Although research on parasite biodiversity has intensified recently, there are signs that parasites remain an underestimated component of total biodiversity in many regions of the planet. To identify geographical hotspots of parasite diversity, we performed qualitative and quantitative analyses of the parasite-host associations in fishes from Latin America and the Caribbean, a region that includes known hotspots of plant and animal biodiversity. The database included 10 904 metazoan parasite-host associations involving 1660 fish species. The number of host species with at least 1 parasite record was less than 10% of the total known fish species in the majority of countries. Associations involving adult endoparasites in actinopterygian fish hosts dominated the database. Across the whole region, no significant difference in parasite species richness was detected between marine and freshwater fishes. As a rule, host body size and study effort (number of studies per fish species) were good predictors of parasite species richness. Some interesting patterns emerged when we included only the regions with highest fish species biodiversity and study effort (Brasil, Mexico and the Caribbean Islands). Independently of differences in study effort or host body sizes, Mexico stands out as a hotspot of parasite diversity for freshwater fishes, as does Brasil for marine fishes. However, among 57 marine fish species common to all 3 regions, populations from the Caribbean consistently harboured more parasite species. These differences may reflect true biological patterns, or regional discrepancies in study effort and local priorities for fish parasitology research.

The host specificity of a parasite is not merely a function of how many host species it can exploit, but also of how closely related these host species are to each other. Here, a new index of host specificity is proposed, one that takes into account the average taxonomic or phylogenetic distance between pairs of host species used by a parasite. The index is derived from measures of taxonomic distinctness used in biodiversity studies. It is easy to compute and interpret, ranging from a minimum value of 1 when all host species are members of the same genus, to a maximum of 5, when all host species belong to different classes. The variance of this measure can also be computed, and provides additional information on the taxonomic or phylogenetic structure of the host assemblage. Using data on helminth parasites of Canadian freshwater fishes, we show that the new index, unlike the mere number of known host species, is independent of study effort i.e. the number of published records of a parasite. Although the index and the number of known hosts are not entirely independent statistically, each captures a different aspect of host specificity. For instance, although acanthocephalans infect significantly more host species than trematodes, cestodes or nematodes, there is no difference in the average index value among these 4 helminth taxa, suggesting that the average taxonomic distances between the host species of a parasite do not vary among these higher taxa. We recommend the use of our new index in future comparative studies of host specificity, in particular when the focus is on the evolutionary history of parasites and of their past colonizations of host lineages.