Dynamic effects of parasitism by an endoparasitoid wasp on the development of two host species: implications for host quality and parasitoid fitness

Summary 1. The study reported here examined growth and developmental interactions between the gregarious larval koinobiont endoparasitoid Cotesia glomerata (Hymenoptera: Braconidae) and two of its hosts that vary considerably in growth potential: Pieris rapae and the larger P. brassicae (Lepidoptera: Pieridae). At pupation, healthy larvae of P. brassicae are over twice as large, in terms of fresh body mass, as those of P. rapae . 2. Clutch size of C. glomerata was manipulated artificially, and the relationship between parasitoid burden and the maximum weight of the parasitised host (= host–parasitoid complex) was measured. In both hosts, the maximum complex weight was correlated positively with parasitoid burden. Compared with unparasitised hosts, however, the growth of P. rapae was increased at significantly lower parasitoid burdens than in P. brassicae . Emerging wasp size was correlated negatively with parasitoid burden in both host species, whereas development time was less affected. 3. After larval parasitoid egress, the weight of the host carcass increased slightly, but not significantly, with parasitoid burden, although there was a strong correlation between the proportion of host mass consumed by C. glomerata larvae during development and parasitoid burden. 4. Clutch size was generally correlated positively with instar parasitised in both hosts, and greater in P. brassicae than in P. rapae . Sex ratios were much more female biased in L1 and L2 P. rapae than in all other host classes. Adult parasitoid size was correlated inversely with host instar at parasitism, and wasps emerging from P. brassicae were larger, and completed development faster, than conspecifics emerging from P. rapae . 5. The data reveal that parasitism by C. glomerata has profound species‐specific effects on the growth of both host species. Consequently, optimality models in which host quality is often based on host size at parasitism or unparasitised growth potential may have little utility in describing the development of gregarious koinobiont endoparasitoids. The results of this investigation are discussed in relation to the potential effectiveness of gregarious koinobionts in biological control programmes.