Development of Optimal Feedback for Zooplankton Seasonal Diel Vertical Migration

The main objective of the research is to develop the optimal feedback for zooplankton diel vertical migrations, taking into account seasonal variations. Feedback mechanisms serve to generate zooplankton movements in response to changes in an environment. Modern approaches to modeling evolutionary stable behavior are based on the involvement of the fitness function that reflect competitive advantages of organisms. In order to model zooplankton behavior, it is crucial to formalize the fitness function and find the feedback settings that maximize this function. To achieve this, the neural network is trained using a dataset encompassing several seasons. The Survival of the Fittest algorithm (SoFa) is utilized as the training method, aiming to maximize the fitness function. The trained network generates migration strategies in different seasons, which are confirmed by field observations. The developed feedback represents an effective tool for studying the diel vertical migration of zooplankton considering seasonal variations.