Résumé
Lifetime recruitment of breeding offspring estimated in small-scale study plots (i.e. local recruitment) is considered to be the best available ecological measure of contributions to following generations, and sufficient for the quantitative study of adaptation in natural populations. Recent investigations suggest that local recruitment of breeding offspring does not always reflect the total recruitment in the whole population, especially in small-scale plots where the majority of locally-born offspring leave these plots to breed elsewhere. We examined in an avian island population whether study plot size has an important impact on different population and fitness measures. We defined around a central nestbox seven plots, varying in radius from 100 to 700 m. We show that in the smallest plots, the local replacement rate of adults by breeding offspring is low, the number of locally-born offspring settling beyond the limits of a plot is high, and relationships between local and total recruitment are weak. This is especially true for daughters as more daughters than sons settle beyond the limits of local plots for breeding. Our interpretation is that the lifetime recruitment of breeding offspring in local plots does not necessarily reflect the lifetime recruitment of breeding offspring in the whole population, especially when plots do not cover the natal dispersal distance. Consequences of dispersal for the quantitative study of adaptation are discussed.