Cladistic and quantitative shape analyses of five new syntopic Sarsamphiascus (Copepoda, Harpacticoida): problems and solutions for diosaccin systematics and taxonomy

Abstract

Five new species of the genus Sarsamphiascus Huys, 2009 were collected from a small Korean beach, which is a record for the number of syntopic copepod congeners. They differ in numerous macro-morphological characters, some of which extend generic boundaries, and are classified in two species groups. Diossacins are widely regarded as taxonomically problematic, and the new discovery presents an opportunity to test characters commonly used in their systematics through a cladistic analysis. Another aim is to test with landmark-based geometric morphometrics the extent of inter- and intraspecific variability of characters with reported significant variation but commonly used in diosaccin taxonomy. The cladistic analysis, based on 14 binary characters, reveals at least three distinct clades among five new species, which, in addition to significant size differences, probably facilitates their syntopy. Some characters previously used to distinguish species groups in Sarsamphiascus, and to distinguish this genus from other diosaccins, are exposed as symplesiomorphies, suggesting that some of these groups might not be monophyletic. Studies of alternative characters are suggested, including cuticular organs on somites, which are surveyed for the first time in this genus. Principal component analysis (PCA) of nine landmarks on the female fifth leg exopod reveals that 70% of variability can be explained by the first two eigenvectors and that in morphospace there is no overlap between Korean species, despite a significant intraspecific variability. Discriminant function analysis with cross-validation shows highly significant results, both in parametric tests and after 10,000 permutation runs, correctly classifying all specimens for all species pairs and suggesting the fifth leg exopod as an ideal taxonomic character here. Evolutionary integration of the morphometric data is tested by plotting PCA scores onto the cladistic tree; randomized permutation test suggests an absence of statistically significant phylogenetic signal.