| Taking Litang alpine meadows as study area, a vegetation survey using quadrat and method of spatial changes in place of temporal changes and a soil sampling with laboratory analysis were carried out on the degraded grassland in this area. Sampling areas of light degradation, medium degradation, heavy degradation and control were selected to investigate vegetation structure, height, cover, abundance, species dominance and living above-ground biomass. Soil organic matter, total and available nutrients were analyzed through lab analysis. Also, discussions on the relationship between vegetation index derieved from a Landsat TM image and field recorded biomass were conducted. This study demonstrated the vegetation characteristics and soil properties of grasslands at different degradation successions; an index system for grassland degradation classification for alpine meadows in this area was concluded and models for biomass monitoring were obtained. The main results are as follows:(1) With the degradation increasing, the vegetation height, cover and above-ground biomass decreased. Comparing to control, height decreased 28%,62% and 72%, respectively and living above-ground biomass decreased 11%,55% and 64%, respectively. The species number of unpalatable plants and grazing-resistant plants increased; Percentage of unpalatable plants living above-ground biomass in total living above-ground biomass increased from 15.97% to 28.40%. Dominance of palatable herbages reduced.(2) There was significant difference among different degradations of soil 0-10cm organic matter content (p<0.05). Comparing to control, it decreased 6.9%,11.3% and 17.7%, respectively. At a whole scale, organic matter was low and the highest content was merely 53.48 g/kg; total nitrogen and total potassium varied inapparently while total phosphorus increased; available nitrogen, available phosphorus and available potassium varied inapparently.(3) Grassland degradation is widely distributed in this area and the medium degradation grassland distributed most widely which is 281318.6 hm2 occupying 55.89% of total grassland area. Grazing altered the vegetation and soil to a certain extent which presented a reverse succession.(4) An index system for grassland degradation classification for alpine meadows in this area was concluded according to the six vegetation and soil indices such as living above-ground biomass, cover, percentage of unpalatable plants and soil 0-10cm organic matter.(5) A maximum likelihood classification was applied to the optimal combination of bands 4-3-2 of a Landsat TM image. Results showed that the total classification accuracy was 87.7285% and the Kappa coefficient was 0.8386. Also grassland occupied 45.96% of the total area which is 5033.574km2 and is the largest. Forestland, occupying 21.55% of the total area, was 2359.821 km2. Wetland was 1526.711km2 and sandy land was 323.764 km2.(6) Using ENVI and ArcGIS computer software, seven vegetation indexes NDVI, RVI, DVI, SAVI, MSAVI, PVI and GVI were obtained. The results of the correlation analysis between vegetation indexes and field above-ground biomass demonstrated that those two variables were significantly correlated and RVI had the highest correlation coefficient 0.861. The optimum model for biomass monitoring:y=14.922x3-106.47x2+328.33x-150.2, which was calculated using linear and non-linear regression was based on vegetation index RVI. The accuracy of this model was 86% which was confirmed using quadrats biomass which weren't used to establish the model. And an estimation was conducted which resulted in the biomass distribution in this area that the high biomass is distributed in the northwest and southeast part while the low biomass is distributed in the northeast and southwest part.
|
PDF Full Text Request