基于MaxEnt模型预测未来气候变化情境下红树秋茄(Kandelia obovata)在中国潜在适生区的变化

Predicting potential suitable habitats of Kandelia obovata in China under future climatic scenarios based on MaxEnt model

红树秋茄为自然分布于我国南方热带与亚热带潮间带的木本植物群落,是淤泥质滩涂上特有的植被类型,由其构成的红树林生境在维持海岸生态系统平衡、湿地生物多样性等方面具有重要作用。全球气候变化影响了红树秋茄的地理分布范围,尤其是其分布北界的变化一直是当前红树研究领域的争论焦点之一。基于MaxEnt模型,筛选了141个秋茄现有分布点,并结合筛选后的14个陆地气候变量和2个海洋表层海洋环境因子,模拟了秋茄种群在不同碳排放情景下的潜在适生分布区域。分析了影响秋茄自然分布的主要环境因子,揭示了其在中国内地分布北界的演化趋势。根据模型结果,当前模拟的潜在分布区与现有分布范围吻合度较高,受试者工作特征曲线下面积(AUC值)为0.990,预测结果优良。通过广义加性模型回归验证表明,主要环境因子与秋茄的生境适生性指数呈显著相关,整体方差解释率为94.7%(R^(2)=0.915),说明MaxEnt模型预测结果可靠稳定。基于环境变量贡献率和刀切法的结果表明海洋表层平均水温、等温性、最暖季度降水量、年平均气温、最热月份最高温度等可能是影响秋茄分布的主要环境因子;等温性(bio_03)为23.43—33.99,最热月份最高温度(bio_05)>31.7℃,最暖季度降水量(bio_18)>740.61mm,海洋表层平均水温(temperature)>24.9℃,将有利于秋茄种群的发育与扩散。根据模型预测,在21世纪中叶和21世纪末,秋茄的潜在分布区发生集中变化的区域在浙江与江苏一带,以秋茄自然分布区域北界变化最为显著。不同代表性浓度路径(Representative Concentration Pathways, RCPs2.6、4.5、8.5)下,秋茄高适生区域均发生北移,低适生区域可至长江口附近,甚至到达江苏沿岸。这为上述区域未来利用红树秋茄进行生态系统修复和增加碳汇储备提供了理论依据。

Kandelia obovata is a rare woody plant that is naturally distributed in the silt mudflats of southern tropical and subtropical intertidal zones.K.obovata forms a mangrove habitat and plays a vital role in maintaining the balance of the coastal ecosystem and the biodiversity of the wetlands.Global climate change has affected the geographical distribution of K.obovata.The evolution of the northern boundary of the distribution of K.obovata has been a hot topic in mangrove research.Based on the MaxEnt model,we screened 141 existing K.obovata distribution sites.Combining 14 terrestrial climate variables and two marine surface environmental factors,we simulated the potentially suitable distribution area of K.obovata using different carbon emission scenarios.The main environmental factors affecting the natural distribution of K.obovata were analysed and revealed the evolutionary trend of its distribution along the northern boundary of China's Mainland.According to the model results,the simulated potential distribution coincided well with the existing distribution area,and the area under the receiver operating characteristic curve(AUC)value was 0.990.The regression validation of the generalised additive model showed that the main environmental variables were significantly correlated with the habitat suitability index of K.obovata,and the overall variance interpretation rate was 94.7%(R^(2)=0.915),indicating that the prediction results of the MaxEnt model were reliable.Among them,the average sea surface water temperature(temperature),annual mean temperature(bio_01),isothermality(bio_03),the maximum temperature of the warmest month(bio_05)and precipitation during the warmest quarter(bio_18)were the main environmental factors affecting the distribution of K.obovata.The environmental conditions were conducive to the development and diffusion of the population when bio 03 was 23.43—33.99,bio_05>31.7℃bio_18>740.61 mm and temperature>24.9℃.According to the model,the potential distribution area of K.obovata will change in Zhejiang and Jiangsu in the middle and end of this century,with the most significant changes in the northern boundary of the natural distribution area.Under different representative concentration pathways(RCPs2.6,4.5,8.5),the suitable area for K.obovata migrated northward.In contrast,a suitable area for K.obovata was found near the mouth of the Yangtze River and as far as the coast of Jiangsu Province,which provides a theoretical basis for ecosystem restoration and increases the carbon sink reserve by using the K.obovata mangroves in these areas.