Short-term interactions of the polyphagous tachinid Winthemia rufopicta (Big.) with its hosts, especially Heliothis zea (Boddie) and H. virescens (Fab.), are considered.
Within a given host species, interacting factors fall into three groups: (1) factors that have little influence on parasite oviposition, or cause little parasite mortality (host responses to ovipositing parasites, previous parasitism, firmness of attachment of eggs and, in the populations studied here, direct competition with other insect parasites); (2) factors that may account for significant limitation of oviposition (habitat, diel activity, and seasonal occurrence of a given host, and species preference of the parasite) or that cause significant parasite mortality (destruction by the host and other mortality of hatching maggots, which in turn depends partly on the position of eggs on the host): and (3) factors that can vary widely in a given host species and are therefore of great potential importance. These are: (a) host feeding behaviour—initial parasitism of H. zea, for example, varies from negligible levels for larvae concealed in corn ears to sometimes very high levels for larvae feeding openly on tobacco leaves; (b) host size—smaller larvae (especially penultimate compared with final instars) receive fewer ovipositions, and fewer eggs per oviposition; (c) host moulting—surface-deposited eggs are lost if the host moults before they hatch, but the rate of moulting varies with instar and foodplant, and interacts differently with parasite egg incubation times at different temperatures (most eggs are lost except in final instars); (d) intraspecific competition—depends on parasite and host densities, but reaches the highest levels in the fall; (e) competition with host pathogens—parasites may be killed during development if the host dies from pathogens.
For different hosts, family, size, surface quality, and spatial and temporal availability are known to have divergent influences on the levels of parasitism, but many other interactions are probably important.
The interpretation of records of total or of successful parasitism in host samples is considered—behavioural, developmental, and numerical interactions may make conclusions drawn from such records unreliable. In particular, levels of parasitism may change very rapidly as host or parasite populations develop; laboratory rearing may distort recorded levels of successful parasitism; and the proportion of attacked hosts that produce parasites, and the number of parasites produced by each host, normally varies directly with the number of eggs per host even when this is large.