Abstract. Inbreeding in local host populations will be a common phenomenon in hostpathogen systems that are characterized by metapopulation dynamics, i.e., frequent extinction and recolonization of local host populations by small numbers of founding individuals. As an example of a pathosystem with metapopulation dynamics we investigated the impact of inbreeding in the host plant Silene alba on its interaction with the anthersmut fungus Microbotryum violaceum. Seeds from eight populations of S. alba were sampled, and five generations of sib mating resulted in 65 inbred lines, with inbreeding coefficients of f ϭ 0, 0.25, 0.375, 0.5, and 0.59 per line. In a first experiment these lines were tested for active, biochemical resistance against fungal infection, by artificially inoculating individuals. The percentage of infected individuals differed significantly among populations, lines, and inbreeding levels, and both population-by-inbreeding level and lineby-inbreeding level interactions were significant. The most striking result was the strong variance in inbreeding effects among lines; inbreeding resulted in increased resistance in some lines and decreased resistance in others. In a second experiment for 12 inbred lines, originating from one population, active resistance and flower traits associated with passive resistance (avoidance) to this insect-vectored, florally transmitted disease were measured. Significant inbreeding depression was demonstrated for petal size and nectar volume. Thus inbreeding might enhance avoidance of spore transmission by insects. For both active resistance and all flower traits, significant line-by-inbreeding level interactions were found. The results indicate that the effect of inbreeding on the interaction between host and pathogen in this pathosystem is unpredictable at the local population level, because: (1) strong genotypic differences in inbreeding effect exist for both active and passive resistance, making the effect of inbreeding at the population level dependent on the genotypic composition of the (founder) population; (2) effects of inbreeding on active and passive resistance were not correlated, making the net effect of inbreeding on field resistance unpredictable; and (3) in several lines, evidence for epistatic effects was found, making the effect of inbreeding dependent on the actual inbreeding level of the genotype. The results underscore that most progress in the study of host-pathogen interactions may be expected from an integrated ecological and genetic approach.

Inbreeding in local host populations will be a common phenomenon in host-pathogen systems that are characterized by metapopulation dynamics, i.e., frequent extinction and recolonization of local host populations by small numbers of founding individuals. As an example of a pathosystem with metapopulation dynamics we investigated the impact of inbreeding in the host plant Silene alba on its interaction with the anther-smut fungus Microbotryum violaceum. Seeds from eight populations of S. alba were sampled , and five generations of sib mating resulted in 65 inbred lines, with inbreeding coefficients of f 0, 0.25, 0.375, 0.5, and 0.59 per line. In a first experiment these lines were tested for active, biochemical resistance against fungal infection, by artificially in-oculating individuals. The percentage of infected individuals differed significantly among populations, lines, and inbreeding levels, and both population-by-inbreeding level and line-by-inbreeding level interactions were significant. The most striking result was the strong variance in inbreeding effects among lines; inbreeding resulted in increased resistance in some lines and decreased resistance in others. In a second experiment for 12 inbred lines, originating from one population, active resistance and flower traits associated with passive resistance (avoidance) to this insect-vectored, florally transmitted disease were measured. Significant inbreeding depression was demonstrated for petal size and nectar volume. Thus inbreeding might enhance avoidance of spore transmission by insects. For both active resistance and all flower traits, significant line-by-inbreeding level interactions were found. The results indicate that the effect of inbreeding on the interaction between host and pathogen in this pathosystem is unpredictable at the local population level, because: (1) strong genotypic differences in inbreeding effect exist for both active and passive resistance, making the effect of inbreeding at the population level dependent on the genotypic composition of the (founder) population; (2) effects of inbreeding on active and passive resistance were not correlated, making the net effect of inbreeding on field resistance unpredictable ; and (3) in several lines, evidence for epistatic effects was found, making the effect of inbreeding dependent on the actual inbreeding level of the genotype. The results underscore that most progress in the study of host-pathogen interactions may be expected from an integrated ecological and genetic approach.