A test of the ideal free distribution with unequal competitors

Summary The distribution of goldfish differing in competitive status approximately mimicked that expected for the ideal free distribution but with slightly too many fish in the poorer site. As predicted by phenotype-limited game theory models of dispersion, the mean rank of fish in a site varied inversely with the number of fish, both in the high and low input sites, and there was no correlation between competitive rank and time spent in the better site. The intake was higher for each individual in the high input site than in the low input site, showing that the distribution was not an exact evolutionarily stable strategy. We suggest that the deviation is due to sampling or perceptual constraints. Analysis of other studies with continous input shows that the discrepancy from the theoretical expectation increases with input ratio.