University of Fribourg Fribourg, Fribourg, Switzerland
Background/Question/Methods
How many, and how, species can be arranged on a niche axis in such a way that they coexist, is a long-standing question in ecology dating back to the works of Hutchinson and MacArthur. Since species interactions are also powerful evolutionary forces, coevolved communities should differ substantially in this respect from random assemblages. More recently, attention has been brought to how evolution affects species coexistence and productivity. Here, we address these questions in a niche-based model, where species are defined by their position on a resource axis. We start with one morph and let biodiversity emerge through successive branching events. We follow the evolution of niche positions, coexistence metrics, and productivity between and following branching points, up to the formation of an evolutionary stable community. Finally, we compare evolutionary outcomes against Monte-Carlo randomized communities.
Results/Conclusions
We find that along the course of evolution, both fitness and niche differences (sensu coexistence theory) increase, but drop at evolutionary branching points. When compared to randomized communities, evolution converges towards singular strategies (branching points and evolutionary stable communities) located on Pareto fronts which optimize the niche difference relative to the fitness difference. In other words, evolution tends to compromise between maximizing the stabilizing effect of niche difference and minimizing the destabilizing effect of fitness difference. This confirms that evolutionary communities are highly non-random. Evolution also leads to an increase in net biomass production in the community, as a result of higher complementarity due to the niche space becoming fully occupied. That being noted, we show that productivity indicators are not strictly maximized except in monomorphic systems. In higher dimensions, the niche-fitness difference trade-off restricts access to even more productive trait configurations.
