Omnivory

Omnivory is the consumption of resources from more than one trophic level by a single species. An omnivore feeds on both plants and animals, or on prey and detritus, or on primary producers and herbivores. Omnivory blurs the clean stratification of the trophic pyramid and introduces feedback loops that can either stabilize or destabilize food web dynamics. High omnivory increases connectance and can provide trophic redundancy — alternative pathways for energy flow when one prey species declines. But omnivory also complicates predictions: a species that preys on both herbivores and plants can produce counterintuitive population dynamics, including the possibility that increasing predator density decreases plant damage. The study of omnivory is essentially the study of why real ecosystems refuse to behave like textbook food chains.

Omnivory as a Network Property

From the perspective of network ecology, omnivory is not merely a dietary habit. It is a topological feature of food webs that determines how perturbations propagate through the network. An omnivorous species creates additional edges in the food web, increasing connectance and creating alternative pathways for energy flow. These alternative pathways are a form of trophic redundancy: if one prey species declines, the omnivore can switch to another, preventing the cascade failure that would occur in a specialized system.

But omnivory is a double-edged network feature. While it provides redundancy, it also creates trophic incoherence — the blurring of distinct trophic levels that makes the system harder to predict and model. A species that feeds at multiple levels can produce counterintuitive dynamics: increasing the predator's density may decrease herbivore pressure on plants if the predator shifts to consuming more plants and fewer herbivores. These feedback reversals are impossible in a linear food chain but common in webs with high omnivory.

The systems-theoretic significance of omnivory is that it transforms a hierarchical network into a heterarchical one. In a strict hierarchy — a chain or a layered web — control flows in one direction, and perturbations propagate predictably. In a heterarchy with omnivory, control is distributed, and perturbations can loop back through unexpected pathways. This is why ecosystems with high omnivory are often more resilient to species loss but more vulnerable to invasive species: the same network flexibility that absorbs internal perturbations can also accommodate novel interactions that restructure the web.

Omnivory and the Stability-Complexity Debate

Robert May's 1972 proof that random networks with high diversity and connectance are unstable posed a paradox: real ecosystems are diverse and connected, yet they persist. One resolution to this paradox is that real food webs are not random — and omnivory is one of the non-random features that may stabilize them. Omnivory reduces the effective dimensionality of the interaction matrix by creating correlated responses: when an omnivore switches prey, it simultaneously reduces pressure on one species and increases pressure on another, producing a damping effect that can stabilize the network.

However, the stabilizing effect of omnivory depends on its distribution. A few highly omnivorous generalists may stabilize a web by providing flexible control at key nodes. But ubiquitous omnivory — every species feeding at every level — can produce the very instability that May predicted. The optimal level of omnivory is therefore a network design problem: enough to provide redundancy, not so much as to eliminate the trophic structure that gives the web its dynamical coherence.

Omnivory is the food web's confession that hierarchy is a fiction. The moment a species refuses its assigned trophic level, the chain becomes a web, and the web becomes a network — with all the complexity, resilience, and unpredictability that networks entail.