The influence of selection, drift and immigration on the diversity of a tropical tree community

Abstract

Ecology is rich in theories that aim to explain why natural communities have as many species as they do. Neutral theory, for example, supposes that a community’s diversity depends on the rate at which it gains species by immigration or speciation and loses them to ecological drift [1–5]. Classical niche theory, by contrast, supposes that diversity is regulated by the complexity of the environment: how many dimensions of resources it has and how finely species can subdivide them [6–10]. These theories are about levels of diversity at equilibrium. But non-equilibrium theory supposes that communities are perpetually buffeted by environmental change so that communities rarely contain all the individuals and species they might [11, 12]. When that happens, some species may profit from their immediate circumstances, but their gains are short lived as the environment changes again, favouring others. Such theories are often seen as competing visions of nature (e.g., [1, 2, 13–20]), but they can also be viewed as collectively describing a set of forces, any of which may be at work at a given time and place (cf. [21]). The relative importance of these forces in shaping the evolution of a community’s diversity can be captured by a small set of parameters: the community’s effective size, Ne, the rate at which it gains new species, μ, and the magnitude and form of species-specific selection coefficients, s [22]. Here we present a way of estimating these parameters using time series data and apply it to the famous Barro Colorado Island Neotropical forest dataset. We show that, for the last thirty years, this community has been dominated by directional selection. We then simulate the evolution of this community in order to disentangle how these forces have shaped the species diversity that we see today. We show that, while species richness can be maintained by a neutral force, immigration, species evenness cannot and argue that it is likely maintained by temporally varying selection driven by environmental change [23–25].