How far is too far when perturbing a food web?

Background/Question/Methods

Anthropogenic impacts on natural systems are steadily increasing, as are conservation and management activities attempting to alleviate their consequences. For these activities to be effective, they require an understanding of which ecosystems and species are most at risk or which perturbations are likely to cause the largest impacts. From a food-web perspective, most existing research has focused on whether a food web is linearly stable and will return to its equilibrium point after a perturbation. However, this approach lacks an understanding of how a food web’s short-term behavior after perturbation relates to its long-term stability and how its vulnerability to pulse perturbations is affected by previous press perturbations.

Here, we investigate how the stability, resilience, and resistance of simulated food webs vary as a function of food-web properties and the properties of species being perturbed. In particular, we manipulated the growth rate of each individual species in the webs---to mimic an external perturbation such as over-harvesting---and identified the points at which (i) the focal species went extinct, (ii) another species went extinct, or (iii) the network was pushed beyond its original equilibrium. We then systematically studied how the web’s structure and dynamic behavior were altered because of these external perturbations.

Results/Conclusions

Food web attributes, such as species richness and the density of links between species, are known to affect a food web’s resilience and the likelihood of the network being linearly stable. Here, we have also found that these properties are intimately related to a network’s reactivity---a measure of the amplitude of a food web’s short-term response to a pulse perturbation---and the level of disturbance a food web can sustain before a species is driven extinct. More importantly, we have found that these different measures of stability rarely show the same pattern across different food webs, so that a food web that may be considered “stable” by one perspective may be “unstable” from another. Moreover, the most resilient networks are not necessarily able to sustain the largest perturbations nor do they need to be exclusively composed of species that are the least vulnerable to extinction. In addition, when a network experiences a long-term press perturbation, such as harvesting, and moves to a new equilibrium, its dynamic behavior can fundamentally change, changing its susceptibility  to future short-term pulse disturbances such as drought or disease. Thus, while some networks become less stable as a result of long-term disturbances, others may increase in stability.