American Society of Ichthyologists and Herpetologists, Norman, OK, May 2004.

The biogeography of Craterocephalus (Atherinidae) in Australia: a test of whether common distributions indicate a common history.

P.J. Unmack

One key question in biogeography is whether organisms with similar distributions have similar biogeographic histories. Models of cladistic biogeography indicate this should be the case, but the question remains to be tested for most groups. The genus Craterocephalus (Atherinidae, silversides or hardyheads) is largely restricted to freshwater environments in Australian and New Guinea. Craterocephalus currently contains 24 described species, which can be divided into four morphological groups, C. eyresii (10 spp.), C. stercusmuscarum (9 spp.), C. honoriae (4 spp., all marine), and C. stramineus (1 sp.). The C. eyresii and C. stercusmuscarum groups are both widespread in northern Australia and species from these two groups often occur sympatrically. Thus, the sympatric ranges of Australian Craterocephalus species provide an ideal test of cladistic biogeography. To examine the hypothesis that organisms with similar ranges should have similar biogeographic histories, each Australian species, including multiple populations, was sequenced for the mitochondrial cytochrome b gene. A subset of samples were also sequenced for the nuclear S7, RAG1, and RAG2 genes to provide an alternative test of relationships and to increase support for deeper nodes. Results support the existence of the four morphological groups with the following relationships among them: (eyresii (honoriae (stramineus (stercusmuscarum)))). Within the sympatric C. eyresii and C. stercusmuscarum groups there was little overlap in biogeographic patterns, indicating they share almost no common biogeographic history. The results suggest the two groups achieved their current distributions at different times, with the C. eyresii group being older than the C. stercusmuscarum group. Thus, each group has likely been influenced by different biogeographic processes. These results refute the hypothesis that common distribution indicates common biogeographic history.