spit is now a narrow dune with a steep seaward face and with well established vegetation, except at the eastern end. There, where the mouth was in 1943, the dune appears lower and has only sparse vege- tation. The 1943 river channel is now the Dead- water. At its western end there is a broad, shallow bar of mobile sand with a well defined deep channel to the river mouth. The eastern end has already silted up, though apparently still not vegetated. 1975 The only significant differences between con- ditions in 1955 and those of the present time are the full development of the fore dune along its whole length and establishment of vegetation on it. Rushes are well established at the eastern end of the Deadwater and the bar at the western end has consolidated and carries some vegeta- tion. A narrow channel keeps open to the north of the bar, and tends to scour its northern shore. Interpreting the changes There have undoubtedly been considerable changes in the estuary and coastline during the Holocene and the Swan Lakes mav be attribut- able to this period; however the close similarity between present topography and that of the period of 1830 to 1925 suggests that this is a stable form for the mouth and coastline at the present time. The events of 1925 to 1945 may have had purely natural causes. It has been suggested that a series of dry years allowed the bar to silt up in summer and so forced the river to erode its way eastward during reduced winter flows. The rainfall record gives little support Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 43 to this explanation; 1925 was indeed a very dry year, but the following year was one of floods and the next eight years all had above average rainfall. Stormy seas may have cut back Point Frederick and transferred the sands to Dukes Head. There was just such a shift early in 1974 and continuation of such a process over several years could be expected to deflect the path of the river and increase the rate of erosion of Point Frederick. Alternatively growth of the spit may have been precipitated by human activi- ties such as destruction of the vegetation on the east bank and consequent mobilisation of the sand: a big fire is said to have gone through the Deadwater area in the early 1920s. A number of possible hypotheses could be advanced. The following explanation is proposed as best fitting the evidence available. Information additional to that recorded comes from members of the Ellis family who have owned property to the east of the Inlet since the turn of the century, and from various residents of Augusta. Much of the land east of the Inlet was, and still is, used for grazing cattle and this includes the Deadwater area. This area provided good grazing behind the fore dunes; it would have been an inter-dune depression at the level of the water table, separated from Swan Lake by low dunes and swampy ground as it is now. Rushy Creek flowed through this, draining fresh water from the lake to the estuary as shown on the 1832 map (Fig. 2>. The lake floor then, as now, must have been below sea level. Wave action would have closed the mouth of the creek with sand each summer; no opening is shown on the chart surveyed April 1925. Water level built up in the lake in winter draining out slowly through Rushy Creek, and in spring the Ellis family cut through the sand bar releasing the water in order to take advantage of good grazing round the lowered lake; this is said to have been done each year for several years, and probably caused scouring of the creek. Eventually some time in the 1920s failure of the sand bar to close allowed sea water to flow back into Swan Lake when there were unusually high tides. The 1943 air photo appears to show a later attempt to close the creek in two places between Swan Lake and the Deadwater. The salt water would have killed vegetation in the lake and flooded low-lying swampy ground between it and the fore dune killing vegetation here too. Add to this destimction of dune vege- tation by fire on Point Frederick and the stage Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 44 was set for natural processes to take over; for Point Frederick to erode, for the mouth to migrate eastwards, and for a substantial sand spit to build out from Dukes Head. What would have happened if a channel had not been cut through in 1945? Water level would have built up until before long the river would have broken through at the weakest point of the spit, perhaps where the cut was made, but perhaps again at the eastern mouth. In the latter event the sand spit might eventually have consolidated, a dune formed on it, and the river continued to flow through the Deadwater channel, but with the bar tending to close each summer because of reduced tidal exchange. How long this essentially unstable condition would have persisted it is impossible to say. The fact that after 1945 mouth and coastline rapidly reverted to their original condition suggests that there was not the necessary reservoir of mobile sand to permit the 1945 topography to consoli- date and for a dune to build up on the spit. Conclusion On the evidence presented it is concluded that the changes to the river mouth which took place during the decade 1925 to 1935 were precipitated by quite small scale human interference: cutting the bar at the mouth of Rushy Creek to drain Swan Lake and possibly burning the fore dune. Natural forces took over and built a substantial sand spit, a spit which diverted the course of the river and forced it to gouge out a new channel that carried the mouth 2.5 km eastwards until halted by high dunes. The Deadwater formed when the sand of the spit was remobilised following the reopening of the original mouth in 1945. This sand formed a bar which closed the connection of the diverted river with the mouth, leaving only a small tidal channel; some of it probably contributed to the shallow bank of mobile sand in the inlet opposite Seine Bay; it silted up the eastern end of the Deadwater; it regenerated Point Frederick and the fore dunes to the east and returned the coastline to an alignment very close to that of 1830, There is no evidence of any net gain or loss of sand to the system. The Deadwater is silting up slowly and, left to itself, it might eventually return to the 1830 condition. Tidal currents tend to keep its mouth open, but bars form both inside and outside this and there have already been requests to have a channel dredged through the mobile sand. Damage to the fore dunes on Point Frederick by beach buggies has been blamed for erosion of the point in 1974, but whether this is sufficient by itself to cause changes of the scale of those of 1925-1935 to be repeated is speculation only. The point has since largely regenerated though it will be some time before the vegetation is re-established. However, it is clear that there is a situation of uneasy balance in which quite small changes to the natural environment may precipitate major movements of the poorly con- solidated sands near the mouth of the river. Acknowledgements. —IX, is a pleasure to acknowledge assistance from present and former residents of Augusta during the course of my enquiries. Dr. E. C. F. Bird kindly read and criticised the manuscript. I am grate- ful to Dr. W. Andrew both for help during the investi- gation and for facilitating preparation of the figures This study was made during the course of an investi- gation of the Blackwood River estuary being made for the Environmental Protection Authority of Western Australia. References Hodgkin, E. P., 1976. Environmental Study of the Black- wood River Estuary. A report to the Estu- arine and Marine Advisory Committee of the Environmental Protection Authority, June, 1976 (unpublished). Lowry. D. C., 1967, Busselton and Augusta, Western Australia. 1:250 000 Geological Series, Explanatory No^es. Geological Survey Gj Western Australia. Public Works Department, Western Australia, 1972. Streamflow Records of Western Australia 1939 to 1970. Volume I. Planning, Design and Investigation Branch, Water Resources Section. Journal of the Royal Society of Western Australia, Vol. 59. Part 2, September, 1976 45 7. — ^The environment of deposition of the Wooramel Group (Lower Permian), Lyons River area, Carnarvon Basin, Western Australia by G. J. McGann’ Manuscript received 19 August 1975; accepted 22 June 1976 Abstract Studies of the Wooramel Group, comprising the Moogooloo Sandstone and the Billidee For- mation, in the Lyons River area indicate that the environment of deposition of the Moogooloo Sandstone was a temperate to cold, shallow marine environment, perhaps deltaic. Sedimen- tation was rapid, with abundant reworking of sediments and steady, fairly strong currents, possibly tidal in part. Conditions changed in late Moogooloo time with the development of euxinic conditions with periodic influxes of sand. The overlying Billidee Formation is believed to have been deposited in shallow water near- shore conditions, with occasional subaerial exposure. Fairly strong currents flowed to the north during deposition of the sandstones. Stagnant anaerobic conditions prevailed during the deposition of the carbonaceous shales. There was rapid alteration between aerobic and anaerobic conditions. The deposition of the Edmondia band was marked by a minor marine transgression. Modern sedimentary analogues are discussed. Introduction This paper presents the results of a study of the environment of deposition of the Wooramel Group (Lower Permian) in the eastern Carnar- von Basin. In the area studied (Fig, 1) the Wooramel Group comprises the Moogooloo Sand- stone conformably overlain by the Billidee For- mation. No previous work has been published dealing with the depositional environments of the Moo- gooloo Sandstone or the Billidee Formation, except for the brief comments made by Condon (1954, 1967). The base of the Wooramel Group is at the sharp change in lithology from fossiliferous calcareous sediments (Callytharra Formation) to arenaceous sediments largely devoid of fossils. The top is at the change in lithology, commonly gradational, from arenaceous sediments to dominant lutite sediments. In the area mapped, the Wooramel Group comprises the conformable sequence of the Moo- gooloo Sandstone and the Billidee Formation. The Moogooloo Sandstone, the lower unit, over- lies the Callytharra Formation, either conform- ably or disconformably. The Billidee Formation is overlain unconformably by the Coyrie Forma- tion (Byro Group). The age of the Wooramel Group is Artinskian because of its position between the Callytharra Formation, of late Sak- marian-early Artinskian age, and the Artinskian, Byro Group. Moogooloo Sandstone In the Lyons River area, the Moogooloo Sand- stone is the basal formation of the Wooramel Group. 1 12 Robin Street, Mt. Lawley, Western Australia, 6050. The Moogooloo Sandstone consists of red, brown and white, fine- to coarse-grained ortho- quartzite, subordinate subarkose (feldspathic sandstone) and orthoconglomerate. Minor inter- bedded silty shale is present in the upper part of the formation. The petrology has been des- cribed by McGann (1974). In the Moogooloo Sandstone, sedimentary structures give the best clues to the environment of deposition. Cross stratification Cross stratification is common in the Moo- gooloo Sandstone with sets ranging in height from 2 cm to 3 m. A subdivision into small-, medium- and large-scale cross stratification (Conybeare and Crook 1968) is used. Medium-scale. — Medium-scale cross stratifica- tion, which is the most common type of cross stratification in the formation, ranges from 60 mm to 2 m. The average height of a set is about 33 cm. A coset comprises from one to four individual sets. The maximum dip is 30°. The medium-scale cross stratification is of Allen’s (1963) Omikron class, formed, he believes, by migration of trains of large-scale asymmetric ripple marks with essentially straight crests. This type of ripple is found in channels or in the open sea at depths many times the wave-ripple height, but still in shallow water. Large-scale . — Trough cross stratification is the most common type of large-scale cross bedding. The large-scale cross beds are isolated sets bounded above and below by “massive” fine- to medium-grained orthoquartzite. The maximum dip is 32° and the average dip is 16°. The large-scale cross stratification is of Theta type (Allen 1963), which is believed to be scour and fill structures formed in a shallow water environment. Small-scale. — Small-scale (less than 6 cm) cross stratification, although not common in the Moo- gooloo Sandstone, is present within small, essen- tially symmetrical ripples of amplitude less than 3 cm. The cross stratification is tabular, solitary and lithologically homogeneous. The small-scale cross stratification is probably Lambda type (Allen 1963), formed by a migration of small- scale straight-crested ripples in a shallow water environment. Small-scale ripple-drift cross lamination is present in one specimen. Walker (1963) suggests that this type of cross lamination is produced by a steady current and abundant sediment supply. The azimuth of 152 foresets was measured, one to approximately every 500 m^ The vector mean is N 3° E, with a high variance of 8 440 (standard Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September. 1976. 47 Figure 1. — Locality plan. Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 48 deviation of 92°). The results are summarized in Fig. 2. The dominant current flow is towards the north with a subordinate flow towards the south. Ripple marks Slightly asymmetric ripples are the only type recognized in the area studied. The ripples are of an amplitude ranging from 1 cm to 5 cm with a wave length between 10 cm and 1 m. Slightly asymmetric ripples suggest an environment of deposition above wave base in a shallow, relatively open sea. Slump structures Two distinct types of slumping are observed in the Moogooloo Sandstone. Slumping within cross stratified^ sets. — The fore- sets of some medium-scale cross stratified sets are distorted. This distortion ranges from minor folds on the foresets to folds where the foresets are parallel to the lower bounding surface. The upper surface of the set is everywhere erosional, indicating that the deformation was penecon- temporaneous with deposition. The deformed sets are confined to particular horizons. Jones (1962) presents five possible causes of this type of deformation, the most acceptable in the present instances being surface thrusting gener- ated by currents while the sand is still in a plas- tic state. CURRENT ROSETTE MOOGOOLOO SANDSTONE ‘'Normar slumping. — “Normal” slumping is present in plane bedded fine- to medium-grained sandstones as minor folds which do not persist vertically, and is often in the same general horizon as the slumping within cross-stratified sets. Shale clasts Two types of shale clast are present, both restricted to stratigraphically variable beds which persist for up to 300 m. Both are present only in fine- to medium-grained orthoquartzites The most common type of shale clast consists of scattered plates of grey fine-silt to clay-sized material roughly parallel to the bedding. The plates are about 2.5 cm in diameter and less than 3 mm thick. ^ A second type consists of tear or ellipse shaped silty-clay clasts about 11 cm long and 4 cm high. Fossils Indeterminate brachiopods, bryozoans and bivalves have been found sporadically in the formation throughout the basin. Despite an active search, however, only fossil wood was found in the area mapped. Trace Fossils. — Tracks are present only in the upper part of the Moogooloo Sandstone, where they are extremely common. The tracks are confined to particular horizons. They consist of smooth, slender, circular tubes about 1 cm in diameter. In places, the tubes, which are essen- tially straight, interpenetrate, but they do not branch. At least one species of Palaeophycus Hall 1847 is represented. The affected beds are extensively ferruginized. Osgood (1970) believes that Palaeophycus tracks were produced by worms and the lack of distinct pattern indicates that they were preda- tors, probably in a shallow water environment. Environment of deposition From the data accumulated in this investiga- tion it is possible to construct an environment of deposition for the Moogooloo Sandstone. Cross stratification types and ripple marks suggest shallow waters with the depositional surface above wave base. Shale clasts indicate subaerial exposure or near-shore marine con- ditions (Conybeare and Crook 1968, p. 20). Palaeocurrent data indicate that the palaeo- slope was towards the north, the high variance indicating that the slope was of low angle. Tidal currents could be the main factor controlling the subordinate south-trending current Sunborg (1956, cited by Lauff 1967) has indicated that currents of approximately 0.3 m/sec. would be necessary to move the average sized grain (0.3 mm). Obviously current strengths were greater at the time of deposition of conglomer- ates and coarse-grained sands. The high degree of winnowing of the sands suggests that the current was not intermittent, but persisted over long periods. Mineralogical and textural maturity, as indi- cated by the low proportion of labile minerals good rounding and sorting, suggests that con- siderable reworking has taken place (McGann Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976 49 1974). A high-grade metamorphic or granitic provenance is indicated. The virtual absence of fossils is probably due to a combination of continual reworking, attri- tion and abrasion, leading to almost complete breakdown of fossil grains. This reworking was probably combined with rapid sedimentation which would have two effects, the first being to produce an environment unfavourable to all but a few fossil groups. The second effect would be to “dilute” skeletal remains with abundant terri- genous clastic material. The overall result is that fossils remain only in isolated patches, probably representing more sheltered parts of the environment. The sparse fossils that are present, including Palaeophycus sp., indicate a shallow marine environment. The abundance of fossil wood indicates proximity to land and is often diagnostic of deltaic conditions (Shepard 1964). The climate was temperate to cold, as inferred from palynological evidence obtained from a shaly unit just above the Moogooloo-Billidee contact. There is no evidence to suggest that climatic conditions changed greatly between Callytharra and Moogooloo time. The climate in Callytharra time is interpreted as cool (Teichert 1941). The environment of deposition changed some- what in late Moogooloo time with the deposition of gypseous carbonaceous silty shale interbedded with coarse-grained orthoquartzite containing Pa'aeophycus sp. This genus is indicative of shallow water conditions, and thus the inter- bedded carbonaceous shale was also deposited in shallow water. A stagnant environment was responsible for low oxidation potentials, which resulted in the high proportion of carbonaceous matter in the shale. Influxes of coarse-grained sand, deposited in aerobic conditions (as indi- cated by Palaeophycus sp.) periodically inter- rupted the euxinic environment. When considering the environment of depo- sition of the Moogooloo Sandstone, the large outcrop area with uniform gross lithology (Teichert 1952) must be taken into account, and only a sedimentary model that is laterally extensive can be considered analgous to the conditions in Moogooloo time. In summary, the available evidence indicates a shallow nearshore marine environment, per- haps deltaic, with rapid sedimentation, abundant reworking and steady, fairly strong currents, possibly tidal in part. Conditions changed in late Moogooloo time with the development of euxinic conditions with periodic influxes of sand. Comparison of the Moogooloo Sandstone with recent analogues is handicapped by the lack of detailed work done on modern terrigenous sedi- ments. Fisher et al. (1969) briefly describe the Gulf of Papua delta, naming it as the type example of a destructive, tide-dominated delta. The facies present in this model are analogous with the facies present in the Moogooloo Sand- stone. The bulk of the formation — the ortho- quartzite suite — is thought to represent the tidal sand-bar facies, the textural and mineralogical maturity, together with the lack of fossils and the sedimentary structures all being produced in a predominantly marine environment with extensive reworking by tidal currents, and to a lesser extent, by waves. The conglomerate in the Moogooloo Sandstone may represent tidal channel facies, the channels running between sand bars. A shaly prodelta facies is not seen in the formation. In late Moogooloo time, euxinic conditions gradually became prevalent, probably by the development of barrier bars with a complement- ary lagoonal-estuarine system. Periodic influxes of coarse sand, perhaps land derived, were intro- duced into the euxinic environment. The change to shallow water conditions could have been produced by a minor regression or, more likely, by normal building up of the delta platform as sediment accumulated. Condon (1954) considered that the Moogooloo Sandstone was the product of slow deposition, caused by a major transgression on to a stable shelf. Major transgression may not have been necessary, however, as the underlying Cally- tharra Formation was laid down in a deep shelf environment, and the Moogooloo Sandstone represents a delta prograding on to this shelf. Thus, no major change in sea level is necessary to explain the Callytharra and Moogooloo facies. Billidee Formation The Billidee Formation consists of ortho- cuartzite and calcite-cemented sandstone (some fossiliferous) and gypseous, carbonaceous silty shale. Only one unit in the Lyons River area persists throughout the area mapped. This unit is an easily identified calcite-cemented sandstone with locally abundant bivalves. The unit is named the Edmondia band, after the large bivalve which is characteristically present. As with the Moogooloo Sandstone, in the Billi- dee Formation, sedimentary structures are useful in reconstructing the environment of deposition. Cross stratification is common in the sandstone units, but poor outcrop makes detailed examina- tion impossible. Both large- and small-scale cross stratifications are present, large-scale being by far the most abundant. The average height of large-scale cross stratifi- cation is about 60 cm and the maximum dip is 30° (average for 42 readings is 19°). The only type observed took the form of single troughs. Small-scale cross stratification is present as cosets containing up to four sets, the height of each set being about 3 cm. The lower bounding surface is erosional, and the foresets are probably trough shaped. The small-scale cross stratifica- tion is the Nu type of Allen (1963), probably formed from the migration of trains of linguoid small-scale asymmetrical ripples in shallow water. The azimuth of 96 foresets was measured, one to approximately every 200 m^. The vector mean is N 16° W with a variance of 3 650 (standard deviation of 59.7°). The results are summarized in Fig. 3. The dominant current direction was towards the north. Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 50 CURRENT ROSETTE Billidee Formation 344 ® Rare symmetric, small-scale ripples were observed in fine-grained orthoquartzite. Shale clasts are abundant in the Billidee Formation, occurring in beds traceable for up to 500 metres. The clasts are essentially planar grey mudstone plates about 3 cm in diameter and about 3 mm thick, and are oriented parallel to the bedding. Vertical ferruginized tubes are common in the Edmondia band. The presence of a central, more ferruginized, column suggests that they are root casts and not burrows. One specimen in the Edmondia band shows well-developed root casts in a clayey lime mud. The casts are about 1 cm in diameter and are infilled with detrital grains supported by ferroan calcite cement. The area surrounding the cast is extensively ferruginized, probably resulting from oxidation of organic matter. The root casts are cut by a number of collophane “tabulae”, which penetrate the car- bonate mud, where they bifurcate. The tabular form of the veins indicates that they are prob- ably dessication cracks, infilled with collophane. A ferruginized clay horizon, about 4 mm thick, is present. It is believed to represent a fossil soil horizon. The carbonate mud contains many sparry ferroan calcite patches up to 1 mm long and 0.02 mm wide. These patches are thought to represent root hairs that have rotted out, the voids being infilled with sparry ferroan calcite. Marine fossils are closely associated with the root casts. Fossils The fossils previously reported from the Billi- dee Formation are: Bivalvia : Nuculopsis (Nuculanella) sp. Nuculana sp. Oriocrassatella sp. Aviculopecten sp. Schizodus sp. cf. S. kennedyensis Dickins Stutchburia n. sp. Gastropoda: Mourlonia (Pseudobaylea?) n. sp.? Macrochilina sp. Warthia sp. Pleurotomaria sp. Bellerophontid Ammonoidea : Propinacoceras sp. Neocrinites sp. Brachiopoda: Neospirifer sp. Aulosteges sp. Strophalosia sp. Chonetid Productid Crinoidea: Calceolispongia sp. An indeterminate bryozoan was also reported. In the area mapped, fossils were found only in the Edmondia band (except for a highly cal- creted bivalve lower in the formation). The fauna is impovished in species and in the number of individuals, and is dominated by bivalves. Poorly preserved Edmondia sp. is common, as broken, disarticulated valves, distributed sporadi- cally throughout the calcite-cemented sand- stones. Oriocrassatella sp. is locally abundant, as disarticulated, and often broken, valves. Other bivalves present are Schizodus cf. kennedyensis and ?Nuculopsis sp. The gastropods Warthia sp. and Mourlonia sp. are also present. Some phosphatic cylindrical fragments up to 6 cm long were collected, and tentatively identified as frag- ments of arthropods. One Stenopora sp. was seen in thin section. Stenopora sp. has not previously been reported from the Billidee For- mation. Silicified fossil wood {? Araucaria) is common, and is distributed randomly throughout the sandstone units of the formation. Palynology. — A carbonaceous shale sample (University of Western Australia sample no. 65501) yielded a rich but not very diverse nor well-preserved assemblage dominated by Sulcati- sporites potoniei (Lakhanpal, Sah and Dube). The other pollen grains are mainly disaccate, nonstriate or monosaccate. The forms present are: Sulcatisporites potoniei (Lakhanpal. Sah and Dube) Leiotriletes directus (Balme and Hennelly) Protohaploxypinus limpidus (Balme and Hennelly) P. sp. cf. P. amplus (Balme and Hennelly) Platysaccus leschiki Hart Microhaculispora tentula Tiwari Acanthotriletes tereteangulatus (Balme and Hen- nelly) Limitisporites sp. Florinites eremus (Balme and Hennelly) Apiculatisporis sp. Cordaitina janakii Potonie and Sah Neoraistrickia ramosa (Balme and Hennelly) Veryhachium trispinosum Deunff. Balme (personal communication) states that this assemblage is typical of early Artinskian coal-bearing sediments from the Perth and Collie Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September 1976. 51 Basinb. It is identical with those from the Irwin River Cnal Measures ^ northern Perth Basin) and the lower coals at Collie. In the Carnarvon Basin, similar assemblages are seen in the Coyrie Formation and also the Keogh and One Gum Formations f formations of the Wooramel Group). The assemblage is unlike the synchronous Poole Sandstone of the Canning Basin, which contains a diverse pteridophyte suite and a more varied gymnosperm pollen, leading Balm? to believe that the Billidee Formation was probably deposited in a cooler climate than the Poole Sandstone. Spores and pollen grains completely dominate microplankton, with only one specimen of V. trispinosum observed. Environment of deposition From a study of field relations, petrology, sedi- mentary structures, fauna and flora it is possible to assemble the following evidence concerning the environment of deposition of the Billidee Formation: (a) Water Depth: The rapid horizontal and vertical variation in lithofacies indicates rapidly changing conditions during deposition. Such conditions would only be expected from a shallow water environment: deep water environments produce much more monotonous and laterally persistent lithofacies. The dominance of spores over microplankton suggests nearshore, thus probably shallow water, conditions. The association of probable root casts with dessication cracks suggests that shallow water plants, subaerially exposed at times, were present during deposition. Shale clasts also indicate subaerial exposure or nearshore conditions 'Conybeare and Crook 1968). The Edmondia band marks a change to slightly deeper water conditions, with the marine influence on sedimentation beceming more noticeable. (b) Currents: Essentially unidir'ctional cur- rents, towards the north, were responsible for the deposition of the cross stratified sands. Tidal currents were probably not significant. Current strengths were strong enough to transport and break bivalve valves, and are estimated at about 30 cm/sec, using the chart of Sunborg a956, quoted in Lauff 1967). Currents were probably responsible for winnowing much of the clay from the sands. During deposition of the carbona- ceous shales, currents were slight, because anaerobic conditions could only be maintained with minimal circulation. Dickins (1963) believes that OriocrassateUa sp. favoured a silt-free environment. Apart from this, however, no evidence is available concerning water turbulence. (c) Biota: Conditions favourable for abundant organisms were not present during deposition of the lower part of the formation, although some broken shell material is believed to have been present on deposition, and subsequently des- troyed by pressure solution. During deposition of the Edmondia band, con- ditions were favourable for the deposition of “shell banks”, as Dickins (1963) believes this was the life habit of OriocrassateUa. The growth of shell banks indicates a favourable combination of salinity, temperature, substrate, etc. Con- ditions were also favourable, during deposition of the Edmondia band for boring organisms. The reason why abundant brachiopods and bryozoans are not present in the Billidee Forma- tion is not clear, but Dickins (1963) states, with- out reasons, that abundant brachiopods and bryozoans are almost never found together with the molluscan association present in the forma- tion. The close association of probable root casts and marine fossils indicates that the plants may have been halophytes, possibly filling the same ecological niche (shallow brackish-marine water with some subaerial exposure) as modern man- groves. Abundant pebbles, up to 14 cm in diameter, are scattered randomly throughout the Edmondia band, in fine- to medium-grained calcite- cemented sandstones. The scattered distribution of these pebbles cannot be explained by “normal” sedimentation, and probably burrowing organ- isms were responsible for reworking coarse- grained beds and distributing the pebbles evenly throughout the member. This same process may also have been responsible for destroying primary lamination, as outlined by Ginsburg (1957). '7 .K / - ' - . . . -j ,-r ,- -, vf- •- 'i . - : - . . V. :-r-' -. , ', . . .,■ 5, I . \ - ^ V- ■"•-■* '■'■ 7* ‘ i ' '' ' ',1 I iRv^'' •/• i •^'«'' *■-■7 - * • •'•. ■ ' ■■ ••;■■ '•■ vV h- 3'^'’ y f y « - I -iB* - -f7?z . 7; 4. 'ft'* ^ i , ‘ r'y( '■y - .•'V,. '/■^l >l’i''----' 7^‘‘>.'':y'y.'’.'t,. /•• ^v^^r'''-•‘••y '777‘^-'. l7 ' yiJT ■••' y -'‘•'•'■'’•:--V, , --y;’ s>sK.r>':y'-"’vyyyy-::;;.yyyi:;7iS:7jyy^ v^Q'itr The Royal Society of Western Australia Incorporated Constitution and Rules and Regulations Constitution 1. The scientific society known as “The Royal Society of Western Australia Incor- porated” (hereinafter referred to as “the Society”) has the objects purposes and powers hereinafter mentioned. Objects, purposes and powers 2. To promote and assist in the advance- ment of science. 3. To print and publish, or join in printing and publishing, one or more journals, periodi- cals, books, newsletters or other documents relating to science. 4. To promote, produce and hold, or join in promoting, producing and holding meetings, lectures and exhibitions in furtherance of any objects of the Society, 5. To found, subsidize or contribute to, for the benefit of any institutions or persons, scholarships, grants, prizes, or monetary or other rewards in connection with all or any of the objects or purposes of the Society. 6. To acquire by purchase, lease, exchange, hire, or by way of loan or otherwise, any real or personal property in furtherance of any objects of the Society. 7. To invest, sell, lease, hire, mortgage, charge, lend, surrender or otherwise dispose of, or deal with all or any part of the assets or property, real or personal, of the Society, or to borrow money and grant security therefor and to liquidate, redeem or discharge any obli- gations undertaken in furtherance of any objects of the Society. 8. To improve, develop or extend all or any of the property or rights of the Society, also to build, erect or alter any buildings or erections, and to furnish, fit up, and maintain the same and provide fittings, equipment and appliances. 9. To amalgamate, co-operate or affiliate with any other society, association or body having objects wholly or in part similar to those of this Society. 10. To protect or assist in protecting the interests or rights of any member or members of the Society. Rules and Regulations Membership 11. The Society shall consist of members divided into the classes hereinafter mentioned, and the members who are, have been or may be duly elected as members of such respective classes shall have the rights and privileges hereinafter specified as appertaining to such respective classes, subject always to the due and punctual payment of the annual subscription or other sum hereinafter provided to be paid by the respective members of each class, namely — (a) Ordinary Members: An Ordinary Member if financial shall have the following rights and privileges: — (i) To be present and to speak and vote at Ordinary, Special or Annual General Meetings of the Society, to vote in postal ballots con- ducted by the Society, and to attend excursions, meetings, lectures or other activities arranged from time to time by the Society. (ii) To be eligible for election as a member of the Council hereinafter referred to and also to any office or position in the Society. (iii) To submit to the Council for consideration for publication papers prepared by any such member on any scientific subject. (iv) To submit to the Council for considera- tion for publication on behalf of authors not resident in Western Australia, papers written by such authors on any scientific subject. (v) To receive the Journal of the Society. (vi) To receive other publications or docu- ments issued by the Society, upon such condi- tions as the Council may from time to time determine. (vii) To borrow books, periodicals, papers, or other documents belonging to the Society, sub- ject to the approval of the Council. (viii) To propose or second candidates for admission as Ordinary, Associate or Student Members of the Society. (ix) With the previous approval or assent of the Council or President, or duly appointed chairman or leader of any meeting held by the Society, to introduce visitors. Such visitors shall not be entitled to vote at any such meeting or excursion, but may express opinions on any matter under discussion at the invitation of the chairman or leader. (X) To propose or second any Honorary or financial Ordinary Member for election to Council. (b) Associate Members: Persons elected as Associate Members shall be entitled to attend and speak at general meetings, excursions and other activities conducted by the Society, but shall not be entitled to vote at any meeting, and if any such member should vote, then that vote shall not be counted. Associate Members shall be entitled to submit papers for publica- tion, to receive publications or documents issued by the Society at a price fixed by the Council from time to time and subject to the approval of the Council, may borrow books, periodicals or other documents belonging to the Society, Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 59 but shall not be entitled to be proposed for, or elected to, any office in the Society, nor to propose or second others for membership or office in the Society. (c) Student Members: The Council may, on application to be made each year in such form it may require, admit as Student Members, persons under the age of twenty five years who are undertaking formal studies in Western Australia. Student Members shall have the same rights and privileges, and shall be subject to the same restrictions, as Associate Members. (d) Honorary Members: The Society at its Annual General Meeting in any year may, on the recommendation of the Council, admit as Honorary Members persons distinguished in Science or as patrons thereof, but only in so far as the number of such members shall not at any time exceed twenty five. Honorary Members shall have the same rights and privi- leges as Ordinary Members, but without liability for any subscription. (e) Honorary Associate Members: The Society at its Annual General Meeting in any year may. on the recommendation of the Council, admit as Honorary Associate Members persons inter- ested in science, but only in so far as the num- ber of such members shall not exceed twenty five. Honorary Associate Members shall have the same rights and privileges, and be subject to the same restrictions, as Associate Members, but without liability for any subscription. (f) Corporate Members: Corporate membership may be made available by the Council to organizations under such terms and with such rights and privileges as the Council may deter- mine. 12. Every person desirous of becoming an Ordinary or Associate Member of the Society shall make application in the form prescribed from time to time by the Council and shall be proposed by an Ordinary or Honorary Member of the Society and seconded by at least two other Ordinary or Honorary Members of the Society, to each of whom the applicant is known personally. Each application shall be accom- panied by the subscription applicable to the class of membership sought and shall be lodged with one of the Secretaries. Each Ordinary and Honorary Member of the Society shall be informed of the application forthwith by such means as the Council may from time to time determine. The Council shall accept or reject each candidate at a meeting of the Council held not less than one month and not more than six months after issue of advice of the candidate’s application to the Ordinary and Honorary Members. Ordinary or Honorary Members wishing to object to admission of any candidate shall communicate their objection and the reasons for it to the President within one month of the issue of such advice. The President shall investigate any such objection and shall report the results of his investigation to the Council, but shall not reveal the identity of the Ordinary or Honorary Member lodging such objection. The Council shall consider the President’s report, and shall determine by vote to admit or not admit the candidate to the class of membership sought. Admission shall be made by a two thirds majority of those Council members present and voting. 13. The Secretary shall inform each candi- date of his admission or non-admission to the class of membership sought. Upon admission, each Ordinary or Associate Member shall receive a copy of the Society’s Constitution and Rules and Regulations. His membership shall be deemed to apply to the whole of the financial year in which he was admitted so that, subject to their availability, he shall be entitled to receive all publications and documents issued during that year to all members of his class. 14. Every member of the Society, of what- ever class, shall be bound to observe and per- form, and not commit any breach of, the Rules and Regulations of the Society from time to time in force. 15. Every member shall notify the Secretary in writing of the address to which he may desire notices or communications to be for- warded to him, and shall from time to time notify in writing any change in such address. If any member shall fail to give such notifica- tion he shall have no claim against the Society for publications or notices of the meetings or other activities of the Society; no meetings or other proceedings shall be invalidated by reason of any such member not having received such notice. 16. Any member, on paying to the Society all subscriptions or moneys owed by him and returning all books, papers, manuscripts, or property of the Society which may have been borrowed or received by him, may resign his membership by giving notice in writing to the Secretary of the Society; and any member ceasing by resignation, death, or otherwise to be a member of the Society shall not, nor shall his representatives have any claim upon or Interest in the funds or property of the Society; but nothing herein contained shall prejudice the right of the Society to recover any moneys owing or property of the Society borrowed, held, or received by such member at any time. 17. Any member whose subscription may be in arrears for a period of at least two years shall, on a resolution of the Council being at any time thereafter passed, be declared to be no longer a member and thereupon shall cease to have the rights and privileges to which he may have been entitled; provided always that nothing herein contained shall prejudice the right of the Society to recover from such mem- ber all moneys or subscriptions due, owing, or payable by him up to the date of such termina- tion of his membership, and also to recover all books, papers, manuscripts, or property belong- ing to the Society which may be held or have been received by such member at any time. 18. Every member in the respective classes of membership in the Society shall use his best efforts to promote the objects of the Society and shall not do or commit any act, deed, or Journal of the Royal Society of Western Australia, Vol. 59, Part 2, September, 1976. 60 thing which may be deemed by the Council to be prejudicial to the interests of the Society. 19. The Council may, by an affirmative vote of two thirds of its total membership, remove or suspend from membership or expel any mem- ber of the Society without being required to assign any reason for such action. Notice of such removal, suspension, or expulsion, shall be sent by registered post to the last known address of the member concerned within seven days after the decision of the Council. Any member against whom such decision of removal, suspension, or expulsion shall be made, shall be entitled to appeal to a Special General Meeting of the Society by notice to be for- warded by him in writing, addressed to the Secretary within two months after the date of such removal, suspension, or expulsion, stating in such notice the grounds of appeal. It shall be the duty of the Council to summon a Special General Meeting of the Society for the purpose of considering any such appeal, and of hearing statements by any Member of the Council or by the member who may have been removed, suspended or expelled, and if a majority of the members present at such meeting uphold the decision of the Council, then the decision of the Council shall be confirmed, but if such majority shall uphold the appeal, then the decision to remove, suspend, or expel such member shall be set aside. No such member shall be entitled to exercise such right of appeal after the expiration of the said period of two months. In the event of any such removal, suspension, or expulsion taking effect, the mem- ber concerned shall remain liable for all moneys or subscriptions due or payable by him as at the date of such removal, suspension, or expul- sion, and for the return of all property belong- ing to the Society. Subscriptions 20. The subscription for each ordinary mem- bership shall be set from time to time on recommendation of Council and approval by a resolution passed by a two-thirds majority of Ordinary Members voting at any Ordinary or Annual General Meeting of the Society, of which at least twenty eight days notice has been given, and in which notice the proposed alterations have been specified. An Ordinary Member, whose subscription is not in arrears, may at any time compound for the subscription for the current year and for all future years during the life of such member on payment of a fee of: Mg>Ca>K an(d that of the anions is Cl>HCO;j>SO.i. This orS 0.5 5 26.5 93 5 g/litre 10.8 0.4 0.4 1 .3 19 4 0.1 ”oeq/litre 77.1 1 .0 :).:3 18 90.2 U'igures for seawater taken from Parker (1972). Lake Table 5 yutrient leveli^ in Lake Le>>chenauUia and three other lake>t on the coastal plain for Mm/ 1975 Depth Nutrient (/xg/litre) 1 Si Tot P PO, P Org- P NH 3 X XO 3 -N 1 >-o.- Leschenaultia surface 274 10 8 ; •> 1 1 4 Iveschenaultia 1 6 m 280 20 8 12 12 5 ; ->n Monger 1 surface 21 <1 21 33 6 594 JcKuidalup surface 38 2 36 14 8 813 McNe.ss surface 7 i <1 7 4.1 01 .) 813 1 12 Journal of the Royal Society of Western Australia. Vol. 59. Part 3. February, 1977. 68 < ortho) phosphate determination were placed in polyethylene bottles previously soaked in a KI/ iodine mixture and determined by the single solution method (Major et al. 1972). Total phos- phorus was determined by the same method after perchloric acid digestion. Surface and depth readings obtained for Lake Leschenaultia during May 1975 are given in Table 5 along with levels found in three other lakes. Levels of phosphorus in Lake Leschenaultia were relatively low. Hutchinson (1957), in a review of a number of geographically-distinct lakes found in humid, temperate regions, noted uniform low concentration of total P in surface waters. Though within 80 km of Lake Leschen- aultia the three coastal lakes, used for compari- son, gave contrasting phosphorus levels. Lake Monger, a well established, shallow alkaline lake, sometimes subject to intense algal blooms and considered to be fairly eutrophic, gave total phosphorus readings during May which were similar to those for bottom water in Lake Lesch- enaultia. However, levels of phosphorus during this period were declining in Lake Monger and it is noteworthy that levels of 60-70 ^g/litre have been reported during this month in this lake in earlier years (Harris 1969). A similar trend was found for total phosphorus in Lake Jcondalup, considered to be mildly eutrophic (Congdon and McComb 1976). Closer agreement w^as found between the values for Lake Leschen- aultia and Loch McNess, the latter being the least disturbed and least productive of the coastal lakes. Inorganic phosphorus comprised a significant part of total phosphorus in Leschenaultia, being 80% of total surface water phosphorus. Much of the phosphorus utilised by phytoplankton in lakewater is in this fraction. From the large amounts of PO4-P relative to total P, it is sug- gested that low phytoplankton numbers, and thus little utilisation of the phosphate fraction, may be responsible. This may be further borne out from comparison with the coastal lakes, where reduced orthophosphate levels occur where algal population numbers are significant. Levels of organic phosphorus were higher at depth, though the orthophosphate fraction remained constant. Phytoplankton levels were minimal during May, and it can only be assumed that organic material was present at depth, derived from benthic or fringing plants. Nitrogen Inorganic nitrogen was determined by indivi- dually testing for ammonia, nitrate and nitrite. Ammonia was detected by the cyanurate method (Dal Pont et al. 1974). Nitrate was detected by an ultra-violet method which takes into account interference by organic nitrogen (Anon. 1971). The nitrite concentration was tested by a diazotisation method which involved coupling diazotised sulfanilic acid with naphthylamine hydrochloride (Anon. 1971). Ammonia was found in similar concentrations at both depths (Table 5), but differences occurred in concentration of nitrite and nitrate. The nitrate was more concentrated at the sur- face. (The lower value of 20 Mg/litre may not be entirely accurate as the ultra-violet technique has a recommended lower detection limit of 40//g/litre (Anon. 1971).) The level reported here is lower than that given by Shipway (1948) (10 mg NOs-N/litre). As no information on the technique used to measure the early sample is available it would be unwise to draw conclusions from the comparison, as many nitrate methods have been found to be unreliable (eg. Vollen- weider 1968). Nitrite was higher in the 6 m sample but the relatively minor role this element plays in the environment (Malhotra and Zanoni 1970) reduces the significance of this change. Using inorganic nitrogen values for Lakes Jcondalup, Monger and McNess (Table 5) it is possible to gain an indication of the relative trophic level of Lake Leschenaultia. The nitrite concentration places Leschenaultia between McNess and Jcondalup and very similar to Monger, but as this compound is so transient in the environment, little emphasis should be placed on this comparison. The nitrate found is considerably lower than levels found in the other three lakes. As the level of nitrate in oligotroph’c lakes rarely exceeds 1 mg/litre NO:)-N (Bayly and Williams 1973), it seems reasonable to classify Lake Leschenaultia as oligotrophic, at least on this one occasion, on this basis. An attempt was made to read nutrient levels, as total P and inorganic nitrogen, onto tables classifying lakes into their trophic status. Using the table of Sakamoto (1966) (cited Vollenweider 1968), Lake Leschenaultia can again be classified as oligotrophic. Thus, from a consideration of nutrient levels (as N, P and Si), phytoplankton populations and physical state of the water using information for other lakes in the area, and information from the literature as a guideline, it is suggested Lake Leschenaultia be designated oligotrophic. Acknowledgements . — We are indebted to Associate Professor A. J. McComb, of the Department of Botany, University of Western Australia, for his invaluable assistance in preparing this manuscript. We also grate- fully acknowledge Mr. R. Leggo. the Mundaring Shire Clerk, and Mr. J. W. Morris, the Publicity Officer of Westrail, for supplying information about the history of Lake Leschenaultia. References Anon. (1971).— “Standard Methods for the Examination of Water and Wastewater”, thirteenth edition. American Public Health Association. American Water Works Association and Water Pollution Control Federation. (Ameri- can Public Health Association: Washington, D.C.) Bayly, I. A. E. (1964). — Chemical and biological studies on some acidic lakes of East Australian sandy coastal lowlands. Aust. J. mar. Freshw. Res. 15; 56-72. Bayly, I. A. E. and Williams, W. D. (1964). — Chemical and biological observations on some volcanic lakes in the south-east of South Australia. Aust. J. mar. Freshw. Res. 15: 123-32. Bayly. I. A. E. and Williams, W. D. (1973).— “Inland Waters and their Ecology”. (Longman: Camberwell.) Congdon, R. A. and McComb, A. J, (1976). — The nutri- ents and plants of Lake Joondalup, a mildly eutrophic lake experiencing large seasonal changes in volume. J. Roy. Soc. W.A. 59: 14-23. Dal Pont, G., Hogan, M.. and Newell, B. (1974), — “Laboratory Techniques in Marine Chem- istry”, C.S.I.R.O.. Division of Fisheries and Oceanography Report No. 55. Journal of the Royal Society of Western Australia, Vol. 59, Part 3. February, 1977. 69 Dimmock. G. M.. Bettenay, E., and Mulcahy, M. J. (1974). — Salt content of lateritic profiles in the Darling Range, Western Australia. Aust. J. Soil Res. 12; 63-69. Gentilli, J. (1948) .—Micro-geography of Lake Leschen- aultia. W. Aust. Nat. 1: 109-110. Harris, P. L. (1969). — “Phytoplankton Ecology of Lake Monger (North Perth)”. Dissertation pre- sented to the University of Western Aus- tralia for the degree of Bachelor of Science with Honours. Hutchinson, G. E. (1957). — “A Treatise on Limnology. Vol. 1. Geography, Physics and Chemistry.” (John Wiley and Sons: New York.) Mackereth, F. j. (1953). — Phosphorus utilization by Asterionella formosa Hass. J. Exv. Bot. 4- 296-313. Maholtra. S. K. and Zanoni, A. E. (1970).— Chloride interference in nitrate nitrogen determina- tion. J. Amer. Wat. Works. Assoc. 62: 568-71. Major, G. A., Dal Pont, G.. Klye, J.. and Newell B (1972).— “Techniques in Marine Chemistry— A Manual”, C.S.I.R.O., Division of Fisheries and Oceanography Report No. 51. McArthur, W. M. and Bettenay, E. (1960). — “The Development and Distribution of the Soils of the Swan Coastal Plain, Western Aus- tralia.” C.S.I.R.O., Australian Soils Publica- tion No. 16. McComb, J. A. and McComb, A. J. (1967). — A prelimin- ary account of the vegetation of Loch McNess, a swamp and fen formation in Western Australia. J. Roy. See. W.A. 50- 105-12. Mortimer. C. H. (1941-1942).— The exchange of dis- solved substances between mud and water in lakes. J. Ecol. 29: 280-329, 30: 147-201. Parker, C. R, (1972). — “Water Analysis by Atomic Absorption Spectroscopy.” (Varian Techtron Pty. Ltd.: Springvale, Victoria.) Peck. A. J. and Hurle, D. H. (1973).— Chloride balance of some farmed and forested catchments in south-western Australia. Wat. Resour. Res. 9: 648-57. Richards, F. A. with Thompson, T. G. (1952).— The estimation and characterization of plank- ton populations by pigment analyses. II. A spectrophotometric method for the esti- mation of plankton pigments. J. Mar. Res. 11: 156-72. Riggert, T. L. (1966).— “A Study of the Wetlands of the Swan Coastal Plain”, Department of Fish- eries and Fauna, Western Australia. Shipway, B. (1948). — Reports on excursions — Lake Leschenaultia. W. Aust. Nat. 1: 107-8. Vollenweider, R. A. (1968). — “Scientific Fundamentals of the Eutrophication of Lakes and Flowing Waters with Particular Reference to Nitro- gen and PhoF.nhorus as Factors.” (Organi- zation for Economic Co-operation and Development: Paris. France.) Welch, P. S. (1948).— “Limnological Methods.” (McGraw- Hill: New York.) Journal of the Royal Society of Western Australia. Vcl. 59, Part 3. February, 1977. 70 10. — Poison plants in Western Australia and colonizer problem solving by J. M. R. Cameron* Communicated by N. C. N. Stephenson Manuscript received 20 April 1976; accepted 22 June 1976 Abstract Because of their leguminous seed pods, poison- ous plants of the genera Gastrolobium and Oxylobium posed major problems for Western Australia’s nascent pastoral industry. Not only did they cause considerable economic loss, but it took nearly twelve years for their lethal pro- perties to be recognised. This delay is attributed here to the nature of these plants and to the confusion generated among colonists by the conflicting explanations offered by Dr Joseph Harris and the botanist James Drummond. The process by which the toxic nature of these plants was established raises important theoreti- cal implications about problem solving and hazard response in unfamiliar environments and throws light on colonizer adjustment. It is concluded that learning can be viewed as a progression along a generalization-differentia- tion continuum where generalization denotes the constraining effect of well established past be- haviours and differentiation refers to a growing sensitivity and responsiveness to previously un- familiar environmental stimuli. Introduction Any relocation in space will induce a period of active adjustment. Nowhere is this more evident, or more crucial than in pioneering situations. Here, people with diverse back- grounds, predispositions and expectations in- variably encounter grossly unfamiliar and often extremely forbidding conditions which contain few of the elements which gave their former behaviour structure and cohesion. Yet, with the exception of Found’s exploratory theorizing of incremental learning (Found 1971, p.139-141), colonizer adjustment has received little atten- tion. No attempt is made here to examine all types of learning evident in colonizer adjustment. Rather, the emphasis is placed on the problem solving activities associated with Western Aus- tralia’s poison plants for these activities repre- sented a deliberate attempt to decrease the disparity between the expected and actual out- comes of pastoral operations. This is a unique exarnple^ but does have important theoretical implications. The paper is therefore structured into three parts which focus respectively on the nature of the problem, the sequence of events leading to its resolution, and the characteristics of problem solving behaviour evident in this sequence. The problem in its context Western Australia has more than 150 endemic plants capable of poisoning stock under some circumstances (Gardner and Bennetts 1956). Of these, the 32 toxic species of the genera 1 Department of Geography, University of New England Armidale. New South Wales. 2351. Oxylobium and Gastrolobium are the most wide- spread and lethal, less than 15 g being at times sufficient to kill an adult sheep (Aplin 1967). As partly indicated in Table 1, they constituted the major hazard for the nascent pastoral in- dustry, losses from them exceeding combined losses from all other hazards including bush- fires, floods, drought, aboriginal depredations and the attacks of native dogs. Stock losses were recorded as early as December 1830 and continued at a high level throughout the 1830s. They created the conditions of stress necessary for inducing accelerated learning and active problem solving (Festinger 1964; Heider 1958; Lewin 1938) but the cause of death was not positively confirmed until May 1841 {Inquirer, 26 May 1841; Perth Gazette (hereafter P.G.). 17 May 1841). Two factors, the nature of the problem, and the nature of the problem solvers, account for this slow resolution. The nature of the problem A recognition of the fact that the genera Oxylobium and Gastrolobium are members of the pea flowered family (Papilionaceae) is funda- mental to understanding delays in identification. Shepherds believed their leguminous structure placed them among the more nutritious local fodder sources and deliberately sought them out. As the plants had not been previously encount- ered in Australia, there could be no warning of their lethal properties. To add further confusion, only 15 of the 59 species recognised by 1864 (Bentham 1864. p. 14-26, 96-207) have since been found to be toxic (Gardner and Bennetts 1956, p. 52-75). All species can be eaten, their palat- ability being greatest in the winter and spring months when toxicity is at a peak. These fea- tures alone sufficiently explain delays but they were exacerbated by the nature and effect of the toxic element. The toxic agent of Oxylobium and Gastrolo- bium is monofluoroacetic acid, found elsewhere only in the gifblaar { Dichapetalum cymosum) of South Africa and the gigyea (Acacia georgina) of Queensland and the Northern Territory (Aplin 1967). More commonly known as the rabbit poison T080’ (its sodium salt derivative), fluoro- acetic acid is odourless, colourless, tasteless, water soluble and extremely stable. When in- gested, it converts by enzyme action into the toxic fluorocitric acid (Peters 1954). As shown in Table 2, only small dosages are required to produce fatal results. There is no known anti- dote. As its presence in nature was not demon- trated until 1943, settlers could not know of its existence. Journal of the Royal Society of Western Australia. Vol. 59. Part 71 3, February, 1977. Table 1 Recorded stock losses from poison plants. 1833-1840 Date reeofded Details of mortality Location Remarks (PG) 21 September 1833 Heavy stock losses: upwards of 100 sheep from 1 Upper Swan flock. (PG) 19 October 1833 4 out of 6 bullocks. Dog eating one of the bull- A'ork Uoad ocks later died. (PG) 15 March 1834 21 sheep Up])cr Swan Scarp face (SD) 17 March 18:}4 300+ sheep. 5 cattle. 3 horses .. King George Sound (PG) 16 Aufiust 18:14 40 sheep. 6 goats . .. Upper Swan (PG) 11 April 1835 55 out of 66 goats York Road Very drv season with the result (PG) 16 May 18:35 93 sheep. 13 goats, 6 bullocks York Road that many flocks were moved to (PG) 20 June 1835 15 sheep York Road i-the inland pastures of the Avon (PG) 14 November 1835 8 out of 10 bullocks York Valiev. Total sheep numbers were then less than 4 000. (PG) 26 November 1836 .... i:30 sheep out of :300 Williams district (PG) 4 August 1838 56 sheep out of 2:)9 . . Williams district \ Exceptionally dry season which (PG) 20 November 1838 . . Major losses of cattle Williams district / persisted until the winter of 1842 (PG) 7 July 1839 Major losses of all stock throughout the colony. (PG) 28 March 1840 250 sheep and 12 cattle out of 750 sheep and 54 Kojonup district I Mortality associated with the cattle. overland movement of imported (PG) 27 June 1840 63 shee]) out of 600. Other losses were even Road from King George >stock from King George Sound greater. Sound to the Avon to the Avon Valiev. Valiev (PG) H) October 1840 33 out of 180 shecj) Upper Swan - Scarp face (INQ) 7 October 1840 . . Total flock of 180 .sheep ., York (PG) 5 December 1840 . All of Graigie’s flock (c. 500) Kojonup district 304 of Tapson’s flock 118 of MacDonald’s flock Source: Perth Gazette (shown as PG); Inquirer (INQ); Spencer Diary (SD). Table 2 Toxiritu of sodium finoroacetate (;')0% morfaliti/) Animal Dosage (a) (mg per kg of lH)dy weight) Dog ,, 0.066 Sheej) 0 . 25 Pig 1.0 Horse 1.0 Kangaroo 8.0 (b) Pigeon 9.0 Frog , . 300.0 (a) All dosages wei’e administered orally. (h) This tiKure is for 100 per cent mortality. Dosage for 50 per cent mortality is not available. Source: Department of Agriculture. Western Australia, n.d.. Sodiinn finoroacetate: A resnme of the available, literature dealimj ivith the poisonimj of human beings and animals (mimeod). Most English poisonous plants, by contrast, are cyanogenetic, or contain poisonous, some- times narcotic, alkaloids. With the exception of foxglove (Digitalis purpurea), their poisonous qualities were suggested through their acrid taste and the foetid smell of their sap and bruised leaves. These characteristics settlers were well aware of and it is not surprising that attention first centred on a number of cyanogenetic plants, particularly the blind grasses ( Stypandra imbri- cata and S. grandifiora) which induce a range of symptoms similar to those induced by Oxylobium and Gastrolobium (Gardner and Bennetts 1956, p. 79-82), and the narcotic Woodbridge poison (Isotoma hypocrateriformis) , a member of the lobelia family. The rarity of fiuoroacetic acid naturally is matched by its variable toxicity, variations de- pending on the species, its location, and the time of the year (Aplin 1967). These variations, in turn, induce a wide range of symptoms from mild agitation, partial paralysis and blindness, to a violent, convulsive death (Came, Gardner and Bennetts 1926). This feature most perplexed colonists because several conditions closely cor- responded to already known stock diseases, par- ticularly ‘hoove’, ‘staggers’, and ‘blood striking’, all of which were caused by an inability to digest over-rich herbage (see Wilson 1852, v. 2, p. 684-6). In addition, not all animals were equally affected (Table 2). Sheep and goats appeared most susceptible, followed by cattle. Horses had a high resistance (Landor 1847, p. 379-380) which was enhanced by their greater body weight. The relative immunity of local fauna was most perplexing. The botanist James Drummond consistently refused to accept that York Road poison (G. calycinuvu had poisonous properties for he had observed pigeons feed on the seeds of the plant with immunity (P.G., 5 December 1840). Problew. solvers While pastoralists were extremely anxious to identify the cause of mortality, and while they were deeply involved in the poison debate, they generally deferred to the superior knowledge and expertise of Joseph Harris, a surgeon with veterinary experience, and James Drummond. Similarly, the Agricultural Society took little direct action although it did provide a forum for discussion and debate. As a consequence, although their initial expertise was only margin- ally superior, Drummond and Harris emerged as the major authorities and the key problem solvers. That they held strongly divergent views greatly added to the confusion surrounding stock deaths. Harris was convinced that stock died from ‘blood striking’ (P.G., 21 September 1833). His conclusions were derived from detailed patho- Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February. 1977. 72 logical evidence and his knowledge of English stock diseases but there were other considera- tions. Apart from being Drummond’s rival for prestige and influence, Harris was concerned that the colony’s already bad reputation would be further degraded by reports of high stock fatalities. This attitude, first expressed in Sep- tember 1833 (P.G., 2 September 1833), was maintained throughout the 1830s with the result that unfavourable evidence was unconsciously suppressed. Drummond, by contrast, early suspected plants because of the similarity of the vegeta- tion associations in localities where deaths were reported (P.G., 19 October 1833; 5 December 1833). Drummond’s approach involved the iso- lation, identification, description and testing of suspected species . When he arrived at Williams, Drummond was contacted by Harris’ son who was extremely perturbed by the continuing fatalities in his flock. Close examination of the area where the sheep had been feeding revealed the presence of York Road poison. The evidence now seemed overwhelming for, not only was this poison common throughout the Kojonup District, but Drummond had earlier observed it alongside the York Road where fatalities had been particularly heavy. Tender young branches were collected, made into a drench and given to a healthy gcat. It died within 14 hours (P.G., 21 November 1840, 5 December 1840). The matter was not yet resolved, however. Harris reluctantly accepted his son’s and Drum- mond’s evidence but refused to accept that more than one species was involved or that York Road poison was widely distributed, (P.G.. 19 Decem- ber 1840). His view was influenced by a desire to defend the pastoral properties of the Kojonup District. In this he had several supporters in- cluding the German botanist Preiss, recently arrived in the colony, who insisted that “legum- inous plants are particularly suited for the food of animals and the human race”. To prove his point, he drank a wineglass of diluted fluid ex- tracted from the leaves of York Road poison (Inquirer, 17 March 1841). Suffering no ill- effects, he recommended the plant to stock holders as “the very best thing they could cultivate as artificial food for stock” (quoted in Erickson 1969, p.59). To resolve all doubts and end the confusion, the Agricultural Society requested Drummond in May 1841 to demonstrate the toxic properties of the plants he had identified. All three test animals died within four hours. Three dogs also died after feeding on the entrails of the poisoned sheep. Harris was now fully convinced Journal of the Royal Society of Western Australia. Vol. 59, Part 3. February. 1977. 74 and agreed that “the plant is a most powerful poison”. He knew no antidote (Inquirer, 19 May 1841; P.G., 17 May 1841). Extended learning Further experiments on 13 August 1841 con- firmed these findings and led to the search for an antidote (P.G., 14 August 1841). Within a week it was reported that washing soda may be beneficial (P.G., 21 August 1841). A considerable hiatus followed which suggests that identifica- tion of the cause substantially reduced settlers’ concern (see P.G., 22 January 1842). A repetition of the May 1841 experiment twelve months later produced negative results and led Drummond to conclude that toxicity was greatest from the beginning of new growth until the onset of the summer drought. January through to May could be considered the quiescent period (Inquirer, 18 May 1842). The range of poison plants was now extended to include rock poison (G. callistachysj and box poison (O. parviflorum) ‘P.G., 14 May 1842). Heart-leaved poison (G. hilohum) growing in the south and particularly around King George Sound was already under suspicion iinquirer, 23 December 1840; P.G., 5 December 1840) . With Drummond’s positive identification, fatalities rapidly diminished and it became com- mon practice for shepherds to carry branches of known or suspected species to aid in their identification. Where belts of poison had to be traversed, stock were driven in haste and were frequently muzzled. Main roads were gradually cleared of poison to a distance of a chain (20 m) on each side (Erickson 1969, p. 60). When moving stock into unfamiliar areas, it became usual to send out scouting parties to identify and mark poison outcrops (P.G., 8 October 1842). On es- tablished properties, aborigines were temporarily employed in grubbing out poison plants in ex- change for a ration of tobacco and flour. The rate at which toxic species in areas set- tled by 1850 were identified, summarized in Table 3, is particularly revealing. The obvious conclusion to be made is that the commonest and most toxic species were among the first to be identified. River poison (G. forrestii) , the major exception, has a restricted location on the rivers of the south coast. Of these, only the Kent and Hay Rivers were stocked before 1850 and then by small flocks grazed mainly on the upper reaches. Similar conclusions may be made for Stirling Range (G. velutinum) , hook-point (G. hamulosum) and berry (G. parvifolium) poisons, all of which grow in rugged, hilly country or poorly grassed sandplains. The late discovery of the low toxicity poisons is almost self explana- tory, but, in addition, crinkle leaf (G. vVlosum) and runner (G. ovalifolium) poisons grow along Table 3 Toxic Species o/Oxy!obiura atul Gastrolobium present in areas occitpied by 1850 Species Toxic category JIaximum (a) Date toxic Date of first detailed toxic reading (ppm) effect determined description (pre 1864) Box poison .... 2 500 1842 1841 {(). parvifforum) Heart-Leaf poison 2 650 1841 1829 (G. hilohum) Rock poison High level of toxicitv 1 000 1842 1844 (G. callistachys) (more than 1 000 ppm) Cluster poison 1 350 1841(b) (G. benneUsianum) River poison 1 200 1926 (G. Forrestii) Champion Bav poison 1 050 1841(b) 1841 (G. oxylobioides) Sandplain poison 600 1841(b) 1839 (G. microcarpum) York Road poison 400 1841(b) 1841 (G. calycinum) Prickly poison 400 1901 1841 (G. spinoHum) Berrv poison Medium level of toxicity 300 1910 1841 (G. parvifolium) (more than 100 ppm) Gilbernine poison 150 1841(b) 1843 (G. rotundifolium Hook-Point poison 100 1920 1843 (G. hamulosum) Stirling Range poison 300 1921 1853 (G. velutinum) Woollv poison n.a. 1955 (G. lomentosum) Crinkle-Leaf poison 11. a. 1900 1841 (G. villosum) Runner poison Low toxicity n.a. 1910 1841 (G. ovalifolium) (less than 30 ppm) Bullock poison n.a. 1921 1841 (G. trilobum) («a) Parts per million of the toxic radical in terms of air dried plant material. (b) No real distinction was made between Champion Bay. York Road, Sandplain, Cluster and Gilbernine poisons because of their basic similarity. The toxic effect of all was effectively demonstrated by Drummond’s experiments with Champion Bay poison (G. oxylobioides) in May 1841. Source: Bentham 1864, v.2, p. 14-26 and 96-107; Gardner and Bennetts 1956, p. 40-77; Aplin 1967-71; Erickson 1969. p. 61-122. Journal of the Royal Society of Western Australia, Vol. 59, Part 3. February, 1977. 75 the ground. Both were suspected in the early 1840’s (Landor 1847, p. 380), but, perhaps be- cause of their prostrate growth form, these sus- picions were rejected. The thorny, unpalatable hardness of the leaves of both prickly and bul- lock poison fG. trilohumi explain their late identification as poisonous plants. The second and more important conclusion is that there was no general knowledge transfer from one harmful species to others within the genera. Otherwise, all presently known toxic plants would have been quickly classified for, with the exception of crinkle leaf and runner poisons, all toxic species are easily recognised by the similarity of their growth form, leaf arrange- ment, and the shape, colour and arrangement of the flowers. Certainly cluster poison, Champion Bay poison, sandplain poison and gilbernine poi- son were identified when encountered but their similarity with York Road poison is very close indeed, clear distinctions only being possible when the plants are in flower. That transfer was minimal indicates that no general framework for identification had yet been determined. To this extent, learning was still at a trial-and-error stage. It is possible, however, that this is indicative of an unwilling- ness to extend the range of harmful plants, and reflects the attitudes earlier adopted by Harris and others. Learning characteristics Three features of the learning sequence above are of particular significance in understanding the colonization of unfamiliar environments, namely: the role of problem solvers in a pioneer- ing community: the relationship between a problem's difficulty and the rate at which it is resolved; and the characteristics of problem solving in pioneering situations.. The role of problem solvers Although Curti (1957, p. 417-440) has demon- strated that individuals rather than formalized institutions are the key problem solvers in pion- eering communities, neither Drummond nor Harris initially knew more about the poison problem than other colonists. They developed their competence through community pressure. Three factors account for their eventual domi- nance: they had superior knowledge in appro- priate areas; they were able to isolate possible causes and solutions, and, as fellow pastoralists, they were viewed as status equivalents. They had credibility because they also suffered severe stock losses and were able to transmit their conclusions through informal channels. The deference of pastoralists was not complete, how- ever, for they clearly reserved the right to ques- tion conclusions when these differed from their own views. Drummond and Harris were used to isolate causes and solutions but pastoralists were the final arbiters. This process minimized their own efforts while maximizing Drummond’s and Harris’ expertise and, as such, is an impor- tant modification of Found's observation that 'man attempts to optimize some utility while minimizing his own effort’ (Found 1971, p.l29). The time factor in problem solving The learning sequence indicates that the tempo of effective learning is not continuous but does accelerate up to the point where a solution is determined. This is a basic characteristic of all problem solving for much initial learning centres on definition and this is often surrounded by considerable confusion. Following definition, problem solvers are able to formulate and test hypotheses and are in a position to reject un- satisfactory solutions or reinforce those which seem suitable. Although new hypotheses may be added, there is a progressive decrease of pos- sible alternatives. With resolution, there is a rapid diversification leading to both the solution of attendant problems and the development of compatible operational procedures. It must be emphasised, however, that problems can be ef- fectively avoided before solutions are defined. This must seriously complicate attempts to identify adaptive processes for what may often be considered problem solving is, in reality, avoidance. That both are in response to stress levels adds a further dimension. Colonizer problem solving Problem solving is far from simple. Gagne <1970, p. 36-69, 214-236) sees it as the apex of an eight-layered hierarchy which ranges from signal learning (essentially Pavlovian condition- ing) at the base through stimulus-response con- nections where the learner acquires a precise response to one or more discriminated stimuli to concept development (common response to a general class of stimuli) and rule or principle learning where causal links between stimuli are clearly established. Problem solving requires competence in most of these learning types. While the Gagne model identifies the stages of most learning situations, not all levels are ap- plicable to the poison problem. The key prob- lem solvers already had well developed learning patterns including single and multiple discrimi- nations, chaining and associations, and, in the case of Harris, clearly established rules regard- ing the relationship between vegetation and stock deaths. What was critical was that these lower orders of the learning hierarchy had been developed in another and quite dissimilar en- vironment and were thus inadequate. Because of this inadequacy, they had to be transformed or, more correctly, extinguished and replaced until they were compatible with the new en- vironment. This was a major inhibitor of the learning rate. The process of transformation may best be discussed by reference to the Lewinian concept of ‘life space’ (Lewin 1946, p.239-40). Lewin views behaviour as a function of the life space which is the product of a person’s interaction with his environment and is essentially the known physical and psychological environment of that person as derived by his whole pattern of behaviour in a specific environmental setting. As the life space is highly structured, behavi- oural directions are well defined. When reloca- tion occurs, a new and unstructured life space is imposed because there has been a complete change in the environmental component. The Journal of the Royal Society of Western Australia, Vol. 59, Part 3. February, 1977. 76 individual is initially confused because of the lack of definite orientations but he institutes a learning sequence whose purpose is the differen- tiation of unstructured areas in the new life space (Lewin, 1942). A similar viewpoint is expressed by Gerritz (1969) in his discussion of the effects arising from gross shifts in the maintaining environ- ment. When the maintaining environment (analogous to Lewin’s life space) is changed, the individual ‘will bring to a new environmental setting . . . behavioural systems that have been maintained (and possibly acquired on the basis of) the stimuli in the setting from which he had come. It follows . . . that initial behaviour in response to the stimuli in the new setting will be a function of the similarity of those stimuli to the stimuli that controlled behaviour in the earlier context’ (Gerritz 1969, p.ll9). If the new context is markedly dissimilar, severe constraints will be imposed on behaviour until the behaviour that does take place provides ‘the basis for a new adaptive learning in con- nection with the stimuli available in the new setting’. Both Lewin and Geiritz hold the view that relocation induces confusion and inhibits res- ponsiveness to environmental stimuli. Yet, it is quite apparent that initial responses in Western Australia were quite pronounced and were ac- companied by a degree of confidence that later responses did not have. Reference to Harris’ assertion that stock died from ‘blood striking’ effectively demonstrates this point. That this was erroneous is irrelevant as it was not seen to be so. This example suggests that, in pion- eering situations at least, differentiation is pre- ceded by another process which can be termed generalization. Generalization may be viewed as the process of interpreting and reacting to newly encoun- tered environmental stimuli from the stand- point of already learned behaviour. Its dominant characteristic is the blanket application of well defined rules formed from past experiences as these are expressed through preconceptions. As such, it is essentially a function of factors in- ternal to the individual. As well as established knowledge and preconceptions, these factors will include motives, goals and expectations. There is little observable interaction with (cognitive responsiveness to) external stimuli. It is only when blanket generalizations prove unsatisfac- tory that the search for the appropriate rela- tionships between stimuli (differentiation) begins. As suggested by the Gerritz quote above, dif- ferentiation refers to the process where specific relationships between environmental stimuli, both internal and external to the individual, are determined. It must be seen as a function of actions centred on the resolution of specific problems or the definition of particular be- havioural orientations, and clearly involves an internal (nature of the problem solver) external (nature of the problem) dichotomy. That is. differentiation requires interaction between the problem solver and those environmental stimuli relating to the problem, the amount of interac- tion and hence differentiation increasing as learning progresses towards the point of ulti- mate resolution. This cannot be viewed as a purely perceptual or even more broadly cogni- tive process for affective components are also influential. As evidenced by Drummond’s atti- tudes towards legumes, these are frequently in- hibiting. Confidence levels may also inhibit (Found 1971, p.l34). It is also apparent that when stress reaches a critical threshold all dif- ferentiation ceases and behaviour is charac- terized by avoidance. From the preceding discussion, it is reasonable to postulate that all learning in colonizing situ- ations progresses from the point where blanket generalizations are made to increasingly selec- tive differentiation, both processes being em- bedded in a cognitive-emotional-motivational matrix. That no true separation of these com- ponents is possible is the problem facing all analyses of adaptive learning. References Aplin, T. E. H. (1967) — The Poison Plants of Western Australia, Journal of Agriculture of Western Australia 8: 42-52. Bentham, G. (1864). — "Flora Australiensis” . Reeve, Lon- don. Came, W. M.. Gardner. C. A. and Bennetts, H. W. (1926) — "The Poison Plants of Western Aus- tralia”. Department of Agriculture, Perth. Curti, M. (1959) — "The Making of an American Com- munity: A Case Study of Democracy in a Frontier Society”. University of California Press, Stanford. Erickson, R. (1969) — "The Drummonds of Hawthorn- den”. Lamb Paterson, Perth. Festinger, L. (1964) — "Conflict, Decision and Dissonance”. University of California Press, Stanford. Found, W. C. (1971) — "A Theoretical Approach to Rural Land-Use Patterns”. Edward Arnold, London. Gagne, R. M. (1970) — "The Conditions of Learning”. Holt, Rinehart and Wilson, New York. Gardner, C. A. and Bennetts, H. W. (1956). — "The Toxic Plants of Western Australia” . Western Aus- tralian Newspapers, Perth. Gerritz, J. L. (1969). — Mechanisms of Social Learning: Some Roles of Stimulation and Behaviour in Early Human Development, in David L. Gos- lin (ed.): "Handbook of Socialization Theory and Research” . Rand McNally, New York. Heider, F. (1958) — "The Psychology of Interpersonal Relationships” . John Wiley, New York. Inquirer. Landor, E. W, (1847) — "The Bushman, or Life in a New Country”. Bentley, London. Lewin, K. (1938) — The Conceptual Representation and Measurement of Psychological Forces. "Con- tribution to Psychological Theory” 1. Lewin. K. (1942) — Field Theory and Learning, in Na- tional Society for the Study of Education: "The Forty-First Yearbook, Part II: The Psychology of Learning”. National Society for the Study of Learning. Chicago. Lewin, K. (1946) — Behaviour and Development as a Function of the Total Situation, in Dorwin Cartwright (ed.) 1951: "Field Theory in Social Science”. Harper and Row, New York. Moore, G. F. (1884) — "Diary of Ten Years in Western Australia”. Walbrook, London. Perth Gazette. Peters. R. (1954) — Biochemical Light Upon an Ancient Poison: A Lethal Synthesis. Endeavour 12, 147-154. Wilson, V. (1852) — "Rural Cyclopaedia”. (4 vols.). Pul- larton, Edinburgh. Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February, 1977. 77’ • ' 1 : * T ' J > IV' ..'a-, . V.' • '. f*'* -•V . ' V: - ’ ' ^ >'!' • •► . T ■ r- ■ r - : f??;: ^ ’? ■ I* V; .■■i'V,v,vi-X. ;-^-:i^,>'p'v''^: ‘ ■ .-'-• •> v’V'V. >V' .'. ••• •. -. f - •-■ j, ■ • ‘ ,i_ ■.'.'•••'• ■ - - ■ f. . • \. • ■ I,' 'I ' •a ■ 11. — Foraminifera of Hardy Inlet, southwestern Australia by Patrick G. Quilty^ Manuscript received 20 April 1976; accepted 22 June 1976 .Abstract Total fcraminiferal faunas were examined from 18 samples from Hardy Inlet and a nearby beach as part of a much broader study of the Black- wood River estuary under the auspices of the Department of Conservation and Environment. Several regimes can be identified on the basis of foraminifera and these correspond closely with geomorphic and hydrological regimes. Faunas in the upstream part of the cidal river regime are dominated by Ammonia beccarii although diversity and foraminiferal number are low. The delta faunas and those from the tidal river near the delta are dominated by agglutin- ated forms, although again, diversity and fora- miniferal number generally are low. Lagoon faunas are dominated by species of Ammobaculites and diversity and foraminiferal number, while still low, are higher than in samples from farther upstream. River mouth, beach and Dealwater faunas are abundant and very diverse with typical shallow marine faunas dominated by Elphidium-Discor- bis-Cibicides. Swan Lakes fauna is dominated by ostracods. The foraminiferal fauna is dominantly Am7no7iia- Elphidium. Introduction Foraminifera consitute a group of skeleton- producing protozoans with a long geological his- tory. Several important summaries of the biology and classification of this important group have been prepared, the most significant being those by Cushman (1948), Glaessner (1945) and especially Loeblich and Tappan (1964). The distribution of recent foraminifera has been the subject of several reviews in recent years and important contributions have been made by Phleger (I960) and most recently by Murray (1973). For many years, the study of the distribution of living foraminifera has been related to the needs of oil exploration companies for recon- struction of past environments. To this end, foraminifera of the Gulf of Mexico and nearby areas have been studied in great detail (see Walton 1964; Seiglie 1970 et seq., Phleger 1951, 1955 etc.). More recently there has been a tendency to study distributions of these organisms in man- made or man-affected situations (e.g. Bandy, Ingle and Resig 1964, et seq.) although the use of foraminifera for documenting changes due to pollution is in its infancy (Schafer 1970). The Hardy Inlet is a relatively small area and could prove an ideal test case for changes due to mining if mining is undertaken in the area. I School of Earth Sciences. Macauarie University, North Ryde, New South Wales. 2113. In Australia, little has been published on distribution of foraminifera from the major river systems but Albani (1968, et seq.^ and colleagues have made significant studies in New South Wales. Apthorpe (1974) has recently studied foraminifera from the Gippsland lakes of Vic- toria. McKenzie (1962) has made the only study in a comparable area from Western Australia. Work is proceeding on a similar study of Swan River foraminifera. Methods This report is based on an examination of 17 box core samples from within the Hardy Inlet and Swan Lakes and Deadwater. In addition, a single beach sand sample was taken immediately south of Deadwater to compare the oceanic and saline lake faunas. Rose Bengal staining of samples was attempted but the attempt can only be regarded as a failure. The results are based on total faunas only. Most samples were taken on 29 June 1974 when the Hardy Inlet proper was approximately at a winter condition. Later sampling at Station 9 (See Figure 1) was done to detect any difference between winter and summer distribution pat- terns. Localities and locality parameters are shown on Figure 1 and on the distribution charts. Also shown on Figure 1 are the sample localities which form part of a broader Hardy Inlet study (See Imberger and Agnew, in press). It was hoped that these sample localities could be used for the study of foraminifera but not all foraminiferal study sample stations are coincident with the standard localities. The results given here are by no means the final study that could be made of the Hardy Inlet foraminifera but give only a preliminary estimate of their distribution. Longer term studies at many more stations are needed. Samples represent the surface 1 cm from the top of each box core sample which was bottled, washed over a 100 pm sieve, and examined. Physical conditions in Hardy Inlet Physiographic units of the Hardy Inlet Hodgkin has identified a series of physio- graphic units which are detailed by Imberger and Agnew (1974). They are as follows: tidal river, lagoon, channel, Deadwater and Swan Lakes, and mouth and sea bar. Throughout this report, these units will be used and are shown on Figure 3A. The lagoon of Hodgkin can be divided con- veniently into two regions for this work. The dominant one is the true delta of the Blackwood- Journal of the Royal Society of Western Australia. Vol. 59, Part 3, February. 1977. 79 Scott system which has a classic delta shape with well marked distributary channel pattern. One of these distributaries has been accentu- ated by dredging. The remainder of what was defined originally as lagoon will be referred to as lagoon proper. To the units documented by Imberger and Agnew (1974) must be added oceanic beach. which is here represented by a single sample (18). Reliability of salinities, bathymetry etc. Detailed measurements of salinity (Figure 2) for Hardy Inlet have been made so far at short intervals over only one year (Imberger and Agnew. in press) so it is not certain that the Journal of the Royal Society of Western Australia. Vol. 59, Part 3, February, 1977. 80 'Metres Figure 2.— Summer and winter salinity and temperature, Hardy Inlet. Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February, 1977 . 81 figures gaven here are typical, although there is no reason to think that they are not. Virtu- ally the entire system is at a salinity of less than 5 Voo in typical winter pattern. The summer pattern is less uniform with surface salinity decreasing at approximately 1 “/oo per km up- stream from the mouth. It is noteworthy that the true delta of the Blackwood-Scott system acts as a shallow barrier between a deeper, more fluvial, channel up- stream, and shallower lagoon proper and channel downstream. As a result of the barrier, a deeper level salt wedge may become stranded in the tidal channel upstream. This is discussed Figure 3. — Parameters of foraminiferal distribution: A. sample type; B. bathymetry; C. dominant elements of biota: D. foraminiferal number. Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February, 1977. 82 further by Imberger and Agnew oo 4- Saline lakes The saline lakes — Swan Lakes and the Dead- water — are saline for a much longer part of the year than the rest of the Hardy Inlet system. As they are so shallow and as water access is only via very narrow channels, they have some features unique in the system. Although it is not properly documented, it is probable that they are slightly warmer than the waters in the rest of the study area. A figure of 4C° warmer than the water in Hardy Inlet itself has been men- tioned (R. Lenanton, pers. comm.). Because of these factors a slightly hypersaline condition ex'sts in part of the summer. Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February, 1977, 83 STATION NUMBER 9 10 8 7 6 5 4 3 II 12 2 1 17 13 15 14 16 18 Dstance(km) from Inlet mouth 8-5 7 8 6-5 4-5 4 4 3-5 7 5 3 2-5 2-5 0 2 1-5 1-5 1-5 Settinq Tidal River Delta Loqoon Channel Lakes Water Depth (m) 7-5 1 1-2 1-5 1-5 1-5 0-7 1-2 1-2 1-2 0 1 5-5 1-2 0-5 0-3 1-5 0 Number of specimens 94 36 II 38 42 II 2 48 27 55 86 48 92 146 55 56 125 182 Foraminiferal Number 300 18 3 13 27 7 1 4 13 25 14 8 500 1200 1000 500 270 2500 SPECIES^ < 1 C2 C.2 <2 C.3 12 6 <5 C.IO II 10 Protoschista finde.ns 1 7 4 1 1 6 Ammonia beccarii 89 6 10 4 1 3 10 6 5 3 38 15 A.tepida 4 4 4 Ammobaculites ogglutinans 7 18 8 3 18 18 22 4 A. sp. 1 16 33 6 10 3 29 50 Bathysiphon sp 1 1 Psammosphaera sp. II 2 1 22 3 Miliammina fusca 1 2 1 Trochammina inflate 1 1 ? F^enardogromia sp 1 PHyperammina sp 1 Quinqueloculina simplex 1 Elphidium crispum 3 3 37 9 16 29 Triloculina inflata 21 1 9 T. laevigata 12 1 3 Quinqueloculina striata 2 Q. lamarckiana 3 4 2 13 7 Q.subpolygona 2 2 Spiroloculina anquiata 1 S. rotunda 1 Miliolinella subrotunda 5 ? Hauerina bradyi 1 Fissurina fasc. carinata 1 Oolina striatopunctata 1 Bolivina sp. 1 2 1 Buliminoides williamsonianus 1 Reussella simplex 1 2 3 3 7 Rosaline vilardeboana 3 Neoconorbina terquemi 1 Pileolina australensis 4 1 4 R patelliformis 2 1 9 1 Mississippina concentrica 1 Spirillina inequalis 1 Elphidium advenum 2 2 1 E. poeyanum 8 22 9 Cibicides refulgens 1 19 2 6 17 22 G.pseudoungerianus 2 Figure 4. — Foraminiferal distribution chart. Part 1. Journal of the Royal Society of Western Australia. Vol. 59, Part 3, February, 1977. 84 s^Sis 9 10 8 7 6 5 4 3 II 12 2 1 17 13 15 14 16 18 Planorb. mediterranensis 1 Globigerina bulloides 1 Turborotalia sp. 2 Gaudryina convexa 3 1 Textularia pseudogramen 2 1 2 4 Marginoporo vertebralis 4 1 Triloculina striatotrigonula 3 1 Discorbis australis 4 1 1 4 D. dimidiatus 40 I 16 15 67 Gen. et sp indet 2 1 Elphidium jenseni 1 2 3 2 1 Vaivulineria rugosa 1 1 Acervulina inhaerens II 4 7 Gypsina globulus 7 5 ? Cassidulina sp 1 Globigerina sp 1 Rectobolivina raphanus 1 Elphidium incertum 3 E. simplex 1 Gypsina vesicularis 1 Cymbaloporetta bradyi 1 Q. seminulum 5 Spiroloculina venusta 1 1 Hyperammina ? cylindrica 1 Oolina sp 1 Bolivinella australis 2 Buliminella gracilis 1 Bolivina striatula 1 Heronallenia lingulata 1 Planulinoides biconcavus 3 2 Pileolina opercularis 4 3 Spirillina decorata 2 1 Calcarina calcar 2 Rotalia perlucida 1 Dyocibicides biserialis 1 Pyrgo lucernula 1 Spiroloculina communis ? Triloculina trigonula 1 Pavonina sp 1 Rotalia trochidiformis 1 Elphidium macellum 1 Cellanthus craticulatus 4 ? Crespinella sp 1 Amphistegina lessonii 1 Figure 4. — Foraminlferal distribution chart, Part. 2. Journal of the Royal Society of Western Australia. Vol. 59, Part 3. February, 1977. 85 A. — Swan Lakes station 15 Sediment: Pink, gelatinous, organic-calcium carbonate mud with superabundant ostracods. The sediment is almost an ostracod coquina. Study of this sample is hampered by difficulty in disaggregation due to high gelatinous organic content. Because of these difficulties the recorded relative abundance of various species may contain some error. Salinity: Winter — low, less than 5 ‘Voo- Summer — fully marine to a lit.le hypersaline. Station 14 (Location 05.5 of Imberger and Agnew) — channel leading from Swan Lakes to the Deadwater. Sea grass covers channel floor. Sediment: Carbonate sand. Salinity: Winter — low, less than 5 «/oo- Summer — fully marine to a lit.le hypersaline. B. — Deadwater station 16 (Location 02 of Imberger and Agnew) — 1.5-2m — weed-covered lake floor. Sediment: Poorly sorted carbonate sand showing much abrasion on the grains. Salinity: Winter — (a) Low, less than 5 atplliiormii 15 (a) 12 April 1970 ' Robert Bav, Peel Inlet 0.000 ± 0.005 15 (b) 12 April 1970 Core Sample (Top) 253 ± 12 15 (c) 12 .\pril 1970 ; Core Sample (Bottom) 74 -h 4 1(5 (a) 12 April 1970 ' Xear Channel, Peel Inlet O.llO 4- 0.007 10(b) 12 April 1970 , Core Sample (Top) 8 ± 1 10 (e) 12 April 1970 .'ore Sample (Bottom) .... 35 ± 3 17 (a) 12 .\pril 1970 Harvev Inlet. West Bank 0.070 ± 0.005 17(b) 12 .\pril 1970 Core Sample (Top) 49 ± 3 17 (e) 12 .\pril 1970 Core Sample (Bottom) 222 -t 11 reduced by adsorption into the underlying sedi- ments of the Peel Inlet. Gardiner (1974) has shown that cadmium is adsorbed on river muds, and that adsorption and desorption processes are likely to be major factors in controlling the concentration of cadmium in natural waters. It is likely that in the future, pollution will in- crease in the Peel River system. It would there- fore be desirable to monitor the cadmium content of the inlet and compare it with the 1975 base values. In April 1976 another three water samples were collected from the Peel and Harvey Inlets. These additional samples gave cadmium concentrations of 0.06, 0.07 and 0.11 ppb. The first two samples are only slightly higher than the average value of 0.04 ppb measured in the previous year. To test the hypothesis that much of the cad- mium is adsorbed on underlying sediments, core samples from the three additional locations were also collected. The core samples were approxi- mately 10 cm long, and 0.5 g portions from the top and bottom of the cores were digested by a HF-HClO.i mixture, and then subjected to an extraction procedure described by Rosman and De Laeter (1974). The results are listed in Table 2. At first sight the data may appear confusing, but the type of material sampled is significant. The Robert Bay core sample was very rich in humic constituents at the surface, whereas the bottom portion contained a higher portion of clay and other silicates. The Harvey Inlet sample was almost the reverse of the Robert Bay sample, the bottom part of the core being richer in humic material, whereas the upper portion contained some small shells. The core sample from the channel comprised white Journal of the Royal Society of Western Australia, Vol. 59, Part 3, February, 1977. 94 sandy material throughout its length, and the low cadmium concentration reflected this pre- ponderance of silicate material. Thus the underlying river mud acts as a sink for the dissolved cadmium, most of the cadmium being adsorbed in regions where humic material predominates. This conclusion is in agreement with the work of Gardiner (1974), who found that the humic constituents of mud are princi- pally responsible for its adsorptive properties. Gardiner has shown that in any fresh water system which has attained equilibrium, the con- centration of cadmium on finely divided solids may be 5 000-50 000 times higher than in solu- tion. In the present case the factor is at the lower level of this range, but it must be re- membered that the water is salty and the inlet is open to the sea for much of the year. Florence and Batley (1976) have shown that a proportion of the cadmium in sea water is pre- sent either as organic chelates or adsorbed on organic or inorganic particles. Colloidal part- icles are not retained by most ion-exchange resins because the resin pore size is too small to allow the colloids to enter the resin network 'Samuelson 1963). It was therefore decided to check if the chemical extraction used in this project was effective in measuring the total cadmium content in the water samples. A sample of water was taken from the upper reaches of the Helena River for this purpose and four 100 ml aliquots from the same poly- ethylene bottle were analysed. The first was spiked with ”'Cd and evaporated to dryness with distilled HNO^. 6M HCl was added to the residue and then taken to dryness. The residue was treated with IM HCl and sub- jected to the normal ion exchange procedure. A Cd concentration of 0.01 ± 0.01 ppb was obtained. The second and third samples were spiked and allowed to equilibrate for 3 and 7 days respec- tively, before processing. The last sample taken was observed to contain some particulate matter. This sample was spiked and allowed the normal mixing time. The second, third and fourth de- terminations yielded concentrations of 0.01 ± 0.01, 0.01 it 0.01 and 0.02 it 0.01 ppb respec- tively. The first three samples gave identical concentrations whereas the higher value obtained for the fourth sample is attributed to the pre- sence of particulate matter. The remaining 7 samples were collected at river crossings between Mundijong and Waroona along the South West Highway. This highway skirts the base of the Darling Scarp and is crossed by some 14 streams and rivers within the 60 km sector sampled. The water samples were collected upstream from the highway and as near to the centre of the stream as possible. Aftrr leaving the Darling Scarp the waterways traverse the coastal plain, and many finally feed into Peel Inlet. The waterways originate in the ranges behind the Darling Scarp in a region of virgin bush. The Serpentine and Waroona reservoirs are situated on two of the major streams. The cadmium concentrations are very low, for the most part being less than 0.1 ppb. Conclusions The World Health Organisation has declared that the maximum advisable concentration limit for cadmium in drinking water is 10 ppb (W.H.O. 1963). The recommended maximum level for irrigation purposes is 5 ppb for continuous use (Committee on Water Quality Criteria 1968). The results of this study demonstrate that the cadmium content in the two major river systems in south Western Australia are of the order of 100 times lower than the W.H.O. limits. The data indicate that the cadmium content tends to de- crease upstream from the mouth of the rivers studied, and that in the reservoir catchment areas the cadmium content is as low as 0.01 ppb in many places. These values compare very favourably with waterways in other parts of the world. Abdullah and Royle (1972) give a value of 0.41 ppb for “clean’' stream water in Wales, whilst in some of the rivers in England the average cadmium content is approximately 25 ppb (Valdez 1975). In North America many streams contain more than 1 ppb cadmium (U.S.G.S. 1970). Dale et al. (1974) have measured the cadmium content of a number of rivers in Victoria, Aus- tralia. Unfortunately their detection limit was 30 ppb and the cadmium content of many of the samples was below their detection limit. However the cadmium content of a number of samples was in excess of 30 ppb. Doolan and Smythe (1973) also found a range of cadmium concentrations (from < 0.02 to 7.7 ppb) in some rivers in New South Wales, Australia. The sensitivity of their technique was 0.02 ppb which compares favourably with ours. It is likely that much of the published work on the cadmium content in water has been based on analytical techniques which were in- capable of measuring accurately below the ppb level. The present study has succeeded in adapt- ing the stable isotope dilution technique using solid source mass spectrometry to the measure- ment of cadmium in river waters, and we be- lieve the measured concentrations establish a definitive set of low level base line determina- tions for cadmium in the environment. ACKnowLeagements.— The authors would like to thank the following graduate students who contributed to the project: I. D. Abercrombie, G. L. Cody. L. P. Costa, H K Cowan, C. B. McKay, D. R. Mills, M. T. Prosser, D. A. Ryan. S. Sandri, R. C. Seinor and D. B. Thornton. References Abdullah, M. I. and Royle, L. G. (1972).— Heavy metal content of some rivers and lakes in Wales, Nature 238, 329-330. Bowen, H. J. M. (1966). — Trace Elements in Bio- chemistry Academic Press. London. Cameron, A. E., Smith, D. H. and Walker. R. L. (1969). Mass spectrometry of nanogram-size samples of lead. Anal. Chem. 41, 525-526. Committee on Water Quality Criteria, (1968).— Report of the Committee, Federal Water Pollution Control Administration. U.S. Government Printer, Washington, D.C. Dale, D. H., Davis, M., Hall. C. T. and Hodgkins D (1974).— Concentrations of metal ions in Melbourne’s Rivers. Proc. Royal Aust. Chem Institute, 241-244. Journal of the Royal Society of Western Australia. Vol. 59. Part 3, February, 95 1977. Doolan, K. J. and Smythe. L. E. (1973).— Cadmium con- tent of some New South Wales Waters. Search 4. 162-163. Florence, T. M. and Batley, G. E. (1976).— Trace Metal Species in Sea Water— I. Removal of trace metals from sea water by a chelating resin. Talanta 23. 179-186. Gardiner. J. (1974). — The Chemistry of Cadmium in Natural Water— II. The Adsorption of Cad- mium on River Muds and Naturally Occur- ring Solids. Water Research. 8. 157-164. Reconnaissance of Selected Minor Elements in Surface Waters of the United States ( 1970).— Geol. Surv. Circular 43, U.S. Geological Survey, Washington. D.C. Rosman, K. J. R. and De Laeter, J. R. (1974). — Mass spectrometric Isotope Dilution Analyses of Cadmium in Standard Rocks. Chem. Geol 13. 69-74. Rosman. K. J. R. and De Laeter. J. R. (1975). — The isotope composition of cadmium in terres- trial minerals. J. Mass Spect. Ion Physics 16. 385-394. Rosman. K. J. R. and De Laeter. J. R. (1976.— Low level determinations of environmental cadmium Nature 261. 685-686. Sarnuelson. O. (1963).— Ion Exchange Separations in Analytical Chemistry. Wiley. New York. Struempler, A. W. ( 1973) .—Adsorption Characteristics of Silver. Lead. Cadmium. Zinc and Nickel on Borosilicate Glass. Polyethylene and Poly- propylene Container Surfaces. Anal. Chem. 45. 2251-2254. Valdez. H. (1975). — Cadmium in River Water. Ecologist 5. 347-348. W.H.O. (1963). — International Standards for Drinking Waters, 2nd Edit. World Health Organisa- tion. Geneva. Journal of the Royal Society of Western Australia. Vol. 59. Part 3. February, 1977. 96 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor, Royal Society of Western Australia, Western Australian Museum, Francis Street, Perth, Western Australia 6000. Publication in the Society’s Journal is available to all categories of members and to non-members residing outside Western Australia. Where all authors of a paper live in Western Australia at least one author must be a member of the Society. Papers by non-members living outside the State must be communicated through an Ordinary or an Honorary Member. Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10 000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white paper; the original and one copy should be sent. Tables and captions for Figures should be on separate sheets. All pages should be serially numbered. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, including the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. No lettering should be smaller than 1 mm on reduction. To avoid unnecessary handling of the original illustrations, which are usually best prepared between li and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. The metric system (S.I. units) must be used in the paper. Taxonomic papers must follow the appropriate International Code of Nomenclature, and geological papers must adhere to the Australian Code of Stratigraphic No- menclature. Spelling should follow the Concise Oxford Dictionary. Fifty reprints of each paper are supplied free of charge. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 59 Part 3 Contents 9. Lake Leschenaultia — an oligotrophic artificial lake in Western Australia. By R. P. Atkins, R. A. Congdon, C. M. Finlayson and D. M. Gordon. 10. Poison plants in Western Australia and colonizer problem solving. By J. M. R. Cameron. 11. Foraminifera of Hardy Inlet, southwestern Australia. By Patrick G. Quilty. 12. The cadmium content of some river systems in Western Australia. By K. J. R. Rosman and J. R. De Laeter. Editor; A. E. Cockbain The Royal Society of Western Australia, Western Australian Museum, Perth 59772/9/76—675 WILLIAM C. BROWN. Government Printer. Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 59 PART 4 JUNE, 1977 PRICE TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Chief Marshal Sir Wallace Kyle, G.C.B., C.B.E., D.S.O., D.F.C., K.St.J., Governor of Western Australia President Vice-Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor COUNCIL 1976-1977 P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. A. J. McComb, M.Sc., Ph.D. C. F. H. Jenkins, M.B.E., M.A., M.A.I.A.S. B. E. Balme, D.Sc. G. Perry, B.Sc. (Hons.) M. W. Perry, B.Sc. (Agric.) (Hons.) S. J. Curry, M.A. A. Neumann, B.A. A. E. Cockbain, B.Sc., Ph.D. P. R. Atkinson, B.Sc. C. E. Dortch, B.S., M.Phil. B. Lament, B.Sc. (Agric.), Ph.D., F.R.H.S. J. K. Marshall, B.Sc. (Hons.). Ph.D. L. J. Peet, B.Sc., F.G.S. P. E. Playford, B.Sc., Ph.D. B. P. Springett, B.Sc. (Hons.), Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. 13. — Middle Holocene marine molluscs from near Guildford, Western Aus- tralia, and evidence for climatic change by George W. Kendrick^ Manuscript received 25 May 1976; accepted 14 September 1976 Abstract Thirty one species of fossil molluscs are reported from a subsurface Middle Holocene deposit near Guildford, Western Australia, 26 km upstream from the mouth of the Swan Estuary. A radiocarbon age of 6660 ±: 120 yr BP (shell carbonate) indicates that the fauna lived near the end of the Flandrian transgression. In the light of their modern distributions, the fossils indicate that in Middle Holocene time, the Swan Estuary was a hydrologically stable arm of the sea, which experienced considerably less winter flooding than at present. A period of regional aridity is indicated, continuing on to some time after 4500 yr BP. Introduction During 1969-70, the Western Australian Public Works Department was engaged in channel clearing and deepening along a part of the estuary of the Swan River below Guildford (lat. 31°54'S, long. 115° 58' E) and about 26 km up- stream from the mouth at Fremantle. The area dredged was about 1 000 m long by 45 m wide, on either side of but mainly below the Helena River confluence (Fig. 1). Spoil was pumped ashore and discharged from a steel pipe line about 0.7 km away on the flood-plain of the Helena River adjacent to Great Eastern Highway and the Guildford State Primary School to provide for an extension of the school grounds. As discharged from the pipe, the spoil com- prised a coarse, greyish-brown, poorly sorted Quartz sand with a high proportion of angular grains, together with fragments of granitic rock and feldspar, mica flakes and occasional small ferruginous nodules. These components suggest a recent origin from the Precambrian rocks and associated laterite of the Darling Range. In addition, scattered sparsely within the spoil, were occasional mollusc shells and small pieces of soft, grey, calcareous, sandy siltstone, mostly shell-bearing. Many shells however were free of sediment (referred to below as “clean”) and a large proportion were also freshly broken or abraded; likewise the pieces of siltstone showed evidence of heavy abrasion as a result of their passage along the discharge pipe. The stratigraphic relationship of the elements of the spoil-pile has not been observed directly but according to the Public Works Department the modern channel bed is composed of “coarse sand”. Dredging removed from 0.3 to 2.1m of the substrate, and the resulting channel bed had a reduced level ranging from 4.4 to 5.6 m below Australian Height Datum (equivalent to mean sea level). It seems probable that the shell-bearing siltstone came from low in the cut beneath the channel sand and, from its relative ^Western Australian Museum, Francis Street, Perth, Western Australia, 6000. scarcity in the spoil, was either of no great thickness or was only slightly or intermittently penetrated by the dredge. Collections of shells from the spoil-pile were made firstly by Mrs. H. E. Merrrifleld in Decem- ber 1969, and subsequently by the writer in January and February 1970 and again in Feb- ruary 1971. The specimens have been accessed into the collection of the Western Australian Museum and provide the basis of this report. Most are in fresh, unweathered condition and, apart from some- recent breakage, are well pre- served. The fossil species Altogether, 35 species of molluscs, as well as crustaceans, polychaete tubes and a bryozoan were represented in the material collected. Identifications are available only for the molluscs, of which 18 species were represented by specimens directly associated with the grey siltstone. The remaining 17 species are con- sidered to comprise two groups, the larger, of 13 species, being mostly fragmentary shells which evidently had been washed and tumbled free of sediment in passage along the pipe. The lesser group of 4 species comprised the bivalves Westralunio carteri Iredale, Xenostrohus securis (Lamarck) and Anticorbula amara (Laseron) and the gastropod Ploticvsis australis (I. and H. C. Lea). With the exception of the first-men- tioned, these are permanent inhabitants of the Swan near Guildford (Chalmer et al. 1976); W. carteri inhabits freshwater tributaries such as the Helena River and Bennett Brook. The shells of these 4 species are believed to be associ- ated with the sand of the channel substrate rather than the calcareous siltstone and to represent the modern fauna at Guildford and upstream. Likewise, specimens of the flat-backed crab Halicarcinus australis (Haswell) (WAM 70.138), collected from the spoil-pile, are con- sidered to be modern. With this adjustment, the fossil molluscs are found to comprise 31 species, of which 16 are bivalves and 15 gastropods. In the following discussion, subdivisions (Lower, Middle and Upper) of the Swan Estuary and modern distri- butions within the estuary are from Chalmer et al. (1976). Other non-estuarine records are from the collection of modern molluscs of the Western Australian Museum. Bivalves Mytilidae Musculus sp. cf. M. nanulus Thiele. Material: numerous specimens of mature size, mostly dis- articulated valves embedded in grey siltstone; Journal of the Royal Society of Western Australia, Vol. 59. Part 4, June. 1977. 97 Figure 1. — Swan River estuary. Localities mentioned in text. Fossil site circled. two “clean” singles. WAM 70.66, 70.73, 70.110, 70.111, 70.143. The relationship of this to Thiele’s species, described from Shark Bay (Thiele 1930) is not clear. Guildford specimens have 35-40 ribs on the posterior slope and are up to 8 mm long. Habitat unknown but probably byssally attached to seagrasses, etc. This species now lives periodi- cally within the Lower Estuary. Pectinidae Pecten modestus Reeve. Material : a substanti- ally complete left valve and a fragment of a right, both “clean”. WAM 70.69, 70.112. The latter measures 55 x 61 mm and is about % mature size. Modern geographic range: southern Australia north to Shark Bay. Not recorded living from the Swan Estuary. An epifaunal species with some swimming ability. Ostreidae Ostrea angasi Sowerby. Material: an articu- lated pair of mature size, filled with grey silt- stone; two small singles embedded in siltstone; eight “clean” singles, mostly mature, four Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 98 attached in a cluster and with adherent Chama valves; the largest 11 x 9cm (slightly damaged). WAM 70.67, 70.68, 70.70, 70.113, 71.483. Modern geographic range: southern Australia, principally In estuaries (Macpherson and Gabriel 1962) but apparently not now living north of Cape Leeuwin. An epifaunal species, attached to stones, other shells, etc. below low water mark. Chamidae Chama ruderalis Lamarck. Material: three articulated pairs and five lefts embedded in grey siltstone; 13 left valves attached to oyster and other shells; five right and three left singles, all “clean”. Most specimens are comparable in size with local modern marine specimens. WAM 70.70, 70.113, 70.115, 71.483. Modern geographic range: South Australia, southern Western Aus- tralia (Cotton 1961), north to about Fremantle. Now lives probably permanently in the Lower Estuary. A sessile, epifaunal species, attached to firm substrates at or below low water mark. Cardiidae Laevicardium (Fulvia) apertum (Bruguiere). Material: two articulated pairs and three single valves embedded in grey siltstone; one fragment with adherent Chama valves; two “clean” frag- ments. Specimens are comparable in size with local marine shells. WAM 70.70, 70.71, 70.72, 70.73, 70.111. Modern geographic range: Indo-SW Paci- fic; in Western Australia south to Cockburn Sound (common) and Geographe Bay (rare). Now lives periodically in the Lower Estuary and part of the Middle Estuary ( in the vicinity of Pt Walter). An infaunal burrower in .‘^andy to muddy substrates. Laevicardium (Fulvia) tenuicostatum (Lam- arck). Material: one small left valve and two fragments from larger valves, all “clean”. WAM 70.74. Modern geographic range: southern Aus- tralia, north to about Fremantle; not recorded living from the Swan Estuary. An infaunal burrower in sandy to silty substrates. Tellinidae Tellina (Tellinangulus) sp. Material: two single valves, one with a little adherent grey calcareous sediment; the larger 5 mm long. WAM 70.83, 70.146. Modern geographic range: not known. Now common in the deeper parts of Cockburn Sound: lives periodically within the Lower Estuary. Probably an infaunal burrower. Tellina (Pinguitellina) sp. Material: a “clean” left valve of mature size (14 x 10 x 3 mm). WAM 70.121. Modern geographic range: Cockburn Sound to Shark Bay; now lives probably per- manently within the Lower Estuary. Probably an infaunal burrower. Tellina sp. Material: 21 “clean” valves, the largest 24 x 15 x 3 mm and comparable in size to local modern marine specimens. WAM 70.81, 70.120, 70.145. Modern geographic range: not known. Now common in the sea near Fremantle, particularly Cockburn Sound, and lives periodi- cally within the Lower Estuary. Of uncertain subgenus, this is an infaunal burrower in fine substrates. Psammobiidae Sanguinolaria (Psammotellina) biradiata (Wood). Material: a fragmentary left valve of medium size. WAM 70.119. Modern geographic range: southern Australia north to about Fre- mantle; lives permanently within the Lower Estuary. A deep burrower in sandy to muddy substrates. Veneridae Dosinia (Pectunculus) sculpta (Hanley). Material: one juvenile, articulated pair embedded in grey siltstone; three left and three right valves, all “clean”, the largest 37 x 40 x 9 mm. WAM 70.75, 71.483. Modern geographic range: northern and Western Australia, south to about Cockburn Sound; living periodically within the Lower Estuary. A burrowing species in fine substrates. Circe sulcata Gray. Material : two articulated pairs containing grey siltstone and a single embedded in the same; one left valve with adherent siltstone; seventeen “clean” singles, the largest 29 x 31 x 7 mm. WAM 70.76, 70.77, 70.116, 70.144, 70.2719. Modern geographic range: Indo- SW Pacific: in Western Australia, south to Albany; living periodically within the Lower and Middle Estuaries. A burrowing species in fine substrates; living specimens sometimes found on the surface of the substrate. Paphia (Callistotapes) crassisulca (Lamarck). Material: one broken left valve with adherent grey siltstone; a complete juvenile right and part of an adult left valve, both “clean”; a shell of Ostrea angasi with the external impression of a mature P. (C.) crassisulca on the lower valve. WAM 70.68, 70.78, 70.79. Shells of this species were common in the original collections and most were used for radiocarbon dating. Modern geographic range: Indian Ocean, northern Aus- tralia (Fischer-Piette and Metivier 1971); in Western Australia south to Cockburn Sound (common) and Cape Naturaliste (rare); prob- ably living permanently in the channel of the Lower Estuary. A burrowing species in fine sub- strates. Irus irus (Linnaeus). Material: two “clean” single valves and a fragmentary single extracted from a cavity of a piece of teredine-bored wood. WAM 70.117, 70.118. Modern geographic range: E. Atlantic, Mediterranean, Indo-W. Pacific (Fischer-Piette and Metivier 1971), Western Aus- tralia south to Cockburn Sound; not recorded living from the modern Swan Estuary. Inhabits crevices of rocks, shells, wood, etc. Hiatellidae Hiatella australis Lamarck. Material: four arti- culated pairs and three singles embedded in siltstone; eleven “clean” singles, the largest 13 X 5 mm, which is small compared with marine specimens. WAM 70.66, 70.68, 70.71, 70.84, 70.111, 70.123, 70.144, 71.483. Modern geographic range: Australia generally (Macpherson and Gabriel 1962): living periodically within the Lower Estuary. A sessile species inhabiting crevices of rocks, shells, etc. Pholadidae Pholas sp. cf. P. australasiae Sowerby. Material: a posterior fragment of a “clean” medium-sized right valve, probably of this species. WAM 70.82. Modern geographic range : Australia generally (Macpherson and Gabriel 1962); not recorded living from the modern Swan Estuary. A seden- tary species confined to burrows near low water mark; often collected near the mouths of estuaries. Gastropods Trochidae Mo7iilea callifera (Lamarck). Material; two shells, one filled with siltstone, the other “clean”. The larger shell measures 12 x 9 mm, about half mature size. WAM 70.124, 70.139. Modern geo- graphic range: Indo-SW Pacific; in Western Australia, south to Safety Bay; living periodically in the Lower Estuary. A herbivore associated with seagrasses in marine bays. Cyclostrematidae Elachorbis tatei (Angas). Material: four shells, one embedded in siltstone and three “clean”, the largest 3 mm in diameter. WAM 70.86, 70.111. Modern geographic range: South and Western Australia north to Shark Bay; living periodically within the Lower Estuary. A herbivore associated with seagrass and algal grow'th in sheltered waters. Diastomatidae Obtortio (Alabina) sp. Material: five shells embedded in siltstone and 25 “clean” shells, generally of mature size. WAM 70.71, 70.87, 70.111, 70.126, 70.143. Modern geographic range: south- western Australia, Albany to Shark Bay; living periodically within the Lower Estuary. Associated with fine substrates in sheltered waters. This may be the species listed by Thiele (1930) as Finella pupoides A. Adams, from Shark Bay and Warnbro Sound. Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 99 Cerithiidae Alaba fragilis (Thiele). Material: three shells embedded in siltstone; two “clean” shells, the largest 6 mm high. WAM 70.88, 70.111, 70.127. Modern geographic range: not known. Lives in Cockburn Sound and permanently in the Lower Swan Estuary, being described originally from Freshwater Bay (Thiele 1930). A herbivore associ- ated with algal and seagrass beds in sheltered waters. Epitoniidae Epitonium sp. cf. E. imperiale (Sowerby). Material: two “clean” shells, the larger 9mm high, which may be either juveniles of E. imperiale or another, closely related species. WAM 70.140. Modern geographic range: E. imperiale occurs in the Indo-SW Pacific and in Western Australia, south to Cape Naturaliste (Wilson and Gillett 1971). In Cockburn Sound it is believed to be associated commensally with the anemone Radianthus concmnata Lager (S. Slack-Smith, pers. comm., April 1975). Modern shells apparently conspecific with the Guildford specimens, are occasionally collected in the Lower Estuary of the Swan, where they may be living periodically. Naticidae Polinices (Conuber) conicus (Lamarck). Material: five “clean” shells, the largest damaged but originally about 4 cm high. WAM 70.89, 70.128. Modern geographic range: Australia generally; not recorded living from the Swan Estuary. An active, infaunal predator on bivalves, etc. A gastropod drill-hole of the bevelled type attri- buted to the Naticidae by Bromley (in Crimes and Harper 1970) was observed on a specimen of Dosinia sculpta (WAM 70.75c). Nassariidae Nassarius rufulus (Kiener). Material: one “clean” fragment from a shell probably about 17 mm high. WAM 70.131. Modern geographic range: south-western Australia, Albany to Geraldton (Wilson and Gillett 1971); not recorded living from the Swan Estuary. An infaunal scavenger/predator in shallow marine habitats. Nassarius pauperatus (Lamarck). Material: 32 shells, one filled with grey sandy siltstone; several incomplete. The largest shell measures 14 x 8 mm. WAM 70.129, 70.130, 70.141. Modern geographic range: southern Australia; in Western Australia north to Geraldton (Wilson and Gillett 1971): permanently living in the Lower and Middle Estuaries. An infaunal scavenger/predator, common in estuaries and marine bays. Nassarius pyrrhus (Menke). Material: a frag- mentary shell, lacking the spire, when intact about 10 mm high. WAM 74.895. Modern geo- graphic range: southern Australia, Victoria to Fremantle (Hodgkin et al. 1966); living period- ically within the Lower Estuary. A scavenger/ predator common in estuaries and marine bays. Pyramidellidae Turbonilla ( Chemnitzia) mariae Tenison Woods. Material; four “clean” shells. WAM 70.92. Modern geographic range : southern Australia ( Cotton 1959); not recorded living from the Swan Estuary. An ectoparasite. Turbonilla (Chemnitzia) sp. Material: one shell embedded in and another extracted from silt- stone; one “clean” shell. WAM 70.134, 71.484, 75.821. Distinguished from the preceding by having more ribs per whorl and a larger proto- conch perched atop poorly ribbed apical whorls; akin to T. iC.) macleayana Tenison Woods (R. Burn, pers. comm., Jan. 1975). Modern geographic range: unknown; not recorded living from the Swan Estuary. An ectoparasite. Agatha simplex (Angas). Material: one shell extracted from a piece of grey siltstone and one “clean” shell. WAM 70.91. 74.1123. Modern geo- graphic range: Queensland — southern Australia — north western Australia (Cotton 1959); not recorded living from the Swan Estuary. An ecto- parasite. Atysidae Liloa brevis (Quoy and Gaimard). Material: six “clean” shells, the largest 6.5 mm high. WAM 70.90, 70.132, 70.142. Modern geographic range: southern Australia, New South Wales to Fre- mantle (Hodgkin et al. 1966); living periodically within the Lower Estuary. A herbivore associated with seagrasses in sheltered waters. Retusidae Retusa sp. A. Material: one shell embedded in grey siltstone. WAM 70.111. A thin-shelled species, differing in shape from Retusa sp. B. Modern geographic range: not known, but living at least periodically in the Lower Estuary (R. Burn, pers. comm., Jan. 1975). An infaunal carnivore on foraminifers and/or small molluscs. Retusa sp. B. Material: one juvenile shell, extracted from a piece of grey siltstone. WAM 70.133. Modern geographic range: not known. Close to R. pygmaea (A. Adams), a southern Australian species (R. Burn, pers. comm., Jan. 1975). An infaunal carnivore on foraminifers and/or small molluscs. Age and correlation One of the more common species in the fossil material was the bivalve Paphia crassisulca, a robust, medium-sized clam, mature specimens of which measure from 2 to 3 mm through each valve. A comparative X-ray diffraction examin- ation of a Guildford fossil (WAM 70.79b) of P. crassisulca and a modern specimen of the same species from Cockburn Sound near Fre- mantle showed that each was composed of ara- gonite; no calcite was detected in either specimen and no significant compositional or crystallographic differences were noted between the two (M. Price and D. Burns, Government Chemical Laboratories, pers. comm., Dec. 1975). Thus the material has not been involved in any detectable carbonate exchange and is suitable for carbon-14 dating. A 200 g sample of these shells from Guildford was submitted for radiocarbon dating and a C'^ age of 6660 ± 120 yr BP (GaK 2874) was determined (Kigoshi et al. 1973). From this it is concluded that the deposit of siltstone presumed to underlie the modern channel sand at the dredged site was formed during Middle Holo- cene time and near the end of the Flandrian or last major glacio-eustatic transgression of the sea (Morner 1976). There is a general agreement between the age and estimated position of the Guildford deposit (about 5 m below datum) and data presented by Thom and Chappell (1975) from Australian sources. However no precise sea level can be deduced from the evidence available at Guildford. Shell beds of similar composition and age occur in the Swan Estuary at Heirisson Island, Perth and Melville Waters, etc., and have been discussed by Maitland (1919, p. 53), Reath (1925), Serventy (1955, p. 71) and Clarke et al. (1967, p. 136). The Guildford occurrence is more distant from the sea than any of these. Other shell beds from excavations at Cannington, Beckenham and Ferndale, on the Canning River well upstream from the modern broadwaters of the estuary, are under study by the writer. These contain mollusc faunas similar to that from Guildford and are considered to be of approximately similar age. Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 100 The Guildford Mid-Holocene molluscs differ markedly in species and preservation from those of the Caversham clay pits of Brisbane and Wunderlich Pty, located some 3 km to the north (Fairbridge 1954). Lying several metres above Datum, the Caversham deposit contains the bivalves Anadara trapezia (Deshayes), Mactra f Diaphoromactra) versicolor Tate and other species unknown in the Guildford fauna. The Caversham deposit is noticeably weathered and evidently is of Pleistocene age (Noakes et at 1967), probably deriving from the high sea levels of the Last Interglacial, approximately 100 000 yr ago (Broecker and van Donk 1970). Comparison of fossil and modern faunas The present Swan Estuary experiences a well defined, two phase, annual hydrologic cycle, which derives from the Mediterranean type climate of the region (Spencer 1956; Wilson 1968, 1969). Reliable, intense winter rainfall results in strong river discharge into the estuary, leading to a sharp drop in salinity and tempera- ture, stratification and deoxygenation of the water body; during each summer drought, this is replaced by marine circulation induced by a weak tidal oscillation. This marked seasonal contrast in the estuarine environment Is reflected in the distribution of the permanent benthic fauna, the species diversity of which declines sharply with distance from the sea. Thus Chalmer et at (1976) report 23 mollusc species living permanently in the Lower Estuary, 10 in the Middle Estuary and only 6 in the Upper Estuary; only 4 species are known to live permanently in the Upper Estuary near Guildford. An analysis of the Guildford molluscs is presented in Table 1. They are grouped into three categories of “permanently resident^’, “periodically resident” and “not recorded” in the present day Swan Estuary. Of the 31 species represented, only 7 are believed to still inhabit the estuary permanently and 6 of these appear to be confined to the Lower Estuary; the seventh ranges further upstream into parts of the Middle Estuary. A second group of 13 species lives from time to time in the Lower Estuary, when conditions are temporarily favourable (i.e., during periods of low river discharge), taut appears to be unable to live permanently in any part of the modern estuary, dying out in times of high winter discharge. The remaining 11 species have not been recorded from the modern estuary as either periodic or permanent inhabi- tants; all are of marine affinity and are either known or presumed to occur in marine environ- ments in south-western Australia. Chalmer et at (1976) recorded 6 mollusc species, all permanent residents of the Middle Estuary of the Swan, which appeared to be more abundant in estuarine rather than normal marine environ- ments. Of these 6, only 1, Nassarius pauperatus, is represented among the Guildford fossils. The modern upstream limit of this species is at about Pelican Point (Fig. 1). The same workers further recognized a group of 5 exclusively estuarine mollusc species characteristic of the Middle and Upper Estuaries, none of which is represented among the Guildford fossils. The differences in range and composition noted between the Guildford fossil assemblage and the modern estuary fauna show that there has been a general contraction seaward by all of the former (grouping of species) since the Middle Holocene, indicating that a substantial environmental change has affected the estuary since that time. The fossils include filter-feeding infaunal and epifaunal bivalves and herbivorous, scavenging, carnivorous and ectoparasitic gas- tropods. Other groups probably also present by inference were seagrasses, one or more actin- arians and other host-species, such as sabellid worms, for a suite of pyramidellid snails. Most species are represented by specimens of average- mature size and the fauna has a balanced diversity consistent with relatively stable, near- normal marine salinity in a marginal, sheltered, gulf environment. The Guildford fossils prob- ably represent a life assemblage or blocoenosis (Schafer 1972). If so, the evidence obtained is not compatible with modern levels of river dis- Table 1 Mollusc species from Guildford grouped according to their modern occurrences in the Swan Estuary com- parative data from Chalmer et al. (1976) and R. Burn (pers. comm., April 1975). (i) Probable permanent inhabitants of the Lower Estuary (7 also living permanently in part of the Middle Estuary. Chama ruderalis Tellina (Pinguitellina) sp. Tellina sp. Sanguinolaria ( Psammotellina ) biradiata species). Asterisk denotes species probably Paphia (Callistotapes) crassisulca Alaha fragilis Nassarius pauperatus* (ii) Periodic inhabitants of the Lower Estuary (13 species). Asterisk denotes species which mav periodically into the lower part of the Middle Estuary. Musculus sp. cf. M. nanulus Laevicardium (Fulvia) apertum* Tellina (Tellinangulus) sp. Dosinia (Pectunculus) sculpta Circe sulcata* Hiatella australis Monilea callifera Elachorhis tatei Obtortio (Alabina) sp. Epitonium sp. cf. E. imperiale Nassarius pyrrhus Liloa brevis Retusa sp. A (advice from R. Burn) range (iii) Not recorded in the modern estuary fauna Pecten modestus Ostrea angasi Laevicardium (Fulvia) tenuicostatum Irus irus Pholas australasiae Polinices (Conuber) conicus (11 species). Nassarius rufulus Turbonilla (Chemnitzia) mariae Turbonilla (Chemnitzia) sp. Agatha simplex Retusa sp. B (advice from R. Burn) Journal of the Royal Society of Western Australia. Vol. 59, Part 4, June, 1977. 101 charge and indicates that a qualitatively different hydrologic regime prevailed in the estuary during the Middle Holocene. Hydrologic regime From Fremantle Harbour to Rocky Bay, the Swan Estuary in its lowermost, inlet section is partly obstructed by an extensive sand-sill, which acts as a barrier to the free circulation of sea water into the Middle Estuary and be- yond (Chalmer et al. 1976). Dredge spoil from this sill has been found to contain a high pro- portion of fresh mollusc shell and other biogenic carbonate and appears to be of marine origin (G.W.K., unpublished data). The Guildford fossil deposit was formed near the culmination of the transgressive phase of the Holocene depositional cycle, at which time the estuary was rather deeper than is now the case. This was particularly so of the lower reaches, as a con- sequence of extensive downcutting by the Swan during the low sea levels of the late Pleistocene (Churchill 1959). The presence of a deep oceanic connection extending through the Lower Estuary and most if not all of the Middle Estuary would have greatly enhanced marine influence throughout the entire estuary but, in conjunc- tion with modern levels of stream discharge, could not, of itself, account for the presence of marine molluscs at Guildford. The marine component of the estuarine Swan environment was clearly stronger than at present during the Middle Holocene but the character of the complementary fluviatile element needs to be considered also. Of the various parameters of river discharge, the two most relevant appear to be volume and concen- tration (i.e. seasonality). A volume of discharge equal to or greater than present levels is ruled out by the fossil and lithologic evidence. The site is located around the Swan-Helena conflu- ence, the latter now being a major freshwater tributary. Geological maps of the district (Low and Lake 1971) suggest that there have been no more than minor changes in the posi- tion of the confluence during the Holo- cene. The presence at Guildford of marine species, such as Pecten modestus, Laevicardium tenuicostatum, Polinices conicus and Nassarius rufulus is highly significant. These species are now living in the sea around Fremantle but apparently are unable to live anywhere within the Swan Estuary under prevailing conditions. Their presence as fossils in the Swan near the Helena confluence is incompatible with active discharge from that tributary, whether this was seasonal (as now) or continuous. The same reasoning applies with almost equal force to other species in the fossil assemblage, for example, Chama ruderalis. Laevicardium aper~ turn, Sanguinolaria hiradiata, Circe sulcata, Paphia crassisulca and Epitonium sp. cf. E. imperiale. These are now able to live either permanently or temporarily in the estuary within a few km of the sea, but nowhere within 20 km of Guildford. The most likely reconstruction of the hydro- logic situation in the Middle Holocene Swan would seem to require a substantially reduced volume of river discharge throughout the entire drainage basin. Under such conditions, the importance of seasonality of discharge would tend to diminish, even to the point of becoming difficult to recognize and assess. The indications are, however, that both discharge volume and seasonality were much lower than at present and, hence, that the climate was relatively dry. Climatic change in the Middle Holocene The time of onset of this postulated dry episode pre-dates 6700 yr BP but is otherwise unknown. Radiocarbon dates from an undis- turbed. partly emergent shell bed at Point Way- len, Melville Water, may help to clarify an upper, terminal date. Mollusc shell carbonate from 10-20 cm above and 60-80 cm below Datum have produced ages of 4500 ± 100 yr BP (SUA 339) and 5940 ±110yr BP (SUA 341) respec- tively (Gillespie and Temple 1976). Faunal studies in progress on the Point Waylen deposit indicate that the mollusc assemblage, with over 60 species, is highly diverse and, like the Guild- ford material, lived under hydrologically stable, marine-gulf conditions. The fossil assemblages from Guildford and Point Waylen are similar and lead to the tentative conclusion that the modern hydrologic seasonality of the Swan Estuary did not develop until some time after 4500 radiocarbon yr BP. Late Quaternary climates in south-western Australia have received little detailed study and interpretations of evidence obtained by Lun- delius (1960) and Churchill (1968) have not supported the concept of a relatively dry Mid- Holocene in the region. Churchill (I960) envisaged a Mid-Holocene extinction of a “Eucalyptus-Casuarina woodland, Xanthorrhoea, Macrozamia and possibly Banksia and Agonis scrubs” on Rottnest Island, seeing this as a consequence of “a marine transgression to at least 9 feet above present sea level in 2000 B.C.”, corresponding to the Older Peron strandline of Fairbridge (1950). Such a post-Pleistocene stand of the sea has not won general acceptance (Morner 1976; Thom and Chappell 1975) and is not supported by the writer’s observations of the Holocene shell beds of the Swan Estuary, located within 50 km of Point Peron. These suggest that the maximum Mid-Holocene trans- gression in this region stood little, perhaps 0.5 m, above Datum. A transgression of that magni- tude or less would tend to favour the alternative suggestion of Grant-Taylor and Rafter (1963) that extinction of Xanthorrhoea on Rottnest may have resulted from “desiccation during the Hypsithermal Maximum”, corresponding to the warm Atlantic climatic phase of the northern hemisphere, which Wendland and Bryson (1974) locate between 8490 and 5060 yr ago. Substantial quantities of rain-derived salt have accumulated in ground waters of lateritic profiles in the Darling Range and the adjacent wheat belt of south-western Australia. Studies by Dimmock et al. (1974) have shown that the concentration of this salt tends to increase sharply with lower rainfall and raised evapora- tion; the lowest salt concentrations occur in the more humid western areas, characterized bv higher and more constant levels of stream discharge. Under the present climate on un- Journal of the Royal Society of Western Australia, Vol. 59, Part 4. June, 1977. 102 cleared land in the Darling Range, salt dis- charge slightly exceeds input, according to Peck and Hurle ( 1973) . Clearly the net rate of salt accession has been greater in the past, and this build-up could be expected to occur in periods of lower rainfall and stream discharge associated with increased evaporation and concentration of ground water salt. If so, this phenomenon may be viewed as an index of past regional climate. It has been pointed out by Dimmock et al. (1974) that only a relatively brief period of time would be required to account for the large quantity of salt stored in ground water at Bakers Hill in the eastern Darling Range. The fossil evidence at Guildford suggests that the regional climate was much drier than present during the Middle Holocene, at least from 6700 yr BP until some time after 4500 yr BP. This dry period, begin- ning at some unknown time, may therefore have been the most recent episode of net salt accu- mulation in the ground waters of the Darling Range. Acknowledgements. — Thanks are due to Mrs. H. E. Merrifield for making her collection of fossils available for this study and to Mr. W. Fleay, Harbours and Rivers Branch, Public Works Department, for the provision of information regarding dredging operations at Guild- ford. Radiocarbon dates from Point W^aylen were generously provided by Mr. R. Gillespie, Department of Physical Chemistry, University of Sydney. I thank Mr. R. Burn of Geelong, Victoria, for identifications and advice on opisthobranch molluscs and Dr. B. Metivier, Museum National d’Histoire Naturelle, Paris, for com- paring specimens of Paphia crassisulca with the type. Dr. D. Merrilees and Mr. A. Baynes criticized the draft and suggested a number of improvements to the text The map was drawn by Ms Jane d’Espeissis. Postscript. — An extension of the present fossil fauna has recently been discovered in core samples from the flood plain of the Swan near Guildford Grammar School, some 4.2 km upstream from the Helena con- fluence. It is associated with a black clay lying between 3 and 6 m below the ground surface. Molluscs are similar to the present assemblage; also present are echinoderm ossicles which represent an asteroid of the genus Astropecten (L. M. Marsh pers. comm.). The core samples, which are uncontaminated by the modern channel substrate of the Swan, do not contain the molluscs Westralunio carteri, Xenostrobus securis, Anticorbula amara and PlotiopsU australis, species excluded from the present study because of their association with the modern channel substrate at Guildford. The new material justifies this exclusion and supports the palaeoenvironmental deductions of this paper. Presentation of this new material by Messrs. H. Grant and J. Backhouse is gratefully acknowledged. References Broecker, W. S. and van Donk, J. (1970). — Insolation changes, ice volumes and the Qis record in deep-sea cores. Reviews of geophysics and space physics, 8: 169-198. Bromley, R. G. (1970). — Borings as trace fossils and Entobia cretacea Portlock as an example, in Crimes, T. P. and Harper, J. C. (Editors). Trace fossils. Geological Journal Special Issue 3. Liverpool, Seel House Press. Chalmer, P. N., Hodgkin, E. P. and Kendrick, G. W. (1976). — Faunal changes within an unstable estuarine environment in south-western Australia. Records of the Western Australian Museum, 4: 383-410. Churchill, D. M. (1959). — Late Quaternary eustatic changes in the Swan River district. Journal of the Royal Society of Western Australia, 42: 53-55. Churchill. D. M. (1960). — Late Quaternary changes in the vegetation on Rottnest Island. The Western Australian Naturalist, 7: 160-166. Churchill, D. M. (1968). — The distribution and pre- history of Eucalyptus diversicolor F. Muell,. E. marginata Donn ex Sm., and E. calophylla R. Br. in relation to rainfall. Australian Journal of Botany, 16: 125-151. Clarke, E. de C., Prider, R. T. and Teichert, C. (1967). — Elements of geology for Australian students. University of Western Australia Press. 4th edition. Cotton, B. C. (1959), — South Australian Mollusca. Archaeogastropoda. Adelaide, Government Printer. Cotton, B. C. (1961). — South Australian Mollusca. Pelecypoda. Adelaide, Government Printer. Dimmock, G. M., Bettenay, E. and Mulcahy, M. J. (1974). — Salt content of lateritic profiles in the Darling Range, Western Australia. Australian Journal of Soil Research, 12: 63-69. Fairbridge, R. W. (1950). — The geology and geomor- phology of Point Peron, Western Australia. Journal of the Royal Society of Western Australia, 34: 35-72. Fairbridge, R. W. (1954). — Quaternary eustatic data for Western Australia and adjacent States. Proceedings of the Pan Indian Ocean Science Congress. Section F: Geography and Oceanography. Fischer-Piette, E. and Metivier, B. (1971). — Revision des Tapetinae (Mollusques : Bivalvia). Memoires du Museum National d’Histoire Naturelle. Serie A. Zoologie. Tome LXXI. Gillespie, R. and Temple, R. B. (1976). — Sydney Uni- versity natural radiocarbon measurements III. Radiocarbon, 18: 96-109. Grant-Taylor, T. L. and Rafter, T. A. (1963). — New Zealand Natural Radiocarbon Measurements I-V. Radiocarbon, 5: 118-162. Hodgkin, E. P., Kendrick, G. W., Marsh, L. and Slack- Smith, S. (1966). — The shelled Gastropoda of south western Australia. Handbook No. 9. Western Australian Naturalists’ Club, Perth. Kigoshi, K., Suzuki, N. and Fukatsu, H. (1973). — Gakushuin Natural Radiocarbon Measure- ments VIII. Radiocarbon, 15: 42-67. Low, G. H. and Lake, R. W. (1970). — Perth and environs geological maps (4 sheets). Geo- logical Survey of Western Australia. Lundelius, E. L. (1960). — Post Pleistocene faunal succession in Western Australia and its climatic interpretation. Report of the Inter- national Geological Congress, XXI Session, Norden, 1960. Part IV. Chronology and climatology of the Quaternary: 142-153. Macpherson, J. H. and Gabriel, C. J. (1962). — Marine molluscs of Victoria. Melbourne University Press — National Museum of Victoria. Maitland, A. G. (1919). — A summary of the geology of Western Australia, in The Mining Handbook. Geological Survey Memoir No. 1. Perth, Mines Department. Mdrner, N. — A. (1976). — Eustatic changes during the last 8 000 years in view of radiocarbon calibration and new information from the Kattegatt region and other northwestern European coastal areas. Palaeogeography, Palaeoclimatology, Palaeoecology, 19: 63-85. Noakes, J. E., Kim, S. M. and Akers. L. K. (1967).— Oak Ridge Institute of Nuclear Studies. Radiocarbon Dates 1. Radiocarbon, 9: 309-315. Peck, A. J. and Hurle, D. H. (1973).— Chloride balance of some farmed and forested catchments in south western Australia. Water Resources Research, 9: 648-657. Reath, J. L. (1925). — The Mollusca from the sub-Recent shell beds of the Lower Swan River. Journal of the Royal Society of Western Australia, 11: 31-41. Schafer, W. (1972). — Ecology and palaeoecology of marine environments. Edinburgh, Oliver and Boyd. Journal of the Royal Society of Western Australia, Vol. 59, Part 4. June. 1977. 103 Serventy, D. L. (1955). — The fauna of the Swan River Estuary, in Swan River Reference Commit- tee. Report on pollution of the Swan River. Perth. Western Australia, Government Printer. Spencer, R. S. (1956). — Studies in Australian estuarine hydrology. II. The Swan River. Australian Journal of Marine and Freshwater Research, 7: 193-253. Thiele, J. (1930). — Gastropoda und Bivalvia, in Michael- sen, W. and Hartmeyer, R. (editors). Die Fauna Sudwest-Australiens. Jena, Gustav Fischer. Thom, B. G. and Chappell, J. (1975). — Holocene sea levels relative to Australia. Search, 6: 90-93. Wendland, W. M. and Bryson, R. A. (1974). — Dating climatic episodes of the Holocene. Quartern- ary Research, 4: 9-24. Wilson, B. R. (1968). — Survival and reproduction of the mussel Xenostrobus securis (Lam.) (Mol- lusca : Bivalvia : Mytilidae) in a Western Australian estuary. Part I. Salinity toler- ance. Journal of Natural History, 2: 307-328. Wilson, B. R. (1969). — Survival and reproduction of the mussel Xenostrobus securis (Lamarck) (Mollusca : Bivalvia ; Mytilidae) in a Western Australian estuary. Part II. Repro- duction, growth and longevity. Journal of Natural History, 3: 93-120. Wilson, B. R. and Gillett, K. (1971). — Australian shells. Sydney, A. H. & A. W. Reed Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 104 14. — Potassium-argon ages of hornblendes from Precambrian gneisses from the south coast of Western Australia by N. C. N. Stephenson\ T. G. RusselP D. Stubbs“, &: G. I. Z. Kalocsai^ Manuscript received 25 May, 1976; accepted 14 September, 1976 Abstract Potassium-argon ages obtained from horn- blendes from seven samples of amphibolite and basic granulite from three localities in the gneiss-granite complex of the south coast of Western Australia range from 1060 to 1160 m.y., with no detectable differences between localities. These results show good agreement with the Rb-Sr isochron age of 1100 ±: 50 m.y. previously obtained from the Albany Adamellite. This could suggest that the K-Ar hornblende ages reflect the period of late-kinematic emplace- ment of anatectic granitic plutons (represented by the Albany Adamellite) which is believed to have occurred during waning metamorphism. Alternatively, they could be regarded as indicat- ing the age of late-orogenic regional uplift and cooling. These are not necessarily conflicting interpretations because it is possible that late- kinematic emplacement of granitic magmas and regional uplift were roughly synchronous events. About 1150 m.y. can be regarded at present as only a minimum estimate of the age of high-grade regional metamorphism in the south coast area. Introduction Precambrian gneisses and granitic plutons outcrop along the south coast of Western Aus- tralia between Point D’Entrecasteaux and Israelite Bay, a distance of roughly 750 km. This area forms part of the Albany-Fraser Province, a narrow arcuate belt wrapped around the southern and southeastern margins of the Archaean Yilgarn Block. This belt appears to represent the site of a younger Precambrian orogenic belt that cut across the Archaean shield (Wilson 1969). A general account of the geology of the region has been given by Clarke et al. (1954). The gneisses are predominantly granitic in character, with intercalated metasedimentary and meta- basite bands; migmatitic types are common. The metamorphic grade varies from upper amphibo- lite to lower granulite facies, with the higher- grade rocks occurring as large enclaves scattered through the amphibolite facies terrain. The gneisses have been intruded by numerous por- phyritic coarse-grained granitic plutons, cut in turn by fine-grained granitic dykes and, later, by occasional dolerite dykes. Stephenson (1973a, h, 1974) believes that the granitic plutons and dykes have been derived from the gneissic country rocks by anatexis during orogeny and high-grade regional metamorphism, and that emplacement and crystallisation occurred during waning metamorphism under syn- to late- kinematic conditions. ^Department of Geology, University of New England, Armidale, New South Wales, 2351. ^Isotope Laboratory, Department of Geology and Miner- alogy, University of Queensland, St Lucia, Queens- land, 4067. The geochronology of the south coast area has not been adequately investigated. Turek and Stephenson (1966) reported a microcline-total- rock Rb-Sr isochron age of 1100 — 50 m.y. for one of the granitic plutons (the Albany Adamel- lite) but there may, of course, have been more than one episode of granitic magma emplace- ment. The main phase of high-grade regional metamorphism responsible for the gneisses is thought to have preceded the emplacement of granitic plutons, but it has not been reliably dated. The only published radiometric data on these gneisses are a chemical U-Pb age (un- corrected for Pb isotopes) of 1390 + 50 m.y. given by allanite from pegmatitic schlieren in chamockitic gneiss from Doubtful Island Bay (Prider 1954), and a Rb-Sr age of 970 m.y. given by biotite from gneiss from the same locality (Wilson et al. 1960). The significance of these dates is not yet fully understood, but the latter is clearly only a minimum estimate of the age of the main metamorphism. The geochronology of the northeastern part of the Albany-Fraser Province (i.e. in the vicinity of the Fraser Range) is better known and possibly of some significance to the south coast region. Compston and Arriens (1968) and Arriens and Lambert (1969) reported a total- rock Rb-Sr isochron of 1330 =t 15 m.y., which is interpreted as a reliable estimate of the age of an episode of granulite facies metamorphism in this area. Preliminary data presented by Arriens and Lambert (1969) and Bunting et al. (1976) suggest that gneisses and granites flanking the 1330 m.y.-old granulites were formed during an earlier metamorphism around 1600-1900 m.y. ago. A period of retrogression and pegmatite emplacement during uplift and waning meta- morphism about 1280-1290 m.y. ago is suggested by a Rb-Sr age of 1280 m.y. obtained by Aldrich et al. (1959) on muscovite from one of the numerous small pegmatites in the 1300 m.y.-old granulites, and by a Rb-Sr total-rock isochron age of 1289 ± 21 m.y. given by muscovite-bearing gneisses and pegmatites associated with gneisses possibly 1900 m.y. old (Bunting et al., 1976). In view of the structural continuity between the south coast and Fraser Range areas (Doepel 1969) it is likely that they experienced regional metamoiphism at least broadly contempor- aneously. Furthermore, it seems reasonable to suppose that the history of metamorphism of the south coast gneisses was possibly as long and complex as it appears to have been in the Fraser Range area. The puipose of this paper is to report and interpret K-Ar ages obtained from hornblendes Journal of the Royal Society of Western Australia. Vol. 59, Part 4. June. 1977. 105 separated from seven samples of amphibolite and basic granulite from three localities in the gneissic complex along the south coast. Two of these samples are from Albany near the Albany Adamellite, two from Forsyth Bluff near the Torbay Adamellite, and three from Cape Riche where there are no known granitic plutons. These localities are shown in Figure 1, and short petrographic descriptions of the samples are given in the Appendix. It was hoped that a comparison of the ages of the Albany and Forsyth Bluff samples, which are expected to reflect the period of granitic magma emplace- ment, with those obtained from the Cape Riche samples would clarify the temporal relations between high-grade metamorphism and granitic magma emplacement in the south coast area. It is realised that the scope of this investigation is far too limited to provide a detailed geo- chronological analysis of the possibly complex metamorphic history of the region. Methods Hornblende was separated from the washed -120 to -f200 mesh fractions of the crushed rock samples using a Frantz isodynamic separ- ator followed by repeated centrifuging in heavy liquids. The purity of the final separates was better than 99%. The methods used in the K-Ar age determina- tions have been described by McDougall (1966). Potassium was determined by flame photometry. Argon measurements were carried out in the Isotope Geology Laboratory at the University of Queensland using the isotope dilution technique, following fusion of the sample and extraction of argon in a high vacuum line. After purifica- tion of the gas the isotopic composition was determined on an A.E.I.G.E.C. — MS20 mass spectrometer. The constants used in the cal- culations were: Xe - 0.585 X lO'^” yr^; x^ - 4.72 X lO'"'* yrh -""K/K - 0.000119. Errors were calculated using the method of Kirsten (1966). The results are listed in Table 1. Discussion of results Table 1 shows that the measured ages range from 1060 to 1160 m.y., with no detectable differences between the three sample localities. Samples R36265 and R36266 were collected from basic granulite bands in the gneiss within the granitised zone around the Albany Adamel- Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 106 Table 1 K-Ar ages on hornblendes from amphibolites and basic granulites from the south coast of Western Australia Sample No.* QA No.t Rock Type K 0/ /o Rad. ^"Ar % Age (m.y.) Error! (m.y.) Latitude (south) Longitude Locality (east) 54563 267 Amphibolite 0-71 29-71 § 1059 ±51 35^05 -r 117^38 -6' Forsyth BlufF 65581 273 Amphibolite 1 -04 99-75 1 151 ±15 35°05-r 117^38-7' Forsyth Bluff R36265 268 Granulite 1-13 97-57 1156 ±24 35^01 -8' 117'55-r King Point, Albany R36266 269 Granulite 1 -04 34-86 § 1 103 ±31 35^01 -5' 117'^55-r Ellen Cove, Albany R36269 270 Granulite I -38 98-64 1133 ±24 34 36-5' 11 8^46 -3' Cape Riche R36270 271 Granulite 1 -45 99-49 1153 ±24 3436-5' 118'^46-3' Cape Riche R3627I 272 Granulite 1-41 97-86 1066 ± 18 34'36-2' 118'46-2' Cape Riche Ae = 0-585 X 10 yr-^ = 4-72 x 10-i« yr’^ ""K/K = 0 0001 19 * Numbers prefixed by R refer to the collection of the Geology Department, University of New England. The remainder refer to the collection of the Geology Department. University of Western Australia, t University of Queensland. Isotope Laboratory catalogue number. % Errors calculated using the method described by Kirsten (1966). § The low values of radiogenic argon percentages are due to a malfunction in the pumping system and they are reflected in the rather large error. For a discussion of the effect of atmospheric contamination see Cox & Dairymple (1967). lite. The rocks within this zone show extensive metasomatic enrichment in Si and K, and partial retrogression to amphibolite facies assemblages, presumably due to the introduction of hydrous fluids from the adamellite magma (Stephenson 1974). Both samples contain a little orthoclase and quartz of probable metasomatic origin (see Appendix), and there seems little possibility that they escaped the thermal influ- ence of the pluton. The ages obtained from these samples (1156 ± 24 and 1103 ± 31m.y.) show good agreement with the Rb-Sr isochron age of 1100 ± 50 m.y. obtained by Turek and Stephenson (1966) for the Albany Adamellite, and they therefore seem to reflect the episode of magma emplacement. Samples 54563 and 65581 are from amphi- bolite bands outcropping 800 m and 1 400 m respectively from the southeastern margin of the Torbay Adamellite. Neither sample shows obvious thermal or metasomatic effects attribu- table to the pluton, but 54563 lies within the zone of appreciable Si and K metasomatism, and 65581 is from the perimeter of this zone. Hence the age obtained for 54563 (1059 ± 51 m.y.) is likely to reflect the emplacement of the Torbay Adamellite. This pluton has not been dated but it shows late-kinematic character- istics similar (though less obvious) to those of the Albany Adamellite (Stephenson 1974), so it is tentatively believed to be of comparable age; i.e. roughly 1100 m.y. The age given by 65581 (1151 ib 15 m.y.), though older, may also have been influenced by emplacement of the Torbay Adamellite, and hence it provides only a mini- mum estimate of the age of high-grade regional metamorphism. The significance of the results obtained from the Cape Riche samples is not entirely clear. Although there are no known granitic plutons in the area, the dates obtained from these samples — namely 1133, 1155, and 1066 m.y. — are not detectably older than the inferred age of granitic magma emplacement. Hence the data have failed to demonstrate that the phase of high-grade regional metamorphism in the south coast area occurred significantly earlier than the emplacement of granitic magmas repre- sented by the Albany Adamellite. However this is a possibility that still cannot be ruled out because the dates obtained from the Cape Riche samples could reflect any of several events in the orogenic cycle following the most intense phase of metamorphism. For example, the em- placement of large masses of granitic magma about 1100 m.y. ago may have been caused by, or given rise to, a period of regional reheating detectable by K-Ar dating even in areas remote from these plutons. The minor retrograde effects which appear to be ubiquitous in granulite facies areas of the south coast should also be considered. The Cape Riche samples from which the hornblendes were separated for dating were chosen for their relative freedom from these effects, but incipient alteration is nevertheless evident in their orthopyroxenes and plagioclases (see Appendix). Although the hornblendes show no visible signs of alteration the dates obtained from them could reflect this minor retrogres- sion. Another possible interpretation of the predominance of Rb-Sr and K-Ar dates around 110()-1150 m.y. obtained from the intrusive and gneissic rocks is that they represent the phase of regional uplift and cooling during the final stages of the orogenic cycle in the south coast area. Even in hornblende, a relatively retentive mineral (e.g. Dairymple and Lanphere 1969, p. 174), diffusion loss of radiogenic argon does not cease until cooling has reached a tempera- ture a few hundred degrees below the tempera- ture of upper amphibolite to lower granulite facies metamorphism to which these rocks have been subjected. Therefore it is unlikely that K-Ar dating will record the age of the most intense phase of metamorphism unless cooling was particularly rapid. The K-Ar ages reported above show an interesting correlation with the total-rock Rb-Sr isochron age of 1150 ± 40 m.y. obtained by Turek and Stephenson (1966) from slates and phyllites from the nearby Stirling Range Beds, and this merits brief discussion. The Stirling Range Beds, together with the Mt Barren Beds to the east, form an east-west-trending strip of Precambrian low-grade metasediments out- cropping discontinuously along the junction between the Albany-Fraser Province and the Yilgarn Block in the south coast area (Pig. 1) Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 107 Contact relations are not well exposed but the low-grade metasediments appear to overlie the gneiss-granite complexes of the Albany-Fraser Province and the Yilgarn Block with a faulted unconformity (Sofoulis 1958a, b). The Stirling Range and Mt Barren Beds exhibit structures interpreted by several observers as indicating thrusting from the south; either thrusting of the metasediments themselves towards and over the adjacent Yilgarn Block, or overthrusting of the metasediments by the gneiss-granite complex of the Albany-Fraser Province, or a combination of both (Wilson 1952; Clarke et al. 1954; Sofoulis 1958a, b). It seems likely that these thrusting movements may have been associated with uplift of the Albany-Fraser Province in the south coast area. Clarke et al. (1954) indicated that the low-grade metamorphism of the Stirl- ing Range Beds was associated with the thrust- ing described above, and Turek and Stephenson (1966) interpreted the 1150 m.y. Rb-Sr isochron obtained by them from slates and phyllites from the Stirling Range Beds as the age of this metamorphism and thrusting. If this low-grade metamorphism and thrusting was associated with uplift of the Albany-Fraser Province as suggested above, then this uplift occurred roughly 1150 m.y. ago. This is consistent with the suggestion that the 1060-1160 m.y. K-Ar ages reported in this paper reflect mainly the post- metamorphic regional uplift and cooling of the gneiss-granite complex of the south coast area. This conclusion does not necessarily conflict with the alternative suggestions that late- kinematic emplacement of granitic magmas or widespread incipient retrogression, or both, have influenced the recorded ages because it is not improbable that these events were roughly syn- chronous with regional uplift. In any case, about 1150 m.y. can probably be regarded at present as only a minimum estimate of the age of regional metamorphism in the south coast area. Because it seems likely that this region has had a long and complex history of deformation, metamorphism, and granitic magma emplace- ment yet to be unravelled, further structural, petrological, and geochronological studies should prove rewarding. Acknowledgements. — N.C.N. Stephenson was financi- ally assisted by the University of New England, and D Stubbs by the Australian Research Grants Com- mittee under a grant to Dr. D. C. Green, University of Queensland. Dr. Green gave advice on several aspects of the project and critically reviewed the manuscript. Mr. J. A. Bunting and Drs. J. R. de Laeter and W. G. Libby kindly made available to the authors their radiometric data on the Fraser Range area prior to publication. Messrs J. S. Cook and N. Petrasz assisted with sample preparation, and Mr. M. R. Bone drafted the diagram. References Aldrich, L.T., Wetherill. G. W., Bass, M. N., Compston, W., David, G. L., and Tilton, G. R. (1959). — Mineral age measurements. Yh. Carnegie Jnstn Wash., 58: 237-250. Arriens, P. A., and Lambert, I. B. (1969). — On the age and strontium isotopic geochemistry of granulite-facies rocks from the Fraser Range, Western Australia. and the Musgrave Ranges, Central Australia. Spec. Pubis Geol. Soc. Aust., 2: 377-388. Bunting, J. A., de Laeter J. R., and Libby, W.G. (1976). — Tectonic subdivisions and geo- chronology of the northeastern part of the Albany-Fraser Province, Western Australia. Ann. Rep. Geol. Surv. West. Aust., 1975,: 117-126. Clarke, E. de C., Phillipps, H. T., and Prider, R. T. (1954). — The Pre-Cambrian geology of part of the south coast of Western Australia. J. Roy. Soc. West. Aust., 38: 1-64. Compston, W., and Arriens, P. A. (1968). — The Pre- cambrian geochronology of Australia. Can. J. Earth Sci., 5: 561-583. Cox, A., and Dalrymple, G. B. (1967).— Statistical analysis of geomagnetic reversal data and the precision of potassium-argon dating. J. Geophys. Res., 72: 2603-2614. Dalrymple, G. B,, and Lanphere, M. A. (1969).— “Pofas- sium- Argon Dating: Principles, Techniques and Applications to Geochronology.” Free- man, San Francisco. Doepel, J. J. G. (1969). — The Precambrian geology be- tween Zanthus and Israelite Bay, Western Australia. Ann. Rep. Geol. Surv. West. Aust., 1968-. 41-42. Kirsten, T. (1966). — Determination of radiogenic argon; in Schaeffer, O. A., & Zahringer, J. (Eds) ‘‘Potassium Argon Dating”: 7-39. Springer- Verlag, Berlin. McDougall, I. (1966). — Precision methods of potassium- argon isotopic age determination on young- rocks; in Runcorn, S. K. (Ed) ‘‘Methods and Techniques in Geophysics”. 2: 279-304. Interscience, London. Prider, R. T. (1954). — The Pre-Cambrian succession in Western Australia. Proc. Pan-Ind. Ocean Sci. Congr., Perth, Sect. C: 69-78. Sofoulis, J. (1958a). — Notes on a reconnaissance of the Stirling Range area, South-East Division. Bull. Geol. Surv. West. Aust., 109: 78-80. Sofoulis, J. (19585). — The geology of the Phillips River Goldfield, W.A. Bull. Geol. Surv. West. Aust., 110 . Stephenson, N.C.N. (1973a). — The petrology of the Mt Manypeaks Adamellite and associated high- grade metamorphic rocks near Albany, Western Australia. J. Geol. Soc. Aust., 19: 413-439. Stephenson, N.C.N. (19735).— The petrology of the Mt Gardner Adamellite, near Albany, Western Australia. J. Roy. Soc. West. Aust., 56: 103-108. Stephenson, N.C.N. (1974). — The petrology of the Albany and Torbay Adamellite plutons, near Albany, Western Australia. J. Geol. Soc. Aust., 21: 219-246. Turek, A., and Stephenson, N.C.N. (1966).— The radio- metric age of the Albany granite and the Stirling Range Beds, south-west Australia. J. Geol. Soc. Aust., 13: 449-456. Wilson, A. F. (1952). — The charnockite problem in Australia. Sir Douglas Mawson Aniv. Vol., 203-224, Univ. Adelaide. Wilson, A. F. (1969). — Granulite terrains and their tectonic setting and relationship to associ- ated metamorphic rocks in Australia. Spec. Pubis Geol. Soc. Aust., 2: 243-258. Wilson, A. F., Compston, W., Jeffrey, P. M., and Riley, G. H. (1960). — Radioactive ages from the Precambrian rocks in Australia. J. Geol. Soc. Aust., 6: 179-195. Appendix — Petrography Saiiiples 54563 and 65581: Amphibolites These are flne- to medium-grained equigranu- lar rocks composed of hornblende (roughly 50%) with X - pale yellowish brown, Y ^ ^ yellowish green, Z ^ green, andesine (40%) showing patchy saussuritisation, and biotite (10%). Quartz, magnetite, sphene, apatite, and zircon are minor accessories. The texture is predomin- antly granoblastic-polygonal, modified by a vague preferred orientation of biotite and horn- blende producing a weak foliation. Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 108 Samples R36265 and R36266: Basic granulites These are fine- to medium-grained grano- blastic-textured rocks containing andesine (roughly 40%), hornblende (30%) (X - ^ light brown, Y ^ brown, Z - dark green), ortho- pyroxene (10%), biotite (5%), perthitic ortho- clase (5%), quartz (5%), and (in R36266 only) clinopyroxene (5%). Opaque oxide, apatite, and zircon are minor constituents. Orthoclase and quartz occur as irregularly distributed poikilo- blasts up to 4 mm in diameter which corrode and enclose the other minerals suggesting metasomatic introduction. In R36265 some of the hornblende and biotite also occurs as poikilo- blasts enclosing pyroxene, plagioclase, and opaque oxide. In both samples plagioclase forms occasional porphyroblasts relatively free of inclusions. Orthopyroxene shows incipient alter- ation to fibrous green amphibole and minor chlorite along fractures and grain boundaries. Samples R36269, R36270, and R36271: Basic granulites These are medium-grained equigranular rocks composed of andesine (roughly 50%), horn- blende (25%) (X - light brown, Y brown, Z brownish green), orthopyroxene (15%), clinopyroxene (5%), and biotite (5%). Quartz and orthoclase are additional minor constitu- ents of R36269. Opaque oxide, apatite, and zircon are minor accessories. Orthopyroxene shows slight alteration to green amphibole and rare chlorite along fractures and grain boundaries. Plagioclase in R36270 and R36271 shows incipi- ent alteration to chlorite, calcite, and sericite. The texture is granoblastic-polygonal. A faint mineralogical banding is evident in R36269. Journal of the Royal Society of Western Australia, Vol. 59. Part 4, June, 1977. 109 15. — In search of the Dibbler, Antechinus apicalis (Marsupialia: Dasyuridae) by P. Woolley^ Communicated by B. K. Bowen Manuscript received 25 May, 1976; accepted 16 November, 1976 Abstract The Dibbler, Antechinus apicalis, was con- sidered to be extremely rare, if not extinct, prior to 1967 when two specimens were collected at Cheyne Beach, Western Australia. Since then several attempts have been made to collect more specimens, both in the locality in which they were rediscovered and in other localities in the vicinity. Altogether, only 7 Dibblers have been trapped and the only known habitat is one small area of scrub situated close to the fishing settlement and camping area at Cheyne Recently, another Dibbler was brought in by a cat on a farm near Jerdacuttup, Western Aus- tralia. Despite an intensive trapping effort on all uncleared areas, of land on or immediately ad- jacent to the farm, and in some nearby areas of bushland, no further specimens were obtained and the habitat of the Dibbler in this region remains unknown. Introduction. The Dibbler, Antechinus apicalis, had not been collected for 83 years when Morcombe (1967) captured two specimens at Cheyne Beach (also known as Hassell Beach), Western Australia. Since then several workers have attempted to collect further specimens of this apparently rare species, with little success. This paper records the results of all attempts to trap the Dibbler, both at Cheyne Beach and at other localities, that are known to the author, from the date of rediscovery. Trapping at Cheyne Beach January 1967 . — Using 10 traps specially con- structed to fit over the flowers of banksias, Morcombe (1967) captured 2 Dibblers in area A (Figure 1) at Cheyne Beach (locality 1, Figure 2). The traps were set for 4 consecutive nights commencing on 25th January. One female A. apicalis was found in a trap on the morning of the 27th and one male on the 29th. This repre- sents a trapping success of 5%. In addition to the Dibblers, 2 bush rats, Rattus fuscipes, were trapped. Morcombe kept the Dibblers in cap- tivity for several weeks to photograph them and observe their habits. April 1967 . — Ride (1970) captured a female Dibbler in area A (Figure 1) on 8th April. It was caught in a Sherman trap (23 x 8 x 9 cm) baited with ‘universal’ bait (Ride, pers. comm.). This animal, together with the two collected by Morcombe in January, were sent to the author in May 1967 for study of their reproductive biology (Woolley 1971). 1 Department of Zoology, La Trobe University, Mel- bourne, Victoria, 3000. August 1967. — On 1st August, Baynes and Kirsch (Baynes, pers. comm.) trapped a female A. apicalis in area A (Figure 1.) Twenty Sher- man traps (23 X 8 X 9 cm) were set for 4 nights and the animal was caught on the third night. This represents a trapping success of approximately 1%. The bait used was either ‘universal’ or beef mince or both. There were no young on the nipples but judging by the appearance of the pouch this female was suck- ling young; seven of the 8 nippies were elon- gated, the mammary glands were enlarged and the pouch fur was a reddish-brown colour. The weight of the animal was 78 g and the pes length 24 mm. It was released at the site of capture immediately after inspection. January /February 1970 . — Between 29th Janu- ary and 4th February Butler (unpublished re- port, Western Australian Department of Fisher- ies and Wildlife) trapped in three areas designated Major Area (area A, Figure 1), Coastal Strip (area B, Figure 1) and Hillside Area, the precise location of which cannot be determined from the report. Elliott (32 x 8 x 10cm), Sherman (23 x 8 x 9cm), cat (60 x 25 X 30 cm), breakback and pit traps (60 cm deep and 35 cm in diameter) were used. The bait used and the number and types of trap set in each area are not given but in a total of 400 trap-nights 5 Sminthopsis murina, 5 Tarsipes spencerae, 12 Rattus fuscipes, 15 Mus musculus and 19 reptiles were captured. No Dibblers were trapped. March 1970, — Burbidge (pers. comm.) trap- ped for 4 nights from the 12th in areas C and D (Figure 1). Twenty Elliott traps (32 x 8 x 10 cm), were set in each area. A bait containing peanut paste, sultanas, rolled oats and bacon was used. The only mammals caught were Mus musculus and Rattus fuscipes. November 1975. — With the objectives of ob- taining a pair of A. apicalis for further labora- tory studies of the reproductive biology of the species and of determining the distribution of the animals at Cheyne Beach the author trapped in 4 areas over a five-day period com- mencing on 25th November. Both small (16 x 5 X 6 cm) and large (23 x 8 x 9 cm) Sherman traps baited with bacon and peanut butter were used. Twenty large Sherman traps were set on 3 consecutive nights and thirty on the fourth night (a total of 90 trap-nights) in the known Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. Ill Figure 1.— Aerial photograph of locality 1 (Cheyne Beach) showing the six trapping areas (A, B, C. D, E and F). Commonwealth of Australia Air Photograph, 1971. Dibbler locality (area A, Figure D. Three A. apicalis, 8 Rattus fuscipes (minimum of 3 in- dividuals) and 1 lizard were trapped. Trapping success for A. apicalis was approximately 3%. Two A, apicalis, one male (number ID and one female (number 12) were captured on the first night of trapping at trap sites approximately 200 m apart. A second female (number 13) was trapped on the second night at the same trap site as female 12. The animals were not in breeding condition when captured; the male was not showing spermatorrhea and there were no young in the pouches of the two females. The body weight and pes length of each of the animals was as follows: — male 11. 60 g, 25 mm; female 12, 48 g, 24.5 mm; female 13, 73 g, 24.5 mm. Examination of the pouches of the females suggested that female 12 had not previously reared a litter (pouch fur pale, nipples minute) and was therefore probably less than one year old. while female 13 had reared a litter (pouch fur reddish-brown, nipples slightly elongated ) and was more than one year old. Male 11 was estimated, by comparison of the body weight with that of laboratory maintained males of known age (Woolley 1971), to be less than one year old. No Dibblers were captured in the 3 other areas in which trapping was carried out. In area D (Figure 1) no animals were captured in a total of 55 trap-nights over 2 consecutive nights using small Sherman traps. In area E (Figure D 1 Rattus fuscipes, 1 Isoodon obesulus and 3 lizards were captured in a total of 60 trap nights over 2 consecutive nights using large Sherman traps. The bandicoot, I. obesulus, which was found dead in the trap on 26th November, was a female with 2 young in the pouch. The three were lodged in the Western Australian Museum (numbers M14364-66). In area F (Figure 1) 50 small Sherman traps were set for 1 night only. One Mus musculus and 1 lizard were trapped. Trapping at other localities Trapping was carried out by the author at 5 other localities in the vicinity of Cheyne Beach in November and December, 1975 and, following the report of a specimen of A. apicalis brought in by a cat on a farm near Jerdacuttup, at another 4 localities in February, 1976. The owners of the cat, Mr. and Mrs. Gold- finch, who were presented with the dead Dibbler at their house on 24th January, 1976, lodged the specimen in the Western Australian Museum (number M13997). The specimen, which has been examined by the author, is an adult male with a body weight (in spirit) of 95 g and a pes length of 27 mm. Histological sections of one testis and epididymis have been prepared and, although the animal had been frozen before preservation in alcohol, spermatozoa could be recognised and were present in both the testis and epididymis. Large and small Sherman and Elliott traps (see above for dimensions) were used in the 1975 trapping period; in 1976 only large Sherman Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 112 and Elliott traps were used. The traps were baited with bacon and peanut butter. Location and reserve numbers of the localities in which trapping was carried out are from Western Australian Government Department of Lands maps. ‘‘Bulla Park” (Plantagenet Location 5310) — Locality 2, Figure 2. — Twenty large Sherman traps were set each night on 26th and 27th November on the road reserve adjacent to the property “Bulla Park” on Manypeaks Road. One Rattus fuscipes was trapped. Off Bluff Creek Road (Reserve No. 30033 — Mining) — Locality 3, Figure 2. — Trapping was carried out for 3 nights on 29th, 30th November and 1st December at this locality. In a total of 278 trap nights (131 using small and 147 using large Sherman traps) 3 Rattus fuscipes were trapped. “Bluff Creek” (Plantagenet Location 6502) — Locality 4, Figure 2. — Twenty small and 20 large Sherman traps were set on the night of 2nd December. No animals were trapped. “Umagalee” (Plantagenet Location 6481) — Locality 5, Figure 2. — No animals were trapped in a total of 96 trap-nights, using large Sherman traps, on the nights of 2nd and 3rd December. Off Cheyne Road (Vacant Crown Land be- tween Plantagenet Location 6501 and coast) — Locality 6, Figure 2. — Using 40 small and 40 large Sherman traps, a total of 160 trap nights in this area on the nights of 2nd and 3rd De- cember yielded 2 Rattus fuscipes and 1 male Isoodon obesulus. Tamarine Road — Locality 7, Figure 3 . — Forty Sherman traps were set along 1.6 km of the road reserve, and 50 Elliott traps in the adjoining property (Oldfield Location 829). In a total of 260 trap-nights over four consecutive days from 18th February, 25 Rattus fuscipes (minimum of 12 individuals), 2 lizards and 1 frog were captured. Flora and Fauna Reserve No. 31128, Jerda- cuttup North Road — Locality 8, Figure 3 . — Fifty Elliott traps were set for 3 nights from 20th February. Three Mus musculus and 2 lizards were trapped. Government Requirement Reserve No. 28110, Middle Road — Locality 9, Figure 3. — Forty Sher- man traps were set on the nights of 21st and 22nd February along part of the western and southern boundaries of this block. One Mus musculus was trapped. '‘Slieve Donard” (Oldfield Location 826) and adjacent land — Locality 10, Figure 3. — In this locality trapping was carried out in the five areas shown in Figure 4. Areas A and B are parts of the only two remaining areas of un- cleared land on “Slieve Donard”. Area C is part of the Jerdacuttup River Reserve; area D, Crown Land at the northern boundary of “Slieve Donard” and area E, the road reserve at the entrance to “Slieve Donard” on Tamarine Road. Sherman traps were used in all areas, and in area C Elliott traps also were used. No animals were caught in areas A (95 trap- nights over 5 consecutive nights from 18th February, B (120 trap-nights on the 18th, 19th, 23rd and 24th February), D (70 trap-nights on 23rd, 24th February) or E (20 trap-nights on 21st, 22nd February). In area C, 1 Rattus fus- Journal of the Royal Society of Weste; cipes, 2 Mus musculus and 2 lizards were caught in 203 trap-nights over 5 consecutive nights from 18th February. Choice of the trapping localities A botanical survey of area A at Cheyne Beach where Morcombe (1967) rediscovered the Dib- bler was carried out in 1970 by Butler (un- published report. Western Australian Depart- raent of Fisheries and Wildlife). The species listed for the area are Banksia attenuata, B. haxteri and B. coccinea, to a height of 2.5 m; Agonis hypericifolia, A. linearifolia, Adenanthos cuneata, Beaufortia micrantha, Cassytha sp, Jacksonia spinosa and Phyllota harhata to a height of 1.2 m and, in the undergrowth, Anarthria gracilis, Andersonia caerulea, Bur- chardia umhellata, Calothamnus gracilis, Casu- arina humilis, Dasypogon bromeliaefolius, Daviesia juncea, D. polyphylla, Haemodorum spicatum, Hakea ruscifolia, Hibbertia triandra, Hypocalymma strictum, Isopogon longifolius, Johnsonia lupulina, Lepidosperma sp, Leptocar- pus sp, Leucopogon (4 species), Lobelia tenui- fiora, Lysinema ciliatum, Melaleuca striata, Australia, Vol. 59, Part 4, June, 1977. 114 Figure 4. — Aerial photograph of locality 10 (“Slieve Donard” and adjacent land) showing the 5 trapping areas (A, B, C, D and E). Commonwealth of Australia Air Photograph. 1971. More land (indicated by diagonal white lines) has been cleared since this photograph was taken. Petrophile longifolia, P. rigida, Pimelea longi- fiora, P. rosea and Stylidium scandens. The vegetation is very dense and the ground litter thick, the area not having been burnt for many years. Figure 5 shows the appearance of area A at the site of capture of male 11 in November 1975. Much of the surrounding countryside was burnt in 1966 (Morcombe 1967) and is now largely covered by lower and less dense vegeta- tion than is present in area A. An exception is the “Coastal Strip” (area B) in which the species listed by Butler are Acacia decipiens, Agonis fiexuosa, Banksia occidentalis, Harden- bergia comptoniana, Lepidosperma gladiatum, Oxylobium lanceolatum, Rhagodia baccata, Sollya fusiformis and Spyridium globulosis to a height of 2.5 m; Agonis linearifolia, Beaufortia micrantha, Muehlenbeckia adpressa, Melaleuca striata and Scirpus nodosus to a height of 1.2 m. Areas C, D, E and F are patches of slightly taller and denser vegetation amid the generally low vegetation surrounding areas A and B. The composition of the flora in these areas has not been analysed. Because Dibblers have only been trapped in the one area at Cheyne Beach which has very dense vegetation dominated by species of Bank- sia a search was made in the vicinity of Cheyne Beach during November and December 1975 for other localities with similar vegetation. Three localities (3. 5 and 6) were selected and, super- ficially, one of these, locality 3, appeared to be similar to the known Dibbler habitat. Here the banksias extended southwards for about 1 km from the point of access on Bluff Creek Road in a narrow zone along the foot of a ridge. At “Umagalee” (locality 5) trapping was carried out in an approximately 200 ha block of un- cleared land adjacent to Flora and Fauna Re- serve No. 27157 and Waychinicup River Catch- ment Area Reserve No. 29883. Locality 6 on Crown Land north of Cheyne Road, was a small area on a hillside between cleared farm land and a swamp. In addition to these 3 localities some trapping was carried out in 2 others in the vicinity of Cheyne Beach. Morcombe suggested locality 2 (“Bulla Park”), a small area of sandplain on v/hich banksias were growing. Access to the area could not be arranged so trapping was only carried out on the fringe area road reserve. Locality 4 was a small area on “Bluff Creek” where the owner, when clearing the land some years previously, had caught an animal which he thought might be a Dibbler. When shown a Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 115 Dibbler he was sure that the animal he had found was not the same, so trapping was dis- continued after one night. Choice of the trapping areas near Jerdacuttup was largely determined by the presence of uncleared land. Trapping was carried out in all areas of uncleared land on or immediatey adjacent to the farm “Slieve Donard", on which a Dibbler was brought in by a cat, regardless of the type of vegetation. In addition to these areas three other localities in the vicinity were selected. Two (localities 8 and 9) were selected because they were sizeable areas of Crown Land with vegetation representative of the patches remaining in much of the largely cleared sur- rounding countryside. The third (locality 7) was selected because the vegetation had the same characteristics as that in the known Dibbler habitat, the tall dominant Banksia species here being B. baxteri and B. speciosa. Discussion Only 7 A apicalis have been trapped since the species was rediscovered in 1967, and the only known habitat is one small area of scrub situ- ated very close to a fishing settlement and camp- ing area. On four of the five occasions when traps were set in this area (A) Dibblers were caught; specimens being obtained in January, April and August 1967 and again in November 1975. The animals were caught in two types of traps, one of special design by Morcombe, and the larger sized Sherman trap. Although Butler was not using Sherman traps his failure to trap Dibblers in the area in late January and early February 1970 cannot be readily explained. Elliott traps, which are of only slightly greater dimensions than the large Sherman traps, would appear to be equally suitable for trapping the animals. Trapping success, which could only be calculated for 3 of the successful trapping periods, was low, the maximum being achieved by Morcombe (5%). It is possible that the small Sherman traps are not large enough for the animals to enter readily and this may have been a contributing factor to the lack of success in catching Dibblers in areas D and E, locality 1 in November 1975, when only small Sherman traps were used. At Cheyne Beach A. apicalis appears to be restricted to one small area (A). None have been caught in any of the other 5 trapping areas in the locality. The vegetation of area A is different from that in the other areas. Here the Banksias are taller and denser than elsewhere, and the ground litter thicker. Morcombe (1967) has commented on the ability of the Dibbler in captivity to climb, and to feed upon the nectar and possibly the insects attracted by the nectar Figure 5 —Dibbler habitat (area A at Cheyne Beach). The photograph shows the part of the area in which male 11 was captured. Journal of the Royal Society of Western Australia, Vol. 59. Part 4. June, 1977. ]16 of Banksia flowers. He has also noted their habit of rapidly burrowing beneath loose leaf litter when disturbed. These observations, together with those on the vegetation of the various trapping areas in the Cheyne Beach locality, suggest that A. apicalis may be dependent on the type of habitat found only in area A. If this is so then there is no obvious explanation for the lack of success in trapping Dibblers in locality 3, near Cheyne Beach, where the vegetation showed a remarkable similarity to that in area A, assuming that the species was widespread prior to the clearing of large areas of land. The habitat of the Dibbler in the Jerdacuttup region remains unknown. Despite an intensive trapping effort (over 1 000 trap-nights) covering all areas of uncleared land on or adjacent to the farm on which a specimen was brought in by a cat, and in other nearby localities, no Dibblers were captured. One possible explana- tion for the lack of success may be found in the timing of the trapping in relation to the breeding season. In the related species, A. stuartii, changes in trapping success through- out the year have been correlated with breeding activity (Woolley 1966). For this species trap- ping success was highest (about 15%) in the 2 to 3 months before the breeding season, and it declined to about 5% as the mating period approached. A. apicalis is known to mate in the laboratory in March and April and the little available evidence suggests that mating also occurs during this time in the field (Woolley 1971). Trapping in late February might there- fore be expected to be less successful than in earlier months in relation to the breeding sea- son. Further attempts to trap the Dibbler in the Jerdacuttup region should therefore be made at another time of the year. Further, although few comparative data are available, A. apicalis appears to be more difficult to trap than A. stuartii and it may be necessary to devise new trapping methods for greater success. Acknowledgements. — The author is grateful to the Department of Fisheries and 'Wildlife, Western Austra- lia for permission to trap and collect A. apicalis, and for making available a vehicle in February, 1976. I wish to thank Dr. Marilyn Renfree and Mr. R. Young of Murdoch University for assistance with the trapping at localities 5 and 6 in December, 1975 and Dr. A. A. Burbidge and Mr. W. K. Youngson of the Department of Fisheries and Wildlife for assistance in February, 1976. I also wish to thank Mr. W. Dunlop, Mr. G. White, Mrs. Boothey and Mr. and Mrs. I. Goldfinch for permission to trap on their properties. Special thanks are due to Mr. and Mrs. Goldfinch for the hospitality extended to Dr. Burbidge, Mr. Youngson and the author at their farm in February. Dr. Bur- bidge kindly commented on the manuscript. Financial support for the work was provided by the Australian Research Grants Committee. References Morcombe, M. K. (1967). — The rediscovery after 83 years of the Dibbler Antechinus apicalis (Mar- supialia, Dasyuridae), W. Aust. Nat., 10: 103-111. Ride, W. D. L. (1970). — ‘A Guide to the Native Mam- mals of Australia” Oxford University Press, Melbourne. Woolley, P. (1966). — Reproductive biology of Antechinus stuartii Macleay (Marsupialia: Dasyuridae). PhD thesis, Australian National University, Canberra. Woolley, P. (1971). — Observations on the reproductive biology of the Dibbler Antechinus apicalis (Marsupialia: Dasyuridae). J. Roy. Soc. West Aust., 54: 99-102. Journal of the Royal Society of Western Australia, "Vol. 59, Part 4, June, 1977. 117 16._Distribution and function of resins and glandular hairs in Western Australian plants by B. DelP Manuscript received 22 June 1976; accepted 19 October 1976 Abstract The taxonomic distribution of glandular hairs and resins is documented. Resinous plants are prevalent in some genera within the Mimosaceae, Euphorbiaceae, Sapindaceae, Boraginaceae, Dicrastylidaceae, Lamiaceae, Myoporaceae, Solanaceae, and Goodeniaceae. With few exceptions there is a correlation between surface resin and glandular hair dis- tribution. The genus Eremophila is discussed as representative of a resinous, arid genus. Some properties of the leaf resins of Beyeria viscosa and Eremophila fraseri are discussed in detail. Resins may have a function in reducing water loss by increasing resistance to cuticular trans- piration and by reducing leaf temperature by increasing radiation reflectance from the leaf. Introduction A systematic treatment of glandular hairs and leaf resins in Western Australian plants has not been attempted previously. Interest in resin formation in some species (Dell and McComb 1975) and the possibility that plant resins may be of use in difficult taxonomic groups (Dell 1975) led to an investigation of the relationship between glandular hairs and surface leaf resins. The significance of resins in plants has remained an enigma, proposed functions having little experimental proof. In Beyeria viscosa the distribution of the resin on the leaf surface varies according to leaf maturity and is closely tied to the early stages of glandular hair forma- tion (Dell and McComb 1974). Incidptal observations that this pattern of resin distri- bution could be altered by temperatures not lethal to some plants, led the author to evaluate the possible importance of the resin in increas- ing reflectance of light from the leaves. Pearman (1966) has indicated the importance of surface features such as hairs and scales in increasing reflectance. Slatyer (1964) and Waggoner (1966) have noted that the possession of shiny leaf surfaces could probably reduce the heat load by 10-15% under stress conditions. This factor together with the high heat resistance of some Western Australian plants (e.g. up to 59°C., Grieve and Hellmuth 1968) could be of impor- tance to plants subject to irregular and often prolonged droughts. Distribution of glandular hairs and resins Collections of plants bearing resins and/or glandular hairs were made in the field. Voucher specimens are housed in the University of West- ern Australia (UWA) (see Appendix I). Sections iSchool of Environmental and Life Sciences, Murdoch University, Murdoch, Western Australia, 6153. of fresh, preserved and in some cases, dried herbarium material, were examined and types of trichomes and their distribution recorded. The results are summarized in Table 1 and a few of the trichome types are illustrated in Fig. 1. It is apparent that the majority of resinous genera are either woody or herbaceous dicoty- ledons. Nearly all plants with external resinous exudations bear glandular hairs beneath the exudate. It can be assumed that the glandular hairs in these species are implemented at least in resin secretion and perhaps also in resin synthesis. Exceptions include some taxa of the Myrtaceae, Celastraceae, Fabaceae, Poaceae and Haemodoraceae. The secretion sites of sticky exudates in some species of Calytrix, Pileanthus, Psammomoya and Burtonia need to be investi- gated further. There is a possibility that epi- dermal cells have a glandular function in these genera. Not all plants with glandular hairs secrete resins (Table 1): some glandular hairs are pig- mented (e.g. Diplopcltis) , others produce volatile oils (e.g. Anthocercis) , mucilages etc. In some plants glandular hairs are confined to the inflorescences (e.g. members of the Proteaceae); in others the trichomes are confined to the leaves, phyllodes and stems (e.g. Acacia) or may occur on both the leaves and the flowers (e.g. Eremophila, Stylidiumi . Western Australian plants with resinous sheets are prevalent in some genera within the Mimosaceae, Euphorbiaceae, Sapindaceae, Bora- ginaceae, Dicrastylidaceae, Lamiaceae, Myopor- aceae, Solanaceae and Goodeniaceae. The genus Eremophila is an example that illustrates the prevalence of resinous species in dry habitats. Approximately 70% of the species occur in Western Australia where they are most abundant in the north and interior regions of the State. About 43% of these have resinous leaves and stems. The resin may exist as a continuous varnish over the leaves (e.g. E. fraseri, E. serrulata) , be confined to one surface (e.g. E. latrobei) , or exist as isolated patches (e.g. E. angustifolia, E. duttonii) . This variation is reflected in the amount of resin expressed as a percentage of leaf dry weight in Table 2. Species with high resin yields usually have con- tinuous sheets of resin, at least on the young leaves. Surface resins in all Eremophila species are produced by glandular hairs. The nearly uni- versal glandular hair has a short stalk with up to eight cells in the head. Many of the species Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 119 Table 1 Occurrence and taxonomic distribution of glandular hairs and resinous plants in IVestern Australia. Voucher specimens are cited in Appendix I. Family Examples Distribution of glandular hairs Surface features* Poaceae Triodia 7 resinous Orchidaceae Elythranthera leaves and stems — Caladenia ... inflorescences — Liliaceae Agrostocrinum inflorescences — Haemodoraceae Conostylis 7 resinous leaf edges Proteaceae Adenanthos inflorescences viscid Grevillea inflorescences viscid Chenopodiaceae Chenopodium leaves and stems — Nyctaginaceae Boerhavia .... leaves and stems — Gyrostemonoaceae Didvmotheca leaves and stems resinous Capparaceae Cleome leaves, stems, and inflorescences viscid Droseraceae Drosera leaves and stems specialized for insect- ivory Byblidaceae Byblis leaves and stems specialized for insect- ivory Mimosaceae Acacia young leaves .... viscid, resinous Caesalpiniaceae Cassia rare on leaves — Fabaceae ? Genus leaves, stems and inflorescences resinous Burtonia 7 resinous Tremandraceae Tetratheca .... rare on young stems — Euphorbiaceae .. Bertva leaves and stems resinous Beveria leaves and stems resinous Ricinocarpos leaves and stems viscid Celastraceae Psammomoya 7 resinous stems Sapindaceae Diplopeltis .... mainly inflores- cences ■ — ■ Dodonaea . .. leaves and stems resinous Malvaceae Abutilon leaves and stems — Hibiscus leaves and stems rarely viscid Myrtaceae Eucalyptus .... rare on leaves 7 Calvtrix 7 viscid Pileanthus .... 7 viscid Plumbaginaceae Plumbago .... inflorescences, fruits ± viscid Boraginaceae Halgania leaves and stems resinous Dicrastylidaceae Chloanthes .... leaves and stems C vanostegia leaves and stems resinous Dicrastrlis .... leaves and stems — Lachnoslachvs leaves and stems — Newcastelia leaves and stems viscid, resinous Pitvrodia leaves and stems — Avicenniaceae - Avicennia .... leaves and stems specialized for salt secretion Lamiaceae Hemigenia .... leaves and stems — Prostanthera leaves and stems resinous Solanaceae Anthocercis leaves and stems resinous Nicotiana .... leaves, stems and inflorescences viscid Scrophulariaceae Gratiola leaves and stems viscid Stemodia .... leaves and stems viscid Verbascum .... leaves, stems and inflorescences — Veronica rare on leaves — Orobanchaceae Orobanche .... stems .... viscid Lentibulariaceae Utricularia .... rare on traps specialized for insect- ivory Myoporaceae Eremophila leaves, stems and inflorescences resinous Mvoporum . leaves and stems resinous Goodeniaceae .. Calogvne leaves and stems — Coopernookia leaves and stems resinous Goodenia leaves and stems viscid, resinous Scaevola leaves and stems — Stylidiaceae Stylidium . . inflorescences, rare on leaves viscid Asteraceae Brachvcome leaves and stems — Helichrysum leaves and stems — Helipterum leaves and stems — Ixiolaena leaves and stems — Olearia leaves and stems resinous Pluchea leaves and stems — * resinous — leaf has a continuous or broken layer of resin, viscid — sticky; resin, when present, confined to glandular hairs, where the glandular hairs do not produce viscid or resinous material this is indicated as — . which do not appear resinous, for example species with a dense tomentum of stellate or branched hairs, also have an understorey of stalked glandular hairs (Fig. 1). Large branched hail’s with some glandular tips are scattered through the tomentum of similar but non- glandular trichomes of E. leuco^hylla and E. turtonii. Distribution of resin on the leaf surface Whilst observing Beyeria leaves under bright lights it was noticed that the surface of the leaf changed from matt-like to mirror-like in appearance (Fig. 2). It was possible to see reflec- tion of images from the mirror surface. Leaves continued to grow after this transformation and presumably were not affected internally by the additional radiant heat. When heated in an oven it was found that at 55 °C. the abaxial sheet of resin coalesces within two minutes. At 51 °C. the youngest leaves form a sheet in about the same time but the half-expanded leaves take up to five minutes to achieve the same resin flow. Twelve minutes exposure at 44°C. causes the resin to run on young and mature leaves as at the higher temperatures. Resin, removed from the leaf surface, melts at about 48°C. to form a thick, viscous liquid. The effect of radiant heat on the leaf surface causes the resin to become mobile and, on the younger leaves, the resin flows together on the abaxial surface and forms a continuous sheet with a smooth surface. On mature leaves the resin droplets are widely spaced and the effect of heat causes the resin to coalesce into ‘rivers’. The composition of the resin on the leaf sur- face is probably determined by genetic factors whereas the amount of resin on the leaf surface is a combination of genetic and environmental factors and is closely related to the distribution and abundance of glandular hairs. The distri- bution of the resin on the leaf surface is depend- ent on such factors as resin composition, amount of resin/unit area and surface topography. If the resin is of adaptive value to the plant, it might be expected that plants growing under some stress conditions would produce more glandular hairs and hence more resin than plants growing under mesophytic conditions. New shoots of plants transferred from the field to glasshouses were always softer and had fewer glandular hairs than those at the time of removal. Reflectance of light from leaves The technique of Pearman (1966) was used. A Bausch and Lomb Spectronic 20 colorimeter with an integrating sphere-reflectance attach- ment was used for measuring the total reflected visible radiation from leaves. The colorimeter produces wavelengths from 340 to 620 nm in bands of 20 nm width, and these were directed onto the leaf with an angle of incidence of 0°. The reflectance spectrum from Beyeria leaves produced a maximum at 560 nm for both sur- faces. Heated leaves showed a small increase in total reflectance especially towards 540 nm. However, for any one wavelength the difference was less than 5%. Removal of resin from the Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June. 1977. 120 Figure 1.— Glandular hairs of some of the genera referred to in Table 1. A.—Helichrysmn rosea. B.~Grevillea eriostachya. C. — Acacia aneura. D. — Acacia glutinosissima. E. — Anthocercis littorea. P . — Scaevola canescpn-^ n Scaevola glandulifera. B..-~Eremophila leucophylla, stellate and shortly stipitate glandular hairs are shown I — Pityrodia hartlingii. Length of bar 50^m. on^wii. x. Journal of the Royal Society of Western Australia, Vol. 59. Part 4. June, 1977. 121 leaf surface of Eremophila fraseri caused a con- siderable reduction in total reflectance from the leaf (Fig. 3). An increase in reflection of light by the resin layer decreases the amount enter- ing the leaf and subsequently being absorbed. It was originally thought that the mirror-like surface would reflect more light than the un- heated surface. Reflectance from a surface, as opposed to reflectance from within a surface, is dependent on surface features alone. When Figure 2. — Effect of heating on the appearance of Beyeria leaves. Control is on the left. measuring total reflected light from a surface with an integrating sphere (see above) the specular (or mirror) reflectance could be close to diffused reflectance depending on the angle of incidence. The control Beyeria leaves have irregular-shaped resin deposits and for low angles of incidence, on leaves naturally held towards the vertical, a considerable amount of diffused reflectance could be directed into the leaf. The distribution and optical properties of the resin can be compared with glaucous and non-glaucous eucalypt leaves where the orienta- tion and type of wax deposits affect the amount Figure 3. — Effect of resin on the reflectance of light from Eremophila leaves. (• — abaxial surface, o — abaxial surface after resin wash, difference between reflectance spectra of control and washed leaf. The graphs are for one leaf. Reflected radiation is expressed as a percentage of the reflectance from magnesium carbonate.) Table 2 Resin yields of some Eremophila species Species Locality Collection No. Resin (% leaf dry wt.) Distribution of resin on leaf surface alternifolia Sandstone 1015 4-5 isolated patches decipiens .. . Boulder . . 1060 18 sheet dempsleri ... Norseman 1066 10 broken sheet drummondii .... . .. Boulder . .. 1058 17 sheet dunonii Wiluna . 1036 8 isolated patches fraseri Yalgoo 56 17 sheet foHosissima Agnew . . 1025 22 sheet georgei Agnew, Leonora 1022. 1051 17-5 sheet granilica .... Sandstone. Leonora 1013. 1048 29-5 sheet interstans . .. Boulder .. 1057 22 sheet latifolia Agnew .... .... . . 1024 18 sheet meiallicoriiin Leonora .... 1046 18 broken sheet nnniata Menzies 1052 7 isolated patches plat veal \'x Barwidgee Station 1038 21 sheet saligna Norseman 1065 3 scarce over glandular hairs serrulata Leonora .. 1078 15 sheet scoparia Payne's Find, Boulder 1008, 1055 1 -5 not resinous willsii Agnew . .. 1027 12 broken sheet Journal of the Royal Society of Western Australia. Vol. 59, Part 4, June. 1977. 122 of light reflected from the leaf (e.g. Cameron 1970). The surface construction of leaves can be important in the reflection of wavelengths other than those measured above. Gates and Tantra- porn (1952) indicate that 80% or more of the infrared radiation is effectively reflected from the outer epidermal surfaces. Wong and Blevin (1967) showed that surface hairs and dry vesic- ular tissues were responsible for slightly higher infrared reflectances in several species. Surface features were shown by Pearman (1968) to reflect an appreciable proportion of the visible spectrum. The distribution of the resin over the leaf surface is important if it is to be considered that the resin has a function in reducing water loss. Reduction in water loss by resins could be achieved in two ways. Firstly, the presence of a sheet of resin over the leaf surface must in- crease resistance to cuticular transpiration. Gardner (1968) considered that resinous leaf coverings in species of Eremophila protect the leaves from the drying influence of wind. Secondly, the presence of resin does not qualita- tively alter the wavelengths of light available to photosynthesis though it may reduce the amount of light entering the leaf. e.g. in Eremo- phila. In this way transpiration could be reduced by a slight lowering of leaf temperature. Acknowledgements . — I would like to thank Dr. A. J. McComb for helpful discussion; C.S.I.R.O. and the University of Western Australia for financial support; colleagues and friends for collections of plant material. References Cameron, R. J. (1970). — Light intensity and the growth of Eucalyptus seedlings II. The effect of cuticular waxes on light absorption in leaves of Eucalyptus species. Aust. J. Bot., 18: 275-84. Dell, B. (1975). — Geographical differences in leaf resin components of Eremophila fraseri. F. Muell. (Myoporaceae). Aust. J. Bot., 23: 889-98. Dell, B. and McComb, A. J. (1974). — Resin production and glandular hairs in Beyeria viscosa (Labill.) Miq. (Euphorbiaceae). Aust. J. Bot., 22: 195-210. Dell, B. and McComb, A. J. (1975). — Glandular hairs, resin production and habitat of Newcastelia viscida E. Pritzel (Dicrastylidaceae). Aust. J. Bot., 23: 373-90. Gardner, C. A. (1968). — Wildflowers of Western Aus- tralia. West. Aust. Newspapers Ltd., Perth. Gates, D. M. and Tantraporn, W. (1952). — The reflec- tivity of deciduous trees and herbage plants in the far infrared to 25 microns. Science, 115: 612-16. Grieve, B. J. and Hellmuth, E. O. (1968). — Eco-physio- logical studies of Western Australian plants. Proc. ecol. Soc. Aust., 3: 46-54. Pearman, G. I. (1966). — The reflection of visible radia- tion from leaves of some Western Australian species. Aust. J. Biol. Sci., 19: 97-103. Pearman, G. I. (1968). — Studies on leaf energetics. Un- published Ph.D. Thesis, University of West- ern Australia. Slatyer, R. O. (1964). — Efficiency of water utilization by arid zone vegetation. Ann. Arid Zone, 3: 1-12. Waggoner, P. E. (1966). — Decreasing transpiration and the effect upon growth in “Plant Environ- ment and Efficient Water Use’’, Am. Soc. Agron. and Soil Sci. Am. Wong, C. L. and Blevin, W. R. (1967). — Infrared reflec- tances of plant leaves. Aust. J. Biol. Sci, 20: 501-8. Appendix 1 Voucher specimens Except where stated otherwise all specimens are housed in the University of Western Aus- tralia and specimens are cited by accession numbers. Triodia pungens 2203; Elythranthera brunonis 2202; Caladenia discoidea 2201; Agrostrocrinum scahrum 2197; Conostylis aurea 2196; Adenanthos meissneri 2199, A. venosa 2200; Grevillea eriostachya 2062, G. excelsior 2198, G. petrophiloides 2052; Chenopodium plantaginel- lum 2195; Boerhavia repandra 2194; Didymotheca thesioides 2193; Cleome viscosa 2192; Acacia denticulosa MURD 27, A. glutinosissima MURD 26, A. kempeana 2190, A. ramulosa 2189, A. rossei 2053, A. tetragonophylla 2188; Cassia phyllodinea 2180, 2181, 2182, 2183; Burtonia scabra 2179; Beyeria drummondii 2177, B. leschenaultii 2176; Ricinocarpos velutinus 2178; Psammomoya choretroides 2048; Diplopeltis huegelii 2175; Dodoiiaea attenuata 2206, D. boroniaefolia 2205, D. bursariifolia 2208, D. caespitosa 2207, D. concinna 2215, D. filifolia 2212, D. inaequifolia 2209, D. larraeoides 2210, D. pini- folia 2214, D. ptarmicifolia 2213, D. stenozyga 2063, D. viscosa 2211; Eucalyptus citriodora MURD 28; Calytrix glutinosa 2173; Pileanthus filifolius 2172; Plumbago zeylanica 2171; Halgania cyanea DELL 127 (PERTH), H. lavendulacea DELL 167 (PERTH), H. viscosa DELL 84 (PERTH), Halgania sp. DELL 113 (PERTH); Chloanthes coccinea 2168; Cyanostegia angustifolia 2054, C. lanceolata 2055, C. microphylla 2056; Dicrastylis micrantha 987; Lachnostachys bracteosa 2170, L. cliftonii 2169; Newcastelia viscida 2057; Pityrodia bartlingii 2167; Hemigenia divaricata 2166; Prostanthera eckersleyana 2164, P. grylloana 2165; Anthocercis aromatica 2159, A. littorea 2217, A. viscosa 2610, DELL 1002; Nicotiana occidentalis 2161, N. rosulata 2162; Verbascum virgatum 2163; Eremophila alternifolia 2086, 2087, E. angustifolia 2088, 2090, E. clarkei 2091, 2098, E. compacta 2073, E. decipiens 2103, 2105, 2109, E. delisseri 2070, E. drum- mondii 2100, E. duttonii 2069, 2102, E. eriocalyx 2072, 2110, E. exilifolia 2111, E. foliosissima 2112, 2113, 2114, E. fraseri 2058, 2060, 2061, E. freelingii 2068, E. georgii 2115, 2119, 2125, E. glabra var. viridifiora 2050, E. granitica 2126, 2127, E. hughesii 2128. E. interstans 2075, E. latifolia 2131, E. latrobei 2065, 2130, E. leucophylla 2134, E. longifolia 2132, 2133, E. macmilliana 2066, 2135, 2136, E. margarethae 2071, E. miniata 2059, E. oppositifolia 2138, 2139, E. platycalyx 2140, 2141 E. platythamnos 2142, E. punicea 2074, E. pustulata 2143, 2144, E. ramosissima 2146, E. saligna 2147, E. scoparia 2051, E. serrulata 2148, 2149, E. spathulata 2150, E. sub- fioccosa 2151, E. virens 2049, E. woollsiana 2152, 2153, E. youngii 2067; Myoporum deserti 2158; Calogyne berardiana 2156; Coopernookia polygalacea 2085; C. strophiolata 2080; Goodenia glandulosa 2084, G. pin- natifida 2079, G. viscida 2081; Scaevola glandulifera 2077, S. restiaceae 2076; Stylidium spathulatum var. glandu- losum 2204; Brachycome sp. 2155; Olearia muelleri 2047. Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 123 17. — Australites from northern Western Australia by R. C. Horwitz^ and D. R. Hudson^ Manuscript received 19 Octo'ber 1976; accepted 13 December 1976 Abstract Recent discoveries of tektites in the Patterson Ranges and western Pilbara regions indicate that australites are more abundant in northern Western Australia than was previously believed. They include the compositional varieties of normal australites and philippinites, indo- chinites, and high-Mg tektites, which supports the existence of a NNW-SSE trending primary distribution pattern extending from the Philippines through northern Western Australia to southern Australia. The present distribution of tektites in Western Australia has been influenced by geomorphological features. This has resulted in numerous tektite occurrences in areas such as the Eastern Goldfields, where concentration has occurred in playas, and relative scarcity of tektites in areas of fast drainage and dissection, such as the Ashburton River valley. Areas of sand cover or areas of recent sediment deposition also contributed to the apparent absence of tektites in northern Western Australia. Introduction Tektites are natural objects of silica-rich glass found in thousands on the surface of certain parts of the Earth. Baker (1959, p. 13) lists eight recognised true tektite provinces; those from Australia are designated australites and are believed to result from a single fall. Radiometric age data for australites are reviewed by Lovering et al (1972, p. 409) and the fall has been dated as about 700 000 years B.P. These authors also review (p. 408) and present new evidence concerning the relative age of the fall to geological and morphological features in the Quaternary. The genesis of shapes and inner structures of australites is described and classified by Baker (1959). Their mode of preservation in present- day fields varies according to occurrences; i.e. well preserved in some scattered occurrences, to abraded pebbles in placer deposits and to dreikantered pebbles in desert areas of central Australia, such as the occurrences described by Johnson (1965) at Lake Wilson near the junc- tion of the three states (W.A., S.A., N.T.). Aus- tralites have a strong mythological association for the Australian Aborigines of parts of West- ern Australia (R. C. Gould, pers. comm. 1965) and are found at aboriginal camp-sites and water holes, sometimes chipped and worked. They occur most abundantly in the banks and on the floors of playas and internal drainage flats, where they are concentrated by drainage, in association with quartz pebbles and rocks of similar density. Baker (1959, p. 31) estimates that 30 000 to 35 000 australites have been found. Cleverley and Dortch (1975, p. 243) refer to a “contin- ental line of known occurrence” which they draw through about south of Geraldton to Lake MacKay and to south of Brisbane; this general northern boundary to the province (Figure 1) had been accepted following Baker (1959, p. 18). Cleverley and Dortch (1975) record that 23 iCSIRO, Division of Mineralogy, Floreat Park, Western Australia, 6014. occurrences of single australites exist north of this line in Western Australia; their paper dis- cusses six australites found at archaeological sites of the eastern part of the Kimberley region. Chapman et al (1964) studied specific gravities for tektites from Australia, Indonesia and Indo- china, grouping several tektite provinces under the heading Australasian tektites. Chapman and Scheiber (1969) established zones with varying chemical characteristics in Australasian tektites attributed by Chapman (1971) to a single fall erupted from Tycho (a Lunar Crater), with a spread on earth containing “streaks” of com- positionally distinct tektites. This general spread differs from the distribution pattern accepted by Cleverley and Dortch (1975) in that it includes northern Australia, and in particular crosses northern Western Australia between about Exmouth and Derby. This note is published to record that — (a) australite occurrences are more abundant in northern Western Australia than was previously believed; (b) in agreement with Chapman (1971) their distribution is best explained as primary in origin and related to NNW-SSE depositional streaks rather than to transport by man, as was believed by Cleverley and Dortch (1975); (c) to account for their relative absence from large areas of northern Western Australia; (d) to record some chemical compositions for northern Western Australian tektites and com- pare them to Chapman’s model. Australites in northern Western Australia A minimum of nine australites, some intact, others worked, were found recently during exploration in the Patterson Ranges (Figure 1). A tektite was collected in 1974 near the mouth of the Sherlock River and some were found in the Teichmans-Pilbara townsite area (125 km ESE of Roebourne) during exploration in 1975 (M. J. Fitton, pers. comm. 1975). According to aboriginal stockmen, they occur in the Pilbara “north of Tom Price but are essentially absent from the Ashburton River valley”. This river is conspicuous by its relative fast erosion com- pared to the areas of classical australite fields in the Eastern Goldfields. The general dissection and fast drainage could have removed tektites and the surfaces on which they were deposited in most areas of the Pilbara and the Kimberley region of Western Australia and could contribute to their relative scarcity in these areas. Our observations in central Australia suggest that tektites get concentrated, as in the Eastern Goldfields and Lake Nabberu-Lake Carnegie regions, in small playas, resulting from natural damming on areas of bedrock (the Lake Wilson occurrence, described by Johnson 1965, is a typical occurrence. Here one of us (RCH) in 1965, found 53 small australites in 25 minutes over 77 m^ marginal to the salt pan, in optimum conditions, with the back to the sun, as recom- Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. 125 mended by Baker 1959, p. 29). They are, how- ever, not found in such concentrated patches in playas in interdunal hollows where the catch- ment area is smaller. A combination of sand cover or erosion with removal to the ocean (or burial in coastal deposits) could be sufficient to produce the apparent absence of tektites in northern Western Australia. Distribution of compositional groups of tektites Chapman (1971) has described the geographic distribution of tektite compositional groups for the southeast Asia-Indonesia-Australia region. He recognised a serious of compositional “streaks” with a general NNW-SSE trend. The most prominent of these in Australia are the high-Ca streak which extends from Alice Springs to the west coast of Tasmania, and the high-Mg streak which curves its way from the northern Philippines through Borneo and Java to cross the Australian coast between Exmouth and Derby and then extends SE towards Ade- laide. These high-Mg tektites together with a ubiquitous population of normal australite- philippinites (some of which approach the high- Ca group) comprise the described tektite popu- ation of Western Australia. With the exception Figure 1. — Map of Western Australia showing geomorphological divisions that have influenced the present dis- tribution of australites. The areas without fast erosion and without sand-dune cover equate broadly with Salinaland and Euclonia of Jutson (1950, p. 22), Nos. 1-12 indicate the location of the analysed tektites referred to in the text. Journal of the Royal Society of Western Australia, Vol. 59. Part 4, June, 1977. 126 of one high-Mg tektite from the Pilbara, all previously analysed Australian tektites reported by Chapman (1971) come from below the line described by Cleverley and Dortch (1975). Compositions of tektites Twelve tektites from localities in central and northern Western Australia have been analysed using the electron microprobe. The analyses were made in order to compare the composi- tions of the new tektite occurrences with those predicted from Chapman’s (1971) distribution model. All analyses were made on polished chips of tektites mounted in epoxy resin. Refractive indices were determined on the chips using a Raynor gem refractometer, and specific gravtiy determinations were made by hydrostatic weigh- ing. Operating conditions for the electron micro- probe are given in Table 1. With the exception of analyses 7 and 8 the totals are generally low, and this combined with a low and variable sodium determination (prob- ably due to the sodium loss under the focussed electron beam) renders the analyses semiquan- titative. Nevertheless, the element ratios are believed to be significant, and enable a classi- fication of the tektites to be made (Figure 2) into the compositional groups of Chapman and Scheiber (1969). Figure 2. — Compositional grouping of tektites, based on the classification of Chapman and Scheiber (1969). Analyses 1 to 9 are typical of the composition of “normal australites and philippinites” with MgO contents ranging from 2.13 to 2.57 weight percent. All three of the analysed tektites from the Lake Nabberu general area (including two tektites from the Canning Stock Route), three tektites from the Patterson Range and a tektite from the Kimberley district are thus of normal australite composition. Analysis 10 is of a tektite from the Kimberleys that falls within the “indo- chinite” compositional group, with low MgO and CaO contents of 1.27% and 1.50% respectively. Both analyses 11 (from Kimberleys) and 12 (from the Pilbara) are “high-Mg” tektites with MgO values of 3.38 and 3.73%. The new analyses thus confirm the continuity of the high-Mg streak through the Pilbara- Canning Basin-Kimberley districts of Western Australia, and increase the known distribution of normal australite-philippinites to northern Western Australia. The occurrence of an indochinite (analysis 10, from Miriwun) in the Kimberley district is puzzling. In Chapman’s study this compositional group was found to be restricted to a small tear-shaped streak over southeast Asia. It is not inconsistent with Chapman’s overall thesis that the indochinite distribution could extend to Australia, but the fact that the Miriwun tek- tite is a glass chip from an aboriginal site means that in this instance we cannot rule out the possibility of transport by man. Acknowledgements. — The authors are pleased to acknowledge the generosity of Mr. P. Robson for the Sherlock tektite, Dr. D. Tyrrwhitt for samples and information on the Patterson Range tektites, and Mr. C. E. Dortch for tektites from aboriginal sites. Mr. K. Gayski made replicas of tektite specimens that were to be damaged during analysis; Mr. W. Cleverly provided information from an unpublished tektite study; Dr. C. R. M. Butt and Mr. T. D. Pearce provided data for the compilation of the map; Mr. C. R. Steel drafted the figures. References Baker, G. (1959). — Tektites. Mem. nat. Mus. Viet. No. 23, 314 pp. Chapman, D. R. (1971). — Australasian tektite geographic pattern, crater and ray of origin, and theory of tektite events. J. Geophys. Res. 76: 6309-6338. Chapman, D. R., Larson, H. K., and Scheiber, L. C. (1964). — Population polygons of tektite specific gravity for various localities in Aus- tralasia. Geoch. Cosmoch. Acta. 28: 821-839. Chapman, D. R., and Scheiber, L. C. (1969).— Chemical investigation of Australasian tektites. J. Geophys. Res. 74: 6737-6776. Cleverly, W. H., and Dortch, C. E. (1975).— Australites in archaeological sites in the Ord Valley, W.A. Search 6: 242-243. Lovering, J. F., Mason, B., Williams, G. E., and McColl, D. H. (1972). — Stratigraphic evidence for the terrestrial age of australites. J. Geol. Soc. Australia 18: 409-418. Johnson, J. E. (1965). — Geological factors in tektite dis- tribution, northwestern South Australia. Quarterly Geol. Notes. Geol. Surv. South Australia 14: 5-6. Jutson, J. T. (1950). — The physiography (geomorph- ology) of Western Australia. Geol. Surv. West. Australia, Bull. 95 (third edition) 366 pp. Mulcahy, M. J. (1961). — Landscapes, laterites and soils in south-western Australia, in Jennings, J. N., and Mabbutt, J. A. (Eds.), “Landform Studies from Australia and New Guinea” 211-230. Aust. Nat. Univ. Press, Canberra. Journal of the Royal Society of Western Australia, Vol. 59, Part 4, June, 1977. ]27 o y *-• c O O VO OO pPl (N oo VO O < CO OO — OOOOPPlOiXO — •'t CV (N — r^opn — inr^oovr^ 'ct >0 CO oofS'XO'^r'i — — O' rJ — I r' — o\ (N — CSOa VO pp5r^Xppior^ — ooooo in r^O— •w^opopoofs-^ vO — C\ r-i — — OOO'O'OC — O — (Ti m VO — r^ooppioop*CTfscovnmr-ooori-ppi o o o 'oxfNr^Or'ji/^ — O'r-i ■'ct m inOOPO© — — O — lo r- — cv ni — o Tj- lO 3 OO — Tj-0r'|pp50r'l«n rs — < — Q\ c y — m — i/vo'^ ra U. (11 ^ rr U< cn 2 ■«.^ c E rt ■«i (U I c u ' u [T-iJ _ o ■J T3 'w ' (U on • — c/1 H 3 CO f- U- ■E> c/3 - t/) • -"E P 3 2S oy J_ 4> a"’ o C _c (j o Cl. t< >6 5- << C- >o •■<£: -^dr 0®-3 a. c < 2 CC “ Ci 3 «J fl OJ ^ 0.-JU J3 u c O J= t i_ Cl O O. (ij O-Z. L. cc ^ZC « • LU X O « J- T3 a 3 a.-3 £ £ CO on rt n d ci: w nj SI ° i — fN Cfl •— 4> C3 3 2 = i •“ s Is o .- o o ^ 3 « P cj _«j - d 2 X) “ 3 X oo JS P "i” 00 0) — E .. O u lA — 3 C5 E ^ u O £ = c o . jO c/5 CC 3 C3 ^ £ CO C 'a p 5 I C/^ — C^ « X -2P n c X -3 00 IT - - -P 43 W — T3 -3 U 0> ■3 ”3 «5 S3 u '_ J3 J3 < < _a w E E CO o S3 .L- ■3 3 4> -o ^ < o *2 c/5 C5 o5 «3 S3 £ o c S3 O ca c 'm 0 3 p CO o “ ch o «j . , •2 E .3 0> ^ -E 2 3 a S3 t/2 _ 3 CB E 2 o C 3 •E -o o .y c/5 «-• a a uj S j: p "O ^ .E o i X o — a • - -- (U ■C — o 'd t/5 (A V S3 M to x: 2 -~ j= S .5? V x: « CL £ 3 i/} U rn 00 00 m po ^ ^ — in pn — Tt OOOOOfsr^r-iooooocr^