Host-Pathogen Dynamics in a Metapopulation Context: The Ecological and Evolutionary Consequences of Being Spatial

The metapopulation concept is useful when considering ecological and evolutionary dynamics of spatially structured populations. However, debate has focused on genetic variation that is neutral rather than under selection. This distinction is particularly important in antagonistic or co-evolutionary interactions such as host-pathogen or predator-prey systems. Plant host-pathogen systems provide some of the best examples of studies in which numerical and genetic dynamics have been investigated in a spatially explicit context, and where genes under selection can be unambiguously identified. Empirical studies of natural host-pathogen interactions have shown that, while in some cases pathogens appear to be locally adapted to their hosts, in others there is no local correspondence between resistance and virulence genes. Recent theory suggests that the dynamics (epidemic vs. endemic) and migration rates of host and pathogen will be important factors in the maintenance of genetic polymorphisms in resistance and virulence. We argue that the relative spatial scales at which hosts and pathogens interact are crucial to understanding the evolution of resistance/virulence structure. Pathogens that disperse further than their hosts will be more likely to show gene-for-gene interactions than pathogens dispersing over spatial scales similar to or smaller than their hosts. We predict that life-history features that influence encounter rates between specific host and pathogen genotypes will be important factors in determining the evolution of resistance-virulence structures. In particular, we would expect correlative and/or causal relationships between pathogen life-history features (e.g. local vs. systemic infection, type (fecundity vs. mortality) and severity of effects], and whether dynamics are endemic or epidemic. Plant host-pathogen systems provide ideal models for investigating the evolution of non-neutral genetic variation in spatial systems. Understanding the co-evolution of such systems will require research programmes that integrate long-term descriptive and experimental studies of multiple populations, with analytical and computer simulation modelling and comparative/phylogenetic studies.