In recent years, there has been an increasing concern over soil erosion in the Mollisols region of northeastern China, which is one of the six water erosion regions in China. This region has become one of the most hazardous regions for potential erosion and desertification for unique physical geographical environment and drastic human activities. There is a close relationship between land use structure and soil loss. Land use changes can increase or reduce surface runoff and soil erosion. The relationship between landscape pattern and soil erosion has been rarely studied.
A total of 30 sub-watersheds were chosen from Shuangyang river basin that is located at the eroded Mollisols region of northeastern China for examining their landscape metrics and soil erosion modulus. Correlation between landscape metrics and erosion modulus was analyzed using SPSS11.5 software. The objective of this study was to provide scientific basis for further erosion controlling and landscape planning in this region. The remotely sensed image (ETM+, August 30, 2002) of the area was classified into landscape map. We calculated landscape metrics with FRAGSTATS3.3 and soil erosion modulus with Revised Universal Soil Loss Equation (RUSLE) for 30 sub-watersheds.
The results showed that the average soil erosion modulus was 6.58 in Shuangyang river basin. Erosion intensity is at low-grade according to the Standards for Gradation of Soil Erosion Intensity in China. But it is higher than that of soil loss tolerance in the Mollisols region, northeastern China. Correlation analysis showed a significant positive correlation between Percentage of Landscape (PLAND) of sloping tillage farmlands and soil erosion modulus; Contagion Index (CONTAG) was also positively related with soil erosion modulus. Significant negative correlation was found between erosion modulus and PLAND of shrub and grassland, as well as Simpson’s Diversity Index (SHDI). Multiple correlation coefficients between the 9 landscape metrics selected in this study and erosion modulus was 0.931, which was higher than single-factor coefficients between any landscape metrics and erosion modulus. There was a linear regression relationship between the 9 landscape metrics and erosion modulus.
Farmlands are main landscape types under erosion in the study area, where more soil and water conservation measures should be taken. It is necessary for soil loss control to increase forest and grassland patches and extend conservation tillage in farmlands.