Rodent Societies: An Ecological & Evolutionary Perspective

data from different populations or species of group-living
rodents should yield a multidimensional cloud of points.
Populations or species occurring in the inner portion of the
cloud (i.e., near the origin) should be those in which indi-
viduals receive only modest fitness benefits from sociality,
while members of those in the outer reaches of the cloud
should gain substantially from group life.
The precise shape of this cloud of points is not yet
known, but should be informative. For example, if the axes
in this model are not truly independent (e.g., benefits to
cooperation are influenced by predation pressure or access
to resources), then clustering of points in different regions
of the graph may reveal previously undetected relationships
among different ecological variables. At the same time, com-
paring the distribution of points for the two general types
of social groups identified previously in this chapter may
help to determine whether plural-breeding species with fe-
male philopatry are qualitatively distinct from singular-
breeding species in which both sexes are philopatric. Fi-
nally, it may be possible to relate the axes of this model and
the relative positions of different species to the typical direct
fitness benefits associated with group living.
As an example of how this model should work, we ex-
pect the placement of African mole-rats and colonial tuco-
tucos on the 3-dimensional graph to differ markedly. Given
that naked and Damaraland mole-rats appear to cooper-
ate extensively in response to ecological challenges, we pre-
dict that they will be positioned far from the origin on the
corresponding axis. In contrast, colonial tuco-tucos are not
highly cooperative and, hence, they would be located much
closer to the origin of this axis. As a result, even if the ef-
fects of resource distributions and predation are compa-
rable for colonial tuco-tucos and naked mole-rats, these spe-
cies will occur in quite different regions of the graph. This
difference, in turn, may have important implications for
other aspects of the social systems of these species, includ-
ing the degree of reproductive and behavioral specialization
among group mates.
Extending these ideas to other group-living rodents,
available data suggest that predation is an important factor
favoring sociality in black-tailed prairie dogs (Hoogland
1995), indicating that this species should be positioned far
from the origin on this axis, but much nearer to the origin
for the other two axes of the model (fig. 21.2). In contrast,
beavers are thought to form groups due largely to benefits
associated with the cooperative construction of lodges and
larders (Busher, chap. 24 this volume), with the result that
this species should be much further from the origin along
the cooperation axis than along the predation or resources
axes. Finally, both resource distributions and predation are
thought to influence yellow-bellied marmots (Blumstein and
Armitage 1999; Armitage, chap. 3o this volume), leading to

our placement of this species far from the origin on the cor-

responding axes of the model.

Given the difficulty of quantifying each of the selective

pressures identified by this model, it remains, for the mo-

ment, a conceptual construct for exploring the effects of dif-

ferent ecological variables on rodent societies. The strength

of this approach is that it brings together several of the

key selective factors thought to promote sociality in one

comparative, at least potentially quantifiable framework. As

studies of rodent social systems continue, we anticipate that

quantitative comparisons of these variables will increase,

leading to a more refined understanding of how ecologi-

cal variables contribute to the behavioral variation evident

among these animals. Although our discussion has focused

on the adaptive bases for group living among subterranean

species, we expect that these ideas can be extended to in-

clude other rodent societies. At a minimum, we hope that

our discussion of rodent sociality will stimulate further stud-

ies of this diverse and fascinating order of mammals.

Summary and Future Directions

Determining why conspecifics live in groups and how those

groups are structured with respect to kinship and repro-

duction is critical to understanding the diversity of rodent

social systems portrayed in this volume. In many species of

rodents, social groups form due to natal philopatry, with

the result that groups are composed of close kin of one or

both sexes. The reproductive structure of groups appears

to covary with patterns of philopatry and kinship; while

plural breeding is typical in species with female-only philo-

patry, singular breeding predominates in species in which

both sexes remain in their natal area. Philopatry is expected

to occur when the confluence of intrinsic benefits of group

living and extrinsic constraints on dispersal dictate that in-

dividuals will achieve greater fitness by remaining in their

natal group than by dispersing and attempting to breed

elsewhere. Although data are available for only a few spe-

cies, the generally negative relationship between direct fit-

ness and group size implies that, among rodents, sociality

may be strongly influenced by extrinsic (ecological) con-

straints on dispersal.

Numerous factors influence an individual’s decision to

remain in the natal area and, accordingly, conceptual mod-

els of sociality should consider multiple selective pressures

that promote group living. Our comparative analyses of so-

cial subterranean rodents suggest that access to critical re-

sources, predation pressure, and the use of cooperative in-

teractions to overcome ecological challenges form the basis

for a multidimensional model of the ecology of group liv-

ing. Specifically, each of these parameters can be viewed as

The Ecology of Sociality in Rodents 253