Roughly 80% of animal species have complex life cycles spanning a major habitat shift, and delayed life history effects play an important role in their population dynamics. Through their effect on size at metamorphosis, factors in the pre‐ metamorphic environment often have profound effects upon survival and fecundity in the post‐metamorphic environment. Here, we adopted a combined experimental and field observational approach to investigate the factors that determine size at metamorphosis in pond‐breeding amphibians, and to predict some of their downstream effects on population stability. We set up ecologically realistic mesocosm communities for the endangered California tiger salamander Ambystoma californiense to test the effects of larval density, prey density and hydroperiod on mean size at metamorphosis. We found significant effects for all three factors, with mean size at metamorphosis negatively correlated with larval density and positively correlated with prey density and hydroperiod. We also used six years of field survey data to identify the most informative model explaining mean size at metamorphosis and thus validate our mesocosm results. The optimal three‐term model identified terms that were roughly analogous to each of the mesocosm treatments and with similar effect sizes, providing strong field confirmation of our experimental results. The field data also provide correlations between each factor and the number of metamorphs recruited to the population, allowing us to predict the effect of each factor on population stability. Finally, we show that these populations of the endangered A. californiense are strongly resource limited, which has important implications for their management and recovery as an endangered taxon.

International audiencePlant-herbivore-natural enemy associations underpin ecological communities, and such interactions may go up to four (or even more) trophic levels. Here, over the course of a growing season, we compared the diversity of secondary hyperparasitoids associated with a common host, Cotesia glomerata, a specialized larval endoparasi-toid of cabbage butterfly caterpillars that in turn feed on brassicaceous plants. Cocoon clusters of C. glomerata were pinned to ~30 Brassica nigra plants by pinning them either to branches in the canopy (~1.5 m high) or to the base of the stem near the ground. The cocoons were collected a week later and reared to determine which hyper-parasitoid species emerged from them. This was done in four consecutive months (June–September). Cocoons placed in the canopy were primarily attacked by specialized winged hyperparasitoids (Lysibia nana, Acrolyta nens), whereas cocoons on the ground were attacked by both winged and generalist wingless hyperparasitoids (Gelis acarorum, G. agilis), although this changed with season. There was much more temporal variation in the diversity and number of species attacking cocoons in the canopy than on the ground; the abundance of L. nana and A. nens varied from month to month, whereas P. semotus was only prevalent in August. By contrast, G. acarorum was abundant in all of the samples placed near the ground. Our results show that hyperparasitoids partition host resources at remarkably small vertical spatial scales. We argue that spatial differences in the distribution of natural enemies can contribute to the diversity patterns observed in the field

International audiencePopulation-wide mating patterns can select for equal parental investment in both sexes, but limiting resources, such as mates or developmental substrates, can increase competition leading to biased sex ratios in favor of either sex. Such competition for resources typically occurs in spatially structured populations, where dispersal is limited. In this laboratory study, we investigate if and how resource competition affects sex allocation, discriminative behaviors and competitive interactions of the wingless hyperparasitoid Gelis acororum, which exploits patchily distributed hosts. We show that G. acororum sex ratios are male-biased and that this is not a consequence of constrained reproduction by virgin females. Our results suggest that this pattern of reproductive investment, which is only rarely observed in parasitoids, is a consequence of resource limitation, in terms of hosts rather than mates. Further, G. acororum appears not to respond to intrinsic host quality or to prior oviposition in its host. When competing inter-specifically for host resources, G. acororum outcompetes its congener Gelis agilis, but does so mainly when ovipositing on the host first. Overall, our results suggest that host resource limitation could be an important environmental factor shaping sex allocation in G. acororum, with competition taking place both intra- and inter-specifically

International audience1. The optimisation theory predicts that insect mothers should oviposit on resources on which they attain the highest exclusive fitness. The development of parasitoid wasps is dependent on limited host resources that are often not much larger than the adult parasitoid.2. In the present study preference and development in three congeneric species of secondary hyperparasitoids attacking cocoons of two congeneric primary parasitoids that differ significantly in size were compared. Gelis agilis (Fabricius) and G. acarorum (L.) are wingless hyperparasitoids that forage in grassy habitats, whereas G. areator (Panzer) is fully winged and forages higher in the canopy of forbs.3. The three species were reared on cocoons containing pupae of a small gregarious endoparasitoid, Cotesia glomerata (L.), and a larger solitary species, C. rubecula (Marshall), both of which develop in the caterpillars of pierid butterflies.4. Adult mass was correlated with initial cocoon mass in all three species, whereas development time was unaffected. Wasps were larger when developing in C. rubecula. However, for a given host mass, wasps were larger when developing on the smaller host, C. glomerata. This suggests that there is a physiological limit to hyperparasitoid size that was exceeded when C. rubecula served as host.5. All three hyperparasitoids strongly preferred to attack cocoons of the larger species, C. rubecula, often avoiding cocoons of C. glomerata entirely. 6. Preference and performance are correlated in the three Gelis species. However, owing to variation in the distribution and thus abundance of their hosts, it is argued that cumulative fitness may be still higher in the smaller host species

International audienceLife histories can reveal important information on the performance of individuals within their environment and how that affects evolutionary change. Major trait changes, such as trait decay or loss, may lead to pronounced differences in life history strategies when tight correlations between traits exist. Here, we show that three congeneric hyperparasitoids (Gelis agilis, Gelis acarorum and Gelis areator) that have diverged in wing development and reproductive mode employ markedly different life history strategies. Potential fecundity of Gelis sp. varied, with the wingless G. acarorum maturing a much higher number of eggs throughout life compared with the other two species. Realized lifetime fecundity, in terms of total offspring number was, however, highest for the winged G. areator. The parthenogenic G. agilis invests its resources solely in females, whilst the sexually reproducing species both invested heavily in males to reduce competitive pressures for their female offspring. Longevity also differed between species, as did the direction of the reproduction-longevity trade-off, where reproduction is heavily traded off against longevity only in the asexual G. agilis. Resting metabolic rates also differed between the winged and wingless species, with the highest metabolic rate observed in the winged G. areator. Overall, these geline hyperparasitoids showed considerable divergence in life history strategies, both in terms of timing and investment patterns. Major trait changes observed between closely related species, such as the loss of wings and sexual reproduction, may contribute to the divergence in key life history trait