Soil nitrogen availability as affected by fallow-maize systems on two soils in Kenya

Abstract

Simple methods for the measurement of nitrogen (N) availability are needed to assess the effect of low-input, organically based land management systems on the N supply of tropical soils. Our objectives were to determine the effect of contrasting land-use systems (LUS) on soil N availability and to identify measures of N availability that correlated with maize (Zea mays L.) grain yield. The LUS at the two sites in Kenya involved growth of a maize crop following 17 months of either: (1) Sesbania sesban (L.) Merr. tree growth (sesbania fallow), (2) natural regrowth of vegetation without cultivation (natural fallow), (3) three crops of unfertilized maize (maize monoculture), or (4) bare uncultivated soil (bare fallow). Soil was collected before the post-fallow maize crop was sown. The LUS had no effect on total soil N or amount of N in the heavy fraction soil organic matter (SOM) (>150 μm, >1.37 Mg m^–3). Sesbania and natural fallows, as compared to maize monoculture, increased the N in light fraction SOM (>150 μm, <1.13Mgm^–3), N in intermediate fraction SOM (>150 μm, 1.13 to 1.37 Mg m^–3), ammonium-N and aerobic N mineralization at a depth of 0–15 cm. Maize yields were highest following the sesbania fallow. Nitrate-N, inorganic-N (ammonium plus nitrate) and anaerobic N mineralization correlated with maize grain yield at both sites. The relationship between maize yield and pre-season nitrate-N improved when the depth of soil sampling was increased to 1 m at one site (an Alfisol), but did not improve at the site with anion adsorption in the subsoil (an Oxisol). The sesbania fallow was more effective than the natural fallow in increasing available soil N. Maize yield was better related to pre-season nitrate than N in size-density fractions of SOM.