Application of the AquaCrop model in decision support for optimization of nitrogen fertilizer and water productivity of soybeans

Modelling of the effects of water and different levels of nitrogen on evapotranspiration and water productivity of rainfed soybeans is very important in optimising resource use in the production of the crop. The objective of the study was to model evapotranspiration, soil water storage and water productivity of rainfed soybeans under different levels of nitrogen fertilizer by using the FAO AquaCrop model. Field experiments were conducted at the Teaching and Research Farms of Obafemi Awolowo University, Nigeria in the rainy seasons of 2015 and 2016. There were five levels of nitrogen, which are 00, 25, 50, 75 and 100% of the recommended nitrogen applications and two varieties of soybeans, which produced a 2 by 5 factorial experimental design. The data of the wetter year 2015, were used for calibration of the AquaCrop model because AquaCrop is water driven. The 2016 data were used for the validation. The AquaCrop model simulated canopy cover with R² and EF > 0.90, d-index ≥ 0.99. The model captured the trend of the soil water storage well, R² and EF ≥ 0.70. The AquaCrop model simulated soil water storage below wilting point for seasonal rainfall less than 600 mm, and it did not overestimate it. The model predicted evapotranspiration with R² and EF ≥ 0.89, d-index = 1.00. Above ground biomass was overestimated even though R² ≥ 0.98. Although, nitrogen stress reduced seed yield and water productivity, there was no under or over estimation of the seed yields. They were predicted with low error under the different levels of nitrogen fertilizer, R² ≥ 0.99, EF and d-index ≥ 0.99. The AquaCrop model is suitable for simulating canopy cover, soil water storage, evapotranspiration, and seed yield of rainfed soybeans with different levels of nitrogen fertilizer under temporal distribution of seasonal rainfall. Therefore, it can serve as a useful tool for smallholder farmers in predicting productivity of soybeans and optimising resource allocation, land and water use in the tropical farming systems.We recommend simulation of the effects of pest on biomass, seed yield and water productivity by subsequent versions of the AquaCrop model. In addition, incorporating an economic sub-unit in the model will enable users to make financial decisions.