Ecosystem carbon and nitrogen losses from temperate agricultural peatland with mineral soil coverage

Draining peatlands for agricultural production induces massive amount of carbon (C) and nitrogen (N) losses, therefore contributing to climate change, environmental pollution and peat degradation. Full rewetting or restoration could decrease peat mineralization, by raising the water table and might save substantial parts of the soil organic matter (SOM) oxidation and halt peatland subsidence. However, with rewetting, intensive agricultural production is in many cases not possible anymore. Hence, there is a trade-off between environmental goals and agricultural production, creating challenges to implementing peatlands rewetting. Peatland management strategies, which could not only sustain the productive use of organic soil but also counterbalance soil subsidence and mitigate peat decomposition and climate change, are urgently needed for those regions where the peatlands surface contribute greatly to food and feed production and economics and thus where restoration of near-natural ecosystem is difficult to implement. Nowadays, in order to counteract the subsidence of drained peatland, mineral soil coverage is an increasingly used practice in Switzerland and other European countries e.g. Norway, Germany and the Netherlands.
The main objectives of this thesis were to investigate the effect of mineral soil coverage on peat decomposition, i.e. C loss and N balance, and on agricultural productivity. The objectives were achieved by in-situ quasi-continuous GHGs observation, an isotope labeling experiment and associated lab incubation. The experimental site was located in the Swiss Rhine Valley, a drained peatland with fen peat ~10 m thick. Drainage with ditches commenced before 1890. An intensive drainage system with drainage pipes and pump was built in 1973. The site was used as pasture until 2013, and since then as an intensively managed meadow. In 2006 to 2007, one part of the field (1.7 ha) was covered with mineral soil material, with a thickness ~ 40 cm, to improve the trafficability and agriculture usability and to counteract peat subsidence. The field experiments were established at this mineral soil coverage site (Cov), and the adjacent drained organic soil without mineral soil coverage was used as reference (Ref). Both sites have the identical farming practice, with 5 – 6 times cuts per year, receiving c. 230 kg N ha-1 yr-1 of nitrogen fertilizer, and similar vegetation. The C losses were determined by in-situ CO2 and 14CO2 fluxes measurement and soil profile-based C and 14C observation. The long-term drainage induced C loss was determined by comparing the hypothetical non-drainage C stock and the measured C stock. This hypothetical non-drainage C stock was estimated by the age gradient between the deeper 14C dated peat layer, the hypothetical non-drainage surface peat layer and its carbon accumulation rate. The source of C loss, i.e. its origin within the soil profile, was determined by Bayesian stable isotope mixing model with the measured soil respired F14CO2 value, atmosphere F14C value, which here represents the radiocarbon signature of fresh plant residues, and the F14C value of topsoil C and middle layer soil C and the measurement uncertainty. The N losses were measured by in-situ quasi-continuous N2O observation by using automatic time integrating chamber (ATIC) systems and an in-situ 15N tracer experiment with labeled mineral fertilizer. The latter allowed to determine the N allocation in plant-soil system, and the N losses by an isotope and mass balances approach.
The results demonstrated that mineral soil coverage moved the source of decomposition of soil organic carbon (SOC) from a higher share of old peat towards a higher share of relatively younger material located in the topsoil. In this study, decadal drainage of organic soil for agriculture induced 41 – 75 kg C m-2 loss, which is equivalent to annual C loss rates of 0.49 – 0.58 kg C m-2 yr-1 and 0.31 – 0.63 kg C m-2 yr-1 for drained peatland with (Cov) and without (Ref) mineral soil coverage, respectively. Correspondingly, the carbon sources of heterotrophic soil respiration (soil Rh) were a mixture of fresh plant residues and soil C. Carbon from peat decomposition contributed around half to the total heterotrophic CO2 from the soil in site Ref, partially stemming also from carbon stored in the subsoil. Mineral soil coverage had no significant effect on the amount of heterotrophic respiration, however at Cov, the radiocarbon signature of heterotrophic CO2 was significantly (p<0.01) younger than at Ref.
The mass balances for 15N tracer in the system was used to account for the quantitatively recovery and loss, the 15N which is not retained in plant and soil system are defined as loss. 15N losses from site Cov was 10 % lower (p < 0.05) than Ref. The lower N losses from Cov might be driven by the higher soil 15N retention, with 20 ± 2% of the added 15N residing in the soil, however, it was only 9 ± 3% at Ref. The plant 15N uptake was not different between the two sites, despite the higher (p <0.05) N uptake at site Cov. The lab incubation results showed that soil Nr_min and 15Nr_min release was ~ 3 times higher at Ref than Cov, however, the specific release per unit soil nitrogen (specific soil Nsr_min) release showed the opposite, indicating a faster SOM turnover rate at Cov. Importantly, mineral soil coverage induced a strong reduction of N2O emissions. During the experimental period, site Ref released 20.5 ± 2.7 kg N ha-1 yr-1 N2O-N, whereas the N2O emissions from Cov was only 2.3 ± 0.4 kg N ha-1 yr-1. At both sites, N2O peaks related to fertilization events contributed more than half of the overall N2O emissions. However, not only the fertilization induced N2O peaks, but also background N2O emissions were lower with mineral soil coverage.
The agricultural productivity was determined over four harvest periods. Grass biomass and N uptake were used as indicators of agricultural productivity for the two sites. During the experimental period grass took up 274.34 ± 22.78 kg N ha-1 from site Cov, more than at the Ref with 229.97 ± 10.56 kg N ha-1 grass N uptake. However, the grass yield was not different for the two sites with 13817 ± 738 kg ha-1 and 13011 ± 290 kg ha-1 dry biomass harvested for site Cov and Ref, respectively. This indicats that mineral soil coverage could sustain the agricultural productivity of drained peatland
In conclusion, the field experiment results demonstrated that mineral soil coverage could maintain the agricultural productivity of drained peatland, while at the same time reducing the peat decomposition rate as indicated by C and N losses. Mineral soil coverage, therefore, seems to be a promising environmental footprint reduction option for intensively used drained organic soils when a sustained use of the drained peatland for intensive agricultural production is foreseen and potential rewetting and restoration of the peatland is not possible.