Biology 441. Animal Behavior.

Lecture 19. Monday, 11 November 1996

PREY IN GROUPS

It is generally thought that optimal group size in prey animals is a compromise between competition for resources, which tends to favor smaller group size, and avoidance of predation which generally favors increased group size (when the environment provides insufficient cover). Many animals form social groups to gain better protection from predators. Open habitats are conducive to formation of groups because social avoidance of predators and defense against predators are most effective in such habitats. Closed habitats offer concealment, where crypticity and spacing out of solitary individualsis the best way to avoid predation.

Why is grouping advantageous in open habitats? Alcock summarizes studies that show that predation attempts of goshawks on wood pigeons are less successful when wood pigeons are in groups, particularly large groups. Vigilance by the group as a whole increases even though individual vigilance decreases with increasing group size. The advantage of larger group size applies primarily to individuals in the center of the group, however. Jennings and Evans (1980. Anim. Behav. 28:634-635) showed that starlings on the periphery of flocks spent much less time feeding and more time in vigilance behavior than those in the center. Even in large groups, individuals on the edge of the group had high vigilance rates (see handout, Figure 13-13). Individuals in the center (and midway) had lower vigilance rates (and thus higher feeding rates) as group size increased. Geometrically area increases exponentially while perimeter increases linearly, thus realtively more individuals will be in the center as size of the group increases. Presumably, individuals on the edge are also more vulnerable to predation.

Even though the effectiveness of vigilance seems to increase as group size increases, larger groups could elicit a greater functional or numeric response by predators. Recall from ecology that the functional response is a change in an individual predator's behavior such that it increases or decreases its hunting intensity on a particular type of prey. The numeric response is a change in the number of individuals feeding on a particular type of prey. In my studies of raven predation on murre eggs at the colony at Bluff in Norton Sound, there is a pronounced functional response to availability of murre eggs by the adult ravens nesting within the seabird colony. However, they are highly territorial and keep other ravens away; therefore, there is a limited numeric response.

Although a large group will be more easily detected by a predator, predation rates per individual probably decrease with group size not only because of more effective vigilance, but also because of the ìconfusionî and ìdilutionî effects. To be effective, a predator generally has to be able to single out an individual prey during the chase to be able to capture it. As a large group escapes and the relative positions of the prey rapidly change, a predator may be unable to track a certain individual. Alcock gives some good examples of how the dilution effect would work: clumping and synchronization of activities are both important.

Mutual defense may be important for some types of prey, e.g., musk oxen.

W.D. Hamilton considered the geometry of prey groups and suggested that individuals should position themselves in the center of the group, i.e., they should use others as a shield. There are few data to test this idea, but it apparently is operating for wood pigeions and starlings (see above). Black-headed Gull nests on the periphery of the colony are more vulnerable to predation of eggs and chicks than those in the center, but this is due to mobbing responses of the gulls which are more effective in deterring predators when there are many, rather than few, mobbing adults.

There might be an upper limit to group size of prey. If group size gets too large, it may be difficult for some members of the group to detect a predator, and they may escape in the wrong direction. Lois Crisler suggested that wolves catch caribou in large herds more easily because many individuals in large herds canít see the wolves approaching.

Alarm calls (covered in Andrea's lecture on signals)

The ability of an animal (predator) to determine the direction of the sound source depends primarily on 3 binaural cues: (a) if the ears are far enough apart, there are differences in arrival time of the sound at the ears and (2) in the phase (point in the cycle of the wave form) at the right and left sides, as well as (3) intensity differences resulting from the sound shadow of the head. Apparently, terrestrial vertebrates depend primarily on differences in arrival time and intensity differences as cues. Differences in arrival time are best determined if the sound has abrupt discontinuities or transients. Intensity differences are best detected if the frequency is high (due to rapid attenuation of high frequencies), and phase differences are best detected if frequency is low (and wavelengths relatively long). The easiest sounds to locate are those that provide cues for all methods, requiring a wide frequency spectrum and sharp discontinuities.

In songbirds there are distinctive call types in relation to predators. ìAlarm callsî are thin whistles which are not strongly punctuated at beginning or end and are very difficult to detect. Alarm calls are designed to minimize cues of location. The time cues are eliminated because the sound fades in and out with no discontinuities. A pure tone is used, pitched above the frequency maximum for phase difference location b ut well below the optimum for generating intensity differences at the ear of potential receivers. Such a sound, a thin whistle, has a ventriloquial quality and is difficult to locate (see handout). Alcock also discusses studies showing that frequencies of alarm calls are those to which other flock members are very sensitive and avian predators are relatively insensitive.

(It's evident from the handout that alarm calls are quite similar among unrelated passerines, indicating some degree of interspecific communication and possibly cooperation. Other calls that show similarities among species and provide evidence of interspecific communication and cooperation include the flight calls of finches that form interspecific foraging flocks during the nonreproductive season. Sounds of many gull species (Larus) used to sound alarm and to attract other gulls to feed are very similar in structure and elicit interspecific responses.) These calls typically are given from cover, and callers probably are at no more risk than non-callers.

In contrast to alarm calls, "mobbing" calls of birds provide excellent cues of location (see handout) as do many other calls and song. These are distinguished primarily by a wide range of frequencies with abrupt starts and stops. Stationary hawks or owls elicit mobbing calls, and these calls attract other birds. Mobbing birds approach with a certain distance, display conspicuously and call loudly. This may serve to drive a predator away from the birds' home ranges.

There are also ìstrong alarmî calls (see handout) that are used to warn relatives (e.g., young in the nest). These provide information on location of the caller and may be used mostly in the context of warning young about location, as well as presence, of predators to which the young are vulnerable but parents arenít.

Individuals in groups may give alarm calls. The individuals giving the calls may be exposing themselves to predators. In his studies of ground squirrels, Paul Sherman found that alarm-calling ìtrillsî increase risk of predation to terrestrial (mammalian) predators, but alarm-calling ìwhistles,î given in response to aerial predators do not increase risk relative to non-alarm calling individuals. Ground squirrel trills are given primarily be females who have female relatives nearby (see handout); whistles show no such nepotism.

Warning coloration

Some defense strategies definitely involve signalling between prey and predators. Many unpalatable animals (particularly insects and amphibians) are brightly colored. Such warning signals are called aposematic. Generally, itís thought that a predator must learn which prey are distasteful. If in learning that a predator species is distasteful, the predator kills one or more individuals, the question of how distastefulness and brighter coloration evolved arises. Ronald Fisher hypothesized that these traits could evolve through kin selection. Toad tadpoles are distasteful and are found with kin. Probably once one individual in the group is sampled by a particular predator others in the group would be less vulnerable to predation.

D.A.S. Smith (1979. Nature 281:215-216) studied bird predation on butterflies. Birds usually catch butterflies on the wing, carry the prey to a perch, and attempt to remove the wings before eating it. If the initial taste is repelling, the butterfly may be released. Such captured and released butterflies have beak marks (missing parts) on their wings. Smith showed that the beak marks are most common in butterfly species known to be distasteful. His work indicates that death of prey is not a necessary corrallary of the system. Thus distastefulness may be beneficial to the individual prey.

Many distasteful prey are brightly colored rather than cryptic. From experimental studies it appears that a predator can learn to avoid and can avoid for longer periods contrastingly colored unpalatable prey more readily than cryptic unpalatable prey.

Gittleman and Harvey (1980. Nature 286:149-150) treated chick food with quinine to make it distasteful and dyed it blue or green. Young domestic chickens were then placed in pens with blue/green floors. In initial trials the birds took more items that contrasted with the background color but then avoided them in later trials, but continued to take distasteful baits that matched the background. Thus, the chicks learned to avoid conspicuous distasteful prey more readily than cryptic distasteful prey.

Sillen-Tullborg (1985. Oecologia 67:411-415) presented cryptic (grey) and aposematic (red) forms of a distasteful bug to hand-reared Great Tits; the birds learned to avoid both prey types, but they learned to avoid aposematic forms more quickly and had a lower frequency of killing in an attack when the prey were aposematic.

Mimicry

Alcock discusses Batesian and Mullerian mimicry. Batesian mimicry, named for the English naturalist Henry Bates in 1864 is the resemblance of a palatable prey species to a noxious prey species, i.e., predator deception by edible or undefended species. Probably every species possessing strong repellants or toxins in resembled by one or more palatable species.

Strong resemblances to noxious species provide the best protection for Batesian mimics. Yet Batesian mimics are not necessarily immune from predation, and a few predators even specialize on mimics. Two wasp species (in Africa and Malaysia) eat noxious ant-mimicking salticid spiders extensively and apparently form very specific search images for them. Yet, most predators are unable to distinguish palatable and noxious look-alikes and tend to avoid both.

The effectiveness of Batesian mimicry depends on how noxious the model is and how abundant it is relative to the mimic. As mimics become more common, more palatable prey will be encountered by predators, which consequently will take longer to learn the association between the warning coloration and the chemical defense of the prey. Also, as the mimic becomes more common, it becomes more worthwhile for the predator to distinguish subtle differences between the mimic and the model. Thus, Batesian mimicry is most effective when the mimics are relatively rare.

Mullerian mimicry is the evolutionary convergence in aposematic signals of two or more noxious species. The benefit would be mutualistic to members of both species, and fewer of both would be sampled by predators before the predators learn to avoid them.

Signalling by prey to predators

Alcock discusses stotting (pp. 366-368) in gazelles. Stotting in gazelles and tail-flagging in white-tailed deer are being interpreted as signals to predators. While the data are consistent with this interpretation, other possiblities cannot be completely ruled out. For example, Tim Caro discounted stotting as a means of surveying the landscape for other predators because gazelles stott both in short grass (where they could easily see predators without stotting) and in long grass (where stotting would probably help them see other hidden predators). Since they stott in both tall grass and short grass, Caro concluded that the gazelles do not stott to better survey their surroundings for hidden predators. However, if stotting is highly beneficial for detecting predators in tall grass and little or no cost in short grass, maybe they end up doing it both because on average it is beneficial. We should not expect adaptations to work perfectly in every situation--they should work best in the most important situations.

Most of the studies of tail-flagging in white-tailed deer have examined the responses of solitary and grouped deer to a human (investigator) or a human with a dog. Generally investigators have concluded that the deer is signalling its awareness of the intruder(s). One idea that appealed to me is that the deer may tail-flag as it goes into cover, providing a conspicuous signal, before making abrupt turns once the predator can no longer see it.

Back to lecture notes index.

Back home.