[Objective] The genetic variations of ancient chestnut trees in Ming and Qing dynasties were studied based on molecular markers. [Methods] 33 accessions were collected from Bohai town at Huairou district of Beijing, and the genetic diversity and polymorphism were detected by using 13 fluorescent polymorphic simple sequence repeats (SSRs). [Results] Thirteen SSR primers amplified 61 polymorphic alleles and 81 amplified bands in the 33 ancient chestnut trees, the number of alleles per locus ranged from 2 to 11 with an average of 4.7, and the amplified bands of 13 pairs of primers were 2 to 16 with an average of 6.2. The mean expected heterozygosity (HE) and polymorphic information content (PIC) were 0.4528 and 0.4103, respectively. PRA 75, PRA 83, PRD 25, CsCAT7 and PRD 53 can distinguish all of the ancient chestnut trees. [Conclusion] The habitats of chestnut trees in Ming and Qing dynasties were identical, the gene flow and gene exchange existed in a small area, which may be closely related to the low genetic diversity level. The study provides references for genetic diversity protection and utilization of chestnut germplasm resource. And the amplified bands and peak figures of 33 chestnut trees could lay the foundation for germplasm identification of the chestnut trees in Ming and Qing dynasties.

目的 为了解决碳储量的不确定性受单一类型变量的影响，而忽略了残差变异导致模型不确定性的问题，研究不同类型变量对碳储量估测模型的不确定性。 方法 以香格里拉市高山松为研究对象，采用随机森林联立蒙特卡洛（RF-MC），基于不同变量组合建立回归模型，判断各模型的不确定性。 结果 表明：（1）光谱和样地数据直接参与RF-MC模型的估测精度及不确定性最差。（2）在上一步的基础上引入纹理特征的模型预测结果优于引入DEM。（3）同时引入DEM和纹理特征的RF-MC模型效果最优（R²=0.892，RMSE=5.539 t·hm⁻²，MAE = 4.319 t·hm⁻², rRMSE=18.7%）。模型拟合度提高了0.343，模型的不确定性降低了19.43%。 结论 基于多特征的RF-MC方法在碳储量估测中效果较好，不同类型的变量对碳储量估测精度及不确定性有一定的影响。