have two types of alarm calls, which elicit two different be-
havioral responses by listeners. Whistles, which are single,
nonrepetitive high-frequency notes, are elicited by fast-
moving, typically aerial, predators and result in evasive be-
haviors such as running to or entering a burrow, and scan-
ning the area only after reaching safety. Trills, which are
composed of a series of five or more short notes, are elic-
ited by slow-moving, primarily terrestrial predators, and
usually cause others to post (a bipedal stance accompa-
nied by visual scanning), with or without changing location
(Sherman 1985; Mateo 1996a). When juvenile S. beldingi
first emerge aboveground at about 1 month of age, they do
not discriminate behaviorally between these calls, or even
among alarm calls and other conspecific and heterospecific
vocalizations. It takes about a week for juveniles to learn to
respond selectively to alarm calls and to show the correct
responses to each type of call, during which time up to
60% of juveniles may disappear, many to predation. This
learning is facilitated by experience hearing the calls as well
as observations of adults’ reactions. Juveniles attend spe-
cifically to their mother’s responses over those of other fe-
males, and tend to adopt her particular response style (e.g.,
more reactive, stay alert longer; Mateo 1996a; Mateo and
Holmes 1997).
Some ground-dwelling squirrels show a significant de-
gree of behavioral plasticity in their survival behaviors. I de-
fine plasticity as the modifiability of behavior as a result of
specific experiences or environmental conditions (e.g., Mil-
ler 1981). The plasticity of alarm-call responses in S. bel-
dingilikely reflects temporal and spatial variation in their
habitats and predators (e.g., meadow versus forest, open
versus closed habitat; presence or absence of snakes or wea-
sels; natal or breeding areas). By acquiring responses that
are appropriate for a given microhabitat, S. beldingican op-
timize both their foraging and antipredator efforts, allow-
ing juveniles (and adults) to gain adequate body weight be-
fore hibernation without expending energy on unnecessary
vigilance (Mateo and Holmes 1999a). Experiences prior to
natal emergence, such as hearing alarm calls or exposure
to maternal stress hormones, can have long-term effects on
the response repertoires juveniles exhibit (e.g., likelihood of
responding, duration of alert behaviors). Yet similar experi-
ences after emergence do not alter juvenile responses signifi-
cantly, suggesting that preemergent experiences prime young
to acquire antipredator behaviors appropriate for the local
predator environments into which they will soon emerge
(Mateo 1995; Mateo and Holmes 1999b). This develop-
mental plasticity is not surprising given that predation can
be such a potent selective agent on both juveniles and adults.
As an example of plasticity in survival behaviors, juve-
nile S. beldingireared at the edge of meadows are more re-
active to alarm calls and remain alert longer than those
from the center of a meadow (Mateo 1996a), which may
reflect increased vulnerability to predators near the edge
(Elgar 1989). The center-edge effect demonstrates plasticity
in the expression of alarm-call responses by both mothers
and juveniles according to spatially and temporally chang-
ing contexts. Juveniles may develop location-dependent re-
sponses by monitoring their mother’s behavior and using
it as a model for their own (Mateo and Holmes 1997).
Females often nest near their mothers in subsequent years
(J. M. Mateo, unpublished data), so adopting location-
specific responses (e.g., increased vigilance by the meadow’s
edge) would be favored. Mothers’ reactions, which serve
as models for juvenile responses, may reflect the mothers’
own vulnerabilities (indirect parental influence), or may be
a form of maternal care, becoming more vigilant if they
locate their natal burrow, and thus their offspring, in a
dangerous area (edge) and less vigilant if in a safer region
(center; direct parental influences). However, conditioning
studies show that adult Spermophiluscan acquire new anti-
predator responses (Mateo 1995; Shriner 1999), so if indi-
viduals disperse to new areas they can adjust their reper-
toires to respond appropriately for their new local predator
environments.
What mechanisms might mediate this behavioral plastic-
ity? As mentioned previously, juvenile ground squirrels are
preyed on by many species, and juvenile emergence tends to
draw predators. Young likely experience stress during this
time, either chronically or acutely. In addition, juveniles are
exposed directly to maternal stress hormones, both pre-
and postnatally. In lab rodents the influence of glucocorti-
coids on learning and memory has an inverted-U-shaped
function. Very low or very high levels of corticoids can
lead to hypo- or hyperarousal and poor selective attention
to input, and thus impair consolidation of new memories.
Moderate levels of corticoids are optimal for attention to
and consolidation of memories (reviewed in Lupien and
McEwen 1997). It follows then that stress-hormone lev-
els around the age of natal emergence may facilitate rapid
learning of appropriate responses, especially in locations
with high predation pressure. Indeed, adult corticoid levels
vary by location and juvenile levels are elevated significantly
around the time of natal emergence (J. M. Mateo, unpub-
lished data). These trends are consistent with my hypothe-
sis that HPA activity reflects geographic variation in stress
due to predation pressure and mediates learning of anti-
predator behaviors (and other survival behaviors) by devel-
oping ground squirrels.Rates of development and acquisition of behaviors
If newly emergent juveniles are highly susceptible to preda-
tion (e.g., Sherman and Morton 1984; Coss et al. 1993;200 Chapter Seventeen