The role of species traits and forest structure on spatial carbon dynamics of temperate forests

森林在全球碳循环中扮演着重要角色。在什么情况下，森林是碳源还是碳汇，人们已经争论了很久。为了在大尺度上确定森林碳储量及其动态，必须考虑地方尺度的异质性。在空间上不同分布的碳储量往往是树种在垂直和水平上构成森林结构的异质分布的结果。然而，到目前为止，树种及其特征对当地和区域森林碳储量和动态的作用尚未得到全面了解。在这里，我们将实地观察与森林建模相结合，以研究物种特征及其变异性如何构成异质森林。对于中国温带森林物种丰富的研究地点(25公顷和24公顷两大样地，4-0.5公顷的13块样地)，我们将描述森林生物量和生产力的空间格局及其与树种特征和森林结构的相关性。通过推广和应用包含新的特征分布方法的基于个体的森林林隙模型，我们将在大范围(例如25公顷)和局部尺度上分析森林碳动态(例如，总初级生产力和净初级生产力、净生态系统交换)，并探索特征分布如何影响森林碳储量和动态。在另一项分析中，我们将通过将我们的研究限制在包括激光雷达活动在内的冠层水平的树木上，在将森林建模与遥感相结合方面向前迈进一步。这背后的愿景是找到通过遥感获得的所需信息的最小集合，以便将森林模型应用于广泛的不同森林类型。

Forests play a major role in the global carbon cycle. There has been a long debate under which conditions forests are a source or sink of carbon. For determining forest carbon stocks and their dynamics at large scales, local-scale heterogeneities have to be considered. Spatially heterogeneous distributed carbon stocks are often a result of heterogeneous distribution of tree species structuring the forest vertically and horizontally. However, the role of tree species and their traits on local and regional carbon stocks and dynamics in forests is so far not comprehensively understood. Here, we combine field observations with forest modelling in order to examine how species traits and their variability structure heterogeneous forests. For study sites of species-rich temperate forests in China (two large plots of 25 and 24 ha, 13 plots of 4-0.5 ha) we will characterize spatial patterns of forest biomass and productivity and its correlation with species traits and forest structure. By extending and applying an individual-based forest gap model including a novel trait distribution approach, we will analyse forest carbon dynamics (e.g. gross and net primary production, net ecosystem exchange) at the large (e.g. 25 ha) and local scale, and explore how trait distributions impact forest carbon stocks and dynamics. In an additional analysis, we will take a step towards advances in combining forest modelling with remote sensing by restricting our studies to canopy-level trees including LiDAR campaigns. The vision behind this is to find a minimal set of required information derived by remote sensing to apply forest models to a broad range of different forest types.