Population dynamics and biological control

Abstract Using simple models for host-parasitoid and host-pathogen interactions, we present a basic framework for examining the outcome of releasing natural enemies against a target pest population in a classical biological control programme. In particular, we examine the conditions for the initial invasion and establishment of a natural enemy species, for the maximum depression of the host population, and for the persistence of the populations in a stable interaction. In these conditions there are close parallels between parasitoids and pathogens. The practice of augmenting an existing natural enemy population by regular mass releases has been widely practised, especially with parasitoids. The conditions for eradication of the pest are very similar in host-parasitoid and host-pathogen models, namely that releases must be greater than the equilibrium production of natural enemies in the absence of releases. Any additional density dependence acting on the host population after the stage attacked by parasitoids can influence the effectiveness of augmentative releases. This is particularly the case with over-compensating density dependence when additional releases can actually lead to an increase in the host population. A theoretical basis for biological control cannot be properly developed simply by considering the dynamics of releasing single natural enemy species. Biological control often involves the interplay among different types of natural enemies affecting the same host population. As a step in the direction of producing more complex, multispecies models, we examine the dynamics of three situations: (1) where the host is attacked by two parasitoid species; (2) by a generalist predator and a specialist parasitoid; or (3) by a parasitoid and a pathogen. The dynamics of these three-species systems can be complex, and with properties not easily foreseen from the separate pairwise interactions. These results caution us against formulating biological control strategies purely in terms of two-species systems. For the main part we examine host-parasitoid interactions with discrete, synchronized generations. These would appear to be less suitable to tropical insects where continuous generations and life cycles of the host and parasitoid of different length are to be expected. We show, however, that cycles (with a period of one host generation) can be obtained from an age-structured simulation model, and that these are promoted by the parasitoids having a life cycle half as long as that of the host. Some implications for biological control are discussed. Finally, we turn briefly to the dynamics of host-parasitoid and host-pathogen interactions where pesticides are also applied, and we discuss the evolution of pesticide resistance within the context of these models.