In population modelling the functional forms of models are at least as important as are parameter values in expressing the underlying biology and in determining the outcome. Here the influence of differing biological assumptions as to predator interference on the forms of predator—prey models is discussed, and it is shown how this aspect of the underlying biology can influence the conclusions to be drawn from multispecies population models as to how "competition from natural predators" affects human harvesting of natural populations. The importance of these issues for the current controversy over the effects of marine mammals on fisheries is stressed.