The high level of biodiversity in tropical rainforest arthropods has been explained with hypotheses of both ecological and historical processes. For the Wet Tropics, previous studies of endemic vertebrate species have suggested that isolation in rainforest refugia has resulted in genetically distinct lineages, although generally these lineages have not become reproductively isolated. In addition, in the Wet Tropics, morphological divergence with gene flow has been recorded across the rainforest-open forest ecotone. Both incidental divergence in allopatry or ecologically-mediated disruptive selection (in allopatry or parapatry) are considered plausible speciation mechanisms for arthropods in the Wet Tropics.
To identify sister-taxa and predict die tempo and mode of speciation, a mitochondrial
DNA phylogeny for the Australian representatives of the dung beetle genus Temnoplectron was produced. This phylogeny shares a number of well supported clades with a previously published phylogeny based on morphological data, and identifies many allopatric sister species, with the location of species' splits coinciding with phylogeographic breaks previously found in vertebrate species. Estimation of sequence divergences between species shows that speciation events are older than the Pleistocene.
The connection between historical phylogeography, gene flow and speciation was further investigated by sequencing mitochondrial DNA for two pairs of sister species from within the genus - T. aeneopiceum + T. subvolitans and T. politulum + T. reyi. There was much stronger phylogeographic structure in T. aeneopiceum and T. subvolitans than in T. politulum and T. reyi. Coalescent analysis indicated historical fragmentation and very low rates of long-term dispersal among regions for both T. aeneopiceum and T. subvolitans, except for recent range expansion (and population expansion) from an isolated population at the southern extreme of the range northwards, while for T. politulum and T. reyi, estimated rates of genetic exchange were typically higher.
To determine the effects of environment on historical demography and speciation, "ecological niche" and the range of suitable habitat were inferred for the same two pairs of sister species, and the combined range of the clade of flightless, montane, rainforest species - T.finnigani, T. monteithi, T. lewisense and T. involucre - using BIOCLIM. Severe habitat contraction during the Last Glacial Maximum was inferred for all species, followed by an increase in suitable habitat during a cooler, wetter period, some decrease during a warmer, wetter period, then expansion to the current habitat range. A tendency for each species' modelled habitat range to include that of its sister species suggests little divergence in overall habitat or physiology during the speciation process.
The two sister-species pairs were examined for morphometries to determine the effects of gene flow and environment on morphology. Three of four species vary significantly in morphometries across subregions. However, these divergences are also consistent with body size clines along an environmental gradient, with larger individuals occurring at cooler, drier locations, although this trend is not statistically significant. This divergence may be due to either genetic drift in isolated populations, or due to environmental effects and either phenotypic plasticity or environmentally-driven selective divergence. Morphometric variation in T. politulum was not associated with either environment or subregion. Overall, the morphometries, along with phylogeography and bioclimatic modelling, suggest that these species probably evolved through non-ecologically based speciation processes, although there is evidence for intraspecific morphometric clines associated with environmental gradients, due to either phenotypic plasticity, or selective divergence, or a combination of both.
I conclude that incidental divergence in isolated populations is the most likely cause of speciation for the Temnoplectron species that were examined in detail. However, this does not necessarily imply that this is the primary speciation mechanism for all tropical rainforest arthropods. Other speciation processes such as environmentally-driven divergence may be expected in other arthropod groups, and in other rainforest regions.