tail to understand some of the dynamics of group living in
desert habitats. These two species are the great gerbil, R.
opimus,in Central Asia (e.g., references in table 31.1 and
Randall et al. 2000; Tchabovsky et al. 2001; Randall and
Rogovin 2002; Rogovin, Randall et al. 2003) and the Mon-
golian gerbil, M. unguiculatus,in Russia and China (Ågren
1984a, 1984b, 1984c; Ågren et al. 1989a, 1989b; Orlenev
1987; Gromov 2000). Both species are arid-adapted, di-
urnal rodents that live in family groups in complex burrow
systems constructed in open habitats. The gerbils coopera-
tively defend nonoverlapping, multipurpose territories from
encroachment by neighboring groups. Animals are active
all year, and members of the family group cooperatively
gather and store vegetation in common stores for survival
during times of the years when food is scarce (Ågren et al.
1989b; Tchabovsky et al. 2001). In R. opimus,the adult
male may store a larger proportion of the food than other
members of the group. Males of both species actively scent
mark the territory by building heaps of soil and then rub-
bing their ventrums on the heaps to deposit a secretion from
a large ventral scent gland (Roper and Polioudakis 1977;
Popov and Tchabovsky 1996; Gromov 2000).
Although there are many similarities, the two species of
social gerbils exhibit some interesting differences in group
composition and behavior. Family groups of M. unguicu-
latusseem to consist of a monogamous male-female pair
and their offspring, except at high densities, when males as-
sociate with several females (Ågren et al. 1989a, 1989b;
Gromov 2000). Mating is not entirely monogamous, be-
cause estrous females enter adjacent territories to mate with
strangers. Mongolian gerbils have a dominance structure in
which males dominate females and larger animals dominate
smaller ones. Ågren et al. (1989a) suggest a reproductive
skew and suppression of reproduction via suppression of
maturation. In contrast, family groups of R. opimusmay be
larger and more complex than those of M. unguiculatus.
Families consist of (at least at higher densities; see the fol-
lowing) one adult male, from one to six adult females, ju-
veniles, and subadults of previous litters. Rhombomy opi-
musseems to have an egalitarian social structure. Males do
not dominate females, and both males and females store
food and give alarm calls. Suppression of reproduction does
not seem to occur with any regularity, and the majority
of females in family groups exhibit signs of pregnancy and
lactation. All females in a family group interact socially
with pups of all litters of different ages (Randall et al. 2005;
Randall and Rogovin unpublished data). Rhombomy opi-
mushas a well-developed system of alarm communication
consisting of both vocal and seismic signals (Randall et al.
2000; Randall and Rogovin 2002).
An 8-year field study of R. opimusrevealed that family
groups of R. opimusare inherently unstable and form and
expand based on the availability of vegetation and subse-
quent changes in population density (Randall et al. 2005).
Reproduction of individual females, as measured by the
number of pups per female emerging from the burrow and
survival for 1 season depended on favorable environmental
conditions in the form of rainfall, mild temperatures, and
availability of green vegetation, the main food source. Re-
production is associated with the production of green veg-
etation in the spring, and gerbils will continue to produce
litters while conditions are favorable. In years with good
production of green vegetation (1996 –1999) the majority
of females lived in multifemale family groups with a resi-
dent male (fig. 31.1). During a drought extending from late
spring 1999 to spring 2002 the abundance and temporal
availability of green vegetation were limited. There was
high mortality and increased predation, causing population
densities to decrease and the social structure to shift to soli-
tary females living alone with their offspring (fig. 31.1). As
dispersion of females changed, adult males altered their be-
havior. Instead of remaining with family groups and pro-
viding parental care by storing food and marking territo-
ries, males traveled the study area visiting solitary females
(Tchabovsky et al. 2001; Rogovin et al. 2003; Randall
et al., in press). Thus far, we have found few differences in
reproductive success in the different-sized groups. Although
solitary females behave differently from females in groups
and spend more time in the burrow and are more vigilant
when foraging (Tchabovsky et al. 2001), they still seem to
have about the same reproductive success as individual fe-
males in groups. Great gerbils are facultatively social. Fe-
males live solitarily under conditions of limited food and
high mortality, which disrupt social behavior and group
formation; they share territories with female kin under fa-
vorable conditions for survival and reproduction when kin
groups can be maintained. Males adjust to the distribution
of females.Environmental Constraints and the Evolution of Sociality in Semifossorial Desert Rodents 377Figure 31.1 Percentage of Rhombomys opimusfemales living in groups of
two or more with a resident male (cross-hatched), in a male-female pair (white)
and solitary (black) in years of high (1996 –2000), low (2001–2002), and recov-
ering population densities. Nfemales in the population at top of bars.