Development and validation of crop yield model

Abstract

A 3-year on-farm trail at Samaru (11°11’N, 07°38’E) was conducted to develop and validate a new model that can predict maize grain yield from different cultivation practices. An empirical relationship was developed to predict grain yield using soil and management factors. The developed mathematical model was validated using field data. The model was evaluated using statistical indices. Modelling efficiency and root mean square error (RMSE) of 0.91 and 0.21 respectively were obtained. Results showed that tillage treatments and planting methods significantly  affected the soil shear strength, soil moisture content and maize grain yield. Deep tilled plots showed higher grain yield than shallow tilled plots.