In addition to CO2 fluxes we also measured greenhouse gas (GHG) emissions, including nitrous oxide (N2O) and methane (CH4). However, N2O emissions were minimal during the first two years - as can be expected on nutrient (i.e. N) poor blanket bog - and only CH4 emissions were further investigated. Methane emissions were measured first only on non-vegetated peat areas (vegetation was removed and cut regularly) but measurements over vegetated areas were subsequently added to capture plant-mediated transport (PMT) impacts from vegetation, especially sedges, as well as oxidation potential from vegetation, especially Sphagnum. Moreover, methane flux measurement methodology adapted over time with availability of state-of-the-art analysers (moving from cover bog GC to LGR GHG and then to LiCor GHG analysers).

Soil GHG flux examples for all three sites (from left to right: July; October; December 2012):

Mossdale - wettest (top), Nidderdale - driest (mid) and Whitendale (bottom)

So far, the GHG fluxes revealed clear differences only for CO₂ and CH₄ (methane) fluxes between sites and less so for treatments (i.e. N₂O emissions were very similar over the first three years and an average value of 0.03 nmol N₂O m^-2 s^-1 or 0.035 g N₂O m^-2 yr^-1 was used in any further GHG calculations; see below left). The annual methane emissions showed considerable inter-annual and general site variability and emissions also showed a huge difference in using mean (much higher) versus median (much lower) fluxes. Moreover, average annual methane emissions (below right) showed considerable management differences (uncut > mown > burnt), and revealed particularly high (median) emission peaks after or during very wet and warm years (i.e. 2015 - 2017) - notable this was similar to measurements reported for another UK blanket bog site (Moor House, NNR; contact: Prof Niall McNamara):

Combining the mean NEE budgets with the median CH₄ and average N₂O fluxes, we estimated the net CO2-equivalent greenhouse gas fluxes (using their respective Global Warming Potentials, GWPs, over a 100 year time period) based on a transformation of CH₄ into CO₂eq = x 25 and N₂O into CO₂eq = x 298.

A key component of the GHG flux is methane. Notably, aerenchymous plants such as sedges act as 'vents' for methane, allowing methane to escape quickly from deeper, anoxic and methane producing (methanogenesis) peat layers to the atmosphere, thus avoiding access to methane oxidising bacteria (methanotrophy). We therefore also measured GHG (including methane) fluxes from plots with or without vegetation and assessed the percentage of sedges within each plot. This allowed us to correlate methane fluxes against %sedge cover and the overall plant mediated transfer (vegetated - peat) was 62%):

Above: Methane emissions were positively correlated with higher sedge cover (left) and Comparison of CH4 (methane) fluxes (in nmol m-2 s-1) using the Los Gatos UGGA analyser on uncut plots on either vegetation free (peat) or vegetated areas (regrown since heather cutting in 2013) for all three sites (centre) or for only Nidderdale and Whitendale (right)

Notably, GHG fluxes are higher from vegetated plots (NEE cut - regrown) versus vegetation free peat (GHG) plots, and the fast re-vegetation at Mossdale with sedges after mowing lead to much higher methane fluxes from mown plots than measured on burnt plots with much less regrowth overall. However, intact (uncut) plots also showed fairly high methane emissions, even under lower overall sedge cover (as it is Calluna dominated).

Below: GHG budgets (i.e. the CO2-equivalent greenhouse gas flux) were compared to those published by the IUCN UK Peatland Program initiative (left), and either with using the mean (centre) or the median (right) methane emissions (note the very good agreement when using median methane fluxes between the project values and the IUCN UK PP estimates (in fact, all three sites are similar to the IUCN UK PP's 'near natural' values:

Finally, the average annual net GHG budgets (i.e. the CO2-equivalent greenhouse gas flux) were estimated over the project period for each site (Nidderdale, Mossdale, Whitendale) and management scenario (uncut, burnt, mown), which included accounting for average emissions from management (fuel for burning and mowing management; combustion loss and charcoal sequestration from biomass burning), assuming a 20 year management rotation. However, we did not include the fluvial C exports as the uncertainties are considerable, both in terms of actual C turnover (CO2 & CH4 production; Billett et al., 2015) as well as C age (young vs old carbon; Billett et al., 2015), although estimates of around 50% C losses of total C export during stream and river export times are likely in a UK peatland/river context (Moody & Worrall, 2017). Notably, the latter study did relate degradation of fluvial C not under stream/river conditions but under air light and temperature conditions (samples stored outside the river on land), degradation rates could therefore be expected to be different, especially under lower light (effective absorbance could be higher in a river) and lower or also higher temperatures (depending on season and dirunal cycling - so similar temperatures but under different light conditions) as shown for a study in Devon (Webb et al., 2003). Uncertainties about C export rates in relation to emissions are thus still very high indeed, requiring further targeted research.

Above: The average (for 2012 - 2020) (left) cumulative (and below time series of) CO2 balance (considering only CO2 fluxes) and (right) the net GHG budget (i.e. the CO2-equivalent greenhouse gas flux considering CO2, CH4 and N2O emissions); both graphs also consider the C losses (combustion) and C gains (charcoal) from biomass burning (with the 'one off' burn emissions or gains calculated as an annual loss or gain over an average 20 year management rotation); the below time series show the moving annual averages comparing the main managements (uncut, mown and burnt). Whilst cumulative C emissions show very similar losses for mowing and burning management, net cumulative GHG (GWP) emissions were lower (since 2017) for burnt plots even considering the high CO2 loss during the biomass combustion in 2013; values are all averages across the three sites.

However, as values for global warming potential (GWP) are context dependent (see Allen et al., 2018: "Ambiguity arises because emissions of cumulative pollutants and [short-lived climate pollutants] (SLCPs) translate into impact on the planetary energy budget in fundamentally different ways: for cumulative pollutants like CO2, radiative forcing largely scales with the total stock (cumulative integral) of emissions to date, while for SLCPs like methane, it scales with the current flow (emission rate) multiplied by the SLCP lifetime"), we considered a range of GWPs. First, we adhered to the IPPC's 4th Assessment values based on the 100 year GWP (for methane and N2O), secondly, we considered the sustained GWP (SGWP; see Neubauer & Megonigal, 2015; Balcombe et al., 2018), this considers that the standard GWP does not adequately represent sustained ecosystem emissions of methane, thirdly, we also considered the SGWP for both, the 100 year and the 20 year time period (i.e. the latter much better representing the actual short-term climate effects, particularly of the short lived methane gas, and the need to achieve urgent reductions in atmospheric GHG emissions); all also considered (for very variable methane emissions) either using the overall median fluxes or the mean of the 8 annual median fluxes (i.e. for all plot replicates throughout a year).

A note of caution: a high CO2-equivalent greenhouse gas flux is equal to the radiate balance (although our calculations do not include all parts of the overall radiative balance, such as albedo and off site fluxes), but it does not necessarily imply a net radiative forcing or warming impact on the atmosphere; this depends on the change over time (if climate or management increased the emissions, then this translate into a net warming impact) - a constant high emission rate does not mean a warming effect as, over time, the global impact can be neutral (see Neubauer, 2021); thus radiative forcing would be applicable when comparing between the different management options (i.e. the difference being equal to the net radiative forcing attributable to a specific management compared to another).

Below: Noticeable are the increasing net GHG emissions to derive an overall net CO2-equivalent greenhouse gas flux (see below tables for 2012-2019) based on the various GWP options (from top to bottom: IPCC < SGWP 100 years < SGWP 20 years) and the overall higher net GHG emissions when using mean methane emissions of overall median emissions (left) vs. annual median fluxes (right) and from mowing over burning after 6 years of post-management. However, burning emissions also need to include the additional emissions from biomass combustion (about 90 tCO2eq m-2 yr-1), which is indicated also for charcoal C capture and mowing emissions.