基于GARP的大薸潜在适生区预测

GARP-based Prediction of Potential Distribution of Pistia stratiotes L. in China

大薸(Pistia stratiotes L.)原产热带和亚热带的小溪或淡水湖中,是入侵风险较高的水生植物之一。利用现有的大薸分布点数据,基于GARP模型采用刀切法选取最优环境因子集合,对其在我国的潜在分布区进行预测,并使用ROC曲线对其预测结果进行精度检验。结果表明,坡度、年平均气温、最冷月气温、多年平均降水量等对其适生区分布影响显著。现有数据表明,预防生物入侵往往比入侵后控制更为经济有效。该研究的目的就是预测大薸的适生区,从而寻找更经济有效的方法阻止它的入侵。

Pistia stratiotes L. originating in the stream or lake of the tropical and subtropical, it is one of the most harmful and widely distribu- ted invasive aquatic species in China. Using the GARP model as the method, based on the existing distribution point data of Pistia stratiotes L. in China and suitable environment layers selected by jackknifing to predict its potential distribution area. The study indicated that layers such as slope, annual average temperature, the coldest temperature and annual average precipitation have great influence on the results. The existing data show that it is more efficient to prevent the biological invasion than to try to control it after invasion. The aim is to predict the potential distribu- tion of Pistia stratiotes L. , and support government to find a more effective way to prevent the invasion.