The potential contribution of terrestrial nature- based solutions to a national ‘net zero’ climate target

1. Many national governments have incorporated nature- based solutions (NbS) in their plans to reduce net greenhouse gas emissions. However, uncertainties per sist regarding both feasibility and consequences of major NbS deployment. Using the United Kingdom as a national- level case study, we examined the potential contribution of three terrestrial NbS: peatland restoration, saltmarsh creation and woodland creation. 2. While there is substantial political and societal interest in these three NbS, they also have strong potential for competition with other land uses, which will be a critical barrier to substantial deployment. We conducted a national mapping exercise to assess the potential area available for woodland creation. We then assessed the combined climate change mitigation potential to 2100 for the three NbS options under a range of ambition levels. 3. In line with the most ambitious targets examined, 2 Mha of land is potentially


| INTRODUC TI ON
Tackling the climate and nature crises is essential to achieve the sustainable development goals (UN, 2016). Most of the world's governments have pledged action via United Nations conventions, particularly the Convention on Biological Diversity (CBD) and the Framework Convention on Climate Change (UNFCCC). Signatories of the UNFCCC Paris agreement aim to reduce greenhouse gas (GHG) emissions to keep global temperature rises well below 2°C (Roe et al., 2019). Numerous actions have been proposed to achieve these goals, combining reductions in fossil fuel use with GHG removal from the atmosphere (Fekete et al., 2021). Among the latter, nature-based solutions (NbS) have gained considerable attention (Fargione et al., 2018;Seddon, Daniels, et al., 2020). If implemented appropriately, NbS provide important co-benefits including flood alleviation, improved livelihoods and biodiversity conservation Di Sacco et al., 2021;. The United Kingdom and its devolved governments have committed to reaching net zero GHG emissions by 2045 or 2050 (Priestley, 2019;Scottish Government, 2018;Welsh Government, 2021). The United Kingdom therefore provides a useful test case of the feasibility of a significant NbS contribution to a national net zero target.
There is growing appetite in the United Kingdom for substantial NbS use to reduce atmospheric CO 2 , especially through peatland restoration, woodland creation and saltmarsh creation (Brandmayr et al., 2019;Climate Assembly UK, 2020). A range of targets have been proposed by governmental and non-governmental organisations (Brandmayr et al., 2019; Climate Change Committee (CCC), 2020; Woodland Trust, 2020). UK devolved governments are already committing funding to peatland restoration and tree planting, with Scotland's £350 million pledge (Scottish , and England's £640 million Nature for Climate fund (HM Treasury, 2020).
However, there have been only limited assessments of the landuse change necessary and the climate change mitigation benefits that might be achieved by implementing NbS at the national scale (CCC, 2019(CCC, , 2020Matthews et al., 2020;Thomson et al., 2018).
Significant questions therefore remain regarding the feasibility and potential contribution of these NbS to meeting net zero emissions targets. Other terrestrial mitigation measures include improved agricultural management of farmland carbon stores (Cardenas et al., 2019;Montgomery et al., 2020). Although these actions might impact agricultural production, they are unlikely to necessitate widespread habitat change. We therefore focus our analysis on peatland, woodland and saltmarsh restoration and creation where there is the greatest potential for competition with other land uses.
Meeting proposed woodland creation targets would necessitate changing substantial areas of land use, but most UK land is already under agriculture or important for semi-natural open habitats. The devolved UK governments incorporate a degree of spatial planning in woodland expansion, with statutory agencies publishing woodland opportunity maps (e.g. Sing et al., 2013;Thomas et al., 2017;Welsh Government, 2020). However, these indicative maps do not consider spatial variation in existing soil carbon stores nor climatic influences on tree growth, both of which strongly affect potential for net carbon sequestration. A more detailed assessment including these factors has recently been published for Scotland (Matthews et al., 2020), but equivalents for the other UK nations are lacking.
Here, we scrutinise existing UK NbS scenarios for climate change mitigation (Table 1) Logistical capacity, economic costs and socio-cultural acceptance of NbS-driven changes in land use are also critical (Foster et al., 2013;Hopkins et al., 2017;Tew et al., 2019), but are not formally assessed here. While accepting that these factors may further limit NbS deployment, it is still important to understand the baseline potential for NbS, as a component of a nation's overall GHG reduction strategy.

| MATERIAL S AND ME THODS
The NbS considered here (Table 1)  When CCC (2020) was published, the Scottish Government was the only UK administration to have made a firm commitment regarding peatland restoration. The CCC (2020), therefore, assumes peatland restoration only occurs in Scotland in the Business-as-Usual scenario, but occurs across the whole United Kingdom in the other scenarios. b Apart from the Business-as-Usual scenario, the CCC gives all other peatland restoration targets as percentages of the total area of degraded peatland in the United Kingdom (Table S1). We determined areas of each category based on figures in ONS (2019), given here rounded to the nearest ha. Annual restoration rate of each category increased over time to represent increasing capacity (Tables S2 and  S7). c We also explored the impacts of restoring 100% of afforested peat (i.e. 439,292 ha) by 2050 in this scenario, see main text.
d There are slight differences between the overall totals and the sum of individual category areas, due to rounding of the latter. government's Climate Change Committee (CCC, 2020), and bespoke saltmarsh creation scenarios based on recent opportunity mapping (RSPB, 2018). In all scenarios, we assumed that restoration and creation began in 2021 at close to current implementation rates and increased linearly each year to the maximum rate shown in Table 1 and

| Peatland restoration
As peatland condition is not comprehensively mapped in the United Kingdom, we could not spatially prioritise restoration. We therefore used the UK area estimates for different categories of degraded peatland from Office for National Statistics (ONS, 2019; Table S1) and restored each category in proportion to country-specific ONS area estimates (Table 1). We excluded areas of near-natural or rewetted bog and fen from our calculations. Deep peat soils covered by cropland and intensive grassland were treated as lowland peat, and domestic fuel extraction and the various degraded bog classes as upland peat ; Table S1). During the 20th century, large areas of peatland were afforested with commercial conifer plantations, which necessitated extensive ground preparation and drainage leading to much greater GHG emissions compared with undamaged peatlands (Table S1). However, CCC scenarios limit restoration of these afforested peatlands to only 20% of the lowest productivity plantations (CCC, 2020). Given the substantial emissions from these areas (Evans et al., 2017; see Supporting Information), we additionally explored the effects of restoring all afforested peatlands. For all scenarios, we offset emissions against the carbon sequestered in the trees on afforested peatlands and their derived harvested wood products (HWP; see Supporting Information).
To calculate restoration impacts, we used GWP 100 values from Evans et al. (2017), except for paludiculture and sustainably managed cropland where we adapted values from CCC (2020 ; Table S1).
Peatland GWP 100 values include direct and indirect GHG emissions and DOC (Evans et al., 2017). Degraded upland areas were assumed to be restored to rewetted bog, and lowland areas to rewetted fen. In line with IPCC (2014), we assumed immediate switching of emissions from degraded to rewetted categories during the year of restoration. The annual area restored was gradually increased each year to represent growing capacity for restoration management (see Supporting Information).

| Woodland creation
We undertook a mapping exercise to identify potentially suitable areas for woodland creation, to assess the feasibility of national woodland creation targets (full methods in Supporting Information).
Areas considered unavailable for woodland creation included higher quality agricultural land and existing woodland (the modification of which could reduce agricultural or timber production, leading to offshoring of emissions), designated sites, priority habitats and peatlands (to avoid perverse outcomes for soil carbon or biodiversity) and existing buildings, infrastructure and archaeological features (Table S3). Some excluded areas were further buffered to limit possible negative spill-over effects of woodland creation, such as on peatland hydrology (FC, 2000), or to reduce predation in sites designated for conservation of wading birds (Wilson et al., 2014; see Supporting Information).
After excluding these areas, we found 530,280 distinct spatial units (henceforth termed polygons) that are potentially available for new woodland (mean 8.8 ha, SD 40.2 ha; see Supporting Information). We considered a representative range of tree species (Table S4), and for each scenario, we randomly assigned polygons to either broadleaves or conifers in line with the target ratios ( To maximise the climate benefits of woodland expansion while minimising impacts on agricultural output, we prioritised the order of polygon conversion to woodland. We categorised polygons as either lower soil carbon risk (mineral soils) or higher risk (organo-mineral soils) for woodland creation, as some recent evidence suggests that woodland creation on richer organo-mineral soils can trigger soil carbon losses that might at least partly counteract the benefits of sequestration by new trees (Friggens et al., 2020). Across all scenarios, we therefore prioritised establishment on polygons with mineral soils and then least productive land (i.e. highest Agricultural Land Classification/Land Capability for Agriculture grade). Broadleaved polygons were further ranked following the 'Lawton principles' (Defra, 2010), that is, prioritising sites adjacent to existing woodland, and then ordered in deceasing size (further details in Supporting Information). Conifer polygons were prioritised in the order of highest yield class, to maximise timber output and then decreasing size.
We calculated carbon sequestration of each polygon using species-and yield-class-specific values extracted from the  Jenkins et al., 2018). Sequestration values include all above-and below-ground biomass and litter, but not soil carbon (Randle & Jenkins, 2011). To account for impacts of woodland creation on soil carbon stores, we used polygon-specific soil carbon estimates  and predicted changes following woodland creation depending on country, current habitat and woodland type (Bradley et al., 2005; see Supporting Information). Following the CCC scenarios, conifers were planted at a density of 1.5 or 1.7 m spacing, broadleaves at 2.5 m (Table S4) and all woodlands were thinned. Carbon in thinnings was assumed to be immediately lost to the atmosphere.
For conifers, we used a 40-year rotation (Moore, 2011) with a 3-year fallow period between rotations but acknowledge that a fixed rotation length is a simplification. Following end of rotation clear-fell, unharvested residues were assumed to be left on-site and decayed over time (Morison et al., 2012). Timber was apportioned to a range of harvested wood product (HWP) pools according to national statistics (FC, 2020). Our scenarios incorporated the carbon stored in these HWPs, with each product class having a different life span IPCC, 2003;Moore, 2011;UNFCCC, 2003).
Each year, a portion of HWP reached end of life and went to landfill, where some emissions continue (IPCC, 2006). Substitution effects of HWP are often considered in analyses like this, though potential emissions avoidance benefits may have been overstated 2-100 fold (Harmon, 2019). We include supplementary analysis with substitution but caution against its use (see Supporting Information).

| Saltmarsh creation
A previous study identified 318 potential managed realignment projects in the United Kingdom, totalling 29,996 ha potentially available for new saltmarsh habitat, replacing a range of terrestrial habitats (RSPB, 2018; Table S6). Area targets in Table 1 were adjusted by subtracting the estimated 105 ha of saltmarsh that will be lost each year to sea-level rise and coastal squeeze (Beaumont et al., 2014;ONS, 2016;RSPB, 2018). Although some mudflats would also be created, we assume that the entire realigned area would be converted to saltmarsh (see Supporting Information). We apply sequestration rates from Burden et al. (2019), which cover below-ground carbon only and do not include plant biomass. We did not spatially prioritise projects, as they will be driven by multiple considerations beyond climate change mitigation, such as hydrological and coastal geomorphology, and public acceptance regarding ceding land to the sea (Foster et al., 2013;Myatt et al., 2003).
For all three NbS, we determined the likely extent of land cover change on existing habitats (see Supporting Information). Our calculations did not include relatively minor emissions from restoration and creation activities such as fossil fuels for machinery, removal of existing site vegetation, seedling production in nurseries, fencing, herbicide, road construction and timber extraction (Lamb et al., 2016;Morison et al., 2012). Climate change is predicted to subject UK habitats to greater risks of pest outbreaks, fire and windthrow (Ray et al., 2010). Although all these disturbances affect carbon sequestration and storage rates, precise impacts are uncertain and complex, and so we did not incorporate future climate change into our scenarios.

| Peatland restoration
Under the Business-as-Usual scenario with minimal restoration, the United Kingdom's peatlands will release a cumulative 1,674 Mt CO 2 e by 2100. Peatland restoration in the Widespread Engagement scenario would reduce this total to 1,011 Mt CO 2 e, hence avoiding the emission of 663 Mt CO 2 e by 2100 (Figure 1a; Table 2). Initially suggested that Sitka spruce growing at the UK mean yield class in such locations is unlikely to sequester more carbon than is emitted from the degraded peat. Increasing the ambition for restoring afforested peatlands from 20% to 100% potentially avoids the emission of a further 74.5 Mt CO 2 e (Figure 1a).

| Woodland creation
After excluding inappropriate areas, over 4.  Table 2). The production forestry-orientated Widespread Innovation scenario had rapid initial growth rates, giving cumulative net GWP 100 of −12.7 Mt CO 2 e, 9% lower than the biodiversity-orientated Widespread Engagement scenario in 2050.
However, by 2100, substantially more sequestration had occurred as trees matured. The Widespread Engagement scenario then had a cumulative net GWP 100 of −111.5 Mt CO 2 e, 23% lower than the Widespread Innovation scenario. The Tailwinds scenario offered maximum woodland sequestration overall, with −602.9 Mt CO 2 e sequestered by 2100.
Most differences in mitigation potential among scenarios stem from woodland management choices. Over the century, greater sequestration is possible with broadleaves managed for biodiversity rather than conifers managed for timber ( Figure S5).
Initially similar per-hectare sequestration rates diverge as intensive forestry management limits long-term sequestration in conifers. Consequently, mean cumulative net GWP 100 over 100 years  (Figure 2b). This is due both to unharvested broadleaves reaching maturity, while production conifer forestry becomes limited by rotational harvesting and losses from HWP.

| Saltmarsh creation
Despite a net loss of 60 ha per year in the Business-as-Usual scenario, higher initial rates of sequestration in new saltmarsh are sufficient to offset these losses and result in net cumulative and annual sequestration (Figures 1c and 2c). However, contribution of saltmarsh creation to the United Kingdom's net zero target is limited by the small areas available. Even under our higher ambition scenarios, with coastal realignment at 11 times the current annual rate, cumulative net GWP 100 is only −2.7 Mt CO 2 e by 2100 (Table 2).

| Combined restoration and creation
Enacting the Widespread Engagement scenario plus 100% restoration of afforested peatland could provide up to −1,326.8 Mt CO 2 e of cumulative climate change mitigation by 2100 (

| D ISCUSS I ON
The  (Seddon, Daniels, et al., 2020). However, the dominant NbS narrative, particularly around woodland creation, strongly emphasises climate change mitigation. Given the modest mitigation potential of even highly ambitious UK NbS targets, compared to national emissions, this emphasis risks undermining the co-benefits provided by natural and semi-natural habitats. Any NbS-driven land use change must prioritise potential co-benefits to avoid perverse outcomes in which relatively modest climate change mitigation undermines nature and other societal objectives.
The delay before substantial benefits accrue highlights both the urgency with which NbS should be deployed but also that these potential land use changes will have to last for decades. A long-term view is essential to understand potential benefits and the full impacts of management choices. Our analysis highlights the critical importance of restoring degraded peatlands, which in their current state will emit more GHGs (1,674 Mt CO 2 e cumulative emissions by 2100) than potential new woodlands could sequester (up to 603 Mt CO 2 e cumulative sequestration by 2100). However, if all afforested peatlands were restored, new plantations would be required elsewhere to maintain current timber production, leading to greater land demand than forecast by CCC (2020). For woodlands to help address the climate crisis, careful deployment means targeting lower risk soils, minimising ground disturbance during establishment and prioritising broadleaved woodlands managed for conservation.
Our woodland opportunity map (Figure 3) is indicative rather than prescriptive. Any site will need careful assessment for potential impacts on local communities, existing biodiversity and soil TA B L E 2 Cumulative net GWP  carbon prior to woodland establishment. Depending on the management aims, it may also be desirable to consider a much broader selection of tree species than we were able to assess here. In particular, Scots pine Pinus sylvestris is a key component of native woodlands in Scotland, and where appropriate could be considered as a candidate species for woodlands managed for biodiversity conservation.
Our mapping suggests there is sufficient potentially suitable land available to meet the CCC's ambitious woodland creation targets, but enacting NbS at the suggested scale will have wide-ranging and long-lasting impacts. Even though our mapping avoids woodland creation on higher grade farmland, conversion of any agricultural land will reduce overall production, unless accompanied by increased efficiencies and/or demand changes elsewhere in the food system . Societal attitudes towards such changes will vary. Removal of land from agriculture can be culturally sensitive, and not necessarily compensated for by increases in other non-exclusive land uses and co-benefits such as flood alleviation and recreation (Hardaker et al., 2021;Kirchhoff, 2012). Moreover, although woodland-associated species should benefit from major woodland expansion, open habitat species may decline (Burton et al., 2018;Lamb et al., 2019;Warner et al., 2021), while increases in restored peatland area should benefit wading birds (Wilson et al., 2014). Wider assessments of such broad-scale NbS-driven land-use change impacts on agriculture and biodiversity are urgently needed .
There are uncertainties in our projections, notably regarding the net sequestration benefits of woodland creation, because impacts on soil carbon stores are poorly quantified. Some recent evidence suggests carbon stores in some organo-mineral soils may be adversely affected by woodland establishment, potentially reducing net sequestration (Friggens et al., 2020), although this may recover over longer time-scales (Vanguelova et al., 2019;Xenakis et al., 2021).
Given the urgency of the climate crisis, a precautionary approach is probably advisable where climate change mitigation is the primary driver of habitat change; in such cases, woodland creation should focus on mineral soils to give the greatest chance of net sequestration this century. However, carbon stores in organo-mineral soils vary widely, so some higher risk areas in our mapping may still be appropriate for woodland creation, especially when motivated by other goals such as flood alleviation or biodiversity conservation.
Semi-natural woodland established for the long term is likely to be better than production forestry for both climate and biodiversity (Burton et al., 2018;Warner et al., 2021) pressure (Gill & Morgan, 2010;Lamb et al., 2016). Coastal realignment has the highest costs of the NbS studied here, although the co-benefits such as reductions in flooding of coastal settlements are particularly important in building resilience to a changing climate (Kiesel et al., 2020;MacDonald et al., 2020).
The UK needs to urgently increase rates of carbon and biodiversity-focussed peatland restoration and woodland creation and improve the long-term use of HWP. However, avoidance of perverse outcomes can be made more certain by emphasising the 'nature' in NbS. Climate change mitigation cannot be used as an excuse for land management practices that damage biodiversity (Girardin et al., 2021). The twin climate and ecological crises are inextricably intertwined, and one cannot be resolved without the other (Leclère et al., 2020). Forthcoming changes to UK agricultural and rural support funding represents a pivotal opportunity to encourage positive change for the natural environment and progress towards net zero, ensuring public money does support public goods. NbS and other negative emissions technologies could be important in mitigating climate change but can only offer a meaningful contribution in the context of rapid and sustained reduction in the use of fossil fuels.

ACK N OWLED G EM ENTS
We thank Eleanor Tew and two anonymous reviewers for helpful comments on earlier drafts of this work. This study also benefitted from input from many people, our thanks Vanessa Amaral-Rogers,

CO N FLI C T O F I NTE R E S T
The authors declare they have no conflicts of interest.

AUTH O R S ' CO NTR I B UTI O N S
All authors were involved in study conception and designed the methodology; T.B.-L. led the data analysis and wrote the first manuscript draft. All authors contributed critically to the drafts and gave final approval for publication.

DATA AVA I L A B I L I T Y S TAT E M E N T
The woodland opportunity map is available from the University of Stirling