Sexing Viking Age horses from burial and non-burial sites in Iceland using ancient DNA

• Published in: Journal of Archaeological Science
• Main Authors: Nistelberger, Heidi Maria, Palsdottir, Albina Hulda, Star, Bastiaan, Leifsson, Rúnar, Gondek, Agata Teresa, Orlando, Ludovic, Barrett, James H., Hallsson, Jón Hallsteinn, Boessenkool, Sanne
• Format: Article in Journal/Newspaper
• Language: English
• Published: Elsevier Science 2018
• Subjects: Iceland
• Online Access: http://hdl.handle.net/10852/68759; http://urn.nb.no/URN:NBN:no-71907; https://doi.org/10.1016/j.jas.2018.11.007

Horses are the most common grave goods found in Icelandic Viking Age graves. Horse skeletons have previously been sexed based on pelvis shape and the presence of canine teeth in male horses over 4–5 years of age. Morphological data has shown that all horses from Icelandic burials that were amenable to sexing were male. Yet the use of morphological methods to determine sex is problematic since they rely on finding a well-preserved pelvis and/or robust canine teeth. Furthermore, quantitative data underlying the features of the horse pelvis used for sexing is lacking and canine teeth have been reported to occur in mares. In this study we build upon and extend recently developed methodologies to make use of shotgun sequencing of ancient DNA (aDNA) for molecular sexing of Viking Age horse remains. With minimal sequencing effort we identified the sex of the largest collection (n=22) of Viking Age Icelandic horses studied to date, sourced from both burial (n=19) and nonburial (n=3) sites. Our results revealed a male to female sex bias ratio of 18:1 in burial sites, versus 0:3 in nonburial sites. These findings support the significant over-representation of male horses in Viking Age graves in Iceland, yet show that –albeit rare– mares could also be selected for ritual burial in Viking Age Iceland. This costeffective method provides statistical confidence to allow for sexing of highly fragmented archaeological specimens with low endogenous DNA content.