most individuals direct fitness declines with group size sug-
gests a compelling conceptual framework for exploring the
ecological reasons for group living in rodents.
Sociality in Subterranean Rodents
As the preceding sections of this chapter reveal, sociality is
a complex phenomenon that likely evolves in response to
multiple selective pressures. Given this complexity and the
diverse array of rodent societies reviewed in this volume,
developing a comprehensive model for understanding so-
ciality in these animals is challenging. Hence, as a starting
point for our efforts, we focus on analyses of sociality in
subterranean rodents — those species in which individuals
spend virtually their entire lives in underground burrows
that, typically, they excavate themselves (Nevo 1979; Lacey
et al. 2000). Subterranean species have played a significant
role in studies of rodent sociality for the past 2 decades
(Sherman et al. 1991; Lacey 2000; Faulkes and Bennett,
chap. 36 this volume), and data obtained from these ani-
mals have substantially influenced thinking on a number of
behavioral issues, including the importance of extrinsic de-
terminants of social structure and the nature and taxonomic
distribution of eusociality (Sherman et al. 1991; Lacey and
Sherman 1997, 2005). While most of the more than 120 ro-
dent species identified as subterranean appear to be solitary
(Nevo 1979; Reig et al. 1990), group living has been docu-
mented for members of two genera of bathyergid mole-rats
(Jarvis and Bennet 1991; Bennett and Faulkes 2000), one
species of ctenomyid (Lacey et al. 1997), and the one spe-
cies of octodontid that is truly subterranean (Spalacopus
cyanus;Reig 1970). Each of these families represents a dis-
tinct evolutionary origin for subterranean sociality and,
hence, comparative studies of these taxa provide important
opportunities to assess the generality of adaptive hypotheses
regarding group living (Bennett and Faulkes 2000; Lacey
and Wieczorek 2003; Lacey and Ebensperger, chap. 34 this
volume). Although the remainder of this chapter emphasizes
subterranean taxa, we believe that the conceptual frame-
work developed here for understanding sociality in subter-
ranean species is applicable to all group living rodents.
Overview of subterranean sociality
The best known of the social subterranean rodents are
the African mole-rats of the family Bathyergidae (Bennett
and Faulkes 2000). This family includes the naked mole-
rat (Heterocephalus glaber), which is one of the most social
mammals known (Jarvis 1981; Sherman et al. 1991; Ben-
nett and Faulkes 2000; Faulkes and Bennett, chap. 36 this
volume). Naked mole-rats live in groups of 200 to 300 in-
dividuals, with a mean group size of 80 animals (Braude
1991; Lacey and Sherman 1997). Groups form due to na-
tal philopatry by both sexes, such that colonies are typically
extended families. A colony can exist in the same area for
many years. Although natal dispersal occurs occasionally
(O’Riain et al. 1996; Braude 2000), it is apparently rarely
successful. Within an established colony, only one female
and one to three males reproduce; reproductive partners
are typically close kin (Reeve et al. 1990). Smaller, non-
breeding mole-rats participate in foraging and colony main-
tenance while larger nonbreeders engage in colony defense
and care of the offspring of breeders (Lacey and Sherman
1991, 1997). The striking parallels between the social sys-
tem of the naked mole-rat and those of social insects have
led to the assertion that cooperative breeding and eusocial-
ity form a continuum of societal structures (Sherman et al.
1995; Lacey and Sherman 2005). At the same time, studies
of naked mole-rats have revealed that both extrinsic (eco-
logical) factors that promote philopatry and intrinsic (ge-
netic) factors that favor cooperation with kin (Hamilton
1964) are important in maintaining such societies (Anders-
son 1984; Reeve et al. 1990; Jarvis et al. 1994; Bennett and
Faulkes 2000).
All species in the bathyergid genus Cryptomysstudied to
date also live in family groups (Bennett and Faulkes 2000;
Faulkes and Bennett, chap. 36 this volume). Although the
social structures of these species are generally similar to
that of naked mole-rats, none exhibits the extremes of col-
ony size or behavioral specialization found in H. glaber.
Within Cryptomys,the Damaraland mole-rat (C. damaren-
sis) is the most social. This species lives in groups of 2 –
40 individuals (modal group size of 12) that form due to
natal philopatry by both sexes (Bennett and Faulkes 2000).
Natal dispersal, however, does not appear to be as severely
constrained as inH. glaber,and the prevalence of philopatry
varies among populations (Spinks et al. 2000b). Within
a colony of C. damarensis,only a single male and female
breed; the remaining group members assist the breeders
in colony maintenance and defense activities, as well as in
rearing young. Unlike naked mole-rats, colonies of C. dam-
arensisdo not tend to persist beyond the lifetime of a breed-
ing pair and reproductive partners are not typically close
kin. Colony sizes for the other species of Cryptomys that
have been studied are smaller and behavioral specialization
among family members is minimal (Bennett and Faulkes
2000; Spinks et al. 2000a, 2000b).
In contrast to social African mole-rats, groups of colo-
nial tuco-tucos (Ctenomys sociabilis) consist of two to six
closely related females and a single, unrelated male (Lacey
and Wieczorek 2004). Approximately two-thirds of year-
ling females are still resident in their natal burrow system,
where they remain for the duration of their life. In contrast,The Ecology of Sociality in Rodents 249