Local adaptation and the geometry of host–parasite coevolution

Gandon, S. (2002). Local adaptation and the geometry of host–parasite coevolution. Ecology Letters, 5(2), 246-256.

The process of coevolution, whether among mutualists, competitors, or hosts and parasites, can depend on the spatial structure of the environment. In particular, when an environment contains multiple populations, each population can experience local adaptation that changes gene frequencies, even when populations are connected by migration. In this paper, Gandon (2002) asks how local adaptation in host and parasite populations differs depending on the relative migration of the two groups, and how parasite specificity and virulence alter these interactions.

In theory, parasites should be “ahead” of hosts in terms of local adaptation, since they have shorter generation times and larger population sizes. However, this theory does not account for spatial structure. Previous simulation results have shown that relative migration rates can be important in determining local adaptation among populations, where the group (i.e. hosts or parasites) with the higher migration rate is more locally adapted. Here, Gandon produces similar results from an analytic solution to a deterministic model. In this model, host resistance is controlled by a single allele that provides resistance to one of two parasite types in the system and both hosts and parasites migrate between populations at a given, distance-independent, rate. Using this structure, Gandon defines local adaptation simply as the difference in host or parasite performance in its “home” and “away” populations.

Overall, Gandon’s analytic results confirm previous results from simulations. Specifically, he finds that the group (host or parasite) with the higher migration rate is more locally adapted, but that local adaptation occurs only below a threshold value of host and parasite migration rates. When local adaptation does occur, host and parasite populations enter a stable limit cycle whose period and amplitude are determined by relative migration rates and parasite virulence. Virulence further amplifies these results by favoring the group that was already “ahead” (i.e. that with a higher migration rate). Gandon concludes that, contrary to usual conceptions of genetic drift as a process that creates homogeneity, genetic drift can maintain spatial variation among populations in this model where the environment is considered homogeneous.