Forest biomass turnover time estimation in China based on spatially explicit root : shoot ratios

Abstract

AbstractAimSpatial patterns of root : shoot ratios (R:S) are critical to better estimate biomass turnover time (τ). Given the difficulty of measuring belowground biomass directly, we still lack an understanding of high‐resolution spatial patterns of R:S in China's forests. Here, we explored patterns of R:S and their association with stand age, tree height, elevation, land cover, and climatic and edaphic variables, and estimated τ for China's forests.LocationChina.Time period2004 and 2010.Major taxa studiedForests.MethodsWe used a machine learning algorithm trained with an R:S dataset of 1,907 observations covering five forest types to upscale the spatial distribution of R:S in China as a function of stand age, tree height, elevation, land cover, and climatic and edaphic variables. We calculated τ using the estimated R:S together with satellite‐based aboveground biomass and net primary productivity data at 1 km × 1 km resolution.ResultsThe estimated mean τ for total biomass in China's forests was years. The longest τ (τ > 20 years) was observed in central China, whereas southern China exhibited the shortest τ (τ < 10 years). Large differences in partial correlations between τ and climatic variables (mean annual precipitation, climatological water deficit, and mean annual temperature) were observed among forest types.Main conclusionsOur data‐driven gridded maps of estimated R:S and τ provide valuable baseline information for assessing model performance and improving future carbon‐cycle assessments.