When experimenting with a large pool of species, a common problem is to determine a priori whether a certain subset of species can coexist when co-cultured. We propose a simple statistical model in which a number of species assemblages are observed, and the coexistence of novel assemblages is predicted out-of-fit. We discuss how this problem connects with popular models for population dynamics, and how the approach can be extended to the case in which only relative abundances are observed, or data are not sufficient to parameterize the full model.

Recording