Quantifying the contribution of soil organic matter turnover to forest soil respiration, using natural abundance δ13C

Abstract

Quantifying the loss of soil carbon through respiration has proved difficult, due to the challenge of measuring the losses associated with the turnover of soil organic matter (SOM) as distinct from autotrophic components. In forest ecosystems the δ13C value of respiration from turnover of SOM (δ13CRSOM) is typically 2–4‰ enriched compared with that from roots and associated microbes (δ13CRROOTS), with that from the litter (δ13CRLITTER) lying between the two. We measured soil respiration at 50 locations in a forest soil and then used differences in isotopic signatures to quantify the proportion of soil respiration arising from the turnover of SOM (fRSOM) at a subset of 30 locations, chosen randomly. The soil surface CO2 efflux was collected using an open chamber system supplied with CO2-free air and the δ13C signature (δ13CRS) measured, giving a mean (±SD) value across the site of −26.1 ± 0.58‰. The values of δ13CRROOTS, δ13CRLITTER and δ13CRSOM were measured at each location by incubation of roots, litter and root-free soil and collection of the CO2 for isotopic analysis. δ13CRSOM became progressively depleted with length of incubation (1.5‰ after 8 h), so CO2 was collected after 20 min. The mean value of δ13CRLITTER was −27.2 ± 0.68 ‰, which was indistinguishable from δ13CRROOTS of −27.6 ± 0.51‰, while δ13CRSOM was −25.1 ± 0.88‰. δ13CRROOTS and δ13CRSOM measured at each location were used as the end points of a two component mixing model to calculate fRSOM, giving a mean value for fRSOM of 0.61 ± 0.28. It was not possible to estimate fRSOM using the total C contents of the roots and soil which were significantly depleted in 13C in comparison to their respired CO2. However, at seven locations the δ13CRS was slightly enriched compared with δ13CRSOM (mean 0.3‰), which was not considered significantly different so fRSOM was constrained to 1.0. If these seven rings were excluded mean fRSOM was 0.49 ± 0.20. We have shown the possibility of using natural abundance 13C discrimination to quantify fRSOM in a forest soil with an input of carbon only from C3 photosynthesis.