EVOLUTION OF SOCIAL BEHAVIOR IN A RESOURCE-RICH, STRUCTURED ENVIRONMENT: SELECTION EXPERIMENTS WITH MEDAKA (ORYZIAS LATIPES).

Abstract

Schooling and aggression in fish are known to be partially inherited traits. Their genetic relationship to growth rate and to each other is, however, not fully understood. In this study we present evidence that schooling, social tolerance, and agonistic behavior in medaka (Oryzias latipes) are altered as an indirect result of selection on growth in two environments that differed in the intensity of social interactions required to obtain access to food. In the high interaction environment food was provided to excess inside a floating cork ring, which limited access to the food and allowed fish to attempt to monopolize it. In the low interaction environment the same amount of food was spread over the container's surface. After two generations of selection on growth the correlation of agonistic behavior and mean growth of broods was negative in the line selected for fast growth, when selection took place in a high interaction environment, but not when it took place in a low interaction environment. School cohesion was higher in the lines selected for fast growth than in those selected for slow growth when selection and rearing environments were both the same, either high or low interaction, but not when they were different. The correlation of social tolerance with growth was significantly more positive in the line selected for fast growth than in that selected for slow growth, but only when selection took place under high social interaction. It appears from these experiments that when resources are aggregated, but unlimited in quantity, competition favors individuals that avoid wasting time and energy on unnecessary and ultimately futile attempts to monopolize food and that also exhibit higher tolerance of nearby conspecifics. The results are interpreted in terms of a hypothesized stimulus-response threshold level for agonistic responses to conspecifics. We suggest that this threshold, which is altered by selection on growth, could provide a common causal (genetic) link between growth and the observed aspects of social behavior. By combining the probabilities from the separate behavioral experiments to obtain an overall significance of our hypothesis we conclude that the probability of no change in threshold is in fact low (P < 0.01).