农作物品种选育通常需要对多目标性状综合选择，依据育种目标性状和权重建立品种选择指数，选择遗传差异鉴别力强和目标环境代表性好的试验点，有助于提高品种选育的效率并降低实验成本。本研究依据国家棉花品种审定标准构建针对性和实用性强的品种选择指数，即SI =0.40´皮棉产量+0.13´纤维比强度+0.09´(纤维长度+马克隆值)+0.11´抗枯萎病+0.09´抗黄萎病+0.10´霜前花率，采用GGE双标图方法，对2000—2013年间39组(含585个单点试验)长江流域国家棉花区域试验中的15个试验点，综合评价品种选择指数的鉴别力、代表性和理想指数。结果表明，湖北黄冈和江苏南京试验环境被评为最理想的试验环境，湖北荆州、湖北武汉和江苏盐城为理想的试验环境，而河南南阳、湖北襄阳、湖南常德、四川简阳和四川射洪试验环境为不理想试验环境。可以看出，理想的试验环境均位于长江流域的中下游棉区，而不理想的试验环境中四川简阳和四川射洪位于上游的四川盆地、河南南阳和湖北襄阳位于长江流域北缘的南襄盆地、湖南常德虽然位于长江流域中游但栽培密度偏低。本研究构建的选择指数采用与国家棉花品种审定中品种评价准则相统一的目标性状和权重分配策略，理想试验环境对我国长江流域棉区的棉花生态育种试验点的选择提供了切实可行的决策方案。%Crop breeding is an integrating selection for multiple desirable traits. Screening locations with high discriminating abil-ity and typical environments in light of cultivar selection index established by target traits and weights facilitates improving breeding efficiency and saving cost. Cultivar selection index was constructed on the basis of the national register criteria for cot-ton variety, namely, SI = 0.40 ´ lint cotton yield + 0.13 ´ fiber strength + 0.09 ´ (fiber length + micronaire value) + 0.11 ´Fusarium wilt+0.09´Verticillium wilt+0.10´harvesting ratio of seed cotton before frost. GGE biplot method was adopted to evaluate the identification, representativeness, and ideal index of cultivar selection index by using 15 locations data from 39 sets of national cotton variety regional trials including 585 trials in the Yangtze River Valley (YaRV) during 2000–2013. The results showed that Huanggang in Hubei Province and Nanjing in Jiangsu Province were evaluated as the most desirable trial locations;Jingzhou and Wuhan in Hubei Province and Yancheng in Jiangsu Province were desirable trial locations;while Nanyang in Henan Province, Xiangyang in Hubei Province, Changde in Hunan Province, Jianyang and Shehong in Sichuan Province were consid-ered as undesirable locations for cotton cultivar selection. The ideal environments were all located in the Middle and Lower Reaches of YaRV, while the undesirable locations included Jianyang and Shehong in the Sichuan basin in the upper reaches of YaRV, and Nanyang and Xiangyang in Nan-Xiang Basin at the northern border of YaRV. Changde in Hunan Province was also considered as undesirable location although it is located in the Middle Reaches of YaRV, probably because of the significantly low plant density used in the farming. In conclusion, this study has established a feasible selection index according to the national cotton registration criteria, and identified desirable test locations for reliable and effective variety evaluation in the area of YaRV. This study is an example of test location evaluation utilizing historical data for the similar studies in other regions and for other crops.
展开▼
