Fish trophic level and the similarity of non-specific larval parasite assemblages.

Whereas the effect of parasites on food webs is increasingly recognised and has been extensively measured and modelled, the effect of food webs on the structure of parasite assemblages has not been quantified in a similar way. Here, we apply the concept of decay in community similarity with increasing distance, previously used for parasites in geographical, phylogenetic and ontogenetic contexts, to differences in the trophic level (TL) based on diet composition of fishes. It is proposed as an accurate quantitative method to measure rates of assemblage change as a function of host feeding habits and is applied, to our knowledge for the first time, across host species in marine waters. We focused on a suite of 15 species of trophically-transmitted and non-specific larval helminths across 16 fish species (1783 specimens, six orders, 14 families) with different sizes and TLs, gathered from the same ecosystem. Not all host species harboured the same number and types of parasites, reflecting the differences in their ecological characteristics. Using differences in TL and body length as measurements of size and trophic distances, we found that similarity at both infracommunity and component community levels showed a very clear decay pattern, based on parasite abundance and relative abundance, with increasing distance in TL, but was not related to changes in fish size, with TL thus emerging as the main explanatory factor for similarity of parasite assemblages. Furthermore, the relationships between host TL and assemblage similarity allowed identification of fishes for which the TL was under- or over-estimated and prediction of the TL of host species based on parasite data alone.