We studied reproductive costs of Black-legged Kittiwakes (Rissa tridactyla) in Prince William Sound, Alaska (USA) by removing entire clutches from randomly selected nests over four successive years, and then contrasting survival and fecundity of adults from manipulated and unmanipulated nests in each subsequent year. To elucidate mechanisms that lead to the expression of reproductive costs, we simultaneously characterized several behavioral and physiological parameters among adults in the two treatment groups. We also examined naturally nonbreeding adults that previously bred to determine their survival and future nonbreeding probabilities.

Food availability varied during the study, being generally poor at the onset, and improving in later years. Adult nest attendance and body condition (assessed late in the chick- rearing period) varied accordingly among years, and between adults raising chicks and adults that had their eggs removed. Adults from unmanipulated nests incurred significant survival costs in all years, although fecundity costs were strongly expressed in only one of four years. Neither survival nor fecundity costs were strongly influenced by body condition or food availability, and no difference in reproductive costs was detected between the sexes. Although unmanipulated breeders survived at lower rates than manipulated breeders due to costs of reproduction, their survival rates were elevated compared to those of natural nonbreeders, presumably due to differences in individual ability. These findings indicate that models of adult survival must consider not only an organism's reproductive state, but also the factors that lead to that state.

Although body condition appeared to be weakly related to survival, it was insufficient to explain the full magnitude of survival costs observed. We suggest that other parameters that were found to differ between treatment groups (e.g., rates of energy turnover, baseline levels of stress, and patterns of allocating body reserves) may be important mechanistic determinants of reproductive costs in kittiwakes, and potentially other long-lived species. Future efforts should move beyond simple assessments of body condition and toward more integrated measures of physiological condition when attempting to identify factors that influence how long-lived species balance the costs and benefits of reproduction.