A Computational Ecological Genetic Model of Phenotypic Plasticity in Species Interactions

Multiple species respond to each other in their co-existing communities. Such reciprocal phenotypic plasticity can shape the behaviour and evolution of ecological communities, but its genetic architecture remains elusive. We address this issue by developing a computational mapping model that combines community ecology and quantitative genetics into a unifying context. Culturing a pair of genotypes from two species in a socially isolated environment (monoculture) and a socialised environment (co-culture) allows for the quantitative estimation of reciprocal phenotypic plasticity, that is, the difference between trait values of one (defensive) species expressed in the monoculture and the co-culture with the other (offensive) species. Classic quantitative genetic theory is implemented to map genetic variants for reciprocal phenotypic plasticity, including defensive loci derived from the defensive species and offensive loci from the offensive species, and estimate the direct effects of the defensive loci, the indirect effects of the offensive loci and horizontal epistasis due to interactions between the two genomes. We design an ecological genetic experiment of monocultures and co-cultures using 100 pairs of genotyped Escherichia coli and Staphylococcus aureus strains, from which the model identifies the existence of defensive and offensive loci, despite a remarkable asymmetry between the two bacterial species. We find that horizontal epistasis between defensive and offensive loci plays a sizable role in mediating reciprocal phenotypic plasticity. The biological functions of these identified loci are annotated via GO analysis. Our model could produce unique results that shed light on the genetic mechanisms of interspecies interactions and their adaptation in ecological communities.