The importance of looking at small-scale patterns when inferring Gondwanan biogeography: a case study of the centipedeParalamyctes(Chilopoda, Lithobiomorpha, Henicopidae)

Abstract

Gondwanan biogeography has fascinated zoologists and botanists for over a century, but most biogeographical work has used continent-scale areas as analytical units. More finely resolved patterns, as can be obtained from small invertebrates with limited dispersal abilities, will be obscured in those studies. A common case is treating Australia as a single biogeographical region. In the present study, the necessity of splitting Australia into multiple microareas is demonstrated using centipedes as an example. The lithobiomorph centipede Paralamyctes is distributed on fragments of Gondwana, with species in southern Africa, Madagascar, southern India, Patagonia, eastern Australia, and New Zealand. A cladogram for Paralamyctes is based on morphology and sequences for four molecular markers for 30 terminals that sample 20 of 26 known ingroup species and four outgroups. Analysis with direct optimization across a range of indel costs and transversion : transition cost ratios identifies two main clades: Paralamyctes (Paralamyctes) unites species from southern Africa, Madagascar, tropical and warm temperate Australia, and New Zealand. The other group includes the temperate Australian/New Zealand Paralamyctes (Haasiella) and Paralamyctes (Thingathinga) and a Chilean clade. Subtree analysis finds that different parts of Australia have closest affinities to other Gondwanan fragments, and some of these relationships (such as that between north Queensland and New Zealand) are based on taxonomically stable clades. Area delimitation for large continental fragments should use sufficiently fine resolution to test the ‘monophyly’ of those fragments and attempt to eliminate spurious geographical paralogy. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 65–78.