Mehlich 3 as an indicator of grain nutrient concentration for five cereals in sub-Saharan Africa

Soil testing for available nutrients is an important tool to determine fertilizer rates, however many standard methods test the availability of a single nutrient only. In contrast, Mehlich 3 (M3) is a multi-element test for predicting crop yield responses to the addition of macro and micronutrients. However, the M3 test has rarely been validated against crop nutrient concentrations, which limits its application for dietary improvement studies in sub-Saharan Africa. The primary objective was to test how well the M3 nutrient concentrations corresponds to grain nutrient concentrations as an indicator of plant nutrient status and grain quality. A secondary objective was to compare the performance of the M3 test with other extraction tests. This study used 1096 paired soil and crop samples of five cereals: maize, rice, sorghum, teff and wheat, covering a broad range of soil types and soil properties in Ethiopia and Malawi (e.g., pH 4.5 - 8.8; Olsen P < 1 - 280 ppm). The samples were selected from a larger collection based on “high” or “low” grain nutrient concentrations in the crop, and the respective soil available nutrients were measured with M3 and other extraction tests: CaCl2 (P, K, Mg, Mn), Ca(NO3)2 (K and Mg), Olsen P, sequential extraction (S), and DTPA (Mn, Fe and Zn). The M3 concentrations followed the trend of the “high” and “low” grain concentrations in nearly all nutrients and crops, and this was statistically significant in teff and wheat for all nutrients. The results were best for macronutrients, and slightly less good for micronutrients, probably because the concentration of micronutrients in the selected soil samples was generally quite low. Compared to the other multi-element extractant (CaCl2), the M3 test corresponded better to grain concentrations of K and Mg, and equally well to Olsen P, sequential extraction (S), and DTPA predictions of P, S, Zn and Fe, respectively. M3 extracted much greater concentrations than the other tests, and this was more pronounced in alkaline soils. Given that the M3 test corresponded well to grain nutrient concentrations across a range of soils and crops in sub-Saharan Africa (SSA), we conclude that it can be considered a universal test for plant nutrients. We also proposed thresholds for M3 values, defining below optimum, optimum and above optimum soil fertility status. These results validate the use of the M3 test to assess soil fertility and develop fertilizer recommendations for improved produce quality to enhance diets in SSA.