Search

One of the single most extraordinary examples of host-parasite co-evolution is shown by the isopod family Cymothoidae, of which all species are obligate parasites of fishes, including many commercially important fish species. Cymothoids are one of the most diverse isopod groups, with 400 species across 43 genera, and they exhibit striking parasitic strategies. Some species, for example, are known to supplant their host’s tongue. In addition to mouth attachment, cymothoids also attach externally to the skin, within the gills, or burrow within the body cavities of their host. The majority of species use only one of these attachment strategies, and attachment location is also largely conserved within cymothoid genera. Yet, there is variation in microhabitat use between species with the same parasitic mode, because distinct locations or orientations are used. As well as specificity in parasitic strategies, cymothoids are highly host specific, with most species restricted to a few host species. Due to their bizarre life-histories and their large size relative to their hosts, cymothoids have been studied since the early 19th century. However, the majority of this work is related to traditional taxonomy and, hampered by high intraspecific variation and lack of an evolutionary framework for the group, species boundaries and relationships between taxa remain unclear. Preliminary molecular evidence suggests that the evolution of attachment in cymothoids has a complex history, but to better understand cymothoid-host co-evolution a more densely sampled phylogeny is required. Here we examine several aspects of cymothoid evolution using museum specimens. First, we use a geometric morphometric approach to quantify morphological variation vAbstract in attachment appendages, and relate this to the different parasitic strategies and phy- logeny. In the process of producing mitogenome reference sequences for cymothoids, we investigate the placement of Cymothoidae within Isopoda, test the monophyly of several isopod subclades, and investigate the origin of isopod terrestrialisation. Finally, using our reference mitogenome sequences, we recover further mitogenome sequences from museum specimens using a low-coverage shotgun sequencing approach, and use these data to reconstruct cymothoid phylogeny. We show that, after accounting for shared ancestry, attachment morphology is strongly influenced by parasitic strategy. Whole mitogenome sequences are not suitable for resolving isopod inter-familial relationships. We find some evidence that cymothoids, and possibly all isopods, have an atypical mitogenome structure (previously characterised), structural variation in which causes compositional biases and long-branch artefacts. Sophisticated modelling and data treatment can alleviate some of these effects but we caution the validity of inferred relationships based on mitogenome data. We successfully obtain sequence data from liquid-preserved museum specimens; find that the evolution of attachment in cymothoids is complex, and that freshwater species from South America likely colonised rivers in a single event.