Résumé
The evolution of resistance by insect and weed pests to chemical pesticides is a problem of increasing importance in applied ecology. It is striking that the evolution of resistance by target pest species in biological control is much less frequently reported, particularly in control involving parasitoids and predators, rather than pathogens. Although it is conceivable that this reflects biases in reporting or frequency of application, we suggest that there is a puzzle here worthy of scrutiny, and we outline several potential underlying causes. In order of discussion (not necessarily of importance), these are: (1) lack of genetic variation; (2) genetic constraints on selection; (3) weak selection; (4) temporally varying selection; and (5) coevolutionary dynamics. We, in particular, focus on the potential for weak selection on the host for increased resistance, despite effective control. The very spatial mechanisms (e.g., refuges, metapopulation dynamics) believed to facilitate the persistence of many natural enemy-victim systems with strong biological control may also incidentally provide an environment where selection is weak on target pests to evolve improved resistance to control agents, thereby biasing coevolution toward the enemy. The basic insight is that in a spatially heterogeneous environment, a strong limiting factor on a population can be a weak selective factor. The hypotheses presented here provide ingredients needed to predict which biological control systems might be evolutionarily stable, and which not. Our aim in this thought piece is to stimulate more attention to the evolutionary dimension of biological control systems.