as in the intervening areas of arid steppe habitat. This dif-
ference in specialization creates very different spatial distri-
butions of suitable habitat for these species, leading La-
cey and Wieczorek (2003) to suggest that the difficulty of
dispersing between patchily distributed mallín habitats is
an important ecological variable favoring sociality in C.
sociabilis.
The adaptive bases for sociality in cururos have not been
explored. S. cyanusoccupies a wide range of habitats in
Chile, including arid mediterranean grasslands and mesic
montane meadows. This intraspecific habitat variation pro-
vides an ideal opportunity to assess the ecological bases for
sociality. Preliminary comparative data from two popula-
tions of S. cyanusindicate that soils are harder to penetrate
at an arid coastal site (Parque Nacional Fray Jorge) than
at a mesic montane site (Santuario de la Naturaleza Yerba
Loca; Lacey, Ebensperger, and Wieczorek, unpublished
data). Although food resource distributions at these locali-
ties have not been quantified, the habitat at Fray Jorge is, at
least superficially, more similar to areas inhabited by social
mole-rats than the habitat at Yerba Loca, which closely re-
sembles the mesic mallínes inhabited by colonial tuco-tucos.
Given these parallels, it is intriguing that preliminary mark-
recapture data suggest that, contrary to the predictions of
the AFDH, groups ofS. cyanusare larger at Yerba Loca (La-
cey, Ebensperger, and Wieczorek, unpublished data).
In summary, the two subterranean examples of sociality
that have been studied in detail yield different pictures of
the ecological factors that favor group living. At a gen-
eral level, these arguments are similar in suggesting that en-
vironmental conditions constrain dispersal, thereby favor-
ing natal philopatry and the formation of social groups.
Resource distributions also appear to be important in both
taxa, although the nature of these resources (food versus
suitable habitat) and the spatial scales over which they are
patchily distributed (within versus between burrows) may
differ between social mole-rats and colonial tuco-tucos.
Data from cururos are not yet sufficient to determine
whether the ecology of group living in this species paral-
lels that of either mole-rats or tuco-tucos, or whether this
example of subterranean sociality reflects a third, distinct
suite of environmental conditions.
An Integrated Ecological Model
for Sociality in Rodents
If the ecological bases for group living differ between the
two best-studied subterranean taxa —African mole-rats and
colonial tuco-tucos — is it possible to develop a general
explanation for sociality among subterranean species, let
alone all rodents? We believe that such an explanation is
possible. Here, we propose a three-dimensional model of the
ecological factors that affect rodent social systems (fig. 21.2).
This model is derived from comparative analyses of the
social, subterranean species reviewed previously in this
chapter, but should be applicable to semi-subterranean or
surface-dwelling rodents as well as to facultatively versus
obligately social species. Our intent is to provide a general
framework for future studies that explore the ecological
bases for group living in rodents. The axes included in our
model are as follows:- Access to critical resources
Access to critical food resources is an integral component
of the AFDH proposed for African mole-rats (Jarvis et al.
1994; Lacey and Sherman 1997; Bennett and Faulkes 2000).
Although the reasons why colonial tuco-tucos are restricted
to mesic, mallín habitats are unknown, these habitat patches
themselves represent important resources, the distribution
of which is thought to favor natal philopatry and group for-
mation in C. sociabilis(Lacey and Wieczorek 2003). More
generally, conspecifics may be required to live together in
order to have access to limited, patchily distributed re-
sources (Alexander 1974; Safran et al., in press), implying
The Ecology of Sociality in Rodents 251Figure 21.2 Three-dimensional conceptual model of the ecological bases for
sociality. The three primary ecological factors thought to influence natal philo-
patry and group living are indicated. Postulated locations for several well-studied
species are shown to indicate how data for these species would be represented
in this model.