摘要
利用遥感、地理信息系统和全球定位系统技术,对黑龙江扎龙国家级自然保护区丹顶鹤巢址选择的"出现点"和"非出现点"初选变量进行空间自相关性检验,采用二项逻辑斯蒂克模型进行回归建模。结果表明:影响丹顶鹤巢址选择的自然环境因子包括芦苇高度、盖度和高程,干扰因子包括巢址距居民点距离、巢址周边堤坝和道路密度网。经计算预测值与观测值之间的匹配系数来验证模型准确性,总正确预测率为80.4%。将土地利用图与模型预测概率图进行叠加分析,得出扎龙湿地丹顶鹤适宜营巢栖息地的面积为223.46km2,主要分布在保护区的中部和北部。模型的预测结果基本反映扎龙湿地丹顶鹤适宜营巢栖息地的分布情况,为开展营巢地的有效保护与管理,并针对性地恢复重要营巢地提供科学依据。
A binomial logistic model was established based on RS, GIS and GPS techniques to simulate and predict the distribution of nest-site of Red-crowned Crane ( Grus japonensis ) in Zhalong National Nature Reserve ( ZLNNR), Hei!ongjiang Province, China. By using the data of present and absent points, a spatial auto-correlation test was conducted ~o screen the variables for the model. The result showed the factors affected the nest-site selection by Red-crowned Cranes included natural environment factors such as reed height and coverage, and elevation, and disturbance factors such as distance to residential area, density of dykes and roads around the nest. The model was tested through calculating the match coefficients of observed and predicted present/absent data. The correctly classified rate reached to 80.4%. Meanwhile, the model predicted that the suitable habitat for nesting by Red-crowned Crane reached to 223.46 km2 , which mainly distributed in the middle and north of ZLNNR. The model prediction basically reflected the suitable nesting habitat distribution, and might provide scientific basis for efficient conservation and management of the nesting habitats, and restoration of key nesting habitats for Red-crowned Crane in ZLNNR.
出处
《林业科学》
EI
CAS
CSCD
北大核心
2009年第7期76-83,共8页
Scientia Silvae Sinicae
基金
中国林业科学研究院森林生态环境与保护研究所所基金项目(CAFRIF200715)
关键词
丹顶鹤
巢址选择
3S技术
模型
扎龙湿地
Red-crowned Crane
nest-site selection
3S techniques
model
Zhalong wetland