Biased 14C-derived organic carbon turnover estimates following black carbon input to soil:: an exploration with RothC

The radiocarbon signature of the atmosphere is frequently used as a tracer to derive and validate turnover estimates of soil organic carbon (SOC). Such models often rely on steady-state assumptions and presume a direct correspondence between the atmospheric signature of CO(2) and that of residues entering the soil with or without a time-lag. The input of combustion products either from recent (i.e. charcoal) or fossil sources (e.g. diesel soot) violates this premise on account of its non-continuous delivery over time, the lack of (14)C in fossil black carbon (BC) and its high inertness compared to plant residues. In this study, possible effects of BC inputs (0.6-5 t BC ha(-1) of different (14)C age and supplied at different dates) on turnover calculations are discussed using hypothetical but realistic scenarios for two sites: cold grassland and temperate cropland. The carbon turnover model RothC 26.3 was used for the simulations. Turnover times may be over- or underestimated by up to 30%, depending on the underlying scenario. In the majority of cases, such effects are more pronounced and longer lasting at the colder site owing to its slower cycling. The models' inherent inert carbon pool improves the match of turnover times compared to runs where inert carbon is not accounted for. This is, however, achieved at the expense of mismatching the amount of BC when the latter is (14)C-young. The impact of BC on (14)C-based turnover calculations is disproportionate with respect to its amount, and an independent record of carbon fluxes and/or BC inputs and characteristics is suggested whenever BC is assumed to play a role in the ecosystem under investigation.