Current opinions: Zeros in host-parasite food webs: Are they real?

As the data have poured in, and the number of published food webs containing parasites has increased, questions have been raised as to why free-living species consistently outnumber parasites, even though most general reviews on the subject of host:parasite species richness suggest the contrary. Here, I describe this pattern as it exists in the literature, posit both real and artifactual sources of these findings, and suggest ways that we might interpret existing parasite-inclusive food webs. In large part, the reporting of free-living species devoid of any associated parasites (termed here in the coding of food web matrices as "zeros") is a consequence of either sampling issues or the intent of the study. However, there are also several powerful explanatory features that validate real cases of this phenomenon. Some hosts appear to authentically lack parasitism in portions of their geographic ranges, and parasites are often lost from systems that are either in early phases of community re-colonization or are compromised by environmental perturbation. Additionally, multi-stage parasite life cycles and broad host spectra allow some parasite species to partially saturate systems without providing a corresponding increase in parasite species richness, leading to low parasite species richness values relative to the free-living community. On the whole, the existing published food webs are sufficient to, at least in principle, determine basic patterns and pathways associated with parasite establishment and persistence in free-living communities because (1) for the purpose of those features, species rarity is roughly analogous to absence and (2) the existing data seem to suggest that the addition of more parasite taxa would reinforce the patterns already observed. This is particularly true for helminth parasites, in which our understanding and the resolution of our work is most robust.