Success of a selection program for increasing grain yield of two-row barley lines and evaluation of the modified augmented design (type 2)

Abstract

The modified augmented design (MAD) (type 2), which adjusts for environmental heterogeneity when large numbers of test lines are being evaluated in non-replicated plots, has been used in the barley (Hordeum vulgare L.) breeding program at the Lethbridge Research Centre. The superiority of test lines selected from two series of F7 test lines selected using the MAD were evaluated in replicated MAD tests (F8) and superior selections were further evaluated in standard replicated designs (F9). When adjustment was indicated for individual MAD tests, the generally high relative efficiencies (RE) confirmed the effectiveness of adjustment in reducing the error variation for yield. The need for adjustment and the adjustment method differed among replicates of the MAD conducted on adjacent blocks of land. The RE from replicated MAD tests indicated the effectiveness of the MAD for reducing the variability in the tests of F8 material. Correlations of the ranks of test lines in different replicates of MAD indicated that more of the same lines in each replicate would be designated as superior with adjustment for heterogenity. The lack of significant positive rank correlations among environments and clearly identifiable improvement in the mean yield was probably the result of the narrowing range in grain yield of the test lines as well as a genotype × environment interaction. Key words: Modified augmented design, Hordeum vulgare, barley, screening