BoundedRationality_TheAdaptiveToolbox.pdf

240
Kevin N. Laland
the social transmission of maladaptive information, neither the behavior of the
fish, nor the general capacity for social learning should be described as
maladaptive. Taking an energetically costly route (indicating similar behaviors,
such as foraging at a relatively low profitability patch) may well be an adaptive
strategy for animals that gain protection through aggregation, provided that
conspecifics are doing the same. Furthermore, although in this instance the information
that was transmitted between fish in the long-route conditions was
suboptimal, in general there is no doubt that animal social learning is typically
adaptive and results in the transmission of "good" information (Galef 1995).
This is, in part, because the germane features of the environment are frequently
sufficiently stable to afford equivalent levels of reinforcement to the transmitter
and receiver of information. It also probably reflects the fact that social animals
have evolved mechanisms that prevent bad information from spreading. For example,
several species of animals can identify cues associated with illness
caused by toxic substances in conspecifics, and consequently will not adopt
toxic diets, while for others the transmission of dietary preferences only works
effectively for adaptive traits (Laland et al. 1996). Nonetheless, the behavioral
tradition for guppies taking the long route may legitimately be described as
maladaptive, since if all of the members of the population were to switch to the
short route, they would all have to expend less energy in locating food, and their
search time would be reduced. The behavioral tradition favoring the long route
is associated with a potential reduction in the long-term rate of net energy gain
relative to a tradition favoring the short route.
The finding that historical as well as economic factors are important in the
foraging decisions of animals is supported by experimental studies of the diet
choices of rats influenced by social cues. Laland and Plotkin (1991) found that
the equilibrium mean proportion of the diet represented by food items for which
there is socially transmitted information cannot always be predicted from ant
mals' consumption of such food items in the absence of social information. Here
diet composition depends on past traditions for exploiting particular resources,
and cannot always be predicted from palatability, profitability, or patterns of reinforcement.
These laboratory findings are mirrored by observations from the
field, which provide compelling, yet generally unproven, evidence for the importance
of historical factors. For instance, the characteristic tool-using repertoires
of particular chimpanzee populations probably reflect cultural traditions,
although differences in the ecological environments may be important. The take
home message is that, while there is no evidence that animal cultures support
maladaptive behavior, we cannot always predict a priori what the adaptive behavior
of social learners will be, since historical and frequency-dependent processes
allow traditions to stray from the optimal.