Food webs, phylogeny, social networks, and statistics: What is the common thread?

Abstract:

A food web is a network of nodes called trophic species. The role of each node as predator or prey determines the trophic or feeding relations that weave the web. Insight into the nature and degree of dependence among nodes has ecological implications, and can facilitate discoveries about trophic levels, stability of the ecosystem under species extinction, etc. In a recent paper, we adapted a novel statistical methodology called latent space modelling - originally developed to study social networks - to yield comprehensive understanding of food web structure and its contributing factors. This unified modelling framework expresses network links as the random response of within- and internode characteristics (predictors). Specifically, phylogeny is shown to have nontrivial influence on trophic relations in many webs, and for each web trophic clustering based on feeding activity and on feeding preference can differ substantially. This talk presents a brief history of latent space modelling and selected results from our recent paper and a related working paper.