Ever since the theoretical work of May (1972) that showed highly complex food webs to be unstable, ecologists have been motivated to determine which food web components most affect its stability and how. Theoretical work since May has revealed two main points: the types of species interactions and the strengths of these interactions greatly affect food web stability. Theoretical evidence suggests that complex food webs are stable when the majority of interaction strengths are skewed weak, especially in the case of omnivory. Furthermore, theory predicts that weak interactions can even stabilize strong negative interactions by dampening their oscillatory population dynamics. To help bridge this gap between theory and empiricism, we experimentally tested the long-term stabilizing effect of weak interactions coupled with strong interactions in a system with two omnivorous predators, two prey, and a basal resource. Using a protist microcosm system, we assembled a food web of four protist species and three bacterial species encompassing three trophic levels. We crossed 13 food web configurations with three productivity levels and replicated each treatment four times. Lastly, we quantified the protist species’ dynamics over the course of 90 days, encompassing between 90 and 270 generations depending on the species’ life history traits.
Results/Conclusions
We show that the coupling of strong interactions with weak interactions stabilizes food web dynamics and that this result holds true across a productivity gradient. Strong interactions typically cause complex oscillatory population dynamics which often lead to the extinction of one or more species. Weak interactions, however, transfer energy away from the strong interaction and act to dampen these unstable dynamics. Our results show that strong pairwise interactions, in which the predator drives the prey extinct, are stabilized by weak pairwise interactions, in which the predator coexists with the prey. The result of coupling these interactions leads to coexistence between the predator and both prey, as predicted by theory. Furthermore, these effects are consistent across a productivity level, indicated by time of persistence, which has yet to be shown to our knowledge. A challenging yet exciting avenue for ecologists is trying to capture the complexity of the natural world in mathematical models that are also empirically relevant. The results presented here show that even the simplest of theories can predict the behavior of a real biological system.