An analysis of continent-wide patterns of sexual selection in a passerine bird.

Patterns of selection are widely believed to differ geographically, causing adaptation to local environmental conditions. However, few studies have investigated patterns of phenotypic selection across large spatial scales. We quantified the intensity of selection on morphology in a monogamous passerine bird, the barn swallow Hirundo rustica, using 6495 adults from 22 populations distributed across Europe and North Africa. According to the classical Darwin-Fisher mechanism of sexual selection in monogamous species, two important components of fitness due to sexual selection are the advantages that the most attractive males acquire by starting to breed early and their high annual fecundity. We estimated directional selection differentials on tail length (a secondary sexual character) and directional selection gradients after controlling for correlated selection on wing length and tarsus length with respect to these two fitness components. Phenotype and fitness components differed significantly among populations for which estimates were available for more than a single year. Likewise, selection differentials and selection gradients differed significantly among populations for tail length, but not for the other two characters. Sexual selection differentials differed significantly from zero across populations for tail length, particularly in males. Controlling statistically for the effects of age reduced the intensity of selection by 60 to 81%, although corrected and uncorrected estimates were strongly positively correlated. Selection differentials and gradients for tail length were positively correlated between the sexes among populations for selection acting on breeding date, but not for fecundity selection. The intensity of selection with respect to breeding date and fecundity were significantly correlated for tail length across populations. Sexual size dimorphism in tail length was significantly correlated with selection differentials with respect to breeding date for tail length in male barn swallows across populations. These findings suggest that patterns of sexual selection are consistent across large geographical scales, but also that they vary among populations. In addition, geographical patterns of phenotypic selection predict current patterns of phenotypic variation among populations, suggesting that consistent patterns of selection have been present for considerable amounts of time.

Imitating the initial evolutionary stage of a tail ornament.

Costs of a sexual ornament in its early evolutionary form and the relationship between these costs and individual condition may be an important influence in the likelihood of possible evolutionary mechanisms involved in the evolution of this ornament. We reconstructed the tail shape in hypothetical ancestors of recent hirundines (Aves: Hirundinidae), from which the elongation of tail feathers under sexual selection might have begun. By elongating the tail in sand martins (Riparia riparia, Hirundinidae), we simulated the early evolution of a long forked tail--the typical ornament of male hirundines. Birds with initial ornament captured smaller insects than controls, which suggests that this ornament imposed a cost in terms of impaired foraging. Furthermore, birds with naturally longer tails were better able to cope with initial ornament than naturally short-tailed birds. If length of tail in sand martins indicates the quality of individuals, our results suggest higher costs of this initial ornament for poorer than for higher quality individuals. We discuss the potential role of the handicap principle and other mechanisms in early evolution of a tail ornament.