| According to clustering way made by Erle C. Ellis, five cells were selected from an IKONOS high resolution image of Jintang county and Jianyang City, Sichuan province, as samples representative to study landscape structure and the soil nitrogen spatial distribution, also the soil organic matter done in this way. The results are obtained as following:1. The village landscapes of the hilly area of the Middle Sichuan Basin are complex. The land utilization is highly dominated by crop land with percentage of 71.61 over the whole landscapes, while the water area is 2.41% and the forest cover 11.22% which is as much as 67.98% of China average level and far lower than world average level.2. The village landscape diversity analysis indicates that the range of landscape diversity indexes vary more largely from 0.9273 to 1.4283 than that of land use and land cover. The evenness indexes are from 0.5176 to 0.7972 with land use evenness the highest values. The indexes of dominance vary from 0.3635 to 0.8644, and that of landform, land use and land cover vary from 0.4944 to 0.8449, from 0.3635 to 0.6985, and from 0.5892 to 0.8644 respectively. There are positive correlation between landscape evenness and diversity in land use and land cover.3. The results of fractal analysis indicate that the fractal dimension of paddy is from 1.1154 to 1.2555, relatively low, reflecting that the structure of paddy is not only affected by landform, but also by human activities greatly. The cropland, forest, hamlet, infrastructure and watercourse have higher fractal indexes (1.3-1.5) and low self-similarity. The landscape fragment indexes are naturally from 0.0240- 0.0424/100m2, relatively low.4. It is obtained from the modeling analysis of soil nitrogen and organic matter spatial distribution that the spatial variations of soil nitrogen and organic matter are fairly fitted with Spherical, Gaussian or Exponential model, and their spatial autocorrelation degrees are different. Total nitrogen mainly accord with Spherical and Gaussian model, the degree of spatial autocorrelation is weak, medium or strong; available nitrogen belongs to Sphericalmodel, having weak or strong degree spatial autocorrelation; organic matter fits to Spherical and Gaussian model mainly, weak or strong degree spatial autocorrelation.Some values and modeling parameters of soil nitrogen and organic matter have strong correlations with village landscape structure, which are followings:1) The landscape structure formed by different land use has significant effects on soil nitrogen and organic matter. There are significant or highly significant positive correlation between the diversity, evenness of land use and soil nitrogen and organic matter standard deviation, nugget and sill, also fragment index and range of available nitrogen and organic matter. Nitrogen and organic matter content and its variance, nugget and sill have significant or highly significant negative correlation with total area and average area of cropland, and significant or highly significant positive correlation with the isolation and fragment index of cropland, the same relationship existed between the total area and fragment index of paddy and available nitrogen content and its variance, nugget and sill. Construction and forest also has effect on soil nitrogen and organic matter spatial variability to a degree.2) The landscape structure of landform and land cover has effect on spatial distribution of soil nitrogen and organic matter to a degree. There is significant negative correlation between evenness of landform and organic matter range. The fragment index of landform, diversity index, evenness and fragment index of land cover are significant or highly significant positive correlating with available nitrogen and organic matter range.3) The correlation between landscape structure and spatial distribution of soil nitrogen and organic matter reflects the close relationship between village landscape structure and function. Optimizing landscape structure can improve nitrogen biogeochemical...
|
PDF Full Text Request