Topography and plant community structure contribute to spatial heterogeneity of soil respiration in a subtropical forest

摘要

Soil respiration is the largest carbon (C) flux from terrestrial ecosystems into the atmosphere. Accurate estimates of the magnitude and distribution of soil respiration are critically important to models of global C cycling and predictions of future climate change. One of the greatest challenges to accurate large-scale estimation of soil respiration is its great spatial heterogeneity at the site level. Our study explored how soil respiration varies in space and the drivers that lead to this variance in a natural subtropical evergreen broadleaf forest in Southern China. We conducted a two-year soil respiration measurement for 168 randomly selected sampling points in a 4 ha plot. We measured the spatial variance of soil respiration and tested its correlation to a variety of abiotic and bi-otic factors including topography, aboveground plant community structure, soil environmental factors, soil organic matter, and microbial community structure. We found that soil respiration was highly varied across the study plot, with a spatial variation coefficient (CV) of 32.75%. The structural equation modeling (SEM) analysis showed that elevation influenced tree species diversity, productivity, and soil water content, which in turn affected soil respiration via soil C content, clay content,fungal:bacterial ratio, annual litterfall, and fine root biomass. 31% of the total spatial variation of soil respiration was accounted for in the SEM, mostly by elevation, soil C content, annual litterfall biomass, tree species diversity as estimated by the Simpson's index, and soil water content, with standardized total effects of 0.31, -0.31,0.29,0.19, and -0.18, respectively. Our data demonstrated that soil respiration was highly spatially varied at the fine scale, and was primarily regulated by factors of topography and plant community structure. More studies investigating the spatial variation of soil respiration are therefore needed to better understand and assess terrestrial ecosystem C cycling.