MAIZE (Zea mays L.) INTERVARIETAL CROSSES STABILITY FOR THE SEMIWARM REGION OF GUERRERO

Orography and agro-ecology of the state of Guerrero, Mexico, are very diverse and virtually no maize (Zea mays L.) improved cultivars and hybrids are planted in the agricultural areas of intermediate altitude (1200 to 1700 m), because the variability of climate and soil conditions throughout the area, makes it difficult to establish a genetic improvement program for maize in each ecologic niche. In these regions, the stability of 20 maize cultivars was examined in 16 environments, during five Summer-Autumn agricultural cycles from 2004 to 2008, with the purpose of identifying at least one stable intervarietal cross, consistent and with good yield potential in several environments in the semi-warm region. The information was generated in an experiment with complete random blocks, with three repetitions per location. In the combined analysis, significant differences were detected between the cultivar averages (p <= 0.05). With this information, the stability analysis suggested by Eberhart and Russell was carried out. The additive main effects with multiplicative interaction model (AMMI), was also applied. The 20 genotypes analyzed included a local control (criollo del agricultor), two groups of parents, and their possible crosses. The male parents adapted to the study region, VE-1 and VE-3, showed stability and consistency throughout the environments; a similar behavior was shown by intervarietal crosses VS-529xCIST, VS-529xVE-1 and SINT-3-HExVE-3, which were also outstanding in grain yield; VS-529xVE-3 showed a greater productive potential, good yield in unfavorable environments and consistency. Intervarietal crosses SINT-3-HExCIST, HEI-1xCIST, VS-521xVE-1, VS-521xVE-3, criollo del agricultor and the female introduced parents SINT-3-HE, VS-521 and HEI-1, showed a good response in favorable environments but were inconsistent, since they were sensitive to unfavorable conditions in the environments. The AMMI analysis allowed understanding the contribution by each one of the sources of variation in the partition of the total sum of squares: cultivars (9.6 %), environments (55.2 %), and cultivar per environment (17.6 %); it also explained better the complex response of cultivars, detecting the ideal, for each environment interaction. The two methods used registered similar results, allowing us to trust the recommendation of cultivars for their grain yield potential and stability in the environments explored.