Theoretical and empirical studies of metapopulations: population and genetic dynamics of the SileneUstilago system

Abstract
Host–pathogen population dynamics may often only be understood by a multifaceted approach designed to understand processes at a regional as well as local scale. We have investigated the regional population dynamics of the anther-smut Ustilago violacea, a pollinator-transmitted fungal disease, and its plant host Silene alba, using descriptive, experimental, and theoretical approaches. A 7-year survey of multiple natural populations revealed persistence of host and pathogen despite a high rate of population turnover. In an experimental metapopulation, disease spread was greater and more rapid in populations that were relatively isolated or had a previous history of disease occurrence. A computer simulation showed that spatial substructuring can drastically alter expectations based on analytical results from single population models of host–pathogen systems. Moreover, the simulation reproduced many of the patterns detected in the long-term survey and predicted that healthy populations should be more resistant than diseased ones, as found experimentally. Key words: metapopulations, host–pathogen dynamics, spatial models, anther-smut diseases.