Research On Evaluation And Analysis Of Shanxi Province Nicotiana. Tabacum Yield

In order to breed new varieties of tobacco,ensure the production of high-quality tobaccoleaf.This study combined with the2012tobacco varieties in Shanxi province regional test dataand in19tobacco varieties for the object.Using four different analysis methods and onecomprehensive evaluation method are analyzed and evaluated tobacco varieties and havemade a detailed analysis of the various methods.By comparing the results,it can be seen that Franics-Kannenberg, adaptive parametermethod model and high stability method can accurately reflect the varieties of high and stableyield and adaptability. However, this does not take into account the interaction between theenvironment and the variety, this is caused by evaluating stability or lack of accuracy. Shuklamodel can make up for the three methods above mentioned shortcomings, considering theinteraction effect of environmental factor and genetic factor of tobacco varieties, theinteraction of environment and varieties, and yield more accurate evaluation of tobaccovarieties. But this method can not reflect the varieties of high yield, interaction value islimited to the sum to zero. The chart of radar can be a comprehensive evaluation of allcharacters included by tobacco varieties, but the large amount of calculation, the calculationprocess is relatively complex.Finally the following conclusions can be drawn:1. In the southern region,high yield,stable yield of good varieties are Yunyan87、6517and Bina1.Comprehensive evaluation of the best varieties is Bina1.The bad varieties areGuiyan1、Xiangyan3、K326and Qingyan96.2. In the northern region,high yield,stable yield and adaptability of bad varieties areCF9010and0408.Comprehensive evaluation of the best varieties is LJ981.The bad varietiesare0207and LJ982.3. Each method has its advantages and disadvantages and for the evaluation of varietiesyield stability,four methods are feasible.But for regional trials for tobacco,the aim is to selectthe best varieties of integrated traits.So the radar chart method is more suitable for regionaltest with processing data.