SOIL-MANAGEMENT OPTIONS TO REDUCE RUNOFF AND EROSION ON A HARDSETTING ALFISOL IN THE SEMIARID TROPICS

Improved farming systems are needed to enhance productivity and reduce degradation on hardsetting Alfisols in the semi-arid tropics. A long-term experiment was started in July 1988 at ICRISAT, Hyderabad, India, to evaluate practices to improve infiltration and reduce erosion by stimulating biological activity and protecting soil in the rainy season. This paper outlines the rationale for the experiment, describes it, and summarises early results. Fifteen treatments in a randomised block design were applied to plots 28.5 m long by 5 m wide on a 2% slope. Nine annual treatments made up a factorial sub-set: tillage by tined implement to three depths (0, 10 and 20 cm) combined with three mulch treatments (no mulch, farmyard manure (FYM) at 15 t ha-1, and rice straw at 5 t ha-1). Pearl millet (Pennisetum glaucum) was sown in the factorial sub-set in July 1988. Six perennial species ley treatments (combinations of perennial pigeonpea (Cajanus cajan), verano stylo (Stylosanthes hamata), and buffel grass (Cenchrus ciliaris)) completed the randomised block design. All treatments were replicated three times making 45 plots in total. Plots were instrumented to measure runoff and soil loss. Millet straw yield was reduced by tillage to 20 cm and grain yield was significantly reduced by tillage to both 10 and 20 cm. Mulches had no effect on millet straw yield but FYM significantly increased millet grain yield. The proportion of rainfall running off plots ranged from 15.8 to 39. 1%. Perennial species treatments tended to have higher runoff than treatments under millet, but this was not consistent as zero tillage without mulch, and tillage to 10 cm either with mulch or with FYM mulch, also had relatively high runoff. Rice straw mulch significantly reduced runoff by comparison with FYM and no mulch. Soil was lost mainly as suspended load, indicating that raindrop detachment was the main erosion process. Effects of perennial species were variable and probably reflected the degree of surface protection provided in this establishment phase. Tillage significantly increased suspended load concentration in the first major runoff event but not in an event later in the season. Rice straw mulch significantly reduced bed load concentration in the first event and reduced suspended load in the later event. The ability of straw mulch to reduce runoff volume and sediment concentration suggests mulch-based systems may be able to contribute to sustainability of farming systems in the semiarid tropics. The lack of beneficial effects from tillage suggests a limited role for tillage-based systems. Results from subsequent years are needed to form definite conclusions and to show effects of slower biologically-induced changes in the soil.