Phylogenetic incongruence and homoplasy in the appendages and bodies of arthropods: why broad character sampling is best

Notwithstanding the rapidly increasing sampling density of molecular sequence data, morphological characters still make an important contribution to our understanding of the evolutionary relationships of arthropod groups. In many clades, characters relating to the number and morphological specialization of appendages are ascribed particular phylogenetic significance and may be preferentially sampled. However, previous studies have shown that partitions of morphological character matrices often imply significantly different phylogenies. Here, we ask whether a similar incongruence is observed in the appendage and non-appendage characters of arthropods. We apply tree length (incongruence length difference, ILD) and tree distance (incongruence relationship difference, IRD) tests to these partitions in an empirical sample of 53 published neontological datasets for arthropods. We find significant incongruence about one time in five: more often than expected, but markedly less often than in previous partition studies. We also find similar levels of homoplasy in limb and non-limb characters, both in terms of internal consistency and consistency relative to molecular trees. Taken together, these findings imply that sampled limb and non-limb characters are of similar phylogenetic utility and quality, and that a total evidence approach to their analysis is preferable.

Phylogeny of Halgerda (Mollusca : Gastropoda) based on combined analysis of mitochondrial CO1 and morphology

A phylogeny of the nudibranch Halgerda Bergh, 1880 is presented based on mitochondrial CO1 and compared with a previously published morphology-based phylogeny. Incongruence length difference (ILD) tests indicated possibly conflicting phylogenetic signals. Both separate and combined analyses resulted in trees that were very similar in topology, both to each other and to the morphology-based tree (albeit with lower consistency and retention indices). Some differences are noted in the topology based on the optimality criteria used: parsimony and maximum likelihood. Results of this study indicate that the CO1 gene contains sufficient phylogenetic signal for inferring species-level phylogeny in the Nudibranchia and also supports many of the clades in the morphological phylogeny. Specifically, the more basal clades of the morphology-based phylogeny are supported with differences noted at the most terminal taxa.