Lead


Lay summary
Dispersal has two major antagonistic effects for the evolution of altruistic cooperation and social behavior. On the one hand, migration by individuals from their native patch decreases relatedness among locally interacting individuals, thus selecting against altruistic cooperation. On the other hand, dispersal decreases local competition for resources and reproduction among kin, which should favor altruistic cooperation. So far, only a few studies have investigated the consequences of these opposing effects when dispersal and cooperation can co-evolve. These studies were based either on social network or game-theoretic models that lacked explicit space and did not consider factors such as overlapping generations (which can lead to competition between parents and offspring) and the ability of individuals to disperse conditionally depending on, for example, the local population density. Population density itself can be influenced by the pattern of dispersal and the level of local cooperation, hence generating a further feedback-loop. The consequences of such feedbacks on the evolutionary dynamics of altruism and dispersal remain unknown.The aim of this project is to develop an individual-based model and an embodied robotic system which explicitly incorporate space, continuous movement of individuals, conditional dispersal and overlapping generations. We will implement a structured environment inhabited by 100 individuals that can freely move between nine patches each containing food sources. We will let foraging strategies, dispersal, and the level of altruism evolve over many generations of experimental evolution. Both dispersal and altruistic cooperation are not encoded by specific genetic traits, but rather emerge from the individuals? behavior that is itself determined by artificial neural networks whose parameters can change by mutation and selection. With this experimental design, dispersal and altruism can --but need not-- be genetically linked. With this system we will address questions about the evolutionary conditions necessary for the emergence, maintenance and disappearance of altruistic cooperation and dispersal in evolving populations. First, we will investigate how variation in the costs/benefits of cooperation, the costs of dispersal, and the level of competition affect the co-evolution of altruistic cooperation and dispersal. Second, we will study the role of condition-dependent dispersal strategies that can be influenced by resource availability and local population density. Third, we will investigate the influence of overlapping generations where parents compete with their offspring. Finally, we will conduct parts of these experiments using a robotic platform to investigate to what extent issues related to embodiment and physical interactions influence the results obtained in numerical and theoretical models. The use of a robotic system also has the advantage of making the studies accessible to a broader audience, including the media which can use videos that illustrate the key results of the scientific research.