A Bayesian approach to ageing sheep/goats from toothwear.

Abstract

For the last 20 years the human osteology literature has contained discussions of statistical problems with traditional skeletal ageing methods, which lead to “age mimicry”, where the estimated age structure partly resembles that of the reference population. This paper discusses how these problems also apply to animal ageing techniques and proposes a method to overcome them in the case of ageing of sheep and goats by toothwear. A form of transition analysis is developed, using the data of Deniz and Payne as a reference population. The method is tested with other published known age animals. Payne’s A–I mandible wear stages are investigated and found to underestimate variability. The method is applied to the large assemblage of sheep and goat mandibles from Arene Candide, Italy to illustrate the extent of uncertainty in estimating survivorship curves. The limitations of the current model and possible improvements are discussed.