Reciprocal altruism in primates
While struggling to understand the evolution of cooperation throughout
the different taxa, one of the most intricate tasks is to study cooperation
in primates. Packer (1977) was the first to report reciprocal altruism
among male olive baboons (Papio anuhis). The formation of short-term
coalitions among adult male P. anuhis during aggressive interaction against
a single adversary is initiated by one male enlisting a partner and completed
by subsequently fighting the opponent. Packer found considerable evidence
that individuals who join an enlisting baboon have previously succeeded
in enlisting that very same individual, suggesting cooperation may be partly
based on reciprocity (Packer 1977). Determining costs and benefits is extremely
difficult, since the payoff matrix may vary according to the different
circumstances under which coalitions are formed (e.g. the strength of the
opponent specifies the costs, while the cause of the aggression specifies
the benefits). However, assuming that most of the coalitions were formed
to take over an oestrous female and that the wounds received were non-fatal,
the resulting sired offspring would be a greater benefit than the less
costly wounds received in a number of consecutive 'plays' if both partners
reciprocated in aiding each other (Packer 1977).
But since only one of the cooperating partners (usually the enlisting
male) obtains the female at issue, two coalitions make a 'game'. This case
is at best the alternating variant of the simultaneous IPD, recently investigated
by Nowak and Sigmund (1994). Noë (1990) argued that this form of delayed
reciprocity is better explained with a 'Veto game' that includes information
exchange, partner-choice and partner competition. The ability of primates
in trading different beneficial acts for other behaviour supporting the
benefactor, further complicates the matter. Seyfarth and Cheney (1984)
looked at coalition-forming in female vervet monkeys (Cercopithecus
aethiops) in relation to preceded grooming. Their data seem to support
the view that vervets are more apt to help an enlisting, unrelated individual
if they have been groomed by those individuals in the recent past (Seyfarth
and Cheney 1984). However, due to the complexity of exchanging acts in
various currencies (especially if more than two behaviours are involved)
and due to the difference in hierarchical rank that breaks the symmetry
of the game, quantification of costs and benefits seems unlikely to be
possible (Noë 1990). Therefore, all results obtained in studies as
the two described above will be persistently disputed (e.g. Noë 1990,
Hemelrijk 1991, Seyfarth 1991).
|