Length of tail streamers in barn swallows

Abstract

Both natural and sexual selection affect the expression of secondary sexual characters¹, so any secondary sexual character will be affected by a mixture of selection pressures. Evans² has compared the flight performance of male barn swallows (Hirundo rustica) whose tail lengths have been altered with that of controls, and concludes that the outermost tail feathers of these birds are mainly the outcome of natural selection.

Main

If tail length has been determined by natural selection, then both elongation and shortening of tail feathers should lead to reduced flight performance. If sexual selection is involved, it will reduce the flight performance of males with elongated tails, so shortening the tail will improve performance. It is widely accepted that sexual selection cannot on its own be responsible for the elongated tail, but it may have an influence, as males have allometrically longer tails and streamers than females³.

The average difference in tail length between the sexes is about 15 mm in northern European barn swallows⁴. Because the female tail presumably represents the aerodynamic optimum in adult barn swallows³,⁴, shortening the male tail by less than 15 mm should provide a test for the sexual selection hypothesis (Fig. 1). But Evans shortened male tails by 20 mm, reducing them beyond the aerodynamic optimum. His experiment could therefore demonstrate reduced flight performance only in support of natural selection, and so could not discriminate between the two hypotheses.

Figure 1: Models for the relation between streamer length and aerodynamic cost in barn swallows, as tested in tail-manipulation experiments.

Filled circles, tail-streamer manipulations and effects described by Evans²; the solid curve shows his model. Our alternative model (broken curve) assumes that female streamer length represents the aerodynamic optimum and associated costs away from this optimum. If male streamer length has been modified by sexual selection, this gives an increased cost in relation to female streamer length. We have assumed that the differences between the sexes are controlled for allometric size differences.

Norberg⁵ has proposed an aerodynamic mechanism to account for the elongated tail streamers in barn swallows, but Evans did not test this. According to Norberg, when the tail is lowered, the tail streamer is bent by air flow, causing an aeroelastic twist through the entire feather that is passively rotated. This rotation turns the outer tail into a flap that provides increased lift of the entire tail. For this to work, the tail must be considered as a finely tuned structure in which the entire tail is a coadapted complex of characteristics. If the tail streamers are manipulated beyond their aerodynamic optimum, other characteristics will also be affected, such as planform, curvature of the feather shaft, flexural stiffness, and torsional rigidity of the outermost tail feathers, which supposedly evolved to compensate for increased pitching moments caused by exaggerated streamers⁵. Any experimental shortening is therefore liable to disrupt performance even in male swallows.

The tail of male barn swallows has some functional utility, but it is the difference between the sexes that is crucial to mate choice. Longer tails in male barn swallows may increase several components of sexual selection (reviewed in refs 4, 6), which has been demonstrated by manipulating apparent tail length without affecting aerodynamic function⁷. All the evidence indicates that the difference in tail length between the sexes is caused by sexual selection.

Reply - Evans