Classification of Diverse Bread Wheat-Growing Environments Based on Differential Yield Responses

Abstract

Bread wheat (Triticum aestivum L.) environments in West Asia, North Africa, and Mediterranean Europe are highly variable in terms of moisture, temperature, and biotic stresses. The present study attempted to divide this region into relatively uniform subregions by cluster analysis, to reduce the large magnitude of genotype (G) ✕ environment (E) interaction. Grain yield data of 21 to wheat lines in 2 yr were analyzed by a hierarchical agglomerative program with the correlation coefficient as a distance measure and average linkage as the clustering strategy. The large and significant entry ✕ trial interaction detected in the region implied that breeding for wide adaptability for this region would be difficult. Based on differential yield responses of the wheat lines, the cluster analysis showed the differences between the irrigated/high-rainfall (IHR) and rainfed low-rainfall (RLR) sites. Besides moisture supply, winter temperature of the sites appeared to be another determinant of the clusters. The IHR clusters consisted of sites with milder temperature than sites in the RLR clusters. The presence or absence of diseases also influenced the delineation of the clusters, but photoperiod of the sites did not appear to have a large influence. The greater variability in RLR sites relative to IHR sites indicated that breeding for RLR environments would be more difficult. We conclude that different breeding strategies should be followed within IHR and RLR environments.