Supersize does matter: the importance of large prey in palaeolithic subsistence and a method for measuring its significance in zooarchaeological assemblages

DSpace Repository

URI: http://hdl.handle.net/10900/114212
http://dx.doi.org/10.15496/publikation-55587
Dokumentart: Buch (einzelnes Kapitel)
Date: 2021-04-14
Language: English
Other Keywords: megafauna
proboscideans
abundance index
Palaeolithic
zooarchaeology
Show full item record

Inhaltszusammenfassung:

Humans consumed megaherbivores, including proboscideans, throughout the Pleistocene. How- ever, there is a high potential for underappreciation of their relative importance to humans’ economy due to their potential relative underrepresentation in Palaeolithic archaeological sites. Relying on our previous work, we discuss the critical importance of large animals in human prehistory. We review four factors that made megaherbivores critical- ly important to humans: high ecological biomass density, lower complexity of acquisition, higher net energetic return, and high fat content. We propose a model that intends to overcome the potential underrepresentation bias by multiplying the MNI (Minimum Number of Individuals) of each animal species by its weight and only then determining the relative biomass abundances. The next step of the model is the accumulation of the relative biomass abundance, beginning with the largest animal. This step enables a comparison of various assemblages in the relative complexity of acquisition, the level of net energetic return, and the level of fat content in the prey. We successfully test the method on an actualistic case of 61 hunts of the Hadza, where the true number and the MNI are known. We then ap- ply the method to three comparisons between two successive cultural periods each, in the Levant, East Africa and Southern France. We find that there is indeed great potential for the underrepresentation of megaherbivores in the analysis of Palaeolithic faunal assemblages. Since the largest animal in our actualistic study was a giraffe, we propose a future avenue of research for better correction of the un- derrepresentation of elephants, which often have partial to no representation in central base sites.

This item appears in the following Collection(s)