Detection of schistosomiasis transmission risks in Yunnan Province based on ecological niche modeling.
Detection of schistosomiasis transmission risks in Yunnan Province based on ecological niche modeling.
复制标题
DOI：
10.12140/j.issn.1000-7423.2020.01.012
复制
发表时间：
2020-01-01
期刊：
影响因子：
--
通讯作者：
Li, S. Z.
中科院分区：
文献类型：
--
作者：
Hu, Xiao-kang;Hao, Yu-wan;Li, S. Z.
目的：基于生态位模型预测血吸虫病传播风险并识别高风险区域，为云南省制定监测和控制措施提供科学依据。方法：收集云南省18个流行县2004—2015年村级血吸虫病疫情数据和气候、地理、社会经济等13个环境变量，采用BIOCLIM、DOMAIN和MaxEnt模型对云南省血吸虫病传播风险进行预测，并以接受者操作特征曲线下面积评估预测的准确性。 （曲线下面积）。采用性能最佳的模型分析环境变量的重要性并预测云南省血吸虫病传播风险的分布。结果： 3个模型对云南省血吸虫病传播风险分布的预测均具有良好的性能，其中MaxEnt模型的预测精度最高（AUC，0.96±0.01），其次是DOMAIN（AUC，0.93±0.04）和BIOCLIM（AUC，0.88±0.01）（三者间P < 0.05）。 MaxEnt模型显示，年平均降水量是影响血吸虫病分布最显着的环境因素（贡献值1.52），其次是国内生产总值和人口密度（贡献值分别为1.06和1.03）。 MaxEnt模型预测，传播风险区占云南省面积的3.1%，主要分布在西北地区，其中中低风险区占2.7%，高风险区占0.4%。高风险地区主要分布在鹤庆县北部、洱源县东部、大理市中部、巍山县东北部、弥渡县北部。结论：基于MaxEnt模型预测血吸虫病传播风险分布是可行的。云南省血吸虫病传播风险依然存在，高风险地区分布呈现聚集性格局。因此，需要有针对性的监测和控制。
Objective: To predict the transmission risks of schistosomiasis based on ecological niche modeling and identify high-risk areas, in order to provide scientific evidence for the formulation of monitoring and control measures in Yunnan Province. Methods: Village-level schistosomiasis epidemic data and 13 environmental variables such as climatic, geographical and socioeconomic factors were collected from 18 endemic counties in Yunnan Province from 2004 to 2015. BIOCLIM, DOMAIN and MaxEnt models were used to predict the schistosomiasis transmission risks in Yunnan Province, and the accuracy of prediction was assessed with the receiver operating characteristic area under curve (AUC). The model with best performance was used to analyze the importance of environmental variables and predict the distributions of schistosomiasis transmission risks in Yunnan Province. Results: All the three models had good performance in predicting the distributions of schistosomiasis transmission risks in Yunnan Province, with the MaxEnt model having the highest prediction accuracy (AUC, 0.96 0.01), followed by DOMAIN (AUC, 0.93 0.04) and BIOCLIM (AUC, 0.88 0.01) (P < 0.05 among three). The MaxEnt model revealed the annual average precipitation as the most significant environmental factor influencing the distributions of schistosomiasis (contribution value, 1.52), followed by gross domestic product and population density (contribution values 1.06 and 1.03, respectively). As predicted by the MaxEnt model, the transmission risk area, which was located mainly in the northwest, accounted for 3.1% of the area of Yunnan Province, comprising 2.7% of middle- and low-risk areas and 0.4% of high-risk areas. The high-risk areas were mainly distributed in northern Heqing County, eastern Eryuan County, central Dali City, northeastern Weishan County and northern Midu County. Conclusion: It is feasible to predict distributions of schistosomiasis transmission risks based on the MaxEnt model. There still remain risks of schistosomiasis transmission in Yunnan Province, and the distributions of high-risk regions show a pattern of clustering. Therefore, targeted monitoring and control is needed.