USING GIS DATASETS TO CLASSIFY MAIZE-GROWING REGIONS IN MEXICO AND CENTRAL-AMERICA

The environments in which maize (Zea mays L.) germplasm originated and in which it is evaluated can substantially affect results from germpl asm evaluations, thus influencing where the germplasm will eventually be used. The adaptation classifications normally used (e.g., temperate , tropical, subtropical, and highland) are imprecise. Our objectives w ere to (i) apply multivariate statistical techniques to spatial GIS (g eographic information system) datasets of agroclimatic data to group s imilar maize-growing regions in Mexico and Central America and then (i i) use the groups to refine the mega environments developed by CIMMYT maize breeders to help manage their germplasm. Data for this region we re extracted from a GIS-compatible global spatial climate dataset. Var iables analyzed (based on long-term monthly averages) included mean ma ximum and minimum monthly air temperatures, absolute maximum and minim um air temperatures based on each year's monthly data, mean monthly te mperatures and precipitation, total precipitation, and mode of the ele vations in the grid. The best grouping of similar regions resulted whe n cluster analysis on 7 mo of growing season data (April through Octob er) was used to obtain 25 groups. The 25 groups were then classified i nto 10 maize ecologies corresponding to CIMMYT's mega environments. Th e ecologies included three lowland, three highland, two subtropical, a nd two transitional from subtropical to highland. The technique will b e an important aid in classifying and using northern Latin America's l arge quantity of diverse maize germplasm.