Circumventing structural uncertainty: A Bayesian perspective on nonlinear forecasting for ecology

摘要: As a consequence of the complexity ecosystems and context-dependence species interactions, structural uncertainty is pervasive in ecological modeling. This particularly problematic when models are used to make conservation management plans whose outcomes may depend strongly on model formulation. Nonlinear time series approaches allow us circumvent this issue by using observed dynamics system guide policy development. However, these methods typically require long from stationary systems, which rarely available settings. Here we present Bayesian approach nonlinear forecasting based Gaussian processes that readily integrates information several short allows for nonstationary dynamics. We demonstrate utility our modeling simulated wide range scenarios. expect will extend systems can be usefully applied.