Pulsed resources affect the timing of first breeding and lifetime reproductive success of tawny owls

International audienceAccording to life-history theory, environmental variability and costs of reproduction account for the prevalence of delayed reproduction in many taxa. Empirical estimates of the fitness consequences of different ages at first breeding in a variable environment are few however such that the contributions of environmental and individual variability remains poorly known. 2. Our objectives were to elucidate processes that underpin variation in delayed reproduction and to assess lifetime consequences of the age of first breeding in a site-faithful predator, the tawny owl Strix aluco L. subjected to fluctuating selection linked to cyclical variation in vole density (typically 3-year cycles with low, increasing and decreasing vole densities in successive years). 3. A multistate capture-recapture model revealed that owl cohorts had strikingly different juvenile survival prospects, with estimates ranging from 0AE08 to 0AE33 respectively for birds born in Decrease and Increase phases of the vole cycle. This resulted in a highly skewed population structure with >75% of local recruits being reared during Increase years. In contrast, adult survival remained constant throughout a vole cycle. The probability of commencing reproduction was lower at age 1 than at older ages, and especially so for females. From age 2 onwards, pre-breeders had high probabilities of entering the breeding population. 4. Variation in lifetime reproductive success was driven by the phase of the vole cycle in which female owls started their breeding career (26-47% of variance explained, whether based on the number of local recruits or fledglings), more than by age at first breeding or by conditions experienced at birth. Females who postponed reproduction to breed for the first time at age 3 during an Increase phase, produced more recruits, even when accounting for birds that may have died before reproduction. No such effects were detected for males. 5. Sex-specific costs of early reproduction may have accounted for females being more prone to delay reproduction. Contrary to expectations from a best-of-a-bad job strategy, early-hatched, hence potentially higher-quality females were more likely to breed at age 1, but then experienced rapidly declining food resources and so seemed caught in a life-history trap set by the multiannual vole cycle