Simulating the Formation of Dual-Patterned Archaeofaunal Assemblages with Experimental Control Samples

Abstract

Attempts to discern bone modification by early hominids from that by carnivores are limited due, in part, to the fact that most inferential models are based on the singular actions of either hominids or carnivores. In order partly to rectify this situation, I conducted two types of two-stage, or dual-patterned, actualistic studies to simulate whole carcass processing by hominids (stage one) followed by carnivores (stage two) in the Serengeti and Ngorongoro ecosystems of Tanzania in 1990 and 1991. Thirty whole-bone to carnivore experiments, which model hominid defleshing, and 39 hammerstone to carnivore experiments, which model the removal of flesh, marrow, the brain, and the mandibular nerve by hominids, were conducted to model variability in the behaviour of early hominids and determine its effect on secondary bone modification by carnivores.Vultures, maribou storks, jackals, and spotted hyenas were observed modifying bones from experiments in both samples; however, the overwhelming majority of bone modification was conducted by spotted hyenas. In stage two, spotted hyenas (1) deleted between 90 and 100% of the axial and compact elements placed in both experimental samples; (2) deleted 79·7% of long-bone epiphyses produced by hammerstone breakage and 98·9% of epiphyses from intact long bones; (3) placed at least one tooth mark on 64·3 and 23·2% of all the specimens recovered from whole-bone to carnivore and hammerstone to carnivore experiments, respectively; and (4) added and subtracted specimens which altered the number of specimens bearing at least one cut mark, scrape mark, and/or percussion mark. These results all reflect the cumulative and variable actions of hominids (via simulation) and spotted hyenas. Therefore, it is not possible to delete traces produced by carnivores without removing data indicative of hominid behaviour as well.