ever, there is no a priori reason to suppose the ancestor was
solitary, as the social traits of a common ancestor could
equally have been lost. This so-called secondary solitarity
has been reported in some species of bees (Wcislo and Dan-
forth 1997).
It is impossible to verify the status of the common an-
cestor directly. In the absence of such data, clues may be
sought in the close hystricomorph relatives of the Bathyer-
gidae. There is evidence that some of the South American
caviomorphs, like tuco-tucos (family Ctenomyidae), exhibit
some form of social grouping (Lacey and Sherman, chap.
21 this volume). However, these New World hystrico-
morphs are divergent from the Old World families, and ac-
cording to the molecular phylogeny of Nedbal et al. (1994)
the closest relatives to the Bathyergidae are Old World por-
cupines (family Hystricidae), cane rats (family Thryonomy-
idae), and dassie rats (family Petromuridae). At least one
species in the Hystricidae, the Cape porcupine (Hystrix
africaeaustralis), has colonial habits. Cape porcupines live
in colonies of six to eight individuals, consisting of an adult
pair and consecutive litters of offspring, which are normally
singletons, occasionally twins, or rarely triplets. Adult males
protect the young, are aggressive towards foreign males and
females, and accompany the young on foraging trips until
6 –7 months of age, after which they tend to become soli-
tary feeders (Van Aarde 1987). However, in the cane rats,
both the greater cane rat (Thryonomys swinderianus) and
the lesser cane rat (T. gregorianus) are generally reported to
be solitary, although individuals may live in close proxim-
ity in reed beds (Skinner and Smithers 1990). Among the
dassie rats, Petromus typicusis reported to live in pairs or
families in the crevices that occur in their rocky habitat, al-
though information is limited (Skinner and Smithers 1990).
However, none of the species in these three families are sub-
terranean, so it remains difficult to extrapolate to make any
definite inferences about the ancestral bathyergid.
If we assume that the ancestral condition of the Bathyer-
gidae was social (node 1 in fig. 36.2), Burda et al. (2000)
suggest that the appropriate questions about the evolu-
tion of eusociality in mole-rats should be: why did certain
species of the bathyergid family become solitary, and why
have Cryptomysand Heterocephalusnot abandoned their
social way of life? While this approach might be of inter-
est from the point of view of phylogeny, it has little or no
effect on the hypotheses that address convergence of life-
styles within the family. All phylogenies published to date
(Allard and Honeycutt 1992; Faulkes et al. 1997; Walton
et al. 2000; Faulkes et al. 2004; Ingram et al. 2004) show
that naked mole-rats and theCryptomysgenus are sepa-
rated by a number of common ancestors leading to lineages
of solitary species, indicating repeated losses and gains of
varying degrees of social elaboration (e.g., nodes 2 and 3 in
fig. 36.2).Social evolution: Ultimate factors
There are two principal ways in which societies may form.
The parasocial route involves nonoverlapping reproductive
generations, “shared nests” composed of groups of related,
or related and unrelated individuals. The second principal
way in which societies may form, and the one generally
agreed as the most likely precursor of cooperatively breed-
ing /eusocial vertebrates, is the so-called subsocial route. In
this case, groups of overlapping generations arise as a re-
sult of natal philopatry, where offspring delay dispersal and
remain to form a family group. Thus a family, as defined by
Emlen (1995, 1997) is a social group where offspring con-
tinue to interact with their parents into adulthood; that is,
beyond the age of sexual maturity. Emlen (1995, 1997) fur-
ther differentiates families into simple or extended. In simple
families only one pair breeds, a situation akin to the social
bathyergids, although with time immigration of nonfamily
members into the group may also occur through dispersal
(Jarvis and Bennett 1993; Jarvis et al. 1994; O’Riain et al.
1996; Bishop et al. 2004; Burland et al. 2004). In extended
families, more than one individual of either one or both
sexes breed.
There has been much debate in the literature regard-
ing the ultimate causes of natal philopatry and sociality in
the Bathyergidae (for review see Bennett and Faulkes 2000;
Burda et al. 2000). Irrespective of where, or even if, the line
between eusociality and other forms of sociality is eventu-
ally drawn, most authors concur that the evolutionary ori-
gins of eusociality should no longer be considered to lie in
intrinsic or genetic factors (e.g., haplodiploidy). Thus while
haplodiploidy might predispose Hymenopteran insects to-
ward eusociality, similar social systems in diploid termites
and mammals discount it as a necessary prerequisite for
the evolution of eusociality. The ultimate evolutionary fac-
tors that have thus far been suggested as important in the
evolution of sociality in mammals are (1) predator vigilance
and protection, e.g., meerkats (Clutton-Brock et al. 1998),
dwarf mongoose (Rasa 1977), naked mole-rats (Alexander
et al. 1991), and (2) increased efficiency in procuring food,
e.g., naked mole-rats (Jarvis and Sale 1971), Damaraland
mole-rats (Jarvis and Bennett 1993), wild dog (Frame et al.
1979), and wolves (Zimen 1976). Several authors have pro-
posed that the distribution, size, and digestibility of the geo-
phytes (underground roots and tubers) upon which mole-
rats feed, as well as the variation and predictability of
rainfall, have played a pivotal role in the evolution of soci-
ality in the Bathyergidae (Jarvis 1978; Bennett 1988; Love-432 Chapter Thirty-Six