Explaining bread wheat (Triticum aestivum) yield differences by soil properties and fertilizer rates in the highlands of Ethiopia

Ethiopia faces major food security challenges. In spite of a modest level of fertilizer use, the percentage of wheat that is imported is substantial. The Ethiopian government has invested in the fertilizer sector, thereby also moving away from di-ammonium phosphate (DAP) to multi-nutrient blends (NPSZnB). Wheat fertilizer experiments were established in seven locations (three replications) in the highlands that represented the wide range of soils found in this area. The crop was exposed to DAP, NPS and to five levels of NPSZnB (50-300 kg/ha). All treatments included 100 kg/ha urea. The average wheat grain yield at the experimental sites, when all fertilizer treatments were averaged, ranged from < 2 to > 7 tons/ha. Soil sampling revealed that organic carbon (28%), total nitrogen and pH, and on the negative side, Fe and Mn concentrations, were significant drivers of yield differences. Fertilizers alone (when averaged for all experimental sites) could only explain 8% of yield differences, proving the ineffectiveness of blanket fertilizer recommendations. Blend fertilizers including micronutrients (NPSZnB) performed slightly but not significantly better than NPS alone or DAP alone. However, since the NP contents in the blend are slightly below those in NPS and DAP (particularly for P), a slight positive effect of Zn or B can be observed. On the other hand, Zn concentration in soils did not correlate significantly to wheat yields. Hence, determining the added effects of Zn and B remains subject for further research. Maximum yield gains to fertilizer application can only be achieved when fertilizers and soil property differences are analyzed jointly. In that case, 79% of yield differences were explained. Grouping soils into 'recommendation windows' then helps to come up with relevant and cost-effective fertilizer strategies. A simple calculation comparing the cost of wheat import with the cost of fertilizers needed to reach the current wheat consumption level in Ethiopia shows that the latter is by far the most cheaper option, but in need of smooth functioning of the entire value chain.