Modeling and protecting global ecological networks

With the rapid urbanization worldwide and ever-increasing impacts of human activities since at least 200 years ago, we are now facing a harsh situation of our biosphere. Building a global-level network model on ecological systems is of great importance, which would be able to provide us predictive and quantitative responses to human activities, leading to viable suggestions to policymakers. In this paper, we propose a multi-layer model for the global ecological network, where a number of local networks are connected via long-range interactions associated with migrant species, which can be induced by human activities or natural migration of wildlife, and each local network is generated by a trophic-level-based stochastic model. Predator–prey dynamics is described by a networked Lotka–Volterra model that accounts for the self-suppression effects on basal species, and the negative feedback loops. Impacts of human activities are modeled by investigating the quantitative changes of biodiversity under certain protecting strategies. We reveal that the global ecological network is organized in a clustered small-world manner, with in-degree distribution more heterogeneous than out-degree distribution. Protecting endangered species, popular preys and predicted-to-be-extinct species is more effective than randomly selected species or influential predators. Protecting after entering the fast extinction stage is more effective than at the beginning for some high trophic level species.