A simple carbon isotope model for reconstructing prehistoric human diet

A compilation of experimental animal data shows that neither δ 13 C collagen nor δ 13 C apatite nor Δ 13 C CO‐AP indicate diagnostic reconstructions of diet, diet energy and diet protein. In contrast, plots of δ 13 C collagen against δ 13 C apatite provide a model of three regression lines (C 3 , C 4 , and marine diet protein) where position on each line indicates the energy source (C 3 , C 4 , or mixed). Neither body size nor trophic position appears to affect these relationships. Modern free‐ranging, terrestrial fauna do not fit the model perhaps because they, unlike the experimental fauna, mainly use fermentation rather than digestion during energy metabolism. Archaeological humans fall as expected based on associated floral and faunal evidence. Foraging people plot at positions expected from associated C 3 fauna and plants. Those from Cahokia plot, as expected, from associated deer, nuts, and maize whereas people from nearby smaller sites plot in positions consistent with eating more fish. Agriculturists from Ontario and Grasshopper Pueblo plot consistent with dependence on fish by the former and on turkeys by the latter. In Tierra del Fuego, people from interior regions ate more terrestrial fauna, as suggested by ethnohistoric reports, than did people from the coast. In the Southwestern Cape in South Africa individuals late in the sequence have pure C 3 diets whereas ones early in the sequence ate marine protein as suggested by independent archaeological evidence. People on San Nicolas Island depended on C 4 plants in contrast to other islands off California's coast. This simple model provides more detailed and precise dietary information than do individual isotopic measures. Am J Phys Anthropol, 2007. © 2007 Wiley‐Liss, Inc.