Phenotypical patterns and coevolutionary dynamics in large landscapes

Background/Question/Methods

Species are formed by groups of spatially structured populations, and the spatial structure can influence the evolutionary outcomes of species interactions. The interplay between selection at local sites and gene flow across space can lead to qualitatively new patterns of phenotypic distributions, affecting the coevolutionary process and its outcomes. Our goal is to understand how the consequences of spatial organization scale up when studying large landscapes formed by thousands of sites. We propose a theoretical model to understand the effects of the spatial structure on the coevolutionary dynamics of species that interact locally via mutualism or antagonism but can migrate to nearby locations. We define the structure of selection between a pair of interacting species as being mutualistic, when both species benefit from the interaction, leading, as a result, to fixation of phenotypes; or antagonistic, when only one species benefits from the interaction leading to polymorphic states. We modeled the coevolutionary dynamics of two interacting species across a spatial landscape composed of a lattice with 10000 sites. The outcome of the interactions is always positive to one partner and may be positive or negative to the other interacting partner.

Results/Conclusions

The evolutionary dynamics leads to the emergence of spatial patterns that are stationary, which are strongly dependent on the rate of gene flow and the proportion of mutualistic sites in the landscape. For low rates of gene flow, phenotypic patterns match what is expected by the structure of selection. However, we found that coupling promoted by gene flow across nearby sites creates complex phenotypic patterns that cannot be predicted from the structure of selection.  In scenarios with an overrepresentation of antagonistic sites, a high fraction of the sites show a polymorphic state, although local overrepresentation of mutualistic sites in some regions leads to the emergence of clusters of sites in which there is fixation or near fixation of phenotypes. Finally, when antagonism is not spatially predominant, there is a clear tendency to fixation of phenotypes. The interplay between spatial structure, gene flow and interaction outcomes can lead to the emergence of clusters of phenotypic patterns that differ significantly from what would be expected by the spatial structure of selection. This implies that neither spatial distribution of phenotypes nor spatial structure of selection alone allows inferring the coevolutionary dynamics of species interactions.