Robustness through regime flips in collapsing ecological networks

Abstract

There has been considerable progress in our perception of organized complexity in recent years. Recurrent debates on the dynamics and stability of complex systems have provided several insights, but it is very difficult to find identifiable patterns in the relationship between complex network structure and dynamics. Traditionally an arena for theoreticians, much of this research has been invigorated by demonstration of alternate stable states in real world ecosystems such as lakes, coral reefs, forests and grasslands. In this work, we use topological connectivity attributes of eighty six ecological networks and link these with random and targeted perturbations, to obtain general patterns of behaviour of complex real world systems. We have analyzed the response of each ecological network to individual, grouped and cascading extinctions, and the results suggest that most networks are robust to loss of specialists until specific thresholds are reached in terms of network geodesics. If the extinctions persist beyond these thresholds, a state change or ‘flip’ occurs and the structural properties are altered drastically, although the network does not collapse. As opposed to simpler or smaller networks, we find larger networks to contain multiple states that may in turn, ensure long-term persistence, suggesting that complexity can endow resilience to ecosystems. The concept of critical transitions in ecological networks and the implications of these findings for complex systems characterized by networks are likely to be profound with immediate significance for ecosystem conservation, invasion biology and restoration ecology.