Abstract

AbstractTerrace benches on the hillslopes of Western Uganda exhibit a crop‐yield gradient: upper portions of terraces produce less than the lower sections. We investigated the soil factors responsible for this yield variation on 30 terraces along 5 toposequences in what was predominantly a Ferralsol. Two levels of spatial analysis were conducted: (1) variation within individual terraces; and (2) differences across hill‐slope positions. A greenhouse experiment further examined this fertility gradient using soils extracted from the upper, middle and lower parts of the terrace. This included a nitrogen (N) treatment of 70 mg N kg−1 soil. In the fields sorghum (Sorghum bicolor) grain yield differed significantly (p ≤ 0·05) across terraces, increasing from 0·4 Mg ha−1 on the upper 40 per cent of terrace to 2·5 Mg ha−1 on the lower 40 per cent. Soil bulk density decreased down the terraces from 1·41 g cm−3 to 1·18 g cm−3, causing an increase of hydraulic conductivity from 1·6 cm h−1 to 7·3 cm h−1, from the upper to the lower part, respectively. Organic carbon (C) and total N increased from upper to lower terrace sections. In the greenhouse, sorghum growing on soils from which the soil physical limitations have been removed did not show significant yield differences across a terrace as observed in the field. Response to N was most pronounced on the upper terrace sections, increasing sorghum dry matter from 3 g pot−1 to 15 g pot−1. On the lower terrace, N amendment increased dry matter from 4 g pot−1 to 11 g pot−1. Technologies to improve the adverse soil physical conditions and the N limitations on upper terrace parts while preventing tillage‐ and soil‐erosion‐induced terrace scouring are required. Copyright © 2005 John Wiley & Sons, Ltd.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call