The Studies On The Distribution Of Fujian Rice Regional Testing Sites And Its Statistical Analysis System

The result of crop regional testing determines the application value and the suitable growth areas of a variety. The number and the locations of the testing sites have important effects on the result. Too many sites would raise the cost, reduce the efficiency and bring about difficulty for management, while too few sites would make the result less representative and could not reveal the characteristics of varieties. The testing sites should have ecological environmental representation of the district and could reveal the adaptabilities of varieties to the district.The Fujian rice regional testing sites, which were selected basing on the administrative district and the traditional rice growth region and were changed location in different years and different rice styles, have less ecologically representation for the effect of the topography and altitude in each sites, less veracity and efficiency for too many sites, more error for the trial data analyzed by hand. So it is necessary to study and change the number and locations of the testing sites and build a statistical software in order to improve the veracity and efficiency of the testing.Using the character data of Shanyou 63 and the weather factor date in each test sites, this thesis has tidied up 14 analysis factors and divided 20 trial sites into 5 districts by means of Cluster analysis, Principal component analysis and Gray relativities degree analysis. The 5 districts is Fuzhou (Fuzhou, Fuan, Fuding 3 sites), Southeast of Min (Putian, Quanzhou, Tongan, Nanjing, Yunxiao 5 sites), Nanping (Nanping, Shaowu 2 sites), Southwest of Min (Zhangzhou, Longyan, Wuping 3 sites), Northwest of Min (Mingxi, Pucheng). It was suggested to dispose one main site and one assistant site in each district in order to cut down the numbers of the test sites to 10.This thesis has built a statistical analysis software with Delphi 6.0 basing Windows 9x for date analysis in each sites. The software manage to do the analysis of variance and analysis of stability. It should improve the management and the accuracy in regional testing.Some questions including the altitude of the main and the assistant site, the site of early rice, middle rice and regeneration rice were discussed in this thesis. There is no analysis of the qualities of the varieties and the other design model in the software because it must be amended after the innovation of the trials distribution in Fujian.