How should we test for the role of behaviour in population dynamics?

The behaviour of an individual affects the probability that it will find food or a mate, and whether it will avoid becoming food itself. Because the birth and death of individuals are the constituents of the birth and death rates of populations, it seems likely that population dynamics is affected by variation in the behaviour of individuals. While intuitively appealing, there are few data to support this contention. I suggest that this lack of support stems from the failure to make use of manipulative experiments to test the hypothesis. Manipulative experiments are not the only approach to testing these hypotheses, but this powerful tool has not been used as effectively as it might. The behaviour of an animal can be manipulated with information-carrying chemicals in its environment, and pharmacologically. Variation in behaviour among individuals that is genetically based can also provide experimental material to test hypotheses about the role of behaviour in population dynamics. Manipulative experiments have the advantages of increased statistical power, and the elimination of unmeasured covariates. They have the disadvantage that they can introduce artifacts into the study system. It seems unlikely, however, that different kinds of manipulations would produce the same kinds of artifacts.