<b>Global change reshapes particulate and mineral-associated organic carbon fractions via passive carbon loss versus active carbon sequestration: A meta-analysis across terrestrial ecosystems</b>

<p dir="ltr">This dataset supports the meta-analysis entitled "Global change reshapes particulate and mineral-associated organic carbon fractions via passive carbon loss versus active carbon sequestration: A meta-analysis across terrestrial ecosystems" (Tan et al., 2026).</p><p dir="ltr">The dataset compiles 2076 paired observations from 423 peer-reviewed publications across global terrestrial ecosystems. It quantifies the responses of soil organic carbon (SOC) and its functionally distinct fractions, particulate organic carbon (POC) and mineral-associated organic carbon (MAOC), to six major global change drivers: land degradation (LD), land restoration (LR), nitrogen addition (eN), elevated carbon dioxide (eCO2), drought (D), and warming (W), including multi-factor interactions.</p><p dir="ltr">For each paired observation, the dataset includes:</p><ul><li>Treatment and control means, standard deviations, and sample sizes for SOC, POC, and MAOC</li><li>Derived ratio variables: POC:SOC, MAOC:SOC, and MAOC:POC</li><li>Land-use transition categories (e.g., natural forest to cropland, cropland to plantation forest)</li><li>Treatment intensity and experimental duration</li><li>Site metadata: geographic coordinates, ecosystem type, mean annual temperature (MAT), mean annual precipitation (MAP), initial soil properties (texture: sand, silt, clay; initial SOC, pH, C:N ratio, total nitrogen, dissolved organic carbon, base cations)</li><li>Moderator variables: aboveground and belowground biomass, microbial biomass (fungal, bacterial, total, MBC, MBN), and extracellular enzyme activities (BG, NAG, LAP, AP, PER, POX)</li></ul><p dir="ltr">Literature was searched in Web of Science and China National Knowledge Infrastructure (CNKI) up to August 2026.

POC and MAOC were unified across studies based on density fractionation (1.6 to 1.85 g/cm3) or physical size separation (50 to 63 um). </p><p dir="ltr">The dataset can be reused for systematic reviews, meta-regression, model parameterization, or validation of terrestrial biogeochemical models related to soil carbon dynamics under global change.</p>