基于MaxEnt模型的美味猕猴桃在中国气候适宜性分析

The Analysis of Climate Suitability and Regionalization of Actinidia deliciosa by Using MaxEnt Model in China

【目的】明确美味猕猴桃(Actinidia deliciosa)在中国的气候适宜性,为品种合理布局提供理论参考。【方法】应用21个与温度和降水相关的气候数据和275条美味猕猴桃地理分布记录,采用MaxEnt生态位模型预测美味猕猴桃在中国潜在适宜分布,运用ROC曲线对模拟准确性进行验证,利用刀切法及相关分析筛选主导气候因子并分析其与存在概率之间关系。【结果】MaxEnt模拟准确性极好,10次重复的AUC平均值为0.961;美味猕猴桃在中国的适生区为N22°～38°,E96°～122°,总面积336.79×10~4 km^2,显著高于美味猕猴桃的实际种植面积;高适生区位于秦岭—巴山一线、四川盆地东部、云贵高原东部、武陵山—巫山和武夷山脉;11月平均温度、最冷季度平均温度、12月最高温度、4月最高温度、2月平均温度和温度季节性变化标准差为限制美味猕猴桃地理分布的关键因子。【结论】研究结果可为美味猕猴桃种植区划提供理论参考。

[Purpose]In order to provide theoretical references for reasonable layout of the Actinidia deliciosa varieties of our country, the climate suitability and regionalization of A. deliciosa should be determined.[ Method] In this study, based on 21 environmental data and 275 distribution records, MaxEnt (the maximum entropy model) was used to simulate the potential distribution of A. deliciosa in China. Response curves which indicated the relationships between climatic variables and the predicted probability of presence of A. deliciosa were drew up by MaxEnt, dominant factors were chosen by Jackknife test, while receiver operating characteristic curve (ROC) was used to evaluate the simulation.[Results]The accuracy of prediction of A. deliciosa in the current period was found to be “excellent”(AUC=0.961) according to the evaluation criterion. Under the present climatic conditions, the potential suitable regions for A. deliciosa were mainly distributed from latitude N25° to N38°, and longitude from E96° to E122°, and occupied 336.79×10^4km^2, which were much larger than the A. deliciosa actual growing area. The highly suitable areas were in the Qinba Mountain area, east of Sichuan Bain, east of the Yunnan-Guizhou Plateau, Wuling Mountain-Wushan Mountian area, and the Wuyi Moutain area. The important environmental factors affecting the distribution of A. deliciosa were mean temperature in November (T11mean), mean temperature of coldest quarter (bio11), maximum temperature in December (T12max), maximum temperature in April (T4max), mean temperature in February (T2mean), and temperature seasonality (bio4).[Conclusion]The results may provide reference for A. deliciosa planting pattern and countermeasures to cope with climate change impact in China.