Prevalence and distribution of parasites in spatially structured host populations

Changes in the size of populations due to habitat loss and fragmentation have important consequences in the ecological interactions among species, including interactions between hosts and parasites. In spatially structured populations, called metapopulations, the spread of a parasite among habitat patches is dependent on migration. However, within patches, the size of host populations may influence parasite prevalence. In experimental metapopulations of red flour beetles (Tribolium castaneum), we found that the spread A. tribolii mites among patches was proportional to the rate of migration among patches. Also, at the metapopulation level, global prevalence was low because spatial isolation alone protected some populations from infectious migrants. That is, when migration rates are low, spatial structure affords some protection from infection at the metapopulation level.;Within patches, we found that the size of the host population was positively related with prevalence, but only for small populations. In our system, populations with more than 50 hosts did not differ in prevalence, but populations with less than 50 hosts experienced reductions in prevalence directly proportional to the number of hosts. The distribution of parasites on hosts was also affected by the size of the host population. This effect was also limited to populations of fewer than 50 hosts where parasites were highly aggregated. Above 50 hosts, parasites were still aggregated, but the degree of aggregation no longer varied with population size.;In a computer simulation where we included the population dynamics of parasites within hosts, a process left out of metapopulation models, we found the same patterns we observed in the experiments. Prevalence was insensitive to host population size except for small populations. In small populations, parasite prevalence and host population size were positively related in small host populations. Also in small populations, parasite aggregation was negatively related to host population size. These results are consistent with the extinction debt model introduced by Tilman and coworkers in 1994.