Résumé
The complexity of food webs and how it depends on environmental variables is a longstanding question in ecology. Food-chain length depends on many factors, such as productive space, disturbance and spatial processes. It is not clear though how food-chain length should vary with adaptive evolutionary changes of the constitutive species. Assuming that each trophic level is subject to a competition-colonization trade-off, we model the adaptive evolution of their colonization rates and its consequences on equilibrium occupancies and food-chain length. When colonization rates are allowed to evolve, longer food chains can persist. Extinction, perturbation and habitat loss all affect the evolutionarystable colonization rates, but trade-off strength (costs of dispersal) has a major role: weaker trade-offs yield longer chains. Our results suggest a strong link between within-trophic level competition, spatial occupancies and food-chain length. Although these eco-evo dynamics partly alleviate the constraint on food-chain length in metacommunities, we show it is no magic bullet: the highest, most vulnerable, trophic levels are also those that least benefit from evolution. Our model generates qualitative predictions regarding how trait evolution affects the response of communities to disturbance and habitat loss. This highlights the importance of eco-evolutionary dynamics at metacommunity level in determining food-chain length.