This study aims to estimate the total biomass aboveground and soil carbon stocks in a Mediterranean riparian forest and identify the contribution of the different species and ecosystem compartments to the overall riparian carbon reservoir. We used a combined field and object-based image analysis (OBIA) approach, based on unmanned aerial vehicle (UAV) multispectral imagery, to assess C stock of three dominant riparian species. A linear discriminator was designed, based on a set of spectral variables previously selected in an optimal way, permitting the classification of the species corresponding to every object in the study area. This made it possible to estimate the area occupied by each species and its contribution to the tree aboveground biomass (AGB). Three uncertainty levels were considered, related to the trade-off between the number of unclassified and misclassified objects, leading to an error control associated with the estimated tree AGB. We found that riparian woodlands dominated by Acacia dealbata Link showed the highest average carbon stock per unit area (251 ± 90 tC ha⁻¹) followed by Alnus glutinosa (L.) Gaertner (162 ± 12 tC ha⁻¹) and by Salix salviifolia Brot. (73 ± 17 tC ha⁻¹), which are mainly related to the stem density, vegetation development and successional stage of the different stands. The woody tree compartment showed the highest inputs (79%), followed by the understory vegetation (12%) and lastly by the soil mineral layer (9%). Spectral vegetation indices developed to suppress saturation effects were consistently selected as important variables for species classification. The total tree AGB in the study area varies from 734 to 1053 tC according to the distinct levels of uncertainty. This study provided the foundations for the assessment of the riparian carbon sequestration and the economic value of the carbon stocks provided by similar Mediterranean riparian forests, a highly relevant ecosystem service for the regulation of climate change effects.

中文翻译：

---

地中海沿岸森林的碳储量估算：结合现场数据和无人机影像的案例研究

这项研究的目的是估计地中海沿岸森林地上总生物量和土壤碳储量，并确定不同物种和生态系统区隔对整个河岸碳库的贡献。我们基于无人飞行器（UAV）多光谱图像，使用了基于场和基于对象的图像分析（OBIA）相结合的方法来评估三种主要河岸物种的C储量。基于先前以最佳方式选择的一组光谱变量，设计了线性鉴别器，从而可以对与研究区域中每个对象相对应的物种进行分类。这样就可以估计每个物种所占的面积及其对树上地上生物量（AGB）的贡献。考虑了三个不确定性级别，与未分类和错误分类的对象数量之间的权衡有关，从而导致与估计的树AGB相关的错误控制。我们发现河岸林地主要由阿拉伯相思树（Acacia Dealbata Link）的单位面积平均碳储量最高（251±90 tC ha ^-1），其次为Alnus glutinosa（L.）Gaertner（162±12 tC ha ^-1）和Salix salviifolia Brot。（73±17 tC公顷^-1），主要与不同林分的茎密度，植被发育和演替阶段有关。木本林区的投入最大（79％），其次是林下植被（12％），最后是土壤矿物质层（9％）。为抑制饱和效应而开发的光谱植被指数一直被选作物种分类的重要变量。根据不确定性的不同程度，研究区域中的总树AGB从734到1053 tC不等。这项研究为评估河岸碳固存和类似地中海河岸森林提供的碳库的经济价值提供了基础，这是调节气候变化影响的高度相关的生态系统服务。