Assessment of nutrient management in major cereals: Yield prediction, energy-use efficiency and greenhouse gas emission

Abstract

Through improved nutrient management, sustainability of cereal production can be achieved by improving crop yield and economic indicators with reduced environmental footprints and energy use. The objectives of this study were to compare the crop yield and economics, and energy and environment related indicators as the measures of sustainability across three nutrient management practices in three major cereals of South Asia. Nutrient Expert (NE) based fertilizer recommendation, government recommendation (GR), and farmers' nutrient management practice (FP) were evaluated in 600 on-farm trials in two eastern Terai districts of Nepal during 2014–2016. The machine learning (Random Forest) approach was used to identify major indicators to predict grain yield variability. The NE-based recommendation had the highest and FP had the lowest grain yield and gross return in all cereals. Farmer's practice had the highest probability of increasing energy-use efficiency (EUE) followed by NE in all cereals. NE had the lowest yield-scaled greenhouse gas emission (greenhouse gas emission intensity; GHGI) in rice and intermediate in maize and wheat compared with GR or FP. Farmers' practice had consistently the highest GHGI in rice and maize while GR had the highest GHGI in wheat. Except for rice in FP, specific energy (SE) followed by EUE were the major predictors of grain yield of all crops in all nutrient management practices. These findings suggest an opportunity for increasing productivity and profitability with lowered GHGI of the major cereals with NE-based fertilizer recommendation and potential for its out-scaling in Nepal and the Eastern Indo-Gangetic Plains of South Asia.

Keywords

• CO2 emission
• Energy productivity
• Multi-criteria assessment
• Random forest
• Yield-scaled emission