Matrix decomposition model for investigating prehistoric intracemetery biological variation

Area of Science:

• Archaeology
• Bioarchaeology
• Population Genetics

Background:

• Analyzing biological variability within prehistoric cemeteries is challenging due to complex population structures.
• Traditional methods like kinship analysis and microchronology often require predefined subgroups, limiting their applicability.
• Previous analyses of intrasite variability struggled with accurately identifying distinct burial populations within a single cemetery.

Purpose of the Study:

• To present a novel matrix decomposition model for estimating the number of distinct burial populations within a single cemetery.
• To overcome identifiability problems common in general variability analyses.
• To provide a method for analyzing intracemetery biological variability without a priori subgroup definition.

Main Methods:

• The model correlates interindividual biological and burial distance matrices.
• It defines a residual matrix based on raw data, assuming a kin-structured burial pattern.
• The proportion of negative or zero residuals estimates the number of hidden subgroupings (lineages).

Main Results:

• The model successfully estimated the number of distinct burial populations in skeletal samples.
• Demonstrated that increased phenotypic variability on the Georgia coast was due to population aggregation.
• Distinguished between population aggregation and genetic admixture as causes of variability.

Conclusions:

• The matrix decomposition model effectively analyzes intracemetery biological variability.
• Population aggregation at missions, not genetic admixture, explains increased phenotypic variability in protohistoric/historic Georgia coast samples.
• This method offers a robust approach to understanding past population dynamics and social structures within cemeteries.