Analytical graphics for multi-environment trials of breed selection

Graphic analysis is an efficient way of data mining for multi-environment trials (MET). In order to fully exploit useful information in trials, make variety evaluation, selection and prediction efficiently and accurately, more analytical graphics that are suitable for various traits, understandable and easy-operating, need be introduced into MET. In this paper, three kinds of analytical graphics-radar, candlestick, and conditional format, accompany with their application methods were put forward based on the national MET data of maize in China. The results showed that: the 3 analytical graphics were suitable for various traits of MET, understandable, and easy-operating. Radar chart was good at comprehensive analysis of multi-index, including reliability and consistency of certain variety or environment, selection and comparison of varieties. Candlestick chart did well in comparing varieties or environments in detail and pattern discovery of single index, such as discriminating ability of test environment, superiority and stability of a variety. Conditional format is a good combination of data viewing and visualization. Using the 3 kinds of graphics will improve the efficiency of analyzing and decision making of MET.