Understanding Variation in Prehistoric Agricultural Productivity: The Importance of Distinguishing among Potential, Available, and Consumptive Yields

Abstract

Based on a model derived from an analysis of contemporary maize yields in Tennessee, Baden and Beekman claim that Mississippian yields would have ranged between 8 bu/acre (501.7 kg/ha) and 30 bu/acre (1,881.3 kg/ha). Using nineteenth-century observations of Native American farmers, I noted in 1999 that available maize yields ranged between 3.7 and 42.67 bu/acre (232.1 to 2,676.3 kg/ha), with a mean of 18.9 bu/acre (1,185.4 kg/ha) for groups that did not have plows. Consumptive yields would have been lower, probably closer to an average of 10 bu/acre (627.2 kg/ha). In this paper, I clarify the differences between potential yields, available yields to illustrate the advantages of my approach. I discuss some factors that affect maize plants prior to harvest, leading to available yields that may be lower than potential yields, and conditions that reduce the quantity of maize kernels available for consumption after the harvest. Baden and Beekman argue that modern agricultural technology provides a more reasonable baseline analog for modeling ancient maize productivity than nineteenth-century Native American technologies. In contrast, I explore agricultural yield data for Native Americans and Euroamericans from a number of tribes and states for 1850, 1867, and 1878. A comparison of these data shows that, overall, yields obtained by Native American farmers tend to be lower than yields for contemporaneous Euroamerican farmers. My approach using agricultural productivity data from nineteenth-century Native Americans, coupled with a consideration of potential, available, and consumptive yields, provides a plausible foundation for the evaluation of late prehistoric yields.