Alley, N.F.¹, McGowran, B.² & Bourman, R.P.^{3 1}Department of Mines and Energy Resources, 191 Greenhill Road, Parkside, South Australia 5063. ²Department of Geology and Geophysics, The University of Adelaide, North Terrace, Adelaide S.A. 5000. ³Faculty of Engineering and Environment, The University of South Australia, The Levels, South Australia 5095.

Changing southern Australian environment since the Late Cretaceous: geology, geomorphology, palaeoclimates and palaeofloras.

Rifting of the Australian plate from the Antarctic was well advanced by the Late Cretaceous. The southern rift basins were the major foci of deposition of sediments eroded from the nearby continents. Major palaeodrainage systems were established on the Australian continent, one draining to the Eucla Basin and another to Lake Eyre Basin area. Southern Australia was at relatively high latitudes (60-75oS) and continued to drift towards the equator during the Tertiary, with a major acceleration in the Eocene. Although deposition was relatively continuous in the southern rift basins, it occurred in three main phases in the interior: late Paleocene-Eocene, mid Tertiary and Pliocene-Quaternary, in part reflecting phases of downwarping and producing major palaeochannel activity.

The northward drift of the continental plate, its geomorphology and tectonic/eustatic history in conjunction with global climatic changes had a profound influence on vegetation distribution and history and soil development.

Widespread podocarp-araucarian-dominated cool temperate rainforest prevailed at least along the southern continental margin in the Maastrichtian (latest Cretaceous) and earliest Paleocene. By the Late Paleocene to Early Eocene rainforest of megathermal aspect existed in the Lake Eyre Basin, indicating that conditions there were warmer than along the southern continental margin where more mesothermal rainforest existed. In the Middle Eocene monsoonal-like conditions prevailed in central Australia and moister conditions in the south where rainforest of meso- to mega-thermal aspect grew, in the latter area extending into the Late Eocene. Throughout the region from the Middle Eocene, Nothofagus became an important part of the forest structure. The ?Late Oligocene-Miocene was a time of development of extensive shallow, alkaline in the Lake Eyre Basin. Lakes in the inland area supported a diverse fauna, including crocodiles. Vegetation had changed to dry, open woodland in the interior of the continent, with marginal monsoonal vegetation confined to wetter valley bottoms. Along the southern continent more sclerophyllous woodland with pockets of rainforest existed in some areas and mixed meso-megathermal rainforest in others. By the Pliocene further drying had produced a chenopod shrub-open woodland containing isolated pockets of forest in edaphically suitable sites in the interior and open woodland with isolated patches of rainforest in the south.

Deep weathering was widespread in the early Tertiary and may be as old as the Early Mesozoic. Later weathering was related to duricrust development. Ferricrete formed probably in the Early Mesozoic, Late Oligocene-Middle Miocene and Late Miocene-Pleistocene. Major phases of silicification occurred in the Late Eocene-Middle Miocene and the Late Miocene-Pleistocene, when significant groundwater silcrete formed. Some silicification was in progress prior to the Late Paleocene. Throughout the Late Cretaceous and Tertiary, weathering and duricrusting was taking place on a landsurface largely of low relief with uplands more or less in areas where they are today.

Austin, A.D. & P.C. Dangerfield. Department of Crop Protection, Waite Campus. the University of Adelaide, P.O. Glen Osmond, S.A. 5064

The phylogeny of Parasitic Wasps - the Basis for interpreting Major Biological Events and Host Relationships.

The parasitic Hymenoptera represent the largest assemblage of insects, as they have radiated to exploit most other groups of insects and other terrestrial arthropods as hosts. They lay their eggs onto (ectoparasitoids) or into (endoparasitoids) a host and the developing larva feeds on the host eventually killing it. Some of the key issues in developing an understanding of hymenopteran evolution are 1) determining the likely origin(s) of parasitism (i.e. primitive basal wasps are phytophagous), 2) tracking the origin(s) and radiation of endoparasitism, the most physiologically complex form of parasitism within the wasps, and 3) determining the pattern of host relationships within various higher taxa. Such questions demand robust phylogenies onto which biological traits such as the type of parasitism and host groups can be mapped. However, because of the small size of many parasitic wasps, morphological studies have largely failed to produce acceptable phylogenies because of their reliance on reductional characters. Recent studies, mostly using DNA sequence data, have generated preliminary phylogenies for major groups (superfamilies and families) of wasps, and within certain families like the Braconidae, that provide some phylogenetic stability. These will be used as examples to explore the evolution of parasitism within the Hymenoptera and the pattern of host associations at various taxonomic levels.

Austin, C. C. Evolutionary Biology Unit, Australian Museum, Sydney, New South Wales

Phylogenetic Relationships of Boid Snakes and Population Structure of Pacific Boas

The enigmatic distribution of Pacific boas (genus Candoia) in the Papuan-Pacific region has been explained by two competing hypotheses:

1. dispersal from the New World via waif over-water dispersal, and

2. the result of a relictual Gondwanan vicariant event.

I present a phylogenetic analysis for boid snakes based on mitochondrial (cytochrome b) and nuclear (c-mos) DNA sequences. These molecular data, in combination with published morphological data, are used to resolve higher level boid relationships and distinguish between the two hypotheses for the distribution of Candoia in the Pacific. Multiple populations for each of the three species of Candoia were sampled to examine levels of population structure and intrageneric relationships. Large intraspecific sequence divergence (up to thirteen percent within C. carinata for cytochrome b) exists within Candoia species demonstrating large population structure.

Barnes, R. W. & R. S. Hill University of Tasmania, Department of Plant Science, GPO Box 252-55, Hobart, Tasmania 7001.

Phylogeny, fossils and biogeography of Cunoniaceae

A preliminary phylogenetic review of a broadly defined Cunoniaceae is presented. The phylogenetic analysis presented includes the genera Brunellia (Brunelliaceae), Bauera (Baueraceae), Davidsonia (Davidsoniaceae) and Eucryphia (Eucryphiaceae) to determine the relationships of these genera to those widely accepted to be included within Cunoniaceae. Corrections on previous analyses are also highlighted. The broadly circumscribed genus Caldcluvia, erected by Hoogland, was not considered valid due to the genus most likely representing an unnatural assemblage which is probably paraphyletic. The analysis is based on character states previously published and several new or modified states which have not been previously considered. Potential macro-fossils representative of the family have also been recovered from Australian deposits both during and prior to this study. Their taxonomic affinities will be discussed along with the biogeographical consequences. Problems associated with defining fossil Cunoniaceae and associated families will also be discussed.

Bayer, R. J. & J. R. Starr. Australian National Herbarium, Centre for Plant Biodiversity Research, GPO Box 1600, Canberra, ACT, 2601 and Dept. of Botany, University of Manitoba, Winnipeg, Manitoba, R3T 2N2, Canada

A tribal phylogeny of the Asteraceae based on two non-coding chloroplast sequences, the trnL intron and trnL/trnF intergenic spacer.

Ubiquitous parallelisms in morphology have made phylogenetic reconstruction and tribal circumscription an area of considerable debate within the Asteraceae. In this study we explored the utility of two relatively short non-coding chloroplast DNA sequences, the trnL intron and trnL/trnF intergenic spacer, for resolving phylogenetic relationships among the tribes. The results of the phylogenetic analysis produced trees that are topologically congruent with prior phylogenetic hypotheses based on both morphological and molecular data sets. The Asteroideae are a monophyletic group, but the Cichorioideae are paraphyletic. The primary clades of the Cichorioideae are the Mutisieae-Cardueae, Liabeae-Vernonieae, and of the Asteroideae, the Inuleae-Plucheeae, Astereae-Anthemideae, Senecioneae-Gnaphalieae, and the Helianthoid clade (Helenieae, Heliantheae s.str. and Eupatorieae). The Inuleae-Plucheeae clade is sister to the remainder of the Asteroideae and the paraphyly of the Inuleae s.l. (Gnaphalieae, Inuleae s.str., and Plucheeae) is firmly supported by our analysis. Using approximately 874 bp we were able to produce a phylogeny that is as resolved as phylogenies based on well-known coding regions such as rbcL and ndhF. For phylogenetic inference at the family level the trnL intron and trnL/F spacer provides similar levels of resolution as longer coding sequences while having the advantage of being much easier to amplify and sequence across broad taxonomic groups due to its short length and universal primers.

Beattie Andrew J. Key Centre for Biodiversity and Bioresources, School of Biological Sciences, Macquarie University, NSW 2109, Australia.

Taxonomic Partnerships in Biodiversity Assessment.

The decline in taxonomy and systematics in terms of resources, personnel and funding is now well documented. Biodiversity offers a means of turning the situation around. The concept is now embedded in many high-profile activities such as ecologically sustainable development, environmental monitoring and conservation planning, all of which require the services of taxonomists and ecologists for biodiversity assessment. Neither can do the job alone. In addition, assessment is often driven by political rather than scientific deadlines. This situation requires the creation of partnerships between taxonomists and ecologists, particularly in situations where the organisms in question are poorly known taxonomically and/or hyperdiverse; or where large or contentious areas of land or ocean are in question. Such partnerships are nothing new, but by imaginatively adjusting the relative contributions of taxonomists and ecologists, parataxonomists and biodiversity technicians, automation and instrumentation and by setting the taxonomic penetration required to get the job done, much biodiversity assessment, monitoring and conservation planning can be accurately achieved. Biodiversity science and its applications offer a practical channel through which the resources, funding and personnel for the taxonomic network, and systematics, will grow.

Benyon F.H.L.¹, Burgess L.W.², & Sharp P.J.^{3 1}Allergens Unit, Institute of Respiratory Medicine, University of Sydney, NSW, Australia, 2006. ²Fusarium Research Laboratory, Department of Crop Sciences, University of Sydney, NSW, Australia, 2006. ³Plant Breeding Institute, Department of Crop Sciences, University of Sydney, NSW, Australia, 2006.

Molecular genetic variation and taxonomy of Fusarium species in sections Discolor and Roseum.

Molecular genetic techniques have been used to examine the genetic relationships between morphologically similar species grouped in sections Discolor and Roseum of the genus Fusarium. The objective is to compare phenotypic and genotypic classifications, from taxon to section. A total of 57 isolates, from F. graminearum Group 1 and Group 2, F. crookwellense, F. culmorum, F. torulosum, F. avenaceum ssp. avenaceum, F. a. ssp. aywerte, F. a. ssp. nurragi, F. heterosporum and F. oxysporum (outgroup) were examined. Restriction Fragment Length Polymorphisms (RFLP) were generated with 33 Fusarium DNA probes (randomly cloned genomic DNA and mt DNA) and six restriction endonucleases. RFLP patterns yielded 957 bands on autoradiograms. These were analysed using Dice's coefficient to calculate pairwise distances in the RAPDistance program. Pairwise distance calculations based on RFLP banding patterns revealed great genetic similarity between F. culmorum, F. crookwellense and F. graminearum Group 2. While F. graminearum Group 1 also displayed genetic similarity with these taxa, F. torulosum was genetically unrelated. The three subspecies of F. avenaceum and F. heterosporum were genetically distinct. However F. a. ssp. avenaceum displayed some genetic similarity to F. torulosum. Genetic variability was highest in F. heterosporum and F. a. ssp. nurragi, and lowest in F. culmorum and F. a. ssp. aywerte. Genetic data supports the grouping of F. graminearum Group 1 and 2, F. culmorum and F. crookwellense into a section, without F. torulosum. However, it does not support the original Roseum grouping. At the taxon level genetic evidence indicates that F. graminearum Group 1 and 2, F. a. ssp. avenaceum, F. a. ssp. nurragi and F. a. ssp. aywerte should be recognised as distinct species.

Bickel, D.J., Entomology Section, Australian Museum, 6 College Street, Sydney NSW 2000.

Tramps and Stowaways from Southern Australia: Biogeographic Processes in a World of Easy Transport

The transoceanic dispersal capability of long-legged flies (Diptera: Dolichopodidae) is well-established, as they are part of the small-sized fauna carried easily by wind or on flotsam. However, human transport is an increasingly important dispersal agent. Can we distinguish between naturally dispersed species and those whose distributions are the result of accidental introduction?

Examples of the increasing influence of human transport on species distribution are presented. Consider the following questions: Why is Perth an "urban island"? What is the origin of a species known only from greenhouses in Holland and France? How did a common temperate Australian species become parthenogenetic in the mountains of the Hawaii? How much more disjunct is possible: New Zealand dolichopodids (and planarians) in Scotland and Ireland?

Blakemore, R.J. Queen Victoria Museum, Wellington St., Launceston, 7250.

The taxonomic status of the earthworm fauna from Lake Pedder, Western Tasmania World Heritage Area - a survey and report commissioned by the NPWS, April, 1996.

This paper describes a total of 24 oligochaete taxa collected during a survey of Lake Pedder in early 1996. These consist of two known and fourteen new megascolecid species (five being subspecies) belonging to eight endemic genera: Vesiculodrilus, Diporochaeta, Perionychella, Atlantodrilus gen. nov., Notoscolex, Aprorodrilus gen. nov., Gastrodrilus gen. nov. and Anisocheta. Five introduced megadrile species were also found (four lumbricids and one megascolecid: Anisochaeta dorsalis (Fletcher, 1887) - this the first record of an introduction from mainland Australia), plus three new species of microdriles (one tubificid and two haplotaxids). These findings attest to the high diversity and high endemicity of oligochaetes in Tasmania.

The purpose of the survey was to determine the conservation and taxonomic status of Lake Pedder earthworms, particularly the only earthworm species previously described from this region, one collected from the original lake's edge prior to its hydro-electric impoundment in 1972. Reinspection of the type, and only, specimen of Perionychella (Vesiculodrilus) pedderensis Jamieson, 1974 showed that it had been inadequately described, was probably of local derivation and was sufficiently unique to warrant separate generic status, in monotypic Atlantodrilus. Although several of the new species described herein are possibly closely-related, occupy a similar habitat and superficially resemble A. pedderensis, none are conspecific. It is therefore concluded that its inclusion on the Parks and Wildlife Service's list of endangered species is justified; most probably A. pedderensis is extinct through loss of its original and unique habitat.

Bower C.C. NSW Agriculture, Locked Bag 21, Orange, NSW 2800 Australia

Defining Cryptic Species of Sexually Deceptive Orchids (Orchidaceae: Caladeniinae) using the Biological Species Concept

Sexually deceptive orchid species in the genus Chiloglottis R.Br. are shown to be reproductively isolated by having specific thynnine wasp pollinators and, by contrast to many other plant groups, conform closely with Mayr's Biological Species Concept. Reproductive isolation by pollinators was used in the Chiloglottis palachila D.L. Jones and C. formicifera Fitzg. groups to distinguish cryptic species which are unlikely to have been detected by morphological analysis. Specific pollinators can be seen as ecological taxonomic characters and highlight the potential value of ecological discontinuities for taxonomic analysis. Speciation in sexually deceptive orchids appears to be based mainly on the generation of differences in volatile chemical attractants, not morphological divergence. Such systems, as in the case of Chiloglottis, are likely to contain many cryptic species. Perhaps uniquely, field pollinator choice experiments with translocated Chiloglottis flowers allowed clear assessments to be made of the biological species status of closely similar allopatric populations. In this regard the objection that the Biological Species Concept cannot be easily applied to allopatric populations does not apply to sexually deceptive orchids.

Braby, M.F. CSIRO Entomology, GPO Box 1700, Canberra, ACT, 2601.

The Australian endemic genus Nesolycaena (Lepidoptera: Lycaenidae)

The family Lycaenidae comprises about one third (35%) of the Australian butterfly fauna, containing 139 species referred to 46 genera. Nearly half of the species (48%) and nine of the genera (20%) are endemic. One Australian endemic genus is Nesolycaena Waterhouse & Turner, 1905, which contains four species. It belongs to the tribe Candalidini which is confined to the Australian zoogeographic region, and is closely related to the genus Candalides. Nesolycaena is distinguished by possessing at least six unique characters. These include a relatively short antenna, a relatively short labial palpus, the venation of the forewing, the presence of bluish-white scales in the male, pupal morphology, and larval food plant specialisation on Boronia (Rutaceae). All four species are allopatric, having limited distributions in the north and east, and the populations of each are very patchy, being restricted to sandstone gorges and escarpments where the larval food plants grow. Until recently Nesolycaena was considered monotypic, however, recent discoveries of populations in the Kimberleys, WA, and northern Qld have uncovered new species. One of these, N. medicea, was discovered whilst I was on field trip in the White Mountains near Hughenden, Qld, in October 1991. The background to this discovery is briefly outlined, and the new species compared with N. albosericea, N. urumelia and N. caesia. The taxonomy, zoogeography and food plant relationships of the genus are reviewed.

Broadhurst, L. M., D. J. Coates & B.H. Tan, Curtin University of Technology, Perth, Western Australia and Dept. of Conservation and Land Management, Perth, Western Australia

Geleznowia verrucosa Turcz. (Rutaceae) - A Monotypic Genus ?

Geleznowia verrucosa Turcz. (Rutaceae) is a little known monotypic genus endemic to the sandplains north and east of Perth, Western Australia. Populations of this woody shrub are small, disjunct and morphologically variable. Although classified as monotypic, three forms can be recognised in the field by height, habit, flower size and abundance; forms rarely co-occur. Sixteen allozyme loci in 19 populations were examined to investigate genetic diversity in disjunct and morphologically variable populations and to determine whether different taxa existed. Single locus diversity measures portrayed Geleznowia as a somewhat genetically depauperate genus (A, 1.3; P, 29.3%) with total genetic diversity (Ht, 0.226) partitioned among (Dst, 0.126) rather than within (Hs, 0.100). Approx. 56% of the total genetic diversity attributable to interpopulational differences. While endemism, bottlenecks and founder effects and small population size have probably contributed to the levels of genetic diversity found within Geleznowia, it is evident from this investigation the genus is probably not monotypic. The partitioning of genetic diversity along morphological lines, morphologically diagnostic alleles and possible differences in reproductive strategy suggest this enigmatic species complex consists of at least two taxa. Additionally, it would also appear hybridisation has occurred independently on several occasions.

Chappill, J.A. Department of Botany, The University of WA, Nedlands, WA 6907

Systematics of Jacksonia (Leguminosae)

Jacksonia is a genus of legume from the endemic Australian tribe Mirbelieae that shows its greatest diversity in northern and southwestern Australia. The genus is defined by the presence of peltate hairs on the stems and calyx and by having hairs inside the ovary. It shares with the bird-pollinated genus Leptosema the reduction of the leaves to small scales in the adult state. Many of the southwestern species are either very spiny or are small rush-like plants and this rather unattractive or inconspicuous appearance when not in flower has led to Jacksonia being largely neglected by taxonomists in the past. Recent study, however, has identified a large number of new species increasing the size of the genus from 37 to 77 species. This talk will discuss the diversity and evolution of this genus.

Conran, J.G., P.G. Martin†, J.M. Dowd. Department of Botany, The University of Adelaide, SA 5005, C.J. Quinn. School of Biological Science, University of New South Wales, PO Box 1, Kensington, NSW 2033, P.A. Gadek. Department of Tropical Plant Science, James Cook University - Cairns Campus, PO Box 6811, Cairns, Queensland 4870 & R.A. Price. Department of Botany, University of Georgia, Athens, Georgia 30602, USA

The relationships of the Gymnosperm genera Lagarostrobos, Manoao (Podocarpaceae) and Phyllocladus (Phyllocladaceae) based on an analysis of the CP gene rbcL.

Analysis of sequences of the chloroplast gene rbcL for 13 genera of the Podocarpaceae, and representatives of the Phyllocladaceae and Araucariaceae were undertaken. The results indicate that the Podocarpaceae appear to be polyphyletic unless expanded to include the Phyllocladaceae, and support the recent removal of the New Zealand species of Lagarostrobos to its own monotypic genus Manoao. There are also a number of revised relationships suggested between the other Podocarpaceae genera sampled in comparison with previously proposed traditional morphological affinities.

Cranston P.S., CSIRO Entomology, Box 1700, ACT 2601.

What's in a species ? Are higher taxa and other composites surrogates for any useful measures?

Two topical issues - biodiversity survey and biomonitoring studies - are increasing user expectations of taxonomists. Demands include wider taxonomic coverage, more responsivity to user needs, more friendly methods of delivery, and rapid identification/validation of voucher materials. These, and other, increased expectations come at a time of unprecedented withdrawal of traditional sources of support for systematics. Criticisms aimed at systematists include the study of "esoterica" such as phylogenetics and biogeography, the slow pace of taxonomic progress. and the failure to deliver in a manner that allows adoption by users without intermediaries to translate. This perception lies behind two short-cuts adopted by some biologists - the use of Recognisable Taxonomic Units in the sense of Oliver and Beattie (dealt with elsewhere and not considered in this talk) - and/or the use of higher taxa as surrogates for species. The latter is advocated, inter alia, for reducing the time taken to identify samples and the training required for personnel, whilst retaining environmental "signal" in the data. I will review some studies in aquatic systems, where responses of higher taxa do appear to act as surrogates for species, and others in which signal is obscured. The remainder of the presentation will concern theoretical issues concerning the circumstances under which higher taxon surrogacy might succeed or fail. What is required still are studies in which assemblage responses to disturbance (for example) at species-level are examined at different taxonomic levels. There seems little justification in biodiversity studies for the approach, phylogenetic diversity notwithstanding.

Crisp, M.D. Division of Botany and Zoology, Australian National University, Canberra, Australian Capital Territory, 0200.

What can hybrids tell us about phylogeny in plants?

Before cladistic analysis became widely used from the 1980s onwards, phylogenetic relationships among sexual organisms were estimated using 'biosystematics'. This approach derived from the neo-Darwinian synthesis which emphasised population genetics and the biological species concept. In this paradigm, reproductive isolation was considered the main criterion for delimiting species. Research was directed towards identifying isolating mechanisms and testing isolation by crossing experiments. Its continuing influence is shown by classical text books that are still in print, e.g. Briggs and Walters 'Plant Variation and Evolution.' With the advent of a range of molecular techniques for investigating gene flow, crossing experiments have gone out of vogue; nevertheless, reproductive isolation or its converse, crossability, is still a widely accepted criterion of phylogenetic relatedness, at least at the lower taxonomic levels. Donn Rosen, an early cladist, challenged this paradigm, pointing out that ability to interbreed is a plesiomorphic condition of populations, and can be (and often is) retained following allopatric speciation. The latter has been known empirically for a long time, and neo-Darwinians erected the ad-hoc hypothesis of 'reinforcement' to account for the awkward delay in reproductive isolation following incipient speciation. Reinforcement has never been demonstrated convincingly. An empirical test of the biosystematic method would be to examine the evolution of reproductive isolation or crossability using a phylogeny. The present paper does this using several examples from plant taxa.

Crisp, M.D. & Gilmore, S. Division of Botany and Zoology, Australian National University, Canberra, Australian Capital Territory 0200.

A phylogenetic analysis of nuclear ribosomal intergene spacers (ITS1 and ITS2) in the Australian legume tribes Mirbelieae and Bossiaeeae

The Mirbelieae and Bossiaeeae are endemic Australian tribes of pea-flowered shrub legumes, together including 36 genera and 700 species. Earlier analyses using morphology suggest that this group is monophyletic, with floral synapomorphies such as red and yellow corolla markings. Its relationship to other tribes is unclear, although morphological analyses have linked it with the 'genistoid alliance', distributed mainly in Europe and Africa. Earlier studies suggested a close relationship between the Brongniartieae (Australia and tropical America) and Mirbelieae/Bossiaeeae, but more recent analyses by Chappill (morphology) and Doyle et al. (rbcL) do not support this. Studies by Van Wyk, as well as that by Chappill, align the African tribes Podalyrieae and Liparieae closely with Mirbelieae and Bossiaeeae.

To test these relationships, we have sequenced ITS1, ITS2 and the intervening 5.8S gene. Our samples include > 100 species representing all the above tribes. We have sequenced all 30 genera of Mirbelieae and most of those in the other two Australian tribes. The ITS regions are relatively conserved in our study group. The sequence length (including 5.8S) varies between 651 and 671 bp. Among closely related genera, e.g. Brachysema, Nemcia and Jansonia, there is very little variation, and the ITS sequences do not resolve their relationships. At higher taxonomic levels variation is informative, and a preliminary parsimony analysis of 30 sequences representing the Mirbelieae gave a tree broadly congruent with that derived by Crisp and Weston using morphology.

Donnellan¹ S.C. & P. Monis²: ¹Evolutionary Biology Unit, South Australian Museum, Adelaide SA 5000: ²Present address: Microbiology Department, University of Adelaide SA 5000.

Molecular systematics of the Australo-Papuan Hylid frogs

The Australo-Papuan treefrogs (Family Hylidae) are a major component of the region's anuran fauna, presently comprising 158 species in three genera, Cyclorana, Litoria and Nyctimystes. It has long been felt that the present generic classification does not adequately reflect either the immense morphological and ecological diversity in the group nor does it reflect our notions of evolutionary relationships within the family because both Cyclorana and Nyctimystes appear to be paraphyletic with Litoria. We present a phylogenetic analysis of nucleotide sequences of two mitochondrial genes from 38 species representing 26 species groups of Litoria, two of the Cyclorana lineages and both Australian and New Guinean Nyctimystes.

Three major clades are present in the mitochondrial gene tree none of which are completely concordant with previous notions of relationships of these frogs and highlight the plastic nature of the body plan and ecological evolution of the group. At the higher systematic level, the data confirm the paraphyly of both Cyclorana and Nyctimystes with Litoria. Furthermore, the Australian N. dayi appears to be related to the torrent frog group (L. nannotis species group) from north Queensland and the New Guinean Nyctimystes are related to the widespread L. infrafrenata and the Australian L. brevipalmata. The polyphyly of Nyctimystes is unexpected and we are presently testing that the mitochondrial sequences obtained are indeed mitochondrial.

Dowton, M.^1,2 & A.D. Austin¹. ¹ Department of Crop Protection, Waite Campus, The University of Adelaide, P.O. Glen Osmond, 5064; and ² Department of Biology, Wollongong University, Wollongong, N.S.W. 2522.

16S rDNA sequence data suggests a novel evolutionary transition from endoparasitism to ectoparasitism in the Braconid (Hymenoptera: Braconidae) wasps.

The Braconidae are a family of ecto- and endoparasitic hymenopteran wasps, which attack narrow host ranges at the subfamily level. As such they provide an important working model in which to examine the transition between these two parasitic lifestyles, as well as the transition between different host groups. We sequenced an homologous portion of the 16S rDNA gene across 59 members of the Braconidae, encompassing 28 subfamilies. Parsimony analyses suggested that endoparasitism of lepidopteran or coleopteran larvae is plesiomorphic for the family. Radiation into hosts other than these (such as Diptera, aphids, other Hymenoptera) occurred relatively recently from an ancestor that attacked lepidopteran or coleopteran larvae. Surprisingly, endoparasitism appears to be plesiomorphic for the family, with our analyses suggesting a single reversion to ectoparasitism in the cyclostome braconid families, in agreement with previous suggestions that the cyclostomes represent a natural, derived braconid group. An alternative explanation, favoured in the literature, is that all of the ectoparasitic ancestors to the earlier diverging Braconidae are now extinct.

Faith D. P. CSIRO Division of Wildlife and Ecology, P. O. Box 84, Lyneham, ACT 2602.

Phylogenetic diversity and conservation evaluation

Phylogenetic pattern has the potential to provide predictions about biodiversity patterns among taxa at the so-called "feature" level. A simple phylogenetic diversity measure, "PD", estimates the relative feature diversity of a set of taxa by the total length of the path spanned by these taxa on the corresponding phylogenetic tree. This measure has desirable properties relative to the many others defined over the past decade. One operational and conceptual weakness, however, may be PD's implicit assumption that branch length estimates are informative over a wide universe of characters of organisms. PD can be used to set conservation priorities within a single group of organisms, such as orchids. Additionally, PD may be useful in setting conservation priorities on areas, if PD can boost the prediction of general biodiversity patterns among areas. In a case study in preparation with Drs. Greenslade and Colloff on Tasmanian invertebrates, PD provided little additional information beyond species-counting. In another study, with C. Moritz, that applied PD to sets of gene trees for vertebrates, PD provided information that suggested general predictivity of biodiversity patterns for the given areas. We concluded that PD will be useful when its values are highly congruent over different phylogenies, suggesting unifying process explanations.

Foley, D.H., Bryan, J.H., Yeates, D.¹ & A. Saul². Tropical Health Program ACITHN and Department of Entomology, University of Queensland, Brisbane, Qld 4072, Australia; ¹Department of Entomology, University of Queensland, Brisbane, Qld 4072, Australia and ²Tropical Health Program ACITHN, Queensland Institute of Medical Research, Brisbane, Qld, Australia.

Evolution and systematics of Anopheles: insights from a molecular phylogeny of Australasian mosquitoes.

Relationships among the genus Anopheles and its many sibling species-groups are obscure despite the importance of anophelines as the vectors of human malaria. For the first time, the interrelationships and origin of Australasian members of the subgenus Cellia are investigated by a cladistic analysis of sequence variation within the mitochondrial cytochrome oxidase subunit II gene. Phylogenetic analysis suggests that two-way exchanges with Oriental mosquitoes rather than only immigration may have been a characteristic of anopheline paleobiogeography in Australasia. Estimated divergence times between many Australasian and Oriental taxa predate the mid Miocene collision of Australasia and Southeast Asia. The Australasian fauna is mostly included in a large clade. The medically important Punctulatus Group is monophyletic and appears derived from Oriental stock. Populations within this group from as far apart as Australia and Vanuatu were in contact in the recent past (i.e. 0.35-2.44 mya), supporting dispersal rather than vicariance explanations. Some support for the monophyly of the Myzomyia, Neomyzomyia, and Pyretophorus Series was found. However, the subgenera Anopheles and Cellia and the Neocellia Series are paraphyletic but branch support at these taxonomic levels was poor. The COII gene shows promise for questions concerning alpha taxonomy but appears to be of limited use for resolving deeper relationships within the Anopheles.

Gowlett-Holmes, K.L., & J.A. Koslow. CSIRO Division of Marine Research, GPO Box 1538, Hobart, Tasmania 7001.

The seamount fauna south of Tasmania.

In January 1997, CSIRO surveyed the benthic fauna on south Tasmanian seamounts that peaked at depths between 550 and -1800m. Seamounts to a depth of -1200m are trawled for orange roughy and oreo dories. Fourteen seamounts peaking at -1500 and 1800m are within a recently-declared interim marine reserve. The aims of the survey were to assess the impact of trawling on the seamount fauna and to determine its diversity, depth distribution and zoogeography. Sampling was carried out with a still camera, benthic sled, dropline and several types of traps. Most of the species collected are still being identified, but preliminary results indicated that many represent new records for Australia or are undescribed. While some species have very wide distributions, such as the colonial stony coral, Solenosmilia variabilis, which forms thick meshlike beds over the summit and sides of most of the seamounts, comparison with material collected from New Zealand seamounts indicate that many species, while related, are not the same. Preliminary comparison of the south Tasmanian material with specimens from orange roughy fishing grounds in the Great Australian Bight show similar results. The marked differences in species composition from seamounts in these three areas have interesting biogeographical and conservation implications, as do the apparent differences in faunal composition from seamounts in different depths and between those that are fished and unfished.

Greenslade, Penelope. CSIRO Entomology, GPO Box 1700, Canberra ACT 2601.

Biogeography and ecology of the Gondwanan genus Setanodosa (Collembola: Brachystomellidae).

The genus Setanodosa currently consists of 14 described species, two from New Zealand, two from Australia, one from both Australia and South Africa, seven from South America, one from Hawaii and one from the subantarctic island of South Georgia. There are several undescribed species known in Australia. A phenetic analysis indicates that the Hawaiian species and one of the South American species may have been recently introduced, possibly from either Australia or South Africa. The species have different and wide ranging climatic and vegetational preferences including warm temperate rainforest, cool temperate sedgeland, semi arid woodland, arid chenopodiaceous shrubland and subantarctic tussock grassland. Within these vegetation types, species are restricted as to habitat but can occur at extremely high densities. A possible physiological basis for the large ecological differences between species will be suggested and its relevance to the evolution of species within the genus will be discussed.

Greenwood¹, David R., Anthony Vadala² and Jack G. Douglas^{3. 1}Dept. of Biological & Food Sciences, Victoria University of Technology, PO Box 14428, MCMC, Melbourne, VIC 8001; ²Cookson Laboratory, School of Botany, The University of Melbourne, Parkville, VIC 3052; ³42 Summerhill Road, Mt. Waverley, VIC.

Victorian Tertiary macrofloras: a conspectus.

Victoria has a rich record of Tertiary vegetation represented in a diverse range of macrofloras ranging in age from Paleocene to Pliocene. Recent systematic analyses of Tertiary plant macrofossils and summations of Australian Tertiary macrofloras have focussed on key sites such as Anglesea and Berwick. However, a review of the literature of the last 100 years, museum collections and a field survey of sites over the past 20 years has demonstrated that the Victorian Tertiary macrofloral record is more extensive than is perhaps realised. These macrofloras range from the well known but poorly documented laterally extensive and thick brown coal and associated interseam clays of the LaTrobe Valley, to largely unknown small clay lenses with leaf impressions or scattered plant fragments in the Dargo and Bogong High Plains. Most sites preserve leaf compression (or mummified), or leaf impression assemblages, or accumulations of seeds, fruits or silicified wood fragments. Rare sites include tree stumps in growth position, in some instances representing fossil forests, and amber containing fossil insects and leaves. The Victorian Tertiary macrofloras are discussed here to provide a comprehensive database as a reference source. Synthetic analyses of such databases will facilitate interpretation of the evolution of southern temperate forests in southeastern Australia.

Groom, P.K. & B.B. Lamont. School of Environmental Biology, Curtin University of Technology, GPO Box U1987, Perth, Western Australia 6845.

Fire response influences biogeography of Banksia and Hakea (Proteaceae) species in southwestern Australia.

Two main life forms have evolved among the Proteaceae in relation to fire - nonsprouters and resprouters. Nonsprouters are killed by fire, relying solely on seed production for post-fire recruitment. Resprouters usually survive fire, regrowing from a lignotuber, epicormic buds or suckers. In southwestern Australia, nonsprouters account for more than half the species (65% and 58% respectively) within the genera Banksia and Hakea, with the highest proportion of nonsprouters (per 50 x 50 km grid cell) occurring in the southern sandplains (Kalgan River - Esperance). The northern sandplains (Eneabba-Mt Lesueur) contain a higher proportion of resprouters. For both genera the greatest proportion of resprouters occurs in the forested region of southwestern Australia (Darling botanical district), a region of high (and very seasonal) annual rainfall. The relative aridities and fluctuations of present and past (Quaternary and Tertiary) climates may explain the much higher proportion of resprouters in the northern than southern sandplains. The high proportion of resprouters in the extreme southwestern corner for both genera might be explained by the elimination of nonsprouters through frequent burning by Aborigines in the late Quaternary.

Haegi L.A.R. Plant Biodiversity Research, Botanic Gardens of Adelaide & State Herbarium, North Terrace Adelaide, SA 5000

Sustaining biological diversity: reciprocal priorities for conservation and systematics

Over the past two decades a steadily increasing stream of literature has promoted a direct role for systematics in the conservation of biodiversity. Calls for modified priorities, usually in systematics, have come from both systematists and conservation biologists. Progress in achieving a co-operative approach has been slow, with a majority of practitioners from both disciplines continuing to accept a passive role for systematics. In Australia demands on systematics institutions by conservation agencies generally do not relate to newly identified needs arising from theoretical advances in conservation biology, but rather to meeting the practical needs such as specimen identification services, the ability to 'mine' the collections selectively for specimen-based information (eg on threatened species) and for what electronic data on taxon distribution is available. Consistent with their traditional role the institutions housing biological collections generally comply with such demands but find increasingly it is at their cost. In a climate of reducing funding they are at risk of decline as their important contribution to biodiversity conservation is obscured and funding is redirected to those producing more visible applied outcomes. At the same time decision making in biodiversity management is compromised through failure to ensure it has an adequate taxonomic basis.

There is an urgent need for biodiversity management agencies and systematics institutions to work in partnership. Their aim should be a strategic coordinated approach for achieving improved outcomes for biodiversity conservation. They have a shared responsibility to promote recognition of systematics as a key discipline in biodiversity conservation, to ensure more effective use of biological collections and to identify improved ways of applying advances in taxonomic knowledge.

Harvey¹, Mark S. and P. Cranston^{2 1}Department of Terrestrial Invertebrates, Western Australian Museum, Francis Street, Perth, W.A. 6000; ²Division of Entomology, CSIRO, GPO Box 1700, Canberra, A.C.T. 2601

Accidental occidentals?: East-west faunal connections in temperate Australia

The ancient, eroded Western Australian landscape is well-known for its highly varied vascular plant flora, especially in the Transitional Rainfall Zone (300-800 mm annual rainfall) of the south-western temperate region. There is little to suggest that similar diversity occurs amongst the terrestrial and inland aquatic fauna, but some groups, notably amongst the invertebrates, are beginning to show patterns of localised endemicity and high diversity.

The faunal assemblages of Western Australia's High Rainfall Zone (800-1500 mm annual rainfall) contain numerous elements with distinct Gondwanan affinities and some are thought to be extremely ancient Pangean relicts. We examine the relationships of selected animal groups found within the High Rainfall Zone and adjacent regions, and attempt to reconcile present distributions with past geological and climatic events. This survey shows a strong relationship between the faunae of south-western and south-eastern Australia, which in turn shows a strong relationship with the faunae of other portions of Gondwana. The presence of some invertebrate groups with sister-species in southern Africa, but not in eastern Australia, highlights the diverse origins of the temperate fauna of Western Australia, many of which are of immense age and have persisted in habitats that have remained stable since the Mesozoic. The 'accidental' nature of the origins and persistence of the south-western Australian temperate fauna are also examined.

The lack of fully resolved cladograms is thought to hamper any serious attempts to analyse the trans-continental relationships and origins of the temperate fauna of southern Australia. In addition, it is not fully known whether the long-term effects of clearing and burning of native vegetation in the High Rainfall Zone of Western Australia will have a long-term deleterious effect on the fauna.

Hill¹, R.S. & L.J. Scriven². ¹Department of Plant Science, University of Tasmania, GPO Box 252-55, Hobart, Tasmania 7001 and ²Darwin Botanic Gardens, PO Box 496, Palmerston, Northern Territory 0831.

Cenozoic climate change and biogeography in south-eastern Australia.

Major climate changes occurred in south-eastern Australia during the Cenozoic, mostly associated with the breakup of Gondwana and continuing continental movement in the Southern Hemisphere. There is little evidence for any kind of long term plant community integrity during these climatic changes, but there is a great deal of evidence for the response of individual taxa.

Four main responses can be identified:

1. Taxa evolved morphologically, apparently in response to climate change, and the descendants retained something like the original range.

2. Taxa have not changed morphologically and retain something like the original range.

3. Taxa, or their descendants, have moved away from their Paleogene range, apparently by "following" a suitable climate.

4. Taxa have become locally or globally extinct and have left no descendant.

It is important to identify taxa within these groups and examine them for trends that may be useful in determining adaptations that allowed successful long term residence in the region. The clearest adaptations are for xeromorphy and scleromorphy, whereas those taxa well adapted to high latitude photoperiods were very prone to extinction.

Hjelmroos M., A. Jones & E.R. Tovey. Institute of Respiratory Medicine, University of Sydney, Sydney N.S.W. 2006.

Differentiation of fresh, airborne pollen grains by image analysis

The globally increasing frequency of asthma and hay fever requires reliable and rapid methods to measure and analyse the atmospheric pollen and spore content. In this study a computerized image analysis system was used to obtain measurements of the airborne allergenic pollen grains from 32 selected species. The investigated material includes pollen grains from 8 grass species (Cynodon dactylon, Dactylis glomerata, Holcus lanatus, Lolium perenne, Phalaris, Phleum pratense, Phragmites and Poa), 4 species of Ambrosia, Artemisia, 3 species of Betula, Alnus, Carpinus, Corylus, Ulmus, Quercus, Olea, Cryptomeria, Castanea, Cupressus, Juniperus, Fraxinus, Pinus, Salix, Parietaria, Urtica, Plantago, Rumex and Chenopodiaceae. Material was collected from well identified, flowering plants in Europe, USA, South-Africa, Japan and Australia. Pollen grains were mounted directly in the glycerine jelly without any laboratory treatment and 200 pollen of each sample were examined. Quantitative measurements of pollen grains from each species were analysed to detect differences in morphological structure between the individual species and between replicates of the same species having different provenance. Discrimination was based on canononical analyses of 7 basic feature measurements obtained from images of individual pollen grains using a high resolution video, an image processing software, and programs developed in-house. By means of these seven basic feature measurements it was possible to classify the pollen grains to the family level. It was also, to some extent, possible to differentiate between the 8 individual grass species, and between the 4 different Ambrosia species.

Hudson,P.J. Department of Zoology, University of Adelaide, South Australia, 5005.

Biogeography of the terrestrial invertebrate fauna of normally dry salt lakes in Australia.

A variety of terrestrial invertebrates such as spiders, scorpions, crickets and beetles occur on the normally dry salt lakes of the temperate region of Australia. Most are endemic to salt lakes and quite a few are flightless or not capable of aerial dispersal. Whilst our knowledge of this fauna is far from complete it is clear that they have an interesting biogeography - some species being restricted to particular lakes whilst others have a regional or far wider distribution.

In South Australia the wolf spider Lycosa eyrei, the scorpion Australobuthus xerolimniorum, the tiger beetle Megacephala whelani and the Cicindela ozellae species group of tiger beetles have similar distributions. These and other examples from South Australian and Western Australian salt lakes will be used to illustrate some of the broad trends in distribution observed to date and to highlight some of the implications for the conservation and management of the fauna.

Jacobs T.V. Department of Botany, University of Transkei, Private Bag X1, UNITRA, Umtata, Republic of South Africa.

Systematics and conservation status of macrophytes of South Africa

Including river mouths, coastal lakes and lagoons, South Africa has 296 estuaries. Besides rainfall patterns, the geomorphology of coastal regions has a profound influence on the biotic diversity of estuarine systems. The flora surrounding the estuaries in subtropical Natal are different from the vegetation of the more temperate south coast. Macroflora inhabiting South African estuaries fall under the following categories: floating macrophytes, emergent macrophytes, salt marsh macrophytes, submerged macrophytes and macroalgae. Species composition of estuarine macrophyte communities varies with salinity. Status of nutrients has a bearing on macrophytes. Submerged macrophytes and emergent reeds act as nutrient sinks during the growing period. Presence or absence of estuarine plant communities is directly linked with estuarine food web.

Some of the factors that affect the health of estuarine ecosystems are: excessive water abstraction from catchments; uninformed agricultural practices that lead to soil erosion and silt deposition; residential, industrial and transport developments in the vicinity of estuaries; various forms of pollution. With the help of a Numerical Index Value (Raal & Burns, 1966), it has been estimated that only 26% of South African estuaries are in a healthy condition. Using biotic diversity as an index, estuaries with high conservation status need to be identified so that the Consortium of Estuarine Research and Management can put pressure on relevant authorities to pay particular attention to the conservation needs of a given estuary.

Keogh, J. Scott. Evolutionary Biology Unit, South Australian Museum, North Terrace, Adelaide, SA, 5000 and School of Biological Sciences A08, University of Sydney, 2006.

Molecular phylogeny and biogeography of the venomous Australo-Papuan elapid snakes.

To discriminate among existing phylogenetic and biogeographic hypotheses for the elapid snakes, portions of the cytochrome b and 16S rRNA mitochondrial DNA genes were sequenced from 19 of the 20 terrestrial Australian genera, six of the seven terrestrial Melanesian genera, a sea krait (Laticauda), a true sea snake (Hydrelaps), and representatives of the American, African, and Asian elapid radiations. These data clarify several significant issues in elapid phylogeny: Afro-Asian cobras form one or more sister groups to other elapids, American and Asian coral snakes form a clade, corroborating morphological studies, Melanesian elapids form sister groups to Australian species, indicating that the ancestors of the Australian radiation came via Asia, rather than representing a relict Gondwanan radiation, the two major groups of sea snakes (sea kraits and true sea snakes) represent independent invasions of the marine environment, and the radiation of viviparous Australian elapids is much older than has been suggested from immunological data. Parsimony analyses were unable to resolve relationships among the Australian radiation, a problem previously encountered with analyses of other (morphological, electrophoretic, karyotypic, immunological) data sets on these species. The large divergences even within genera suggests that the reason for this continued difficulty lies in the timing of speciation events.

Ladiges, Pauline Y. School of Botany, The University of Melbourne, Parkville, Vic., Australia 3052

Phytogeography of continental Australia: west-east area relationships across the southern zone

For Burbidge (1960) patterns of distribution of endemic and non-endemic plants and high levels of endemicity indicate that genera in the southern temperate zone have a long history and represent ancient forms. Crisp et al. (1995) published the first area cladogram for continental Australia based on various plant groups. They included genera with representatives in the west and east of temperate southern Australia as well as taxa with northern distributions. They found a significant level of congruence among taxa and concluded that: (1) differentiation of Tasmania was the earliest vicariance event; (2) monsoon tropical areas (Kimberley, Arnhem, Northern Desert, Cape York, New Guinea) are related and as a group (track 1) differentiated from a group of areas (track 2) along the east and south coasts of Australia; (3) within track 2, differentiation of the south-west group (South-West Western Australia and related Western desert) was the earliest event.

Reanalysis of Crisp et al.'s data using subtree analysis and the program TASS, suggests some differences in the sequence of events, with the oldest vicariance event being isolation of the south-west sensu lato (South-West and Western Desert, also related to Pilbara and Eastern Desert), followed by Tasmania, then the Kimberley, Arnhem and Northern Desert regions from the eastern and southern regions, which group as a track (New Guinea, Cape York, Atherton and Eastern Queensland being related as a group to Eyre, Adelaide, Victoria, MacPherson-Macleay and eastern New South Wales). Analysis of Boronia section Valvatae, which occurs in many of these areas but not all, is congruent with this pattern and supports the hypothesis that vicariance of western and eastern Australia across the southern zone is one of the earliest events in Australian biogeography.

West-east relationships possibly have their roots in the Cretaceous. Taxa of different ages will reflect different histories since that time with on-going changes in climate, soils and sea-levels, and expansion and contraction of the boundaries of biota. Geographic variation in the soap mallee, Eucalyptus diversifolia, the only monocalypt eucalypt that traverses southern Australia from west to east, is an example of a pattern younger than that of Boronia.

Lambkin C.L. Systematic Entomology Laboratory, Entomology Department, University of Queensland, Brisbane, Queensland 4072.

Characters, Congruence and Bee Flies (Diptera: Bombyliidae)

In recent years, phylogenetic signals from biochemical and morphological data sets have been evaluated. Data sets in agreement have been termed congruent, and those in disagreement have been termed incongruent. Information from two independent character systems can be synthesised in two separate but related ways: compare the phylogenies derived from the two data sets separately using tree based methods of comparison, taxonomic congruence; and compare the information in a phylogeny derived from a combination of the two data sets using character based methods of comparison; character congruence. The opposing viewpoints of taxonomic and character congruence have become one of the most important issues in modern systematics. I used character congruence to examine character agreement in a group of Australian bee flies, between each of seven character partitions and the rest of the morphological data using several incongruence indices, including IM (Miyamoto 1981), and IMF (Mickevich and Farris 1981) and several derivations of IMF.

I will discuss character congruence using IMF and measures of the significance of character incongruence using XARN (Farris et al 1995). IMF found higher levels of incongruence for the wing pattern partition when compared to the rest of the morphological data. XARN revealed that the wing pattern and vestiture partitions were significantly incongruent with the rest of the morphological data. This incongruence supports the findings of taxonomists in studies of the group in other parts of the world. While the wing patterns are easily observed, appear consistent and distinctive, species-groupings based on wing patterns are rarely supported by other morphological characters. Quantitative cladistic methods such as character congruency studies can provide a measure of the taxonomic value of a character system for species-grouping.

Leijs, Remko. School of Biological Sciences, Flinders University of South Australia, GPO Box 2100, Adelaide, SA 5001

Mating strategies and speciation rate in large carpenter bees (Xylocopa).

The genus Xylocopa or carpenter bees consists of 36 subgenera and include world wide about 500 species. The different subgenera exhibit diverse mating strategies. In some taxa, mating occurs predominantly at sites that contain resources for females (flowers, nesting sites). Males may or may not defend territories in the neighbourhood of such resources. In other subgenera, males defend non resource territories. Leks may be located on hill tops or between branches of trees. The subgenera that are basal in the phylogeny of Xylocopa all have resource based mating systems. Lekking seems to be a derived character that has evolved at least three times in the group. In subgenera with resource based mating systems, sexual colour dimorphism is usually not very marked, whereas taxa with lekking behaviour usually show strong sexual dimorphism. Furthermore female choice seems to be more predominant in lekking species than in non lekking species. This creates potential for accelerated speciation through drift and sexual selection, more so for lekking than for non lekking species. Large carpenter bee subgenera that exhibit lekking behaviour are more speciose than the non-lekking groups. It is likely that this difference is due to sexual selection and drift operating more strongly in lekking species.

The various possibilities will be compared and discussed.

Linder H.P. Bolus Herbarium, University of cape Town, Rondebosch 7700, South Africa

Case studies in adaptations in the Cape Flora, South Africa.

Adaptations imply a changed morphology (or behaviour) in response to a particular environmental factor. It is very difficult to substantiate that a morphological change is an adaptation, rather than simply a morphological change. There appear to be two approaches: the first is phylogenetic, and attempts to reconstruct the evolutionary history (at least partially) of a lineage. The second approach is more geographically bound, and searches for explanations for morphological peculiarities in particular geographical areas. I will work through three examples, largely from the Cape Flora. In the first example I explore the evolution of the annual habit in the Danthonieae and Ehrharta (both south-temperate grass groups), showing that the habit is restricted to a portion of southern Africa, where it has evolved in several different lineages. The second example concerns the evolution of dispersal modes in the African Restionaceae, where dispersal modes range from atelic, to wind or ants. Here I explore the associated features of the syndrome, especially in seedling types. I will show that it is difficult to determine the selection force for myrmecochory, and suggest that the forces driving the diversification of dispersal units are remarkably complex. In the third example I will explore the alpine adaptations of the flora of the upper slopes of the Klein Swartberg mountains. Although there is little structural change in species from lower to upper mountain slopes, and phylogenetic approach is useful in testing whether there are any adaptations at all. These comparative methods allow us to test hypotheses in a comparative way, and to develop conceptual models of the evolution of particular organs, or the diversification of taxa.

Lockhart, P. & D. Penny, Plant Biology and Biotechnology, Massey University, PO Box 11222, Palmerston North, New Zealand.

New Problems and New Opportunities with Sequence Data

The rapid increase in the rate of obtaining DNA sequences is increasing the range of ideas to be tested at the same time it is increasing the difficulties of an objective analysis. Longer sequences mean that there are smaller limits on sampling errors but does not remove the problem of systematic errors in the analysis that can lead to very high bootstrap values for an incorrect result. At the same time, many more sequences, particularly within a species may lead to relatively low bootstrap values, even though the tree is relatively stable and new approaches to the analysis of large numbers of sequences will be described. Because trees are becoming more stable with longer sequences it is possible to test a wider range of hypotheses. Two examples to be discussed are the relative role of Gondwana origins versus mass extinctions in early bird and mammalian evolution, and the relative importance of the Oligocene reduction of land area, mountain building during the Pliocene, Ice Ages, refugia and long distance transport in the relation between some New Zealand and Australian plants. The overall theme will be approaches to testing a wider range of systematic questions.

Macdonald, J.A. ¹, Gillings, M. and Holley, M. ^{2 1} NSW Agriculture, Orange Agricultural Institute, Forest Rd, Orange 2800. ² Centre for Biodiversity and Bioresources, School of Biological Sciences, Macquarie University 2109

Use of DNA fingerprinting in systematic studies on mites(Acarina) of agricultural importance.

Both identifications and systematic investigations of mites normally involve considerable time and difficulty in the preparation of material for study. Identification often provide very limited information on important agricultural characteristics of the mite species(eg source of origin or resistance status) that may be critical for the purposes of quarantine or management. Further, morphological characters, in the broader context, often fail to provide relevant information for studies of systematic relationships between, and within, species.

In response to these difficulties, preliminary investigations into DNA fingerprinting research for mites of agricultural importance was proposed. In particular, our studies focused on three objectives which were; firstly, the possibility of developing a suitable methods transporting and storing mites for molecular studies; secondly, the development of a generic fingerprinting method for mites; finally, the investigation of this methodology for distinguishing different mite taxa.

These studies proved successful for all three objectives and results are presented and discussed in our presentation.

Mackay, D.A. & Whalen, M.A. School of Biological Sciences, Flinders University, GPO Box 2100, Adelaide, S.A. 5001

Variation in seed characteristics of some ant-dispersed Australian euphorbs

Seed dispersal by ants or myrmecochory is common among sclerophyllous shrubs in Australia and South Africa. Ant-dispersed seeds typically bear an appendage called an elaiosome that is attractive to ants. Several authors have suggested that morphological and chemical characteristics of elaiosomes may evolve to preferentially attract ants that are efficient as seed dispersers, rather than those ants which act primarily as seed predators. The elaiosome and seed sizes and the ratio of the elaiosome to diaspore mass are thought to influence the probability of seed transport by ants, as is the fatty acid composition of the elaiosomes. Little information is available, however, on the extent of variation among populations and among individual plants in these traits. We present information on the relative amounts of variation in these traits among plants and populations within the genus Adriana and we compare the fatty acid chemistry of elaiosomes of Adriana with that determined from other plant genera.

Matthews, E.G. South Australian Museum, North Terrace, Adelaide, South Australia 5000.

Biogeography of an austral tenebrionid beetle group.

The tribe Adeliini comprises 45 genera in Australia, Chile, New Caledonia and New Zealand, inhabiting mostly mountain forest leaf litter. In Australia the tribe is widely distributed, however there are two distinct areas of endemism, one in north-eastern Qld (the Wet Tropics) and one in south-eastern Australia as far north as southern Qld. Genera which are endemic to each area occur mostly in rainforest. Vicariance analysis of the whole tribe, based on a genus-level phylogenetic reconstruction, reveals only two well-supported sister-area relationships, which are to be seen mainly between endemic rainforest genera: one close one between genera of south-eastern Australia and of Chile, and one less close between those of the Wet Tropics and of New Caledonia. Sister-group links between the northern and southern forested areas of Australia occur only at species level within genera common to both areas. New Zealand taxa are not involved in convincing monophyletic relationships with those of any other single area. The patterns observed are consistent with a scenario where New Zealand separated first, by itself, from the remaining areas and where the break between the northern and southern forested areas of eastern Australia occurred before the isolation of the continent from New Caledonia and Antarctica/South America.

Morrison, David A. Key Research Centre for Molecular Parasitology, University of Technology Sydney, Westbourne St, Gore Hill, NSW 2065.

After Tree-Building: Assessing and Comparing Phylogenetic Trees

We can never know whether a cladogram actually represents the true phylogenetic history of the taxa. So, there are three further issues that need to be dealt with once a cladogram has been constructed:- assessing the magnitude of the phylogenetic signal in the cladogram; assessing the robustness of the monophyletic groups represented on the cladogram; and comparing the structure of multiple cladograms from the same taxa. Even randomly-generated data will lead to the construction of a tree, and it is thus important to assess the magnitude of that part of the character variation that is potentially informative about the history - is the support for a tree any better than would be expected from a random dataset?

Several classes of techniques exist: optimality measures; skewness indices; analytical statistical tests; and randomization (or permutation) statistical tests. A cladogram is interpreted in terms of the monophyletic groups of taxa that it hypothesizes, and it is thus important to assess the degree of support for each branch in the tree - is the support for a particular group any better than would be expected from a random dataset?

Many classes of techniques exist: analytical procedures, such as confidence limits, branch-length variances, and likelihood-ratio tests; resampling procedures, such as the bootstrap, the jackknife, and topology-dependent permutation; and non-statistical procedures, such as the decay index, clade stability, and spectral signals. It is unlikely that different tree-building methods will always produce identical trees for the same dataset, and multiple optimal trees can be produced by a single tree-building method. Further, we may wish to compare cladograms produced using different sets of characters. It is thus necessary to quantitatively compare the structure of multiple trees from the same set of taxa - which clades are supported by which set of trees? Several classes of techniques exist: consensus trees; largest common pruned trees; consensus indices; tree comparison metrics; analytical statistical tests; and randomization tests.

Orlovich, D.A.¹, A.N. Drinnan & P. Y. Ladiges. School of Botany, The University of Melbourne, Parkville, Victoria 3052. ¹Present address: Department of Microbiology, The University of Queensland, Brisbane 4072.

Patterns of flower development in the Myrtaceae

We have investigated floral development in the informal 'Metrosideros group' (Myrtaceae) and compared this with floral development in Melaleuca and Callistemon. Flowers form on an invaginated floral apex and exhibit non-acropetal initiation of floral parts, ie: the stamen primordia are initiated after the invagination of the gynoecium has begun. Two general patterns of androecial development are identified. The multistaminate fascicles of Melaleuca spp., Lophostemon spp., Welchiodendron longivalve and Tristaniopsis laurina develop from stamen primordia initiated on antepetalous pre-staminal bulges. In Xanthostemon spp., Lysicarpus angustifolius and Callistemon spp., which generally have a complete ring of unfused stamens, often with more stamens in the antepetalous region, pre-staminal bulges do not develop prior to stamen initiation; stamen primordia are initiated directly on the floral tube. In some Metrosideros group taxa, differential hypanthial expansion results in the spreading-out of antesepalous stamens whilst antepetalous stamens remain clustered in front of the petals. Callistemon viminalis and C. glaucus are like Melaleuca leucodendra and M. cajuputi in that neither have a distinct pre-staminal bulge. The lack of a PSB is homoplasy in this case. Staminal arrangement in all the taxa studied is a result of the interaction between the timing of androecial initiation and the available space on the hypanthium.

Patterson, David J. & Alastair G. B. Simpson, School of Biological Sciences, University of Sydney, NSW 2006

On the endemism of free-living flagellates in Australia

Our perceptions of endemism is in part due to the actual distribution of organisms as determined by biological and other factors, but is also influenced by factors extrinsic to the distribution of the organisms. In the case of flagellated protozoa, extrinsic factors include inaccurate identifications, inconsistent taxonomic practices, non-discriminatory (lumping) taxonomy, incomplete records, erratic sampling methods, and the subjective analysis of data. As a result, our current understanding of the biogeography of free-living flagellates is largely anecdotal and unreliable. We have sought to reduce the influence of extrinsic factors in a series of studies of free-living heterotrophic flagellates from marine and freshwater sites world-wide. We have reduced variable taxonomic practices by applying consistent and articulated criteria and by placing a heavy reliance on uninterpreted records. We have adopted standard sampling protocols, used replicates, and have subjected data to objective analysis by clustering algorithms. We have been unable to unamibiguously identify endemism in complete communities, nor do we have convincing evidence that a holophyletic taxon more than one species of flagellate is restricted to a single geographical location. Intrinsic causes of cosmopolitanism may be the long evolutionary history of heterotrophic flagellates or high rates of dispersal arising from large populations, contiguous nature of habitats, and physiological tolerances. An extrinsic factor leading to the same conclusion is an inadequate species concept. Our observations of the occurrence of the same species in ephemeral, exotic and non-contiguous habitats raises unresolved questions as to the mechanisms of dispersal.

Ponder W.F. Australian Museum, 6 College Street, Sydney, NSW 2000.

Small-scale speciation and conservation strategies.

Groups of organisms with very poor dispersal capabilities sometimes show very small-scale patterns of speciation. There are many examples in terrestrial and freshwater molluscs, including several in Australia. These include spectacular radiations of landsnails in arid limestone hills in north western Australia and freshwater snails in arid-zone artesian springs and in streams in parts of south eastern Australia and Tasmania. Some species have ranges of less than a few hundred metres, others a few kilometres. Some are confined to a single gorge or spring. Possible conservation strategies to deal with such cases are suggested.

Work on the systematics and genetics of some of the freshwater snails is outlined and other examples are given that would provide excellent subjects for evolutionary studies in the future.

Rawlings L.H.^1,2 & S.C. Donnellan. ¹South Australian Museum, North Terrace Adelaide, South Australia 5000. ²Department of Genetics, University of Adelaide, North Terrace, Adelaide, South Australia 5000.

Snake tales: comparative phylogeography of snakes of Papua New Guinea.

A powerful approach to intra-continental biogeography is comparative phylogeographic analysis of co-distributed taxa. We sampled mitochondrial gene trees and allozyme phylogenies from four widespread PNG snake species from two families, Boidae and Pythonidae. Each species has an island-wide, lowland and mid-montane distribution. MtDNA and allozyme trees were totally concordant in all four species and all showed a distinct genetic discontinuity across the central mountain spine. This wide-spread genetic subdivision suggests that there may be a historical global factor (eg. tectonic movements) that caused a barrier to gene flow along the east-west axis of eastern New Guinea.

Rouse, Greg. W. School of Biological Sciences, University of Sydney, N.S.W. 2006 Australia.

Trochophore no more? Downstream feeding, ciliary bands, and evolution of

larvae in annelids and other phyla.

The idea that a free-swimming 'trochophore' with an opposed-band feeding system is a primitive larval form for the 'Eutrochozoa' (Mollusca, Sipuncula, Annelida, and Echiura etc.) or 'Lophotrochozoa' has a long tradition. It has also been proposed, and is generally accepted, that in the evolution of marine animals there is a biased transition from feeding larvae to non-feeding larvae.

The homology hypotheses of the trochophore, downstream larval feeding and various larval features are combined with a matrix of over 100 other morphological characters developed for the Polychaeta and other 'eutrochozoa'. The results suggest the trochophore is a rare and homoplastic form; that lecithotrophic larvae are plesiomorphic for the Polychaeta; that various ciliary bands have evolved several times and that downstream feeding has also evolved several times. There is no sign of any biased transition to non-feeding larvae.

Many polychaete groups have a variety of reproductive modes. Hence, clade by clade analyses are necessary to assess the plesiomorphic polychaete reproductive mode. An example is shown, using the Sabellidae, where external fertilization is secondary and brooding larvae is primitive. Finally, it is discussed that life history data are not independent and statistical analysis of covariability must take into account phylogeny.

Rowett A.I. Department of Mines and Energy Resources, South Australia. PO Box 151 Eastwood, South Australia. 5063.

The Tertiary Silcrete floras of Stuart Creek.

To date, 142 fossil leaf types, 50 fruit and seed types and two wood types have been identified from four fossil leaf sites at Stuart Creek, central South Australia. The floristic composition of all sites is very similar, interpreted as a marginal monsoonal forest with well defined tropical and sclerophyllous components. Lobed-leaf forms comparable to the modern genera Brachychiton (Sterculiaceae), Cochlospermum (Cochlospermaceae), Orites (Proteaceae) and a serrate-margined form comparable to the Cunoniaceae dominated the tropical component. These plants probably grew along the watercourses where permanent water was available. The sclerophyllous component dominates the flora, mainly linear to lanceolate leaf forms, some forms appear indistinguishable from modern Eucalyptus leaves. These plants grew on the exposed, drier plains in more open forest communities and probably on edaphically drier parts of the flood plains. Sedimentological investigations indicate that deposition occurred in a broad, shallow, meandering to braided channel system that flowed to a depocentre in the current Lake Eyre area. Stratigraphic investigations undertaken, has revised the stratigraphic relationship between the fossil-bearing Willalinchina and Watchie Sandstones with the former unit now found to younger than the latter. Although younger than the Watchie Sandstone, the precise age of the Willalinchina Sandstone is still not known. The Willalinchina Sandstone is here correlated with a leaf-bearing, late channel phase of the (?) Late Oligocene to Pliocene Etadunna Formation in the Poole Creek Palaeochannel along the southern margin of Lake Eyre South although further work is required to test this correlation. The Willalinchina Sandstone could therefore be Miocene - Pliocene in age.

Schimanski, L.J. & R.S. Hill. Department of Plant Science, University of Tasmania, GPO Box 252-55 Hobart, Tasmania. 7001.

Reconstructing past distributions of Grevilleoideae (Proteaceae) based on leaf morphology.

The family Proteaceae has an extensive extant distribution across Australia and a well defined macrofossil record. This family is often associated with scleromorphic characteristics such as thick leaves, revolute margins and dense trichomes. These gross morphological leaf characteristics as well as several others were examined in the subfamily Grevilleoideae (Proteaceae). The biogeographic regions considered ranged from the tropical rainforests of Far North Queensland to the arid zones of Western Australia and the subalpine areas of Tasmania. Many of these characteristics varied markedly across the major environmental gradients of rainfall and soil type. The gross morphology of Proteaceous macrofossils from Tasmanian deposits were also examined in this study. These characteristics indicate the possible environments in which these taxa grew, and can be directly related to their past distribution.

Schodde R. & I.J. Mason. Australian National Wildlife Collection, CSIRO, Canberra, A.C.T. 2602

The Geography of Speciation in the bird fauna of Bassian Australia

The temperate sclerophyllous vegetation of Australia, from cool wet eucalypt forest to heath and mallee, supports an assemblage of bird species that occurs right around the southern rim of the continent wherever this vegetation is found. The assemblage totals almost 120 species, over a quarter of the Australian land bird fauna, and includes such familiar birds as the yellow- and white- tailed black-cockatoos, blue-cheeked rosellas, Powerful Owl, most thornbills, and the firetail finches and currawongs. This assemblage is the Bassian bird fauna.

Since Spencer in 1896, the Bassian biome has been thought to extend from southwest Australia to southeast Qld (the Macpherson-Macleay overlap of Burbidge 1960). Although some bird species range throughout the biome, most are more localised, many with coincident limits. By 1990, five speciation-sponsoring "faunal barriers" had been identified from these coincidences, including the Nullarbor Plain and Bass Strait.

Morphological analysis of extensive new collections gathered by the Australian National Wildlife Collection over the last two decades reveals that the geography of speciation in the Bassian avifauna is far more complex. At least ten "faunal barriers" of one kind or another have operated at one stage or another to promote allopatric differentiation to different levels, driven by Quaternary climatic pulses; and a number of unsuspected refugia have emerged, such as Kangaroo Island, the Otway-Strzelecki ranges, and the Atherton-Windsor uplands in Qld. Implications for rates of evolution, as well as for biodiversity conservation, are canvassed.

Shiel¹, Russell J. & Stuart A. Halse^{2 1}Murray Darling Freshwater Research Centre, PO Box 921, Albury, NSW 2640; ²Dept of Conservation and Land Management, PO Box 51, Wanneroo, WA 6065

Regionalism in the aquatic microfauna of southern Australia

Scanning electron microscopy, light microscopy and image analysis were used to determine the taxonomic composition of a suite of protists, rotifers and microcrustacea from a series of wetland surveys. Areas surveyed include billabongs and ephemeral floodplain waters of the Murray-Darling Basin, wetlands and lake littorals of SW Western Australia, dune lakes and wetlands of King and Flinders Island (Bass Strait), and lake littorals and wetlands of Tasmania. Summary data are presented from >200 sites sampled and >500 taxa identified to date. Systematic resolution of the different groups is patchy, however for the better known (e.g. rotifers and microcrustacea) there is clearly a higher degree of endemism than was previously considered. Even in the purportedly cosmopolitan protists there is evidence of regionalism in the Australian microfauna. The dichotomy between eastern and western assemblages appears to have been amplified by adaptive radiation in the aquatic microfauna of southwestern W.A.

Skevington, J. & D.K. Yeates. University of Queensland, Brisbane, St.

Lucia, Queensland 4072.

Relationships of the Syrphoidea (Diptera) and a preliminary look at the

systematics of the Australian Neodorylini (Diptera, Pipunculidae)

Some authors suggest that the Syrphidae (Diptera) may be paraphyletic with respect to the Pipunculidae (Diptera). We are assembling a molecular and morphological data set to test this idea. Pivotal to this analysis will be the addition of the subfamilies Nephrocerinae (Pipunculidae) and Microdontinae (Syrphidae).

The Neodorylini is the largest tribe of pipunculids in the world with 471 described species worldwide (36 percent of the world pipunculid fauna). This lineage is the last taxonomic wilderness in a family that has had a great deal of recent attention. Although there have been recent efforts to divide the Neodorylini into more manageable groups, generic concepts within the clade are questionable and there is no region of the world where species can be confidently identified. Australia hosts several distinctive species groups within this tribe and taxa here should help to provide answers to some of the questions of higher phylogeny. A mere 16 species of Neodorylini have been described from Australia, but this greatly underestimates the diversity and significance of this fascinating group of flies and preliminary work suggests that over 100 species exist here. The Pipunculidae are under-represented in museum collections, and although these collections form the basis for taxonomic work such as this, collecting is necessary to augment this material. Fortunately, Australian collectors have discovered that pipunculids congregate on hilltops and are thus easy to collect. We have found as many as 50 specimens of 9 species and 6 genera on a hilltop in one morning of collecting.

Snow, N. Queensland Herbarium, Meiers Road, Indooroopilly, Queensland, 4068.

Application of the phylogenetic species concept: a case study with Leptochloa P. Beauv. sensu lato (Poaceae, Chloridoideae).

Theoretical advancements in the phylogenetic paradigm of comparative biology are useful only to the extent to which they can be applied. During preparation of a recent monograph of the grass genus Leptochloa I developed a working methodology by which phylogenetic species can be recognized by the botanical monographer who must rely primarily on herbarium specimens. Necessary for the recognition of species from this perspective is the fixation (100% occurrence) of a character or combination of characters. The approach borrows from a methodology developed by J. Davis (Cornell University) called Population Aggregation Analysis, but adds an element of the late A. Cronquist regarding the need to recognize species by ordinary means. Thus, a species is defined as an aggregation of sexual or asexual semaphoronts consistently diagnosable by a fixed character or combination of characters recognizable by ordinary means. The requirement of character fixation for the designation of species from populations and/or herbarium specimens will often overlook significant levels of non-fixed phenetic variation. An allowance for the recognition of infraspecific taxa, which are not expected to be hierarchically related, is therefore deemed to remain desirable. However, the geographic occurrence and presumed "historical fate" of populations are seen as irrelevant for the recognition of infraspecific taxa. Application of the above definition of phylogenetic species resulted in the reduction in rank of several species of Leptochloa to subspecies.

Snow, N. & G. P. Guymer. Queensland Herbarium, Meiers Road, Indooroopilly, Queensland, 4068.

Conservations and systematics of Lithomyrtus F. Muell. and Myrtella F. Muell. s.s. (Myrtaceae).

Although undescribed species of Myrtella sensu lato have been known to occur in the sandstone escarpments of the Northern Territory for some time, the genus never has been studied comprehensively. Our recent fieldwork in the region with colleagues from the Northern Territory Herbarium has confirmed the existence of seven new species, most of which are restricted to the sandstone ranges and escarpments of Kakadu National Park and Arnhem Land. Four of the new species occur in the vicinity of the Jabiluka mine site, making their conservations status a matter needing further attention. Myrtella sensu stricto is comprised of two species mostly confined to New Guinea, which differ from the Australian species formerly assigned to the genus in a number of characters. The Australian species formerly in Myrtella appear to require the generic name Lithomyrtus F. Muell. Characters that diagnose the new species, which will be illustrated with photos taken in the field, include growth habit, presence/absence of abaxial oil glands on the leaf, presence/absence of petioles, leaf shape, leaf pubescence, leaf size, sepal shape, sepal length, bracteole length, fruit shape and fruit pubescence. Species of Lithomyrtus have considerable potential as ornamental shrubs.

Stork, Nigel E. Director, Cooperative Research Centre for Tropical Rainforest Ecology and Management, James Cook University, Cairns Campus, P.O. Box 6811, Cairns 4870, Queensland.

The role of systematics in ecology and conservation

Systematics has many crucial roles to play in the ecological and economic sustainability of Australia's environment. Systematists need to be more aware that we have customers and that we have to deliver goods and outcomes. Traditionally taxonomists produce taxonomic products for other taxonomists and usually only a small group at that. If resources for systematics are to be maintained or increased then we need to assess who we are serving and the value of the work that we do. In many instances the value of some systematics research has been grossly undervalued or neglected. In other instances there are very few or no identified benefits. We need to look at our users needs, decide what we want to achieve, make priorities of which things to do and what products to produce and how quickly.

Systematics has the potential to provide answers to some critical questions in biology and the environment. We live and work in a multispecies environment and so the studies that other biologists carry out on single species or simple systems need to be translated in more species-rich environments. For example, the role of termites in the global carbon cycle is little understood and yet so far measurements of methane production is based on a few species in simple environments. As a biologist I find it embarrassing that we cannot say to within a factor of ten how many species there are and where they are found. I find it even more embarrassing that we cannot identify more than 1500 species of plants and animals that are known to have become extinct in the last few hundred years and yet some leading biologists are talking about unprecedented rates of global extinctions. How can we expect to be taken seriously? Systematists have an important part to play in providing the answers to these problems.

In the last fifty years we have seen a number of distinct fields of biological research and application, such as systematics, ecology and conservation, develop from what was the classical study of natural history. As these fields have matured there has been a tendency for them to grow apart and yet advances in important areas of human concern involve all three areas. In this talk I will look at how the gaps between these fields are being filled and how they are leading to new insights into the distribution and plight of biodiversity. Today we are concerned with human impacts on the environment, particularly in Australia where introduced species have created so much change. Systematics can and does play a critical role in the environment. If the resources for the study of systematics are to be increased then we need to demonstrate their value to society.

North Queensland is home to the largest area of tropical rainforest in Australia. Although some of these forests have World Heritage status their long-term conservation and sustainable use depends on outcome driven research that will provide solutions to management problems. In this talk I will use examples from the research being carried out by researchers who collaborate in the Cooperative Research Centre for Tropical Rainforest Ecology and Management. In particular I will talk about the systematics work and its significance for management of the Wet Tropics World Heritage area.

Teo R. Department of Zoology, James Cook University of North Queensland, Townsville, Queensland 4811.

Systematics of the Criocerinae

The Criocerinae are leaf beetles, with about 1400 described species in only 20 genus rank taxa. They primarily feed on monocotyledonous plants and some are commercially important pests of orchids, Solanaceae and grasses. A few species are herbivores of cyads. The larvae are external feeders. They lack defensive spines or glands, but make up for this by coating themselves in poo. The position of the Criocerinae in the phylogeny of Chrysomelidae is unresolved. Sister-groups proposed include Donaciinae, Hispinae and (Chrysomelinae + Galerucinae). The phylogeny of the criocerine genera is also unresolved and is unlikely to be resolved without recognition of the most likely sister-group. New morphological evidence for the relationship of Criocerinae to other subfamilies and the internal arrangement of genera is presented here.

Thompson I.R., Ladiges P.Y.& Ross, J.H. School of Botany, The University of Melbourne, Parkville, Victoria, 3052

Systematics of Hovea (Fabaceae)

Hovea is a genus of up to 50 species with a widespread distribution in Australia. Its taxonomy in eastern Australia has historically been unsatisfactory. Placement of the genus into the correct tribe is also problematic; it has recently been transferred along with three other Australian genera (Templetonia, Lamprolobium, Plagiocarpus) to the tribe Brongniartieae from the tribe Bossiaeeae based on morphological data. Tribe Brongniartieae previously contained just two tropical American genera.

Approximately 20 new taxa for eastern Australia were identified by conventional morphological taxonomy and morphometric analysis. These taxa formed the basis of a phylogenetic study within Hovea using DNA sequence data. The 5S ribosomal DNA intergenic spacer region was compared for 40 eastern taxa. Only partial resolution was achieved within this group due to a high similarity in the nucleotide sequence for this region. A further 30 taxa from Hovea and other genera in the Brongniartieae and Bossiaeeae were sequenced. Wide variation in the length of the 5S rDNA spacer region was found and alignment proved difficult. The nucleotide sequence of a more conserved spacer region, ITS1 of the rDNA cistron, was then examined for 35 taxa. Several robust monophyletic groups are indicated but a lack of information disallows complete resolution of the tribe. The monophyly of Hovea is ambiguous.

Trueman J. W. H. Research School of Biological Sciences, Australian National University, Canberra, ACT 0200

Molecules: Clocks or Parasites?

Molecular systematics covers a range of techniques but in recent times has come to mean phylogenetic inference from DNA/RNA molecules. Many labs are involved, vast(??) databases are being developed, and the new discipline of bioinformatics is attempting to link what the gel-jockeys produce to what the phylogeneticists thought they would get for their money. Meanwhile, funding for traditional systematics is in decline.

Current views on the usefulness of molecular systematics range from "molecules change in a clock-like way and tell the branching pattern plus the time" to "molecules are parasites and a gene tree may say nothing about host-cell relationships". The clock analogy leads to substitution-rate-model-based methods for phylogenetic inference. The parasite analogy suggests biogeographic methods and a one-gene-one-noisy-character approach. Neither extreme view may be realistic and both these sets of tools are incomplete. As well, the great divide between distance, ML and parsimony tree-inference methods as used in molecular systematics would benefit from a close second look.

At this time, phylogeneticists would be wise not to embrace molecular systematics as The Fount of All Wisdom. Neither can they ignore it completely, because it won't go away.

Vadala, A.J. & Drinnan, A.N. School of Botany, The University of Melbourne, Parkville, Australia 3052.

Late Paleocene Banksieaephyllum praefastigatum :Proteaceae, or a possible enigma in the past flora of Australia?

Late Paleocene sediments from Cambalong Creek in the Southern Highlands of New South Wales contain a flora of mummified leaves. One of these leaf types has been identified as a new taxon, Banksiaephyllum praefastigatum. Its cuticular micromorphology and leaf architecture are very similar to some previously-described taxa of the genus Banksieaephyllum Cookson & Duigan, which was erected for fossil leaves with architectural characteristics of the subtribe Banksiinae. This newly-discovered fossil is equal in age to the oldest previously-described species of Banksieaephyllum, a genus that had a spatial range extending from south-western Western Australia to north-eastern Tasmania, and a temporal range of at least 35 million years, from the late Paleocene to the Oligocene/Miocene. An alternative interpretation of this newly discovered fossil, is that it (and some related species of Banksieaephyllum) are not Proteaceae. This is based on a different interpretation of important micromorphological characters, namely, the stomatal structure and trichome base structure of the fossils. Such an alternative interpretation would imply that these taxa constitute a group that was once diverse, widely distributed and prominent in the fossil record of south-eastern Australia, and that became extinct at some time after the Oligocene/Miocene.

Vijverberg K. and K. Bachmann. University of Amsterdam, Amsterdam NL-1098SM318, The Netherlands.

Adaptive radiation of Microseris (Asteraceae) in Australia and New Zealand.

The allotetraploid species (complex) Microseris lanceolata (Walp.) Sch.-Bip./M. scapigera (Forst.) Sch.-Bip. (Asteraceae, Cichorieae; native dandelion or yam daisy) is a variable taxon of perennial plants distributed over south (eastern) Australia, Tasmania and New Zealand. According to morphological, cytological and molecular data, it arose in western North America by hybridisation of an annual and a perennial diploid species followed by polyploidisation and long-distance dispersal. Adaptive radiation resulted in various morphological types including tuberous lowland, alpine and self-fertile forms on the Australian mainland, and self-compatible and -incompatible low- and highland forms on Tasmania and New Zealand. The chloroplast and nuclear DNA of 52 Australian and New Zealand populations has been investigated to reconstruct the evolution of the taxon.

A phylogenetic tree based on 23 HinfI RFLPs (Restriction Fragment Length Polymorphisms) and three trnL(UAA)-trnF(GAA) intergenic spacer length variants in the chloroplast DNA shows a putative basal group consisting of all New Zealand, Tasmanian and self-fertile Australian populations and a derived group containing the Australian tuberous and alpine forms. The groups correspond to the previously recognized taxa M. scapigera and M. lanceolata respectively. Within the latter, three subgroups reflect more their geographic distribution than morphological entities.

An unrooted network based on AFLPs (Amplified Fragment Length Polymorphisms) in the nuclear genome shows the Australian self-fertile biotype, the New Zealand populations and one of the Australian subgroups found in the chloroplast analyses as well defined groups. The Australian alpine populations show a weak clustering and are closely related to the Australian tuberous form. Tasmanian populations are dispersed among the Australian tuberous and alpine populations.

The group is a test case for methods of phylogenetic reconstruction. The singular origin of the taxon should eventually produce a tree rooted at the site of arrival from North America and a distinction between morphological specialisations arising once or repeatedly under similar conditions.

Waycott, Michelle. Tropical Environment Studies and Geography, James Cook University, Townsville, 4811 and Smithsonian Tropical Research Institute, PO Box 2072, Balboa, Republic of Panama.

Marine angiosperm reproductive characters and evolution: extremities of convergence.

The marine angiosperms (seagrasses) have been characterised as having evolved certain reproductive characters (dioecy) and a breeding system (outcrossing) to maximise survival in evolutionary terms in their unusual and restrictive habitat, the ocean. However, phylogenetic relationships recently determined independently of these reproductive and genetic characters based on DNA sequences is shedding light on the validity of this model. Seagrasses have evolved in at least 3 separate lineages within the subclass Alismatidae and a growing body of data on the levels of genetic variability within species shows no correlation with sexual status (dioecious/monoecious/hermaphrodite). Previous hypotheses about evolution in the seagrasses have been based on morphological and anatomical characters such as pollen form, sexual status and leaf blade form, all of which are found to be convergences when viewed in the context of the molecular phylogeny.

Other features of seagrass evolution may have a more important impact on the observed high levels of dioecy and low levels of speciation than the selection for outcrossing. Low levels of speciation may be the result of extreme clonal growth and genet longevity resulting in a slowing of the evolutionary rate. In some groups, dioecy may be the result of the past evolutionary history in the family such as is seen in the Hydrocharitaceae, a mixed marine/freshwater family. Factors such as these need further investigation in understanding the evolutionary processes occurring in this unusual group of angiosperms and may help to better understand evolutionary processes within the angiosperms.

Winterton Shaun L. and David K. Yeates. Department of Entomology, University of Queensland, St. Lucia, Brisbane, Queensland, 4067.

Insights into a diverse but poorly known continental fauna: Systematic studies on Australian Therevidae

The Therevidae (Insecta: Diptera) are a cosmopolitan group of Orthorrhaphous flies found in a variety of habitats, with greatest diversity in semi-arid regions and woodland. Adults are nectarfeeders of blossoms while the larvae are voracious, fossorial predators of other soil arthropods in sandy, friable soils. The highly endemic Australian therevid fauna has been poorly studied, with numerous species yet to be described. There is no established subfamilial and tribal classification for the Therevidae, but it is apparent that the Australian fauna represents a separate subfamily with relatives in southern Africa, not in South America.

We are investigating the taxonomy and systematics of the Australian Therevidae as part of a collaboration with United States researchers studying the world Therevidae. Initial studies are concentrating on several large genera (e.g Agapophytus Guérin-Méneville) which appear to be centers for species radiation, firstly describing the numerous new taxa, and secondly investigating the systematic relationships of the groups using morphological and molecular data. One endemic genus with 20 species has been described and Agapophytus a large endemic genus (ca. 80-100 spp.; 13 described spp.) of relatively large therevids is presently being revised. Males of several species are known to sit on trunks of smooth bark Eucalyptus trees waiting for females.