Trisha Spears, Ronald W. DeBry, Lawrence G. Abele, Katarzyna Chodyla
Proceedings of the Biological Society of Washington 118 (1), 117-157, (1 April 2005) https://doi.org/10.2988/0006-324X(2005)118[117:PMAIPI]2.0.CO;2
Peracarids are a large group of malacostracan crustaceans whose systematics and phylogeny are uncertain. The present phylogenetic study of peracarids is, to our knowledge, the first that includes full-length nuclear (n) small-subunit (SSU) ribosomal DNA (rDNA) sequences for representatives of every peracarid order. Sequence length varied substantially (1807–2746 base pairs), and two variable regions (V4 and V7) contained long expansion segments. Variable regions also exhibited significantly greater heterogeneity in nucleotide frequencies than did core regions. Maximum-parsimony, maximum-likelihood, Bayesian, and distance phylogenetic estimates indicated a monophyletic Peracarida that excluded the Mysida and included the Thermosbaenacea. This peracarid clade received strong support under Bayesian, maximum-parsimony, and distance analyses, but not maximum likelihood. Further, the thermosbaenacean lineage does not occupy a basal position relative to other peracarids, as suggested by many morphology-based phylogenies. The phylogenetic position of the Mysida within the Malacostraca remains uncertain, but a sister-group relationship between Mysida and Lophogastrida (i.e., a monophyletic Mysidacea) was consistently rejected, a result that also differs from those of most morphology-based phylogenies. High Bayesian clade-credibility support was obtained for an (Isopoda Tanaidacea Cumacea) clade and a (Lophogastrida [Spelaeogriphacea Amphipoda]) clade. Within the peracarid clade, internal branches were considerably shorter than terminal ones, so relationships among some peracarid lineages were equivocal. Data partitions corresponding to stem and loop regions in a secondary-structure model for nSSU rRNA had congruent phylogenetic signal but differed in nucleotide composition and evolutionary model. Maximum parsimony and Bayesian phylogenetic estimates based on the loop partition generally shared more nodes in trees inferred from combined data than did those based on stems, even though the stem partition had roughly twice as many characters.