"Revisiting the link-species scaling relationship in ecological networks"
Timothée Poisot, University of Montreal
Tuesday February 25, 12-1pm
Montreal Neurological Institute, deGrandpre Communications Centre
Abstract: Community ecologists are fascinated by counting things. It is therefore no surprise that the early food web literature paid so much attention to counting species, counting trophic interactions, and computing the relationship between these numbers. More species always means more interactions; this scaling between species richness and number of interactions is universal. In fact, this scaling underlie most means of describing a food web. Accurately predicting both is essential when trying to model food web structure. The number of links in a food web does not simply scale with the number of species: it also must obey constraints fixed by biology. These constraints determine both the maximum and minimum number of links, but have not been used in previous attempts to model the relationship between links and richness. Using probabilistic programing, we show how the relationship between interactions and richness boils down to a single parameter, which can in turn be used in generative models of ecological networks.