The Relationship Between Landscape Change And Vegetation Productivity In Horqin Sandy Land

The landscape of Horqin sandy land changed dramatically under the effect ofnatural and human activities. Landscape change is one of the important aspects oflandscape ecology, and it can directly depict landscape change in study area and theinfluence of landscape change on vegetation productivity (a proxy, NDVI). Therefore,choosing Naiman banner in Horqin sandy land as the study area, by using “3S”technique combine with landscape ecology theories and methods, the landscape andthe NDVI change were analyzed, and also the driving force of the landscape changeand the NDVI were analyzed. In order to analysis the relationship of them, seventowns were chose by using random selection method. The aim of this paper is toprovide theoretical basis to the regional ecological and environmental improvementand sustainable development in the study area.Four Landsat TM satellite images of Horqin sandy land were preprocessed byusing Erdas9.2software, and then interpreted it using Arcgis9.3software to obtain thelandscape types database. The annual NDVI data were using maximum valuecomposite (MVC) to process the MODIS NDVI from June to September, and theaverage NDVI of different seasons were using average value composite method toprocess the MODIS NDVI in given seasons. There exists correlations betweendifferent landscape indexes, therefore typical landscape indexes at patch level andlandscape level were chosed, and these indexes were calculated by using Fragstats3.3software. In order to reflect the change of different landscape types, landscape dynamicdegree was used to reflect the change of a landscape type under a certain time scale,and GIS spatial overload method was used to process two scene landscape type data toobtain transformation matrix of different landscape types of2002-2005,2005-2008and2008-2013. The driving forces of landscape and NDVI change were analyzed fromnatural and social economic aspects, and also analyzed the relationship betweenlandscape change and NDVI by using random selection method.The main conclusions are as follows:1. The main landscape types remained the same and the cropland, fixed sand duneand forest land were still the main landscape types at the patch level from2002to2013.The area of cropland, fixed sand dune, residential area and gully increased, while other landscape types decreased. The fragmentation of cropland, wetland, residential areaand gully increased, and other types decreased, bare mountain and unused landunchanged. The shape of residential area, floodplain and wetland increased, other typesdecreased. The cropland became dispersed in the study area, while other types becameconcentrated distribution. The dynamic degree of fixed sand dune, gully, forest landand residential area were positive value, others were negative value. The dynamicdegree of residential area was the maximum, and medium coverage grassland was theminimum. The fragmentation of the landscape alleviated, landscape became dispersedand landscape diversity decreased.2. Different landscape types transform from one to another and the area of movingsand dune decreased, while the area of fixed sand dune increased, the area ofsemi-moving sand dune and semi-fixed sand dune has a fluctuation change from2002to2013. The area of cropland transformed to forest land, wetland, fixed sand dune andresidential area, while water, wetland, river and gully transformed to cropland, andgrassland transformed to cropland and forest land. The increased area of cropland wasmainly from occupying forest land, fixed sand dune and grassland.3. The NDVI of different landscape types had a similar change, and the differenceswere small in spring, while big in summer from seasonal variation aspect from2002to2013. The average NDVI of forest land, wetland, floodplain, residential area, meadow,medium coverage grassland and gully were high, while others were low. The totalNDVI of cropland was the highest, and account for48.22%-58.71%of total NDVI.The total NDVI of bare mountain and unused land was the lowest, and account for0.59%-0.83%of total NDVI. The total NDVI of different landscape types and its’ areahad a similar change, while the total NDVI of fixed sand dune and cropland and its’area changed differently.4. The climate of the study area was cold and wet from2000to2013, and thelocal government takes active action to protect local environment from20th century80years, and these efforts have important effect on the landscape and NDVI change. Therelationship between precipitation, evaporation and productivity don’t remarkable,while the relationship between average temperature and NDVI remarkable, and thecharacter of the climate benefit to the enhancement of the NDVI. The relationship between the main landscape type cropland and moving sand dune were analyzed byusing random selection method, and the results showed that the NDVI of cropland andmoving sand dune and landscape indexes (number of patches, landscape shape indexand Interspersion Juxtaposition Index) had different relation at different stages. Thecropland dispersed in the study area has a positive influence on the enhancement of thecropland NDVI, and the number of moving sand dune and it’s dispersed and shapecomplexity in the study area had positive influence on its NDVI.