棉花品种区域试验适宜试验点数量的抽样估计

Sampling Estimation of Suitable Quantity of Test Sites in Cotton Variety Regional Trials

基于长江流域国家棉花区域试验的对照品种泗棉3号、湘杂棉2号、湘杂棉8号和鄂杂棉10号在2000-2011年期间84～270次试验中早熟性、农艺性状、产量性状和纤维品质性状表现,构建了对照品种抽样总体,采用随机抽样方法估计了在不同精确度水平下鉴别棉花品种主要性状表现所需要的试验点数量,旨在为长江流域棉区国家棉花区域试验的试验点数量设置提供理论依据,并为其它棉区乃至其它作物区域试验中适宜试验点数量的估计提供参考方法。研究结果表明:试验点数量的需求与目的性状选择及变异度密切相关,皮棉产量的表型标准差最高,在同等精确度水平下需要的试点数量也最多。目前,长江流域棉花区域试验设置18个试点,对皮棉产量的估计精确度为90%;增加16个试点,估计精确度将提高到93%;而随后再增加试点数量对提高估计精确度收效甚微。

The sampling populations of four check varieties, namely, Simian 3, Xiangzamian 2, Xiangzamian 8 and Ezamian 10 were constructed and analyzed based on their performances in earliness, yield, fiber quality and agronomic traits in 84 to 270 time tests in national cotton regional trials in the Yangtze River Valley during 2000-2011 period. And the adequate and neces- sary quantity of regional trial sites to identify cotton main characteristics were estimated using random sampling method within the cultivar population in accordance with the responding sampling estimation accuracy ratio, so as to investigate the theory and methodology in determining the suitable quantity of cotton regional trial locations in the Yangtze River Valley, and also act as a useful demonstration for other cotton planting regions and even other crops in regional trial sites determination. Research results showed that the number of sites demanded was closely related with the target selection trait and its variation degree, so the lint yield with the highest degree of variation demanded most quantity of trial locations under the same estimation precision level. At present, the estimation accuracy of lint yield reached 90% in national cotton regional trials with eighteen trial sites in the Yangtze River. In case the estimation accuracy was set to 93%, additional sixteen test sites must be arranged, and the subsequent increase of test sites was estimated to be less effective in estimation accuracy enhancement.