Phylogeny of Habronattus jumping spiders (Araneae: Salticidae), with consideration of genital and courtship evolution

Abstract

Abstract. DNA sequences from the mitochondrial (including ND1, 16S) and nuclear (EF‐1α) genomes of about ninety‐four species were obtained to reconstruct phylogenetic relationships of Habronattus jumping spiders. Maximum parsimony trees were sought with both separate (mitochondrial, nuclear) and combined analyses; maximum likelihood trees were sought with both separate (ND1, 16S, EF‐1α introns, EF‐1α exons) and combined (mitochondrial, nuclear) analyses. All analyses agreed on some fundamental aspects of the tree, including the monophyly of the previously recognized agilis, amicus, dorotheae and americanus species groups. The deep phylogenetic structure is well resolved, placing the agilis, amicus, tranquillus and dorotheae groups basally. Several other previously unrecognized clades were well supported, including a newly formulated decorus group. The large group of species with modified male first and third legs was supported as monophyletic except for the surprising placement elsewhere of three species of the group. The phenotypic similarities between these three and the others are so detailed and precise that convergence in ornamentation can probably be ruled out. There are hints of phylogenetically distant genetic introgression involving the coecatus group. The combination Habronattus paratus is restored based on the species falling within Habronattus. Regarding patterns of character evolution, there was consistent support for the basal placement of several species groups with a long embolus, suggesting that there were more evolutionary reductions in embolus length than postulated in a previous morphological phylogeny. This is in accord with the expectation that there is a bias to an overly conservative interpretation of a character's evolution if it is interpreted on a phylogeny based in part on that same character. In contrast, the molecular phylogeny did not suggest any instances of the evolutionary transformation of one complex style of courtship into another, a possibility that could have been difficult to detect using the morphological phylogeny because of the same bias to conservativism.