ido MEMOIRS OF THE QUEENSLAND MUSEUM GILL AREAS AND EVOLUTIONARY restrial habitats without an intervening supralittoral phase. RELATIONSHIPS IN TALITRID AMPIUIPODS Those Group 4a species that did exhibit a clear reduetion in Bousfield (1984) has identified 4 eco-moarphological gill area are apomorphic species (Friend, 1987). groupings of talitrids; palustral lalitrids (1); beach eas (2); However, consideration of the partitioningo fa rea between sandhoppers (3); and tandhoppers (4). Within the landhap- the five pairs of gills ruises questions. Landhoppers show pers, two sub-groups are recognised, one of whieh (4a) is very marked reduction in the area of gills3 ,4 and 5 compared believed to share ancestry with palustral amphipods, the other with the palustral species, presumably because they carry (4b) is considered a more recent group rejaled (u beachfleas, fewerand lurgereges [han aqualicumphipods, The similurily The groupings, although polyphylette, provide a reasonable in total ared belween Group 4a and palustral talitrids is overview of the probable pathways of talitrid evolution. achieved by a pronayunced increase in the area of pills 2 and Tasmania's diverse range of lemperate habiluls provides 6, this is inconsistent With the view that gull areas in (hese representatives of all the major talitrid groups. including animals represent 4 <curry over9 from aquatic origins and about 20 species of landhopper, suggests that terrestrial selection pressures have been in- Published work (Moore and Taylor, [984; Spicer umd volved, Taylor, 1986) provides data for gill surface areas in beach- [tis nat immediately apparent how the large area of gills fleas, sandhoppers and one landhopper from Group 4a. We 2 and 6 might be adaptive since it is Counterproductive tor have added data fora palustral talilrid. Group 4a landhoppers contral of walter Juss and is unlikely to be necessary for (6 spp.) and a Group 4b landhopper, allowing us ly review improved axygen uplake or ummonia excretion in these all of Bousfield9s talitrid groupings in the context of gill permeable animals. Preliminary work suggests that, like surface acea relationships. aquatic amphipods, landboppers use their gills for ion uptake, and jon availability may be a majordelerminant of terrestrial Results and Discussion survival (Spicer et a/., 1987). Since permeability restricts Relationships between gill area and dry body weight are landhoppers to moist microhabitats, it may also be thal it hus described by the typical allomeiric relationship (y = ax9) , proved casicr to retain the low blood PCO2 characteristic of enabling statistical comparisons of regression slapes and aquitlic Organisms, and increase the areas of gills 2 und 6 lo elevations to be carried out on double log transformed data facilitate COz removal, than to evolve efficient blood bulter- for ali & species investigated. Only 2 species (within one ing capacity with high internal CO> concentrations us in other genus) differed in slope, indicating that allometric relarion- (errestrial taxa, ships were similar in all species and groups; visual compari- sun indicates thal this holds for published data also. However, Literature Cited most comparisons of elevation were statistically significant, Boustield, E.L. 1984, Recentadvances in the systematics and indivating that differences hetween species are maintained biogeography of landhoppers (Amphipoda: Talitridae) {hroughoul the size range. oft the Indo-Pacific region. In F.J. Radovsky, P.H. Raven To relate these data to Bousficld9s hypotheses, gill areas und $.H. Somer (eds) <Biogeography of the tropical were calculated from our and published data for a stand- Pacific9. Bishop Museum Special Publication 72: 171- ardised, 2mp dry weight, animal 4 a realistic size for our 210). species allhougha small beachflea or sandhopper. Calculated Friend, JA, 1987. The terrestrial amphipods (Amphipoda: areas are consistent with the proposed phylogenies. Thus.the Talitridae) of Tasmania: systematics and zoogcography, semi-aqualic palustral species exhibited the largest area; Records of the Australian Museum Supplement 7; |~85. although no data for Hyalidae exist the close correspondence Moore, P.G. and Taylor, A.C, 1984. Gill area rélationships with aqualic gammarids (calculated from Moore and Taylor, in an ecological series of gammaridean amphipods (984) suggesis that the palustral species resemble the aquatic (Crustacea). Journal of Experimental Marine Biology ancestral condition. In comparison sandhoppers (Spicer und und Ecology 74: 179-186. Taylor, 1986) have by far the smallest areas, in keeping with Spicer, J.1., Moore, P.G. and Taylor, A.C. 1987. The physi- the intense desiccatory stress imposed by their fong evalu- ological ecology of land invasion by the Talitridae iongry history on sandy beaches. The beachfleas (Spicer and (Crustacea. Amphipoda). Proceedings of ihe Royal Taylor, 1986), commonly found under wrack and pre- Sociely of London, B, 232: 95-124, sumably subjected to reduced stress, show a lesser, but still Spicer, J.J, and Taylor, A.C. 1986. A comparative study of considerable, reduction. The area calculated for the Group 4b the gill area relationships in some talitrid amphipods. specios was Wilhin the beachflea range and was the smallest Journal of Natural History 20; 935-947. of any landhopper investigated, consistent wilh Lhe proposed Roy Swatn and Alastair M.M. Richardson, supralittoral phase in (he group's history. Most Group 4a Jandhoppers approximated the palustral species, supporting Department of Zoology, University of Tasmania, the argument that they entered moist. non-desiccatory ter- GPO Box 252C, Hobart, Tasmania 7001, Australia,