Desempenho de genótipos de milho utilizados pela agricultura familiar no Território Zona Sul do RS: o modelo CERES-Maize no auxílio à tomada de decisões

Abstract

Maize has a significant socio-economic importance for family farmers in the Território Zona Sul in RS, as it may be characterized as a croft crop by the extent of the cultivated area in the territory, although low yields have been reported regarding this culture. Management strategies may be used to increase the yield. Modeling may be a very useful tool in this case, as it allows a simulation of various management scenarios of production factors in addition to enabling the understanding of the dynamics of processes involved in the system and also the isolation of some factor’s effects over the yield of the crop. The aim of this study was to analyze the performance and viability of different maize genotypes employed by family farmers in the Território Zona Sul in RS, using the CSM-CERES-Maize model as a simulation tool for management scenarios and for support in the decision -making. Thus, an experiment for the parameterization of specifics genetic coefficients for the genotypes Amarelão, AL 30, AG 5011 and AG 122, which was conducted in a farm in Pelotas, RS, under optimized management conditions. The experiment was designed in random blocks, with four replicates. Subsequently, data was collected from a farmer, in the city of Cangaçu, RS, in order to evaluate the model’s predictive capacity. The plots for data collection were randomly distributed with four replicates throughout the plantation. The quality of the simulations was assessed by using the statistics index “d” for agreement and the root of the mean square error (RMSE). The seasonal analysis tool was used to evaluate the inter-annual and seasonal effect of weather conditions over the performance of maize genotypes in a dryland regimen, in Cangançu, RS. The hybrids AG 5011 and AG 122 have presented the highest yield values in the experiment for the parameterization, as the lowest values have been found for the variety crioula Amarelão. RMSE values remained under 17.8%, and the agreement d index over 0.83, which indicates good simulation quality and high agreement among the simulated and observed data. In the experiment for the assessment the yields were adequately simulated by the model, RMSE values remained under 15.4% and d over 0.96. Taking into account a break in grain productivity of up to 10%, sowing windows are longer for the variety AL 30, extending from December 12th to February 13th and are shorter for the two hybrids, AG 5011 and AG 122, which extends from December 26th to February 6th. For silage yield, all genotypes may be sown between July 26th and December 19th with no chances of missing the deadline for harvest (June 21st). Energy per unit of weight of harvested biomass (EPUWHB) presented values higher than 0.91UFL kg -¹, which indicate good silage quality for all genotypes.