REGULATION OF APHID POPULATIONS BY APHIDIID WASPS - DOES PARASITOID FORAGING BEHAVIOR OR HYPERPARASITISM LIMIT IMPACT

Aphidiid parasitoids (Hymenoptera: Aphidiidae) of aphids generally exploit only a small percentage of the available host resources in the field. This limited impact on aphid populations has often been explained as a consequence of hyperparasitism. We propose that a wasp's reproductive strategy, as opposed to hyperparasitism, is the dominant factor in aphidiid population dynamics. A wasp's foraging efficiency and oviposition decisions are influenced by several variables, including searching behaviour between and within patches, host choice (as modified by the aphids' defensive behaviours), and plant structural complexity. Two broadly different patterns of host exploitation have evolved in aphidiid wasps in relation to ant-aphid mutualism. Firstly, in species that are exposed to predation and hyperparasitism, a female may leave a patch before all suitable hosts are parasitized. Because predators and hyperparasitoids tend to aggregate at high aphid or aphidiid densities, or in response to aphid honeydew, this strategy enables females to reduce offspring mortality by ''spreading the risk'' over several host patches. Secondly, in species that have evolved mechanisms to avoid aggression by mutualistic ants, females are able to exploit a hyperparasitoid-free resource space. Such species may concentrate their eggs in only a few aphid colonies, which are thus heavily exploited. Although hyperparasitism of species in the first group tends to reach high levels, its overall impact on aphid-aphidiid population dynamics is probably limited by the low average fecundity of most hyperparasitoids. We discuss the foraging patterns of aphidiid wasps in relation to aphid population regulation in general, and to classical biological control in particular. We argue that a parasitoid's potential to regulate the host population is largely determined by its foraging strategy. In an exotic parasitoid, a behavioural syndrome that has evolved and presumably is adaptive in a more diverse (native) environment may, in a more uniform (managed) environment, result in suboptimal patch-leaving and oviposition decisions, and possibly increased resource usage.