Direct manipulation of behaviour reveals a mechanism for variation in growth and mortality among prey populations

Understanding the relation between behavioural processes and their population- or community-level consequences is fundamental to developing a mechanistic understanding of ecosystems. Rarely are such links made, particularly outside the laboratory, and rarely is behaviour manipulated directly through genetic or hormonal manipulation. The area that provides the greatest promise is the relation between foraging behaviour and its subsequent growth and survival consequences. In whole-lake experiments, we used domestic and wild strains of rainbow trout, Oncorhynchus mykiss, differing in intrinsic activity and antipredator behaviours in the laboratory, to determine whether differences in activity and habitat use between strains exist in the field and could provide a behavioural mechanism for previous studies showing a trade-off between growth and survival. Domestic trout made up the majority of gill net catches (greater overall activity), particularly in deep and pelagic habitats that had abundant food but offered little or no physical habitat structure to escape from predation. Thus, greater activity rates (inferred) and use of risky habitats (directly observed) by domestic trout provides a behaviourally mediated mechanism for previous large-scale studies showing trade-offs between growth and mortality.