Indexed on: 26 Nov '10Published on: 26 Nov '10Published in: Molecular Phylogenetics and Evolution
Previous efforts at understanding the evolution of the genus Phyllodesmium, based on morphological analyses, have been plagued by poorly supported phylogenies (Ortiz and Gosliner, 2008; Moore and Gosliner, 2009, in press). It has been suggested (Moore and Gosliner, 2009) that a molecular phylogeny might provide more insight into this history than can be easily discovered using morphological data. In this study, 658bp of the cytochrome c oxidase subunit I gene (CO1), 441bp of the mitochondrial large ribosomal subunit (16S) gene, and 328bp of a protein-coding nuclear gene (histone 3) were sequenced for 18 species of Phyllodesmium and six outgroup species. A total of 464 parsimony informative sites were used for parsimony, maximum likelihood, and Bayesian inference of phylogeny analyses. All three analyses produced similar topologies, with the exception of a single difference within the parsimony analysis. Bootstrap values and posterior probabilities provided strong support at many shallow nodes, and the monophyly of Phyllodesmium was supported in every case. Three distinct clades of Phyllodesmium are evident in this analysis. One of these represents the majority of asymbiotic taxa. Phyllodesmium poindimiei, an asymbiotic species, is clearly a member of a symbiotic clade and appears to have secondarily lost its symbiotic relationship with zooxanthellae. There was moderate support confirming similar topological trends seen in earlier morphological phylogenies, including the hypothesis that symbiotic species associating with zooxanthellae have evolved more recently than non-symbiotic species. Despite the inclusion of a presumably conservative nuclear locus, some deep nodes are still unresolved or are not well supported. Future inclusion of additional taxa and more slowly evolving loci will likely improve resolution of these deeper nodes. The subsequent phylogeny supports previous hypotheses by Rudman (1991), Kempf (1991) and Burghardt et al. (2008b) that evolution of more complex digestive gland structures is related to increased complexity of symbiosis with zooxanthellae and greater efficiency of photosynthetic activity. Our phylogeny also demonstrates that this symbiosis has evolved only once in Phyllodesmium and that azooxanthellate species are sister, rather than basal, to zooxanthellate species.