Consequences of recruitment decisions and heterogeneity on age-specific breeding success in a long-lived seabird.

An individual's age at first reproduction and investment in successive reproductive attempts are involved in mechanisms that can impede somatic repair, resulting in a decline in reproductive abilities with age (reproductive senescence). We used long-term data from the Black-legged Kittiwake, a long-lived seabird, to address the relationship between recruitment age, age-specific breeding success (BS), and reproductive senescence, while accounting for breeding experience and temporal variation in BS. We first detected late-life improvement in BS across all recruitment groups, which we recognized as "within-generation selection" or the selective disappearance of "frail" phenotypes. When such heterogeneity was accurately accounted for, we showed that all individuals suffered reproductive senescence. We first highlighted how different combinations of pre- and post-recruitment experience across recruitment groups resulted in maximal BS at intermediate ages. BS increased in early recruits as they gained post-recruitment experience, whereas late recruits gained pre-recruitment experience that led to high BS at recruitment. Only individuals recruiting at intermediate ages balanced their pre- and post-recruitment experience. Consistent with the "cumulative reproductive cost hypothesis," we also observed a faster decline in BS in early recruits at advanced ages, whereas individuals delaying recruitment experienced the slowest decline in BS with age. Early recruits, however, reached the highest levels of BS at intermediate ages, sensus stricto (10-13 years old), whereas individuals delaying recruitment experienced the lowest at similar ages. These divergent trajectories may reflect a "delayed trade-off" balancing a maximization of midlife BS against reproductive senescence at advanced ages. Additionally, annual variation in BS had a greater effect on individuals early in life, suggesting that experienced individuals were able to buffer out the effects of temporal variation on BS, which can ultimately improve fitness in stochastic environments. Our findings stress that (1) both observed and unobserved heterogeneity are important in detecting within-individual senescence, and (2) short-term trade-offs may be rare in long-lived species; thus, cumulated reproductive costs should be invoked as an alternative mechanism underlying reproductive senescence.