The production component—Odors as cues to relatedness
Although it is widely recognized that odors can mediate
kin recognition in Rodentia (Mateo 2003), little is known
about howodors convey the information that allows ani-
mals to recognize their relatives. Odors serve multiple func-
tions in rodents, such as indicating an animal’s individual
identity, sex, relatedness, genetic quality, health, reproduc-
tive status, or even location (Brown and MacDonald 1985;
Halpin 1986; Penn and Potts 1999). Assessment of related-
ness via odors may be facilitated by the major histocom-
patibility complex (MHC), a large group of genes involved
in vertebrate immune function, which also influence odor
production. These highly variable genes, some having over
100 alleles per locus, regulate recognition of self versus
non-self tissues and pathogens. Because of the large number
of alleles involved, the likelihood of two random conspe-
cifics having the same MHC haplotypes is very low, but
since relatives, by definition, share many genes in common,
the MHC can serve as an accurate correlate of relatedness,
which could be important to nepotism. In addition, the ex-
treme variability of the MHC is thought to facilitate an in-
dividual’s ability to resist a broad range of rapidly evolv-
ing parasites and pathogens, and therefore could be used in
mate-choice contexts to reduce inbreeding and promote
heterozygosity in offspring if adults prefer MHC-dissimilar
mates (Penn 2002). Besides coding for glycoproteins, the
MHC also influences odor production, likely through an in-
teraction between MHC by-products and bacteria on gland
surfaces (leading to unique secretions) or in gastrointestinal
tracts (creating unique urine and fecal odors; Schellinck
et al. 1997). Thus families share similar MHCs and similar
odors (Boyse et al. 1991; Brown and Eklund 1994), which
means that MHC-derived odors could serve as kin labels,
which may be learned from family members during early
development (Yamazaki et al. 1988, 2000; Penn and Potts
1998b).
Laboratory rodents (rats and mice) can discriminate be-
tween the odors of highly inbred animals that are geneti-
cally identical except for oneMHC allele (Beauchamp et al.
1985; Brown et al. 1990). In mate-choice trials, they often
prefer to mate with MHC-dissimilar individuals, although
preference varies depending on the sex of the choosing in-
dividual and the methods used in the trials (Manning et al.
1992a). Thus besides conferring immune protection, the
MHC may facilitate mate choice (Potts et al. 1991; Carroll
and Potts, chap. 5, this volume). It is unknown, though,
whether MHC-derived odors mediate recognition, kin pref-
erences, mate choice, and reproductive success in animals
that are notinbred. We note, however, that Manning et al.
(1992b) studied the nesting habits of female house mice re-
sulting from crosses between wild-caught mice and inbred
strains, and found that females tended to nest commu-
nally with MHC-similar females. Conesting reduced infan-
ticide compared with single-nesting females, and resulted in
greater direct- and indirect-fitness benefits for conesting fe-
males (Manning et al. 1995; Dobson et al. 2000). Further,
MHC-based odors of inbred rodents can be influenced by
diet and health status (Schellinck et al. 1997; Ehman and
Scott 2001), yet we do not know if free-living animals at-
tend to genetic differences among odors that also reflect for-
aging habits and infectious conditions. Field data on MHC-
based odors, including how they develop and their effects
on social interactions, are needed to test functional hy-
potheses about the role of these odors in nepotism and mate
choice.
Much of the research on the production component in
rodent kin recognition has emphasized MHC-based odors,
but recent work has shown that another urinary compound
warrants attention because it influences chemical identity
signals and may be involved in social recognition. Major
urinary proteins (MUPs) are large, abundant, nonvolatile
compounds that bind and carry the more volatile odor mol-
ecules that are traditionally thought to mediate recognition.
The large molecules, which are costly to synthesize, extend
the longevity of an odor (Hurst et al. 1998), lengthening its
functional life, which may be especially important for ani-
mals that cannot demarcate and defend all areas of their ter-
ritory simultaneously. MUPs may also be involved in signal-
ing individual identity, as a male house mouse excretes urine
with 10 –12 different genetically encoded MUPs, which can
be used by other mice to discriminate among individuals.
MUPs appear to express distinct cues themselves, as the
nonvolatile compounds are sufficient for individual recog-
nition (Nevison et al. 2003; Thom and Hurst 2004). Thus
in mice, and perhaps other rodents, MUPs and MHC by-
products contribute to the kin-distinct urine odors (Hurst
et al. 2001; Thom and Hurst 2004).Production cues from other modalities
Most examples of rodent kin recognition implicate con-
specific discrimination via olfactory cues (Mateo 2003), in
part because odors vary with genetic relatedness and thus
offer evolutionarily reliable cues to assess kinship (see also
Todrank and Heth 2003). In theory, relatedness could be
assessed through other modalities, such as vibrational sig-
naling (Randall 2001), visual-behavioral cues (Michener
1973a; Terrazas et al. 2002), or vocalizations (Hare 1998a),
but if such cues were to mediate recognition by phenotype
matching then these production cues must covary with re-
latedness. Nonolfactory kin labels might be important if
odor differences were unreliable or inaccurate indicators of
relatedness, such as when a nest of social insects containsKin Recognition in Rodents: Issues and Evidence 225