Memoirs of the Museum of Victoria 56(2):275-280 ( 1997) 28 February 1997 https://doi.org/10.24199/j.mmv.1997.56.14 BIODIVERSITY AND BIOGEOGRAPHIC RELATIONSHIPS OF SELECTED INVERTEBRATES FROM URBAN BUSHLAND REMNANTS, PERTH, WESTERN AUSTRALIA M.S. HARVEY 1 , J.M. WALDOcK 1 , R.A. How', J. DELL1 AND E. KosTAs2•3 1 Westem Australian Museum, Francis Street, Perth, WA 6000, Australia 2School of Environmental Biology, Curtin University of Technology, GPO Box U1987, Perth, WA 6001, Australia 3Present address: 52 Lewis Road, Wattle Grove, WA 6107, Australia Abstract Harvey, M.S., Waldock, J.M., How, R.A., Dell, J. and Kostas. E., 1997. Biodiversity and biogcographic relationships of selected invertebrates from urban bushland remnants. Perth, Western Australia. Memoirs of the Museum o.fVictoria 56(2): 275-280. The results ofa continuous pitfall trapping programme designed to collect ground-dwell­ ing invertebrates on the Swan Coastal Plain, Perth, is documented. Of the groups identified to date (Arachnida. except Acarina and some Arancae, scolopcndrid and scutigerid centi­ pedes, cockroaches and baeinc wasps), 181 species have been identified, indicating a rich assemblage. Average linkage clustering analysis (UPGMA) of similarity indices derived from species abundance data is presented for some groups, which show differing patterns in relation to vegetation and land form. Larger bushland remnants have higher species richness and abundance. Introduction land patches on the Swan Coastal Plain, a region characterised by a series of landforms of Pleis­ The depletion of native vegetation during tocene age and a diverse array offloristically rich urbanisation in many regions of the world has vegetation types. We here present preliminary had an obvious and deleterious effect on much data on a selection of invertebrates, including of the original biota, raising concerns for the predacious (arachnids and centipedes), detrivi­ long-term welfareof not only individual species, rous (cockroaches) and parasitic (baeine wasps) but ofentire ecological communities. The rapid groups, and examine the distribution of some growth of Australian cities since European species, their probable origins, and the biogeo­ settlement has led to large-scale clearance and graphic significanceof these groups on the Swan degradation of substantial areas of pristine Coastal Plain. This work is seen as a necessary bushland such that concerns have been precursor to any investigation into the effect expressed over the long-term viability of con­ that urbanisation might have on the remaining stituent species. biota. Perth is the largest city in Western Australia and is mostly situated on the Swan Coastal Plain centred on a latitude of approximately 32°S. It Methods experiences a predominately Mediterranean cli­ During the present study, which is the first ofa mate consisting of hot, dry summers and wet 3-year program designed to investigate 12 or winters where more than halfof the annual rain­ more locations on the Swan Coastal Plain, a total fall of 837 mm falls between June and August. of 15 sampling sites were selected from four The coastal heaths and woodlands of the Swan locations. These were selected so that a broad Coastal Plain and surrounding regions are rec­ array of soil type and vegetation communities ognised internationally for their high diversity were sampled representing locations on four of both plants (Marchant, et al., 1987) and rep­ major landforms, including the more recent tiles (How and Dell, 1994). However, the taxo­ dune systems of Bold Park in the west, to the nomic and biogeographic relationships of the sands and colluvial clays and gravels of Talbot invertebrate fauna of this area is generally Road Reserve in the east. poorly documented. Two ofthe four sites in Bold Park (BP) are on A survey was commenced in 1993 to describe the Spearwood Dune System. Site BP3 is a Bank­ and interpret the ground fauna, both invert­ sia attenuata/8. menziesii low woodland over ebrate and vertebrate, of remnant urban bush- low heathland on a dune top. Site BP4 is a Tuart 275 E. KOSTAS 276 M.S. HARVEY, J.M. WALDOCK, R.A. HOW. .1 DELL AND Biodiv (1993) Eucalyptus gomphocephala woodland in an the Renkonen Index from the interdunal swale. Sites BP1 and BP5 are low package. = min(P„, P heathlands on the coastal Quindalup Dune I I, lk ) System. where P,j is the proportion of the /th species in ' /'th and P is the proportion of the /th All three sites at Tuart Hill (TH 1-3) are Jarrah the sample lk Eucalyptus marginata open woodland with a species in the kih sample. The derived matrix Banksia attenuata and B. menziesii low wood- was clustered using the UPGMA method. land understorey over diverse heathland on pale Dune System. brown sands of the Spearwood Results and discussion Three of the four Perth Airport sites (PA) are on quartz sands of the Bassendean Dune Sys- A total of 181 species was collected and ident- harvestmen, 4 tem. PAS is a dense low heathland, PA7 is a ified, comprising 103 spiders. 3 winter-inundated Melaleuca preissiana wood- pseudoscorpions, 3 scorpions, 6 scolopendrid cock- land. PAS is a low woodland o\ Banksia altenu- centipedes. I scutigerid centipede, 34 ata and B, menziesii over low heathland. PA6 is roaches and 27 baeine wasps. a mixed shrubland apron of the Spearwood Sand A total of 66 species (36.5%) was collected Svstem over Bassendean Sands. from only a single location and, of these. 34 single 'Of the four Talbot Road sites (TR). sites TR1 species ( 1 8.8%) were collected from only a and TR4 are on quartz sands of the Bassendean site, although some of these have been pre- Dune System. TR1 is an open woodland of Jar- viously recorded from outside of the study areas rah Eucalyptus marginata with an understorey and should not necessarily be used to indicate of Banksia menziesii and low heathland. TR4 is extreme endemicity. For example, the pseudos- a B. menziesii low woodland over a low hea- corpion Geogarypus taylori Harvey is widely dis- thland. TR2 is a Marri E. calophylla woodland tributed across much of southern Australia over low heathland on laterite/clay soils. TR3 is (Harvey, 1 986), and yet was only collected at the a E. calophylla woodland over Hakea lissocarpa four Bold Park sites during the present study. shrubland on quartz sands similar to TR1 and However, it has been taken from two other sites TR4. on the Quindalup dunes. Woodman Point and The invertebrate fauna was sampled solely by Mount Claremont, during later aspects of this the use of pitfall traps placed flush with the study, and the species may be restricted to this ground at each of the sites. Ten 2 litre plastic ice- dune system in the Swan Coastal Plain. The acti- cream containers set with approximately 400 ml nopodid spider Missulena granulosa Hogg was of ethylene glycol were placed 5-10 m apart. only represented by a single male from BP5, and They were covered with linoleum floor tiles set yet is widespread in south-western Australia approximately 10 cm above the top of the con- (based upon records in the Western Australian tainer to exclude rain and leaves. The traps were Museum). In addition, some predominately cleared every 6 weeks for a 12 month period, and arboreal species were occasionally caught in the the samples stored in 75% ethyl alcohol until pitfall traps, and the apparent endemicity of they were sorted in the laboratory. these species is probably due to a form of sam- Several target groups were selected for pling bias. For example, of the five species of detailed identification (ordinal names follow Heteropodidae collected, at least one (Isopeda Harvey and Yen, 1989). which included arach- leishmanial Hogg) is known to be extremely nids belonging to four orders (Araneae. Scor- abundant throughout the Perth metropolitan pionida. Pseudoscorpionida, Opiliones). centi- region and elsewhere in southern Australia pedes belonging to two orders (Scolopendrida (Hirst. 1992). and Scutigerida), cockroaches (Blattodea) and Our preliminary documentation of invert- the parasitic wasps of the tribe Baeini (Scelion- ebrate biodiversity shows considerable vari- idae). Several arachnid and centipede groups ation in the total number of species recorded were excluded from the analysis due to uncer- from each site, ranging from 34 species at TR3 to tainties with species level identifications: these 71 species at BP1 (Table 1). included all of the several Acarina and families The 1 8 1 species identified from 1 6 sites at the of Araneae, such as the Gnaphosidae (Arach- four locations sampled were subjected to simi- nida), and the Geophilida and Lithobiida (Chi- larity analyses of their composite assemblages. lopoda). The data matrix is presented in more Similar data for the Baeini were also evaluated. detail in How et al. (1996). These data are preliminary in that the project Similarity between sites was calculated using intends to document assemblages from over 30 • i' INVERTEBRATES FROM URBAN 8USHLAND REMNANTS 277 < h rn «N O O * r~- oo - H — n ^t m i^ -nrs —< tC C* —<— — ON H i i(Nrno- no TD 03 o O Qd 5t CC o_^h- -OOHMOONt o ON H nOnnino-OiO rn 5 H O —< <N f> K _ H o X (N ON u- X ON 033 E- -'(N-OMO-'ON H i I O — <N oo H rn (N < m cn o oo — oo < Cu —>r*-><NCN|i/->OooNO *— '— ao t~~ i— < o cu <N m — OmOM — < o "* XI ^o tn < (N OO NO O CU NO Cu l/~> < O no m OO Cu r<im —i —i Tl" O —' — NO ir-> 0. O cn| rN © 03 NO .* Tf I- a. nO \D— oo 03 r-l —' (N — (N o ' no cn -a <^m 3 cu nO DO ^3" 03 m — — r-4 —i —i on nO , CU no f — 03 m — r) n t -^ — — 03 T3 C 03 O 03 X! '§• a 03 '— « S.-2 T3 1) a 03 c .a y c C c (D o 2 T3 P a.