LETTER Effects of local negative feedbacks on the evolution of species within metacommunities Nicolas Loeuille, 1 * and Mathew A. Leibold 2 1 IEES Paris, UMR7618, UPMC-CNRS, Universit e Pierre et Marie Curie, 7 quai St Bernard, 75005 Paris, France 2 Department of Ecology and Evolution, University of Texas at Austin, Austin, TX 78705, USA *Correspondence: E-mail: nicolas. loeuille@upmc.fr Abstract Local negative feedbacks occur when the occupation of a site by a species decreases the subse- quent fitness of related individuals compared to potential competitors. Such negative feedbacks can enhance diversity by changing the spatial structure of the environment. The conditions, how- ever, involve dispersive, environmental and evolutionary processes in complex interactive ways. We introduce a model that accounts for four mechanisms: colonisation-competition-extinction ecological dynamics, evolutionary dynamics, local negative feedbacks and environmental averag- ing. Three qualitatively distinct dynamics are possible, one dominated by specialists, another dom- inated by generalists and an intermediate situation exhibiting taxon cycles. We discuss how metacommunity diversity, macro-ecological patterns and environmental patterning are linked to the three qualitative dynamics. The model provides classical shapes for morph-abundance distri- butions, or diversity-area relationships. Diversity can be high when specialists dominate or when taxon cycles happen. Finally, local negative feedbacks often yield fine-grain environments for taxon cycle dynamics and coarse-grain environments when generalists dominate. Keywords Eco-evolutionary feedbacks, environmental grain, Janzen-Connell hypothesis, metacommunity, metaecosystem, niche construction, species abundance distribution, species area relationship, taxon cycle. Ecology Letters (2014) INTRODUCTION Natural communities have a capacity to maintain high diver- sity that is often surprisingly large (Hutchinson 1961; May 1973). While there is no shortage of possible explanations (Wilson 1990; Roy & Chattopadhyay 2007), a number of these mechanisms work by producing local negative feedbacks where the fitness of particular species decreases disproportion- ately as a function of their local density or occupancy. Local negative feedbacks were initially proposed to rely on the accumulation of pests and pathogens on sessile adult organ- isms such as trees that inhibit survival of their offspring that settle nearby (Janzen 1970; Connell 1971), but the general process can involve a number of other possible mechanisms. Particularly, consumption constraints can also yield local negative feedbacks, for instance when a consumer mostly takes up or retains its most limiting nutrient, thereby creating stoichiometric constraints for its future competitive ability (Daufresne & Loreau 2001), or when a predator mostly depletes its preferred prey. There is increasing experimental evidence for the presence of local negative feedbacks (Klironomos 2002; Diez et al. 2010; Mangan et al. 2010; Hawkes et al. 2013), as well as for their potential to structure natural communities at different scales. It has been suggested that local negative feedbacks can play a major role in the invasiveness of natural communities (Klironomos 2002; Diez et al. 2010), in affecting their species abundance distribution (Mangan et al. 2010) and in constrain- ing the phylogenetic structure of species assemblages (Burns & Strauss 2011). While the experimental tests of negative feedbacks are usually implemented at the individual scale, these effects on community patterns suggest that they also apply to larger spatial scales (Kardol et al. 2013). While there are probably important differences that result from the specifics of the situation under study, there are also important commonalities among these processes that are likely to determine to what degree local negative feedbacks actually lead to enhanced diversity. These include the dispersal ability of the organisms involved, the spatial diffusion of the feedback effect (i.e. ‘spillover’), the frequency of disturbances that disrupt the feedback process, the rate at which the nega- tive effects decay after extinctions and the adaptive evolution- ary responses of the organisms involved. Dispersal ability is important because it determines the fitness cost of the local feedback; highly dispersive organisms (relative to the spatial extent of the local feedback effect) will not be as strongly affected by local negative feedbacks as are less dispersive ones. Spillover of the feedback effect is important in affecting the spatial extent of the effect. It interacts with dispersal ability; high spillover enlarges the scale of the local feedback, thus counteracting the effects of dispersal but it only does so to a point because spillover can also serve to homogenise and destroy the spatial structure that would otherwise exist. The frequency of disturbances is important because it breaks up the spatial patterning and scale of the feedback effects; under high disturbance rates local feedbacks are less likely to be important. The decay rate of the feedback effects determines how long the ‘legacy’ effects of the local feedbacks last; slower decay implies that current local feedbacks are weak (Kardol et al. 2013). Finally, adaptive evolution is important for © 2014 John Wiley & Sons Ltd/CNRS Ecology Letters, (2014) doi: 10.1111/ele.12258