The high-risk invasive apple snail Pomacea canaliculata has greatly threatened the agriculture, ecosystem integrity and public health. In order to provide scientific evidence for effective prevention and control of P. canaliculata, a most suitable ecological niche model was selected to predict the potential suitable distribution areas of P. canaliculata in China. Based on 377 reported occurrence points and 19 bioclimatic variables, four ecological niche models, MaxEnt, GARP, BIOCLIM, and DOMAIN, the potential geographic suitable distribution areas were predicted for the invasive snail. Then, the results of different models were analyzed and compared with two statistical criteria, the area under the Receiver Operating Characteristic curve (AUC) and Kappa value. The results showed that all of the four ecological niche models could simulate the snail's distributions very well. More specifically, the MaxEnt model outperformed the others in all aspects of predicting the snail's potential distribution (AUC=0.955±0.004, Kappa=0.845±0.017), followed by GARP and DOMAIN. Although BIOCLIM offered the lowest prediction accuracy, its AUC was 0.898±0.017 and its Kappa value was 0.771±0.025. Based on the MaxEnt model, the prediction results showed that the potential suitable distribution areas of P. canaliculata were mainly located in the south of 30° N in China, but there was some regions spreading over the north of 30° N. The potential areas accounted for 13.2% of the national land in area. Notably, Guangdong, Guangxi, Hunan, Chongqing, Zhejiang and the coastal areas of Fujian were potentially high-risk areas. In conclusion, this study would be an important reference for the prevention and control of the invasive apple snail P. canaliculata and it also would be an example of predicting the potential distribution of aquatic alien species on large scale.
高危性外来入侵种福寿螺严重危害我国的农业生产、生态系统完整性和人体健康.为制定有效的防控策略提供科学依据,本研究通过选取最适的生态位模型以预测福寿螺在我国的潜在适生区.结合福寿螺在我国的337条分布记录和年均温、年降水量等19个生物气候变量数据,本文采用MaxEnt、GARP、BIOCLIM和DOMAIN等4种生态位模型分别模拟预测了福寿螺在我国的潜在适生区,并利用受试者工作特征曲线(ROC)和Kappa统计量分析比较不同模型的预测效果.结果表明: 4种模型均能较好地模拟福寿螺在我国的分布,其中MaxEnt模型的模拟准确度最高(受试者工作特征曲线下的面积AUC=0.955±0.004,Kappa=0.845±0.017),其次是GARP和DOMAIN,准确度相对较小的是BIOCLIM,但其平均AUC也达0.898±0.017,平均Kappa值为0.771±0.025.MaxEnt模型的预测结果显示,福寿螺的潜在适生区主要分布在30° N以南地区,但其中也有部分地区地处30°N以北.适生区面积占国土面积的13.2%,广东、广西、湖南、重庆、浙江和福建沿海地区具有高度潜在入侵风险.本研究可以为福寿螺的科学防控提供参考,并且对大尺度上外来水生生物的适生区预测具有一定的借鉴意义.
Keywords: Pomacea canaliculata; adaptability; biological invasion; model evaluation; population dispersal; risk analysis.