Habitat Changes Of Red-crowned Cranes Derived From Remote Sensing Data In The Yancheng Coastal Wetland During 1989-2019

The Yancheng Coastal Wetland,as a UNESCO(United Nations Educational,Scientific and Cultural Organization)Natural Heritage Site,is located in the central part of China's coastal zone.It is the largest intertidal nature reserve.Its unique geographical location and ecological environment make it a resting place for migratory birds and becomes the important wintering habitat for red-crowned cranes in the world.In recent decades,due to the natural factors such as climate change and invasion of spartina alterniflora and human activities like costal development,the structure and function of wetland ecosystem have changed significantly,resulting in a reduction of the area and quality of habitat.This thesis explored the wetland pattern changes and its impacts on habitat for red-crowned cranes.Identifying and restoring their key nodes are of great international significance for the ecological conservation,especially for the biodiversity conservation.Based on the multi-temporal Landsat TM and OLI satellite remote sensing data during 1989-2019,the pattern changes of the Yancheng coastal wetland in the past 30 years were analyzed.In order to explore the impact of wetland changes,the landscape connectivity of suitable habitats and important habitat patches were analyzed and the key nodes were identified.The coastal wetland is located in the transition zone of land and marine ecosystem,so it has high dynamics and spatial heterogeneity.Therefore,it is difficult to meet the needs of high precision in wetland monitoring and mapping by using low spatial resolution(30 meters)and single seasonal remote sensing data.In this study,the multi-spectral and multi-seasonal of Sentinel-2 data with a spatial resolution of 10-20 meters was used to extract the spectral features,index features and texture features.Then the optimal classification method was explored by comparing the experimental schemes with different feature combinations to achieve the high accurate classification.The main conclusions obtained from this study are as follows:(1)In the past 30 years,the natural wetlands in coastal wetlands have degraded rapidly.The area of natural wetlands decreased to 961.85 km~2,and the area of alkaline land and Suaeda salsa marsh,suitable habitat for red-crowned cranes,reduced to137.41 km~2.During 1989 to 2002,the coastal wetland had changed dramatically.The natural wetland was occupied by the artificial one,and the area of Suaeda salsa marsh and reed marsh had decreased by 88% and 68%,respectively.During 2002 to 2019,the natural wetland was stable,and the area of artificial wetland increased continuously through reclaiming tidal flats.The hydrological connectivity increased,and the center of reclamation migrated southward and toward to sea side.(2)The structural and functional connectivity of the suitable habitat showed a decreasing trend,and the habitat fragmentation was severe.The structural connectivity can be used to describe landscape heterogeneity.The calculated results show obvious change of habitat in two aspects: the large patches broke and small patches gradually disappeared,and connection of whole habitat has broken into small parts.During 1989 to 2002,the indexes of integral and probability connectivity were only 6.76% and 2.91% of those in 1989.From 2002 to 2019,the small habitat patches were surrounded by heterogeneous patches and finally disappeared.In past 30 years,the functional connectivity indexes,representing the potential habitat connectivity,are less than 0.01 and the habitat network has completely broken.The patch importance index and betweenness index showed that the key nodes of habitat restoration mainly distributed in the core area of heritage site and Tiaozini wetland.(3)At present,the core area is the key habitat node of red-crowned cranes.The weeds,Suaeda salsa and Spartina alterniflora distributed in the core area.Similarly,the wetland patterns of core area had changed dramatically in past 30 years.Some of natural wetlands were developed into aquaculture ponds and paddy fields.The Spartina alterniflora extended toward inland with the tidal motion and the Suaeda salsa wetland was occupied by artificial wetlands and Spartina alterniflora wetland.In 2014,the ecological programs,such as returning aquaculture ponds and paddy fields to wetland,were implemented in order to promote restoration of wetland eco-environment.The aquaculture ponds in the northern part and alkaline land in the southern part were restored to natural wetlands.Up to 2019,the area of Spartina alterniflora has reduced to 64.25% and the area of natural wetlands has increased to89.41% of total area in the core area.Change simulation of landscape pattern in core area shows that the area of Spartina alterniflora marsh is larger and its expansion is more disordered.In 2024,the area of Spartina alterniflora marsh will increase to13.30 km~2 and the area of Suaeda salsa marsh will decrease to 11.55 km~2.(4)The index,texture and temporal features of Sentinel-2 data can effectively improve the accuracy of coastal wetland classification.By the feature selection,top25 features as an optimal feature subset can improve the model accuracy.Based on the optimal seasonal characteristics,the coastal wetland classification accuracy is 87.55% and the overall Kappa coefficient is 0.856.Comparing with the classification result derived from single season data,the overall accuracy increased by 13.86%.The classification precision of land cover types has also been improved and they are closer to the actual distribution of wetland.(5)Currently,the area of natural wetland in the Yancheng coastal wetland is961.85 km~2 and the core area is 187.75 km~2.Due to the continuous expansion of Spartina alterniflora marsh,the suitable habitat area reduced to 43.81 km~2.In order to promote the sustainable development of UNESCO Yancheng coastal wetland heritage site,the reeds should be selectively harvested in autumn and winter.Meanwhile,building the Tiaozini coastal nature reserve can also provide more larger resting space for red-crowned cranes and other migratory birds to accelerate the formation of coastal habitat connection network.