Abstract
In an urban setting, the association between clutch size and a fitness component differed between four scales of analysis. At the individual level, birds with larger clutches produced on average more fledglings per nest, but clutch size was not associated with the average proportion of eggs that produced fledglings. At the study population level, the intermediate-sized clutches produced cumulatively most of the hatchlings, fledglings and locally born breeders compared to the smallest and largest clutches. Therefore, at this scale of analysis, clutches with an intermediate size made locally the highest contributions to next generations. However, the estimated lifetime replacement rate of adults by locally born offspring was very low, whatever the size of the clutch. None of the clutch sizes were productive enough to maintain the size of the local study population in the absence of massive immigration. Thus, interpretations of processes associated with local adaptation might change according to which fitness components will be highlighted in a study. During the breeding phase, costs associated with egg production, incubation and/or parental rearing efforts might ultimately determine which clutch sizes will be maintained across generations. After the breeding phase, the probability to recruit breeding offspring is ultimately influenced by unpredictable and rarely identified events that also might determine which clutch sizes allow population maintenance. In this framework, the best adaptive breeding strategy would be to lay an excessive number of eggs aimed to anticipate offspring loss in a landscape with unpredictable living probabilities.