Guidance - Evidence on carbon and nature

Purpose

This technical document aims to provide an indication of carbon stores and sequestration (per hectare) for a range of habitats. The information was originally meant to be a support for staff promoting nature-based solutions. It should be used judiciously depending on the context and the application.

This is a ‘live’ document which the Climate Change Technical Team in NatureScot update at regular intervals as new research is published. 

The document includes (see table of contents)

1. A summary table highlighting key numbers
2. What the evidence says about each habitat
3. Detailed tables showing the range of figures found in the literature

Metrics

• Stores are shown in tC (tonnes of carbon)
• Sequestration is shown in in tCO2e (tonnes of carbon dioxide equivalent)
• Conversion from C to CO2 is x3.67
• CO2 only measures carbon dioxide. CO2e stands for "carbon dioxide equivalent" and measures CO2 plus all other greenhouse gases including methane, nitrous oxide, and other ones which are less important for us (100-year global warming potentials). In peatlands, methane emissions will be significant and CO2 and CO2e emissions of peatlands will have different values, as the latter will also include methane and nitrous oxide emissions.

Note:  Other than freshwater and some coastal habitats where methane and nitrous oxide need to be taken into account as significant, for other habitats studies usually consider carbon dioxide only and there is lack of evidence on other greenhouse gases. In this document, the values for sequestration/emissions are provided in CO2e or in CO2. Overall, across habitats, except in wetlands (incl coastal), we assume that GHG sequestration/emissions are mostly linked to carbon dioxide - so that it should be good enough for our purposes to use values in CO2/ha/year for CO2e/ha/year even if not totally accurate. Studies for freshwater habitats often consider methane and nitrous oxide alongside carbon dioxide but some do not – in which case, it should be highlighted. 

Acronyms

• Carbon (C)
• Carbon dioxide (CO2)
• Dissolved Organic Carbon (DOC)
• IPCC (Intergovernmental Panel on Climate Change)
• Methane (CH4)
• Nitrous oxide (N2O)
• Particulate Organic Carbon (POC)
• Soil Organic Content (SOC)

If you see any figure or other evidence which looks odd to you, or you would like to make a comment, please get in touch [email protected]

The work is not a systematic review or a meta-analysis. Confidence is a judgement by staff, based on the review, climatic and relevance of other environmental conditions, number of studies, comparison and agreement between studies.

Summary Table - Carbon Stores and Carbon Sequestration per Habitat

Notes on estimates

• Peatlands - Carbon store numbers are based on the National Soil Inventory of Scotland (2007-2009) for which a repeat a few years later did not suggest a change in stores (as a result we consider confidence medium); carbon sequestration rates/net GHG emissions come from evidence that informed the GHG inventory. We consider the confidence in peatlands estimates in Scotland to be medium. There is however large spatial variability across sites.
• Upland heath - Carbon store numbers are based on the National Soil Inventory of Scotland and we consider the confidence to be low-medium. There is large heterogeneity in soil type x habitat in the uplands. Evidence on carbon sequestration/GHG emissions is insufficient.
• Woodlands - These values are based on the methodology in the Woodland Carbon Code, which is based on the CARBINE model which is used for the UK GHG inventory so the confidence is medium. These values can vary further than is allowed in the model, depending on how open the woodland canopy is and the amount of understorey vegetation. We consider the confidence to be medium.
• Scrub - In Europe, the research on carbon accumulation in scrub was undertaken in abandoned agricultural land in mountain environments of Central and Southern Europe. The evidence in inconclusive on the impact of vegetation succession from grassland to shrubland on soil organic carbon stocks (contradictory findings). Unsurprisingly, there is an increase in above ground carbon stocks in the transition from grassland to shrubland. There is a lack of research in the UK or similar environments. There is a lack of evidence on carbon and vegetation succession on carbon-rich soils. We consider we cannot use existing estimates and consider the evidence to be very low.
• Hedgerows -There has been a few empirical studies in the UK but more studies in temperate regions. Estimates made on a hectare basis assume a full hectare of hedge and are not based on a set hedgerow density within the landscape – the actual value in a field will be lower, hence estimates are also provided in km. We decided to use one recent empirical study undertaken in the NW of England as representing conditions closest to Scotland, and a meta-analysis of hedgerows in temperate climates, as well as a Rapid Evidence Assessment for ClimateXChange. The values in the meta-analysis and Cumbrian study were very similar on carbon sequestration. The type of hedgerow management has a strong influence on values, which introduces complexity in terms of understanding the potential sequestration under different design and management intensities. As a result, more studies are needed and we consider the confidence in the evidence available to be low for Scotland.
• Agroforestry - There are few empirical and modelling studies of agroforestry systems in temperate environments and even less evidence in Scotland. Besides the limited evidence, there are varied agroforestry systems with very different tree densities, resulting in substantially different carbon sequestration rates per hectare. Location (soil, temperature etc) will also affect sequestration. Hence it is difficult to make comparisons between estimates. For shelterbelts, estimates made on a hectare basis assume a full hectare of shelterbelt and are not based on a set shelterbelt density within the landscape – therefore the actual value in a 1-ha field will be lower. There is hardly any agroforestry in Scotland other than ancient wood pasture and hedgerows. A few farmers have however implemented new agroforestry on their farms. We consider the confidence in the evidence to be low for Scotland.
• Grasslands - Carbon store numbers for acid grasslands are based on the National Soil Inventory of Scotland. We consider the confidence in the evidence on carbon stores to be low-medium as there is large heterogeneity in soil type x habitat in the uplands. Evidence on carbon sequestration is insufficient while GHG emissions will be affected by the livestock component. Carbon stock estimates for neutral grassland are based on the Countryside Survey 2007, and based on previous surveys, carbon stocks do not seem to change much, so the confidence is medium. Note the value is for soil at 15 cm depth, unlike acid grassland values above which are for 1 m depth. Not enough sampling of calcareous grassland for measuring carbon storage has been undertaken. Grasslands are an agricultural land use and the impact of various types of management on carbon in semi-natural grasslands is not well understood. We consider the confidence in the evidence on carbon stores in Scottish grasslands to be medium.
• Freshwater - Freshwater play an important role in the carbon cycle by emitting methane, carbon dioxide, storing carbon and transporting carbon. However the evidence is rather patchy and due to the complexity of the processes, and variability, it is difficult to provide conclusive estimates with some exceptions. Floodplains potentially play an important role in the carbon cycle by emitting methane, carbon dioxide, storing carbon and transporting carbon. Many floodplains are no longer functional. There is agreement that a functional floodplain can accumulate large stores of carbon but due to the complexity of the processes, the variability, it is difficult to provide conclusive estimates with some exceptions. We lack evidence on wetlands other than peatlands. We consider the confidence in the evidence to be very low for Scotland.
• Coastal and marine - The carbon stored in the coastal and marine environment is known as blue carbon. The IPCC definition is “All biologically-driven carbon fluxes and storage in marine systems that are amenable to management can be considered as blue carbon.” Within Scotland that means, saltmarsh and seagrass. Vegetates sand dunes, including machair may also be included, however there are very limited data for carbon values within these habitats. Other habitats that are potentially amenable to management but do not have the evidence base to support this yet, include kelp and subtidal sediments. Finally, there are habitats within the marine environment that can be considered as supporting blue carbon habitats, for instance the calcifying aggregations such as maerl, serpulid reef, and native oysters. While the process of calcification produces CO₂, these 3D structures can support enhanced deposition of organic carbon in underlying sediments. They also support high levels of biodiversity. This is an emerging area of science, and the evidence base is quickly evolving and values change fairly regularly. Various initiatives, including the Scottish and UK Blue Carbon Forums and the UK Blue Carbon Evidence Partnership exist to improve the evidence base and provide a coordinated and standardised approach to blue carbon evidence. Carbon stores are generally highest within the sediments of blue carbon habitats with stocks usually reported to 0.1 m depths. Scottish subtidal sediments provide the largest stores of organic carbon on account of their sheer; it is not clear if or how disturbance of surface sediments contributes to loss of carbon. There are very few data on organic carbon burial rates within subtidal sediments, beyond the Scottish fjords. 

Review of evidence for carbon and habitats

Evidence was found in peer reviewed journal articles and synthesis reports, including ClimateXChange’s Rapid Evidence Assessment reports and Natural England’s literature review (Gregg et al, 2021), which provides review of evidence drawing from research undertaken in the UK, and other parts of the world were conditions are not too dissimilar. This review is not a systematic review. We initiated the review starting from the Natural England’s report published in 2021, and largely then focused on publications published thereafter. Journal articles were identified using Scopus with search words including the habitat and carbon or greenhouse or climate; for some habitats, we also carried out searches on nitrous oxide and methane. With regards to the geographical spread, we have been more conservative for actual figures than for other aspects of the evidence. The figures are for habitats in good condition and/or degraded and this is differentiated where applicable.

The habitats covered on land include a range of broad semi-natural habitats and linear features. A short annex is provided on soil carbon stocks in cultivated land, but due to the variability in management practices, and management of cultivated land being outside the scope of this work, it was decided not to consider carbon sequestration. Agroforestry is however covered in the main document with a particular focus on the tree component. The review of species-rich grasslands does not include the livestock component of grassland systems, but it is noted that an understanding of the latter is essential to assess net GHG emissions from grasslands.

On coastal and marine, the document provides a summary of estimates and points to a NatureScot Commissioned Report for more detail.

Note on soil carbon: in order to fully assess changes in soil organic carbon over time, it is important to take into account changes in dry bulk density and horizon/soil thickness. Sampling at fixed, shallow depths is unlikely to be able to take account of changes in the carbon-rich topsoil thickness. However soil organic carbon is measured at various depths across studies making comparison difficult; attention must be paid to the depth of sampling when using figures.

1. Peatlands

• Peatlands are the largest terrestrial soil carbon store. Peat accumulations have been built up over thousands of years from the partial decomposition of Sphagnum mosses, other plants and other organic matter. Pristine peatlands hence sequester carbon, and organic matter will accumulate, in theory indefinitely (all other things remaining equal) i.e. carbon stores can continue to grow.
• However, peatlands in Scotland are widely degraded because of historic drainage, extraction of peat, plantation forestry, controlled fire/muirburn and wildfire and nitrogen deposition. Degraded peatlands have low water tables, resulting in losses to the atmosphere through oxidation of organic carbon in the peat and losses of water-bound carbon into streams, rivers and other bodies of water (Bruneau and Johnson, 2014).
• Peatland habitats have a layer of peat >40 cm. Estimates of peat carbon stocks are uncertain due to the variation in the depth of peat soils. Based on the National Soil Inventory of Scotland, peat soils down to 1 m hold between 273 and 823 tC/ha with an average of 547 tC/ha (Rees et al, 2018). Areas of deeper peat (2-8 m) will of course hold bigger carbon stocks. Although most peats are deeper than 60 cm, the contribution to total carbon stocks within each depth range decreases with depth and after 150–200 cm is considered small (i.e. there are many very small areas of very deep peat) (Aitkenhead and Coull, 2019). No field studies have been found that estimate the carbon stored in the living Sphagnum layer.
• Depending on their condition, peatlands will emit carbon dioxide, methane and nitrous oxide, as well as water-bound carbon. Peat accumulation depends on a high-water table close to the surface, which in turn creates the anaerobic conditions required for methanogenesis (the microbial formation of methane). While increased emissions of methane will offset some of the carbon gains from peat accumulation, raised water levels generally result in a lower emissions overall than drained sites (Gregg et al, 2021).
• The net GHG emissions from peatlands will vary ranging from 0.32 tCO2e/ha/year for near natural bog, 3.32 tCO2e/ha/year for drained heather/grass dominated peat, 15.88 tCO2e/ha/year in (drained) extensive grassland - and up to 37.17 tCO2e/ha/year for cropland on peat (of which there is very little in Scotland). Only near natural fens’ high rate of carbon dioxide sequestration from the atmosphere outweighs methane and nitrous oxide emissions, and indirect carbon losses via leaching. Over long time horizons (100s to 1000s years), the shorter atmospheric lifetime of methane (12 years) compared to carbon dioxide (100s of years) means that near natural bogs and fens have a strong net cooling impact (Evans et al, 2017).
• The mean annual effective water table depth - WTD_e; that is, the average depth of the aerated peat layer, seems to have a significant effect of management measures on GHG fluxes. Data from flux towers in the UK and Ireland combined with 49 published studies showed that boreal/temperate peatlands are predicted to act as net CO2 sinks when Water Table Depth (WTDe) < 20 cm. Peatlands with WTDe > 25 cm were overwhelmingly net CO2 sources. Between a WTDe of 5 cm and 13 cm, the cooling effect of CO2 sequestration exceeds the warming impact of CH4 emissions (based on 100-year global warming potentials, GWP100), implying that peatlands in this range (which is typical of natural systems) will have a small cooling impact on a 100-year time horizon (Evans et al, 2021).
• Peatland restoration currently achieves a reduction in GHG emissions rather than carbon sequestration. Eventually as the peatland is fully restored it may start sequestering; there are indications in some older restoration sites in Scotland (>10-15 years) that these function as carbon dioxide sinks, but we don’t have sufficient data for methane, other emissions or stability over time of this sink function (pers. comm, Rebecca Artz, James Hutton Institute).
• There is ongoing research into assessing accurately the carbon sequestration from peatland restoration work. A meta-analysis suggests that rewetting results in a decrease in carbon dioxide emissions and an increase in methane emissions, respectively by −1.43 ± 0.35 tCO2/ha/year, and +0.033 ± 0.003 tCH4/ha/year overall. Presence / absence of vegetation in the baseline, and changes in vegetation cover following rewetting, significantly influence carbon dioxide and methane emissions, especially after a longer period following rewetting. Long-term monitoring period is lacking following rewetting interventions to capture trends in GHG emissions post-rewetting (Darusman et al, 2023).
• Various figures for sequestration/emissions post rewetting can be found in the literature, though more evidence is needed in Scotland. The methodology in the Peatland Carbon Code provides a standard method for quantification of GHG benefit in projects for funding from the sale of climate benefits and uses emission factors to assess the net effect in changes to GHG emissions in moving between Peatland Code’s condition categories. For example, restoring from ‘actively eroding’ to ‘drained’ would save 19.30 tCO2e/ha/year while restoring from ‘modified’ to ‘near natural’ would save 1.46 tCO2e/ha/year (Peatland Code, 2023; Smyth et al, 2015).
• Peatland formation is climate-dependent, and as a consequence of climate change, some areas in Scotland may not be able to support peatlands. The potential additional emissions could be significant (Ferretto et al, 2019). However the assumption is that healthy peatlands will be more resilient to changing conditions than degraded ones. For example, raising water levels in drained peatlands reduces their susceptibility to deep burns; however without the recovery of a new moss layer rewetting alone is not sufficient to reduce the risk of deep burning. Improving peatland management practices and restoration is key to avoidance of catastrophic wildfires (Granath et al, 2016).
• Peatlands are exposed to climate risks through extreme weather events, drought, heat wave, heavy precipitations and wildfire. Under warmer conditions, while there will be higher carbon uptake from the atmosphere by plants, based on up-to-date knowledge, the result of climate-induced changes in peatland vegetation phenology and composition will be lower carbon accumulation by peatlands. In the longer term, currently pristine peatlands may become net emitter. There is however a lot of uncertainty and evidence gaps in the response of vegetation to change in temperature and rainfall (Antala et al, 2022).

2. Upland Dry Heath and Wet Heath

• Like peatlands, wet heaths are dominated by mosses, primarily Sphagnum, under waterlogged conditions and are found on carbon-rich soils. Upland wet heath is generally a semi-natural community derived from woodland and blanket bog through a long history of burning and grazing (Rodwell, 1991). Wet heaths are usually found on peaty soils with peat layer<40 cm. Soil carbon stores down to 1 m are less than for peatlands, ranging from 114 to 784 tC/ha with an average of 313 tC/ha. Upland soils are complex and variable soil systems. This is because the soils are highly heterogeneous, and also carbon dioxide and methane can be transferred between the soils and the atmosphere. There has been a lack of studies on the GHG fluxes of heath and acid grassland habitats hence it is difficult to quantify emissions/sequestration (Baggaley et al, 2021).
• Due to waterlogging, as in peatlands, wet heath may represent a source of methane and may not always contribute to the net removal of carbon dioxide from the atmosphere, or may be ‘climate neutral’. The large carbon stocks contained in wet heaths make them susceptible to carbon loss as a result of land management practices, therefore it is likely that some wet heath are losing carbon.
• Upland dry heath is found on a range of soil types with a shallow peaty layer. Nearly all dry heath is semi-natural, being derived from woodland through a long history of grazing and burning. Carbon stores down to 1 m range from 47 to 648 tC/ha with an average of 205 t/ha. Less carbon is stored than for wet heath likely due to the relative thickness of the organic layer. As for wet heath, there is a lack of evidence on the net GHG balance in dry heath, and how habitat condition influences it (Baggaley et al, 2021).
• There is limited evidence on the carbon stored in heathland vegetation. The Countryside Survey reported 2 tC/ha in dwarf shrub heath and a study in upland Scotland reported a sequestration rate of 5.6 t CO2e/ha/year (Gregg et al, 2021). Overall there is not enough evidence to even propose a range of figures for heathland vegetation.
• There is strong evidence from studies in both the UK and Europe that, within dry upland and alpine heathland communities, grazing pressure (from both wild and domestic animals) has a negative impact on above-ground biomass carbon stocks, but the understanding of the impact of below ground carbon stores is limited. The effect of grazing of wet heath carbon stores is poorly understood (Baggaley et al, 2021).
• There is limited and inconclusive evidence on the impacts of muirburn on carbon stocks and net GHG emissions, with few studies undertaken, and mostly focusing on single component of the carbon budget. There is moderate evidence that muirburn results in reduced peat accumulation. Peat after a burn is more likely to be subject to erosion than peat that has not been subject to a prescribed burn The carbon balance will depend on the time it takes for the ecosystem to recover via fresh plant growth and biomass accumulation during the inter-fire interval. Wildfire will also have an impact and there is evidence that muirburn directly causes a proportion of wildfires that occur, however there remains uncertainty regarding this proportion. There are no primary studies which directly studied whether variation in fuel loads resulting from muirburn influence the subsequent occurrence or likelihood of wildfire on moorland (Holland et al, 2022).

3. Woodlands

• Native woodlands in Scotland make 22% (over 300,000 hectares) of woodlands and 4% of the land area (Scottish Forestry, Native Woodland Survey 2006-2013, 2023). The potential for expansion is much larger reaching 2.96 Mha outwith designated areas. In particular, part of the Scottish uplands would be suitable for woodland expansion through natural regeneration, including scrub-type woodland up to a naturally established tree line, and montane and other scrub habitats on highly exposed land (currently very depleted) (Fletcher et al, 2021).
• Native woodlands in Scotland are predominantly overgrazed, more than half are compromised by grazing impacts which would prevent successful tree and shrub regeneration, and have little or no understorey (Scottish Forestry, Native Woodland Survey 2006-2013, 2023) due to deer principally at present. Old growth forests such as the Atlantic rainforest are in poor condition, and therefore likely far from their carbon storage potential; we are lacking quantitative evidence on the potential from restoration, both in terms of carbon sequestration and carbon storage longer term. Deer influence above- and below-ground carbon through browsing, fraying and trampling of vegetation. Furthermore, by consuming plant biomass, herbivores are removing vegetation that would otherwise photosynthesise and store yet more carbon (Hirst, 2021).
• Carbon stores in woodlands will include both soil carbon stocks and large above ground biomass, and will vary with the type of species mix and location. Broadleaved woodlands planted on mineral soil can reach up to 600 tC/ha after 100 years. Expansion through natural regeneration will be slower than after planting and after 100 years, the same species mix on organo-mineral soils may have accumulated about 100 tC/ha less (Gregg et al, 2021; Woodland Carbon Code, 2024). At a country level, the average carbon stocks per hectare are 31% higher in ancient woodland than all woodland (Reid et al, 2021): Most woodlands in Scotland are actually conifer plantation forestry (71%) (Forest Research, 2023) and are therefore young trees, harvested on a cyclical basis (circa 40 years).
• The Countryside Survey 2007 showed that upland woodland hold more carbon in their soils (84.7 tC/ha) than lowland woodland (68.3 tC/ha) (0-15 cm) (Emmett et al, 2010), while other values found in the literature show soil carbon stocks in woodland on mineral soils may range from 107 to 173 tC/ha (depth of 1 metre) (Gregg et al, 2021). Carbon stocks in carbon-rich soils range from 224-362 tC/ha (organo-mineral) to 539 tC/ha (peat soil) (depth of 1 metre) (Vanguelova et al, 2013).
• A study based on the Native Woodland Model identified 58 woodland types, and above ground carbon stocks were calculated based on percentage canopy cover. This suggests a range of values depending on the native woodland types such as 32 tC/ha for mixed mountain scrub, 80 tC/ha for upland oak-birch with bilberry, 84 tC/ha for Scots pine with heather, 32 tC/ha for basin bog woodland, 10 tC/ha for scattered juniper (Fletcher et al, 2021). This also suggests that scrub in montane and highly exposed habitats could provide additional sequestration potential.
• A study in England suggests the understorey (woody and ground foliage vegetation) in native woodlands could make up ~15% of the carbon stores above ground excluding deadwood and litter i.e understorey divided by (understorey+overstorey) - for wood, foliage and roots, suggesting an additional carbon sequestration potential (Patenaude et al, 2023). Another study from Poland in Scots Pine plantation shows that the total carbon sequestered by the six dominant species in the understorey was three times greater in the older plantations (80+ years) than in younger plantations (Woziwoda et al, 2014).
• Research in Western Europe show similarities in total carbon stocks (aboveground wood + coarse roots + soil) between two species-mixed stands and the most performing monocultures (beech, oak and pine) (Osei et al, 2022). Research in Canada suggests that that soil C and N accumulation in both organic and mineral horizons can be substantially enhanced by fostering tree evenness and functional diversity (for example, mixed forests that include both broadleaf and coniferous species) (Chen et al, 2023).
• Old growth forests tend to hold large carbon stores above ground and below ground, in both live and non-living biomass, and hold the biodiversity potential associated with complex structures (Leuschner et al, 2022). Research in an old growth oak forest in NW Spain (maritime climate) with trees over 400 years old and limited disturbances suggests that oak recruitment was variable but rather continuous for 500 years. Carbon turnover times ranged between 153 and 229 years and mean carbon ages between 108 and 167 years. Over 50% of above ground biomass persisted ≥100 years and up to 21% ≥300 years (Martin-Benito et al, 2021). Evidence from a literature review and data from hundreds of pilot studies in boreal and temperate forests suggest that old-growth forests can continue to sequester carbon over time (Luyssaert et al, 2008; Gundersen et al, 2021; Luyssaert et al, 2021). A global analysis of hundreds of tropical and temperate tree species showed that for most species mass growth rate increases continuously with tree size. This suggests that big old trees actively sequester large amounts of carbon compared to smaller trees (Stephenson et al, 2014).
• Above ground carbon storage is higher in old-growth forest development stages, particularly in large stems and dead wood, and there is a positive relationship between carbon stocks and biodiversity potential. Carbon sequestration is to be highest immediately following disturbance, bolstered by the rapid growth of younger trees. Natural disturbances in old growth forests followed by early growth stages also show a positive relationship between carbon sequestration and biodiversity. Analysis of data from the Canadian forest inventory (boreal and temperate) suggest in addition that linkages between background climatic conditions and soil C and N accumulation rates are mediated by tree diversity and identity, suggesting that tree composition plays a key role in controlling climate–soil interactions (Mikolas et al, 2021; Chen et al, 2023). A study in secondary forests in New Zealand following natural regeneration (<240 years) showed that above ground carbon stocks were correlated with species-richness (Carswell et al, 2012).
• Disturbing soil to plant trees can cause soil carbon emissions, especially on soils with an organic layer, which is a particular issue in Scotland.  Planting replicated stands of two native tree species onto heather moorlands was associated with significantly lower or no change in soil organic carbon stocks. Despite increased above-ground carbon associated with tree biomass, the loss of soil organic carbon in planted plots resulted in no net increase in ecosystem carbon stock at any site over the duration of the experiment (40 years) (Friggens et al, 2020). Soil carbon loss is driven by oxidation, erosion, decomposition and leaching. Tree establishment can also alter the soil microbial and mycorrhizal fungi communities, influencing the carbon balance (Warner et al, 2021).
• Minimising soil disturbance at planting minimises this emissions source on organo-mineral soils.  Extensive establishment of lower yielding trees on low-quality ground, with organo-mineral soils could result in net emissions that persist for decades (Matthews et al, 2020). Overall, a large portion of Scotland offers limited capacity for tree planting to offset emissions due to poor growth rates and loss of carbon from soils. New forests could be planted in the uplands and achieve meaningful carbon sequestration within the net-zero time scale (2045) only if the lowest disturbance preparation methods are adopted (Baggio-Campanucci et al, 2022).
• Woodland that is regenerating naturally (rather than planted) does not result in the same level of soil disturbance, though growing trees will interact with the microbiota in the soil, soil moisture and temperature among other things (van Noordwijk et al, 2023). A site experiment in England found that tree planting on pasture with minimal soil disturbance was associated with a loss of soil organic carbon in the uppermost soil layer; two possible explanations are the loss of perennial grasses in the understorey and an increase in soil respiration due to lower soil water contents (Upson et al, 2016). A study in the alpine zone in Norway, which compared carbon fluxes and pools in heath, upland meadow, and one Salix-shrub community, suggests that Salix-shrub expansion into an alpine meadow and heath would result in increased rates of litter decomposition and combined with the efficient cycling of nutrients with the help of ectomycorrhiza might reduce the organic soil carbon pools. Research carried out in the Cairngorms along a transect from young very sparse naturally-regenerated trees to open heath (hence not qualifying as a woodland) showed that the carbon gained in the tree biomass was negated by soil carbon losses around the tree for the first couple of decades, and soil carbon stocks were less where trees were found than in open heath (Housego et al, 2025). As time progresses, the carbon stocks in the wooded ecosystem will become greater than in the open heath. The site in the study presented trees are very low density, and the results might be different if regeneration over the same short timescale had resulted in actual high tree density representative of a woodland.
• Using the methodology underpinning the Woodland Carbon Code to generate planting scenarios, carbon sequestration on organo-mineral soils for a mix of broadleaved species is about -2 tCO2e/ha/year in 25 years (low to medium soil disturbance at planting) and about -5 tCO2e/ha/year over a 100 years, but result in net emissions in the 10 years which range depends on the cultivation method (from slightly above 0 up to circa 2tCO2e/ha/year for the most disturbing soil preparation). For natural regeneration, no net emissions are reported, but sequestration is marginally slower due to the more hit and miss nature of expansion through natural regeneration.
• Carbon sequestration outcomes over short timescales are more sensitive to variations in tree growth rates, silvicultural practices (thinning) and soil carbon stock changes related to woodland establishment. Modelling work by Forest Research suggests that, in the period 2022 (woodland creation) to 2050, carbon sequestration (in the carbon pools of trees, deadwood, litter and soil) in the broad-leaved woodland options is in the range 0.9 to 1.6 tCO2/ha/year; for the coniferous woodland options the range is 1.8 to 12.0 tCO2/ha/year. Over longer time horizons (e.g. 2022 to 2100) total net CO2 uptake in the different woodland options are closer to one another. This occurs because most of the faster growing conifer woodlands are under management for production and areas of trees are being felled by thinning or clearfelling, diminishing the rate of carbon sequestration when this occurs. The slower growing and relatively lightly managed broad-leaved woodland options continue to grow and sequester carbon in later decades during this period (Matthews et al, 2022).
• Estimates of soil carbon sequestration based on historical data do not provide an accurate account of future sequestration rates under a changing climate. There are fundamental questions about the fate of the carbon stored in forests and assumptions on the permanence of the carbon stored. Climate-driven forest dieback caused by fire, drought, biotic agents, and other disturbances threatens to release CO2 emissions earlier than anticipated using carbon calculators and expectations on longevity (Anderegg et al, 2020).
• There is not necessarily a correlation between an increase in biodiversity and carbon. Planted monocultures of sitka spruce on mineral soils (or with very light preparation on organo-mineral soils) offer high carbon sequestration and outputs (for timber or pulp) but a poor environment for biodiversity to flourish. More mixed productive forestry would offer more opportunities for biodiversity and bring more resilience which will be critical in a changing climate. Woodland expansion in the uplands, including natural regeneration and appropriately located planting of naturalistic woodlands, would contribute to multiple benefits such as biodiversity conservation, flood alleviation and reduction of soil erosion, with increasing carbon storage a possibility longer term (i.e. beyond 2050, 2100) (pers. comm, Kate Holl, Jeanette Hall and Duncan Stone, NatureScot ).

4. Scrub

• Secondary vegetation succession on disturbed sites will eventually lead to scrub development which will grow from seeds being in the soil or being transported from elsewhere. In the context of nature restoration in the uplands, scrub would form a transition habitat, part of a mosaic of habitats which would also include woodlands, heath, peatland, grassland, tall herbs. This section considered studies on scrub encroachment in grasslands/heathlands but not wetlands. Scrub encroachment on arable soils could be expected to increase above ground and soil carbon stocks over time but this is not a likely scenario in Scotland.
• There is a lack of research on scrub and carbon in the UK, or on the implications of natural regeneration of woodland and scrub on carbon stocks of open habitats. The development of scrub on grasslands might result in increased emissions from soil and reduced soil carbon stocks, but there are contradictory findings between studies. The growth of woody vegetation will draw carbon from the atmosphere and increase above ground carbon stocks (Gregg et al, 2021). More recent studies do not provide more light on the impact of vegetation succession on soil carbon stocks from grassland.
• A study in mountain grasslands of the Central Pyrenees compared soil carbon stocks in meadows, grasslands, young and old shrubs, and young forests. Meadows had the biggest carbon stocks at 113.4 tC/ha compared with 58.9 tC/ha for young shrub (open components of the habitat) and 67.1 tC/a (under the canopy). Only when the forest stage is reached are carbon stocks similar to the meadows (Nadal-Romero et al, 2021).
• In the Massif Central which has also experienced significant agricultural abandonment, above ground carbon stocks were 1.1 tC/ha for herbaceous vegetation and increased to between 100 and 200 tC/ha for the scrub successional stages (with 326 tC/ha in old forests>80 years). Below ground biomass carbon stocks increased from c. 5tC/ha in herbaceous vegetation to 19.57-42.49 tC/ha in scrub. In this study, there was no significant differences in soil carbon stock (0-20cm) between the different succession stages and other studies suggests that there is no clear pattern in the evolution of soil carbon stocks during succession from grassland to shrub to forest (Weissgerber et al, 2024).
• A review of 344 studies across the globe (incl. temperate climes) showed that soil carbon stocks increased up to 30% following shrub encroachment on grasslands. Soil organic carbon also increased in acidic soils after shrub encroachment, especially with Leguminosae shrubs. However this review did not include studies on carbon-rich soils or in cool oceanic climates (Du et al, 2024).
• Vegetation succession elicits changes in the soil microbiome. A study of vegetation succession the Polish Carpathians shows that during succession from open meadows to succession woodlands, there was less efficient use of organic substrates by soil microbes (Sokolowska et al, 2022). A study in mountain grasslands of the Cantabrian Mountains (NW Spain) following livestock grazing abandonment showed that the resurgence of woody species and changes in the productivity of herbaceous species provoke changes in the abundance of several bacterial taxa compared with grazed grasslands (Fernández-Guisuraga  et al, 2022).
• A study in the alpine zone in Norway, which compared carbon fluxes and pools in heath, upland meadow, and one Salix-shrub community, suggests that Salix-shrub expansion into an alpine meadow and heath would result in increased rates of litter decomposition and combined with the efficient cycling of nutrients with the help of ectomycorrhiza might reduce the organic soil carbon pools (Sørensen  et al, 2018). This is a tundra environment, and it is not clear how this would translate to Scotland’s upland environment.

5. Hedgerows

• To date, few studies have undertaken direct biomass measurements to quantify hedgerow C in temperate climates. Some studies used destructive sampling while others used allometric equations. However for the latter accuracy depends on appropriate coefficients for the site and species, and it seems that no such adapted functions for hedgerows are available in the temperate climate zone (Mayer et al, 2022).
• Additional research on the topic has relied on models, based on data from alternative sources such as woodlands and agricultural sites. While information is available on C sequestration and C stores, there is a clear gap in evidence regarding other greenhouse gases (Baggaley et al, 2022). There are differences in growth dynamics and carbon storage of hedgerow-grown trees compared to forest-grown trees. Trees grow in narrow strips where they receive more light than in woodlands, are more exposed to the wind resulting in increased branch production, wider and deeper rooting. The microclimate in hedgerows is warmer and drier during the growing season compared to a forest (Van der Berge et al, 2021). Introducing trees and shrubs on agricultural land may increase the overall above- and belowground carbon input into the soil by pruning residues, litterfall, root turnover and rhizodeposition, mostly within the sphere of influence of woody components (Mayer et al, 2022).
• All studies below are on hedgerows on mineral soils where specified or assumed to be so. They do not all consider all components of carbon stores. Estimates made on a hectare basis assume a full hectare of hedge and are not based on a set hedgerow density within the landscape – therefore the actual value in a field will be lower; estimates can also be reported in kilometres.
• Hedgerow carbon stock estimates in the literature were found to vary with hedge height, width management, structure and hedgerow species. Values range from 25 to 42 tC/ha in vegetation and 43-166 in the soil (Gregg et al, 2021; Crossland, 2015). A meta-analysis of studies on carbon stores in hedgerows in temperate climates (which excludes studies relying on allometric equations and remote sensing) suggests an average of 92 (40) tC/ha (soil+vegetation) (Drexler et al, 2021).
• A study in Cumbria in intensive grasslands across dairy farms showed above ground carbon stocks ranging from 8.34 (3.12-13.57) tC/ha to 33.4 (19.31-47.51) tC/ha depending on the age up to 33 years. Hedgerows being linear features, it is also possible to express these figures per kilometre which would come to 1.25 to 5tC/km in this study (based on 1.5 m width) (Biffi et al, 2023).
• In hedgerows and tree rows, all trees are ‘overstorey’ trees and individual competition between trees will not result in a decrease in productivity, as light is never scarce. The conditions in hedgerows encourage persistent tree growth i.e. growth does not stagnate with the aging of the trees (van der Berge, 2021a). Biomass stocks seem higher in irregular hedges (trimming every 4-5 years) due to an increased hedgerow width, suggesting that the biomass accumulation potential is higher in irregular shapes hedges. Comparison of two 50+ years old hedges in Ireland showed carbon stock in above ground biomass in irregular hedges being around double regular ones (Black et al, 2023).
• Compared to increasing the hedge height, a study in England showed that widening hedges was more efficacious at sequestering carbon into above ground biomass than increasing the height i.e increasing width by 1.6 m resulted in additional 7.5 tC/km compared with 4.2 tC/km for 1.6 m taller (Axe et al, 2018). Another study in England showed that if untrimmed for 3 years hedges could store up to 42 (±3.78)tC/ha in their above ground biomass. It also showed that a 4.2 m wide hedge contained 30 tC/ha more above ground biomass carbon stock than a 2.6 m wide hedge (mean height 3.5 m) (Axe et al, 2017). This suggests that the current classic ‘box hedge’ shape found in Scotland, which is due to overtrimming is not optimal for carbon storage, besides offering a poor quality habitat for biodiversity. As nature-based solutions, hedgerows should be trimmed no more than every three years to maximise these benefits.
• More than 80% of the additional C stocks of hedgerows, compared with cropland, is found in the biomass (Drexler et al, 2021). However, in addition to sequestering carbon within their biomass, hedgerows can increase sequestration of carbon in the surrounding soils. This outcome is a combined function of deeper rooting systems than surrounding vegetation, and by the addition of falling leaves, deadwood and other organic matter which enriches the soils beneath it (Gregg et al, 2021). The addition of trees placed periodically within a hedgerow will increase sequestration rates further. Warner et al (2011) estimated the potential increase in sequestration to be 1.6 tCO2e ha-1 (0.432 t C) if assuming 2 trees per 100 metres of hedge.
• A German study that included sampling of above ground and below ground vegetation in old hedgerows (300 years) showed a below ground to above ground carbon stock ratio at 0.7 highlighting the importance of below ground biomass. Combined with a higher stump biomass due to coppicing, this means that the carbon stocks held in hedgerows could be higher than otherwise thought (Drexler et al, 2023). Another study of hedgerows in the South of England, untrimmed for 3 years, showed below ground biomass carbon stocks to be about half of above ground biomass’ (Axe et al, 2017). Studies of carbon stocks in below ground vegetation are otherwise limited and the Irish study shows a lower ratio ranging from 0.2 to 0.7 (Black et al, 2023), suggesting a dependency on the type of management and most likely other factors such as soil type. Default estimates for root:shoot ratio in perennial woody vegetation recommended by IPCC (2006) is 0.26 in temperate regions (Cardinael et al, 2018).
• Few studies have quantified soil carbon stocks beneath woody features in farmland and comparisons among those that have can be difficult to extrapolate from due to differences in hedgerow tree species, structure and management, climatic conditions, soil type and sampling depth. Overall evidence suggests that soil carbon stocks under hedges are larger than those in adjacent arable fields (Biffi et al, 2022). A review suggests that the establishment of hedgerows compared with grassland may not have a statistically significant effect on soil carbon stocks (though it is unclear what is rotational or permanent grassland) but with considerable heterogeneity between studies (Drexler et al, 2021). Some studies suggest greater soil carbon accumulation in field margins with a hedge and grass strip than with a hedge alone though again more research is needed (Lesaint et al, 2023).
• Research in Western Belgium showed soil carbon stocks in the hedgerow margins to be significantly higher (28.6 ± 10.1tC/km) compared with ‘ghost hedgerows’ i.e. where removed in the past (20.3 ± 4.9 tC/km) (van der Berge et al, 2021b). A study in NW France with old hedgerows and younger hedgerows showed that soil carbon stocks accumulate mainly in the first 30 cm and laterally within 1 m of each side of the hedge. Hedges’ influence on soil carbon concentrations and stocks extended at least 1 m at 0-60 cm depth and up to 3 m at 60-90 cm depth. Soil carbon stocks were higher in old hedgerows than in young ones, but there will be as much variability in soil carbon stocks between sites (Viaud et al, 2021). In a study in Cumbria across dairy farms, soil carbon stores ranged from 130.8 (98.7-163) tC/ha (young hedges) to 175.3 (148.7-201.8) tC/ha (37 year old hedges). All the soils in the studies referred above seem to be mineral soils.
• It was also found that soil type had a significant effect on soil carbon stocks e.g. stocks were significantly higher in stagnosols than cambisols (Biffi et al, 2023). The Belgian study, which looked at ‘ghost’ hedgerows, showed that the removal of established hedgerows results in the rapid loss of their associated soil carbon stocks in arable fields. Leaves, seeds, fruits, tree fine roots, pruning residues and the herbaceous vegetation growing in the hedgerows contribute to a higher input of organic carbon to the soil compared to a treeless field margin (van der Berge, 2021b). This means that the historic removal of hedgerows over the last few decades have resulted in larges losses of soil carbon from agricultural landscapes, particularly in arable areas. Consequently, it is as important to protect existing hedgerows as to plant new ones.
• A meta-analysis across 83 sites in temperate zones concluded that total C sequestration with the establishment of hedgerows on cropland could be between 7.7 and 19.1 tCO2/ha/year for a period of 50 and 20 years, respectively (Drexler et al, 2021).
• Regular trimming has an impact on net biomass carbon sequestration, and hedges trimmed every year will have reduced average biomass carbon sequestration rates. Sequestration rate in above ground biomass across dairy farms in Cumbria was highest in young hedges <6 years (7.65 (2.86–12.45) tCO2/ha/year with no trimming). It then decreased as hedges grew at 6.7 (2.76–10.64) tCO2/ha/year for <12 years; and reached maturity being regularly trimmed (3.14 (1.82–4.47) tCO2/ha/year or 60% lower) i.e. 1-2 years. In managed hedges, sequestration will oscillate between minimal levels in regularly trimmed hedges and maximum levels during periods of fast growth following laying or coppicing (Biffi et al, 2023).
• There can be high variability in soil carbon sequestration rates potential between sites. Soil carbon accumulation is highest in young hedgerows and continue for decades though the sequestration rate diminishes. In the Cumbrian study, soil carbon sequestration rates (0-50 cm) ranged from 13.71 (-7.63–34.9) tCO2/ha/year (2-4 years taking into account initial disturbance of the soil) to 5.43 (2.72–8.15) tCO2/ha/year (37 years old) (Biffi et al, 2022) – which, combined with the biomass carbon stock values, is not dissimilar to the values found in the meta-analysis by Drexler et al (2021). A large French review for the government proposed a soil carbon sequestration rate of 2.75 tCO2/ha/year for 26 years old hedgerows (30 cm depth), hence less than the 4 tCO2/ha/year found in the Cumbrian study at 30 cm depth (Pellerin et al, 2020). The mean carbon sequestration rates of hedgerows planted on cropland will be higher than on grassland due to the higher soil carbon gains in the former compared with the baseline (Burgess et al, 2022). 

6. Agroforestry

• Agroforestry may include hedgerows (covered above) as well shelter belts, alley cropping, wood pasture and parkland. At present, there is very limited agroforestry in Scotland, but there is an aspiration to increase its extent. Ancient wood pasture were a form of agroforestry, as well as hedgerows.
• Agroforestry can sequester carbon and this will depend on the species, density of planting, soil type, elevation. The biggest potential in terms of carbon storage and productivity is in the lowlands, however there are also opportunities for silvo-pastoral systems in the uplands e.g. shelter belts, wood pasture (Perks et al, 2018).
• The majority of evidence available originates from tropical, Mediterranean or modelling studies. Few studies have assessed the impact of agroforestry systems on carbon storage in soils in temperate climates (Baggaley et al, 2022).
• Existing evidence on carbon sequestration in agroforestry is based on a disparate range of studies and the findings are specific to local contexts and methodologies used. This compromises the extent to which the findings can be compared (Saraev et al, 2022).

6.1.  Orchards

• Higher carbon stocks were reported in traditional orchards where trees were allowed grow large and accumulate a high amount of woody biomass, than in intensive orchards where trees are grown at higher densities but are managed for maximum fruit production. Carbon stocks in above ground biomass range from 13.9 to 33.2 tC/ha depending on the intensity of management and 41 to 111 tC/ha in soils.
• Traditional orchards accumulate 0.42 to 2.60 t CO2/ha/year in their biomass while intensively managed orchards ranged 3.21–4.42 t CO2/ha/year. In intensive production systems, a large proportion of the carbon stocks is removed every year through harvest and trees do not grow longer than 30 years reducing their capacity for carbon storage in wood. Trees that are removed, followed by re-planting of new stocks, results in soil disturbance hence likely leading to loss of soil carbon. There is a lack of long-term studies on soil carbon accumulation in traditional orchards. Initial rates of soil carbon accumulation in young orchards will depend on the previous land use. Soil carbon sequestration on six different orchards in Southern England were estimated to range from 0.06 tCO2e/ha/year for traditional orchards up to 0.52 tCO2/ha/year for an intensively managed orchard (Gregg et al, 2021; Robertson et al, 2012).

6.2. Silvo-arable alley cropping

• The carbon storage by silvo-arable systems vary greatly and is dependent on the location, the tree species, the tree spacing between the crop alleys, and the tree management. In an English silvo-arable system with poplars created in 1992 in Bedfordshire, aggregated over 80 years, a soil carbon model, alongside evaluation of timber production, predicted that trees would store 400 tC/ha in the most dense systems (156 poplars/ha) and 240 tC/ha in the less dense (56 trees/ha) compared with arable without agroforestry (Crous-Duran et al, 2020). With poplars at 156 trees/ha, after 19 years the carbon stocks in vegetation was estimated to be 33.6 tC/ha (Upson, 2014). A meta-analysis suggests that soils (all mineral soils) under alley cropping hold significantly lower carbon stocks than hedgerows (Mayer et al, 2022).
• On the same English site, with c.150 trees/ha, another modelling study estimated mean carbon sequestration, as in carbon stored as timber to be 4.0 tCO2/ha/year, relative to the arable system (Garcia de Jalon, 2018).
• In another alley cropping system in England including single apple tree rows intercropped with 24 m wide arable crop alleys (c.60 trees/ha), the net reduction in carbon dioxide emissions of the agroforestry system compared with the equivalent annual crops only arable system ranged from 0.31 to 0.55 tCO2e/ha/year, largely due to carbon sequestration by apple trees. This was calculated using the Farm Carbon Toolkit, and the amount of carbon dioxide sequestered by apple trees, including below ground sequestration, was assumed to be 3.3 or 5.0 t CO2e/ha/year (Staton et al, 2022).

6.3. Silvo-pastoral systems

• In Glensaugh (upland Scotland), a silvo-pastoral system with 400 trees/ha (sycamore, larch, Scots Pine) reached c. 110 (±20) (sycamore) to 175 (±10) tC/ha (larch) in the trees after 40 years including both soil (0-45 cm), and vegetation (incl. woody component). In conifers, both larch and Scots Pine, carbon stores are broadly equivalent in the vegetation (above+below ground) and soil (top and sub soil), while in sycamore there was more carbon stored in the soil than in the vegetation (Beckert et al, 2016).
• In a trial site in Bedforshire with control pasture plots, silvo-pasture (14 year old silvo-pastoral system with 64 poplar trees/ha) and woodlands (14 years old), increment the soil carbon stock stored below the pasture (91.3 (±2.6) tC/ha) was greater than that below the silvopastoral trees (85.2 (±3) tC/ha) and the woodland (73.8 tC/ha). Tree storage was 4t C/ha in the silvo-pasture and 35.9 (±2.8) tC/ha in the woodlands (Upson et al, 2016).
• Tree density in silvo-pastoral systems should not be too high to avoid shading the grassland and compromising grass productivity. Agroforestry trials usually test tree density ranging from < 100 to 400 trees/ha. However at 400 trees/ha, canopy closes in after a decade, removing grazing value, while at 100 trees/ha, the pasture is not much affected (Soil Association, 2019).
• A study based on four modelling sites in Alpine Austria predicts that conversion to agroforestry (80 cherry trees/ha) would profoundly change the carbon dynamics of the agroecosystem. Carbon storage in the vegetation reaches 67.5 tC/ha at a mean tree age of 30.5 years, while the average carbon sequestration rate in the trees in this study was computed at 3.3 and 7.7 tCO2/ha/year for the period of 2020–2050 (Bertsch-Hoermann et al, 2021). In the Bedfordshire trial, the sequestration rate was estimated to be 1.02 tCO2/ha/year over 14 years while in Glensaugh it was 4.9 tCO2/ha/year over 40 years (Burgess et al, 2022). A meta-analysis of soil carbon sequestration in agroforestry systems revealed statistically significant higher mean soil carbon sequestration rates in hedgerows compared with silvo-pasture. If grassland soils are close to saturation, the planting of trees as part of a silvo-pastoral system will not promote further carbon sequestration and tends to show a small loss of carbon (mineral soils) (Mayer et al, 2022). In a systematic literature review, carbon sequestration in silvo-pasture in cool temperate Europe was estimated to be about 10.7 tCO2/ha/year including 0.7 (±0.55) in soil (at 260 trees/ha) and 10 (±1.33)in vegetation (at 225 trees/ha) (Cardinael et al, 2018).

6.4. Shelterbelts

• Shelterbelts are tree rows, some of which were planted for that purpose, while others may be abandoned hedges that successfully grew into trees. Carbon stocks in such tree lines were found to be similar to stocks found in forests on a per tree area basis. Based on modelling over a 50 year time series, a row of trees with 2 m width of grass to the side of each tree line resulted in a carbon gain of 46 t C/ha, equivalent to 3.4 tCO2/ha/year. Using the Woodland Carbon Code, assuming a Yield Class 6 for beech, but reduced to half of the value, assuming that on average 3 m out of 6 m is taken up by the tree row, mean carbon storage over 40 years would be 69 tC/ha, equivalent to 6.3 tCO2/ha/year (Burgess et al, 2022).

7. Semi-Natural Grasslands 

• The evidence on grassland management and carbon is both complex and incomplete. Semi-natural grasslands usually require grazing by livestock or cutting, and are an agricultural land use. The current management of a grassland site influence its flux status, but so does its historic management, which can exert a legacy effect many decades after a land management or land use change (Gregg et al, 2021). While many grassland studies focus on the topsoil, the zone where management actions interact with the soils and vegetation, considerable carbon stocks may be held below 15 cm depth.
• Many upland acid grasslands are likely derived from dwarf-shrub heath, blanket bog, or wetland vegetation, due to long-term over-grazing, burning and drainage. Hence a lot of Scotland’s peatlands are under agricultural use, largely grassland areas. Agricultural peat with the poorest capability also presents the greatest opportunity for emission reductions. Eroded peat, drained grassland and drained heather contribute the most to emissions. The relationship between livestock density and emissions from peatlands is not known (Aitkenhead et al, 2021).
• Soil carbon stores in upland acid grassland range from 237 to 282 tC/ha (average 240 tC/ha) for Molinia grassland and 99 to 823 tC/ha (average 337) in other upland grassland with Molinia (including communities with Nardus species and communities with bracken). There is little evidence on net GHG balance/carbon sequestration (Baggaley et al, 2021). Molinia tussocks are dense stores of carbon, with one study suggesting up to 6.85 tC/ha with no grazing, but this was reduced by half with 2.7 ewes/ha/year (Smith et al, 2014).
• Soil carbon stores in other semi-natural grassland are estimated to be 79.3 tC/ha in neutral grassland in Scotland (15 cm i.e. top soil), while not enough evidence is available for calcareous grasslands (Emmett et al, 2010). Values found in studies in continental Europe can reach 75 tC/ha but carbon stocks are likely to be lower in calcareous grasslands in Scotland such as machair which lie on shallow soils (Gregg et al, 2021).
• Additional carbon stocks gained through changes to management are vulnerable to loss, and physical disturbance of soils such as ploughing or re-seeding of pasture species can lead to rapid loss of soil carbon (Baggaley et al, 2022). Some long-established grasslands may be at equilibrium in which case additional carbon storage would be limited. However, it is important that these grasslands are not converted to another land use involving disturbance of the soil and emissions of carbon dioxide.
• Grazing has an impact on carbon sequestration and soil organic carbon over time. The balance of biomass removed by grazing animals on the one hand and the addition of manure inputs on the other hand will influence the soil carbon balance. The relationship between grazing and soil carbon balance is also influenced by the climate. Generally grazing stimulates pasture growth, but high grazing densities will have a negative impact on soil carbon content particularly in cool and humid climates. So, for example, in upland sheep grazing, no or low intensity grazing is necessary to enhance plant and carbon sequestration. Increasing grass productivity by adding more N fertiliser combined with higher grazing densities, can increase soil organic content, but also result in nitrogen losses (Abdalla et al, 2018). One study in the Highlands considered the impact of wild deer grazing on the topsoil between grazed and ungrazed plots and showed no difference on topsoil carbon; this result would need further studies to be confirmed (Warner et al, 2021).
• Most studies on the effects of grazing changes on grassland biodiversity and carbon stocks in the UK have been in degraded upland semi-natural acid grasslands dominated by mat-grass and purple moor-grass. These studies included sites in Scotland and in very similar habitats in northern England. In these grasslands, grazing removal or reduction increased above ground biomass and benefitted cover of dwarf-shrubs at the expense of grasses. Effects on soil carbon stocks were small and difficult to detect, but modelling suggests an increase (Baggaley et al, 2022).
• Mob grazing, (short duration high density grazing with a long recovery period for grass) has been proposed as a type of grassland management that will help increase soil carbon storage. In principle, the longer recovery period allows grass to grow at a faster rate thus increasing carbon capture; however, the evidence for this is still being gathered. Trampling has been put forward as a pathway to enhance the transfer of plant matter into soil matter although evidence for this is inconclusive (Oyesiku-Blakemore, 2022).
• Information on the effects of management on carbon sequestration in semi-natural grassland is limited, particularly for grasslands on organic soils and under Scottish climatic conditions. There are gaps in knowledge around the effects of management actions in calcareous and neutral semi-natural grasslands and in species-rich lowland grassland. Degraded permanent grasslands are likely to be net emitters (Gregg et al, 2021).
• Some studies point to carbon storage benefits from increasing plant species richness. High plant diversity has been shown to enhance soil carbon stocks by elevating belowground carbon (i.e., root biomass and root exudates) inputs and promoting microbial growth, turnover, and entombment of necromass (Bai et al, 2022). As degraded grasslands lose carbon, it makes sense that restoration of grassland cover and biodiversity is an effective strategy for promoting soil carbon storage. There is an evidence gap of the direct effects of increasing species-richness on carbon sequestration in Scotland’s grasslands (Baggaley et al, 2022).
• Carbon sequestration in grasslands are usually legacy effects from transitioning from previous land uses (e.g. restoration of degraded land, conversion from arable to grasslands) and the potential for sequestration cannot continue indefinitely. At some point, the soil will reach an equilibrium when no more carbon is sequestered (IPCC, 2018).

8. Freshwater

• Naturally functioning freshwaters produce their own sources of carbon through the net primary production of aquatic plants and algae. In addition, freshwater ecosystems receive carbon from land, some of which will be lost as gases while other forms of carbon will be transported downstream, some of which will be deposited in floodplains, some eventually reaching the ocean. Water-bound carbon loss occur as Dissolved Organic Carbon and Particulate Organic Carbon, as well as dissolved carbon dioxide and evaded carbon dioxide (emitted). Dissolved Organic Carbon (DOC) can be labile and be mineralised quickly leading to dissolved carbon dioxide while recalcitrant DOC resist mineralisation for much longer in freshwater systems and the oceans. Most terrestrial carbon transferred to freshwater is emitted to the atmosphere as carbon dioxide from the surface of rivers and other water bodies, while a smaller portion reaches the world oceans. In large rivers and lakes, residence time in water is long enough for most carbon to be emitted as carbon dioxide but in faster moving stream with short residence time, more carbon will be transported downstream though turbulence will also enable carbon dioxide to evade water (various authors in Vachon et al, 2023).
• Freshwater ecosystems form a key contributor to global methane emissions alongside anthropogenic sources of emissions (agriculture and fossil fuel industry mostly). A 2021 study found that aquatic ecosystems emit more than half of the global CH4 emissions Methane fluxes can be ebullitive, diffusive or plant-mediated (ebullition means bubbles of methane reach the atmosphere fast from sediments while diffusion means slower release at the water-air interface from dissolved methane). A metadata analysis of global relevance suggests that methane emissions increase from natural to modified aquatic ecosystems and from coastal to freshwater ecosystems (Rosentreter et al, 2021; Global Carbon Project, 2024).
• Freshwater ecosystems can also be a source of nitrous oxide due to incomplete denitrification or nitrification. Emissions of nitrous oxide will be controlled by nitrates in the water as well as organic carbon availability (Speir et al, 2023). Nutrient pollution from anthropogenic sources have been shown to favour in-situ processes of nitrification and denitrification in aquatic environments (various authors in Mwanake et al, 2024).

8.1. Floodplains

Carbon and carbon dioxide

• Floodplains arise from dynamic hydrogeomorphic processes which includes flooding, erosion, and sedimentation and which in turn interact with biotic processes to create diverse habitats (Steiger et al, 2005). Channel migration redistributes floodplain sediments and can result in bend abandonment by cutoff. This is responsible for both the storage and remobilisation of sediment (and associated carbon) across the channel and floodplain (various authors in Quine et al, 2022).
• Floodplains are a significant carbon stock but with large spatial variability among the different geomorphic units (channel-floodplain connectivity, connectivity with eroding slopes, variation in substrate, vegetation, topography, hydraulics) (various authors in Swinnen et al, 2019).
• Floodplains would naturally hold a mosaic of habitats including wet meadows, wet woodlands, bogs, fens, other wetlands, ponds as well as habitats on drier grounds, distributed according to natural hydrological pathways and floodplain microtopography. Natural active floodplains are essentially wetlands that exchange water, nutrients and sediments with their neighbouring rivers (Hoffmann, 2022).
• However, floodplains within the wider farmed and urban environments tend to be modified as a result of recent and historical drainage systems and land use changes (draining of wetlands, channelisation and urbanisation, disconnection of rivers from floodplains, longitudinal disconnection from dams and weirs, mills, straightening affecting hydrodynamics, removal of riparian woodlands). This also includes the construction of dikes that separates the river from the floodplain.
• Because of the variability in land cover and land uses in floodplains, and the significant spatial variability in sedimentation rates, it is difficult to provide an estimate for their value as a carbon stores and for carbon sequestration. This will depend on habitat composition and condition, whether the floodplain is still connected to the river, the proportion of cultivated land and land sealed for urbanisation. However research into carbon in floodplains highlight the very important role they can play in carbon sequestration if functional. As a result of stabilisation by fine soil particles and preservation by oxygen scarcity, soil carbon stocks have the potential to be very high in temperate floodplain grasslands and forests (various authors in Heger et al, 2024).
• While vegetation contribute to a floodplain’s carbon stocks, and sequestration occurs through living vegetation, most of the carbon is held in the soil/sediment and dead wood. Carbon accumulation is the result of litterfall and root exudates (a combination of allochthonous catchment material, and autochthonous vegetation sources), fluvial deposition, sediment residence time, soil respiration, level of water table. Anthropogenic sources include manure and slurries, soil erosion from farmed land, septic tanks (various authors in Wohl et al, 2022). Changes in the floodplain hydrology can affect autochthonous as well as allochthonous carbon input. Sedimentation of nutrient-rich material can boost the net primary productivity of the vegetation, if it is allowed to grow, which in turn supports autochthonic carbon input (Heger et al, 2021). In the deeper layers, buried organic rich sediments are a significant store of old carbon (> 1000 years to >10,000 years old) (Sear et al, 2023).
• In an analysis for England using existing geospatial data, the average soil carbon stock in English floodplains over the first 15cm is 56.1 (±13.8) tC/ha. The table below proposes estimates of the distribution of carbon stores in the UK floodplains.

From Sear et al (2023) for floodplains in the UK

• River management strategies that aim to reconnect channels and floodplains hydrologically are the most effective way to enhance carbon sequestration, as shown in comparisons of floodplain soil organic carbon stock in managed versus unmanaged river and floodplains (various authors in Wohl, 2022). A study of hardwood forests of the Elbe river suggested that carbon sequestration potential was up to 42% higher in the active floodplain than in the zone behind dikes, the latter may be due to disconnection from nutrient-rich flooding pulses (‘active’ means active floodplain whereby the habitat is still connected to the river). This suggests that dike relocation and de-embankment could increase carbon sequestration in trees and woodlands located on floodplains (Shupe et al, 2022). Another study from the Elbe showed that active connection to the river increase soil carbon sequestration potential, and soil carbon stocks were 33% higher in the active floodplain compared with the former floodplain (separated by dikes) (Heger et al, 2021). Wetlands connected to river channels in floodplains store high amount of carbon and research from the US shows the legacy of now disappeared wetlands on the carbon content of floodplain soils in the American prairies (Wohl et al, 2018).
• Topography in the floodplain influences soil organic content. A study of floodplains in the Elbe, where habitats were divided by hydrologic situation (elevation and exposure to flooding) and vegetation classes, showed that soils in low forests of the active floodplain stored 50% more carbon than their high counterparts. Low grasslands stored 63% more carbon than their high counterparts. Mean carbon stocks for all active forests were 124 tC/ha, and 116 tC/ha for grasslands. Soil carbon stocks in the low forests of the former floodplain (i.e habitats separated from the river by dikes) were 33% smaller than the low forests of the active floodplain (Heger et al, 2021). These are mineral soils. Where peatlands are present soil carbon stocks would be higher.
• In a modified floodplain in Devon, carbon accumulated at significantly greater rates in the active channel belt (river and levees) compared with the floodplain (4.7 ± 0.3 tC/ha/year compared to 0.9 ± 0.1 tC/ha/year). Decomposition rates were less in the active channel belt than in the floodplain, thus resulting in reduced carbon loss. The authors suggested that by increasing river channel complexity, the active floodplain area could increase thus sequestering more carbon (Quine et al, 2022).
• Disconnected rivers result in an increased degassing of carbon dioxide due to exposure to air, increased respiration and breakdown of organic material. Storage depends on the maintenance of anoxic conditions. Management practices that stabilise river channels and / or disconnect them from the floodplain reduces the potential for carbon sequestration by reducing transit times. Conversely, soil erosion due to agricultural practices and run off, including organic deposits from eroding peat, have resulted in an accumulation of sediments in floodplains (a 2500% increase compared with the Mesolithic), some of which will be mineralised and some of which will be stored longer-term. Floodplains have become a repository of old carbon from eroded soils in the water catchment while erosion followed by burial of modern carbon will result in avoided emissions (this is not to suggest that erosion due to land management is a good thing, but rather than the soil inevitably lands somewhere, and might end up being buried). The consequence of erosion and loss of carbon to act as a source or sink of atmospheric carbon depends ultimately on the preservation of carbon that is buried in floodplain deposits, or whether it is lost through soil respiration, or exported from the catchment (Hoffmann, 2022). Floodplains can therefore switch between being a carbon ‘sink’ and an emitter depending on land cover and management. While floodplains’ soil carbon is affected by upstream sources of organic matter, and burial of eroded sediment results in avoided emissions and provide stability over long timescales, it is the biomass on floodplains in the conditions provided by restored / natural floodplain environments which result in carbon sequestration by drawing carbon dioxide from the atmosphere (Craft et al, 2024; Sear et al, 2023).

Adapted from Sear et al (2023) * former floodplain - these are not net carbon sequestration estimates and do not account for oxidation, mineralisation and biological activity in organic matter all of which contributes to emissions of carbon

• Mean carbon storage across 78 transects in the mountain headwater floodplains of the Upper Dee in Scotland is reported as 323.27 (±12.58) tC/ha, which is a bit below other average carbon stocks found in temperate floodplains. Floodplains in the Upper Dee showed great variability in soil carbon storage depending on the hydromorphic units and floodplain type. The floodplain soil carbon stocks ranged from 1.26 to 3091.99 tC/ha, with carbon stocks at the higher end being linked to peat layers and low energy floodplains with an anastomosing channel. This is because peat formation is favoured in a floodplain where there is limited stream power (and the climatic conditions are suitable). On the contrary, where extreme events carry hillslope sediments to the floodplain resulting in braided channels and channel mobility, it is more difficult for peat to be formed. The Upper Dee also includes high and medium energy floodplains, which store much less carbon. Despite the presence of peat, the average carbon stock was lower than the average temperate floodplain due to the shallow depth of deposits and lesser erosion within the catchment. Within the study area in the Upper Dee, there had been no direct anthropogenic disturbance of the river channel, hence the floodplain can be considered active. The authors argued that mapping of the geomorphic floodplain type can be an effective technique to estimate the floodplain sediment and soil organic carbon storage - see also Nanson et al (1992) for further information on floodplain types) (Swinnen, 2019). We may speculate that this situation may be similar in other upland floodplains in Scotland.
• In a study of six river floodplains in Southern England, with a mix of upland and lowland floodplains, and main land use is permanent pasture, carbon sequestration rates ranged between 2.5 to 4.2 tCO2/ha/year. The study did not consider connectivity to the river (Gregg et al, 2021). A study of flooded woodlands in the Danube showed that while soil carbon dioxide emissions stop during inundation in regularly flooded areas, these emissions increase and are higher than in more elevated areas (not flooded) at drier periods. It is speculated that the higher respiration rates are due to the abundance of organic matter brought during the flood event and access by heterotrophic decomposers in the dry period (Schindlbacher et al, 2022).

Methane and nitrous oxide

• Wetlands are an integral part of the natural global methane cycle. Floodplains, when flooded, emit methane like all wetlands. However, in healthy wetlands, the benefits of carbon sequestration outweigh methane emissions’ contribution to net GHG emissions (Evans et al, 2023). Linkages between the inundated parts of a floodplain and the river will bring methane to the river channel (various authors in Bussman, 2022).
• The level of the water table as it fluctuates throughout the year drives methane emissions alongside vegetation type. The duration has an impact on methane emissions, with observations in floodplain forests that experience only short burst of flooding (few days) showing them to remain methane sinks (Jacinthe, 2015). The vegetation could however have an impact on pulses of methane from inundated areas in floodplains.
• A study in Southern England based on a mesocosm experiment, and field measurements in a wet meadow, showed that methane emissions were higher at high water table (15 cm), which seems to be an approximate threshold for methane emissions. Inundated periods led to high emissions which overrode the non-flood period methane sink. Plants have a significant influence on methane emissions by acting as a conduit for the gas, and the level of emissions will vary with plant assemblages; plants with aerenchymatous tissue in particular can act as chimneys. Methane exchange also appears to respond to soil temperature with emissions being notably higher at temperatures ≥19C. The maximum methane emissions rate observed was 160 mgCH4/m2/d, while the average in vegetated mesocosm was 3.47 ± 0.8 mgCH4/m2/d or 0.013 ± 0.003 tCH4/ha/year (0.39 ± 0.09 tCO2e/ha/year) (Peacock et al, 2024).
• Nitrous oxide from natural and anthropogenic sources of nitrogen is another greenhouse gas that might be emitted from riparian zones in floodplains when inundated. Sediment heterogeneity seems to control the location and magnitude of emissions of nitrous oxide production as a byproduct of nitrogen cycling within floodplains. Organic matter and increased DOC seem to facilitate nitrogen transformations and production of nitrous oxide (Wallace et al, 2021).

8.2. Freshwater wetlands (other than peatlands)

Carbon and carbon dioxide

• Some wetlands are on deep peat (= peatlands), some on shallow peat, some on mineral soil and some wetlands contain areas of both peat and mineral soil with a mosaic of wetland communities. Over the past 300 years about 21% of global wetlands have been lost due to climate change and anthropogenic disturbances (Fluet-Chouinard et al, 2023). There is some empirical data on carbon budgets for different types of constructed wetlands, which show that they can be both carbon sources and sinks. The consensus is that this is highly dependent on the type of wetland and on its management. The review below is only concerned with natural wetlands.
• Wetlands (freshwater and marine) contain 20-30% of the total soil carbon pool (~535 Gt), a disproportionally large amount given their areal coverage is only 5-8% (7 million km2) of land surface. In the UK, the total areas for wetlands, excluding peatlands, is not known and there are no estimates for carbon in wetlands other than for peatlands (Chambers et al, 2024).
• Freshwater (inland) wetlands contain significant carbon stocks but have been highly susceptible to disturbance and conversion to farmland, making them at risk from losing their carbon stocks and capacity to sequester carbon. The net cumulative potential of freshwater mineral wetlands, such as freshwater marshes, for climate change mitigation may be underestimated given the lack of data on organic carbon stocks in deeper soils. The recovery of stored carbon stocks in restored wetlands is highly variable and driven by factors related to landscape and land use. A study in the Lake Erie Basin, Southern Ontario, Canada, found that the average soil carbon stocks at a 0-30 cm depth for actively restored site soils was 16-115 t C/ha. Underlying parent material was the strongest driver of soil organic carbon (clay and silt: 107 ± 9 t C/ha, sand: 39 ± 19 t C/ha). The mean short-term rate of organic carbon accumulation was 5.285 tCO2/ha/year (range of 1.615-10.643 tCO2/ha/year), rates were highest shortly after restoration (1964) and gradually decreased over time (Loder et al, 2023).
• Wetland plant biomass differs across climate zones and wetland types. A global study revealed that below ground biomass in equatorial, temperate, arid, snow and polar regions ranged from 14.70 – 18.07 t C/ha. As plant biomass increases in wetlands, the soil organic carbon pools also increased, in temperate deltas the soil organic carbon pool content was between 5 – 40 t C/ha at above ground biomass levels of 2.5-7 t C/ha, and suggesting that above ground carbon was a relatively inefficient carbon source for generating stable organic carbon. Changes in below ground biomass had a greater impact than above ground biomass in influencing soil organic carbon, when below ground biomass was below 1 tC/ha the wetland organic carbon pool increased by 5.7 tC/ha (Pan et al, 2024).
• Wetland carbon flux can be strongly positively influenced by high levels of rainfall (increased wetland CO2 flux with high soil moisture), compared with much smaller variations in carbon fluxes during periods of no or little rain (Ouyang et al, 2021).
• Samples taken from 33 drained wetlands in Saskatchewan, Canada found that following drainage implementation there was a loss of organic carbon and annual carbon sequestration rates were negative across all sampling depths (i.e. the wetlands were net GHG emitters) (Chizen et al, 2024).

Methane and Nitrous Oxide

• The partitioning between carbon burial and GHG emissions (CO2, N2O and CH4) is key to whether a wetland acts as net emitter. Anoxic conditions in sediments favour microbe-mediated methanogenesis while eutrophication will favour the development of anoxic conditions increasing methane production. Plant species also play a role whereby plant that allocate more oxygen to their rhizosphere will impede methanogenesis; plants also provide an emission pathway with methane moving from sediments to the atmosphere through plant tissues. The alternance between dry and wet phases in wetlands means that there is seasonal variability in emissions to the atmosphere (Malerba et al, 2022).
• Natural wetlands (freshwater and marine) are the largest natural source of atmospheric methane contributing to 30-40% of the total global emissions. Freshwater wetlands emit more methane than saltmarshes. This is due to salinity and the presence of sulphate reducing bacteria in saline waters that utilise the primary substrates of methanogens. Converting freshwater wetlands back to saltmarshes by restoring tidal flows where these have been restricted (e.g. embankments) could reduce methane emissions while increasing carbon sequestration (Rosentreter et al, 2021; Arias-Ortiz et al, 2024).
• A study of a freshwater, estuarine marsh, Lake Erie, Ohio, USA found that on average the cove acted as a net carbon dioxide and methane source, primarily driven by water inundation (Hassett et al, 2024).
• Variations in wetland methane emissions is suggested as the primary driver for observed interannual variability in global methane fluxes. A driver of methane emissions has been linked to water-table fluctuations but this has high variability (i.e. fluctuations do not consistently increase with water table rise but appear to be tied to a critical threshold, below which there is correlated increase and above which there is high variability) and is highly dependent on the specific site (Cui et al, 2024).
• In wetlands, most methane and 70% of carbon dioxide efflux originate from the decomposition of organic matter by soil microorganisms. A study of 3022 paired carbon dioxide and methane observations from 159 field sites worldwide found that emissions rise exponentially with seasonal temperature. Decreased amount of C and N in wetland soils, through anthropogenic activities was found to increases SOM decomposability which then increases carbon dioxide and methane emissions. C and N soil content was found to be a better predictor of emissions than water-table depth. Therefore, increasing stable soil organic carbon components is likely to decrease emissions from wetlands (Hu et al, 2024).
• The anaerobic environment of wetlands favour methane production by soil methanogenic microorganisms. A study of data from 38 wetland sites across 160 years taken from the FLUXNET-CH4 dataset (Global Carbon Project) found that multiyear methane emissions from all sites averaged 1.035 t CO2/ha/year (min: 0.025, max: 5.947). When the water-table was between 0.4 and 0.8 m the average annual methane emissions increased to 3.034 t CO2/ha/year, the largest rates of emissions. Emissions from wetlands dominated by vascular plants was greater and more variable than emissions from moss dominated wetlands (Li et al, 2024).
• Wetland ecosystems are more susceptible to and affected by invasive plants. Although presence of invasive species increased methane emissions in wetlands in China, this impact was not significant in freshwater wetlands. Nitrous oxide fluxes were also not impacted by the establishment of invasive species in freshwater wetlands (Bezabih Beyene, 2022).
• A study of 800,000 km2 of freshwater wetlands in north-central USA and south-central Canada found differences in average nitrous oxide fluxes depending on land use and drainage (see table 3 below). These differences are due to agricultural soil management and the use of nitrogen fertilizers. A greater proportion of cropland wetlands were dry compared to grassland wetlands, due to differences in management conditions, which could account for the increase in N2O emissions due to the combination of nitrogen additions and dry conditions (Tangen et al, 2022).

8.3  Ponds

• Ponds are considered greenhouse gas hotspots, with higher GHG fluxes than for example lakes. This may be due to higher temperature, higher shoreline: surface water ratio, with greater accumulation of organic matter. A study in Denmark showed variations in emissions of greenhouses gases during the day and an increase between winter and summer of methane formation in the sediments (6 fold in forest ponds and 4 fold in open landscapes). Because of the GHG emissions on the one hand and carbon burial on the others, the contribution of ponds to climate mitigation has been open for debate. The authors argued that ‘’the debate of pond construction as a climate mitigation strategy should be evaluated as part of carbon budgets for the entire hydrological network at relevant time scales’’ (Sø et al, 2024).
• Having multiple ponds in the landscape is important for freshwater connectivity as they offer stepping stones between freshwater habitats. Through this connectivity ponds have an influence over the carbon cycle. Emissions include nitrous oxide and methane which can be emitted through diffusion or/and ebullition while carbon accumulates through allochthonous and autochthonous organic matter. They can be net emitters depending on the balance between carbon burial rates and GHG emissions. This will depend on surrounding land use, levels of eutrophication, vegetation, temperature, length of time the pond is covered in water. There is high temporal variation which means it is difficult to propose estimates on net emissions (Cuenca-Cambronero, 2024). A study from Minnesota on urban ponds showed that with stratification of water in ponds, methane that form in anoxic conditions is rapidly exchanged with the atmosphere unlike in deeper systems (Rabaey et al, 2024).
• A review of studies relating to 55 ponds in England (from North to South) suggests a burial rate of 1.42 tC/ha/year (which in carbon dioxide would translate to 5.21 tCO2/ha/year) (Jeffries et al, 2023).

8.4. Rivers and streams

Carbon and carbon dioxide

• Quantification of GHG emissions from rivers is difficult due to spatial and short-term temporal variability. Quantification of GHG fluxes typically relies on measurements of GHG concentrations in a limited number of water samples taken along the river. The fluxes in streams and rivers are large relative to their surface area, hence they are hotspots for the exchange of gases with the atmosphere (Gu et al, 2021; Koschorreck  et al, 2023).
• Rivers carry multiple forms of terrestrial carbon, but the dominant forms are particulate organic carbon (POC) and dissolved organic carbon (DOC) occurring through erosion and overland flow, and direct input from trees and other riparian vegetation. Rivers can deposit large amounts of sediment in floodplain soils that then act as a carbon sink. Inorganic forms of carbon also enter river systems from the weathering of bedrock and mineral precipitation. The cool temperate climate of the UK can favour high organic carbon concentrations in floodplain sediments, particularly in catchments fed by peatland headwaters. Evidence shows that carbon transported through rivers to the oceans is a fraction of the carbon entering rivers from land. This suggests that rivers are a net source of carbon dioxide, as a proportion of terrestrially derived organic matter is mineralised during transport (Gregg et al, 2021).
• However the proportion will vary between types of rivers, and cumulatively, there are uncertainties regarding how much DOC is lost to the atmosphere from land to sea. DOC may also be generated within rivers, via photosynthesis, by desorption of DOC from suspended sediment and in anthropogenic inputs including wastewater and agricultural leaching and run off (Tipping et al, 2022).
• Lateral carbon fluxes from the land into rivers offset some of the carbon accumulation in terrestrial ecosystems. Besides being a natural process, land management and changes in land use, wastewater discharge and anthropogenic hydrological modifications also have an impact on these riverine carbon exports. Typically evaluations of carbon accumulation on the land do not take account of carbon exported to rivers. A modelling study applied across Europe found that on average, 14.3 (±10) MtC/year of DOC is leached from land into European rivers, which is about 0.6% of the terrestrial net primary production (NPP). On average, 12.3 MtC/year of the leached DOC is exported to the coast via the river network, and the rest is respired during transit. The model did not consider peatlands, hence is an under evaluation (Gommet et al, 2022).
• In another study, it was found that rainfall and organic soils are the most important drivers affecting riverine DOC export, and it is argued that under-representation of smaller river systems draining peatlands could lead to underestimation of the DOC export from land to the ocean. In Britain, the main anthropogenic factor influencing the spatial distribution of DOC exports appears to be upland conifer plantation forestry, and this is amplified on peat soils, which is estimated to have raised the overall DOC export by 0.17 MtC/year for the whole of Britain (about 15% of current estimated net C uptake by growing trees in the UK.). Urban systems can also impact the fluvial carbon cycle, and a study on the River Kelvin Scotland suggests that the concentration of DOC is likely influenced by sewage overflows with a weak relationship with summer and autumn hydrology, suggesting a potential link between urbanization and riverine carbon export. Assuming a worst case scenario where all DOC was mineralised and CO2 degassed, this would represent about 1% of UK GHG emissions. In reality, a proportion of DOC is also likely to be buried in freshwater and marine sediments through flocculation, while another proportion of mineralised DOC may enter the stable ocean DIC pool (Williamson et al, 2021; Gu et al, 2021).
• River carbon dioxide fluxes are primarily controlled by the gas exchange velocity at the water-air interface and the gradient between the water and atmospheric partial pressures of carbon dioxide. There is a considerable diurnal variation in fluxes of carbon dioxide with higher fluxes at night due to the impact of light on the balance between photosynthesis and respiration in aquatic plants (Attermeyer et al, 2021). A study over 1 km stretch in the Elbe River in Germany in the summer resulted in the following measurements for carbon dioxide: 2.2 (1.86–2.21) μmol/m2/h in the middle to 1.5 (1.07–3.13) μmol/m2/h on the side of the river (this is equivalent to 0.008 (0.007-0.008) tCO2/ha/year and 0.006 (0.004-0.0120) tCO2/ha/year (Koschorreck et al, 2021).
• Evidence from 80 catchments in the UK shows that one tonne of POC entering rivers gives a median emission factor of 5.5 tCO2/year. POC entering rivers could only become a carbon sink if re-accumulation from atmosphere to land was high and the erosion rate low (Gregg et al, 2021).

Methane and nitrous oxide

• There are multiple climatic, biological and geomorphological drivers of methane emissions from streams and rivers methane emissions that operate across land–water boundaries. Landscape processes will have a bigger influence in rivers than standing waters, such as at high latitudes, methane emissions from rivers being associated with the presence of large stores of organic carbon in soils and the hydrological connection between these sources and the river. Pollution from agricultural and urban land increases methane emissions from rivers (Rocher-Ros et al, 2023).
• There are large spatiotemporal variabilities that make quantification of emissions difficult. Any methane input to a stream from riparian habitats is lost to the atmosphere within a short distance. A study involving sampling at multiple points along the Elbe river (laterally and longitudinally over 580 km) in the summer indicates the river is a source of methane emissions with the means estimated at 251 μmol/m2/d (equivalent in our calculations to 0.015 tCH4/ha/year or 0.43 tCO2e/ha/year) and a range from 66 to 3709 mol/m2/d in hotspots. Weirs created key methane hotspots and they affect both upstream and downstream river; the trapping of sediments upstream results in extreme methane emissions from both ebullition and diffusion. Taking into account the surface area affected by a weir tends, the contribution from dams to a river’s methane emissions can be disproportionately high relative to the length of the river (Bussman et al, 2022).
• In contrast to carbon dioxide, there is not diurnal variation in methane emissions, but greater spatial variation. Because of the temporal and spatial variability across a stretch of river between day and night and the middle and the sides of the river, total emissions in another Elbe River study were very different from what the emissions per square metre were suggesting, with methane only a small fraction of carbon dioxide emissions (Koschorreck  et al, 2023).
• Point source pollution in urban environments is an important cause of methane emissions from rivers, with a study along the Clyde river showing methane concentrations in water 20 times higher in the urban environment compared with samples taken upstream where semi-natural habitats dominated. Low oxygen levels, higher residence times and higher temperatures in summer also increased methane concentrations by stimulating methagenosis. Diffuse pollution from agriculture was the main source of nitrous oxide in the river though urban pollution also played a role (Brown et al, 2023).
• Nitrous oxide emissions from rivers have been increasing due to pollution. Nitrates concentration control the production of nitrous oxide, however it is not the only driver. Carbon availability and temperatures are other drivers but the interaction between these is not yet well understood (Speir et al, 2023).

8.5. Headwater streams draining peatlands

• Streams and rivers draining peatlands are saturated with CO2 and contain high concentrations of dissolved organic carbon (DOC); these are often associated with large lateral (downstream) and vertical (evasion) fluxes i.e. gaseous emissions from surface water, which may produce significant changes in the sink/source relationships of individual peatlands. Largely, dissolved organic carbon export from soils to rivers is natural, and thus an intrinsic component of the Earth’s carbon cycle. However, human activities such as agriculture and changes to drainage systems have increased the fluxes of DOC (Williamson et al, 2021).
• Aquatic carbon fluxes will be highest from drained peatland compared with non-drained. For example, DOC in highly degraded peatlands in the Southern Pennines reached a DOC flux of 0.30 tC/ha/year, while typical values tend to <0.10 tC/ha/year. There is deeper carbon loss (i.e. from older layers) in degraded peatlands than in intact peatlands (Evans et al, 2022). In the Flow Country, aquatic carbon fluxes were highest from the drained upstream catchment and lowest from the non-drained upstream catchments at 0.86 and 0.03 tCO2/ha/year respectively, with variability between the upstream and downstream sites within each catchment being very low. The same study showed much greater variability in the restored catchment (0.37 to 0.77 tCO2/ha/year) (Pickard et al, 2022). A meta-analysis on the effects of rewetting peatlands showed rewetting seems to have uncertain effects on DOC concentration, in view of the varied effect sizes in the studies, which were not considered significant (Darusman et al, 2023).
• Estimates of evasion (degassing) of carbon dioxide and methane from the water surface to the atmosphere is less studied. Gases are lost from the aquatic pathway as surface water gas concentrations progressively equilibrate with lower concentrations in the atmosphere. Isotopic evidence from four UK peatlands suggests that whilst the age of DOC released in the drainage system of peatlands is consistently young, the age of carbon dioxide lost by evasion from the water surface seems older. This suggests that the release of carbon dioxide into the aquatic system is related to significantly older carbon pools than for DOC (Billett et al, 2007).
• A study undertaken at multiple sites across the UK suggests that while DOC tend to be from recent decomposition of plant matter, evasion carbon dioxide has also other origins including geogenic sources and old biogenic sources (deep peat layers), and a significant proportion of carbon dioxide lost by evasion is not derived from within-stream breakdown of DOC. Their best current estimate for the size of the evasion flux term from UK peatland headwater streams was 0.85 (±0.25) tCO2/ha (catchment)/year. The aquatic and land–atmosphere systems are not synchronised in terms of carbon fluxes and turnover, and there may be significant lags in the components that make up the carbon balance of a peatland, with parts of the carbon cycle responding at different rates to change (Billett et al, 2015).
• A study from Quebec in an undrained peatland suggests that 86.5% of the dissolved carbon dioxide and methane present in peatland headwater stream was emitted to the atmosphere within the catchment boundary rather than being exported downstream. About 81% of carbon dioxide present in the stream was from porewater lateral discharge (i.e. from peat soil) while only 17% was from in-stream production (i.e. within stream break down of DOC) (Taillardat et al, 2022) which is in line with the UK study above. 

8.6. Peatland pools

• Pools act as recipients for dissolved and particulate organic carbon fed either through or over the surrounding peat. Fluxes of methane and carbon dioxide are lacking for the UK. Evidence seems to suggest that carbon dioxide and methane losses are smaller in deeper pools compared with shallow ones. It also appears that vegetation cover has an impact with Sphagnum mosses associated with lower methane fluxes than Eriophorum cotton grass cover (Gregg et al, 2021). Research in the Flow Country suggests DOC was on average larger across all sites and sampling times in restoration pools than in natural pools. All pools were supersaturated with dissolved methane and carbon dioxide, with carbon dioxide concentrations 10 times higher in restoration pools than atmospheric levels while those in natural pools were typically only just above atmospheric equilibrium. This could be due to higher soil respiration in the surrounding peat, lesser area of open water, shorter water residence times and lesser depth. Hence, there could be a peatland degradation legacy from drainage inherent in the biogeochemical functioning at restoration sites. Where natural pools are present within a catchment, they may play an important role in controlling downstream aquatic carbon chemistry (Chapman et al, 2022).

8.7. Freshwater lochs

• The sediments of lakes globally contain more organic carbon than the entire rest of the terrestrial part of the biosphere. But the accumulation rate is slow, and lakes are a modest global source of carbon dioxide into the atmosphere. Inputs to lakes’ OC budgets include allochthonous (externally derived) dissolved and particulate OC, (DOC and POC respectively) carried in inflowing surface and in groundwater sources. Also included in these inputs are allochthonous inputs from precipitation, and wind-blown litter fall from the lake shore. Organic C budgets may also include autochthonous (internally derived) DOC and POC from organisms such as bacteria, phytoplankton, and zooplankton. Most DOC though appears to be from terrestrial origin. There are three possible fates for this OC within the lake: mineralization to inorganic carbon, sedimentation to the lake’s bed, and export via surface water. The DOC may tend to flow through the system and out to the ocean. Heavier carbon particulates, brought in by the rivers in suspension settle out to contribute to sediments in the lake. POC export rates vary enormously in peatland catchments (Prairie et al, 2022; Doyle et al, 2023).
• Carbon deposition in lake sediments is a long-term process that is part of both the fast and the slow carbon cycle. Excess matter is deposited and permanently stored in bottom sediments, while in the fast carbon cycle, carbon moves through the biosphere (Skwierawski, 2022). The balance between carbon sedimentation (carbon reaching the sediments) and carbon burial (carbon remaining permanently buried) is variable; overall the larger the ecosystem size, the smaller the burial rate (Prairie et al, 2022).
• A study in Estonia across multiple lakes suggests that soil type is a key variable in determining DOC, and in particular peatlands are a major source of DOC in lakes in regions with peat soils (Sepp et al, 2022).
• Carbon dioxide exchange occurs at the surface water while lakes are also large emitters of methane which can be emitted through ebullition (bubbles arising from sediments) and diffusion at the water surface. Numerous variables can have an impact on methane and carbon dioxide fluxes including trophic status, lake productivity, depth, shape, oxygen availability at the sediment surface, temperature, wind speed, wave turbulence. Methane emissions tend to increase with lake productivity. In one study, wind speed was key to explaining variation in eight of 13 sites (Golub et al, 2023). Spatial and temporal (diel, seasonal, annual) variation is high within and between lakes, and there are no reliable estimates of carbon storage and GHG fluxes, partly due to inconsistent sampling protocols (Sieczko, 2020; Sø et al, 2023).
• The flux of carbon dioxide is a significant and underestimated component of regional and global carbon cycles. Lakes mostly behave as net carbon sources and are super saturated with CO₂ in the surface waters, but lake sediments store a large amount of organic carbon over the long term; only a minority of lakes are net sinks for carbon dioxide, largely eutrophic ones. An analysis of lake sediments records in the UK showed a carbon accumulation rate of 7.4±5.5g C/m2/year during the Holocene (equivalent to 0.07 tC/ha/year or 0.26 tCO2/ha/year). These numbers which tend to be more typical of oligotrophic lakes are very significantly lower than accumulation rates in for example agricultural ponds, so estimates from lakes cannot be applied to other standing waters. The balance between evasion through carbon dioxide and sedimentation will depend on the source and type of organic material. In oligotrophic lakes, carbon dioxide diffusion at the air-water interface will dominate whereas sedimentation and burial will dominate in eutrophic lakes. Burial rates will be higher in deeper lakes than in shallow lakes where degradation goes faster (Du et al, 2023).
• Surface waters:
  • Dissolved organic carbon concentration in lakes varies broadly from 1-300 mg C/L although most lakes fall below 30 mg C/L. This carbon is central to regulating lake pH and alkalinity. There are only 3 possible fates for aquatic carbon: fluvial outflow, outgassing of CO₂ and sedimentation. The importance of each of these pathways are highly variable and not fully understood. In oligotrophic lakes, CO2 diffusion at the air-water interface will dominate while sedimentation and burial are more important is eutrophic lakes (Prairie et al, 2022).
  • A study of 30 ponds and shallow lakes (Europe and North America) found that on average waterbody CO₂ dissolved gas (pCO2) concentration was 7 times higher than in the air, indicating they were net CO₂ sinks. An increased pCO₂ was correlated with a smaller surface area, a decreased perimeter, less floating vegetation, less emergent plant cover and less dissolved phosphorus (Ray et al, 2023).
• Sediments:
  • Sediments in lakes contain more organic carbon than the entire rest of the terrestrial part of the biosphere, with a concentration of 3-30% of sediment weight, however, the accumulation rate is slow (Prairie et al, 2022).
  • Lakes are a long-term carbon sink and can store carbon for millennia and therefore should not be overlooked as a carbon sink. Dissolved inorganic carbon is consumed as part of photosynthesis by heterotrophs and oxidised into dissolved and particulate organic carbon (Du et al, 2023).
  • Organic carbon is sequestered 0.5-0.6m below the surface. This process is highly dependent on turbulence, which affects the ability to mineralize terrestrial dissolved carbon, and increases with increased water residence time, increased temperature, eutrophication and, potentially, in smaller lakes (Vachon et al, 2023).
  • Draining lakes can negatively impact carbon sequestration. A study of 8 eutrophic to hypertrophic lakes in Poland demonstrated that after being restored, they accumulated 2.75 tC/ha/year on average, which is equivalent to 10.11 tCO2/ha/year removed from the atmosphere (Skwierawski, 2022) and represent a high estimate.
  • Other studies suggest burial rates of 1 tC/ha/year in eutrophic lakes and 0.125–0.545 tC/ha/year in non-eutrophic lakes (Gregg et al, 2021).

Methane

• Methane can be emitted from lakes to the atmosphere via different pathways, including diffusion, ebullition, release from storage, and flux mediated by emergent plants. The main pathways are diffusion and ebullition which exhibits high spatiotemporal variability. Diffusive flux depends partly on the methane concentration gradient between the water and the air, and the turbulence in the surface water which in turn influences gas transfer rates. Ebullition is highly dependent on methane production in the sediments and any factors triggering the bubble release such as pressure, wind, temperature, precipitation or radiation – these occur on very short timescales contributing to temporal variability (Sieczko et al, 2020).
• Methane is produced both in anoxic conditions and in the oxygenated water layers (Ordóñez et al, 2023). The accumulation of methane in oxygenated waters is known as the ‘methane paradox’, as methanogenesis is usually expected in anoxic conditions. The production of methane in oxic conditions seems linked to aerobic production by cyanobacteria, phytoplankton and zooplankton. Transport from benthic zone and dissolution of methane bubbles is another explanation. A study from a lake in the Alps suggests that accumulation of methane in oxygenated pelagic waters is the result of an interplay between physical and biogeochemical processes, and these processes are likely to be lake-specific (Bartosiewicz et al, 2023).
• A study of a large lake (> 500km2) in Sweden found a significant negative relationship between surface area and average surface water methane concentration, which is in agreement with other studies that find inverse relationships between lake area and methane emissions (Peacock et al, 2023).
• A study in northern Germany of a medium sized 70m deep lake showed that methane concentrations in sediments increased from the littoral to the intermediate to the deep zones, with further variability depending on sediment depth of sampling. The concentration gradient in the sediment result in diffusive methane fluxes from layers of high to low gas concentrations in sediments and at the interface with water column. Methane moves through the water column through both diffusion and ebullition. Other studies showed that methane ebullition from sediments in larger lakes tend to be higher on the littoral while the opposite can be observed in small shallow lakes. High methane concentrations in sediments can be expected where there are high organic inputs, with refractory humic substances at depth and fresh organic input that is more labile at the littoral. However heterogeneity in spatial structure and resuspension can affect these outcomes. Most methane is likely to be oxidised in the water column, particularly at depth, and methane fluxes to the atmosphere seem linked to methane dynamics in the upper water layer. In the German study, methane concentrations in the surface water were similar at the various sampling sites, suggesting a decoupling from the production sites in the sediment (Li et al, 2021).
• Lakes with high organic matter have higher methane fluxes than lakes with low organic content. In a Danish study, ebullition rates were highest in deeper, hypoxic water (5–7m) while diffusive methane fluxes were 4-fold lower and spatially less variable than ebullitive fluxes. Diffusive methane flux showed higher diel (day/night) variation than the ebullitive methane. The high spatiotemporal variability in lakes challenges the identification of drivers of carbon dioxide and methane fluxes and make upscaling findings very difficult (Sø et al, 2023).
• Based on a dataset containing estimates of diffusive and ebullitive methane fluxes, a review of over 200 lakes at different latitudes suggested depth and surface area are the greatest predictor of methane fluxes, while chlorophyll was not a good predictor (Deemer et al, 2021). Methane ebullition tends to be favoured by the availability of easily degradable organic matter. A study on a small lake in Germany suggested that spatial patterns of methane ebullition may be different between deep and/or large lakes and small shallow lakes. Overall there is a poor understanding of the drivers of observed spatial variability of methane ebullition in lakes (Praetzel et al, 2021).
• The contribution of shallow lakes and ponds to global emissions are poorly understood, but could be high due to their ubiquity, and evidence of high emissions relative to their size compared with larger standing waters ecosystems. Methane emissions tend to be higher in the centre of small ponds and shallow lakes than at the littoral. There is a lack of understanding of the spatial and temporal variability in those small aquatic systems. A study of 30 small standings waters in Europe and northern USA suggests that partial pressure of methane and carbon dioxide is inversely correlated to size and positively correlated to floating vegetation cover (Ray et al, 2023). Methane flux tends to increase in the summer months and sediment temperature seems to be more important than surface water temperature, and that temperature sensitivity is more manifest in small and shallow lakes (Schmiedeskamp et al, 2021).
• Intense vertical mixing during autumn overturn redistributes the dissolved methane, which accumulates in anoxic deep water over the summer months. This process leads to enhanced methane concentrations at the water surface and results in increased diffusive methane fluxes, which can contribute up to 80%to the annual diffusive methane emissions. During the overturn, studies indicate that between 54% and 94% of the methane is oxidised. There is however large variations in methane storage in the anoxic layer and annual emissions between years (Ragg et al, 2021). A study of small lakes in Canada suggests that oxic methanogenesis is an important component of methane emissions in small lakes (Thottathil et al, 2022).
• Research in Lake Geneva and the Rhône River, which is one of the main tributaries to the lake, suggests that tributary river discharge during the summer season may provoke internal waves and stir up sediments, promoting mixing, and influencing the release of dissolved methane in upper layers of the lake up to several kilometres across the lake (Khatun et al, 2024).
• Methane-oxidizing bacteria (methanotrophic bacteria) play an important role in lake methane budgets and carbon cycling, as well as mitigating methane emissions to the atmosphere. These bacteria use methane as source of metabolic energy and structural carbon, producing biomass and carbon dioxide from the oxidation of methane. A study of 6 temperate lakes in Québec suggests a positive relationship between methanotrophy in lakes and DOC. Less light penetration in lakes with high DOC can contribute to shallower thermal stratification and a greater portion of the water column dominated by methanotrophy. This relationship may not verify in lakes with low DOC, or where DOC quantity and water colour do not correlate. The bulk of aquatic bacteria in lakes that grow on molecules of DOC are known as heterotrophs and they also produce biomass and carbon dioxide (Reis et al, 2022).
• Overall, there is a range of mechanisms underpinning methane production and emissions from lakes, oxidation and various transport pathways, which are also interacting (Schroll et al, 2023). There are lots of uncertainties regarding emissions from lakes, and the representativeness and utility of observed flux observations seems at this point limited (Johnson et al, 2022).
• In lakes and ponds, microbial denitrification in sediment is understood to be the main nitrous oxide production pathway. A mesocosm experiment showed a significant positive effect of nutrient enrichment on nitrous oxide concentrations and emissions (Audet et al, 2024), thus pointing to the role of nutrient load from agriculture and sewage in increasing nitrous oxide emissions.
• A global modelling study suggests that there had been a 133% increase in nitrous oxide emissions from small lakes in the period 1850s-2010s, and this is approximately twice that of large lakes. In the 2010s, agricultural nitrogen was the primary factor responsible for enhancing nitrous oxide emissions from lakes and ponds in Europe. Emissions range from 30.6 ± 5.0 mgN m−2 year−1 (0.08 ± 0.01tCO2e/ha/year) down to 17.6 ± 3.8 mgN m−2 year−1 (0.05 ± 0.01 tCO2/ha/year using 273 factor) and are inversely proportional to lake size. By comparison, in the 1850s, the nitrous oxide emissions per unit area from small lakes was 13.2 ± 0.1 mgN m−2 year−1, and 10.4 ± 2.9 mgN m−2 year−1 for large lakes (Li et al, 2024). A review of papers on nitrous oxide emissions across 10 lakes in Poland suggests that there is seasonality in nitrous oxide emissions, which a peak in winter and early spring (Woszczyk et al, 2023).

Note on reservoirs behind dams: Carbon burial rates can be higher in human-made systems such as reservoirs than natural ones, due to catchment instability and high erosion rates, and these can become large sources of carbon dioxide and methane where sediment builds up behind dammed areas (Gregg et al, 2021).

9. Coastal and Marine

For marine and coastal habitats, check Commissioned Report 1326 https://www.nature.scot/doc/naturescot-research-report-1326-scottish-blue-carbon-literature-review-current-evidence-scotlands#Background

References

Abdalla, M, Hastings, A, Chadwick, D et al (2018) Critical review of the impacts of grazing on soil organic carbon storage and other soil health indicators in extensively managed grasslands, Agriculture, Ecosystems and Environment, Vol 253.

Aitkenhead M, Coull M (2019.) Mapping soil profile depth, bulk density and carbon stock in Scotland using remote sensing and spatial covariates, European Journal of Soil Science, Vol 71 (issue 4).

Aitkenhead, M, Castellazzi, M, McKeen, M et al (2021) Peatland restoration and potential GHG savings on agricultural land, ClimateXChange, 

Anderegg, W, Trugman, A, Badgley, G et al (2020) Climate driven risks to the climate mitigation potential of forests, Science, Vol 368 (issue1327).

Antala, M, Juszczak, R, van der Tol, C et al (2022) Impact of climate change-induced alterations in peatland vegetation phenology and composition on carbon balance (review), Science of the Total Environment, Vol 827.

Arias-Ortiz, A, Wolfe, J, Bridgham, S et al (2024) Methane fluxes in tidal marshes of the conterminous United States, Global Change Biology, Vol 30 (issue 9).

Attermeyer, K, Casas-Ruiz, JP, Fuss, T et al. (2021) Carbon dioxide fluxes increase from day to night across European streams. Nature Communications Earth Environ, Vol 2 (issue 118). 

Audet, J, Levi, E, Jeppesen, E et al (2024) Nutrient enrichment—but not warming—increases nitrous oxide emissions from shallow lake mesocosms, Limnology and Oceanography.

Axe, M, Grange, I, Conway, J (2017) Carbon storage in hedge biomass- case study of actively managed hedges in England, Agriculture, Ecosystems & Environment, Vol 250. 

Axe, M, Grange, I, Convey, J (2018.) Hedgerow agroforestry in England and Wales: increasing width to sequester additional carbon. European Agroforestry Conference-Agroforestry as Sustainable Land Use, 4^th, EURAF.

Baggaley, N, Britton, A, Barnes, A et al (2021) Understanding carbon sequestration in the uplands. ClimateXChange Report.

Baggaley, N, Britton, A, Sandison, F et al (2022) Understanding carbon sequestration from nature-based solutions, ClimateXChange, Edinburgh.

Baggio-Campanucci, A, Ovando, P, Hewitt, R et al (2022) Barking up the wrong tree? Can forest expansion help meet climate goals?, Environmental Science and Policy, Vol 136.

Bai, Y and Cotrufo, MF (2022) Grassland soil carbon sequestration: Current understanding, challenges, and solutions, Science, Vol 377.

Bartosiewicz, M, Venetz, J, Läubli, S et al (2023) Detritus‐hosted methanogenesis sustains the methane paradox in an alpine lake, Limnology and Oceanography, Vol 68 (issue 1).

Beckert, M, Smith, P, Lilly, A et al (2016) Soil and tree biomass carbon sequestration potential of silvopastoral and woodland-pasture systems in North East Scotland, Agroforestry Systems, Vol 90.

Bertsch-Hoermann, B, Egger, C, Gaube, V et al (2021) Agroforestry trade-offs between biomass provision and aboveground carbon sequestration in the alpine Eisenwurzen region Austria, Regional Environmental Change, Vol 21 (issue 77).

Bezabih Beyene, B, Li, J, Yuan, J et al (2022) Non‐native plant invasion can accelerate global climate change by increasing wetland methane and terrestrial nitrous oxide emissions, Global Change Biology, Vol 28 (issue 18).

Biffi, S, Chapman, P, Grayson, R, Ziv, G (2022) Soil carbon sequestration potential of planting hedgerows in agricultural landscapes, Journal of Environmental Management, Vol 307.

Biffi, S, Chapman, P, Grayson, R, Ziv, G (2023) Planting hedgerows: biomass carbon sequestration and contribution towards net zero targets, Science of the Total Environment, Vol 892.

Billett, M, Garnett, M, Harvey, F (2007) UK peatland streams release old carbon dioxide to the atmosphere and young dissolved organic carbon to rivers. Geophysical Research Letters, Vol 34 (issue 23).

Billett, M, Garnett, M, Dinsmore, K (2015) Should Aquatic CO₂ Evasion be Included in Contemporary Carbon Budgets for Peatland Ecosystems?, Ecosystems, Vol 18 (issue 3)

Black, K et al (2023) Biomass carbon stocks and stock changes in managed hedgerows, Science of the Total Environment, Vol 871. 

Brown, A, Bass, A, Skiba, U et al (2023) Urban landscapes and legacy industry provide hotspots for riverine greenhouse gases: A source-to-sea study of the River Clyde, Water Research, Vol 236.

Bruneau, P and Johnson, S (2014) Scotland’s peatland-definitions & information resources, Scottish Natural Heritage Commissioned Report No701.

Burgess, P, Graves, A (2022) The Potential Contribution of Agroforestry to Net Zero Objectives - Report for the Woodland Trust, Cranfield University.

Bussman, I, Koedel, U, Schutze, C et al (2022) Spatial Variability and Hotspots of Methane Concentrations in a Large Temperate River, Frontiers, Vol 10.

Cardinael, R, Umulisa, V, Toudert, A et al (2018) Revisiting IPCC Tier 1 coefficients for soil organic and biomass carbon storage in agroforestry systems, Environmental Research Letters, Vol 13 (issue 12).

Carswell, F, Burrows, L, Hall, G, et al (2012) Carbon and biodiversity gain during 200 years of woody succession in lowland New Zealand, New Zealand Journal of Ecology, Vol 36 (issue 2). 

Chambers, L, Mirabito, A, Brew, S et al (2024) Evaluating permanganate oxidizable carbon (POXC)’s potential for differentiating carbon pools in wetland soils. Ecological Indicators, Vol 167.

Chapman, P, Moody, C, Turner, E et al (2022) Carbon concentrations in natural and restoration pools in blanket peatlands, Hydrological Processes, Vol 36.

Chen, X, Taylor, A, Reich, P et al (2023) Tree diversity increases decadal forest soil carbon and nitrogen accrual, Nature, Vol 618.

Chizen, C, Helgason, B, Weiseth, B et al (2024) Soil carbon dynamics in drained prairie pothole wetlands, Frontiers in Environmental Science, Vol 12.

Craft, C and Li, S (2024) Short- and long-term legacies of carbon sequestration, and nutrient burial in floodplain wetlands of agricultural and forested catchments, Indiana, United States, River, Vol 3 (issue 1).

Crossland, M (2015) The carbon sequestration potential of hedges managed for woodfuel, The Organic Research Centre.

Crous-Duran J, Graves AR, García de Jalón, S et al (2020) Quantifying Regulating Ecosystem Services with Increased Tree Densities on European Farmland, Sustainability, Vol 12 (issue 16).

Cuenca-Cambronero, M, Blicharska, M, Perrin, J et al (2023) Challenges and opportunities in the use of ponds and pondscapes as Nature-based Solutions, Hydrobiologia, Vol 850 (issue 15).

Cui, S, Liu, P, Guo, H et al (2024) Wetland hydrological dynamics and methane emissions, Communications Earth & Environment, Vol 5 (issue 1).

Darusman, T, Mudiyarso, D, Impron, Anas, I (2023) Effect of rewetting degraded peatlands on carbon fluxes:a meta‑analysis, Mitigation and Adaptation Strategies for Global Change, Vol 28 (issue 10).

Deemer, B, Holgerson, R (2021) Drivers of Methane Flux Differ Between Lakes and Reservoirs, Complicating Global Upscaling Efforts, Biogeosciences, Vol 126 (issue 4).

Doyle, B, de Eyto, E, McCarthy, V et al (2023) The organic carbon budget of an oligotrophic temperate peatland lake, Limnology and Oceanography, Vol 68 (issue 3).

Drexler, S, Gensior, A, and Don, A (2021). Carbon sequestration in hedgerow biomass and soil in the temperate climate zone, Regional Environmental Change, Vol 21(issue 3).

Drexler, S, Thiessen, E and Don, A (2023) Carbon storage in old hedgerows: The importance of below-ground biomass, GCB Bioenergy Vol, 16 (issue 1).

Du, M, Jones, V, van Hardenbroek, M et al (2023). How much organic carbon have UK lakes stored in the Holocene? A preliminary estimate, The Holocene, Vol 33 (issue 6).

Du, Z, Zheng, H, Penuelas, P et al (2024) Shrub encroachment leads to accumulation of C, N, and P in grassland soils and alters C:N:P stoichiometry: A meta-analysis, Science of The Total Environment, Vol 951.

Evans, C, Artz, R, Moxley, J et al (2017) Implementation of an emissions inventory for UK peatlands, Report to the Department for Business, Energy and Industrial Strateg, Centre for Ecology and Hydrology.

Evans, C, Gauci, V (2023) Wetlands and Methane - Technical paper, Wetlands International.

Evans, M., Alderson, D, Evans, C et al (2022). Carbon loss pathways in degraded peatlands: New insights from radiocarbon measurements of peatland waters, Journal of Geophysical Research: Biogeosciences, Vol 127 (issue 7).

Emmett, B. A, Reynolds, B, Rowe, E et al, C. (n.d.) (2010) Soil Report, Countryside Survey.

Evans, C, Peacock, M, Baird, A et al. (2021) Overriding water table control on managed peatland greenhouse gas emissions, Nature, Vol 593.

Fernández-Guisuraga, J, Calvo, L, Ansola, G et al (2022) The effect of sheep grazing abandonment on soil bacterial communities in productive mountain grasslands, Science of The Total Environment, Vol 851 (issue 2).

Ferretto, A, Brooker, R, Aitkenhead, M et al (2019) Potential carbon loss from Scottish peatlands under climate change, Regional Environmental Change, Vol 19.

Fletcher, T, Scott, C, Hall, J et al (2021) The carbon sequestration potential of Scottish woodlands, Environmental Research Communications, Vol 3.

Fluet-Chouinard, E, Stocker, B, Zhang, Z et al (2023) Extensive global wetland loss over the past three centuries, Nature, Vol 614.

Forest Research (2023) Forestry Statistics 

Friggens, N, Hester, A, Mitchell, R et al (2020) Tree planting in organic soils does not result in net carbon sequestration on decadal timescales, Global Change Biology, Vol 26 (issue 9).

García de Jalón, S, Graves, A, Palma, J et al (2018) Modelling and valuing the environmental impacts of arable, forestry and agroforestry systems: a case study, Agroforestry Systems, Vol 92 (issue 2).

Global Carbon Project website, Methane Budget, [retrieved 2024]

Golub, M, Koupaei-Abyazani, N, Vesala, T et al (2023). Diel, seasonal, and inter-annual variation in carbon dioxide effluxes from lakes and reservoirs, Environmental Research Letters, Vol 18 (issue 3).

Gommet, C, Lauerwald, R, Ciais, P et al (2022) Spatiotemporal patterns and drivers of terrestrial dissolved organic carbon (DOC) leaching into the European river network, Earth System Dynamics, Vol 13 (issue 1).

Granath G, Moore, P et al (2016) Mitigating wildfire carbon loss in managed northern peatlands through restoration, Nature Scientific Reports 6.

Gregg, R, Elias, J, Alonso, I et al (2021) Carbon storage and sequestration by habitat: a review of the evidence (second edition) Natural England Research Report NERR094, Natural England, York.

Gu, C, Waldron, S, Bass, A (2021) Carbon dioxide, methane, and dissolved carbon dynamics in an urbanized river system, Hydrological Processes, Vol 35 (issue 9)

Gundersen, P, Thybring, E, Nord-Larsen, T et al (2021) Old-growth forests carbon sinks overestimated, Nature, Vol 591 (issue 7851) & Luyssaert et al (2021) Reply to Old-growth forests carbon sinks overestimated, Nature.

Hassett, E, Bohrer, G, Kinsman-Costello, L et al (2024) Changes in inundation drive carbon dioxide and methane fluxes in a temperate wetland, Science of the Total Environment, Vol 915.

Heger, A, Becker, J, Vásconez Navas, L et al (2021) Factors controlling soil organic carbon stocks in hardwood floodplain forests of the lower middle Elbe River, Geoderma, Volume 404 (issue 8).

Heger, A, Becker, J, Vásconez Navas, L et al (2024). Drivers for soil organic carbon stabilization in Elbe River floodplains, Journal of Plant Nutrition and Soil Science, Vol 187 (issue 3). 

Hirst, C (2021) Deer and climate change in Scottish woodlands, ClimateXChange.

Hoffmann, T (2022) Carbon sequestration of floodplains in Treatise on Geomorphology (2d Edition), Vol 9.

Holland, J, Pollock, M, Buckingham, S et al (2022) Reviewing, assessing and critiquing the evidence base on the impacts of muirburn on wildfire prevention, carbon storage and biodiversity, NatureScot Research Report 1302. 

Housego, N, Parker, T, Street, L et al (2025) Natural tree colonisation of organo‐mineral soils does not provide a net carbon capture benefit at decadal timescales, Journal of Applied Ecology, Vol 62 (issue 3).

Hu, H, Chen, J, Zhou, F et al (2024) Relative increases in CH4 and CO2 emissions from wetlands under global warming dependent on soil carbon substrates, Nature Geoscience, Vol 17 (issue 1).

International Panel on Climate Change (2006) IPCC Guidelines on GHG inventories 

International Panel on Climate Change (2018) IPCC – Agriculture Forestry and Other Land Use Chapter 

Jacinthe P (2015) Carbon dioxide and methane fluxes in variably-flooded riparian forests, Geoderma, Vol 241–242 In Schindlbacher, A, Heinzle, J, Gollobich, G et al (2022) Soil greenhouse gas fluxes in floodplain forests of the Danube National Park: effects of flooding and soil microclimate, Biogeochemistry, Vol 159 (issue 2)

Jeffries, M, Gilbert, P, Taylor, S et al (2023) Organic carbon in British lowland ponds: estimating sediment stocks, possible practical benefits and significant unknowns, Hydrobiologia, Vol 850 (issue 15).

Johnson, M, Matthews, E, Du, J et al (2022) Methane emission from global lakes: New spatiotemporal data and observation-driven modeling of methane dynamics indicates lower emissions, Journal of Geophysical Research: Biogeosciences, Vol 127 (issue 7).

Khatun, S, Berg, J, Jézéquel, D et al (2024) Long‐range transport of littoral methane explains the metalimnetic methane peak in a large lake, Limnology and Oceanography, Vol 69 (issue 9).

Koschorreck, M, Kamjunke, N, Ködel, U et al (2023). Diurnal versus spatial variability of greenhouse gas emissions from an anthropogenic modified German lowland river, Vol 21 (issue 6).

Lesaint, L, Viaud, V, Menasseri-Aubry, S (2023) Influence of soil properties and land use on organic carbon storage in agricultural soils near hedges, Soil Use Management, Vol 39 (issue 3)

Leuschner, C, Feldmann, H, Pichler, V et al (2022) Forest management impact on soil organic carbon; a paired-plot study in primeval and managed European beech forests, Forest Ecology and Management, Vol 512.

Li, L, Fuchs, A, Herrero Ortega, S et al (2021) Spatial methane pattern in a deep freshwater lake: Relation to water depth and topography, Science of The Total Environment, Vol 764 (issue G2).

Li, J, Hao, T, Yang, M et al (2024) Analysis of methane emission characteristics and environmental response in natural wetlands, Atmospheric Environment, Vol 334.

Li, Y, Tian, H, Yao, Y et al (2024) Increased nitrous oxide emissions from global lakes and reservoirs since the pre-industrial era, Nature Communications, Vol 15 (issue 942).

Loder, A, Zamaria, S, Arhonditsis, G et al (2023) Controls on organic carbon stocks among restored wetland soils in the Long Point region of southern Ontario, Canada, Restoration Ecology, Vol 31 (issue 5).

Luyssaert, S, Schulze, ED, Bonner, A et al (2008) Old-growth forests as global carbon sinks, Nature Letters, Vol 455.

Malerba, M, Friess, D, Peacock et al (2022) Methane and nitrous oxide emissions complicate the climate benefits of teal and blue carbon wetlands, One Earth, Vol 5 (issue 12).

Martin-Benito, D, Pederson, N, Feriz, M et al (2021) Old forests and old carbon: a case study on the stand dynamics and the longevity of above ground carbon, Science of the Total Environment, Vol 765.

Matthews, K, Wardell-Johnson, D, Miller, D et al (2020) Not seeing the carbon for the trees? Why area-based targets for establishing new woodlands can limit or underplay their climate change mitigation benefits, Land Use Policy, Vol 97.

Matthews, R, Henshall, P, Beauchamp, K et al (2022) Quantifying the sustainable forest carbon cycle, Forest Research.

Mayer, S, Wiesmeier, M, Sakamoto, E et al (2022) Soil organic carbon sequestration in temperate agroforestry systems – A meta-analysis, Agriculture, Ecosystems & Environment, Volume 323.

Mikolas, M, Svitok, M, Bace, R et al (2021) Natural disturbance impacts on trade-offs and co-benefits of forest biodiversity and carbon, Proceedings of the Royal Society B, Vol 288 (issue 1961).

Mwanake, R, Imhof, H, Kiese, R (2024) Divergent drivers of the spatial variation in greenhouse gas concentrations and fluxes along the Rhine River and the Mittelland Canal in Germany, Environmental Science and Pollution Research, Vol 31.

Nadal-Romero, E, Rubio, R, Kremyda, V et al (2021) Effects of agricultural land abandonment on soil organic carbon stocks and composition of soil organic matter in the Central Spanish Pyrenees, CATENA, Vol 205.

Nanson, G, Croke J (1992) A genetic classification of floodplains, Geomorphology, Vol 4.

Ordóñez, C, DelSontro, T, Langenegger, T et al (2023) Evaluation of the methane paradox in four adjacent pre-alpine lakes across a trophic gradient, Nature Communications, Vol 14 (issue 2165).

Osei, R, del Río, M, Ruiz-Peinado, R et al (2022) The distribution of carbon stocks between tree woody biomass and soil differs between Scots pine and broadleaved species (beech, oak) in European forests, European Journal of Forest Research, Vol 141.

Ouyang, W, Wang, P, Liu, S et al (2021) Rainfall stimulates large carbon dioxide emission during growing season in a forest wetland catchment, Journal of Hydrology, Vol 602.

Oyesiku-Blakemore, J, Dindini, M (2022) Managing permanent grassland for carbon sequestration, ClimateXChange

Pan, Y, Liu, J, Zhang, M et al (2024) The below-ground biomass contributes more to wetland soil carbon pools than the above-ground biomass-a survey based on global wetlands, Journal of Plant Ecology, Vol 17 (issue 5).

Patenaude, G, Briggs, B, Milne, R et al (2003) The carbon pool in a British semi-natural woodland, Forestry, Vol 76 (issue 1).

Peacock, M, Davidson, S, Kothawala, D et al (2023), Spatial and seasonal variations in dissolved methane across a large lake, Journal of Geophysical Research: Biogeosciences, Vol 128 (issue 8).

Peacock, M, Lawson, C, Gowing, D et al (2024) Water table depth and plant species determine the direction and magnitude of methane fluxes in floodplain meadow soils, Ecology and Evolution, Vol 14 (issue 3).

Pellerin, S, Bamière, L, et al (2020) Stocker du carbone dans les sols français, Quel potentiel au regard de l’objectif 4 pour 1000 et à quel coût ? Rapport scientifique de l'étude, INRA (France).

Peatland Carbon Code Guidance (2023) https://www.iucn-uk-peatlandprogramme.org/sites/default/files/2023-03/Peatland%20Code%20Guidance%20V2%20FINAL%20WEB_5.pdf

Perks, M , Khomik, M, Bathgate, S et al (2018) Agroforestry in Scotland: potential benefits in a changing climate, ClimateXChange.

Pickard, A, Branagan, M, Billett, M et al (2022) Effects of peatland management on aquatic carbon concentrations and fluxes, Biogeosciences, Vol 19 (issue 5).

Praetzel, L, Schmiedeskamp, M, Knorr, K (2021) Temperature and sediment properties drive spatiotemporal variability of methane ebullition in a small and shallow temperate lake. Limnology and Oceanography, Vol 66 (issue 7).

Prairie, Y, Cole, J (2022) The carbon cycle in lakes: a biogeochemical perspectives In Encyclopaedia in Open Waters 2d edition.

Quine, T, Cressey, E, Dungait, J et al (2022) Geomorphically Mediated Carbon Dynamics of Floodplain Soils and Implications for Net Effect of Carbon Erosion, Hydrological Processes, Vol 36 (issue 9).

Rabaey, J, Cotner, J (2024) The influence of mixing on seasonal carbon dioxide and methane fluxes in ponds, Biogeochemistry, Vol 167.

Ragg, R, Peeters, F, Ingwersen, J et al (2021) Interannual variability of methane storage and emission during autumn overturn in a small lake, Journal of Geophysical Research: Biogeosciences, Vol 126 (issue 12).

Ray, N, Holgerson, M, Andersen, M et al (2023) Spatial and temporal variability in summertime dissolved carbon dioxide and methane in temperate ponds and shallow lakes, Limnology and Oceanography, Vol 68 (issue 7).

Rees, B, Buckingham S, Chapman S et al (2018) Soil carbon and land use in Scotland, ClimateXChange.

Reid, C, Hornigold, K, McHenry, E et al (2021) State of the UK's Woods and Trees 2021, Woodland Trust.

Reis, P, Thottathil, S, Prairie, Y (2022) The role of methanotrophy in the microbial carbon metabolism of temperate lakes, Nature Communications, Vol 13 (issue 43).

Robertson, H, Marshall, D, Slingsby, E et al (2012) Economic, biodiversity, resource protection and social values of orchards: a study of six orchards by the Herefordshire Orchards Community Evaluation Project, Natural England Commissioned Reports, Number 090.

Rocher-Ros, G, Stanley, E, Loken, L et al (2023) Global methane emissions from rivers and streams, Nature, Vol 621.

Rodwell, J.S. (ed) 1991. British Plant Communities Vol 2. Mires and heaths. Cambridge University Press. 

Rosentreter, J, Borges, A, Deemer, B et al (2021). Half of global methane emissions come from highly variable aquatic ecosystem sources. Nature Geoscience, Vol 14 (issue 4).

Saraev, V, Low WH, Beauchamp, K et al (2022) The potential for agroforestry to reduce net GHG emissions in Scotland through the Woodland Carbon Code, ClimateXChange.

Schindlbacher, A, Heinzle, J, Gollobich, G et al (2022) Soil greenhouse gas fluxes in floodplain forests of the Danube National Park: effects of flooding and soil microclimate, Biogeochemistry, Vol 159.

Schmiedeskamp, M, Praetzel, L, Bastviken, D et al (2021) Whole‐lake methane emissions from two temperate shallow lakes with fluctuating water levels: Relevance of spatiotemporal patterns, Limnology and Oceanography, Vol 66 (issue 6).

Schroll, M, Liu, L, Einzmann, T et al (2023) Methane Accumulation and its Potential Precursor Compounds in the Oxic Surface Water Layer of Two Contrasting Stratified Lakes, Vol 903.

Scottish Forestry, Native Woodland Survey of Scotland (2006-2013),  [retrieved 2023]

Sear, D, Speck, I, Pears, B (2023) Carbon storage in river and floodplain systems: A review of evidence to update and inform policy development for riverine Nature based solutions, Southampton: University of Southampton.

Sepp, M, Koiv, T, Noges, P et al (2022) Catchment soil characteristics predict organic carbon, nitrogen, and phosphorus levels in temperate lakes, Freshwater Science, Vol 41 (issue 1).

Shupe, A, Jensen, K, Oldeland, J et al (2022) Droughts decrease and floods increase carbon sequestration rates of Quercus robur in hardwood floodplain forests, Trees, Forests and People, Vol 9.

Sieczko, A, Thanh Duc, N, Schenk, J et al (2020) Diel variability of methane emissions from lakes, PNAS, Vol 117 (issue 35).

Skwierawski, A. (2022). Carbon sequestration potential in the restoration of highly eutrophic shallow lakes. International Journal of Environmental Research and Public Health, Vol 19 (issue 10).

Smith, S, Vanderberghe, C, Hastings, A et al (2014) Optimising carbon storage through a spatially heterogeneous upland grassland through sheep grazing management, Ecosystems, Vol 17 (issue 3).

Smyth, M, Taylor, E, Birnie, R et al (2015) Developing Peatland Carbon Metrics and Financial Modelling to Inform the Pilot Phase UK Peatland Code, Report to Defra for Project NR0165, pp18-19, Crichton Carbon Centre, Dumfries.

Sokołowska, J; Józefowska, A; Zaleski, T (2022) Impact of Natural Forest Succession on Changes in Soil Organic Carbon in the Polish Carpathian Mountains, Forests, Vol 13 (issue 5).

Soil Association (2019) The Agroforestry Handbook, First Edition, Edited by Ben Raskin and Simone Osborn.

Sø, JS, Sand-Jensen, K, Marinsen, K et al (2023) Methane and carbon dioxide fluxes at high spatiotemporal resolution from a small temperate lake, Science of the Total Environment, Vol 878.

Sø, J, Martinsen, K, Kragh, T et al (2024) Hourly methane and carbon dioxide fluxes from temperate ponds, Biogeochemistry, Vol 167.

Sørensen, M, Strimbeck, R, Nystuen, K et al (2018) Draining the Pool? Carbon Storage and Fluxes in Three Alpine Plant Communities, Ecosystems, Vol 21 (issue 2).

Speir, S, Tank, J, Taylor, J et al (2023) Temperature and carbon availability interact to enhance nitrous oxide production via denitrification in alluvial plain river sediments, Biogeochemistry, Vol 165 (issue 2).

Staton, T, Breeze, T, Walters, R et al (2022) Productivity, biodiversity trade-offs, and farm income in an agroforestry versus an arable system, Ecological Economics, Volume 191.

Steiger, J, Tabacchi, E, Dufour, S, et al (2005) Hydrogeomorphic processes affecting riparian habitat within alluvial channel-floodplain river systems: a review for the temperate zone, River Research and Applications, Vol 21 (issue 7).

Stephenson, N, Das, A, Condit, R et al (2014) Rate of tree carbon accumulation increases continuously with tree size, Nature, Vol 507.

Swinnen, W, Daniëls, T, Maurer, E et al (2019) Geomorphic controls on floodplain sediment and soil organic carbon storage in a Scottish mountain river, Earth Surface Processes and Landforms, Vol 45 (issue 1).

Taillardat, P, Bodmer, P, Deblois, C et al (2022) Carbon dioxide and methane dynamics in a peatland headwater stream: Origins, processes and implications, Journal of Geophysical Research: Biogeosciences, Vol 127 (issue 7).

Tangen, B, Bansal, S (2022) Prairie wetlands as sources or sinks of nitrous oxide: effects of land use and hydrology, Agricultural and Forest Meteorology, Vol 320.

Thottathil, S, Reis, P, Prairie, Y (2022) Magnitude and Drivers of Oxic Methane Production in Small Temperate Lakes, Environmental Science & Technology, Vol 56 (issue 15).

Tipping, E, Elias, J, O. Keenan, P et al (2022) Relationships between riverine and terrestrial dissolved organic carbon: Concentration, radiocarbon signature, specific UV absorbance, Science of The Total Environment, Vol 817.

Upson, M (2014) The Carbon Storage Benefits of Agroforestry and Farm Woodlands, Unpublished PhD thesis, Cranfield University in Burgess, P and Graves, A (2022) The Potential Contribution of Agroforestry to Net Zero Objectives - Report for the Woodland Trust, Cranfield University.

Upson M, Burgess, P and Morison, J (2016) Soil carbon changes after establishing woodlands and agroforestry trees in a grazed pasture, Geoderma, Vol 283.

Vachon, D, Sponseller, R, Rosvall, M et al (2023). Controls on terrestrial carbon fluxes in simulated networks of connected streams and lakes, Global Biogeochemical Cycles, Vol 37 (issue 3).

Van der Berge S, Vangansbeke, P, Baeten, L et al (2021a) Biomass increment and carbon sequestration in hedgerow-grown trees, Dendrochronologia, Vol 70.

Van Den Berge, S, Vangansbeke, P, Baeten, L et al (2021b) Soil carbon of hedgerows and ‘ghost’ hedgerows, Agroforestry Systems, Vol 95 (issue 3).

van Noordwijk, M, Aynekulu, E, Hijbeek, R et al (2023) Soils as carbon stores and sinks: Expectations, patterns, processes, and prospects of transitions, Annual Review of Environment and Resources, Vol 48.

Vanguelova, E, Nisbet, T, Moffat, A et al (2013) A new evaluation of carbon stocks in British forest soils, Soil Use and Management, Vol 29 (issue 2).

Viaud, V, Kunnemann T (2021) Additional soil organic carbon stocks in hedgerows in crop-livestock areas of Western France, Agriculture, Ecosystems and Environment, Vol 305.

Wallace, C, Tonina, D, McGarr, J et al (2021) Spatiotemporal Dynamics of Nitrous Oxide Emission Hotspots in Heterogeneous Riparian Sediments, Water Resources Research, Vol 57 (issue 12).

Warner, D., Tzilivakis, J., Lewis, K. (2011) A revisit to previous research into the current and potential climate change mitigation effects of environmental stewardship (Defra BD5007), University of Hertfordshire.

Warner, E, Lewis O, Brown, N et al (2021) Does restoring native forest restore ecosystem functioning? Evidence from a large-scale reforestation project in the Scottish Highlands, Restoration Ecology, Vol 30 (issue 3).

Weissgerber, M, Chanteloup, L and Bonis, A (2024) Carbon stock increase during post-agricultural succession in central France: no change of the superficial soil stock and high variability within forest stages, New Forests, Vol 55.

Williamson, J, Tye, A, Lapworth, D et al (2021) Landscape controls on riverine export of dissolved organic carbon from Great Britain, Biogeochemistry, Vol 164.

Wohl, E, Pfeiffer, A (2018) Organic carbon storage in the floodplain soils of the U.S. prairies, River Research and Applications, Vol 34 (issue 5)

Wohl, H, Vox, R (2022) A first-order approximation of floodplain soil organic carbon stocks in a river network: The South Platte River, Colorado, USA as a case study, Science of the Total Environment, Vol 852 (issue 5).

Woodland Carbon Code – calculations update with latest version 2024 – see templates https://woodlandcarboncode.org.uk/landowners-apply/template-documents

Woszczyk, M, Schubert, C (2023) Methane and Nitrous Oxide Emissions From an Anthropogenically Transformed Lake (Lake Licheńskie, Poland), Journal of Geophysical Research: Biogeosciences, Vol 128 (issue 12).

Woziwoda, B, Parzych, A and Kope, D (2014) Species diversity, biomass accumulation and carbon sequestration in the understorey of postagricultural Scots pine forests, Silva Fennica, Vol 48 (issue 4).

Annex 1 & 2