Modeling wheat development considering different cardinal temperatures and methods for the temperature response function calculation

Abstract

The objective of this work was to improve the prediction of developmental stages of Brazilian wheat (Triticum aestivum) genotypes made using the WE-Streck model. Different combinations of cardinal temperatures and methods of calculating the temperature response function were evaluated. Data regarding the dates of emergence, terminal spikelet, anthesis, and physiological maturity of six Brazilian wheat genotypes sown on 11 dates during three years (2005, 2006, and 2007) at Santa Maria, RS, Brazil, were used to estimate the WE-Streck model coefficients and to evaluate the different combinations of cardinal temperatures and methods for calculating the temperature response function. For the genotypes BRS Louro, BRS 177, CEP 51, CEP 52, and Nova Era, the simulation of the development with the WE-Streck model is better with higher values of optimum and maximum cardinal temperatures compared to those originally used in the model. For the genotype BRS Tarumã, the cardinal temperatures from the original version of the WE-Streck model should be used. It is recommended that the minimum and maximum daily temperatures be used to calculate the temperature response function when the WE-Streck model is selected for simulating the development of Brazilian wheat genotypes.