Zhou Hang, Li Ao, Luo Xuequn, Wang Jiafeng, Xie Yihong, Lin Zhongping, Hua Donglai
School of Resources and Environmental Engineering, Mianyang Teachers' College, Mianyang, China.
School of Life Science and Technology, Minshan Biodiversity Monitoring and Analysis Laboratory of Giant Panda National Park, Mianyang Teachers' College, Mianyang, China.
Front Plant Sci. 2025 Mar 11;16:1528255. doi: 10.3389/fpls.2025.1528255. eCollection 2025.
Rupr. is distributed in regions such as China, Kyrgyzstan, and Tajikistan. Owing to the impacts of climate change, it is increasingly threatened by habitat fragmentation, resulting in a precipitous decline in its population. Currently listed as endangered on the Red List of Trees of Central Asia, this species is predominantly found in the Tianshan Mountains. Examining the influence of climate change on the geographical distribution pattern of is crucial for the management and conservation of its wild resources.
This study employed two models, maximum entropy (MaxEnt) and random forest (RF), combined with 116 distribution points of and 27 environmental factor variables, to investigate the environmental determinants of the distribution of and project its potential geographical distribution areas.
The MaxEnt model and the RF model determined the primary environmental factors influencing the potential distribution of . The MaxEnt model showed that the percentage of gravel volume in the lower soil layer and elevation are the most significant, while the RF model considered elevation and precipitation of the wettest quarter to be the most crucial. Both models unanimously asserted that elevation is the pivotal environmental element affecting the distribution of .The mean area under the curve (AUC) scores for the MaxEnt model and RF were 0.970 and 0.873, respectively, revealing that the MaxEnt model outperformed the RF model in predictive accuracy. Consequently, the present study employed the estimated geographical area for modeled by the MaxEnt model as a reference. Following the MaxEnt model's projected outcomes, is mainly located in territories such as the Tianshan Mountains, Ili River Basin, Lake Issyk-Kul, Turpan Basin, Irtysh River, Ulungur River, Bogda Mountains, Kazakh Hills, Lake Balkhash, Amu River, and the middle reaches of the Syr River.Within the MaxEnt model, the total suitable habitat area exhibits growth across all scenarios, with the exception of a decline observed during the 2041-2060 period under the SSP2-4.5 scenario. Remarkably, under the SSP58.5 scenario for the same timeframe, this area expands significantly by 42.7%. In contrast, the RF model demonstrated relatively minor fluctuations in the total suitable habitat area, with the highest recorded increase being 12.81%. This paper recommends establishing protected areas in the Tianshan Mountains, conducting long-term monitoring of its population dynamics, and enhancing international cooperation. In response to future climate change, climate refuges should be established and adaptive management implemented to ensure the survival and reproduction of .
Rupr.分布于中国、吉尔吉斯斯坦和塔吉克斯坦等地区。由于气候变化的影响，它越来越受到栖息地破碎化的威胁，导致其种群数量急剧下降。该物种目前被列入中亚树木红色名录中的濒危物种，主要分布在天山山脉。研究气候变化对Rupr.地理分布格局的影响对于其野生资源的管理和保护至关重要。
本研究采用最大熵模型（MaxEnt）和随机森林模型（RF），结合116个Rupr.分布点和27个环境因子变量，研究影响Rupr.分布的环境决定因素，并预测其潜在地理分布区域。
MaxEnt模型和RF模型确定了影响Rupr.潜在分布的主要环境因素。MaxEnt模型表明，下层土壤层中砾石体积百分比和海拔是最显著的因素，而RF模型认为海拔和最湿润季度的降水量是最关键的因素。两个模型一致认为海拔是影响Rupr.分布的关键环境要素。MaxEnt模型和RF模型的曲线下面积（AUC）平均得分分别为0.970和0.873，表明MaxEnt模型在预测准确性方面优于RF模型。因此，本研究以MaxEnt模型模拟的Rupr.估计地理面积为参考。根据MaxEnt模型的预测结果，Rupr.主要分布在天山山脉、伊犁河流域、伊塞克湖、吐鲁番盆地、额尔齐斯河、乌伦古河、博格达山、哈萨克丘陵、巴尔喀什湖、阿姆河和锡尔河中游等地区。在MaxEnt模型中，除了在SSP2-4.5情景下2041-206年期间观察到适宜栖息地总面积下降外，所有情景下适宜栖息地总面积均呈现增长。值得注意的是，在同一时间框架的SSP5-8.5情景下，该面积显著扩大了42.7%。相比之下，RF模型显示适宜栖息地总面积的波动相对较小，记录到的最高增幅为12.81%。本文建议在天山山脉建立保护区，对其种群动态进行长期监测，并加强国际合作。针对未来气候变化，应建立气候避难所并实施适应性管理，以确保Rupr.的生存和繁殖。
