Soil organic carbon stock is closely related to aboveground vegetation properties in cold-temperate mountainous forests

The soils of cold-temperate mountainous forest ecosystems contain large carbon pools that may respond quickly to changes in climate and land use. A better understanding of the temporal–spatial distribution of soil organic carbon (SOC) stock is the first step toward quantifying its potential responses to global changes in climate or land use. The primary objective of this study was to understand the spatial distribution of SOC within a representative mountainous landscape in the cold-temperate zone in Japan. The two dominant tree species in this region are beech (Fagus crenata, a climax species in this zone), and birch (Betula ermanii). In 15 permanent plots located along an altitudinal gradient on the Mount Naeba and Mount Yakio, relationships were determined between SOC and climate (as represented by mean annual temperature), topography (aspect, slope, and altitude), aboveground vegetation properties (stand age, leaf area index, aboveground biomass, mean tree height, tree diameter, stand density, tree basal area, litter depth, and site index), and soil properties (soil bulk density, soil hardness, soil water content, soil root content). Our results indicated that SOC stocks in Japanese cold-temperate forest ecosystems are relatively high, ranging from 12 to 20 kg m− 2 in the top 30 cm of soils. Correlation analysis that included data from all 15 plots (and therefore data from both beech and birch forest and both mountains) suggested that the aboveground vegetation properties, such as aboveground biomass, mean tree height, and stand density, were strongly correlated with the SOC stock. In contrast, topographic factors (aspect and slope) explained SOC spatial distribution for the plots within the same mountain but not for plots from both mountains. Regression models based only on aboveground vegetation properties explained 76.7% of the variability in SOC stocks in beech ecosystems (represented by 10 plots) and 74.9% of the variability in all 15 plots. The strong correlation between aboveground vegetation properties and SOC suggests that measurement of vegetation properties by remote sensing could represent a feasible and rapid method for estimating SOC distribution in a rugged terrain.