Volume 49

Research Grant Report

Polyphyly across oceans: a molecular phylogeny of the Chromodorididae (Mollusca,Nudibranchia)

Lucy Turner

School of Biological Sciences, University of Bristol, Woodland Road, Bristol, BS8 1UG, UK. Email: lucy...@bristol.ac.uk

Having done my first degree in marine biology (Marine Biology BSc, University of Wales, Swansea, 2003) where I completed my dissertation on opisthobranch behavioural ecology and being a keen diver I had for some time been interested in opisthobranchs. Upon the start of my MRes studies at the University of Plymouth (where I was based at the Marine Biological Association) I started researching topics for my nine month research project and came upon a particularly intriguing phylogenetic problem, one that the more I researched it, the more fascinating it became; that of the phylogenetics of one of the largest and most spectacularly coloured groups of opisthobranch molluscs, the Chromodorididae.

History of the problem
Representatives of the Chromodorididae are found in all ocean basins, at all latitudes, with the highest diversity occurring in tropical regions (Figure. 1 will open in a separate window when you click). The family is known to comprise over 300 described species although it is thought that there are many more members yet to be discovered, and many more await description. The classification and phylogenetic analysis of the Chromodorididae have an extremely confused history. Early attempts at the classification of the group were based primarily on the examination of morphological characteristics and external colouration, but there has remained a poor knowledge of the comparative anatomy of the many species within the family. Rudman (1984) carried out a comprehensive revision of the family, using previously untapped reproductive characters, together with external colouration and radular morphology to produce a detailed discussion about potential evolutionary trends. Rudman proposed that the Chromodorididae comprised 16 genera which could be split into three subgroups: the Cadlina subgroup, which he considered to contain the three genera exhibiting the most primitive characters (Cadlina, Cadlinella and Tyrinna); the Chromodoris subgroup, containing six genera (Chromodoris, Ardeadoris, Glossodoris, Noumea, Pectenodoris and Verconia); and the Hypselodoris subgroup, containing seven genera (Hypselodoris, Ceratosoma, Digidentis, Durvilledoris, Mexichromis, Risbecia and Thorunna) which he considered to possess the most advanced characters, having characters derived from those found within the Chromodoris subgroup.

Before I started this project Rudman's work remained the best revision of the family to date, although it only addressed representatives from the Indo-west Pacific, despite the fact that the Chromodorididae have a worldwide distribution. Since Rudman's 1984 paper a few studies have been conducted examining some of the 17 genera of the family in phylogenetic detail (Ceratosoma, Hypselodoris and Chromodoris (Wilson & Lee, 2005)). Wilson and Lee's paper suggested that Chromodoris was paraphyletic, although broad geographic sampling was not undertaken. However, in general, the comparison of congeners from different ocean basins had not yet been addressed with appropriate outgroups, and the monophyly of broadly distributed genera remained largely untested. Also, despite recent studies suggesting new sources of taxonomically informative characters for the Chromodorididae, such as histological structures, sperm ultrastructure and egg mass shape there has been no further work on the phylogenetics of the Chromodorididae as a complete family and the phylogeny of the group has remained unresolved.

Master's work
For my master's project I decided to use molecular methods to investigate the phylogenetics of the Chromodorididae, specifically by sequencing two genes from members of this family: the mitochondrial 16S gene and the nuclear 18S gene and then use these results to: (a) reconstruct chromodorid phylogeny; (b) to establish the position of the basal genera Cadlina, Cadlinella and Tyrinna either within the Chromodorididae or outside the family; (c) to test the monophyly of the more widely distributed genera of the Chromodorididae (i.e. Chromodoris, Glossodoris and Hypselodoris).

However, a stronger phylogenetic signal was observed from the 16S gene compared to that of the 18S gene and it was concluded that the phylogenetic information obtained from the 18S gene was not enough to produce a fully resolved hypothesis for the phylogeny of the Chromodorididae. I was therefore keen to examine another gene for phylogenetic information and after further research decided that the mitochondrial COI gene was a good candidate. It was here that the money awarded by the Malacalogical Society was invaluable as it was sufficient to sequence the mitochondrial COI gene from the specimens I had already sequenced 16S and 18S for, as well as enabling me to sequence 16S and COI for further species that I previously did not have access to material for.

Results and Discussion
In total sequences were used from 61 species of chromodorids, including at least one representative from each genus. Where possible the type species were used. Some additional sequences, where they already existed were taken from GenBank. An additional 7 species (from the Actinocyclidae, Dorididae and Discodorididae) were used as outgroups to root the trees produced.
Our most resolved tree is shown in Figure 2 (click here to open this in a separate window). This tree was produced using a combination of the 16S and COI data. Some key findings included: