Ecological theory predicts that keystone species have greater impacts on food webs than would be expected based on relative abundance. Along these lines, examples of mutualism and facilitation with dramatic impacts on ecosystems are increasingly being referred to as “keystone interactions”. However, despite the popularity of this term, a robust definition or model of a keystone interaction has not been established. We propose a simple model that could be applied to food webs with putative keystone interactions. We evaluated our model with two classes of positive interactions which have demonstrable and dramatic impacts in multiple terrestrial ecosystems: (1) Ant-sap-feeder interactions, which drive suppression of herbivores and initiate trophic cascades. (2) Vector borne plant pathogens, which restructure community-wide plant herbivore interactions and impact ecosystem-wide plant productivity.
Results/Conclusions
We conclude in our models that keystone interactions are a special class of “interaction modifier”, in which their presence modulates consumptive interactions more frequently than would be assumed based on occurrence in a larger food web. This model provides an analogue for the abundance-impact relationship used to define classical studies in keystone species, and should provide a useful quantitative criteria for evaluating future examples of keystone interactions.