vole population’s temporal distribution of activity. A uni-
form level of activity across time may thwart the predators
by precluding predictable periods of very high and very low
activity. A further reduction in activity synchrony in frag-
mented habitats may further overdisperse activity. Alterna-
tively, other factors may primarily influence the circadian
rhythm of microtines, with the predators adjusting their be-
havior around that of the voles (Halle 1993).
Activity and moonphase
Among nocturnal rodents a strong pattern of moonlight
avoidance has been documented experimentally (Kotler
1984; Brown et al. 1988) and via field data (Lockard and
Owings 1974; Price et al. 1984; Rosenzweig 1974). White-
footed mice, in an experiment, adjusted their behavior to
both moon phase and the crunchiness of their leaf substrate
(Fitzgerald and Wolff 1988). Moonlight seems to favor
owls over their rodent prey. The pattern of mortality, how-
ever, may be complex. Radio-tagged Merriam’s kangaroo
rats suffered higher predation mortality rates from typically
diurnal predators (e.g., shrikes) during full moon nights but
higher predation mortality from typically nocturnal preda-
tors during the new moon (Daly et al. 1992). This seem-
ingly odd pattern of predation risk may be a consequence
of the foraging game between rodents and their predators
(Brown et al. 2001).
With the full moon, kangaroo rats may reduce activity,
remain in safer microhabitats, and become more crepuscu-
lar and thus more exposed to diurnal predators. Bannertail
kangaroo rats (Dipodomys spectabilis) show more activity
at new than full moon, range farther from their burrows,
and show equal or more crepuscular activity on full moons
compared to moons (Rosenzweig 1974). In aviary experi-
ments with owls and heteromyid rodents, more rodents
were captured on dark nights than nights with artificial il-
lumination (Kotler et al. 1988). This apparently anomalous
result can be explained through activity: rodents were much
less active on nights with illumination than without. When
active, however, rodents suffered much higher mortality
rates per unit of foraging activity on nights with illumina-
tion. Gerbils at a Negev Desert site exhibit the greatest ap-
prehension toward predators early in the night and near full
moon (Kotler et al. 2002). Interestingly, apprehension in-
creases again near dawn, in accord with Daly et al.’s (2000)
suggestion that crepuscular activity incurs risk from shrikes.Experiments with giving-up densities
Giving-up densities in natural or experimental food patches
provide an effective tool for measuring spatial and tempo-
ral variability in rodents’ perceived predation risk (fig. 28.2;
see also Brown and Kotler 2004). When presented with
a depletable food patch, a rodent should forage less thor-Fear and the Foraging, Breeding, and Sociality of Rodents 331Figure 28.2 Seed trays such as this are used for experimentally measuring giving-up densities (GUDs) in response to
predation risk. The tray and burrow are on semistabilized sand dunes at Bir Asluj, Negev Desert, Israel. One night of
foraging at this patch revealed spoor from the Egyptian sand gerbil, Gerbillus pyramidum,Allenby’s gerbil, G. a. allenbyi,
and the common jerboa, Jaculus jaculus. Photo by Joel Brown.