opposed to 35%–41% in captivity). This low frequency of reconciliation was found in
relation to male–male, male–female, and female–female conflicts. In fact, aggressive
incidents are not as frequent in the Sonso community (and other wild chimpanzee com-
munities) as they often are in captive situations. When reconciliation did take place it
fitted the ‘valuable relationship hypothesis’ under which individuals who spend more
time together in affiliative contexts are expected to reconcile more often. Surprisingly,
however, oestrous females were not more likely to achieve reconciliation with males
than anoestrous ones. In relation to status, reconciliation was almost always initiated by
the lower ranking individual after a conflict. Of all successful attempts at reconciliation
54% were preceded by some form of submissive signal, most commonly the pant-grunt,
which is a universal signal of subordinacy in chimpanzees. Gestures used before recon-
ciling included holding out a hand or foot or presenting the rump, but mouth to mouth
kissing in this context (de Waal and von Roosmalen 1979) was never seen, and nor
was any form of consolation, which the above authors have reported seeing in captive
chimpanzees. Summarizing, Arnold and Whiten conclude that only one dimension of
relationship significantly influences the likelihood of reconciliation: compatibility. Thus
a history of friendly contacts constitutes a relationship that is sufficiently valuable to be
worth restoring after a conflict.
Finally, on this subject, we have the finding by Seraphin (2000) of a positive correla-
tion between male rank and cortisol production (the cortisol was derived from faeces).
We can interpret this to mean that higher ranking males have higher stress levels than
lower ranking males, in which case the fluctuating strategies of rivals and the need to
keep a constant watch on rivals may constitute, as Seraphin suggests, a cost on high
rank. A study of the Kanyawara community of the Kibale Forest by Muller and
Wrangham (2004) also found this correlation, which they attributed to elevated energ-
etic expenditure rather than to psychosocial factors. As they point out, aggressive
displays, characteristic of dominant males, are energetically expensive.
Status among females
With the qualification that females may influence males in a variety of ways, there is no
doubt that at Sonso females are subordinate to males. They are highly sociable and often
congregate together or in mixed parties of males and females (Fawcett 2000; Bates
2001; Munn 2003). Between themselves there is not so much a rank order as a division
between ‘central’ females and ‘peripheral’ females. This refers partly to spatial location
and partly to social position. If we compare the ranging patterns of central and peri-
pheral females, some females, such as Ruda or Ruhara, are rarely seen in the central area
of the community; they seem to keep to themselves and prefer to be in small groups of
females and juveniles or with lower ranking males away from the action. By contrast the
central females such as Nambi and Zimba are very often found with the dominant males
of the community in the centre of the range.
We have already noted (Chapter 5) that at Taï females form stronger bonds than at
other sites studied. At all sites, however, there is some evidence for dominance-striving
Status among females 131