Nitrogen fluxes in forests from atmospheric deposition to soil: use of water flux monitoring and stable isotopes to close gaps in nitrogen transfers
Temperate forest ecosystems are significant sinks for atmospheric nitrogen (N) deposition (Ndep) yielding benefits such as protection of waterbodies from eutrophication and enhanced sequestration of atmospheric CO2. Many uncertainties remain about the fate of Ndep due to the different input fluxes and their spatial and temporal variation, the transformation between different N forms, the complexity of interactions between N and different forest ecosystem components, and the different methods used to quantify N stores and fluxes. Previously, many studies on the interaction between Ndep and forests have focused on Ndep at the soil level, assuming that the interaction between Ndep and the tree canopy is negligible. However, in the last 20 years an increasing number of studies showed how canopy uptake, calculated as the difference between the Ndep input and the fluxes below the canopy (throughfall and stemflow), accounted for a significant fraction of the total input. This could lead to an underestimation of the effects of Ndep on forest carbon sequestration. Moreover, transformations of N passing through the canopy might occur which can change the N dynamics and N availability in soil. Previous studies have shown evidence of biological nitrification and Ndep processing and retention at the canopy level. However, this was reported only at sites where Ndep levels were high or where low background levels were experimentally raised (up to 18 kg N ha-1 y-1). The aim of this research was to resolve in a low Ndep area, some uncertainties related to Ndep processing by forest canopies. The case study area was Griffin Forest (Perthshire, Scotland), a typical Sitka spruce plantation of the UK uplands, characterised by a generally low Ndep (5-9 kg ha-1 year-1 ). Field monitoring was conducted for 5 years of N fluxes in water onto and below the canopy, litter transfer from the canopy to the soil, and nitrous oxide (N2O) fluxes from forest soils. Comparison of rainfall (RF, or bulk precipitation) and cloudwater (CW) with throughfall (TF) and stemflow (SF) measured below the canopy suggests strong transformation and uptake of Ndep in the forest canopy. The annual mean canopy uptake (CU) of N (calculated as a balance of RF + CW – TF – SF) at Griffin Forest was 70%, and varied between 60% in 2014 and almost 80% in 2012. The data showed a significant long term decreasing trend in bulk deposition of NO3 - with peaks during the growing season and a significant strong positive correlation between bulk deposition and CU for NO3- N, NH4-N and total N. These results and a seasonal difference in results were confirmed through a labelled simulated Ndep experiment, where the top of the canopy of three selected trees was sprayed with a 15NH4 15NO3 (98%) - NH4NO3 solution on two occasions, one during the growing season and one in winter. Background RF and TF and SF below the trees were collected and analysed for 15NH4 and 15NO3. The N CU in summer (N% = 78±4.5; 15N% = 85.7±2.9) is much higher than the amount recovered in winter (N% = 51.3±11.9; 15N% = 43.7±14.2) suggesting at least a partial retention by the plant, together with possible transformations from inorganic to organic N and N gaseous losses. To assess actual plant retention more effectively direct application of a 15N solution to target branches in situ showed that ~14% of the applied N was recovered in needles and twigs after a period of 24 hours. The short time scale in which this recovery occurred and the particularly dry conditions during the experiment could lead to an underestimation of the actual potential N retention by the canopy as foliar uptake depends on leaf wetness, as literature suggests. The fate of organic N transfer to forest soils through litter was addressed through 15N-labelled litter plots sampled 2 and 4 years after the litter replacement. Results show that the different soil features typical of a forest plantation (ridge, undisturbed soil, and furrow) had different δ15N and estimated 15N recovery over time. The estimated maximum recovery was ≃52% in 2017 of which ~16% was found in roots. N2O-N losses from soil measured on a 3-year period showed a significant increase in time and they were positively correlate to reduced N bulk deposition. Their order of magnitude was similar to N losses through streamwater and represented a small portion of the atmospheric inputs. This research has shown that the effects of forest canopies on N deposition occurs at two levels. Firstly, at the canopy level there is consistent uptake of the N input, with only 30% or less directly reaching the soil as inorganic N. A second indirect effect is that the uptaken N, either directly by the plant or through bacterial/fungi sequestration at the phyllosphere level, is likely to be transferred to the soil as organic N via litter and here rapidly used by the plants. The water and gas flux monitoring showed that no major leaching occurs, indicating that the forest acts as a N sink. The research results confirm the highest figures in the literature of nitrogen canopy uptake. At the relatively low deposition rates present in the UK uplands, Ndep represents an important extra source of N to the forest N cycle. Lower N fluxes measured under the canopy, excluding the canopy effect and those taken under high 15N-N tracer additions, could underestimate the extra carbon sequestration induced by the Ndep. The research results and findings are relevant to understanding and modelling N cycling and its impacts on forest growth and carbon storage in similar forest systems in the UK uplands and at a broader scale under similar Ndep conditions.