On methods of spatial analysis for genotyped individuals

Area of Science:

• Ecology
• Genetics
• Spatial Statistics

Background:

• Spatial autocorrelation methods are widely used in individual-based spatial genetic studies.
• The properties and relationships among these spatial statistics require careful examination.

Purpose of the Study:

• To reformulate widely used spatial statistics within a point process framework.
• To clarify the interpretations of Moran's I and the number of alleles in common (NAC) for genetic data.
• To provide a unified theoretical basis for spatial analysis in molecular ecology.

Main Methods:

• Reformulation of spatial statistics using point processes.
• Analysis of Moran's I and NAC statistics for discrete allele frequencies.
• Application of the point process framework to molecular ecological data.

Main Results:

• Moran's I and NAC statistics can be expressed as weighted sums of join-count statistics for allele frequencies.
• Moran's I amplifies minor genotype distributions based on allele frequency, while NAC uses constant weighting.
• The point process framework enables spatial analysis across different genetic levels (genotype, allele).

Conclusions:

• The point process framework offers a unified approach for spatial genetic analysis.
• This reformulation enhances the understanding and application of spatial autocorrelation statistics in molecular ecology.
• The methodology is validated through application to Fagus crenata population dynamics data.