Predator-prey studies in semi-arid eastern Australia demonstrated that populations of rabbits (Oryctolagus cuniculus) could be regulated by predators. The functional, numerical and total responses of foxes (Vulpes vulpes) to rabbits and the numerical response of feral cats (Felis catus) to rabbits, are described. Measurement of the rabbit component of foxes' stomach contents indicates a Type III functional response. The size of the fox population in summer was dependent on the availability of rabbits over the immediately preceding rabbit breeding season but there appeared to be no density-dependent aggregation of young foxes in areas of surplus food. The total response of foxes, estimated using the short-term numerical response of dispersing foxes, was directly density-dependent for low rabbit densities and inversely density-dependent for high rabbit densities. Two states are possible with this form of total response: a state with low rabbit densities regulated by predators and a state with high rabbit densities which occurs when rabbits escape predator regulation. The boundary between regulation and non-regulation by predators was demonstrated by a predator-removal experiment. In the treated areas, predators were initially culled and rabbits increased to higher densities than in an untreated area where predators were always present. When predators were allowed back into the treated areas, rabbit populations continued to increase and did not decline to the density in the untreated area. This is the critical evidence for a two-state system. When predators were present, rabbits could be maintained at low densities which were in the density-dependent part of the total response curve for foxes. Exceptionally high rabbit recruitment, or artificially reduced predation, could result in rabbits escaping predator-regulation. Under these circumstances, rabbits could move into the inversely density-dependent region of the total response curve for foxes.
