) COMPARISON OFARTHROPOD SPECIES RICHNESS IN EASTERN AND WESTERN AUSTRALIAN CANOPIES: A CONTRIBUTION TO THE SPECIES NUMBERDEBATE J.D. MAJER, H.F. RECHERAND AC. POSTLE Major, J.D., Recher, U.V. & Postle, A.C. 1994 06 30: Comparison of arthropod species richness in eastern and western Australian canopies: acontribution lo the species number debate. Memoirsofthe QueenslandMuseum 36(1): 121-131. Brisbane. ISSN 0079-K835. mm Apart I forest pest species,ourknowledgeofEucalyptuscanopy arthropods isrudimen- tary.Thishascontributedloalackofapprcctaiionofthedifference*inarthropodabundances, biomass and richness on different species ofcucalypts and in different forests throughout Australia. A three year chemical knockdown study has been earned oui in one western Australianforest, wherejarrahEucalyptusmarginataandmarri E tulophyllawere sampled andoneeastern Australian forest,where narrow-lea\t:*d ironb.irk /:, < nhroand givv bOJI £ moluucanu were sampled. The arthropods from one year ofsampling have been soncd to morphnspecies. This paper documents the range ofspecies found and compares arthropod speciesrichness withinordersand familiesandbetween the iu.i i'imvst typW Hyim-implCTa, Colcoptcra,DipteraandArancaeweretherichestinspecies.Ninehumliedandsevyiiiv-M/vvu species m 173familieswere tound in theeastern Australian forest, while691 species tn 176 familieswerefoundinthe western Australian forest. Only _VWroffamilieswerecommon to both Forests, but almost half the families iccorded were represented by fewer than live ICS. Reasons forthese patterns arc briefly discussed and arthropod species richness in cucalypt communities is contrasted to thai In othci temperate ami trpplcal forests. The implications of forest and land management practices fot the conservation of arthropod richness are presented \jjm>rnehnur\. inset ts, arthropods* forest Eucalypius diversity t ./. 0. Major&A. C. Postle, SchoolofEnvironmentalfin>log\, Curtin fh<i\,->uivnfTet haul- o,ev era Box U1987. Penh. WA 6001, H F Recher* Deparmwni qfEcosystem Manage- men!, UniversityofNe\s England, Armidate. NSW235J; 12July I9S&. Erwin's (1982) seminal paper on rainforest problemsinextrapolating fromasinglesampleto canopy invertebrates, which included the global provide global figures. Arthropod species rich- estimate of 30 million arthropods, resulted in U ness in the canopy may not exceed that of the steady flow of papers on canopy arthropods. demonstrably lich soil and litter fauna and they Some supported Erwin's estimate and some maynot betotallyseparate faunas(sec,ft.fi,,Aths. regarded his figure as an overestimate (May, L988; Hammond, I9*J0), A furtherproblem with 1988; Stork. 198X; Gaston. 1991) To recopiM* (his debate is thai most eStirmUCS SUV based upon late. Erwin (1982) identified beetles from the samples taken in the rainforests; the implicit as- canopy ofone speciesofPanamanian lire Then. siimphon is ih;i( mosl of ihe world's biological using an estimate of the proportion of beetle diversity occurs in the tropics. Clearly, if the Species which were specific to individual tree conflict of opinion about global arthropod th.ti Species, the number of tropical tree species ness is to be sensibly resolved, we need reliable worldwide and the proportion of the total data on arthropod richness from temperate areas arthropod fauna represented by beetles, he ex- as well. trapolatedtoprovideanestimate oftotal arboreal Australian Eucalyptus forests represent a arthropod species richness of 20 million On the vegetation type for which few data on arthropod assumption that arthropod richness is twice as richness exist, Eucalypt communities are of par highinthecanopyasontheground. Erwin( 1982 tieular interest in relation to rainforests because went on to estimate a global arthropod richness they arc evergrgftn and Seasonal cwremes w, Of 30 million species temperature are not as grcai as in temperate The assumptions on which Erwin based his forestsol the northern hemisphere Thus, interna estimatesaresubjecttoqucstion.Forinstance,the ofthese features, they arc into"mediate hetwivi- degree of host specificity that Erwin assumed rainforests and temperate deciduous forests. may notbecorrect(Gaston, 199J )andtheproju>r- which have had arthropod richness documc i lion of key utxa in a sample may vary (ran by a seriesofdetailed investigations (e.g. South* community to community (e.g.. Abbott ct all, wood ct ah. 1982a, 1982b; Erwin, 1983a. 1983b; 1992; Kitching ct ah, 1993). thus leading to Adisct al.. 19X4; Hijii, 1984; Stork. 1991). 1m- 112 MEMOIRS OFTHEQUEENSLANDMUSEUM 600 £ * 2f>Nal-. t9o <aBm »c(B ?*7i a>> aG3 <e—Un »aa5c>. a> a8. -ooEaO soO5S 3«?B >3) a<oan9a. 3» -DC 3aCr>o sco eom eeEon 5caq aS. Qa. XiOCb—fL X31E>oc. Xcoi XFQT. aaa soao Ht>o. 3ao 8 ai 2aa. Zaa3o. X23oa|. E a% iEu Fig. I.Numbersofindividualsindifferentarthropodordersongreybox(£.moluccana,black)andnarrow-leaved ironbark(£.crebra,stippled).Valuesare meanarthropodspertree(n= 10trees)summedoverthefourseasons (Table 1) portantly, eucalypt forests are dominated by a analysis of the species richness of the arboreal single genus, Eucalyptus, which in most habitats invertebratefaunas ineucalyptforests.Thenum- is represented by only a few species. Therefore, bersofspeciesinthetwoforesttypessampledare studiesofeucalyptcommunitiesnotonlyprovide comparedtothenumbersreportedforotherforest an opportunity to test predictions of global communities. Subsequentpapers willanalysethe species richness but can be used to investigate similarity of species composition between the various assumptions aboutthe distribution ofin- eastern and western faunas, the extent of tree sect species between habitats and theirdegree of species specificity within each forest type, the host specificity Theevergreenness ofeucalypts, variation in faunal composition within a tree coupled with moderate seasonal changes in speciesand the extent to which seasonal changes temperature and rainfall also allows an assess- in community composition contribute to overall mentofthe contribution to species richness aris- patterns ofspecies richness. ing from temporal changes in community composition as distinct from spatial and habitat METHODS variation. In 1985,weinitialedstudiesineucalyptforests Sampling was done seasonally from February onthe relationshipbetweenarboreal invertebrate 1987 through January 1988 at Scheyville, New communities, foliage nutrient levels and tree South Wales (33°53'S, 150°51'E), where we species selection by foraging birds (Majer & sampled invertebrates on co-dominant narrow- Rccher, 1988; Majerel al., 1990, 1992, in prep.; leaved ironhark {Eucalyptus crebra F.Muell.) Keener et nl, 1991, 1993; Recher & Majer. in and grey box {E. moluccana Roxb.) and from press). Arboreal invertebrates were sampled April 1987 through November 1989 at Karragul- seasonally on eachoftwo species ofeucalypts in len. Western Australia (32°04'S, 116°07'E) on a marrt-jarrah forest in western Australia and a co-dominant marri (E. calophylla R.Br.ex box-ironbarkforestineasternAustralia.Asubset Lindley) and jarrah (E. tnarginata Donn. ex of the samples has now been sorted to mor- Smith). During each season, samples weretaken phospecies. Here, we present a preliminarv from the canopy (>7 m) and subcanopy (<7 m). * ARTHROPODSPECIES RICHNESS IN AUSTRALIAN CANOPIES 123 200 g 160 CO | 140 TaCa03<4ooo33-A3>. XCc©©to©CErDt £©orkaa(-P.;A <0c*a3 £0F0©Oa©aii31. aooaeCiirT-nJni. o©0ac3e—q3a. £HcEaMcaar>aa.* ^"0(aucaaOiq)A. UVTEuai r*c§aa©©>-d>. 2*©aar?a£°©—e.. -o—5©°©ac©o2"o2. I<gco 5S«icsa £<>o. EgDacQ. 3Tn3£CC0&oa3A3Q. ©©nEFI— •amoCuDO<oa3gAA. "243C(aoE0BD1OA1. £o©8DC3 LoaaEai 2'3To13J1 =o©©©£°©=. T2a2©i©93o©r Fig.2. Numbersofindividualsindifferentarthropodordersonjarrah(E.marginaia,black)andmarri (S cahphylfo, suppled). Valuesa/cmeanarthropodspertree(n= 10trees)summedoverthefourseasons(Table I.t. Because we were specifically interested in level; resulting data were described in papers foliage-associated arthropods, we avoided sanv quoted above. Subsequently, except for the en- pling trees which were flowering. dopterygote larvae, the arthropods from each of Branch clipping and chemical knockdowns the fourseasonsand fourtreespeciesweresorted were used to sample invertebrates but only the to species. All animals were assigned lo families data from materia) obtained by chemical knock- and were labelled with code numbers for each down are presented here. Details of the proce- species. Because ofthetaxonomiccomplexity of duresusedandthehabitatssampledare presented dealing with many juvenile spiders and of the in Majer&Recher(1988)and Majeretal. (1990, extremely high richness of Hymcnoptera, we 1992). Briefly, in each season we selected 10 sorted these two groups for the first two seasons trees of each species and stratum for sampling. only. In addition, because of the uncertainty in Notreewassampledmorethanonce.Withineach deciding whether individuals from eastern and tree,wesuspendedten0.5m".funnel-shapednets westernAustraliawere thesame species,weused using acherry-picker. Netswere positioned soas a separate numbering system for the material not to overlap and lo sample all parts ofthe tree from the two areas. The putative species repre- canopy. After a period of equilibrium (usually sentativesarccurrentlybeing senttotaxonomists overnight), the trees were sprayed witha fast-ac- in order to obtain generic and, where possible, ting pyrethrin insecticide synergised with specific names piperonyl butoxide. Spraying was done only undercalm conditions during early morning. In- RESULTS vertebrates collected by the nets were stored in 70% ethanol until sorted. Limited lime has allowed material only from Ordinal profiles derived from the numbers of the upper canopy samples and for the samples arthropods collected in each taxon are presented taken from April 1987 through January 1988 to for each tree species and for the two forests be sorted to species (i.e. once for each season sampled (Figs. 1,2). The current status of data- from autumn through summer in both States). codingpreventssegregation ofarthropodspecies The invertebrates were sorted initially to ordinal by tree speaes. so the number ofspecies in each 124 MEMOIRS OFTHEQUEENSLANDMUSEUM «8 '0 gz3 . . I J'.Jl.fl.rMi.iijfa. to n n 2 £ 0ouI) QC3JL <c2i g<aEv UaB5 02ca1 .'f5bf2=l1 c = o* Sa ai. om2- O5 ? < £ I 1 « -5S -SC «! ? i 3 Xc IB X g a a. < F iL'.3.Totalnumberofspecieswithineacharthropodordersampledon80easternAustraliantreeslE, moluccana and E. crebra, stippled)and 80Western Australian trees{E. mar^inata and E*. culophyllu), black). taxon can be compared between eastern and Hymenoptera, Hemiptera. Coleoptera, Diptera western Australia only. and Araneae were the most abundant orders of A total of 67,400 individual arthropods were arthropods in both forests (Figs. 1, 2). These obtained from the 160 upper canopy trees werefollowedbyPsocoptera,Thysanoptera,Col- sampled. Arthropods sampled numbered 50.900 lembola, LepidopteraandAcarinain thatorderin in the eastern forest, but only 16,500 in the west. westernAustralia,andby Acarina,Thysanoptera, Theyweremoreabundantonnarrow-leavediron- Lepidoptera, Psocopteraand Collembola ineast- bark and grey box than onjarrah and marri in all ern Australia. seasons sampled (Table 1). Narrow-leaved iron- Whiletherewassomeconsistencyintheranked barksupportedconsistentlymorearthropodsthan abundanceofordersbetweeneasternandwestern grey box and, apart from spring, theirabundance Australia, their relative abundance on different was higher on marri than on jarrah. The most species of cucalypts was more variable. In pronounced differences were in the many more Western Australia, Hymenoptera, Hemiptera, psyllids,otherHemiptera,DiptcraandHymenop- Coleoptera, Diptera, Araneae and Psocoptera in tera (excluding ants) on narrow-leaved ironbark that order were the most abundant arthropodson and ants and adult Coleoptera on grey box (Fig. jarrah, while Hymenoptera, Coleoptera. Psocop- 1 )- Marri had many more adult Coleoptera. ants tera, Araneae. Hemiptera and Diptera were most and Psocoptera than jarrah, while psyllids mid abundant on marri. In eastern Australia, Hemip- other Hemiptera were more abundant on jarrah tera. Hymenoptera,Diptera.Coleoptera,Araneae (Fig.2). andLepidopterain thatorderweremostabundant on narrow-leaved ironbark. On grey box. OrdinalProfiles Hymenoptera, Coleoptera, Diptera, Hemiptera, Overall, arthropods from 23 orders of insects Araneae and Lepidoptera were mostabundant. (Heteroptera and Homoptera counted as one order), arachnidsandcrustaceanswerecollected, SpeciesProfiles with 20 sampled in western Australia and 17 in A total of691 species ofarthropod were iden- eastern Australia (Figs. 1, 2). tified from western Australia and 977 from east- M«THROPODSPECIESRICHNESSINAUSTRAL!ANCANOPIES Season 1ret Species. Autumn 1987 Winter I9K7 Spring 1987 Summer 1988 Mean SD Mean SD Mean SD Mean SD Grcvbox 411.40 48.68 416.10 38.20 605.50 9ZIW 742.00 118.28 Narrow-leavedironbark 760.10 131.21 490.30 42.57 844.50 54.00 823.50 154.39 ' Jacruh 242.20 I7JS3 180.20 35.54 206.60 40.70 125.60 14.64 '-i mm >:--' :7.s: Mi 10 Sv | . is-i SI- 22.74 21630 42.74 Tabic I.Meanvalues(andstandarddeviation)oftotal numberofarthropodindividn.iKs.miplcd pertree(n = 10) Lin Australia. Some species would undoubtedly df=18, p< 0.001 tot western Austialia, r=Q.9|, be 'tourists' which have temporarily alighted or df=I3.p<0.001 toreastern Australia*. been carried onto the trees (e.g. the lycosid Of the richest foe oidcis, all vm c rcprcsfifitcd NptiJer). Although the two reference collecttt I by more species in eastern Austral|ia than in the haVfl not yet been homologised. our knowledge west. The differences in specicg numbers weie of ihc material and early determinations from greatest among the Hymenoptcra, with specialists indicate littleoverlapbetweenthetwo species in eastern Australia and 167 in the west; faunas.Thetotalnumberofarthropodspeciescan the Coleoplera with 222 species versus 141; the conservatively be estimated to exceed 1500 Araneae with 112 species versus 56; and llie species, ofwhich some 1300are insects. Homoptera with 87 species versus 63. The A total of 229 families were encountered, o( Psocoptera with 35 species versus 23 was also which 176 were represented in western Australia substantially neherin the eastern thanihe we.sfrn : and 173 in eastern Australia (Table 2j Tim- forest. Numbers of species in the less rich laxa although there are about 4i)r<- nunc species on exhibited broadI\ Minilai ttends, with ihe oni\ easlernthanonwesternAustraliantrees,the num- taxa having more species in western than eastern bers of families in the two forests arc relatively Australia he", iin i oltentboki, Heteroptera, similar. There is a 53$ overlap in the families Ncuroptera and Onhoptera. The remaining tax:* sampled in the two forests, suggesting a high were represented by relatively few species, so die level ofbiological richnessatfamily level. Forty- differences could well be an artifact ofsampling. seven families represented inthe eastern samples Thus, although geneially morearthropodspei i were absent in the west, while 56 families were were sampled in the eastern ih;m the western sampled solely in the western Australian forest forest, the inconsistent ttends in the less spCCIOjSC Ofthe families which were confined to one par- Uixa meant thai rfchlKiW W ithil) ordci value l ticular forest, only four contained five 01 more tween the two forests were not significantly dil species. Ofthe 22°- fannies. 95 were represented M (paired lest) I by fewer than three species. Thetotal species within each ordci (Fig 3)UTU DISCUSSION rankedbythe averageoftheamalgamatedeastern and western Australian counts. The inosi diverse groups in both forests were the Hymenoptcra (a Thearthropodsdiscussedhere v.civ sampledby total of 450 species), the Cpleopiern (363) and chemical kmu kdoun from JOticesofeachot IWCJ Dipteia (252), in that order. Araneae (168) and species in western ami such of iw© vpecieg \$\ Homoptera I 150) wete the next lichcst taxa. The eastern Austiaiia. The work of Abbott et al. richness ol species in Hvmrnoplcin and Aianeae 1992), also performed in prrah-marri forest, in- 1 is based on samples from two seasons and we dicah hcmical knockdown satjiplesoi estimatethattheirspeciesriehnesswould be partofthecanopy fauna.Thenetswete hung ncai greater had the full set of samples been sorted the extremities of hranches of non-flowering This docs not alterthe position oithe llymenop- trees, so animals collected were largely those teraasthemostspecics-riehtaxonbutcouldmean associated withthe foliage.Theonlyexception to that the Araneae are the third most species-rich this amongst the most abundant elements of ihe group.The numbersofspecies wiihm cneh ofthe Fauna was the Psocoprcftf, which tend to be ;'s orders in the two forests were strongly correlated goeiajed with the hark. Within eastern and (r-fl9S, df=26 p<0 001 There were significant western Australia respectively ihcse anmi.ils T ) . correlations between the number of individuals were most common on narrow-leaved ironbark and of species unhm the various taxa (r=0.')2. and marri and, r>l the pans ofuee species in each A 126 MEMOIRS OFTHEQUEENSLAND MUSEUM WA NSW WA NSW CRUSTACEA Trombidioidea 3 Isopoda I Acarinaindet 6 1 ARACHNIDA DIPLOPODA Pseudoscorpionida Pseiaphognatha 1 Cnernetidae i COLLEMBOLA Araneae Brachystomellidae 1 1 Araneidae 14 32 Dicyrlomidae 1 1 Clubionidae 4 6 Enlomobryidae 4 3 Connnidue 2 Hyp«?nastrundae 2 Cienidae 1 Isotonudue 2 Dcsidae 5 Neanuridae 1 Gnaphosidae 2 7 Sminthuridac 3 2 Hahniidae INSECT ] Hersiliidae I ! Thvsanura Heieropodidae T T Lepismatidae 1 Linyphiidae 1 1 Odonata !.vcosidae 1 Coenagrionidae 1 Micropholcommatidae ! Lestidac 1 i ixvopidae 2 Plecoplera Pararchaeidae 1 Gripoplerygidoe 1 Philodromidae 1 Orthopiera Salticidae 9 19 Gryllacrididac -) 2 Se^eslriidae 1 Gryllidae 5 Tetragnathidae 9 Tettigoniidae o 2 Theridiidae 9 14 Blaltodea Thomisidae 5 7 Blatlellidae 4 3 indel. 5 14 Blaltidae 1 Atarina-Mesosligmata Isoplera Phytosehdae i Rhinotermitidae 1 Acarina-Oribatida Dermaptera Ceratozetoidea -) pygidicrunidae 3 Cymbaeremaeidae Phasmatodea 1 ?Cymbaeremaeoidea 1 Phasmatidae 2 Oribatulidae 1 Manlodea 1hihatuloidea 1 5 Amorphoscelidae 1 Plaeremaeidae ] Mantidae I Plateremaeidae 4 Psoloptera indel. 1 1 Caeuliidae -j Acarina-Prostiftmaia Eclopsocidae f Anystidae -> 5 Elipsocidae 2 VAnyslidae 1 Lepidopsocidae 1 Bdellidae 7 3 Myopsocidae I Erythraeoidea 3 8 Peripsocidae 2 'l;rvihraeoideu 1 Philotarsidae 7 1 Oribatuloidea 1 Pseudocaeciliidae 1 5 Table 2. Numbersofspecies found within varioousarthropod families sampled from trees in a western and an eastern Australian forest. ARTHROPODSPECIES RICHNESS IN AUSTRALIANCANOPIES 127 Table2. continued WA NSW WA NSW Anisopodidae 1 1 Anthomyziriac 1 Psocidae 3 4 Asilidae 1 indet 4 III Aulacipastridae ] Homopiera Rnmbvliidae 1 Aehilidae 2 Calliphondae 1 1 Aleyrodidae 2 I Cecidomyiidae 4 10 Aphididae 5 3 Ceralopogonidae 9 9 Aphrophoridae 1 Chamaemyiidae 'j I Cercopidae i Chironomidae 11 7 Cicadellidae 17 34 Chloropidae 15 24 Cicadidae > Chyromyidae 1 Cixiidae 2 1 Crvploehaelidae 2 Coccoidca 2 Dolichopodidae 3 4 Eurybrachyidae T ! Drosophilidae 5 4 Eurymelidae A T Empididae 17 9 Flatidae 2 3 Ephydridatr 3 ] Machaerotidae 4 3 Fer^usoninidae ! 7 Membracidae 2 I Heleomyzidae 1 Psvllidae 16 36 Lauxaniidae 3 7 Hcteroptera Longchaeidae 1 Alvdidae 2 Milichiidae 2 2 Anlhocoridae 3 3 Muscidae 3 7 Ceratocontbidae 2 1 Mvcelophilidae 6 4 Lygaeidae 7 3 Phoridae 3 8 Miridae 7 15 Piptinculidae Peniatomidae 6 3 Platvsiomatidae Reduviidae 3 1 Pseudopomvzidae Thaumastocondae -> PsvcliodidiiL* 1 Tingidae 2 Scalnpsulae I Thysanopiera Sciaridae 5 Aelothripidae 3 4 Sepsidae 3 Phlaeothripidae 6 10 Simuliidae Thripidae 8 3 Stratiomyidae Neuroptera Svrphidae t Chrysopidae 2 1 Tnbamdae 1 Coniopterygidae 3 3 rachinidae 5 Hemerobiidae ~j I Therevidac 1 Manlispidac 2 3 Tipulidae 2 Myrmeleonlidae 1 tndL'l Lepidoptera Cnleoplera Yponomeulidae 1 Aderidac 1 indeL 1 Alleculidae 3 Tnchoplera Anobiidae 9 Lepioceridae i Anthicidae -> Diplera Anlhribidae 1 A^romyzidae 1 Alielahidae 2 8 128 MEMOIRS OFTHEQUEENSLANDMUSEUM Table2. continued WA NSW Silvanidae 1 : WA NSW Spercheidae i Belidae 1 3 Staphylinidae 4 9 Bostrichidae 1 Tenebrionidae 7 7 Buprestidae 3 3 Throscidae Canthandae 7 10 Trogossitidae 1 Carabidae 5 9 Zopheridae 1 Cerambycidae 5 indet. 4 Chrysomclidae 12 30 Hymenoptera Ciidae 1 ! Anthophoridae 1 Cleridae 5 6 Aphelinidae 9 9 Coccinellidae 9 18 Apidae 3 Colydiidae 2 Bethylidae 7 4 Corylophidae 2 3 Braconidae 14 31 Cryptophagidae 7 2 Ceraphronidae 4 1 Cucujidae 1 Chalcidae 2 Curculionid.je 37 43 Charipidae 1 Dascillidae 2 Collelidae 1 Dermestidae *» 5 Diapriidae 1 Dytiscidae ! Dryinidae 1 Elateridae 5 Elasmidae 1 Endomychidae 1 Encyrtidae 16 60 Endomychidae? 1 Eulophidae 32 43 Histcridae 7 1 Eupclmidae 5 4 Hvdraenidae 1 Eurylomidae • 3 Laemophloeidae 3 Fi^itidae 1 Lagriidae 1 Formicidae 22 33 Lathridiidae 3 2 Ichneumonidae 4 Leiodidae 1 Meeaspilidae 1 Melandryidae 1 2 Mymandae 2 13 Melyridae 5 Pergidac 1 5 Mordellidae 2 Plaiypasteridae 6 18 Mycetophagidae 1 Pompilidae 1 Nitidulidae 7 7 Pteromalidae 18 26 Oedomeridae Scelionidae 3 10 ] Phalacridae 1 Sphecidae 4 3 Phloeostichidae 1 Thysanidae 1 Pselaphidae 2 Tiphiidae 1 Ptiliidae 1 1 Torymidae 5 4 Pythidae I Trichogrammatidae 2 Salpingidae 3 2 Vespidae 1 Scarabaeidae 7 6 indet. 4 5 Scraptiidae 2 1 Totalfamilies i76 173 Scydmaenidae Totalspecies 691 977 1 . ARTHROPODSPECIES RICHNESSINAUSTRALIANCANOPIES 129 area, thesearetheoneswhichretain a thick bark always yielded a considerable number of addi- layeron theirbranches tional speciesand that thisseasonalturnoverwas MoAnsagwhealnl,apserss,peccoimesm.a)s,sofclioawteerds,wfirtuihtsbaanrdk ((hJe, raicmhanjeosrsifnacotuorrscaomnptlreisb.utWiengbetolitehveethhiagththsipseIcsiaens wood oftrees,thereisalsothatcomponentofthe importantcomponentofdiversity which needs to biota associated with othertree species, with the be considered in future studies and that it is im- psOhtrtuSbsoafndthweitehcotsheysstoielmanadlsolitsteurplpaoyretr*aThreisceh cpoormtpaontneenntouogfhditvoerbseityc.onWsiederreefderastoatshiespanraetwe arthropod fauna inthese two areas ofAustralia. component as sigma (o~)diversity Forinstance, Postleetal. (1991)sampledthe soil Using current estimates of between 10&,0OU and litter arthropods in jarrah-marri forest at and 145,000 species of Australian insects Dwellingup, some 50 km south of Karragullen (Taylor, 1983; Nielsen & West, in press),, our and found 290 animal species in nine small samples represent some 0.9-1.2% of the total sample plots. Thus, the total arthropod species Australianinsect fauna. Wesampledonly fourof richnessforourtwosamplesitesisaconservative the 600 or so Australian cucalypt species, estimate; the actual total would be considerably sampled only thecanopy andsampled only from higherthan the figure we obtained. twoextremely localised sites. We thus feel thai it The high correlation between the ordinal and is unlikely that we have sampled as great a per- speciesprofilesindicatesthattheformerprovides centageasthisofthe Australian insect fauna This some reflection of the species richness of a leaves us with no older conclusion than that the sample or a site. Indeed, our finding: that species number of Australian insect species has been nebness is far higher in the eastern than the grossly underestimated, western foresthad alreadybeenalluded toby the Our figures for arthropod species nchness are substantially higher abundance of arthropods in intermediate between the high values for the the eastern than the western site (Majer el al., canopv oftropical forests ( Erwin, 1982. 1983b; 1990. in prep.; Rccher ct al., 1991). The reason SUd^ 1987; Basset ft Arlington, 1992) and the fo/ this difference between forests has not yet much lower values for deciduous temperate been conclusively found. However, Majer et ai. forests (Southwnod ct al, 1982a, 1982b), Most (1992) found substantially higher levels uffoliar contributions to the debate on global arthropod nitrogenandphosphorusin theeasternthaninthe speciesrichness are based on data obtained from westernAustraliantreesand,byreferringtoother thetropics.Limitedconsideration isgiventodata trendsinfoliarnutrientsbetweentreespeciesand from temperate forests. Our data support the within tree canopies, suggested that the abun- statement that Australia is one of the 12 dance of arthropods might be a response 1,0 mcgadivcrsc countries that togetheraccount for nutrient levels. Ifthis is thecase, thedifferences 15% ofthe total biodiversity of the planet (Mc- may well apply to eastern and western Australia Neely ct aL 1992) and concur with Platnick's asa whole. (1991) statement that more attention should be Reasons fox the high richness ofarthropods in given to the temperate regions when estimating these two forests is not here discussed. The de- global biodiversity. 1/this were done, it is likely gree of host plant specificity [Fo* A Morrow,. that current estimates of arthropod species rich- 1981)andalsothegeographicalrangeofthehost ness wouldbe elevated toeven higherlevels. plant(Strong, 1979)couldbecontributoryfactors The richness of the canopy arthropod fauna but the data have not yet been processed to the fromonly twositesand fourspecie* ofeuralypt* extent required to investigate this aspect of confirms the need to include a consideration of cucalypt-associated invertebrates. Tbis paper invertebrates when planning and managing I aimsonly tointroducea plannedscriesofpapers nervation reserves*The 1500orso species which on arthropod community structure in Eucatypws we sampled represent only part of the invcr- canopies However, one immediately evi- ichr.iie Epecies richness ot* the forests where we i dent component of richness is the. as yet LSI- worked. analysed, seasonal variation in community There were no apriori reasons for expecting composition. This component ol richness has either site 10 have a nch canopy invertebrate generally been overlooked by canopy workers, fauna. Neitherforesthasafloristicallydiverseor most of whom base their nchness counts on a complex structured canopy. Five species of single season of sampling. It was evident from eucalypt occuron the Scheyville site and fourat out samples that each season which we sported Karragullen but the two cucalypts sampled on MEMOIRS OFTHEQUEENSLAND MUSEUM eachareadominatethecanopy (>90%offoliage) burning, affect the whole fauna rather thanjust and understorey vegetation. Both forests have a the vertebrate fauna. long history of disturbance (e.g.. logging, ACKNOWLEDGEMENTS changed fire regimes and, in the case of Scheyvilie,grazing)andoccuranrelatively pour soils. Scheyvilie ret;»n.\adiverge,albeitdepleted This project was undertaken by means of a avifauna (>70 breeding bird species) (H.F grant from the Australian Research Council. We Recher,unpublisheddata)and priortoEuropean would also like to thank the many IftXOflOmK settlementwould have had arich mammal fauna specialists who have helpedorwho arecurrently (Recher& Hutchings. 1993).Jarrah forest,ofthe helping us with the determination ofspecies. Dr type represented at Karragullen, has a relatively Robert Floyd and Philip Groom kindly assisted poor avifauna (about 45 breeding bird species) with the data base andgraphical analyses. (H.F-Rechcr, unpublished data), afeature which LITERATURECITED is typical ofdry, open eucalypt forests. The Karragullen site is part of a Western ABBOTT. BURBIDGE, T. WILLIAMS, M. & Australian State Forest which is managed by the VAN [H.,EURCK, P. 1992. Arthropod fauna of government for timber and firewood production vtrrah {Eucalyptusmar^tnata)foliagein mediter- and, as such, is relatively secure from develop ranean forest of Western Australia: spatial and meat. As one ofthe largest remaining fragments temporal variation in abundance, hiomass. guild ofanoriginally extensive* woodland on the Cum structure and species composition, Australian berland Plain. Scheyvilie has been proposed for Joum.il ol Ecology 17: 263-274, nature reserve status since the late 1960*5 It is ADIS, J. 1988. On the abundance and density ol ter- also Crown Land (i.e government owned) but restrial arthropods in Central Amazonian dryland forests.JournalofTropical Ecology4: 19-24, only a small part has been reserved and the AIMS) K l.UBIN, YD. &. MONTGOMERY, GO remainderhasbeenproposedfordevelopmentas 1984. Arthropods from the canopy of inundated ahousingestate. Failuretoreservetheentirearea and terrafinite forests near Manaus, Bray.il, with b ill part a failure to appreciate the biological critical considerations on the pyrethmm fogging richnessand inpartaconsequenceofaparadigm technique; Studies on Neotropical Fauna and the that emphasises large, predominantly natural or BASSEEnTv,iroYn.me&ntA1R9:T2H2I3-N2G3T6.ON. AH L992. The wild areas with little economic value (or nature arthropod community 01 an Australian rainforest conservation. Such attitudes do not consider the tree: Abundanceofcomponent laxa, speciesrich- possibility that invertebrate communities may ness and guild structure. Australian Journal of persist relatively intact or at least retain high Ecology 17:K9-9X. speciesrichness,regardlessofahistoryofdistur- ERWIN. T.L. 1982. Tropical forests. Their richness in banceandhabitatfragmentation.Thediversityof Coleoplera and other species. Coleopterisis Bul- die flora and the number of vertebrate, species letin 36. 74-75. may also no* be good predictors of invertebrate I9Sjj Tropical forestcanopies: thelast hjotic fron- tier. Bulletin of the Entomological Society of species richness. This isparticularly soifhistori- America30: 14-19. cal changes to the flora and vertebrate fauna arc 1983b. Beetles and other insects of tropical forest notconsidered. canopies ai Manaus, Rra/.il, sampled by mscc- The richness and abundance of CttbOpjf ticidal fogging.InSutlonS.L.,Whitmore,T.C.& arthropods at Scheyvilie and Karragullen arc Chadwick, A.C. (eds) 'Tropical rain forests: compelling arguments for the use of broader Ecologyandmanagement' pp59-75. (Blackweil criteria when planning and managing conserva- FOX,.SLc.iRe.nti&ficMPOubRliRcOatWio,ns"PAOx.fo1r9dX)1,. Specialisation: tion reserves Areas such as Scheyvilie that rep- species property or local phenomenon*7 ScictJCfl nt the only remaining fragments of formerly 211:887-693 extensivebahiuns, may retain mostofIheoriginal GASTON. K.J. 1991. The magnitude of global insect fauna,although muchofthe vertebratefaun;. species richness. Conservation Biolocy 5: 2X2- have become extinct. As zuch, these arca> have considerable conservation value regardless of HAMMOND, P.M. 1990. Insect abundance anddiver- sity in the Dumoga-Bonc National Park, N their size and the lack of wilderness values The Sulawesi, with special reterencc to the beetle management ofmore extensive habitats, such as fauna of lowland dry forest oftheTorant region. that at Karragullen^ needs to consider how In Knight, W.J. &. Holloway, J.D. (eds) 'Insects management practices, for example prescription andrain forests M'southeastAsia(WallaceiV pr