Simulated weather variables effects on millet fertilized with phosphate rock in the Sahel

Abstract

The Sudano–Sahelian agroecological zone is characterized by low and variable rainfall regimes and P deficiency. The present study complements previous research efforts and the objective was (i) to use the Newhall Simulation Model (NSM) to characterize three ICRISAT research sites, and (ii) to use output of NSM to develop an empirical model to guide efficient use of rainfall and fertilizers. The results show that length of the periods that rainfall exceeded evapotranspiration was larger in Bengou than in Gobery and Sadore. Total positive moisture balance during the three growing seasons was 85.7 mm at Bengou and 19.7 mm at Sadore. The model explained 52% of the variability in millet yields based on curvilinear response to P fertilizer, standardized May–June (Rmj) rainfall, and the number of wet days in the year (BW3). Yields appear more sensitive to BW3 than to Rmj. Their respective elasticity coefficients (Ec) were 0.62 and 0.09. Assessment of the model using R2=0.76 and the D-index = 0.85 showed reasonable agreement between model estimation and actual field yields. The study demonstrates the application of simulation models as a cost-effective means in terms of time and funds to agronomic research.