Dynamic processes and the iterative effect of the finest roots on forest soil carbon accrual in the Northern Hemisphere

The dataset includes four parts: (1) decomposition records of each harvest from original observations and the fitting of decomposition rates and curves of the finest roots (i.e., absorptive roots) ("decomposition.origin.xlsx", "decomposition_ rate_fitting.R"); (2) the uncertainty analysis of standing biomass, turnover rate, and productivity of absorptive roots, as well as the iterative effects of absorptive root and leaf litters on soil carbon accural ("data.xlsx", "Data analysis.R"), which also includes the grouping of two mycorrhizal types (i.e., arbuscular mycorrhizal and ectomycorrhizal species) and in two biomes [ i.e., (sub)tropical, and temperate forests] ("data_group.xlsx", "Data analysis.R"); (3) the influence of biotic and abiotic factors on absorptive root dynamics ("factors.xlsx", "Data analysis.R"); (4) relationships between absorptive root traits and absorptive root processes ("trait_process.xlsx", "Data analysis.R"). This dataset provides biomass, turnover, and decomposition rates of absorptive roots across major forest types in the Northern Hemisphere. Original observations of absorptive root biomass, turnover, and decomposition rates were documented in "factors.xlsx", collected from publications via the ISI Web of Science, Google Scholar, and the China National Knowledge Infrastructure. The main codes of the study contain two parts: (1) decomposition curve optimized with the nonlinear least squares and the mean decomposition curve estimated by the maximum likelihood regression of absorptive root litters ("decomposition_rate_fitting.R") where two more scripts are used for curve fitting and plotting ("fun_scan_3par.R", "plot_func.R"). (2) statistics analysis of the datasets ("Data analysis.R") where one more script is used for the calculation (“analysis_func.R"). The users can run the two main R codes in the R environment of version 3.1.4. The results of decomposition curve fitting and samplings will be generated into two result folders: "result_decomposition" and "result_boot" separately. Please note that all the original data files and codes should be put in the same folder and the users need to set the workspace in the data folder first before running the codes.