Carbon control on terrestrial ecosystem function across contrasting site productivities: the carbon connection revisited.

Understanding how altered soil organic carbon (SOC) availability affects microbial communities and their function is imperative in predicting impacts of global change on soil carbon (C) storage and ecosystem function. However, the response of soil microbial communities and their function to depleted C availability in situ is unclear. We evaluated the role of soil C inputs in controlling microbial biomass, community composition, physiology, and function by (1) experimentally excluding plant C inputs in situ for 9 yr in four temperate forest ecosystems along a productivity gradient in Oregon, USA; and (2) integrating these findings with published data from similar C-exclusion studies into a global meta-analysis. Excluding plant C inputs for 9 yr resulted in a 13% decrease in SOC across the four Oregon sites and an overall shift in the microbial community composition, with a 45% decrease in the fungal : bacterial ratio and a 13% increase in Gram-positive : Gram-negative bacterial ratio. Although gross N mineralization decreased under C exclusion, decreases in gross N immobilization were greater, resulting in increased net N mineralization rates in all but the lowest-productivity site. Microbial biomass showed a variable response to C exclusion that was method dependent; however, we detected a 29% decrease in C-use efficiency across the sites, with greater declines occurring in less-productive sites. Although extracellular enzyme activity increased with C exclusion, C exclusion resulted in a 31% decrease in microbial respiration across all sites. Our meta-analyses of published data with similar C-exclusion treatments were largely consistent with our experimental results, showing decreased SOC, fungal : bacterial ratios, and microbial respiration, and increased Gram-positive : Gram-negative bacterial ratio following exclusion of C inputs to soil. Effect sizes of SOC and respiration correlated negatively with the duration of C exclusion; however, there were immediate effects of C exclusion on microbial community composition and biomass that were unaltered by duration of treatment. Our field-based experimental results and analyses demonstrate unequivocally the dominant control of C availability on soil microbial biomass, community composition, and function, and provide additional insight into the mechanisms for these effects in forest ecosystems