Coexistence of Habitat Specialists and Generalists in Metapopulation Models of Multiple-Habitat Landscapes

Abstract

In coarse-grained environments specialists are generally predicted to dominate. Empirically, however, coexistence with generalists is often observed. We present a simple, but previously unrecognized, mechanism for coexistence of a habitat generalist and a number of habitat specialist species. In our model all species have a metapopulation structure in a landscape consisting of patches of different habitat types, governed by local extinction and colonization. Each specialist is limited to its specific type of habitat. The generalist can use more types of habitat, has a lower local competitive ability but can exploit patches left open by the specialists. Our modeling shows that coexistence is easily possible. The mechanism amounts to a colonization/competition trade-off at the landscape level, where the colonization advantage of the inferior competitor does not arise from a higher colonization rate but from its ability to use more types of habitat. Habitat availability has to be intermediate: when there are few patches of each habitat, only the generalist is able to maintain itself and when there are many patches, high propagule pressure of the specialists excludes the generalist. Habitat selection or temporal variations in relative habitat quality are not necessary for coexistence. Increased niche-width, colonization rate or local competitive ability of the generalist enhances its performance compared to the specialists. Various types of habitat degradation favour generalism. When able to use a broad range of habitats, generalists can generate so much propagule pressure that only a low level of local competitive ability is needed to globally exclude the specialists. Hence, in a reversal of the original problem, the question is why there are so many specialist metapopulations?

Appendix: critical γ for specialist existence

Without specialists $g^{*} = N - \frac{e}{hc}$. Eq. (1a) then gives $\frac{{{\text{d}}s}}{{{\text{d}}t}} = sc(N - s - \gamma \left[ {N - \frac{e}{hc}} \right]) - es - c\gamma hs\left[ {N - \frac{e}{hc}} \right]$. When N is large the specialist cannot invade ($\frac{{{\text{d}}s}}{{{\text{d}}t}} < 0{\text{ for small }}s$) when $\gamma > \frac{1}{1 + h}$.