means for suppressing the accumulation of scent signs for
searching predators, lemmings (Dicrostonyx groenlandicus)
defecate in special underground latrines during the summer,
but not during the winter when snow protects their outdoor
toilets (Boonstra et al. 1996).
Mobility and scent marking are interconnected by sexual
advertisement. Despite the negative result of Ylönen et al.
(2003), several studies demonstrate a connection between
rodent scent marking and increased risk of predation. High-
marking mice (Mus musculus) decreased marking intensity
when exposed to ferret (Mustela putorius) scent, but not
when exposed to scent of naked mole-rats (Heterocephalus
glaber) in a laboratory experiment (Roberts et al. 2001).
In an unpublished study, D. Dudek and H. Ylönen failed to
find any difference in urine marking of male bank voles ex-
posed to either weasel scent or moose (Alces alces) scent as
a control. Similarly, Wolff (2004) found that prairie voles
(Microtus ochrogaster) did not decrease scent marking in
response to odors of mink (Mustela vison) or a bullsnake
(Pituophis melanoleucus) in the field or the laboratory. In
the field, however, artificially increasing vole scent mark-
ing led to increased predation and lower survival of radio-
collared voles (Koivula and Korpimäki 2001).
Acoustical and auditory cues
To their predators, nocturnal desert rodents likely emit
acoustic signatures, and perhaps olfactory, visual, and —
to pit vipers —heat signatures as well. Barn owls (Tyto alba)
are thought to use acoustic cues to detect rodents at a dis-
tance and then to use visual cues to fine-tune strikes. Canids
(Canis sp., Vulpes sp.) likely follow the scents of rodents
in addition to responding to visual and acoustical cues. The
presence of fox scent raised the GUDs of Merriam’s kan-
garoo rats, although this effect was only significant in the
winter and in the bush microhabitat (Herman and Valone
2000). Snakes may use olfactory cues to determine areas
with higher likelihoods of encountering rodents (Duvall and
Chiszar 1990). Sand vipers in Israel are known to stake out
the burrow entrances of gerbils. In an experiment with the
Australian house mouse (Mus musculus), mice did not uti-
lize seed trays closer than 20 m to rabbit nests occupied by
a brown snake (genus species). The absence of mice from
these trays was due either to direct predation by the snake
or, more likely, to the fear caused by the snake’s presence
(Ylönen et al. 2002).
In response to predators, nocturnal desert rodents may
rely heavily on hearing to detect owls. Relative to other ro-
dent taxa, nocturnal rodents often possess inflated auditory
bullae, which aid in the detection of low-frequency sounds
such as those emitted by owls in flight (Webster and Web-
ster 1971). Kotler (1984, 1985) found that a species’ use of
the open microhabitat increased with the size of its auditory
bullae. Furthermore, in samples collected from owl pellets,
the owls’ diets favored heteromyid species with smaller au-
ditory bullae.
Variation in noise level is one of the most striking fea-
tures of rodent-owl experiments in aviaries. Nights with-
out owls provide a cacophony of sounds. Kangaroo rats
(Brown et al. 1988) or gerbils (Kotler et al. 1991) can be
heard noisily scurrying from brush pile to brush pile. In ad-
dition, clearly audible sound signatures reveal the digging
and scratching activities of rodents within the food patches.
A “deathly” stillness descends over the aviary during nights
with owls. The overall silence is punctuated on occasion
by the crash of owl into a brush pile and the sudden noisy
scurry of an escaping rodent. Then all is silent again. Yet, in
the morning, even when owls have been present, footprints,
evidence of digging, and seed removal from food patches
indicate considerable rodent activity. While these findings
have never been quantified, it appears in these aviary ex-
periments that the amount of sound emitted per seed har-
vested is much lower on nights with owls than on nights
without owls.
Auditory cues probably influence predation rates on
boreal voles as well. Trapping success of voles typically in-
creases with rainfall in both forest and field habitats (Si-
dorowicz 1960; Ansorge 1983). Similarly, field and labora-
tory studies with Peromyscusshowed that animals are more
active on rainy nights or on wet substrate than on dry nights
or on leaves that are crunchy (Fitzgerald and Wolff 1988).
The conclusions from these studies are that rain dampens
auditory cues emitted by foraging rodents, cues that are
used by predators to locate their prey.Fear- versus mortality-driven predator-prey systems
Boreal voles and desert rodents can have similar preda-
tors (owls, foxes, and snakes), and they show behavioral
response to predation risk that involves microhabitat selec-
tion and shifts in activity patterns. But here the similari-
ties may end. Predator-prey interactions are population-size
driven (N-driven) when the prey pay the cost of predation
by high fecundity and feeding rates. Classical predator-prey
models apply to N-driven systems in that doubling the num-
ber of predators doubles the prey’s predation risk. Predator-
prey systems are driven by fear (m-driven) when prey opt
to forgo feeding and breeding opportunities in the face of
predators. Increasing the numbers of predators results in
more vigilant or less active prey that are harder to catch;
doubling the number of predators does not double the
prey’s predation risk. N-driven systems are likely to see rela-334 Chapter Twenty-Eight