Heterogeneous hosts: how variation in host size, behaviour and immunity affects parasite aggregation.

Infection heterogeneity is one of the most fundamental patterns in disease ecology, yet surprisingly few studies have experimentally explored its underlying drivers. Here, we used large-scale field assessments to evaluate the degree of parasite aggregation within amphibian host populations followed by a novel experimental approach to assess the potential influence of host size, behaviour and immunity in reproducing such heterogeneity. Among 227 wetlands, 2468 hosts and seven parasite species, infections were consistently aggregated among host individuals within populations of the Pacific chorus frog (Pseudacris regilla). For each parasite species, the relationship between the log-mean and log-variance of infection load was strongly linear (R(2): 0·91-0·98) with a slope between 1·37 and 1·67, indicative of aggregation relative to the expected Poisson slope of unity. In laboratory trials with P. regilla and the most virulent trematode (Ribeiroia ondatrae), experimental reductions in either host immunity (through corticosterone exposure) or antiparasite behaviours (through anaesthesia exposure) increased parasite infection loads in isolated hosts by 62-102% relative to unmanipulated individuals. In a second experiment designed to test how variation in host immunity, behaviour and body size affected variation in infection load within small groups (dyads), a reduction in immune function or behaviour of one host significantly amplified infection heterogeneity within the group, effectively doubling the variance-to-mean ratio. However, immunity affected aggregation only in the absence of behavioural manipulation, and changing the size distribution of hosts did not appreciably affect aggregation. Using Taylor's Power Law to integrate field and laboratory data, we found that only treatments involving behavioural reductions achieved aggregation levels comparable to natural host populations. Thus, despite their short duration, our treatments generated heterogeneity in infection loads similar to natural observations. These results emphasize how, alongside extrinsic variation in parasite exposure risk, individual host attributes generally and behaviour in particular have the potential to influence infection success and parasite aggregation. Continued integration of infection heterogeneity research across space, among host species, and over time has important implications for understanding and managing human and wildlife diseases.