Continuous cultivation of soils of the semiarid tropics has led to signi®cant land degradation. Soil erosion and nutrient loss caused by high runoff volumes have reduced crop yields and contributed to offsite damage. We compared a number of soil management practices (tillage, mulch and perennial/annual rotational based systems) for their potential to improve crop production and land resource protection in an Al®sol of the semiarid tropics of India. Runoff and soil erosion were monitored and surface soil and sediment were analysed for nitrogen and carbon to determine enrichment ratios. Amelioration of soils with organic additions (farmyard manure, rice straw) or rotating perennial pasture with annual crops increased soil carbon and nitrogen contents and reduced runoff, soil erosion and nutrient loss. Soil erosion totalled less than 7 t ha ±1 , but enrichment ratios were often greater than 2 resulting in up to 27 kg N ha ±1 and 178 kg C ha ±1 being lost in sediment. Up to an extra 250 mm of water per year in®ltrated the soil with organic additions and was available for crop water use or percolation to groundwater. The results show that there are good opportunities for reducing degradation and increasing productivity on farms.

No abstract

No abstract

Techniques to simulate effects of surface cover and tillage on runoff and erosion are described. Data for 15 soil management treatments on an Alfisol in the semi-arid tropics, India, were used to modify existing procedures of runoff prediction using USDA runoff curve numbers. A relationship between surface cover and curve number was developed to account for the effects of surface cover on runoff. Impact of shallow or deep tillage was predicted using functions that relate curve number to cumulative rainfall since tillage. The derived relationships were applied to adjust curve number due to the effects of cover and tillage on a daily basis and were incorporated into the cropping systems model called PERFECT-IND. Results of model validation showed that PERFECT-IND explained between 71 and 91% of the variation in daily runoff volumes. The model also provided accurate predictions of average annual runoff ranging from 33 to 217 mm for the 15 soil management treatments. Runoff was reduced to a much greater extent by surface cover compared with surface roughness. Surface cover reduced runoff curve number by a maximum of 35 units. The maximum reduction in curve number due to surface roughness was 5 units for shallow tillage and 10 units for deep tillage. Erosion predictions were acceptable but the lack of erosion data for all years in the experimental data limits the confidence in model output. Model calibration and validation have provided a set of parameters that can be coupled with historical climate records to provide a long-term perspective of the effects of soil management on runoff and erosion.

No abstract