Effects of six years atmospheric CO2 enrichment on plant, soil, and soil microbial C of a calcareous grassland

Stimulated plant production and often even larger stimulation of photosynth esis at elevated CO2 raise the possibility of increased C storage in plants and soils. We analysed ecosystem C partitioning and soil C fluxes in calca reous grassland exposed to elevated CO2 for 6 years. At elevated CO2, C poo ls increased in plants (+23%) and surface litter (+24%), but were not alter ed in microbes and soil organic matter. Soils were fractionated into partic le size and density separates. The amount of low-density macroorganic C, an indicator of particulate soil C inputs from root litter, was not affected by elevated CO2. Incorporation of C fixed during the experiment (C-new) was tracked by C isotopic analysis of soil fractions which were labelled due t o C-13 depletion of the commercial CO2 used for atmospheric enrichment. Thi s data constrains estimates of C sequestration (absolute upper bound) and i ndicates where in soils potentially sequestered C is stored. C-new entered soils at an initial rate of 210 +/- 42 g C m(-2) year(-1), but only 554 +/- 39 g C-new m(-2) were recovered after 6 years due to the low mean residenc e time of 1.8 years. Previous process-oriented measurements did not indicat e increased plant-soil C fluxes at elevated CO2 in the same system (C-13 ki netics in soil microbes and fine roots after pulse labelling, and minirhizo tron observations). Overall experimental evidence suggests that C storage u nder elevated CO2 occurred only in rapidly turned-over fractions such as pl ants and detritus, and that potential extra soil C inputs were rapidly re-m ineralised. We argue that this inference does not conflict with the observe d increases in photosynthetic fixation at elevated CO2, because these are n ot good predictors of plant growth and soil C fluxes for allometric reasons . C sequestration in this natural system may also be lower than suggested b y plant biomass responses to elevated CO2 because C storage may be limited by stabilisation of C-new in slowly turned-over soil fractions (a prerequis ite for long-term storage) rather than by the magnitude of C inputs per se.