Nonequilibrium Theory for Adaptive Systems in Varying Environments (2506.19018v1)

Abstract: Biological organisms thrive by adapting to their environments, which are often unpredictably changeable. We apply recent results from nonequilibrium physics to show that organisms' fitness parses into a static generalist component and a nonequilibrium tracking component. Our findings: (1) Environmental changes that are too fast or too small are not worth tracking. (2) Well-timed anticipatory tracking enhances fitness in coherent environments. (3) We compute and explain the optimal adaptive strategy for a system in a given environment, such as bet hedging or phenotypic memory. Conversely, (4) We compute and explain the optimal way for an environment to control a given system, for example for designing drug regimens to limit the growth of pathogens. By connecting fitness, adaptive strategy, and environmental variability, this work provides the foundations for a generic physical theory of adaptivity.