| As an important component of terrestrial ecosystems,the dynamics of carbon stock and carbon cycle in forests play an important role in regulating global climate change.Existing methods to estimate large-scale forest biomass,such as satellite remote sensing or forest inventory,are good at estimating aboveground component,but rely on forest R/S ratio(rootshoot ratio)for estimates of root biomass.It is therefore critical to understand spatial patterns of forest R/S ratio and their driving factors.Despite recent advances in understanding forest R/S variations at individual level,systematic analysis of stand-level forest R/S ratio across the globe and its driving factors remain elusive.Therefore,understanding keys questions of spatial patters of forest R/S ratio and how environmental and growth-related factors control it across the globe is important.This is useful for accurately estimating forest carbon stocks and calibrating carbon cycle models.Our work intends to address this important question.In view of this,based on the large dataset of forest biomass and forest structure observation data from873 sites around the world,this study analyzed the effects of biotic and abiotic factors on the variation of forest R/S ratio at the site scale,and compared the differences in the mean forest R/S ratio between different climatic zones,forest types,forest regeneration methods and soil textures.Based on this,the study created a model of global forest R/S ratio variation using multivariate stepwise regression,compared the mean differences in R/S ratio among countries,continents,and biomes,further combined multiple datasets(Global Forest Heights,Global Consensus Land Cover,Global Forest Loss,EDB-CWD,and latitude)to create quantitative and spatially explicit map of the spatial pattern and uncertainty of global forest R/S ratio,then to explore the causes of differences in the distribution of forest R/S ratio.The main conclusions are as follows:(1)At the site scale,the forest R/S ratio decrease significantly with increasing growth-related factors(stand age,DBH,canopy height and basal area),MAT and MAP,and increase significantly with increasing CWD.Forest R/S ratio differs significantly among climate zones,soil types,and regeneration methods.(2)Coniferous and broadleaf forests showed consistent trend in response to growth-related variables;R/S ratio decrease significantly with growth in naturally regenerating forests,but changes insignificantly in planted forests.In smaller forests,plants grown under low temperature(low precipitation or high-water deficit)generally had larger R/S ratio than plants grown under high temperature(high precipitation or low water deficit),but the statistical trend of this result is weak at larger growth stage.(3)Globally,canopy height,CWD,forest type,regeneration method,and latitude have the greatest influence on the site-scale R/S ratio,explaining 37% of the variation in R/S ratio.Among them,forest type,canopy height,and CWD are the three main explanatory variables,while regeneration method and latitude are the secondary explanatory variables.(4)Forest R/S ratio varies significantly between continents and countries.Among the different divisions,forests R/S ratio in the tropics are generally lower,but in areas with more pronounced seasonal characteristics of rainfall,R/S ratio are higher.R/S ratio are generally moderate in temperate and boreal regions,higher than in the humid tropics but lower than in regions with strong seasonal rainfall,regulated mainly by species composition and regeneration patterns.(5)Uncertainty in the spatial pattern of R/S ratio have a distinct latitudinal profile,with the tropics being significantly higher than the temperate and northern regions.The uncertainty of R/S ratio prediction are moderately low in the humid tropics and increase in areas with significant seasonal rainfall.The uncertainty of R/S ratio prediction is low in most of the temperate and boreal regions,but moderate in the boreal plantation forests. |
