Discussion
Representatives of the isopod family Cymothoidae are diverse, remarkably large fish parasites that are known to infect the skin, mouth, and gills of their marine, brackish, and freshwater hosts. To date, evolutionary biologists have been intrigued by this family, and have sought to understand how this family of parasites has evolved. Because one of the more obvious ways to categorise these taxa is by their attachment location, most molecular phylogeny papers seek to understand their phylogeny with respect to attachment site. Interestingly, these papers have not reached a consensus on the evolution of attachment. It is most likely that the disparate phylogenetic results for the family is related to the use of genetic data that have been extracted from specimens which have not been correctly morphologically identified. To work towards a more complete, accurate, and reproducible molecular phylogeny for this family, the aim of this paper was to use cymothoids from Moreton Bay, Australia and the southern coast of South Africa as a case study to provide best practice guidelines for generating a morphological and molecular cymothoid dataset for future phylogenetic work. Because morphological and molecular phylogenies of the Cymothoidae have been documented to contradict each other, we seek to specifically discuss which aspects of morphological and molecular taxonomic data should be paired for the most accurate results. In order to achieve this aim, sequences of the partial mitochondrial genes cytochrome c oxidase subunit 1 (COI) and 16S, and the partial ribosomal gene 18S were generated for 25 cymothoid taxa. Based on the criteria formulated in the present study (e.g., peer-reviewed publication, reliable identification at species level, suitable sequence length), just 40 cymothoid sequences available in Genbank (out of 407) were included in the phylogenetic analyses. The newly generated sequences were aligned with these 40 sequences, using Aegidae and Corallanidae as outgroups. Three alignments were created: COI (43 taxa, 678 bp long), 16S (30 taxa, 500 bp long), and 18S (14 taxa, 1521 bp long). Phylogenetic analyses were run under Bayesian and Maximum Likelihood inference, generating trees for each alignment. Phylogenetic relationships for COI and 16S showed a pattern congruent with the morphological classification of the genera
Anilocra,
Ceratothoa,
Cinusa,
Cymothoa,
Mothocya,and
Nerocila, in which congeneric species clustered together in supported clades. However, species of
Elthusa did not cluster together in the COI analyses, which corroborates morphological evidence that this genus should be split into at least two. At present, there are few ribosomal gene sequences for cymothoids. Based on the available data, the 18S tree depicted two main clades: one clustering together two freshwater species from Brazil, and the other uniting 12 marine species. Despite these promising results, there is still a deficiency of ribosomal gene sequences for cymothoids. This is the first study to compare the phylogenetic relationships of cymothoid species using the most current taxonomic classifications available, as well as the first to include 18S sequences for seven species of cymothoids. This research provides guidance for moving forward towards good practice in morphological and molecular studies on cymothoids.
