availability of suitable crevices (Pearson 1948; George and
Crowther 1981; Channing 1984; Moses and Millar 1992;
Prakash and Singh 2001). This effect of crevice number on
population density has been confirmed in the bushy-tailed
woodrat by experimental manipulation of den and nest
sites (Hickling 1987, as cited in Moses and Millar 1992). In
the case of the dassie rat, competitive exclusion from shel-
ters by the rock hyrax led to a 97% difference in population
densities between the two species (George and Crowther
1981). Both intraspecific and interspecific competition for
crevices limits the number of rock-dwellers that can occupy
any one site.
Social structure is also influenced by the size and distri-
bution of crevices. Small groups of petrophilic rodents are
found where crevices are small or few in number, whereas
large groups tend to occupy outcrops that have many large
crevices (Anderson et al. 1983; Hickling 1987, as cited
in Smith 1997; K. Nutt, pers. obs). In studies of the com-
mon gundi, territory size was comparable among all social
groups, yet the smallest group (of three individuals) occu-
pied a territory that contained only two small primary shel-
ters, whereas the largest group (with over twenty gundis)
occupied an isolated territory that contained a vast number
of sizeable shelters (K. Nutt, pers. obs.). Interspecific com-
petition for shelters may also influence group size in the
ctenodactylid rodents. Field observations indicate that Val’s
gundi may live in smaller-sized social groups because the
common gundi prevents it from occupying better outcrops
with larger crevices (Gouat 1988b; Gouat 1988a). More
detailed quantitative studies are needed to verify this ef-
fect of crevice size and distribution on group size in rock-
dwelling rodents.
How distribution of rocky substrate may influence a species’
geographic distribution
Many rock-dwelling mammals inhabit rock outcrops that
are disjunct and patchily distributed (Smith 1980; Hoeck
1982; Christian and Daniels 1985; Kilpatrick and Crowell
1985; Churchill 1996; Kim et al. 1998; Braun and Mares
2002; Walker et al. 2003; Spotorno et al. 2004). These rock
islands vary greatly in size and proximity to one another,
and together form what is known as a metapopulation. The
occupancy of each rock outcrop within a metapopulation
varies over time as some populations go extinct and others
are recolonized. Models have shown that patch occupancy
of mountain viscacha outcrops is primarily determined by
the abundance and depth of crevices at a given site (Walker
et al. 2003). Alternatively, metapopulation studies of the
American pika have shown that patch occupancy depends
on size of rocky habitat (smaller habitat patches going ex-
tinct more often than larger patches) and the distance be-
tween patches (more isolated islands of rocky habitat hav-
ing slower rates of recolonization; Smith 1980). Additional
studies have shown that low rates of recolonization in some
large patches may lead to long-term population declines
(Moilanen et al. 1998; Clinchy et al. 2002). The knowledge
of such a process should prove invaluable to conservation
efforts of rare, endangered, or threatened rock-dwelling
rodents (Oliver 1977; Balcom and Yahner 1996; Jimenez
1996; Zahler 1996).SummaryOver one-fourth of all rodent genera contain at least one
species that utilizes rocky substrate. These species do so
because of the highly complex nature of rocky habitat,
to avoid or detect predators, for the thermoregulatory and
water-retaining properties of rocks, to avoid habitat dam-
age by fires and grazers, to take advantage of unique micro-
environments within rocky habitat, and to obtain year-
round resources.
Approximately seventy-seven species of rodents can be
considered to be obligate rock-specialists. Although very
little is known about the sociobiology of most saxatile ro-
dents, documented social behavior ranges from highly co-
lonial to solitary, with the majority of species tending to live
in social groups. Several factors undoubtedly influence the
type of social behavior exhibited by each petrophilic ro-
dent including shared ancestry, the need for extended pa-
rental care in harsh environments, predation pressure, and
the limited distribution of favorable rocky habitat. The spe-
cial features of rocky habitat provide a unique opportunity
to study the selective pressures influencing the social behav-
ior of rock-dwelling mammals. Hopefully the very limited
but suggestive information currently known about saxatile
rodents will serve to motivate future studies of the relation-
ship between ecological constraints and social evolution in
this select group of mammals.426 Chapter Thirty-Five