Establishment and validation of site specific fertilizer recommendation for increased barley (Hordeum spp.) yield, northern Ethiopia

Shimbahri Mesfin; Mitiku Haile; Girmay Gebresamuel; Amanuel Zenebe; Abera Gebre

doi:10.1016/j.heliyon.2021.e07758

Establishment and validation of site specific fertilizer recommendation for increased barley (Hordeum spp.) yield, northern Ethiopia

Heliyon. 2021 Aug 12;7(8):e07758. doi: 10.1016/j.heliyon.2021.e07758. eCollection 2021 Aug.

Authors

Shimbahri Mesfin^{1

2}, Mitiku Haile¹, Girmay Gebresamuel¹, Amanuel Zenebe^{1

2}, Abera Gebre³

Affiliations

¹ Land Resource Management and Environmental Protection, Mekelle University, Mekelle, Ethiopia.
² Institute of Climate and Society, Mekelle University, Mekelle, Ethiopia.
³ Soil Resource and Watershed Management, Aksum University, Shire, Ethiopia.

Abstract

Establishing model based balanced nutrient requirements for barley (Hordeum spp.) in the northern Ethiopia can solve the fertilizer recommendation problems and enhance crop yield. The Quantitative Evaluation of Fertility of Tropical Soils (QUEFTS) model was used to estimate balanced nitrogen (N), phosphorus (P) and potassium (K) requirements for barley production in Alaje, northern Ethiopia. The objectives were to (i) quantify soil N, P and K supply and recommend fertilizers using QUEFTS model; (ii) investigate response of QUEFTS fertilizer application on yield and nutrient uptake and (iii) validate QUEFTS model performance. The experiment had four treatments: (T1) model based fertilization; (T2) blended fertilization; (T3) farmers' fertilization practices and (T4) control/no fertilizer. Soil information of the experimental plots were analyzed and used as model input to estimate soil nutrient supplies and recommend fertilizer. Yield and agronomic data were recorded and nutrient uptake and use efficiencies were analyzed. Model performance and accuracy were also checked using root mean square error, coefficient of determination, index of agreement and percent bias. The result revealed that the N, P and K soil supply ratio in the field experimental plots were 9:1:6. The higher grain yield of 4747 kg ha^-1 was recorded in the QUEFTS based fertilization plots. Validation results indicated that there is a good correlation between the QUEFTS predicted and observed grain yields implying that the QUEFTS model can be a base for development of simple and cost-effective decision support tools for nutrient management and fertilizer recommendations. Thus, the model performance and prediction accuracy is promising and can help farmers to adjust fertilizer application rates based on crop requirements.

Keywords: Fertilizer recommendation; Model performance; Soil fertility; Validation.