We surveyed nine populations of the three-spined stickleback infected by the diphyllobothriidean cestode Schistocephalus solidus from south-central Alaska for two apparent forms of tolerance to infection in females capable of producing egg clutches notwithstanding large parasite burdens. Seven populations exhibited fecundity reduction, whereas two populations showed fecundity compensation. Our data suggest that fecundity reduction, a side effect resulting from nutrient theft, occurs in two phases of host response influenced by the parasite : host body mass (BM) ratio. The first is significantly reduced ovum mass without significant reduction in clutch size, and the second one involves significant reductions in both ovum mass and clutch size. Thus, ovum mass of host females who are functionally being starved through nutrient theft seems to be more readily influenced by parasitism and, therefore, decreased before clutch size is reduced. This inference is consistent with expectations based on the biology of and effect of feeding ration on reproduction in stickleback females. Fecundity compensation appears to be uncommon among populations of three-spined stickleback in Alaska and rare among populations throughout the northern hemisphere. Fecundity reduction seems to be common, at least among stickleback populations in Alaska.

The objective of this study was to investigate the means by which Schistocephalus solidus might reduce annual fecundity in female threespine stickleback fish (Gasterosteus aculeatus) through processes of oocyte development. Histological examinations of specimens from one lake in Alaska in 2000 and 2001 were used to analyse the effects of S. solidus on recruitment of primary growth oocytes into vitellogenesis, atresia of vitellogenic oocytes, and the interspawning interval. The ratio of primary growth to late secondary growth (late vitellogenic) oocytes was significantly greater (P<0·01) among infected fish than uninfected ones in early-season samples from 2000 and 2001, revealing a decrease in recruitment of oocytes from primary growth into vitellogenic oocytes among infected females. The difference was marginally non-significant (P=0·087) in a mid-season sample from 2001 due to reductions in the entire pool of vitellogenic (early and late secondary growth) oocytes recruited prior to the spawning season in this determinate-fecundity species. Atresia among all vitellogenic oocytes was low and did not differ between infected and uninfected females. Histological estimations of the interspawning interval using post-ovulatory follicles showed no significant differences between infected and uninfected fish, suggesting that the number of spawnings in stickleback females each spawning season is unaffected by S. solidus infection. Thus, annual fecundity appears to be reduced only through recruitment of oocytes into vitellogenesis.

Infection with developing metacestodes of the rat tapeworm, Hymenolepis diminuta, is known to retard the accumulation of the yolk protein, vitellin, in the terminal ovarian follicles of the intermediate host, Tenebrio molitor. It is probable that this is the result of competitive inhibition of juvenile hormone binding at a microsomal binding site in the beetle follicular epithelium. Experiments were designed to test the hypothesis that inhibitor molecules were circulating in the haemolymph of infected beetles. Whole haemolymph, collected from male or female beetles at various stages post-infection, was injected into non-infected female recipients 2 days post-emergence. Ovaries were removed 3 days later and the vitellin content of the same sized follicles measured using an ELISA. The vitellin content of follicles from recipients of haemolymph from females infected with metacestodes at stage 1 and stage 3–4 was significantly reduced (24 and 27·9%) compared to sham-infected females. However, haemolymph from females infected with mature metacestodes did not affect the vitellin content. Results were thus comparable to those obtained by monitoring ovarian vitellin levels in female T. molitor with bona fide infections. Haemolymph from infected males did not affect ovarian vitellin content. These results indicate that molecules that can modulate vitellogenesis may be present in the haemolymph of females infected with developing metacestodes but that these factors disappear later in infection.