AGROMETEOROLOGICAL MODELS TO ESTIMATE MAIZE YIELD AS A FUNCTION OF WATER AVAILABILITY IN THE STATE OF RIO GRANDE DO SUL

Abstract

Water deficit is the main factor responsible for reductions in maize production at Rio Grande do Sul State,Brazil. Therefore, quantifying the relationship between water availability and grain yield will give to us important informationsto help improve production in thc state. In this study, we looked at the relationships between grain yield and water availabilityusing a multiple regression models and Jensen model, which relates productivity to relative water use (ETr/ETm) at differentstages during the crop cycic. The data on grain yield and phenology were obtained from experiments conducted at Taquari,São Borja, Santo Augusto and Veranópolis experimental stations, during the periods from 1975/76 to 1989/90. The analisyswas performed for two maturity groups separately (early and normal), with a total of 89 data points for each of the groups.Based on the available data, both models were good estimators of com yield. Comparing the models results to the actual data,we notice a strong relationship between the relative yield observed and estimated with correlation coefficient varying from0.893 to 0.946.