a b s t r a c t a r t i c l e i n f oSoil organic carbon (SOC) is one of the key components for assessing soil quality. Meanwhile, the changes in the stocks SOC may have large potential impact on global climate. It is increasingly important to estimate the SOC stock precisely and to investigate its variability. In this study, Yangjuangou watershed was selected to investigate the SOC distribution under different land uses. We found that SOC concentration decreased with increasing soil depth under all land uses and was significantly different across the vertical soil profile (P b 0.01). However, considering effect of land use on SOC, it is only significant (P b 0.01) in the topsoil (0-5 cm) layer. This indicated that land use has a large effect on the stocks of SOC in the surface soil. The stratification ratio of SOC N 1.2 may mean that soil quality is improving. The order of the SOC density (0-30 cm) under different land uses is forestland N orchard land N grassland N immature forestland N terraced cropland. The SOC stock is found to be as large as 2.67 × 10 3 t (0-30 cm) in this watershed. Considering time effect of restoration, the slope cropland just abandoned is more efficient for SOC accumulation than trees planted in the semi-arid hilly loess area.

Understanding of the relationships between vegetation and soil and topography would be very important for ecosystem restoration and management efforts in the dry valleys of Himalayan region but how to clarify the complicated relationships and figure out key factors for practical purpose is a challenge. The main objective of this research was to propose a four-staged procedure by combining several multivariate statistical techniques to detect the relationships between vegetation and soil and topography, and thereby identify the key factors for the degraded ecosystem restoration and management. Forty-three plots (5 m × 5 m) were selected for the field survey of the vegetation, soil and topography variables in the dry warm river valley of the upper Minjiang River, Sichuan Province, China. Cluster analysis (CA) demonstrated that high plant diversity, cover and height were associated with good soil quality and favorable topographic positions with lower solar incident radiation, runoff and soil erosion potential. Correlation analysis (simple correlation analysis and canonical correlation analysis) and multiple linear stepwise regression analysis affirmed that plant diversity was mainly correlated with soil water content, and soil water content was mainly determined by soil texture (clay content). Soil clay content alone could explain about 70% of the total variance. Identifying the favorable topographic position and the distribution pattern of soil texture and its controlling mechanisms is thus very important for restoration practices. In the process of ecosystem restoration, we should promote the co-evolution of vegetation and soil, and follow the natural succession sequence. Some relevant conservation polices are also needed to reduce human disturbance on ecosystem.

The small-scale effects of plant morphology in improving soil quality and reducing runoff and soil loss have remained unclear, especially in some arid environments with sparse vegetation. We selected three representative species with contrasting morphologies (Artemisia gmelinii; Ajania potaninii; Pulicaria chrysantha) to examine the effects of plant morphology on soil quality, runoff, and soil loss in the dry-warm river valley of the upper reach of Minjiang River, SW China. Runoff events were monitored from July through October 2006 using runoff plots on a micro scale (<40 · 40 cm 2 ) on a south-facing slope. The observation duration for rainfall and runoff events can be divided into two stages. Higher runoff depth, but lower soil loss per event occurred at the second stage as compared with the first stage due to the differences in rainfall, plant, and soil surface characteristics. The two herbs, A. gmelinii and P. chrysantha, had greater improvements on soil quality yielding high soil nutrient content and low soil compactness, while the effectiveness of the small shrub, A. potaninii, was minimal. Relative to bare surface (control treatment), the effectiveness of reducing runoff depth per event was 64.9, 66.6 and 38.0%, and reducing soil loss 65.5, 59.3 and 69.9% for A. gmelinii, A. potaninii, and P. chrysantha, respectively. All three plant species can improve soil quality and reduce runoff and soil loss, but their effects vary, which implies that plant morphology has to be considered while selecting species for ecosystem restoration.