Exploring optimal farm resources management strategy for Quncho-teff (Eragrostis tef (Zucc.) Trotter) using AquaCrop model

•Unlike previous findings, teff Normalized Water Productivity (NWP) fits to C4 crops.•Teff yield and biomass responded well to high fertilizer and full irrigation.•Early sowing was found to be the best strategy for growing teff in study area.•Two irrigation applications also substantially improved the yield of late sown teff.•This study suggests regional fertilizer recommendation for teff need to be revised. Teff is a major staple food crop in Ethiopia. Moisture and soil fertility are the two major factors limiting teff yield. Studies were conducted across three sites in Ethiopa [Mekelle (MK) in 2012 and 2016, Ilala (IL) in 2012 and Debrezeit (DZ) in 2009 and 2010]. The objectives of these studies were (1) to assess the response of Quncho-teff to different fertilizer and irrigation levels; 2) to quantify irrigation water productivity (IWP), and (3) to collect data to calibrate and validate AquaCrop model for simulating yield and evaluate optimal irrigation and sowing date strategy for Quncho-teff at different locations in Ethiopia. The different fertilizer levels were: 1) 64kgN and 46kg P/ha (N2P2); 2); 32kgN and 23kg P/ha (N1P1); 3) 0kgN and 0kg P/ha (N0P0) and 4) 52kgN and 46kg P/ha (N3P3). The four irrigation treatments were: zero (rainfed), two, four and full irrigation applications. Findings showed that full irrigation in combination with high fertilizer (N2P2) could give better yield. However, during abnormal rainfall, spreading the available fertilizer at a rate of 32kgN and 23kg P/ha may be preferable to applying 64kgN and 46kg P/ha. This study also indicated that the regional fertilizer recommendations for teff need to be revised taking in to account the soil characteristics, climate and irrigation water availability. The AquaCrop model was able to simulate the observed canopy cover, soil water, biomass and yield of teff satisfactorily. Canopy cover was simulated with normalized root mean square error (NRMSE), index of agreement (I) and R2 of 7%, 0.5 and 0.8, respectively. Soil moisture during the season was simulated with NRMSE of 11.4–15.7%, I of 0.99 and R2 of 0.85–0.9. Simulated final aboveground biomass values were in close agreement with the measured (NRMSE, 7.8%, I, 0.89 and R2, 0.66). There was also good agreement between simulated and measured grain yield with NRMSE, I and R2 values of 10.9%, 0.93, 0.80, respectively. Scenario analysis indicated that early sowing was the best option to maximize teff yield with the least amount of irrigati