Ancestry determination using geometric morphometrics

Ancestry identification is an important aspect in forensic anthropological investigations concerning the biological and social identity of unidentified human remains. Traditionally, ancestry is usually assessed by the visual examination of the shape and form of the features of the skull and/or measured by linear distances between certain cranial landmarks. In this research, three-dimensional Procrustes based geometric morphometric methods are employed to investigate whether these techniques can classify ancestry in a forensic context using the cranium. Cranial landmark and semilandmark configurations, focused specifically on features routinely used to make ancestry assessments using traditional morphological and morphometric techniques, are used to classify ancestry using discriminant function analyses with leave-one-out cross validation. Landmarks and semilandmarks were captured on 3D cranial images of 31 sub-Saharan African and 31 European individuals and variables of size, shape, and form were used to assign crania into respective groups based on different configurations of points. Results demonstrate that variables of shape and/or form (i.e. size and shape) are more accurate at classifying ancestry than size alone, and that an overall aspect of the cranium is more accurate at classifying individuals into their respective ancestry groups than semilandmarks recorded along specific features, such as the border of the orbits, the edge of the lower nasal aperture, and along the vault on the midplane.

Name dc:type.qualificationname

Master of Forensic Anthropology

Level dc:type.qualificationlevel

Master's Thesis

Grantor dc:publisher.institution

University of Dundee

Year dc:date.issued

2012

Author dc:creator

• Gillick, Hana

• Forensic anthropology
• Ancestry
• Geometric morphometrics

Language dc:language

eng

Identifier

oai:discovery.dundee.ac.uk:studenttheses/31cdae88-c1db-4dd3-9df9-be3f5f073854

OAI identifier oai:identifier

oai:discovery.dundee.ac.uk:studenttheses/31cdae88-c1db-4dd3-9df9-be3f5f073854