Browsing by Author "Piepho, Hans-Peter"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Soil and farm management effects on yield and nutrient concentrations of food crops in East Africa
    (Science of the Total Environment, 2020) Fischer, Sahrah; Hilger, Thomas; Piepho, Hans-Peter; Jordan, Irmgard; Karungi, Jeninah; Towett, Erick; Shepherd, Keith; Cadisch, Georg
    Crops that grow on soils with higher fertility often have higher yields and higher tissue nutrient concentrations. Whether this is the case for all crops, and which soil and management factors, or combinations mostly affect yields and food nutrient concentrations however, is poorly understood. Here, the main aim was to evaluate effects of soil and management factors on crop yields and food nutrient concentrations in (i) grain, fruit and tuber crops, and (ii) between high and low soil fertility areas. Total elemental concentrations of Mg, P, S, K, Ca, Fe, Zn, Mn and Cu were measured using a portable X-Ray Fluorescence Spectrometer (pXRF) in maize grain (Zea mays; Teso South, Kenya: n = 31; Kapchorwa, Uganda n = 30), cassava tuber (Manihot esculenta; Teso South: n = 27), and matooke fruit (Musa acuminata; Kapchorwa, n=54). Soil properties measuredwere eCEC, total N and C, pH, texture, and total elemental content. Farmmanagement variables (fertilisation, distance to household, and crop diversity)were collected. Canonical Correspondence Analyses (CCA) with permutation rank tests identified driving factors of alterations in nutrient concentrations. Maize grain had higher correlations with soil factors (CCA N 80%), than cassava tuber (76%) or matooke fruit (39%). In contrast, corresponding correlations to management factors were much lower (8–39%). The main soil properties affecting food nutrients were organic matter and texture. Surprisingly, pH did not play an important role. A positive association of crop diversity with nutrient concentration and yield in lower fertility areas was observed. Considering, food nutrient composition, apart fromyield, as response variables in agronomic trials (e.g. fertilisation or soil improvement strategies), would contribute towards discounting the notion that crops growing on fertile soils always produce healthy and high quality foods