The effect of clonal integration on plant competition for mosaic habitat space

Physiological integration between ramets has been observed in several clonal plant species. But consequences of integration on the competitive ability and spatial development of genets have received little attention so far. This study is an attempt to examine the population- and community-level implications of integration in a spatially explicit model. We have simulated different resource patterns in a cellular automata, varying the proportion (p) and size (s) of resource-rich patches and the average resource level in the area (h). We compared the efficiency of integrator to splitter genets in exploring and utilizing the resource patches. In the integrator, ramets that belonged to the same genet were physiologically connected and shared the resource taken up by any part of the genet. In the splitter, all ramets were autonomous. i.e., survival and reproduction of each ramet only depended on the local resource conditions. Thus, the integrator sensed and responded to environmental heterogeneity on a larger spatial scale than the splitter. Integration can be interpreted as sharing the risk of genet mortality among interconnected ramets, while splitting represents risk spreading among autonomous ramets. In each simulated habitat type, we followed the course of competition between splitter and integrator genets for 500 generations. First, we examined the effect of environmental heterogeneity on the result of competition. We fixed the average resource level in the area and changed the spatial distribution of the resource by varying p and s. We found that the smaller the proportion of favorable sites, the better it is to integrate. But already a small deviation from random to clumped resource distribution (i.e., relatively small s) made the integrator competitively inferior to the splitter. In the next step, we varied the overall resource richness in the area. High productivity of the habitat promoted the dominance of integrators. The simulations demonstrated that there were habitat types in which one of the strategies (the integrator or the splitter) rapidly excluded its competitor. But at the medium range of environmental parameters, long-term coexistence of the strategies became possible. As integration proceeded simultaneously with horizontal growth and exploration of new patches, integrators remarkably differed from splitters in their pattern of space occupation. A spatial statistical analysis revealed that splitters maintained strong spatial association with the resource: splitting promoted habitat selection. Integration, on the other hand, led to gap-filling, fugitive spatial behavior. For a range of realistic habitat patterns, we need not expect the exclusion of any of the pure strategies-splitters and integrators can coexist.