Introduction

Upon sufficiently close inspection, virtually all animals will show spatial variation in their behavior. Most studies of behavioral geography have focused on local adaptation, in part based on genetic differences in learning predispositions (e.g., Foster and Endler, 1999). Only a few studies have assumed that the geographic variation in behavior was affected by social learning; that is, it was traditional rather than genetic in origin (e.g., Galef, 1976, 1992, 1998). Primate studies, however, are clearly the exception to this rule: social transmission is often thought to be important (e.g., Wrangham, de Waal, and McGrew, 1994). Unfortunately, descriptive field studies face various obstacles, making it very difficult to demonstrate unequivocally that differential invention and social transmission underlie the pattern of geographic variation.

Local variants can be defined as behaviors that show geographically patchy distribution (Table 11.1.) Following Galef (1976, 1992) and Fragaszy and Perry (Ch. 1), it appears that three criteria must be met to decide that a local variant qualifies as a tradition: (a) the local variant must be common, shown by multiple individuals (cf. McGrew, 1998); (b) it must be long lasting, probably persisting across generations; and (c) it must be maintained by some form(s) of social learning.

Field studies can yield information on condition (a) and, with some patience, on condition (b); however, the processes underlying the acquisition of behavior (condition (c)) are notoriously difficult to study in the wild.