Structural forecasting of species persistence under changing environments

The persistence of a species in a given place not only depends on its intrinsic capacity to consume and transform resources into offspring, but also on how changing environmental conditions affect its growth rate. However, the complexity of factors has typically taken us to choose between understanding and predicting the persistence of species. To tackle this limitation, we propose a probabilistic approach rooted on the statistical concepts of ensemble theory applied to statistical mechanics and on the mathematical concepts of structural stability applied to population dynamics models – what we call structural forecasting. We show how this new approach allows us to estimate a probability of persistence for single species in local communities; to understand and interpret this probability conditional on the information we have concerning a system; and to provide out‐of‐sample predictions of species persistence as good as the best experimental approaches without the need of extensive amounts of data. We propose a probabilistic approach rooted on the statistical concepts of ensemble theory applied to statistical mechanics and on the mathematical concepts of structural stability applied to population dynamics models – what we call structural forecasting. We show how this new approach allows us to estimate a probability of persistence for single species in local communities, to understand and interpret this probability conditional on the information we have concerning a system, and to provide out‐of‐sample predictions of species persistence as good as the best experimental approaches without the need of extensive amounts of data.