Modelagem do desenvolvimento em cultivares crioulas e melhoradas de milho

Abstract

Maize has high economic and social importance, because it is both produced on a large scale as well an agricultural commodity, as well as in small family farms for subsistence. The objectives of this dissertation were (i) to compare the simulation of tip leaf number with the CSM-CERES-Maize model (linear) and the model of Wang and Engel (non-linear), and (ii) to calibrate and to test the of Wang and Engel model for simulating the phenology of landraces and improved maize varieties. The cultivars used were two landrace varieties 'Cinquentinha' and 'Bico de ouro' and two improved cultivars 'BRS Planalto' and 'AS 1573PRO'. Field experiments were conducted in the 2013/2014 and 2014/2015 growing seasons at the experimental area of the Department of Plant Science of UFSM, Santa Maria, RS. The experimental design was a complete randomized block with four replications. Models calibration to simulate leaf appearance was with the data from the 11/04/2013 sowing date for cultivars 'Cinquentinha', 'Bico de ouro' and 'AS 1573PRO' and from the 12/13/2014 sowing date for 'BRS Planalto' whereas for calibrating the phenology data from the 12/13/2014 sowing date for all cultivars. The best estimate of the tip leaf number for both improved and landrace varieties of maize was with the Wang and Engel model with a root mean square error (RMSE) of 0.9 leaves compared to CSM-CERES-Maize with RMSE of 1.5 leaves. The Wang and Engel model estimated with good precision all developmental stages of the landrace and improved varieties with a RMSE for emergency was 1.4 days, 4.5 days for silking and 4.3 days for physiological maturity.