Biotypes of the seed beetle Callosobruchus maculatus have differing effects on the germination and growth of their legume hosts

• 1 Populations of the seed beetle Callosobruchus maculatus (F.) exhibit considerable differences in body size and larval behaviour. We examined whether such variation modifies the relationship between beetle infestation and host plant performance.
• 2 Larvae from African and Asian biotypes were reared in seeds of four hosts that represented an almost four‐fold variation in seed mass. We estimated mass lost to larval consumption, and compared germination rates and seedling growth between infested and control seeds.
• 3 In seeds bearing a single larva, the larger‐bodied, contest‐competing larvae of the Asian biotype caused a 38–47% greater reduction in seed mass compared with the smaller‐bodied, scramble‐competing larvae of the African biotype. The amount of seed mass lost per larva remained similar in seeds with one or two scramble‐competing larvae but decreased significantly in seeds bearing two contest‐competing Asian larvae.
• 4 Differences in larval consumption and behaviour produced striking differences in the frequency of germination. Germination of singly‐infested mung bean (i.e. the smallest host) was 71% for African‐infested seeds versus 11% for Asian‐infested seeds. In cowpea (i.e. the largest host), 76% of Asian‐infested seeds germinated, whereas the germination rate of African‐infested cowpeas (92%) was similar to that of uninfested seeds.
• 5 Effects of beetle origin persisted after germination. Seedlings derived from Asian‐infested seeds had greater cotyledon damage 7 days after germination, and displayed lower height and less biomass 15 days after germination. Cotyledon damage was a good predictor of seedling performance (i.e. better than seed mass consumed) 15 days after germination.
• 6 Previous studies have suggested that population differences in larval size and burrowing behaviour (‘centripetal tendency’) reflect adaptation to different‐sized seeds. The present study demonstrates that these differences in turn influence the impact of larval feeding on host viability. Strong biotypic variation makes it difficult to generalize about pest impacts at the level of pest species.