Predicting recoverability of collapsed food webs through perturbation and dimension reduction

Biodiversity collapse, driven by increasing environmental changes, poses significant threats to ecosystem stability and the provision of essential ecosystem services. Understanding the recoverability of collapsed food webs thus is crucial for devising effective conservation strategies. This study delves into the theoretical exploration of the recoverability of food webs from a collapsed state. Through simple tools like dimension reduction, propagation of species-specific perturbation, and dynamical simulations, we explore whether simple tri-trophic food webs can be recovered from a collapsed state. Our study examines in detail the topological features of food webs that could either facilitate or impede their recovery. We demonstrate that the recoverability of complex food webs can be predicted by using a simple dimension-reduced model, with certain structural factors that could constrain the full recovery of collapsed food webs. Furthermore, we found that such a simple dimension-reduced model can accurately capture the rate of recovery for complex collapsed food webs. In addition, dynamic simulations highlighted the significance of topological features such as connectance and the number of predator links in determining recoverability. Our dimension-reduced modeling framework offers insights into the feasibility of restoring entire complex predator-prey networks through species-specific interventions. This study contributes to a deeper understanding of ecosystem resilience and aids in the development of targeted conservation strategies.