Mega-Environment and Multivariate Statistical Analysis of Hard Winter Wheat Grown in Washington State.

Multi-environment trials of crop breeding lines are conducted annually to identify superior cultivars for the target region and also to develop understanding of the target region. The objectives of this study were to determine whether subdividing the hard winter wheat (Triticum aestivum L.) production areas in the eastern of state of Washington into mega-environment and to classify wheat genotypes and existing relations among agronomic traits. The information that was used in this study regarding grain yield, percentage of protein, plant high and heading date were obtained for 17 winter wheat genotypes from WSU field trials that were conducted in 17 locations of the eastern region of Washington State during 2000-2012. The genotype and genotype x environment (GGE) biplot method was used to subdivide the wheat production areas into subregions. In addition, Principal Component Analysis (PCA) and Cluster analysis by Ward method use of standardized means was carried out to study the main source of variability of winter wheat cultivars. All statistical analysis was performed using the Statistical Analysis System (SAS) V.9.2. The preliminary results showed that the GGE biplot analysis based on the averaged seven varieties that were common to trials in 2000 to 2006 and the 10 varieties that were common to trials in 2007 to 2012 separated Walla Walla from other locations, indicating that the subdivision of wheat production areas into mega-environment is justified for eastern Washington. Therefore, eastern Washington should be considered as two winter wheat mega-environment: a minor mega-environment (Walla Walla) and a major one. Multivariate Statistical Analysis using PCA identify that two components explained 81.6 percent of the variation and both components were associated with yield. The first one was also associated with protein concentration and the second one with plant height. The weight of protein concentration in the first component was (-0.594300) and (0.503976) for yield; the weight of the second component was (0.628284) and (-0.546083) for plant height and yield, respectively. Yield and protein concentration responded to environmental conditions. However, the variation in protein concentration is not always associated with high yield. The cluster analysis of data placed the genotypes into three groups. Comparison of traits average in the given groups showed that existing genotypes in the first group have the highest yield followed by medium and low yield.