Journal of Anthropological Archaeology

0278 © 2002 All rig Archaeological tests of hypotheses drawn from foraging theory face a unique set of challenges. Simple foraging models, such as the diet breadth model, rely on assumptions that are clearly violated in the human case. Testing is complicated by the indirect nature of the observations used to reconstruct environment and behavior and by the cumulative nature of the archaeological record. However, the negative impact of these issues on understanding can be ameliorated by adding and valuing research strategies that go beyond those those designed to test hypotheses derived from the model against archaeological evidence. One such strategy is to probe the model’s failures by manipulating constraints and variables. The model’s performance under varying environmental conditions constitutes a partial test of alternative explanations of behavior. The value of such an approach is illustrated by a case study involving plant use by early food producers who lived in the rugged hill country of eastern Kentucky during the early 3rd millennium B.P. Archaeobotanical data suggest changes in the dietary contributions of different mast-producing tree species during the transition to food production. Possible explanations for these changes were evaluated using linear programming. Running the model under varying conditions of resource availability showed that the broad-based mast diet inferred from nutshell assemblages was probably not energetically optimal. Although chestnut is profitable to exploit when hickory is limited, acorns are too costly to use under most environmental conditions unless efficent processing techniques are used. The substitution of starchy seeds for nuts with similar nutritional characteristics would have been inefficient, although seed crops are potentially important sources of macronutrients and energy when mast supplies are depleted. These findings point out vulnerabilities in economic efficiency-based explanations for the origins of agriculture in eastern North America. © 2002 Elsevier Science (USA)