Phylogeny of suckers (Teleostei: Cypriniformes: Catostomidae): further evidence of relationships provided by the single-copy nuclear gene IRBP2

Abstract

The order Cypriniformes and family Catostomidae, the Holarctic suckers, have received considerable phylogeneticattention in recent years. These studies have provided contrasting phylogenies and classifications to historical,morphology-based phylogenetic and prephylogenetic hypotheses of relationships of species and the naturalness ofhypothesized genera, tribes, and subfamilies. To date, nearly all molecular work on catostomids has been done using DNAsequence variation of mitochondrial genes. In this study, we add to our previous investigations to identify single-copynuclear gene markers for diploid and polyploid cypriniforms, and to expand sequences of nuclear IRBP2 gene to 1,933 bpfor 23 catostomid species. This effort expands our previous studies using only partial sequences of 849 bp. The extendedgene fragment consists of nearly the complete gene across exon1 to exon 4 and is used in two analyses to inferphylogenetic relationships of the currently, or formerly, recognized genera, tribes, and subfamilies. One analysis includes23 ingroup species for which the larger fragment of IRBP2 could be obtained; these taxa were also included in a secondanalysis of 67 samples of 52 species for the shorter fragment. As is typical of other nuclear genes examined to date forcypriniform species, variation in IRBP2 provided strong nodal support for some supra-specific groupings and speciesrelationships. The two analyses revealed slightly different relationships, yet are largely consistent with one another. Theresulting tree from variation in the shorter fragment for 52 species is somewhat inferior to the tree derived using theextended fragment in that not as many nodes were resolved, and few have strong support. Relationships from the latteranalysis are, however, consistent with inferred relationships that are more robustly supported in the smaller taxon analysisusing the larger fragment, lending credence to the use of more complete sequence data of genes in phylogenetic analyses.The current classification of the family (e.g., Nelson 2006) is not fully supported herein. The Ictiobinae is monophyletic,but some ambiguity exists as to relationship of this group relative to Cycleptinae and Myxocyprininae, as well as the needto recognize the latter two subfamilies. Catostominae is monophyletic. Catostomus is clearly not monophyletic;unnaturalness of the genus is supported herein as well as in multiple, consistently repeated and highly supported studiesresolving Deltistes, Chasmistes, and Xyrauchen within Catostomus. We herein synonymize the former three genera intothe latter genus; their recognition as distinct genera has been based on historical methods of classification based strictlyon “distinctiveness” or anagenesis of each lineage alone and not phylogenetic relationships relative to species ofCatostomus. The monophyly of Erimyzonini is strongly supported within the analysis of the longer sequence data set. Themonophyly of Thoburniini is ambiguous, but Moxostomatini, including “Scartomyzon,” is monophyletic in both analyses.The proposed recognition of Scartomyzon as a monophyletic group separate from Moxostoma is again falsified but with evidence from the nuclear gene IRBP2.