sitive to the value of the contested resource (the pups; cf.
Parker and Rubenstein 1981). Relative to other adults,
females with young pups spend more time confronting
snakes and are more discriminating with regard to the level
of threat posed by the rattlesnake (Swaisgood, Owings, and
Rowe 1999; Swaisgood, Rowe, and Owings 1999; Swais-
good et al. 2003).
We have only limited data on the defensive behavior of
rattlesnakes while dealing with ground squirrels. But we do
know that rattlesnakes also proceed in ways generally char-
acteristic of social conflicts, defending themselves in risk-
sensitive ways while dealing with squirrels, experimental
squirrel models, and other sources of danger. For example,
rattlesnakes defend themselves less readily by rattling and
striking when colder body temperature renders them less
able to follow through on such a threat (Rowe and Owings
1990; Rowe and Owings 1996; Owings et al. 2002; also un-
published observations). On the offensive end, these rattle-
snakes are active exploiters of prey-related cues, choosing
hunting sites rich in both prey odors and microhabitat
features preferred by their prey (Theodoratus and Chiszar
2000). These snakes also use active probing in an apparent
quest for squirrel reactions that leak cues about whether a
female has nearby young. Mother squirrels resist such leak-
age, standing their ground while engaging in little addi-
tional activity until forced into it by a persistent rattlesnake
that may eventually get dangerously close to the burrow
containing the pups (Hennessy and Owings 1988).
The relation between ground-dwelling sciurids and mus-
telids of about the same size appears to provide another ex-
ample of relative parity between predator and prey. For ex-
ample, black-tailed prairie dogs effectively mob and harass
black-footed ferrets (Mustela nigripes) and plug the bur-
rows they invade (Henderson et al. 1974; Martin et al.
1984). Similarly, female Belding’s ground squirrels attack
and chase long-tailed weasels (Mustela frenata;Sherman
1977).
How Social and Antipredator Systems Acquire Their
Functional States, Linkages, and Similarities
Biological systems are adjusted via a variety of processes
that act in immediate, tonic, developmental, and evolution-
ary time frames (Coss and Owings 1985; Owings 1994).
These proximate and ultimate processes dovetail in the pro-
duction of an adaptive organism-environment relationship,
typically acting in complementary rather than alternative
ways (Cairns et al. 1990; West et al. 1994; Stamps 2003).
Proximate and ultimate contributions interact with each
other rather than working independently. For example, evo-
lutionary processes shape development through hetero-
chrony, that is, by changing the relative rates of develop-
ment in different systems (Gould 1977; Mason 1979). In
turn, developmental processes are central to generating the
heritable variation upon which natural selection acts (Bate-
son 1988; Cairns et al. 1990; Mateo, chap. 17 this volume).
A central feature of this latter point is that parents shape the
development of their offspring by providing not only genes
but also many reliable environmental contexts for develop-
ment (West et al. 1994). Such experientially induced effects
can either amplify or reduce the behavioral manifestation
of genetic variation. For example, allowing mice to have
only one social encounter can increase the expression of
variation in aggressiveness. This variation can be used in ar-
tificial selection experiments to generate very rapid diver-
gence of lines of high and low aggressiveness. Conversely,
these selected differences in aggressiveness can be masked
if each mouse has multiple social encounters, which gener-
ates increased aggressive activity in the low-aggression line
and therefore developmental convergence with the high-
aggression line (Cairns et al. 1990).
Proximate and ultimate processes also interact with each
other when the current structure of behavioral systems cre-
ates constraints and opportunities for subsequent evolution-
ary change. As a consequence, adjustment in biological sys-
tems has a cascading quality. Self-grooming in an agonistic
context by California ground squirrels provides a case in
point (Durant et al. 1988; Bursten et al. 2000). When males
of this species engage in a territorial boundary encounter,
agonistic contact may be interspersed with breaks during
which one or both individuals use forepaws and mouth to
self-groom in a highly stereotyped cephalocaudal pattern.
The focus of this grooming starts with the muzzle and moves
in a caudal progression across the head, down the body, and
along the tail to its tip. Such grooming is associated with
other noncontact agonistic activities, including tail pilo-
erection, staring at the adversary, and scent-rubbing against
stationary objects. Agonistic encounters containing such
grooming bouts are less likely to escalate to fighting than
those without grooming.
In other words, this cephalocaudal self-grooming has be-
come a social signal, perhaps through the following sce-
nario derived by Spruijt and colleagues (1992) from the lit-
erature on grooming. They argue that grooming originated
evolutionarily to care for the body surface in part by de-
fending against ectoparasite infestation (see also Hart et al.
1987). This defense requires that individuals not only groom
in immediate reaction to irritation from such sources, but
also engage in tonic grooming as prophylaxis against infes-
tation (programmed grooming; Hart et al. 1992). However,
tonic grooming does not produce the short-term payoff of
reduced irritation, and so requires a more endogenous form
of payoff to sustain it, such as arousal reduction and en-312 Chapter Twenty-Six