Invertebrate predation of 15N-marked prey in semi-field wheat enclosures

One of the most famous examples of successful, classical biological control in Japan is the introduction of the parasitoids Coccobius fulvus and Aphytis yanonensis against the citrus pest arrowhead scale Unaspis yanonensis. Together, they comprise a host‐parasitoid system that has been demonstrated to be stable. To test the conventional theory that successful biological control of pests occurs through the establishment of a low stable equilibrium, brought about by the density‐dependent responses of natural enemies to the pest species, sampling was carried out at five sites in the field during 2000 and 2001 to examine the relationship between the rate of parasitism by C. fulvus and the density of its host. The data were analysed using three statistical techniques at nine spatial scales. Contrary to conventional theoretical predictions, each method of analysis detected very little density‐dependence at any spatial level in this study. Parasitoid aggregations independent of host density were not sufficient to stabilise host–parasitoid interactions. Our results suggest that neither spatial density‐dependent nor density‐independent parasitism is necessary for successful biological control, or for the stability of the host–parasitoid system. We propose an alternative mechanism: a spatial refuge induced by parasitoid introduction may stabilise a system.