不同恢复模式下红壤区森林植被盖度变化的阈值效应分析

植被恢复过程中乔木层与灌草层变化的阈值效应往往决定着生态恢复工程的成败,若忽视这一效应可能会引发其他生态问题。为了进一步促进我国森林生态系统植被恢复建设,填补相关研究空白,研究了典型红壤区自然和人工恢复模式下乔木层和灌草层盖度的动态变化特征,基于前沿的生态阈值评估框架,利用线性模型和三种阈值模型（包括分段回归模型、阶跃函数模型以及通用模型）评估了不同恢复模式下两种盖度指标之间的阈值效应。模型评估结果的后验分布形式表明：在人工恢复乔木层与灌草层盖度之间存在阈值效应,而在自然恢复模式下则不存在;通过比较各个模型评估结果的精度可知,分段回归模型结果更加准确;进一步利用分段回归模型得出,当前人工恢复样地中乔木层盖度的阈值为42.85%（95%的置信区间为41.57%~44.13%）,此时灌草盖度为58.75%,若超过乔木层盖度阈值会使得灌草层盖度开始降低。研究结果对促进我国红壤区人工林生态系统的结构改善具有较高的参考价值。

Threshold effect analysis of the change of forest vegetation coverage in red soil region under different recovery modes

The threshold effect of the change of canopy and shrub & grass cover during vegetation restoration often determines the success or failure of ecological restoration. If this effect is ignored, other ecological problems may arise. In order to further promote the restoration of forest ecosystems in China and fill in the relevant research gaps, we study the dynamic characteristics of vegetation coverage of tree layer and shrub & grass layer under natural and artificial restoration patterns in typical red soil regions. Based on the advanced ecological threshold assessment framework, we evaluated the threshold effects between the two coverage indicators for different recovery modes through the linear model and three threshold models (including piecewise regression models, step function models, and general models). The results show that, according to the form of posterior distribution of the ecological threshold, there is a threshold effect between the canopy cover and shrub & grass cover under artificial restoration, but it does not exist in the natural recovery mode. Compared with the accuracy of the evaluation results of each model, the results of the hockey stick (HS) model are more accurate. Using the HS model, the threshold of canopy cover in the current artificial restoration plot is 42.85% (95% confidence interval is 41.57%-44.13%), and the shrub & grass cover is 58.75%. If the threshold is crossed, the coverage of the shrub & grass will begin to decrease. Our results can provide an important scientific basis for the structural improvement of plantation ecosystem in red soil region of China.