Loaves and fishes: a stable isotope reconstruction of diet in medieval Greece
By Sandra Jean Garvie-Lok
PhD Dissertation, University of Calgary, 2001
Abstract: The historical sources on medieval Greek diet provide extensive information on the identity of foods consumed, but are less informative regarding the proportions in which they were consumed. Thus, there are a number of outstanding questions regarding diet in medieval Greece. These include the importance of animal proteins, especially marine resources, the importance of maize and millet, and the existence of ethnic or gender variation in food use. This dissertation sheds light on these issues through analysis of archaeological human remains.
Collagen and carbonate stable isotope analyses are performed on human bone from ten medieval and early modern Greek cemeteries, representing Greek Orthodox, Frankish and Ottoman Muslim populations. The results do not support extensive legume use by any of the populations. Instead, the diet likely included substantial amounts of animal protein, perhaps in the form of dairy products and eggs. There is no evidence for general marine dependence by any of the ethnic groups. However, a relationship between collagen values in two island populations likely reflects the consumption of some marine resources. Carbonate 5 C values suggest reliance on C3 grains, with supplementary consumption of C4 grains. This consumption is greatest in Ottoman groups, perhaps reflecting the introduction of maize to the Mediterranean at that time. Ottoman values are also more variable than those of earlier groups; this may reflect increased population movement in the Ottoman era.
The dissertation also addresses issues in the preparation of archaeological bone carbonates. The results of an experiment examining the effects of different acid treatment solution concentrations and treatment times on samples indicate that shorter treatment with a more dilute solution will produce more dependable data. They also suggest that differences in solution concentration lead to systematic differences in sample S13 C; this has implications for the comparison of data in the literature. Finally, the dry yield by weight of the preparation process is shown to serve as a proxy of bone organic content. As such, this measure has potential value as an indicator of sample quality.