inS. lateralismay have functional value in the context of
mate choice, but they appear to have no value in the context
of nepotism (Mateo 2002). Current evidence for ground
squirrels and other rodents as well provides little support
for the belief that simple differences in social organization
can be used to predict fine-grained differences in recogni-
tion abilities (Mateo 2003). This contrasts with the situa-
tion in some birds, in which colonial species have a more
complex and elaborate recognition system than noncolo-
nial species (e.g., Beecher 1991).
The recognition behavior of some ground-dwelling squir-
rels appears paradoxical in light of evolutionary theory
(Hamilton 1964) or, as Hoogland (1995) writes about prai-
rie dogs, “The nondiscriminatory aspects of kin recogni-
tion among prairie dogs seems puzzling” (p. 215). Both
black-tailed prairie dogs (C. ludovicanus;Hoogland 1986,
1995) and Gunnison’s prairie dogs (C. gunnisoni;Hoogland
1996b) distinguish between kin and nonkin, as revealed by
differences in the frequency of amicable interactions (C. lu-
dovicanus) and the likelihood of giving alarm calls (C. gun-
nisoni). However, field data show that prairie dogs fail to
treat close kin and more distant kin differently despite the
fitness benefits that would seem to accrue if such discrimi-
nation were made. For example, femaleC. ludovicanustreat
distant kin, such as female full (r 1/8) and half (r 1/16)
first cousins, just as amicably as they treat their daughters
(r 1/2) and full (r 1/2) sisters (Hoogland 1995). To
evaluate kin discrimination independentlyof kin favorit-
ism, it would be useful to use a neutral measure like differ-
ences in olfactory investigation to determine whether prai-
rie dogs can discriminate between close and more distant
kin, even though nepotism is not directed differentially to-
ward these kin classes in nature. Prairie dogs provide a re-
minder that close kin may be intense competitors (Hoog-
land 1985) and thus inappropriate targets for nepotism,
because they do not satisfy Hamilton’s rule. Accordingly,
it is risky to assume that the absence of nepotism implies
an inability to recognize kin (fig. 19.1; Beecher 1991), even
though there is often a positive association between kin rec-
ognition and nepotism (e.g., Griffin and West 2003).
The kin-recognition abilities of Columbian ground squir-
rels may also be unusual, at least according to some inves-
tigators (Hare 1992; Hare and Murie 1996).S. columbianus
are the most social members of the genus Spermophilus
(Michener 1983a) and display kin-differential behavior in
nature (Waterman 1988; King 1989a, 1989b; Wiggett and
Boag 1992). Juveniles, for example, encounter both their
littermates and unrelated juveniles during the first 3 weeks
after emerging from their natal burrow, and play more of-
ten with their littermates than with nonlittermates (Water-
man 1986), which indirectly indicates that juveniles can
discriminate between these two kin classes. Hare (1992),
however, disputed the claim that juvenile S. columbianus
could recognize their littermates and, following Grafen
(1990), argued that what appears to be kin (littermate) rec-
ognition may be an artifact of colony member (familiar) ver-
sus noncolony member (unfamiliar) discrimination. Hare’s
recognition studies included juveniles of unreported ages
(Hare 1992) or juveniles that had been aboveground at
least 4 weeks before recognition tests were done (Hare and
Murie 1996). We believe that to reveal the kin-recognition
abilities of juvenile S. columbianus,including how these
might change ontogenetically, it would be valuable to use a
neutral bioassay (e.g., olfactory investigation) and conduct
a longitudinal study, beginning soon after natal emergence
and continuing throughout the juvenile summer, to better
assess their skills in social recognition. That juveniles play
preferentially with littermates (Waterman 1986) suggests
that kin can be recognized even if nepotism is not com-
mon in adult S. columbianus(Hare and Murie, chap. 29
this volume).Sex differences in kin recognition
Ground squirrel social organization is based on female kin
groups; females are nepotists, whereas males are not (Mich-
ener 1983a; Murie and Michener 1984). Thus it seemed
that kin-recognition abilities might differ between the sexes,
which was initially confirmed in S. beldingi(Holmes and
Sherman 1982). In paired-encounter tests, females dis-
criminate between unfamiliar (reared apart from) litter-
mates and unfamiliar nonlittermates, whereas males do not
(fig. 19.3b), and females, but not males, discriminate be-
tween unfamiliar paternal half-siblings and unfamiliar non-
kin (Holmes 1986a). In subsequent studies, however, using
odor-discrimination tests rather than paired-encounter tests
that monitored agonistic interactions, both females and
males produced kin labels and both sexes displayed similar
discrimination abilities in response to these labels (Mateo
and Johnston 2000a, 2000b; Mateo 2002, 2003). We can-
not explain the apparent disjunction between sex differ-
ences in nepotism and the absence of such differences in kin
discrimination. However, we encourage investigators to an-
alyze their recognition data for a sex effect whenever nepo-
tism is sex limited, and to recall that kin recognition might
serve one function in females (e.g., mediating nepotism)
and another in males (e.g., mediating mate choice).Pressing Questions and IssuesWe now examine some intriguing issues surrounding ro-
dent kin recognition, and explain why pursuing them will
pay dividends for understanding recognition at both the
proximate and ultimate levels of analysis.224 Chapter Nineteen