Evolutionary Dispersal of Ecological Species via Multi-Agent Deep Reinforcement Learning

TL;DR

This paper employs multi-agent deep reinforcement learning with starvation-driven diffusion models to simulate and analyze species dispersal strategies in heterogeneous environments, offering new insights into ecological migration mechanisms.

Contribution

It introduces a novel application of MARL with DQN to simulate species dispersal incorporating SDD models, advancing understanding of evolutionary dispersal strategies.

Findings

Revealed emergent dispersal strategies through simulations

Validated traditional mathematical models of species dispersal

Demonstrated advantages of SDD models for species survival

Abstract

Understanding species dynamics in heterogeneous environments is essential for ecosystem studies. Traditional models assumed homogeneous habitats, but recent approaches include spatial and temporal variability, highlighting species migration. We adopt starvation-driven diffusion (SDD) models as nonlinear diffusion to describe species dispersal based on local resource conditions, showing advantages for species survival. However, accurate prediction remains challenging due to model simplifications. This study uses multi-agent reinforcement learning (MARL) with deep Q-networks (DQN) to simulate single species and predator-prey interactions, incorporating SDD-type rewards. Our simulations reveal evolutionary dispersal strategies, providing insights into species dispersal mechanisms and validating traditional mathematical models.