Trophic Network Complexity and Ecosystem Stability

Abstract

Trophic networks -- the feeding relationships that connect primary producers to consumers across multiple trophic levels-- determine how energy and nutrients flow through ecosystems and fundamentally constrain ecosystem stability, productivity, and resilience to perturbation. The classical May (1972) paradox -- that mathematical random food web models predict stability to decrease with complexity (more species and stronger interactions) while empirical observations suggest more complex ecosystems are often more stable -- has driven decades of food web ecology research seeking the structural features that allow complexity to confer stability in real ecosystems. This study constructed and analysed 42 empirical trophic networks from European freshwater, woodland, and grassland ecosystems (2,840 trophic interaction records; 284 species; 2019-2022) quantifying network complexity metrics (connectance, linkage density, trophic level diversity, omnivory, chain length) and ecosystem stability indicators (temporal biomass CV, resistance to simulated species loss, recovery rate after experimental perturbation). Network complexity was positively correlated with ecosystem stability across all ecosystem types (r = +0.74-0.84; p < 0.001) when weak interaction strength predominated -- confirming the McCann et al. (1998) prediction that weak interactions stabilise complex food webs. A Trophic Network Stability Index (TNSI) integrating connectance, omnivory index, weak:strong interaction ratio, and trophic diversity predicted observed ecosystem stability with r = +0.84 and AUC = 0.884, and identified trophic simplification from biodiversity loss as the primary mechanism by which species loss destabilises ecosystem function