Research Yancheng Wetland Nature Reserve Landscape Pattern And Model Simulations Predict

Coastal wetland located in the transition zone connecting land and ocean system had abundant ecological resources. To achieve ecological resources sustainable utilization in coastal wetland, knowing and planning their landscape pattern was becoming very necessary. Landscape metrics were the main indicators to represent landscape pattern quantitatively. But there was not a one-to-one correspondence between landscape metrics and landscape pattern, because landscape pattern may be affected by the correlations among Landscape metrics. So the correlations study of landscape metrics become a primary task for most researchers. Simulating the changes of landscape pattern could effectively reveal the effects of human activities on the landscape pattern, and it was helpful to understand the relationship between landscape pattern and ecological processes.In this paper, landscape pattern change and its driving forces of the Yancheng National Natural Reserve were analyzed by using RS and GIS technology. Date collected from remote sensing images of TM in1992,1999and2008were analyzed to figure out landscape pattern change of Yancheng National Natural Reserve from Xinyang port to Xingang port coastal wetland during the16years.The results indicated that:between1992and2008, the dominant type of landscape matrix transformed from natural landscape whose area was occupied78.04%to artificial landscape including dry field and paddy field whose area was occupied61.81%in entire zone; natural landscape still dominated in core zone, although its area had a little change that decreased from86.51%to84.79%; natural landscape area changed from83.86%to32.61%in buffer zone and changed from66.60%to15.83%in experimental zone, the change of two zones whose natural landscape area was decreasing were agreement with the entire zone. In the landscape fragmentation, the patch density index in entire zone increased from27.90to46.15, landscape fragmentation was serious; the density index decreased from27.60to26.60and the average patch area increased in core zone, landscape fragmentation slightly increased but not significant; the change of buffer zone and experimental zone were agreement with entire zone.In the landscape diversity, the diversity index in entire zone increased from1.35to1.65, the landscape pattern become more complicated and the degree of heterogeneous increased; landscape diversity slightly increased but not significant in core zone; the diversity index which transferred from1.01to1.62 increased significantly in buffer zone; the landscape diversity index decreased from1.40to1.36and the degree of heterogeneous decreased in experimental zone. The landscape pattern in entire zone was developing into the equalization direction; the core zone and buffer zone were agreement with entire zone, but the experimental zone changed into simplification direction. In the driving force of the landscape pattern change, the natural factors were dominant in core zone and human factors were dominant in buffer and experimental zone. And the correlations among Landscape metrics may directly influenced the results of reclamation landscape pattern analysis.TM image as data resource covered1764km2over study area, which located in core area of Yancheng National Natural Reserve. GIS was used to generate remote sensing classification map of landscape type and then4types among32landscape metrics were selected for calculating landscape metrics and correlation coefficient, respectively. As the change of scale, the index had obvious scale effect except the diversity index. In addition to the landscape isolation, segmentation, largest patch index, most of the index had not obvious turning point scale. The reasons of the existence of the phenomenon was mainly changes in particle size change the complexity of patch boundary, segmentation and incorporates some plaques, changed the original landscape pattern characteristics. From different scales, the interrelationship was analyzed between each landscape metrics. The results demonstrated that (1) when scale increased from0.5km to14km, based on significant level the order of range effect of correlations among landscape metrics were area, perimeter and density metrics, contagion metrics, shape metrics and diversity metrics.(2) Based on the result of range effect derived from calculating the correlations among metrics of different landscape types, the area, perimeter and density metrics and others possessed significant correlation, the shape metrics and diversity metrics had a very low correlation.(3) An obvious quantitative change occurred when scale increased to3km×3km, and accompanying with scale became larger, the majority of metrics would be uncorrelated. This research provided an effective way for selecting landscape metrics from different scales.By the methods of quantitative analysis in landscape ecology and the simulation method with CA-Markov model, the scenarios simulation landscape pattern in2016and2014were analyzed. Firstly, Chi test was utilized to validate simulation accuracy of CA-Markov model. The results showed that the value of Chi test was far outweigh x20.05(36) and the kappa index was greater than0.75, both of which satisfied the stabilized requirement of CA-Markov. This indicated that CA-Markov model could be used to predict landscape pattern changes of Yancheng National Natural Reserve. Secondly, CA-Markov model was constructed based on transition rules determined by transition probability matrix, which was calculated by Markov model. Then CA-Markov model was applied to simulating prospective transition direction of landscape types of Yancheng Natural Reserve. After validating with current landscape types, we found the simulation accuracy reached79.66%. It means that the prediction results could reflect the dynamic changes of landscape pattern. Sea water, Spartina alterniflora and bared intertidal flats possessed relatively higher prediction accuracy (kappa>0.77), but to river and breeding the accuracy of both was not too high, which lowered the overall accuracy. Finally, the landscape types of Yancheng Nature Reserve were predicted during the period from2016to2024.The forecast results show that reed area tends to be stable, mudflat and spartina area are going to decrease, breeding including artificial water lake and breeding pond area will increase, patches fragmentation of landscape types trends to increase. The problem of patches fragmentation cannot be ignored.