Evaluate The Locations Of Maize Cultivar Trials (Even & Hill Group) In Sichuan Province

In this study, 9 statistics and combined anovas analysis were used to analyse the material of locations (even hill group) of maize regional trial in Sichuan province and also compared each statistical method in detail. Discrimination ability of locations, representation of locations and precision of locations are used to carry out quality synthetic evaluation of the locations of maize regional trial in Sichuan province. The results showed that:1. genetic change coefficient(GCV), regression coefficient(b), discrimination index (Y_m) and distance between location and zero point (D) based on additive main effects and multiplicative interactions (AMMI) model and sites regression (SREG) model, these five statistics are reflect the location distinguishes the cultivars ability. GCV with only a single location's yield to estimate the discrimination ability of the location, had not considered yearly material of maize regional trial as a whole, this made GCV lack of adequate information to estimate the discrimination ability of the location; Regression coefficient is to establish interaction with the additive effect of environmental index showed a linear relationship between the foundation, moreover the error variance component distribution is not independent, so the regression analysis is not a ideal model to estimate the discrimination ability of locations when the interactions between locations and cultivars are rich; Y_m is highly effected by cultivars comparison precision and can't explain more information about the interaction between locations and cultivars; the D value based on AMMI model merely considered the interaction between locations and cultivars which can not reflect the variation of the cultivars under the locations, this made the D value based on AMMI model to estimate the discrimination ability of the location is worth discussing; the D value based on SREG model simultaneously considers in the variation of the cultivars and the interaction between locations and cultivars, In addition, D value based on SREG considered yearly material of maize regional trial as a whole to study the discrimination ability of locations, so that can reflect the discrimination ability of locations accurately.2. Correlation coefficient(R), the angle (A_n) between the locations' first two component value and the average first two component value for all locations under the CCL bipplot based on SREG, Both R A_n can reflect the representation of locations and it was shown that coefficient of rank correlation between R and A_n were significant at the 0.01 level. In this study, R was used as a quantification index to evaluate the representation of locations.3. The experimental precision(CCV) and the cultivar comparison precision(RLSD) were consistent when used to analysis the precision of single location by yearly, each of them could be used to calculate the synthetic evaluation index.4. Select D value based on SREG model as the discrimination ability of locations, correlation coefficient(R) as the representation of locations and the experimental precision as the precision of locations to calculate the location quality synthetic evaluation index. The results showed that Neijiang and Yaan were listed as fifth level location in 2006, Chengdu were listed as fifth level location in 2007, Bazhong, Chengdu and Wusheng were listed as fifth level location in 2008, two years were listed as below fourth location were Emeishan, Bazhong and Chengdu, three years were listed as below fourth location was Yaan.