725487.fig.002a
(a)
725487.fig.002b
(b)
725487.fig.002c
(c)
Figure 2: Facilitation acting as a cohesive force and reducing the chance of ecological speciation. (a) represents the population divergence at neutral loci as a function of the size of patch. (b) represents the same population divergence at selected loci. (c) represents the average fitness in each environment as a function of the patch size (Box  1). Consistent with our hypothesis, in the absence of benefactors (i.e., no facilitation, patch size of 0), populations were locally adapted and gene flow was reduced between the core and the marginal habitat (higher FST on both neutral and selected loci). In the presence of benefactors, adaptation to the marginal environment was reduced by the increased gene flow from the core population and especially from populations occupying the mild patches created by the benefactors. Individuals occupying these benefactor patches have an increased fitness because of the low population density in the neighboring maladapted marginal environment that inflates the logistic growth part of the fitness function. As the size of the patches increased, migration and gene flow decreased between the patches created by the benefactors and marginal environment (increased FST on both neutral and selected loci). Because of this reduction in gene flow, the average fitness in the marginal environment increased, and consequently the fitness advantage in the benefactor patches decreased. Additionally, the FST for selected loci between the core and the benefactors, even though still high, is reduced at large patch size. This is because each patch has a subpopulation that can achieve the same phenotype using different genotypic combinations. Thus, as the patch size increased the number of patches decreased, the number of subpopulation decreased, the genotypic variance within an environment decreased, and so did the FST.