Response to F3 Selection for Grain Yield in Durum Wheat [Triticum turgidum (L.) Thell. ssp. turgidum conv. durum (Desf.) Mac Key] under South Mediterranean Conditions

An experiment was carried out at the Field Crop Institute, Agricultural Research Station of Setif (ITGC-ARS, Setif, Algeria), to assess the response to early generation selection in durum wheat (Triticum durum Desf.). Results indicated the presence of sufficient variability to warrant selection. Correlation coefficients of biological yield, plant height, spike number, kernels per spike, harvest index and days to heading with grain yield were significant. Path analysis emphasized the direct and indirect effects of biomass yield, spike number and harvest index on grain yield, suggesting that F3-selection based on one or simultaneously on all these traits would result in significant direct and correlated responses. However, intergeneration correlation and parent-offspring regression coefficients predicted that F3-selection based on biomass yield, spike number, harvest index, 1000-kernel weight and grain yield will be ineffective while plant height and days to heading selection will be efficient. Direct and correlated responses to F3-selection for biomass yield, spike number, harvest index, 1000-kernel weight, plant height, days to heading and grain yield confirmed predictions drawn from intergeneration correlation and parent-offspring regression coefficients. Inefficiency of F3-selection was attributed to genotype x environment interaction. The overall results suggested that early generation selection should be restricted to plant height and days to heading while selection to improve grain yield directly or via related traits should be delayed to later advanced generation.