SEARCH

SEARCH BY CITATION

Keywords:

  • Andes;
  • intron;
  • mtDNA;
  • Percichthys;
  • Percilia;
  • phylogeography;
  • Pleistocene glaciations;
  • vicariance

Abstract

We used molecular evidence to examine the roles that vicariance mechanisms (mountain-building and drainage changes during the Pleistocene) have played in producing phylogeographical structure within and among South American fish species of the temperate perch family Percichthyidae. The percichthyids include two South American genera, Percichthys and Percilia, each containing several species, all of which are endemic to southern Argentina and Chile (Patagonia). Maximum-likelihood phylogenies constructed using mitochondrial DNA (mtDNA) control region haplotypes and nuclear GnRH3-2 intron allele sequences support the current taxonomy at the genus level (both Percichthys and Percilia form strongly supported, monophyletic clades) but indicate that species-level designations need revision. Phylogeographical patterns at the mtDNA support the hypothesis that the Andes have been a major barrier to gene flow. Most species diversity occurs in watersheds to the west of the Andes, together with some ancient divergences among conspecific populations. In contrast, only one species (Percichthys trucha) is found east of the Andes, and little to no phylogeographical structure occurs among populations in this region. Mismatch analyses of mtDNA sequences suggest that eastern populations last went through a major bottleneck c. 188 000 bp, a date consistent with the onset of the penultimate and largest Pleistocene glaciation in Patagonia. We suggest that eastern populations have undergone repeated founder-flush events as a consequence of glacial cycles, and that the shallow phylogeny is due to mixing during recolonization periods. The area of greater diversity west of the Andes lies outside the northern limit of the glaciers. mtDNA mismatch analysis of the genus Percilia which is restricted to this area suggests a long-established population at equilibrium. We conclude that patterns of genetic diversity in these South American genera have been primarily influenced by barriers to gene flow (Andean orogeny, and to a lesser extent, isolation in river drainages), and by glacial cycles, which have resulted in population contraction, re-arrangement of some watersheds, and the temporary breakdown of dispersal barriers among eastern river systems.