The foundational concepts behind the persistence of ecological communities have been based on two mathematical properties: dynamical stability and feasibility. The former is typically regarded as the capacity of a community to return to an original equilibrium state after a pulse perturbation, and is linked to the strength of interspecific interactions. The latter is the capacity of the community to sustain positive abundances on all its constituent species, and is linked to both interspecific interactions and species demographic characteristics. Over the last 40 years, theoretical research in ecology has emphasized the search for conditions leading to the dynamical stability of communities, while the conditions leading to feasibility have been overlooked. However, thus far, we have no evidence of whether species interactions are more conditioned by the community's need to be stable or feasible. Here, we introduce novel quantitative methods based on structural stability analysis and use empirical data to investigate the extent to which the persistence of mutualistic networks depends on their dynamical stability and feasibility.
Results/Conclusions
We demonstrate that the persistence of mutualistic networks does not depend on stability as much as it does on feasibility. First, we find that the more nested the species interactions in a mutualistic network are, the lower the mutualistic strength that the community can tolerate without losing dynamical stability. Second, we find that high feasibility in a mutualistic network can be reached either with high mutualistic strength or with highly nested species interactions. Importantly, we find that mutualistic networks have a tendency to increase feasibility through their nested species interactions, promoting feasibility over stability.These findings reveal that mutualistic networks may not need to be highly dynamically stable to persist. This supports the idea that allowing high mutualistic strengths can push communities to shift from a weak to a strong mutualistic regime, which can easily take the community to rather unpredictable dynamics. This also supports empirical observations that mutualisms among free living species are of low specificity, which is compatible with the combination of coevolutionary convergence and complementarity. In both cases, mutualistic communities may have a higher opportunity to increase persistence through higher levels of feasibility instead of stability by changing the organization of their interactions rather than by increasing the overall mutualistic strength. This calls for a stronger research program on the factors modulating feasibility conditions in ecological networks, as they can hold the key for a general theory of community persistence.