Surplus Killing in the Hunting Strategy of Small Predators

Abstract

Listen

Translate article

Studies on the impact of vertebrate predators on a prey population are often based on estimates of the numbers of predators and prey (e.g., Pearson 1966, 1971; Goszczynski 1977; Phelan and Robertson 1978; King 1980; Erlinge et al. 1983). This approach includes assumptions about the functional response of predators, that is, about the changes in hunting effort for different prey species and about the factors that determine the handling time (Holling 1965). Changes in the preference for different prey species, often called switching, have an important effect on the functional response curve, especially at relatively low prey densities (Greenwood and Elton 1979). The lower part of the functional response curve also will probably be influenced by transit time (Murdoch 1977; Oaten 1977) and by the costs of activity (Abrams 1982). The upper part of the functional response curve is largely determined by handling time, that is, the time between the capture and the resumption of searching behavior (Holling 1965). The extent of switching can be estimated by using scat analysis (e.g., Day 1968; Phelan and Robertson 1978; Tapper 1979; Erlinge 1981), and weight loss of predators can be used for estimating inadequate feeding rates. Handling time could be estimated readily if all predators were characterized by rigid hunting behavior in which a prey is attacked when seen within a fixed attack radius and consumed before the predator starts to search for new prey items. If, however, predatory behavior is more flexible, handling time becomes more difficult to estimate. Such flexibility may be created, for example, by selective feeding and consequently decreased consumption per prey with increased prey density (Mysterud 1980; Stenseth 1981; Abrams 1982). It is also possible for predators to kill prey without immediately consuming them (Nyholm 1961; Kruuk 1972; Curio 1976; Mysterud 1980; Elgmork 1982). Such behavior substantially shortens the handling sequence by eliminating the time used for feeding activities. This can make the functional response curve nearly linear or only slightly convex at even high prey densities. Consequently, the numerical predator-to-prey ratio advocated by Pearson (1966, 1971) might not reflect the intensity of predation, and estimates of predation based on identifiable prey remains found from predator scats could be far too low.