American Society of Ichthyologists and Herpetologists, St Louis, MO, July 2007.

Phylogeography of trichomycterid catfishes of Chilean Patagonia.

Unmack, Bennin, Habit, Victoriano, and Johnson.

We investigated phylogeographic patterns in the trichomycterid catfish Trichomycterus areolatus, which is one of the most widespread fish species in central Chile. A total of 45 populations and 300 individuals across its entire range were sequenced for 1074 base pairs of cytochrome b. We also examined several outgroups: these included the monotypic genera Bullockia and Hatcheria, and two other Trichomycterus species. We investigated three principal biogeographic hypotheses: (1) are geographic breaks congruent with previously identified faunal breaks; (2) has extensive glaciation impacted genetic diversity within populations; (3) does the narrow continental shelf act as a barrier to movement during periods of low sea levels. We found considerable phylogeographic structure which included six distinct clades (composed of 120 different haplotypes) that largely separated based on geography. The largest phylogeographic break was not congruent with any faunal break, but two small breaks were, suggesting there may be some common biogeographic factors limiting biotic exchange in these areas. There appeared to be little correspondence between the degree of glaciation within various river basins and current genetic diversity, which overall is high. It seems likely that the narrow continental shelf is acting as a strong barrier to movements between basins. However, it is clear that some recent movement has occurred, possibly across drainage divides rather than via a coastal route. One unexpected finding was that two of the outgroup taxa, Bullockia and Hatcheria, were deeply nested within T. areolatus. This raises the question of whether this may be due to introgression, incorrect taxonomy, or some other factors. Both genera are morphologically quite distinct and can easily be distinguished from T. areolatus. We have also sequenced several nuclear genes (RAG1, S7, Growth Hormone) and these data support a phylogeny consistent with the topology of the cytochrome b data, suggesting the patterns found may have complicated origins, possibly involving long-term introgression.