SAMPLING STRATEGIES FOR DEVELOPING INDIAN SESAME CORE COLLECTION

Abstract

Various stratified random sampling strategies were compared for developing a core collection of
indigenous sesame Sesamum indicum L accessions. Twenty diversity .groups (strata) representing 3,129
Indian sesame accessions were used for the study. The Shannon Diversity Index (SOl) pooled over
19 descriptors was used as the measure of the diversity. Simple random sampling alongwith six
stratified random sampling strategies were compared for varying sample fractions ranging from· 5 to
30%. Stratified random sampling was always superior to simple random sampling. Of the six
stratified sampling strategies, a genetic diversity dependent strategy with sampled accessions from
various strata direcdy proportional to the product of the diversity and logarithm of the size of the
strata (GL strategy) appeared to be the best strategy. A 10% sample fraction could be determined
as the suitable sample fraction, invariably for all sampling strategies, serving both objectives of
germplasm management and use.