Résumé
Iteroparity is an adaptive response to uncer- tainty in reproductive success. However, spreading repro- ductive success over multiple reproduction events during a lifetime is constrained by adult mortality and the stochas- ticity associated with interactions between external factors and physiological states. The acquisition of information about environmental conditions during the growth of prog- eny and sufficient resources during the non-reproductive period are key factors for breeding success. Consequently, we hypothesized that long-lived animals may skip a breed- ing opportunity when information about unfavourable environmental conditions is available. In addition, nutri- tional constraints could prevent an animal from replenish- ing its reserves sufficiently to invest in the current breeding period. We investigated these questions using capture– recapture data from a 5-year study on a large population of yellow-bellied toads in a forest in north-eastern France. We took advantage of various advances in multi-state cap- ture–recapture models (e.g. unobservable states and mix- ture models) to test our hypotheses. Our results show that the combined effects of rainfall deficit and the breeding/ non-breeding state of individuals during the past breeding season affect breeding probability during the following breeding opportunity. We also found that females breed less frequently than males, suggesting that the overall energy cost of reproduction differs between genders. Finally, the results indicate that toad survival appears to be negatively influenced by rainfall deficits. We discuss the yellow-bel- lied toad’s reproductive behaviour in term of bet-hedging strategy and life history trait evolution.