Evaluating 50 years of time-series soil radiocarbon data: Towards routine calculation of robust C residence times
Baisden, W. T., Parfitt, R. L., Ross, C., Schipper, L. A., & Canessa, S. (2013). Evaluating 50 years of time-series soil radiocarbon data:Towards routine calculation of robust C residence times. Biogeochemistry, 112(1-3), 129-137.
Permanent Research Commons link: https://hdl.handle.net/10289/7397
In 1959, Athol Rafter began a substantial programme of monitoring the flow of ¹⁴C produced by atmospheric thermonuclear tests through New Zealand's atmosphere, biosphere and soil. By building on the original measurements through ongoing sampling, a database of over 500 soil radiocarbon measurements spanning 50 years has now been compiled. The datasets, including an 11-point time series, allow strong focus on the robust quantification of residence times ranging from years to decades. We describe key aspects of the dataset, including the ability to identify critical assumptions inherent in calculating soil C residence times. The 3 most critical assumptions relate to: (1) the proportion of old C ("fraction passive"), (2) the lag time between photosynthesis and C entering the modeled pool, and (3) changes in the rates of C input (i. e., steady state). We demonstrate the ability to compare residence times in contrasting sites, such Andisols and non-Andisols, and the ability to calculate residence times across a range of soil depths. We use ¹⁴C in a two-box model to quantify soil carbon turnover parameters in deforested dairy pastures under similar climate in the Tokomaru silt loam (non-Andisol) versus the Egmont black loam (Andisol), originally sampled in 1962, 1965 and 1969, and resampled again in 2008. The ¹⁴C-based residence times of the main soil C pool in surface soil (~8 cm) are ~9 years in the Tokomaru soils compared to ~17 years for the Egmont soils. This difference represents nearly a doubling of soil C residence time, and roughly explains the doubling of the soil C stock. Passive soil C comprises 15% of the soil C pool in Tokomaru soils versus 27% in Egmont soils. A similar difference in residence times is found in a second surface soil comparison between the Bruntwood soil (Andisol) and the Te Kowhai soil (non-Andisol) with residence times of 18 and 27 years, respectively. The comparisons support evidence that C dynamics do differ in Andisols versus non-Andisols, as a result of both the mineral allophane and Al complexation. Expanding our calculations beyond surface soil, we show that thickening the calculation depth by combining horizons allows robust residence times to be calculated at a range of depths. Overall, the large and systematically collected dataset demonstrate that soil C residence times of the main soil C pool can be routinely calculated using ¹⁴C wherever samples collected 10 or more years apart in New Zealand grassland soils are available, and presumably under similar circumstances in other soils worldwide.