that this axis represents a logical component of any ecolog-
ical model of social behavior.
- Predation
Predation has frequently been invoked as a selective pres-
sure favoring group living (Hoogland and Sherman 1976;
Hoogland 1979a, 1981; Krause and Ruxton 2002). Sur-
prisingly, however, predation on social subterranean ro-
dents has received little attention, despite compelling anec-
dotal evidence (see the following) that these animals are
preyed upon by multiple species. In particular, the AFDH
does not consider predation, an omission that we address
by including predation as one of three axes in our model.
Predation is often difficult to quantify. For example, even
though social African mole-rats are known to be preyed
upon by birds, mammals, and snakes (Jarvis and Sherman
2002), no data are available regarding the frequency of this
predation. Jarvis et al. (1994) indicate that, qualitatively,
predation (i.e., the identities of predatory species) does not
appear to differ between solitary and social species of mole-
rats, but this observation does not preclude quantitative dif-
ferences in predation pressure that may favor sociality in
some species but not in others. Behavioral studies of naked
mole-rats indicate that colony members exhibit a conspicu-
ous, cooperative response to threats within their burrow
systems (Lacey and Sherman 1991), implying that, over evo-
lutionary time, these animals have regularly encountered
predators in their burrows. Collectively, these lines of evi-
dence suggest that predation may be an important compo-
nent of mole-rat ecology.
Similarly, predation has not been considered as part of
comparative studies of the ecology of solitary and social
tuco-tucos. Building from the AFDH, these analyses have
focused on the same suite of ecological factors identified as
important to bathyergid mole-rats (Lacey and Wieczorek
2003). Colonial and Patagonian tuco-tucos are preyed upon
by foxes (Dusicyon australis) as well as by several species of
birds (e.g., barn owls, Tyto alba;caracaras, Milvago chi-
mango;Lacey and Wieczorek 2003). Data regarding pre-
dation are comparable to those for African mole-rats; al-
though the suite of predators affecting C. sociabilisand C.
haigiis the same, it is possible that quantitative differences
in predation contribute to the differences in social structure
between these species. C. sociabilisroutinely gives alarm
calls in response to predators, suggesting that, evolutionar-
ily, predation pressure has been sufficient to favor a specific
behavioral response. In sum, despite the lack of quantita-
tive information regarding differences in predation on soli-
tary versus social subterranean rodents, we believe that
predation is an important ecological factor that should be
included in our model.
- Cooperation as a response to ecological challenges
The adaptive benefits of cooperation may vary according
to the nature of the ecological factors that favor group liv-
ing. For example, if (as postulated by the AFDH) temporal
constraints on the excavation of new burrows (due to soil
hardness in dry seasons) are a primary factor underlying so-
ciality among African mole-rats, then cooperative digging
by these animals should be favored because it increases the
probability of locating additional food resources during the
limited period when burrow excavation is possible. In con-
trast, if the difficulty of dispersing from one mallín patch
to another is the primary ecological factor favoring social-
ity in colonial tuco-tucos, it seems unlikely that cooperation
among burrow mates would provide individuals with an
advantage in responding to this environmental challenge;
although the animals could disperse in groups, this would
not require that they share a burrow system, nor does it
seem likely to increase the chances of locating a new patch
of suitable habitat. This difference between social African
mole-rats and colonial tuco-tucos suggests that the ability
of cooperative behavior to resolve ecological challenges may
be an important component of social structure.
At first glance, this axis appears to differ from the pre-
vious two in that it represents a consequence, rather than
a cause of group living. Cooperative solutions to ecologi-
cal challenges, however, may be critical to the persistence of
groups that form due to extrinsic environmental conditions
(Alexander 1974; Brown and Brown 1996). Indeed, in some
species, cooperation may function as a form of group aug-
mentation (Kokko et al. 2001) that contributes significantly
to the maintenance of sociality. At a minimum, the degree
of cooperation among group mates provides an important
basis for comparing different social systems and, thus, in-
clusion of this axis in our model should facilitate distinc-
tions between highly cooperative societies (e.g., naked mole-
rats) and societies in which individuals live together but do
not engage in elaborate forms of cooperation (e.g., colonial
tuco-tucos).
The axes that form the basis for our model closely par-
allel the benefits generally ascribed to group living (Alexan-
der 1974: Hoogland and Sherman 1976; Krause and Rux-
ton 2002; Safran et al., in press). Although it is implicit
in many previous treatments of this subject that sociality
likely reflects the combined effects of multiple environmen-
tal factors, our model makes this explicit by treating each
of the relevant variables as one of three distinct factors con-
tributing to sociality. Heuristically, the model can be viewed
as a three-dimensional graph, with the distance from the
origin along the x-, y-, and z-axes increasing as the strength
of each selective factor (access to resources, predation pres-
sure, benefits of cooperation) increases (fig. 21.2). Plotting
252 Chapter Twenty-One