Parasite community structure within and across host populations of a marine pelagic fish: how repeatable is it?

The geographical variation in parasite community structure among populations of the same host species remains one of the least understood aspects of parasite community ecology. Why are parasite communities clearly structured in some host populations, and randomly assembled in others? Here, we address this fundamental question using data on the metazoan parasite communities of different host size-classes of four distinct populations of a small pelagic fish, the Argentine anchovy, Engraulis anchoita, from the South West Atlantic. Within each fish sample, fish length was correlated with both the total intensity of parasites and species richness among infracommunities. More importantly, average fish length correlated with mean infracommunity richness and mean total intensity across the fish samples, indicating that the characteristics of parasite assemblages in a fish population are strongly influenced by the size of its fish in relation to those in other populations. Nested subset patterns were observed in about half of the fish samples. This means that the presence or absence of parasite species among fish individuals is often not random; however, no repeatability of nestedness among component communities was observed. Average fish length did not influence directly the likelihood that a parasite assemblage was significantly nested. However, variables influenced by average fish length, namely mean infracommunity richness and mean total intensity, determine the probability that a nested hierarchy will be observed; host size may thus indirectly affect parasite community structure either itself or via its influence on host movement and feeding patterns. To some extent, this apparent link may be due to the sensitivity of nestedness analyses to the proportion of presence in a presence/absence matrix; this in itself is a biological feature of the parasite community, however, which is associated with mean host length.