Abstract:

Objective: A quantitative acknowledge on the spatial variation and scale effect of soil organic carbon (SOC) content provides basis to accurately and conveniently estimate the average SOC content of the whole slope, comprehensively understand the status of soil resources and the ecosystem carbon cycle, and propose high-quality soil management programs. Method: Three adjacent typical slopes formed by the project of returning farmland to forestland were selected in the small watershed of Zhongzhuang within the semi-arid loess hilly region of Ningxia. After multiple sample points were set up in succession from the top to the foot on slopes, the land use, vegetation characteristics and site conditions were investigated, the surface soil samples (0-20 cm) were collected to test the SOC content, and the slope aspect difference and slope position variation were analyzed. Taking the horizontal distance or relative distance from slope top as the independent variable, the slope scale effect was quantitatively described by the slope moving average of the surface SOC content as the dependent variable, and the ratio of the surface SOC content of any point on the slope to the average value of the slope as the dependent variable to realize the scale from ‘point’ to ‘slope’. Result: The surface SOC content had obvious slope difference, position variation and scale effect in the study region. The average surface SOC content was the highest on the southern slope (7.60 g·kg^?1), followed by the eastern slope (6.42 g·kg^?1) and the western slope (5.65 g·kg^?1). However, the variation range of surface SOC content of the eastern slope (15.95 g·kg^?1) was the largest, followed by the western slope (11.34 g·kg^?1), and the smallest was the southern slope (9.72 g·kg^?1), indicating that the slope effect was the strongest on the eastern slope, followed by the western slope and the southern slope. The pattern of position variation in surface SOC content was roughly the same among the three slopes, which gradually decreased from the slope top, and tended to be stable after the horizontal distance from slope top reached 200, 150 and 280 m (relative distance were 0.73, 0.45 and 0.76), respectively. It was mainly due to the spatial distribution pattern of land uses (Upper part: natural slopes; Lower part: terraced fields), vegetation types (Upper part: forest and grass; Lower part: crops) and vegetation restoration years (Upper part: long time; Lower part: short time). For every 100 m increase in the horizontal distance from slope top, the moving averages of the eastern, western, and southern slopes increased ?3.40, ?2.50, and ?1.51 g·kg^?1, respectively. For every 0.1 increase in the relative distance from slope top, the moving averages increased ?0.96, ?0.75 and ?0.55 g·kg^?1, respectively. The quantitative relationships between the ratio of surface SOC content at different slope positions to the slope average in three slopes was well constructed with the increase of horizontal distance or relative distance from slope top (R²>0.7, P<0.001), then the value of slope average can be accurately and conveniently estimated from the data of any point on a slope. Besides, the surface SOC content measured at a relative distance of 0.4 from slope top was the most similar with as the average value of slope. Conclusion: The surface SOC content decreases first and then stabilizes from the top to the bottom of slopes in semi-arid loess hilly region, which closely relates to the spatial distribution pattern of land uses, vegetation types and restoration years. The slope variation characteristics and scale effect of surface SOC content can be quantitatively described by taking the increase of (relative) horizontal slope length as the scale variable, so as to accurately and conveniently estimate the average value of surface SOC content on a slope.

Key words: loess hilly region, site condition, land use, soil organic carbon content, slope change, scale effect