A
fter detecting a predator,individuals of
numerous avian and mammalian species give alarm
calls that warn conspecifics. But why should an in-
dividual incur the cost of attracting a predator’s attention to
itself by vocalizing (table 37.1)? This question has puzzled
evolutionary and behavioral ecologists for 50 years (see
reviews in Sherman 1977, 1985; Blumstein, chap. 27 this
volume). Breakthroughs came when Christopher Dunford
(1977a) and Paul Sherman (1977, 1980a) discovered that
alarm calls of round-tailed (Spermophilus tereticaudus) and
Belding’s ground squirrels (S. beldingi), respectively, func-
tion primarily to warn kin, especially juvenile offspring.
Similar results soon appeared for numerous other ground-
dwelling squirrels (Smith 1978; Yahner 1978b; Noyes and
Holmes 1979; Schwagmeyer 1980; Davis 1984a; Owings
et al. 1986; MacWhirter 1992). To explain alarm calling,
some behavioral ecologists (e.g., Shields 1980; Blumstein
et al. 1997) have focused on the importance of direct selec-
tion versus indirect selection (Brown 1987)— i.e., calling to
warn offspring and grandoffspring versus calling to warn
nondescendent kin. Other behavioral ecologists (e.g., Sher-
man 1980b; Hauber and Sherman 1998; see also Dawkins
1979) have de-emphasized this dichotomy. Either way, we
can all agree that observing marked individuals of known
genealogies is pivotal for a better understanding of the
adaptive significance of alarm calling. In this chapter I sum-
marize information on alarm calling for three species of
prairie dogs. As expected, kinship is once again a big part
of the story, but I discuss two other factors that also affect
alarm calling: demography and obesity. Demography is an
ultimate factor that also has been important in the evolution
of alarm calling among Belding’s ground squirrels (Sherman
1977, 1980a). Obesity is a proximate factor, whose rele-
vance to alarm calling has not been previously considered.
Male reproductive success in most species varies directly
with the number of inseminations (Trivers 1972; Emlen and
Oring 1977), and multiple mating by males is widespread.
Sperm numbers from a single insemination are usually suf-
ficient to fertilize all a female’s ova (Bateman 1948; Petrie
et al. 1992; Birkhead 2000), however, so the observation
that females of so many species routinely mate with several
males is puzzling (Birkhead and Moller 1992; Kano 1992;
Solomon and Keane, chap. 4 this volume). Multiple mating
by females entails both costs and benefits (table 37.1). Eval-
uating these costs and benefits is difficult because it requires
information on both the copulatory and rearing success for
the same females. In this chapter I discuss two factors that
affect the frequency of multiple mating by female prairie
dogs: the ability of males to monopolize females, and the ef-
fect of multiple copulations on litter size.
Nonparental infanticide — i.e., the killing of a conspe-
cific’s juvenile offspring — is one of the most intriguing, con-
troversial, and misunderstood issues in behavioral ecology
and population biology (Hrdy 1979; Sherman 1981b; Haus-
fater and Hrdy 1984; Ebensberger 1998d; Ebensberger and
Blumstein, chap. 23 this volume). Infanticide involves both
costs and benefits (table 37.1), and seems to be especially
common and observable among rodents such as gray-tailed
voles (Microtus canicaudus;Wolff et al. 2002), ground-
dwelling squirrels (Sherman 1981b; Hoogland 1985), Mon-
golian gerbils (Meriones unguiculatus;Elwood and Oster-
meyer 1984b), and house mice (Mus musculus;vom Saal