50年来长沙市湿地时空变化及空间自相关分析

Spatial-temporal Changes and Spatial Autocorrelation Analysis of Wetland in Changsha City in Recent 50 Years

为了更合理地规划长沙市湿地,利用空间统计分析的方法对长沙市湿地面积和分布规律进行时空2个维度分析。结果表明:(1)1955—2007年期间,人类活动在湿地的变化过程中起最关键的作用,长沙所有湿地围绕城区呈环状分布,湿地面积呈现了不断增加后减少的趋势;(2)水塘和沟渠的面积在1955—1990年间大幅增加,1990—2007年间大幅减少;湖库面积呈波动变化,1955—1972年间不断减少,1972—1990年间不断增加,1990—2007年间不断减少;河流面积变化小,较为稳定;(3)通过Moran’s I指数、Moran散点图分析可得:在这4个时期内,水塘的分布是集聚的,河流的分布是分散的,湖库和沟渠的分布是随机的,且1972年、1990年和2007年的水塘分布是高值聚集形成的空间自相关。其中,每个时期里大部分水塘的分布呈现正的局部空间自相关。总之,研究快速城市化地区湿地面积和分布的变化规律对城市湿地生态管理至关重要。在城市化水平提高背景下,城市人口和空间增大过程中,确定城市发展过程中湿地容易流失的位置,剖析湿地经常破坏的类型和原因,可以帮助城市湿地进行资源管理,有关部门在城市化发展之前做好规划,从而避免城市扩张建设中对重要湿地破坏造成的损失。

In order to provide valuable ecological benefits to urban ecosystems and properly design plan and use the wetland in Changsha,spatial statistics（geostatistics） was employed in analyzing the changes of size and distribution of wetlands during the 1955-2007.The results showed that：（1） from 1955 to 2007,human activities played a key role in spatial-temporal change of wetland of Changsha.All wetlands ring distributed around the city of Changsha.The wetland area was increased firstly and then decreased.（2） The area of ponds and ditches were increased significantly from 1955 to 1990,however,were substantially decreased from1990 to 2007.The area of lake and reservoir was fluctuated,it decreased ceaselessly from 1955 to 1972,and increased from 1972 to 1990,then decreased from1990 to 2007.The change of the area of river was small.River was more stable.（3） In the four period,according to Moran＇s I index,Moran scattered point diagram,the distribution of the ponds was concentrated,the distribution of the river was dispersed,the distribution the lakes and canals was random.And in 1970,1990 and 2007,the distribution of the ponds was high value formed by the aggregation of spatial autocorrelation.Furthermore,the distribution of a lot of ponds presented positive local spatial autocorrelation in each period.In summary,an examination of changes in wetland size and distribution was useful in wetland management,especially in urban and fast-growth regions.Locations where wetland losses were occurring most frequently could be identified,as well as the types of wetlands most frequently destroyed and the predominant causes of loss.This information,in conjunction with information on projected population increased,could be used to help resource managers and regulators anticipate where increased demands on wetlands were likely to occur.These areas could then be targeted for protection prior to urbanization and the consequent wetland destruction.