棉籽蛋白质和油分含量预测的生态模型

Ecological model for predicting the protein and oil contents in cottonseed

为研究棉籽蛋白质和油分含量与环境因子间的定量关系,选择不同熟性棉花品种在长江流域下游棉区和黄河流域黄淮棉区各2个地点进行分期播种和施氮量试验,分析了品种、主要气象条件和施氮量对棉籽蛋白质和油分含量的影响。结果表明:棉籽蛋白质和油分含量的主要影响因子除品种外依次为铃期平均温度、太阳辐射量和施氮量;棉籽蛋白质和油分形成的最适宜铃期日均温分别为26.1℃和25.7℃;充足的光照不利于棉籽蛋白质和油分含量的提高;增加施氮量提高了棉籽蛋白质含量,降低了油分含量。建立了棉籽蛋白质含量与油分含量的生态模型,模型以品种、铃期平均温度、铃期日均太阳辐射量和施氮量为模型输入,简便易行。模型对棉籽蛋白质含量和油分含量预测的根均方差(RMSE)分别为2.03%和2.54%,具有较好的预测性。

Taking different cotton cuhivars with different maturity stage as test materials, a field experiment of two sowing dates and three N application rates was conducted in the lower reaches of Yangtze River basin （Nanjing and Huai＇ an） and Yellow River basin （Xuzhou and Anyang） , with the effects of cuhivar, main meteorological conditions, and N application rate on the cotton- seed protein and oil contents analyzed, aimed to understand the quantitative relations of cotton- seed protein and oil contents with environmental factors. Besides cotton cultivar, the major factors affecting the cottonseed protein and oil contents were in the sequence of the mean air temperature at bolling stage, solar radiation, and N application rate. The optimum mean air temperature for the production of cottonseed protein and oil at bolling stage was 26.1 ℃ and 25.7 ℃, respectively. Sufficient illumination did not favor the enhancement of cottonseed protein and oil contents. Increasing N application rate increased the cottonseed protein content but decreased the cottonseed oil content. Based on the statistical analyses, an ecological model for predicting cottonseed protein and oil contents was established, taking cuhivar, mean air temperature and solar radiation at bolling stage, and N application rate as the input parameters, being easy to be operated. The root mean square error （RMSE） of the prediction of cottonseed protein content and oil content by the model was 2.03% and 2.54%, respectively, showing a good predictability of the model.