Abstract

Resource defence theory predicts that defence and monopolization increase as the spatial clumping, spatial predictability, and temporal predictability of resources increase and as temporal clumping of resources decrease. Previous tests of resource defence theory have manipulated only one aspect of resource distribution at a time. I tested whether there was an interaction between the effects of the spatial and temporal clumping of food by allowing groups of ten Japanese medaka (Oryzias latipes) to compete, over a 4 week period, for food in a 2 x 2 factorial design. As expected, the intensity of aggression increased as the spatial clumping of food increased and as the temporal clumping of food decreased. However, there was also a significant interaction between the effects of the spatial and temporal clumping of food on the intensity of aggression and total aggression. The number of fish in the feeding patch, a measure of monopolization, decreased as the spatial clumping of food increased, but was not affected by the temporal clumping of food. Although measures of aggression were high in the economically defendable trials, there was no apparent cost to resource defence, since mean growth rates did not differ significantly among treatments. Growth depensation increased over time, but was not influenced by the temporal or spatial distribution of food. The power of my experiment to detect a significant effect of food distribution on growth depensation was low. Therefore, although I failed to reject the null hypothesis, I cannot confidently conclude that the spatial and temporal clumping of food have no effect on the magnitude of growth depensation. My results suggest that care must be taken when extrapolating the results of single factor experiments to multi-factor or field situations.