This post is by Haskel J. Greenfield, lead author of the recently published paper “Estimating the Age- and Season-of-Death for Wild Equids: a Comparison of Techniques Utilising a Sample from the Late Neolithic Site of Bad Buchau-Dullenried, Germany“. Greenfield is a Professor in the Department of Anthropology, University of Manitoba, Canada.
The genesis of this study was an odd email request from a German colleague (Karlheinz Steppan) to see if someone was willing to conduct a dental cementum analysis of wild equid specimens from the Neolithic settlement at Bad Buchau-Dullenried, Germany. In my usual exuberance (and naiveté), I thought that this was an interesting opportunity to learn about horse cementum, while collaborating on the analysis of Neolithic material (always a passion of mine). I had already done previous cementum work with my students on sheep and goat. How difficult could it be to do equids, I thought. Well, I should have spoken with my good friend and colleague, Arianne Burke at the University of Montreal, before I embarked on what became a six year quest. On the surface (pun intended!), it was actually much easier to thin-section equid teeth than sheep and goat. Equid teeth were larger and the cementum layers longer and thicker. I thought it should be a “no brainer” to thin-section them and do the seasonality and ageing analyses.
A horse skull. Image from Wikipedia.
In fact, nothing proved to be simple other than recruiting Collin Moore (at that time, a MA student in Biological Anthropology at the University of Manitoba; currently, finishing his PhD at the Max Planck Institute for Evolutionary Anthropology, Leipzig) to help me with the project. Collin was a particularly apt collaborator for this research since his specialty was the analysis of dental microstructures (albeit in humans) with new digital technologies.
Once Karlheinz had sent us the specimens, we began to prepare our documentation and analytical protocol. The thin-sectioning stage went relatively easily once we had selected, documented, and prepared the specimens. Once the slides were ready for analysis, we began by independently examining them in a traditional light optical microscope. Of course, the results were contradictory. Then, Collin suggested that we use the new digital microscope that my colleague, Robert Hoppa, had installed in the Bioanthropology Digital Image Analysis Laboratory at the University of Manitoba. This opened up a wealth of visual and data analytic techniques for mapping and analysing the cementum increments. The problem was: how to interpret the results? This is where the science and the art begin to separate in our discipline. The scientific techniques were there, but the interpretation becomes the art. It is in the art that all the techniques and analysts begin to differ, we discovered.
Four of the samples used in the study. From Greenfield et al. 2015.
Given our initial consternation at having come up with different results for ageing and season of death of the samples, we decided to use the specimens to test the various ageing and seasonality techniques for equids that have been published and become widely accepted by zoologists and zooarchaeologists. This forced us to backtrack and begin the analysis again, without any preconceptions. First, we analysed the remains using traditional veterinary and zoological techniques, such as incisor tooth wear. Then, we tried using Marsha Levine’s commonly applied cheek tooth height formula. Finally, we began to analyse the specimens using well-established and age verified Arianne Burke’s protocol for dental cementum. These were then compared with our results from the new automated technology.
The results were surprising—and disconcerting. The tooth eruption and wear and crown height measurement techniques yielded the lowest age estimations and were not useful for seasonality determination. Simply based upon comparison with modern tooth eruption analogues, the age of any individual analysed in this manner would be underestimated. The manual increment counting age estimation technique yielded wider variation in age estimation than the other techniques, and older age estimation than tooth wear/eruption or crown height measurements. The line histogram technique yielded narrower range of age estimation than manual observations, but still yielded a slightly older age range. The automated age estimation technique yielded the narrowest range of age estimation. However, age values were slightly higher than with the manual or line histogram results.
In contrast, all of the thin-sectioning techniques yielded a similar range for the season-of-death of the animal. However, we found that it was most efficient to use the automated technique to pinpoint when in the cold or warm seasons that the specimen may have died. Overall, the automated technique was by far the most efficient in terms of time and effort, and the most accurate when comparing teeth from the same individual.
The automated technique provides the narrowest age- and season-of-death range. It is the most effective, as well as the most efficient in terms of time and energy, technology to use. The application of the new digital technologies is the way out of the intellectual trap of “which technique is superior” because it removes much of the “art” (or human observer variability) from analysis. What do we learn from this experiment? The most important result is to never simply to use one technique of analysis. Which technique is closest to the truth? Who knows, but we learned a lot while doing it!
You can read the full article, completely free, in our latest issue of Open Quaternary.