of S. galili, 2n52 and S. golani, 2n54 (Nevo et al.
1982). Conversely, aggression decreases southward partly
in S. carmeli,the 2n58 species, but particularly in the
S. judaei,the 2n60 species, presumably due to the low
available resources —hence, sparser populations, in the
more xeric southern habitats (Nevo 1985; Nevo et al. 1986;
fig. 25.4; and fig. 47 in Nevo et al. 2001). In the latter, cli-
matic selection presumably operates to minimize overheat-
ing as well as water and energy expenditure (Nevo and
Shkolnik 1974). This pattern supports the optimal activity
hypothesis where selection maximizes fitness by optimizing
net energy gain per unit activity (Nevo et al. 1982b, in
Nevo 1999). Within populations, natural selection resulting
from intrapopulation conflicts among aggressive behavioral
phenotypes, militants, intermediates, and pacifists lead to
genetic equilibria. The equilibria comprise a mixture of be-
havioral phenotypes (fig. 25.5b). This mix is evolutionarily
stable according to the game theory definition of Maynard-
Smith and Price (1973) due to the ecological heterogeneity
of the habitat (Nevo et al. 1986). Militants dominate rich
vegetational patches while pacifists are pushed to periph-
eral, poorer vegetational patches (Nevo, unpublished obser-
vations). However, these equilibria are presumably based
on flexible and dynamic changes in the proportion of be-
havioral phenotypes in accordance with regional and local
variations in ecological diversity caused by macro- and mi-
croclimatic variations.
The active ecological speciation and adaptation of S. eh-
renbergiand its adaptive radiation into increasingly arid
environments (also increasingly arid environments within298 Chapter Twenty-Five
Figure 25.5a Discriminant analysis of eight categories of aggressive behavior (“total aggression”) based on all four species displaying multi-
peak variation of aggression in the Spalax ehrenbergisuperspecies. Frequency of animals appears on the Y-axis. The analysis is the basis of
differential polymorphism among species described in figure 25.5b and figure 25.5c (from Nevo et al. 1986).Figure 25.5b Genetic polymorphism in patterns of aggressive behavior in both sexes of each species of the Spalax ehrenbergisuperspecies
in Israel: Spalax galili, 2 n52; S. golani, 2 n54; S. carmeli, 2 n58, and S. judaei, 2 n60. The Y-axis represents percentage of animals
in each of the three behavioral phenotypes (from Nevo et al. 1986).