The scientific identification of unknown human skeletal remains in forensic contexts relies heavily on the estimation of demographic parameters (i.e., sex, age, stature, and population affinity). Population affinity, or the likelihood of group relatedness to a defined population of a decedent, can be estimated using measurements and observations from the cranial and postcranial skeleton. These estimations may be less accurate among populations which have been pooled together based on convention. Latin American individuals-with geographic origins widely distributed throughout Central and South America-are broadly pooled together under the blanket term Hispanic with little regard for the immense cultural and biological diversity represented by these groups. Consequently, forensic anthropologists may be unintentionally disregarding genetic diversity, population structure, and population history and their impact on the formation and morphology of these groups. The purpose of this dissertation is to investigate variation in craniofacial morphology and develop population affinity models for Latin American groups using cranial metric and nonmetric data. The intent is to move beyond a single classification level (i.e., Hispanic) to more refined levels based on geographic origins (e.g., Guatemala, Southeast Mexico). The broad category of Hispanic was adopted by forensic anthropologists in large part because it is still used in medicolegal death investigations in the U.S. to describe individuals with familial origins in Latin America, Spain, and the Caribbean (U.S. Census Bureau 2021). Since the term Hispanic does not narrow down the region of origin for unidentified human remains, it is uninformative for identification and repatriation purposes, particularly regarding forensic investigations along the southern U.S. border. In this context, population affinity estimation benefits from refinement of a broad category to a more focused, population-level group. Craniometric and cranial macromorphoscopic (MMS) data are collected from samples in Guatemala City, Guatemala and Merida, Mexico-with strong support from the forensic anthropologists in these countries-to capture aspects of skeletal variation associated with these regions. Biological distance and population affinity models are assessed and comparative data from other Latin American and U.S. populations are used to assess how well these model skeletal variation. Biological distance analysis demonstrates that Latin American populations, including the Meridian and Guatemala sample are distinct. Classification models obtain varying accuracy rates; the combined craniometric and cranial MMS model had the highest classification accuracy (70.7%). This study provides further support for the refinement of this broad category and is important for future investigations involved in identification efforts along the U.S.-Mexico border.
Read
