How Do Foragers Assess Danger

No living forager can really know its odds of being killed by a predator. Therefore, we should not expect foragers to make accurate estimates ofdanger, but the details oftheir estimates may have profound effects on their behavior and ecology (Sih 1992; Houtman and Dill 1998; Brown et al. 1999). What they might know and how they might know it are largely open questions. We can divide these questions into subquestions by remembering that predation involves encounter, attack, and capture stages. Since each stage requires the previous one, foragers generally experience many more encounters than attacks, and more attacks than captures. Here I present some examples of how different foragers may estimate the probabilities of encounter, attack, and capture.

Gerbils in the Negev Desert in Israel react to noises that indicate the presence ofbarn owls (Abramsky et al. 1996). Prey may know the general density of predators by detecting signs of predators. Part of this information comes directly from the inevitable sights, sounds, and smells that predators create. Furthermore, predators often betray their presence through their territorial behavior or other social interactions. Lions roaring and wolves howling are two examples that are obvious even to humans. Among seabirds, petrels limit their exposure after eavesdropping on the territorial calls of predatory skuas (Mougeot and Bretagnolle 2000).

When potential predators are nearby, foraging animals must assess the likelihood ofattack. For fish and other aquatic organisms, chemical cues may provide detailed information on the capture of similar prey in the area (see Kats and Dill 1998; Wisenden 2000). Gerbils and other small rodents behave more cautiously on moonlit than on dark nights (Daly et al. 1992; Vasquez 1994). Although moonlight probably helps rodents detect attacks, it seems to help predators more in detecting prey. Moonlit nights are dangerous because ofthe increased probability ofattack should predator and prey encounter each other.

Finally, how can a foraging animal assess its likelihood of escape if it were attacked? Likelihood of escape is intimately linked to detection behavior. We can even define detection to mean detection of an attack in time to avoid capture (see Bednekoff and Lima 1998b). Time needed to escape is a basic variable ofes-cape that animals might assess. Time needed to escape depends on the distance and the structure of the habitat between the forager and safety (see Blumstein 1992). As mentioned earlier, small birds forage differently when farther from a refuge. The same applies to burrowing animals such as marmots, except that the refuge is a burrow rather than a bush or tree. Townsend's ground squirrels flee less quickly through shrub habitats than across open ground (Schooley et al. 1996), although shrub vegetation may also make predators harder to spot. Fox squirrels also react to escape substratum (Thorson et al. 1998).

Overall, foragers can respond to direct cues to attack rate or conditions that indirectly give cues to relative danger levels. Animals react to factors that are likely to affect their probability of escape. Differences in the probability ofescape may drive many reactions to small differences in exposure. Because probability of escape after capture is difficult to estimate from experience, I

Figure 9.7. The symmetry of the overall fitness function with z, the exponent of the curve relating foraging effort to predator detection. This relationship determines whether prey should over- or underestimate predation risk. (k = 4, k = 1.)

0 0.2 OA 0.6 CLP I.C Foraging effort

Figure 9.7. The symmetry of the overall fitness function with z, the exponent of the curve relating foraging effort to predator detection. This relationship determines whether prey should over- or underestimate predation risk. (k = 4, k = 1.)

speculate that learning about escape probability is less important than learning about probabilities of encounter or attack.

Was this article helpful?

0 0

Post a comment