Optimization of vegetation carbon content parameters and their application in carbon storage estimation in China

As a key parameter for C pool and flux assessments, vegetation carbon (C) content can be used in ecological models to predict climate-induced changes in the C sequestration capacity of vegetation. However, the differences in methods for upscaling C content from the organ to the community scale and their impact on regional C stock estimates have been ignored. Based on a comprehensive community structure survey of 72 typical natural ecosystems in China and 27,905 measured samples of plant organs (leaves, twigs, trunks, and roots), we first quantified the differences among scaling-up methods for vegetation C content. These methods included the community or dominant species-weighted mean, geometric mean, arithmetic mean, and traditional empirical coefficients (45 % and 50 %), and their impact on C storage estimation at the regional scale. Comparing the accuracy, variability, and response patterns of the different scaling-up methods, the dominant C species biomass-weighted mean (C_DWM) method had the highest similarity to the community-weighted C mean (C_CWM) method. Concerning vegetation C storage estimation in China's natural terrestrial ecosystems, the relative errors of the other methods ranged from -2.6 % to 8.22 % compared with that of the C_CWM method (18.39 Pg C). The empirical coefficients had the highest uncertainty, with a 45 % empirical coefficient underestimating the vegetation C stock by 2.60 %, and a 50 % empirical coefficient overestimating it by 8.22 %. The C_DWM method proposed here has high reliability for C storage estimation (overestimated by only 0.44 %), making it a preferable sampling and scaling-up method for regional C content and stock assessment. Additionally, our study provided the C content of plant organs for China's provinces and typical vegetation types based on the C_CWM, which could be used for regional C stock assessment and C cycle models.