Impact of the mesoscale structure of a bipartite ecological interaction network on its robustness through a probabilistic modeling

The robustness of an ecological network quantifies the resilience ecosystem it represents to species loss. It corresponds proportion that are disconnected from rest when extinctions occur sequentially. Classically, is calculated for a given network, simulation large number extinction sequences. link between structure and remains open question. Setting joint probabilistic model on sequences allows analysis this relation. Bipartite stochastic block models have proven their ability bipartite networks e.g. plant-pollinator networks: divided into blocks interaction probabilities determined by membership. Analytical expressions expectation variance obtained under model, different distributions primary impact analyzed through set properties numerical illustrations. collection us compare empirical approach our approach, illustrates relevance latter comes computing partially observed or incompletely sampled network.