COS 26-5
Quantifying the resilience of a Caribbean food web
Common ways of measuring the fragility of food webs is to examine their long-term stability in terms of species persistence or community dynamics. A food web is considered to be of high stability when most species survive in the long run. Such approaches overlook, however, another important element of community fragility: that of the existence of multiple alternative states. If a community converges to different endstates under the same conditions, it runs the risk of shifting from one to another even due to small disturbances. Here, we quantify the fragility of a Caribbean food web by mapping its potential endpoint communities. We simulated the dynamics of the species in the food web starting from different biomasses to identify alternative endpoint communities. We estimated the risk of shifting from the current state to any other alternative state by measuring the relative change in biomass needed to incur the shift. We used the magnitude of this change to quantify the resilience of the current state of the Caribbean food web
Results/Conclusions
We found that there are more than one endpoint communities in the Caribbean food web. These communities differed in their richness, composition and network properties. The risk of shifting to any of these alternative states from the current state of the Caribbean food web depended on which species we disturbed. We reported higher risks of shifting when disturbances were targeted at species belonging to high trophic levels, although shifting to an alternative attractor was more sensitive to multiple species than single species disturbances. Our results show that complex ecological communities may run the risk of shifting between alternative states even without experiencing radical changes in external environmental conditions. While our framework might offer a possible way of quantifying this risk, it also highlights an additional source of community fragility that needs to be considered in conservation management.