Ulf Dieckmann, Krisztian Magori, Géza Meszéna, Beáta Oborny

Paper #: 03-08-047

Plant species show great variation in the degree of physiological integration between developmental units (modules). When this degree is minimal, individual modules are self-supporting and compete with other modules. When the degree of integration is higher, modules remain physiologically connected and “cooperate” by sharing resources like water, nutrients, and photoassimilates taken up from their local environments. In such a manner, local differences in habitat quality can be diminished within a group of modules. Here we investigate how the evolutionarily optimal degree of integration depends on habitat type--with habitats being characterized by the proportion of resource-rich and resource-poor sites and by the turnover rate between these. Two main questions are addressed: First, how does spatial heterogeneity influence natural selection for or against integration? Second, can adaptation, under reasonable ecological conditions, stabilize partial integration? A non-spatial version of the model, which assumes well-mixed populations, predicts the complete physiological independence of modules as the only evolutionarily stable outcome in any realistic habitat type. By contrast, a spatially explicit version of the model reveals the adaptive advantage of integration in typical high-risk habitats, where resource-rich sites are sparsely distributed in space and transient in time. We conclude that habitat diversity without spatial population structure suffices to explain the evolutionary loss of physiological integration. But only the additional consideration of spatial population structure can convincingly explain any backward transition, and the stable existence of partial integration.