Gastropod phylogeny based on six segments from four genes representing coding or non-coding and mitochondrial or nuclear DNA

Abstract

Significant differences remain between gastropod phylogenetic hypotheses based on morphological and molecular datasets. We collected additional data from three gene segments (28S rDNA expansion region D1 (36 taxa plus two from GenBank), cytochrome c oxidase subunit 1 (35 species plus one from Genbank) and small nuclear RNA U2 (24 species)). These were combined with data available for the same species for histone H3 and two other segments of 28S rDNA. Analyses of these data using cladistic, maximum likelihood or Bayesian methodologies were conducted in an attempt to resolve some of the differences between current hypotheses of gastropod relationships based on morphological and molecular data. The results were of particular interest in four areas. (1) Patellogastropoda in most analyses are included in a derived clade with some Vetigastropoda. In an analysis with Nautilus as the sole outgroup, transversions weighted threefold as costly as transitions and, with third codon position data ignored, Patellogastropoda are excluded from an otherwise monophyletic Gastropoda. (2) Cocculiniformia was never monophyletic in our analyses, although this possibility is not statistically rejected. (3) Nerita, the only representative of Neritopsina in this dataset, is placed anomalously in most analyses, but is, in a few cases, shown as a sister-group to the Apogastropoda, in accord with some morphological hypotheses. (4) Heterobranchia is rarely monophyletic in our analyses owing to the variable placement of the architectonicoid Philippea. This genus, even judged by the high levels of divergence within Heterobranchia, has undergone extreme rates of substitution. The Euthyneura is invariably monophyletic and nearly always included in a clade with the valvatoidean Cornirostra as its sister-group.