大兴安岭湿地潜在分布与环境因子的相关性  

Correlations between potential distribution of wetlands in Great Hing’an Mountains and environmental variables

作　　者：周昊[1] 布仁仓[1] 胡远满[1] 闫红伟[2] 刘宏娟[1] 冷文芳[1] 许申来[3] 

机构地区：[1]中国科学院沈阳应用生态研究所,沈阳110016 [2]沈阳农业大学,沈阳110161 [3]中国科学院生态环境中心城市与区域国家重点实验室,北京100085

出　　处：《生态学杂志》2007年第10期1533-1541,共9页Chinese Journal of Ecology

基　　金：中国科学院知识创新工程项目(KSCX2-SW-133);国家自然科学基金资助项目(40671013)

摘　　要：为了研究大兴安岭湿地在现行气候下的潜在分布,以大兴安岭地区的沼泽、沼泽化草甸、草甸为研究对象,选择对湿地可能有影响的地貌因子(海拔、坡度、坡位、综合地形指数、照射坡向)、水文因子(到河流的距离、到湖泊的距离)和气候因子(年平均温度、年降水量),利用SPSS软件建立Logistic逐步回归模型,探讨沼泽、沼泽化草甸、草甸分布与9种环境因子的相关关系。结果表明:模型模拟的湿地分布在大兴安岭北部和南部都具有较高可信度,其相对特征曲线值(ROC)分别为71.70%和73.40%;同时,模型模拟的沼泽化草甸、草甸在大兴安岭南部的分布也都具有较高的可信度;但在大兴安岭北部,沼泽、沼泽化草甸、草甸的Logistic回归模型的相对特征曲线值则不高。由于其分布形式趋于随机,模型模拟结果也是可信的。相关性分析表明,在区域尺度上,大兴安岭南部的草甸分布对温度最敏感;北部的草甸分布及其它类型湿地的分布均对地形因子最敏感。Taking the marshes, marsh-meadows and meadows in Great Hing＇ an Mountains as test objects, and selecting landform factors （altitude, slope, slope position, compound topographic index and irradiation slope direction）, hydrological factors （distances to rivers and lakes）, and climate factors （annual average temperature and annual precipitation） as potential affecting factors, a stepwise Logistic regressive model was established by SPSS to probe into the correlations between these wetlands and environmental factors, aiming to predict the potential distribution of wetlands in this area under cunrent weather conditions. The results indicated that the wetland distribution in the north and south of Great Hing＇ an Mountains simulated by the model had a high reliability, with the ROC value being 71.70% and 73.40%, respectively. The simulated distribution of marsh-meadows and meadows in the south of Great Hing＇ an Mountains had a high reliability, but for the marshes, marsh-meadows and meadows in the north of Great Hing＇ an Mountains, Logistic model had a moderate reliability. Though the test wetlands had a random distribution, the simulated results were still reliable to some degree. Correlation analysis showed that at regional scale, temperature was the most important factor affecting the distribution of meadows in the south of Great Hing＇ an Mountains, while landform factors were the important ones affecting the potential distribution of meadows and other types of wetlands in the north Great Hing＇an Mountains.

关 键 词：大兴安岭 LOGISTIC回归模型 地形因子 水文因子 气候因子 

分 类 号：Q14[生物学—生态学]