An appraisal on the usefulness of R-matrix analysis and the model of Harpen
ding and Ward (1982) to study the population structure of Indian population
s is made in the light of prevalent marriage patterns, caste structure, and
cultural specificity of the region. With the help of available data on the
migration histories and geographical backgrounds of marine fishermen on th
e east coast of India, and through the historical analysis of marriage patt
erns, it is demonstrated that the routine inferences based on the regressio
n plots of average geterozygosity versus genetic distance from the centroid
may not generally be apt for the Indian situation. Increased heterozygosit
y among migrant fishermen seems to have resulted from mating patterns withi
n this community rather than from external gene flow. In either case, the g
enetic implications are supposedly identical. Nevertheless, when making inf
erences regarding local population structure it is important to have detail
ed knowledge of mating patterns and the cultural context of the region unde
r study.