A spatial-temporal diffusion study of maize variety in China using a new two- dimensional representation

The temporal diffusion patterns of crop varieties have been widely modeled; however, their spatial diffusion patterns remain less investigated due to a lack of available robust analytical tools. Based on the specific diffusion patterns of different varieties under unique time and spatial scales, we can more accurate inform decisions about variety extension and seed production. In this study, we proposed a simple method that converts three- dimensional (3D) dynamic data of planting area into two-dimensional (2D) static data to study the spatial-temporal diffusion of crop variety on the meso scale. Using this method, we studied the diffusion patterns of major maize varieties released between 1982 and 2008 in China. Contrasting the general displaying methods of 3D dynamic data, the 2D representation was more efficient and directly to capture the characteristics of spatial- temporal diffusion of corn varieties. The major corn varieties in China were divided into 5 classes or diffusion stages (A through E) according to their annual planting area. By looking at the variety diffusion pattern in a static figure, we found that both the average number of provinces covered and the average dissemination time of varieties varied along the variety's diffusion stages. Class E usually forms large diffusion centers and the diffusion centers migrate among dominant regions or provinces over time, whereas Class D and C usually form small diffusion centers and those diffusion centers rarely migrate.