Predation strategies in aerial feeding birds

The predation strategies of four species of aerial feeding bird, the Swallow, Hirundo rustica, Sand Martin, Riparia riparia, House Martin, Delichon urbica and Swift, Apus apus, are examined during their breeding season in Britain. The abundance of aerial insects upon which they feed increases in April and reaches a plateau in May which is maintained until September. Aerial insect density is higher nearer the ground, the spring rise in abundance occurs earlier, a greater proportion of larger insects are available and insect numbers fluctuate less than at higher levels. The arrival times of the different predator species into the breeding area are staggered and this is linked to their preferred prey and feeding station, and to the distribution of aerial insects. Throughout the breeding season the four species of bird are ecologically isolated with respect to feeding station in the air-space and various characters of the insect prey, primarily size and mobility. It is suggested that to coexist they segregate along a combination of axes of the feeding niche, although segregation in air-space may be the most effective. Morphological adaptation to feeding niche is shown for tail shape, wing-length and bill shape, which respectively influence manoeuvrability, mode of flight and prey handling ability of the predators. Both Sand Martin and Swallow show increases in their prey size and mobility niche breadths in favourable feeding conditions whilst all four species simultaneously show decreases in air-space niche breadths. It is postulated that both within and between species there is a tendency to switch between patch specialisation and prey specialisation. Current theories of optimal patch and prey choice are discussed in relation to this hypothesis. Under very adverse conditions prey size overlaps are small and, when they increase, air-space overlaps simultaneously decrease. This suggests competition between aerial feeding birds in adverse conditions. Swallows deliver more meals per unit time to larger broods, with an increase in the feeding rate of the male most evident. Males also increase their feeding rate to second broods, but these receive the same amount of food as first broods indicating a seasonal decline in meal size. There is a seasonal decline in size of prey items even though a higher density of available large (> 5 mm body length) insects occurs during second brood feeding. It is shown that the largest insects are in patches too far from second brood nests to be profitable. For first and second broods an optimal foraging strategy is to choose patches with the largest mean insect size, if the patches are within a given distance from the nest. Selection of an optimal size for fast-flying taxa is influenced by time and energy costs of pursuit. It is suggested that rearing larger than normal broods is constrained by the requirement of equal parental investment and the extra reproductive cost likely to be incurred by the male bird. Based on a very small sample of nests it is tentatively suggested that in colonial nesting Sand Martins there is an advantage in nesting early, with more birds feeding communally to aid location of food concentrations. Because individuals with centrally placed burrows are more able to minimise time spent deterring predators, and because early centre nesters suffer less time loss through competition at the nest-site, these same individuals have more time for locating other feeding birds and food concentrations. Thus the early season, centre colony nesters may collect more food per unit time and raise more off-spring. Comparison of aerial feeding bird communities in three zoogeographical regions indicates that selection for characters contributing to mode of flight and manoeuvrability are more important than selection for bill character. In Africa the Palearctic species maintain essentially the same feeding stations as in their breeding season although the air-space niche breadths show contraction and overlap is reduced except between Swift and House Martin. The indigenous species have different feeding stations and small air-space niche breadths, and where overlap is high the species involved apparently do not breed during the "winter" of the Palearctic species, possibly because of food shortage. The prey size niche breadth of the Swallow, H. rustica is very similar between Britain and Africa. With many more species utilizing this niche axis, compensation is likely along another axis, probably air-space