The GrassSyn dataset: Soil organic carbon stocks in Brazilian grassy ecosystems

Although ecosystem management and restoration are known to enhance carbon storage, limited knowledge of ecosystem-specific soil organic carbon (SOC) stocks and processes hinders the development of climate-ready, biodiversity-focused policies. Baseline SOC stocks data for specific ecosystems is essential. This paper aims to: (i) examine SOC stock variability across major grassy ecosystems in Brazil and (ii) discuss data limitations and applications. We compiled the Grassland Synthesis Working Group dataset, which comprehensively aggregates SOC stocks data from published studies on main Brazil's grassy ecosystems. Our dataset results from systematic literature review and regional soil sampling datasets. The dataset provides spatially explicit SOC stocks, physical soil properties, and ancillary information from 182 studies (1996-2021) across 803 sites, spanning 35° latitude and 28° longitude. The dataset, structured in relational tables, reports soil C stocks and ancillary soil parameters at depths up to 100 cm. SOC stocks vary by grassy ecosystem types and sampling depth, with subtropical grasslands (Campos Gerais, South Brazilian highland grasslands, and Pampa) showing the highest SOC stocks across all depth layers (SOC 0-30 cm: 64.5-162.8 Mg C ha^-1; SOC 0-100 cm: 137.6-224.7 Mg C ha^-1). The tropical Cerrado and Amazon grassy ecosystems exhibit high SOC stocks, particularly in subsurface layers (SOC 0-30 cm: 53.6 and 38.3 Mg C ha^-1; SOC 0-100 cm: 109.8 and 121.4 Mg C ha^-1, respectively). Our data analysis shows high carbon stocks in natural/seminatural ecosystems, but some ecosystems are undersampled. The dataset on SOC stocks in grassy ecosystems could greatly aid Brazil's national greenhouse gas inventory.