PREDATORS IN GROUPS

Generally, when predators form groups it is thought that they do so to enhance rates of prey intake. Jackals hunting gazelle fawns and hyenas hunting wildebeest calves are also more effective hunting in pairs because one can counter the attacks of the mother while the other kills the defenseless calf. Youíve seen a similar example when two cheetahs hunt a wildebeest calf: one chases the mother while the other runs down the calf. Groups of killer whales have been seen surrounding and hemming in porpoises. Likewise, white pelicans surround fish schools. Gotmark's experimental study of gulls have shown that individual foraging success on schooling fishes is higher in groups; this is because the foraging of other gulls disrupts the school and makes it easier for a gull to single out and capture an individual prey (Discussion article, also see Alcock). These examples show a clear advantage to cooperation in hunting. Foraging success of wolves hunting large ungulates (e.g., wapiti, moose) also appears to depend on group cooperation.

Social mammalian predators can generally capture larger prey than solitary predators. Social carnivore species characteristically prey on animals their own size or larger while solitary mammalian carnivores typically prey on animals considerably smaller than themselves.

These comparisons also hold within species. Small packs of African wild dogs (Lycaon pictus) kill mainly wildebeest calves and gazelles; larger packs specialize on larger prey, e.g., adult zebras. Solitary wolves live primarily on carrion and small game, whereas packs can feed on moose and smaller ungulates. Lions in small hunting groups do not attack water buffalo, but lions in groups of four or more do hunt water buffalo and achieve highest per capita rates of food intake when specializing on water buffalo.

Social carnivores are likely to be able to take large prey not only because several individuals can better pull down a large victim, but also because a social predator can afford injuries that would seriously hamper the survival of a solitary predator. Both Brian Bertram and George Schaller have recorded injured lions surviving for months on food killed by other pride members.

Grouping may reduce variability in individual rates of food intake in unpredictable environments. Ekman and Hake (1988. Behav. Ecol. Sociobiol. 22:91-94) showed experimentally that individual variation in food intake is less when individuals feed in groups of 2 than solitarily when food is patchy and patches arenít monopolized by the finder of the patch. However, when individuals were food-deprived, they tended to be risk prone, choosing to forage solitarily.

Optimal group size

Several theoretical approaches to evaluating optimal size of animal groups have been attempted. Wittenberger (1980. Am. Nat. 115:197-222) suggested computation of lifetime reproductive output of group members as a function of group size. Costs and benefits enter the model by affecting the adult mortality rate or the recruitment rate. For example, as group size increases, mortality rate either may decrease because protection from predators increases or may increase because competition for food intensifies.

Group size of mammalian carnivores

George Schaller found that success rate of lions hunting gazelles, zebras and wildebeest was doubled when two or more hunted together. A prey animal escaping from one lion may run within the range of another which it has not detected. Most studies have assumed that group hunting predators do little to coordinate their hunting, but some recent studies have shown that group members may coordinate their hunting and likely increase their success by doing so. For example, Stander (1992. Behav. Ecol. Sociobiol. 29:445-454) observed 484 coordinated hunts by lions. Some lionesses would act as "wings" and circled prey while others, "centers", waited for prey to move toward them (see handout, Fig. 1a,b). Stander found that individual lionesses tended to assume the same positions in different hunts, and hunts had the highest probability of success when females did occupy their typical positions. Also, Alcock summarizes the coordinated hunting movements of white pelicans.

On the basis of Schallerís observations, several studies suggested that food intake rates of lions on a per individual basis were highest for hunting group size of 2 throughout the year. Because group sizes are typically larger than 2 (53% of lions hunting Thomsonís gazelles and 79% of lions hunting wildebeest and zebra were in groups of 3 or more, (see handout, Table), it appeared from such early studies that lions often hunt in larger groups than would be optimal for each individual. Packer et al. (1990. Am. Nat. 136:1-19) recently summarized a far more extensive data set. During the season of prey scarcity, hunting groups of 1 and 5-6 had the highest foraging success (kg/day/female). These results suggest that females in prides of 4 or fewer females should hunt alone, but Packer et al. found that females in small prides typically foraged in the largest groups possible (e.g., in half of the sightings of prides of 4 females, all of the females were hunting together--see handout, Figs. 1 and 2). Tendencies of females to group seem to be related primarily to (1) defense of cubs against (a) non-pride males (see handout) and (b) hyenas and (2) competition with females on neighboring territories (larger groups repel smaller groups in territorial disputes). Packer et al. concluded that females live in prides and hunt in groups primarily to protect cubs and maintain their territory. These groupings may be suboptimal in terms of foraging success, i.e., the rate of food intake/individual. At least in Packer's study area, female lions seem to group primarily to protect cubs and defend their territory, and male lions group to compete with other males.

Do group sizes of other predators fit predicted optima? At least for wolves, group sizes also are larger than the optimality models predict. Individual foraging success of wolves peaks in packs of three when deer are the prey and in packs of five when larger cervids are the primary prey (Nudds, T. D. 1978. Am. Nat. 112:957-960). In packs that Nudds examined, 51 of 66 wolves (77%) hunted in groups that were larger than the predicted optima. Rodman expanded the analysis and calculated the size of the group per wolf (see handout) -- sizes of groups in which most wolves occurred were considerably larger than would be optimal in terms of the hunting success of the individuals in those groups.

The optimal group size models assume that individuals choose or influence group size in ways that maximize their fitness in terms of rate of food intake. However, demographic constraints may also affect group size. For example, individuals likely are not be free to move between groups to find one of optimum size. Each social group may be viewed as a discrete population characterized by its own birth rate, death rate, immigration rate and emigration rate. If migration rates are low, the group should reach an equilibrium size determined by the carrying capacity of the environment rather than the optimal size, insofar as the interests of individuals in the group are concerned. Since migration rates are low in many mammalian species, observed group sizes may not agree with predictions of the above models. Pride size of lions and pack size of wolves seem to be limited by the carrying capacity and is larger than optimal in terms of prey intake rates of individual group members.

Should relatedness of group members affect optimal group size? We have seen that group sizes of lions and wolves are larger than optimal in terms of rate of resource intake by the hunting individuals. In both species the beneficiaries are relatives that may not have much chance of eventual reproduction unless they stay at home. Rodman (1981. Am. Nat. 118:275-283) provides a theoretical basis for the argument that optimal group size might be expected to increase as the degree of relatedness within the group increases (see handout, Fig. 1). Thus a dominant animal's inclusive fitness may be enhanced by permitting a subordinate relative to stay but not a subordinate individual who is unrelated.

Carnivore group size and territory size

Kruuk and MacDonald (1985. Pp. 521-536 in R.M. Sibly and R. H. Smith. Behavioural ecology. Blackwell) suggest that mammalian carnivores that are group-living can be divided into two classes. Expansionists are species which expand territory size as group size increases (e.g., coyotes, wolves; see handout, Fig. 31.2); such species typically occupy more-or-less homogeneous habitat. Contractors are species which maintain constant territory size, regardless of group size (see handout, Fig. 31.3). Eurasian badgers are clearly contractors. They are omnivorous but prey primarily on earthworms; they forage individually but live in clans. The number in the clan is correlated with the biomass of earthworms on the territory, i.e., the territoryís carrying capacity. Declines in clan size and removals of clans did not alter neighboring boundaries over at least 6 years. The key difference between expansionists and contractors seems to be the patchiness of the environment. In patchy environments it may be necessary for an individual to defend a large area so that all necessary habitat types are incorporated into its territory. The territory then may be large enough to support additional individuals (see handout, Fig. 31.1).

Back to lecture notes index.

Back home.