Abstract

M. Golabadi, A. Arzani, and S. A. M. Mirmohammadi Maibody

Wheat improvement for drought tolerance requires reliable assessment of drought tolerance variability among segregating populations. One hundred and fifty-one F3 and F4 families of durum wheat derived from a cross between Oste-Gata (as drought tolerance) and Massara-1 (as susceptible) genotypes were evaluated both under moisture stress (E1) and non-stress (E2) field environments using a randomized complete block design for each environment and growing season (2003-04, 2004-05) . Entries of E1 were subjected to moisture stress at grain filling period. Five drought tolerance indices comprising: stress tolerance index (STI), stress tolerance (TOL), stress susceptibility index (SSI), mean productivity (MP), and geometric mean productivity (GMP) were used. The indices were adjusted based on grain yield under drought (Ys) and normal (Yp) conditions. Analysis of variance for each individual year showed that there was a significant genetic variation among families for all criteria with the exception of SSI. The combined analysis of variance over seasons indicated the genetic diversity of lines, significant variation of seasons and differential response of genotypes over seasons for all indices with the exception of SSI. The significant and positive correlations of Yp and (MP, GMP and STI) and Ys and (MP, GMP and STI) under both the seasons as well as significant negative correlation of SSI and TOL in E1 revealed that selection could be conducted for high MP, GMP and STI under both environments and low SSI and TOL under E1 conditions. The calculated correlation coefficients revealed that STI, MP, and GMP are the superior criteria for selection of high yielding genotypes both under E1 and E2. Cluster analysis of families using Ys, Yp and five other indices categorized genotypes into five groups each of which having 37, 56, 13, 34 and 11 genotypes in year 2003-04 growing season, respectively. Based on 2004-05 growing season data, six groups each of which having 25, 9, 25, 45, 10 and 37 genotypes were obtained, respectively. Cluster analysis distinguished groups contains superior lines for both E1 and E2, superior lines for only E1 conditions and superior lines for E2 conditions, considering their yield performance (Yp and Ys). Results of calculated gain from indirect selection indicated that selection from moisture stress environment would improve yield in moisture stress environment better than selection from non -moisture stress environment. The comparison of the number of families in common within the top 25% families at E1 in year 2004-05 and those selected using various indices indicated that drought tolerant indices could perform comparable with yield performance (Yp and Ys).