Abstract

The network approach to the long-drawn but increasingly society-relevant ecosystem-complexity-stability debate started by Robert May is a sizzling hot topic today, with three major papers in Science, Nature and Ecology Letters within the past month. A chronic symptom identified by these papers is that most preceding research has examined ecological networks as entities frozen in time, as subgraphs of a single type of species interaction, as collections of binary switches, or by extrapolating from observations of small subsets of species. These drawbacks have often been overcome in isolation but rarely all at the same time. I will present a theoretical model that brings almost everything together yet keeps it simple. We will hear that this paradox is not a paradox and how and why the model differs from apparently similar products in the market. Controversially, it does not ‘predict’. I will explore how such a model could enrich a culture in which people justifiably want results of immediate ‘applicability to the real world’, and discuss how my particular formulation may help answer important questions.