Some taphonomic effects of scavenging canids on the bones of ungulate species : some actualistic research and a Romano - British case study.

Bones of dogs are found on most archaeological sites of holocene date in several continents. The presence of tooth marks on the bones of other species often suggests that a recovered assemblage has undergone scavenging by canids and may be taphonomically biased. An actualistic study monitored the destruction, weathering and burial of bones of modern sheep and deer that had died naturally and been scavenged by foxes. The assemblages recovered after three years are biased severely towards certain element types and have suffered the preferential loss of young bones or epiphyses. Subcollections within the sheep assemblage indicate that relative frequencies of elements are different in residual and carnivore transported assemblages. Comparisons with work by other researchers indicate that: (1) inherent factors influence element survival rates, and (2) the patterns of element frequencies in assemblages from carnivore-scavenged carcasses are consistent across a wide range of environmental settings. The results of the actualistic study were applied to some Romano-British material from a military site in northern Britain. Several of the ungulate bones show tooth marks and patterns of breakage that are very similar to those observed in the actualistic study. It is very likely, therefore, that this assemblage was scavenged by dogs (bones of which were also recovered). The relative frequencies of elements of the sheep-goat and cattle assemblages match those of the modern sheep assemblage, suggesting that whole carcasses of both species were deposited at the site. The paucity of certain element types can be explained by the activities of scavenging canids together with a bias against the recovery of smaller bones, and need not be the result of cultural practices such as trading. A new method is suggested for the demographic analysis of archaeological faunal assemblages that overcomes some of the biases caused by the preferential loss of unfused epiphyses