From species survival to community persistence in mutualistic networks

Background/Question/Methods

Recent work has shown that the architecture of mutualistic networks increases species richness by minimizing competition relative to facilitation. It is, however, currently unknown whether each species contributes similarly to---and benefits in equal portions from--- this relationship between network structure and network persistence. Knowing the sign and magnitude of these relationships is important to understanding the trade-offs faced by species when they form part of a network. In this talk, we address this issue by combining a suite of structural and dynamic methods with an ensemble of plant-animal pollination networks. Specifically, we relate the contribution of each species to the network’s overall nested architecture with that species’ survival in a dynamic simulation of the entire mutualistic community.

Results/Conclusions

First, we find a heterogeneous distribution of species contributions to network architecture across our empirical data set. Thus, while some species contribute very much to the nested structure of the ensemble, others contribute comparatively little. Second, we find that species' contributions to the collective architecture and robustness are inversely associated with their survival. That is, the highest contributors have the lowest probability of survival and vice versa. As a consequence, species in mutualistic networks may face an important trade-off between contributing more to increase the persistence of the network or contributing less to increase their own survival probability.

Our results illustrate how the tragedy of the commons is manifested in heterogeneous networks of mutual cooperation. They also describe how information from just the network of interactions can be used to estimate the expected survival of component species. We illustrate how this can provide a quantitative assessment of the likely persistence of an invasive species and its effects on the overall persistence of the network. Finally, we confirm the generality of these results by studying socio-economic cooperative networks for which there are temporal information to estimate the extinction risk of a node.