untitled

food items are rarely mixed homogeneously across the landscape visited by the for-

ager. This is particularly true of mobile prey which may be attempting to avoid pre-

dators. If alternative prey species tend to occur in different microhabitats, differ in

their activity patterns, or have a different capacity for avoidance of predators, then

simple frequencies of abundance may be a poor predictor of diet composition. In these

cases, more complex models may be needed to predict optimal diet patterns.

In some cases, particularly herbivores, foragers need to maintain a balanced intake

of particular nutrients, rather than simply maximizing energy gain in whatever way

possible (see Chapter 4). For example, howler monkeys tend to choose a diet more

heavily laden with leaves than should be optimal, perhaps because they must balance

nutrients (Milton 1979). Similarly, the marine gastropod Dolabella auriculariagrows

several times faster on a mix of algal species than when fed on a single species of

algae (Penning et al. 1993).

The optimal diet model represents a simplistic view of foraging. Most species have

biological features that introduce additional elements into the decision-making pro-

cess. For example, many species forage outwards from a central place, whether that

place is a den, perch, or resting site. If the forager sallies forth, retrieves one or more

prey items and then returns to the central place before feeding on the items, then

this additional travel time and energetic expenditure must be accommodated to make

useful predictions (Orians and Pearson 1979). Thus, beavers forage on a variety of

woody plants on the shore surrounding the ponds or streams where they build their

lodges. Several studies have shown that beavers feed more selectively the further out

food items occur from the lodge, as predicted by central place foraging models ( Jenkins

1980; McGinley and Witham 1985; Fryxell and Doucet 1993). The greater energetic

cost of travel to more distant food requires that animals be more selective. As a

consequence, the handling time of potential food items is dependent on distance,

implying different patterns of diet selectivity.

Optimal patterns of foraging can have important effects on the rate of attack by

consumers. The rate of consumption of prey type 1, f(N 1 ), by an optimal forager is

predicted by the following multispecies functional response:

THE ECOLOGY OF BEHAVIOR 63

100

90

80

70

60

50

40

0 0.05 0.10 0.15 0.20

Rate of encounter with big prey

% Big prey in diet

Fig. 5.2Krebs et al.’s

(1977) laboratory

experiment on diet

selection by Parus

major. Two sizes of

mealworms were

delivered via conveyor

belt in random order to

the forager. The

horizontal axis shows

the rate of encounter

with big prey. As the

rate of encounter with

big prey declined, the

birds expanded their

diet to include small

prey.

5.2.2Optimal diet

selection: effects on

feeding rates

WECC05 18/08/2005 14:42 Page 63