腾格里沙漠东南缘4种灌木的生物量预测模型

灌木生物量模型是预测灌木生物量最有效的方法。选择腾格里沙漠南缘荒漠生态系统中常见的4种灌木（驼绒藜(Ceratoides latens)、盐爪爪(Kalidium foliatum)、珍珠猪毛菜(Salsola passerina)、红砂(Reaumuria soongarica)）为研究对象,以株高（H）和冠幅（C）的复合因子灌木体积（V）为自变量,通过回归分析,分别构建了4种灌木和混合物种的叶、新生枝、老龄枝、地上部分、地下部分和整株生物量的预测模型。通过决定系数（R2）、估计值的标准误（SEE）和回归检验显著水平（p＜0.05）筛选出了最优的生物量估测模型。结果显示:4种灌木的生物量模型主要以幂函数W=aVb为最优模型,少数以三次函数W=a+bV+cV2+dV3为最优模型。灌木生物量与V之间呈极显著的相关关系（p＜0.001）,决定系数较高,分别为:叶片（0.775＜R2＜0.866）,新生枝（0.694＜R2＜0.840）,老龄枝（0.819＜R2＜0.916）,地上部（0.832＜R2＜0.917）,地下部分（0.74＜R2＜0.808）,全株（0.811＜R2＜0.912）,说明预测模型可以应用于此4种灌木的生物量估算。不同物种之间及不同器官之间的生物量模型存在差异,在实际使用中,要根据物种来选择相应的模型。生物量模型的建立有助于全面估算荒漠生态系统的生物量,并进一步评估生态系统不同碳库的碳存储量与碳循环。为有效提高荒漠草地碳储量、合理实施生态系统管理和人为干预提供科学依据。

Biomass Estimation Models of Four Shrub Species at Southeastern Edge of the Tengger Desert

Biomass estimation modeling is one of the most important methods in biomass estimation of shrubs. In this paper, shrubs volume (V) was used as independent variable to establish the models between V and the leaf, current year branch, aging branch, above-ground, under-ground and total biomass for four single common shrubs and their mixed samples in desert area, respectively, at the southeastern edge of the Tengger Desert, China. The optimal models were selected according to the largest determination coefficient (R2), the smallest standard error of estimate(SEE) and significance level (p＜0.05). Results showed that the optimal biomass models for all organs, above-ground, under-ground and the whole plant of single species and mixed species were mostly expressed by power functions (W=aVb), only a few can be expressed by cubic equations (W=a+bV+cV2+dV3). The p values of all models reached a significant level (p＜0.05) and R2 ranged between 0.69-0.92. Those models can be used to estimate actual biomass of shrubs due to their good effectiveness. However, there existed differences in optimal models among different species or organs. Models should be chosen according to the practical conditions in estimating biomass of shrubs. The effects of mixed models are lower than that of single species. The optimal biomass models are expected to contribute to estimate biomass and carbon storage of different carbon pools for desert ecosystems.