Abstract: Atpresent, arithmetic mean method is commonly used to evaluate the yield stability and adaptability of oat (Avena sativa) varieties, and analysis of variance is used to estimate the error in national regional trials. However, it is difficult to evaluate the varieties accurately, because of the discrepancies across tests and years. In this study, GGE-biplot method was used to analyze the date from the national regional trials for oat from 2006 to 2008. Twelve lines of naked oat were tested in ten sites, and six lines of oat were tested in six sites. The results showed that Bayou 8, Bayou 3, and Zhangyan 2 had higher yield and better yield stability than other lines. Keyan-1 and Baoluo possessed special adaptabilities. For naked oat, Zhangbei of Hebei province, was the best representative testing site; Baicheng of Jilin, Wuchuan of Inner Mongolia, Urumchi of Xinjiang, Chongli of Hebei, and Guyuan of Gansu had better discrimination than other sites, where as, Huining of Gansu and Youyu of Shanxi were the worst. For the oat test, Wuchuan of Inner Mongolia had the best representative; Xining of Qinghai had the best discrimination, but Zhangbei of Hebei was the worst. This article provides a simple and effective method for analyzing data, and evaluating the yield stability and adaptability of oat varieties involved in national regional trials oat and favorite test sites.

Key words: GGE biplot, Oat, Variety stability, Variety adaptability