ties to interact only with kin before they encounter other
conspecifics, then “familiarity” is a reliable proxy for re-
latedness (Bekoff 1981). In Spermophilus,young develop
in underground nests, one litter and one mother per nest,
which provides a developmental-ecological explanation
(West et al. 2003) for the effects of prior association on lit-
termate recognition in S. beldingiand S. parryi(Holmes
and Sherman 1982), S. richardsonii(Sheppard and Yoshida
1971; Davis 1982b; Hare 1998b),S. lateralis(Holmes 1995;
Mateo 2002) andS. tridecemlineatus(Holmes 1984b), and
mother-offspring recognition in S. richardsonii(Michener
1971, 1974) and S. beldingi(Holmes 1984b). In various
other species of rodents, both littermate recognition (e.g.,
Kareem and Barnard 1982; Sun and Müller-Schwarze 1997;
Todrank et al. 1998; Mateo and Johnston 2000b) and
mother-offspring recognition (e.g., Hepper 1987; Polan and
Hofer 1998; Yamazaki et al. 2000) are mediated by the
prior association mechanism.
We emphasize that the familiarity produced by prior as-
sociation is adescriptiveterm that only indicates whether in-
dividuals have encountered each other previously, and thus
has limited explanatory value. For example, the descriptor
does not identify a mechanism like perceptual learning or
associative learning, which might generate familiarity (Hep-
per 1986; Alexander 1990), it does not indicate when dur-
ing development familiarity is established (Jackel and Trill-
mich 2003), nor does it suggest whether familiarity, once
established, is relatively permanent (Kendrick et al. 2001;
Insley 2000) or requires repeated exposure to former asso-
ciates to be maintained (Porter and Wyrick 1979; Paz y
Miño C. and Tang-Martínez 1999a). Following prolonged
social isolation, for instance, S. beldingiretain the ability
to recognize previously familiar kin, but appear unable to
recognize previously familiar nonkin (Mateo and Johnston
2000b). Demonstrating that learned familiarity mediates
differential treatment is a necessary first step in understand-
ing how kin are recognized by prior association, but, as we
note earlier in this paragraph, many questions must be an-
swered to achieve a full understanding of how this mecha-
nism operates (Porter et al. 1984; Holmes 2001).Kin Recognition in Rodents: Issues and Evidence 221Figure 19.3 Laboratory (A and B) and field (C) data on littermate recognition (A and B) and mother-offspring recognition
(C) in Belding’s ground squirrels. Laboratory paired-encounter tests with cross-fostered animals revealed that pairs comprising
familiar (reared together) young were significantly less agonistic than pairs comprising unfamiliar (reared apart) young (A). How-
ever, when unfamiliar pairs were categorized by sex-of-pair, pairs of unfamiliar littermate females were significantly less agonis-
tic than pairs of unfamiliar nonkin females, which demonstrate that female S. beldingican recognize their unfamiliar kin (B). Fi-
nally, free-living mothers readily retrieved into their natal burrow unrelated young placed at their burrow entrance, but only until
their own young reached about 25 days of age, which coincides with when mothers’ own young emerge aboveground from
their natal burrow (C). From Holmes and Sherman (1982).