Constraining models for C Exchange in Permafrost and Peatland Soils: Soil radiocarbon and its utility for C turnover

Permafrost and peatland systems generally accumulate carbon (C ) upward as the carbon also turns over. This turnover-accumulation paradigm sets these systems apart from other soils partially because C pools and their average ages have a vertical rather than mixed structure. Past rates of C exchange in permafrost and peatland soils are more common than turnover models and while helpful, such rates are inherently biased toward events and periods in which there was a net positive exchange onto land and for periods or places in which C is preserved. For example we assume that slow rates of peat accumulation correspond to periods when net losses or smaller gains persisted. Preserved char material holds hope for indicating periods of net C loss via combustion, but periods of enhanced decomposition have few, if any, direct and datable indicators that link C loss to past climate events at spatial scales that are meaningful to soil-plant-atmosphere studies. Models greatly expand the opportunity for linking net C exchange to climate conditions of the past, but model testing by peat, macrofossil, or C data is limited conceptually and quantitatively by not addressing the entire soil C pool and its dynamic nature. We approach this problem with hypothesis testing. For hypothesis formulation, we turned to multi-year modern flux measurements to look for triggers of C loss or lower accumulation rates (via net ecosystem carbon balance (NECB) or net ecosystem production) in periods or places in which (1) water tables are more variable or are drawn down to aerate more peat, and (2) active layer thickness is deeper, resulting in greater ratio of thawed: frozen substrates. Keeping in mind that significant or persistent changes in seasonal factors could trigger (1) or (2), we tested for times or areas in which NECB was reduced. For hypothesis testing, we then used two approaches. The first approach compared total inventories of bomb-enriched Cs-137, unsupported Pb-210, and bomb enriched C-14 and used controls sites for establishing regional variations in fallout. The second approach uses Cs-137 and Pb-210 chronologies and then uses soil C-14 data over those chronologies to estimate turnover times of bulk soil C. Preliminarily, C turnover in soils was fastest in well drained landscapes with deepest active layers and was slowest in mature forest stands with thinner active layers. This result is consistent with hypotheses that water tables and active layers play a leading role in governing soil carbon fate in high-latitude regions.