The distribution of fauna in Southeast Asia stimulated many of the pioneers of biogeographic theory including Wallace, Huxley, Lydekker, Weber, Diamond and Mayr. From their observations, these early biogeographers proposed imaginary transition lines that delineated subregions according to differences in faunal distribution. Most famously, Wallace proposed a transition line that separated the Australasian and Oriental faunas, passing between Bomeo and Sulawesi and between Bali and Lombok - islands separated by only a few kilometres in some cases. Although some of the explanations these biogeographical pioneers put forward have subsequently been rejected (Wallace, for instance, thought that Peninsular Malaysia and Sumatra were once under water), the idea of historical vicariance has become a major theme in evolutionary and biogeographical theory. Remarkably, however, although the importance of historical vicariance has subsequently been tested in many other regions, modem approaches have only rarely been applied to the biogeography of the Southeast Asian region - arguably the crucible of biogeographical theory.
The overall aim of this study was to use a combination of macrogeographic and microgeographic approaches to test the relative roles of various mechanisms that have been proposed to drive diversification in natural systems. First, I used a macrogeographic approach to test the modem vicariance model of Southeast Asian biogeography. This model is based on changes in sea level due to glaciation cycles and provides a mechanistic explanation for Wallace's Line. I tested this model by quantifying patterns of overlapping in the avifauna of the mainland and islands of the Southeast Asian region. Second, I used a microgeographic approach to look at the relative roles of vicariance and selection in promoting variation within a single species - the little spiderhunter (Arachnothera longirostra). At this microgeographic level, morphological and genetic markers were used to determine the geographic variation in little spider hunters within and between the Malay Peninsula and Bomeo populations. The morphological markers include tarsus length, bill length, head and bill length, wing length, tail length, bill depth, bill width and weight. For genetic markers, the mtDNA control region was utilized.
My analysis of macro-geographical distribution patterns of avifauna provided general support for the classic, glacial model of biogeographic divergence in Southeast Asia and the proposed faunal transition lines of past zoogeographers. Most transition lines did not, however, follow the exact path of Wallace's Line, suggesting that divergence occurred across many glacial stages rather than just at full glacial retreat (as assumed by Wallace's line). The glacial model was also partly supported when I split species up according to their degree of habitat specialisation. The traditional model predicts that primary forest specialists should show the strongest patterns of differentiation because the formation of "savanna-like" forests on the land masses exposed due to the lowering of sea levels during glaciation periods would allow more generalist species to move between islands and subregions. My results supported the hypothesis that different sorts of habitat specialists would be differentially affected by glacial changes in sea levels, although the distinction between primary and secondary specialists was not as clear as predicted by the glacial model.
At a microgeographic level of variation, both morphological and genetic analysis led to a defined pattern of variation and differentiation within and between populations in the Malay Peninsula, Bomeo and the Philippines. With respect to my morphological analyses, both univariate and multivariate analysis revealed significant differences between the Malay Peninsula and Bomeo populations, and also between populations within the Malay Peninsula. Within Bomeo, pairwise comparison of three selected populations revealed significant morphological differences except between Kubah National Park and Poring Hot Spring. Both leg and bill morphology showed significant variation in little spider hunters between the Malay Peninsula and Bomeo populations while bill morphology alone differed between populations within the Malay Peninsula and leg morphology alone within the Bomeo populations. No significant differences in morphology were found with respect to habitat differences.
My molecular genetics analyses of microgeographic variation revealed evidence of substantial gene flow between the Peninsular Malaysia (Taman Negara National Park) and Bornean populations. Such a level of gene flow was indicated by the sharing of common haplotypes between these subregions and low levels of nucleotide divergence. Conversely, my results suggest low gene flow between the Thailand population and the joint Peninsular Malaysia - Bornean populations, as indicated by lack of shared common haplotypes. Overall, therefore, my genetic analysis did not support the hypothesis that neutral genetic divergence accumulates according to a distance model. Instead, my mitochondrial data suggests that the little spiderhunter populations are most likely to be derived from multiple ancient refugia. The "gallery" forests which formed along the North Sunda River and its tributaries during glaciation periods seem to be the most likely locations of these historical refugia.
At a higher phylogenetic level, the rate of mtDNA control region sequence divergence between little spiderhunter subspecies (Arachnothera longirostra longirostra, A. I. flammifera, said A. I. dilutor) is concordant with the historical land connections and separations during the glaciation period. The mtDNA ATPase6 sequence analysis revealed that subspecies flammifera has diverged from the other two subspecies by approximately the same amount as A. crassirostris and A. affinis have diverged. Again, these observations support the hypothesis of glacial refugia, in this case between the Philippines and the rest of Southeast Asian region.
In conclusion, the study has demonstrated the roles of historical processes of vicariance and natural selection in shaping the pattern of macro- and micro-geographic variation in Southeast Asia and Wallacea. As a pioneer in bird phylogeography and population genetics in this region, this study has opened up more avenues for further investigations into the complexity of the rainforest bird species.