Yield evaluation and identification of drought tolerant lines based on stress tolerant indices, ranking method and multivariate analysis in sweet potato (Ipomoea batatas Lam).

• The correlation analysis highlighted that the indices STI, MPI, GMP, YI and HM are suitable for selecting drought-tolerant genotypes as they exhibited strong correlations with yield under both stress and non-stress conditions. • Based on ranking method, cluster analysis and bi plot analysis, the genotypes, Dhenkanal local-2, 84 × 14, SB21/57, Howrah, S-783 and 84 × 1 were identified as the most drought tolerant. • In stress conditions, storage root yield was negatively associated with SSI and YR, suggesting that genotypes with low SSI and YR values were drought tolerant. • PC1 had high positive coefficients for indices such as Yp, Ys, STI, MPI, GMP, YI and HM, has effectively identifying tolerant genotypes with high storage root yield. PC2 had high positive coefficients for TOL, SSI and YR, for identifying susceptible genotypes. Sweet potato is a vital food crop in tropical and sub-tropical regions, but its sustainable production is hampered by drought stress. This study was conducted at ICAR-CTCRI, Regional Centre, aimed to evaluate 27 sweet potato genotypes to identify lines with drought tolerance. The experiment employed a split-plot design with a randomized complete block arrangement under both irrigated and drought stress conditions over two years, during Sep-2017 to Jan-2018 and Sep-2018 to Jan-2019. The results of the combined analysis of variance revealed significant differences among main effects and their interactions for storage root yield over the two years. The stress tolerant indices, correlation analysis, ranking methods, cluster analysis, principal component analysis and biplot analysis were employed to identify drought tolerant genotypes. The stress tolerance index, mean productivity index, geometric mean productivity, yield index and harmonic mean productivity, were found to be the best indicators for identifying high yielding genotypes under both conditions. The genotypes Dhenkanal local-2, 84 × 14, SB21/57, Howrah and S-783 emerged as the most drought tolerant genotypes based on mean rank, standard deviation of ranks and rank sum. The cluster analysis grouped the genotypes into three distinct groups, highlighting SB21/57, S-783, 84 × 14, Dhenkanal local-2, Howrah, 84 × 1, Baster-11 and X-79 in group II as tolerant genotypes. The principal component analysis demonstrated that the first and second components accounted for 99.042 % of the variations. The biplot analysis further confirmed that the genotypes S-783, 84 × 14, SB21/57, Dhenkanal local-2 and 84 × 1 displayed significant drought tolerance. Among all the genotypes, Dhenkanal local-2, 84 × 14, SB21/57, Howrah and S-783 were identified as the most drought tolerant. These genotypes are recommended for breeding for drought tolerance in sweet potato. [ABSTRACT FROM AUTHOR]