Detecting the most probable transition phenomenon in a nutrient-phytoplankton-zooplankton system

Transitions between metastable states, induced by the interaction between uncertainty and nonlinearity, play an important role in understanding ecological systems. In this study, we investigate a three-dimensional nutrient-phytoplankton-zooplankton (NPZ) model subject to random perturbations. For an appropriate set of parameters, the NPZ model without random perturbations exhibits bistability, characterized by a stable equilibrium and a stable limit cycle. We show that, in the presence of random perturbations, the NPZ model may undergo noise-induced transitions between these two states. By combining the Onsager-Machlup action functional with the neural shooting method, we identify the most probable transition time, the most probable transition pathway and the most probable transition probability of the system. Our work provides a framework for predicting regime shifts in plankton communities and contributes to a deeper understanding of noise-induced transitions in complex biological systems. This article is part of the theme issue 'Critical transitions and intelligent control in complex systems'.