托里桉在我国南部的潜在地理适生区预测

Predicting Potential Geographic Distribution of Corymbia torelliana in Southern China

本文通过收集托里桉的实地栽培点地理数据,运用最大熵软件(Maxent)和地理信息系统技术(ArcGIS),结合12个气候因子数据,预测划定了托里桉在我国南部的潜在地理适生分布区。结果表明:Maxent模拟托里桉的潜在地理分布准确性较高,模型预测训练子集和验证子集AUC值均大于0.87。托里桉最适宜分布区集中在广东和海南沿海、广西沿海和中部及福建南部沿海,面积依次为广东(8.63×10~4 km^2)、海南(2.65×10~4 km^2)、广西(2.22×10~4 km^2)、福建(1.18×10~4 km^2);适宜区集中在广西中南部、广东中北部、福建南部、江西中南部,面积依次为广西(10.17×10~4 km^2)、广东(6.15×10~4 km^2)、福建(2.52×10~4 km^2)、江西(1.43×10~4 km^2)。Maxent刀切法(Jackknife)分析结果表明:影响托里桉适生区分布的主导气候因子为年平均气温、≥10℃积温、极端低温、最冷月最低温、最冷月均温、≥10℃的天数。

The areas climatically suitable for Corymbia torelliana plantations in southern China were analyzed using maximum entropy modeling(Maxent)and ArcGIS spatial analysis technology in this study,by combining data on the current distribution of C.torelliana and related climatic factors in China.The results showed that the Maxent model is able to predict the potential distribution of C.torelliana with a high accuracy– the correspondence between the training data and test data was over 0.87.The majority of area highest suitable for this species is concentrated in the coastal regions of Guangdong(8.63×104 km2),Hainan(2.65×104 km2),the coastal and central regions of Guangxi(2.22×104 km2)and the southern coastal regions Fujian(1.18×104 km2).The area highly suitable for the species is concentrated in the central southern regions of Guangxi(10.17×104 km2),central northern regions of Guangdong(6.15×104 km2),southern areas of Fujian(2.52×104 km2)and central southern regions of Jiangxi(1.43×104 km2).Jackknife analysis of Maxent showed that the dominant climatic factors influencing the potential geographic distribution of C.torelliana are annual mean temperatures,accumulated temperature of 10oC or more,extreme lowest temperature,lowest temperature of the coldest month,average temperature of the coldest month and the number of days with temperature above 10℃