摘要
基于生态位理论,结合气候变化背景下的适宜性预测需求,对湖北省黄连(Coptis chinensis)的潜在适生区进行了动态预测与分析。以最大熵(MaxEnt)模型为基础,融合地理信息系统(GIS)空间分析技术,引入环境变量进行多重共线性筛选(基于Pearson相关性分析)及模型参数优化(正则化乘数1.5,特征组合LPQH)。采用筛选后的11个生物气候变量与地形数据,结合k折交叉验证(k=10),预测当前及未来(SSP-245情景下)湖北省黄连潜在适生区分布格局。结果显示,海拔(贡献度52.8%)、年平均降水量(贡献度11.1%)和最冷月最低温(贡献度10.5%)为黄连分布的主导环境变量。当前高适生区面积为6643.29 km^(2),主要集中于恩施、十堰和神农架地区。未来至2061—2080年,高适生区呈“先增后稳”趋势(峰值7048.14 km^(2)),中适生区面积波动下降,适生区质心在30.92°N附近呈小幅东西向迁移。模型AUC值为0.911±0.027,表明预测精度较高。研究提出的优化方法与生态解释框架可为中药材种植适地性评估及气候变化背景下物种资源保护提供参考。
Based on niche theory and considering the need for suitability prediction under climate change,a dynamic prediction and analysis of the potential suitable habitats of Coptis chinensis in Hubei Province were conducted.Using the MaxEnt model combined with GIS spatial analysis,environmental variables were screened for multicollinearity(via Pearson correlation analysis)and optimized(regularization multiplier=1.5,feature combination=LPQH).Eleven bioclimatic variables and topographic data were selected,and k-fold cross-validation(k=10)was applied to predict the current and future(under SSP-245 scenario)distribution patterns of Coptis chinensis.Results indicate that elevation(contribution rate:52.8%),annual precipitation(contribution rate:11.1%),and minimum temperature of the coldest month(contribution rate:10.5%)are the dominant environmental factors influencing its distribution.The current high-suitability area spans 6643.29 km^(2),primarily concentrated in Enshi,Shiyan,and Shennongjia.By 2061—2080,the high-suitability area shows an“initial increase followed by stabilization”trend(peak:7048.14 km^(2)),while the medium-suitability area fluctuates with a declining trend.The centroid of suitable habitats shifts slightly along an east-west axis near 30.92°N.The model’s AUC value(0.911±0.027)confirms high predictive accuracy.The proposed optimization method and ecological interpretation framework provide insights for assessing the adaptability of traditional Chinese medicinal plants and conserving species resources under climate change.
作者
刘少华
张祯志
陈小慧
LIU Shaohua;ZHANG Zhenzhi;CHEN Xiaohui(College of Geosciences,Yangtze University,Wuhan 430100,Hubei;College of Agriculture,Yangtze University,Jingzhou 434025,Hubei)
出处
《长江大学学报(自然科学版)》
2025年第5期123-130,共8页
Journal of Yangtze University(Natural Science Edition)
基金
国家自然科学基金项目“基于沉积模式的辫状河储层构型建模方法”(41872129)。
