but limit dispersal will result in a trend toward larger group
size, provided other variables (e.g., predation) are fairly
constant. Further, in areas where the food density is low or
of low energetic value the average body size within a given
species will be reduced. Evidence for the latter has been
shown in the naked mole-rat (Jarvis 1985), where the mean
size of nonbreeders was smaller compared to those for non-
breeders in colonies living in areas where food resources
were more abundant. Spinks (1998) reports a similar trend
in Cryptomys h. hottentotus,where individuals in arid re-
gions were smaller than those from mesic parts of their dis-
tribution. These observations support the utility of a com-
bination of group size, reproductive skew, and a measure
of philopatry in defining the degree of sociality in African
mole-rats.
The occurrence of solitary dwelling mole-rats in arid re-
gions would argue against the AFDH, which is the case for
Bathyergus janetta,a species that ranges into areas of very
low rainfall in the northwestern Cape region of South Af-
rica and parts of southwestern Namibia. However, this ap-
parent contradiction to the AFDH is an oversimplification,
as the AFDH depends on a combination of aridity and food
distribution. Not only is B. janettaoften found in associa-
tion with subterranean moisture seepage areas, which mit-
igate against the low rainfall, but due to the unique flora of
Namaqualand, with many plants with swollen roots, sub-
terranean stems and tubers, food resources are abundant in
these otherwise very arid regions (Herbst 2002). Recently,
a study of H. argenteocinereusin Malawi has suggested
that this solitary species is also successfully exploiting a re-
gion of low geophyte density (Sumbera et al. 2003). While
the reported mean geophyte density was very low (2.5 geo-
phytes /m^2 ) and comparable with areas inhabited by naked
mole-rats, the mean geophyte biomass was relatively high
at 192 g /m^2 , and falls within the range of other solitary
mole-rats (e.g.,G. capensis: 120-600 geophytes /m^2 ;DuToit
et al. 1985; Bennett 1988). Following on, it has also been
argued that social or eusocial species of mole-rats should
not occur in mesic areas (e.g., C. mechowi;Burda and Ka-
walika 1993). However, the AFDH does not propose that
social species cannot occur in mesic areas, and the limited
data available suggest that in such areas social group dy-
namics fit the predictions of the AFDH (e.g., C. h. hot-
tentotusin mesic versus arid habitats; Spinks et al. 1998,
2000). Nevertheless, unambiguous data on group sizes, kin
structure, and patterns of dispersal /philopatry are needed,
especially on the diverse populations of social Cryptomysin
Zambia that inhabit regions of high rainfall.
Burda et al. (2000) argued against the AFDH, and sug-
gested that the ultimate evolutionary reason for eusociality
in mole-rats is that it is a phylogenetically constrained phe-
nomenon resulting from a social common ancestor rather
than a response to ecological conditions, and that a solitary
lifestyle is the derived trait in some species of Bathyergidae.
While this is an interesting point to consider, this proposal
merely sidesteps the issue of what selective factors ultimately
may have given rise to natal philopatry and cooperation in
ancestral bathyergid species in the first place. It also ignores
the comparative approach, whereby phylogeny can be used
to make valid independent contrasts between sociality and
ecological factors irrespective of the status of the common
ancestor (e.g., Faulkes et al. 1997).
Other factors that have been suggested as causative in
the social evolution of some mole-rats include their appar-
ent inability to store fat and their increased postnatal de-
velopment rates. It is suggested that these lead to an inabil-
ity of the breeding female to rear young on her own, hence
the retention of helpers to provision food during the long
periods of gestation and lactation (Burda 1990; Burda et
al. 2000). In the absence of a comprehensive study on fat
stores in mole-rats, the relevance of fat storage ability to
the evolution of sociality remains unclear. However, field
observations of autopsied, freshly captured breeding and
nonbreeding females of C. damarensis, C. darlingi, C. h.
hottentotus,and H. glabershow clear evidence of stored fat
(Bennett and Jarvis, unpublished data). Lack of an ability to
store fat might predispose such species to the retention of
offspring, but it does not explain why such a social system
evolved in the first place. Females that cannot store fat
but that can nevertheless maintain a positive energy balance
throughout reproduction because of ample food supplies
need not retain either their mates or their offspring. In con-
trast, females that live, for example, in an arid region where
food resources are difficult to obtain may be under selection
to retain their partners and their offspring to ensure contin-
ued mating and to lower the costs and risks with obtaining
such food. An ability to store fat might be a proximate fac-
tor favoring the evolution of sociality, but seems unlikely as
an ultimate evolutionary factor.
Burda et al. (2000) also argue that postnatal develop-
mental rates are a cause rather than a consequence of euso-
ciality in naked and Damaraland mole-rats, although there
is no consistent evidence for this being a causative factor.
Developmental length differs dramatically in naked and
Damaraland mole-rats, both of which are considered to
be eusocial. Naked mole-rats are comparatively altricial at
birth and progress to solid foods at approximately 2 weeks
of age, and begin to leave the nest voluntarily at 3 weeks. In
contrast, Damaraland mole-rats are relatively precocious,
and may eat solids within a few days after birth and explore
the burrow system as early as days one to five. Develop-
mental length may well be phylogenetically constrained in434 Chapter Thirty-Six