Forest regeneration on European sheep pasture is an economically viable climate change mitigation strategy

The net present value (NPV) of sheep pasture and forests receiving payments for carbon was calculated. This represents the benefits minus costs over a given time period; a positive NPV indicates net profit whereas a negative NPV signifies net loss. Following HM Treasury (2018), NPV was calculated over a 25 yr horizon with a discount rate of 3.5%. Rates of 7% and 10% were also applied to reflect higher discount rates in the private sector (Moran et al 2008, HM Treasury 2017). This range is consistent with existing literature (Nijnik et al 2013, Gilroy et al 2014, Warren-Thomas et al 2018).

NPV was calculated, without any subsidies, for sheep that lamb in spring. Income and cost values were taken from the 2018 John Nix Pocketbook for Farm Management, which is produced by the Andersons Centre of Farm Business Consultants (Redman 2017). The pocketbook provides estimates of costs and income for low-, average- and high-performing farms. Values are based on the UK Farm Business Survey using 2017/18 prices and presented separately for upland and lowland farms due to their distinct characteristics. Upland farms deemed 'Less Favourable Areas' are characterised by lower-stocking densities and poor-quality land. Lowland farms are generally more intensive and have better access to markets (Hardaker 2018). The values in Redman (2017) were used to establish the range of farms in the UK, with the resulting distribution resampled 10 000 times to account for uncertainty and variability within these parameters. The underlying distribution of farm productivity across the UK was unknown, and consequently a uniform distribution was adopted, bounded by the values for high- and low-performing farms in each case. Productivity and NPV were then calculated for each iteration as:

$$\begin{equation}{\text{SheepNPV}} = \mathop \sum \limits_1^n \frac{{\left( {O - V - F} \right)}}{{{{\left( {1 + r} \right)}^n}}}\end{equation} \tag{ 1 }$$

where O is output, the income from lamb and wool sales less depreciation costs, calculated using a fixed price for lamb and wool; V is variable costs, those that vary with output and include feed concentrates, veterinary and medicine, forage and miscellaneous costs (e.g. contract shearing etc); F is fixed costs, which do not vary with output, including labour, power and machinery, and overhead costs. NPV is calculated both with and without labour costs. Rent was excluded from the analysis because prices can be distorted by subsidies (Patton 2008, DEFRA 2018a). r is discount rate and n is year (up to 25 yr).

The quantity of claimable carbon sequestered was determined using look-up tables provided by the WCC, based on forest carbon models (Randle and Jenkins 2011). Carbon sequestration was estimated for native woodland over 25 yr and accounted for the soil carbon loss through disturbance, depending on the establishment method adopted (SI Text 1).

The amount of claimable carbon per ha was then multiplied by the price of carbon to estimate income from forests. Payments were assumed to be made each time carbon sequestration is verified: after the first 5 yr, then at 10 yr intervals. Forestry costs include carbon validation and registry use (table S1). Values were provided by Dr Vicky West, a member of the WCC executive board. When planting trees (Scenario 3, table 1), there are additional establishment costs; these were taken from Haw (2017) and cover the first 15 yr (table S1). NPV was then calculated as

$$\begin{align}{\text{ForestNPV}} & = \frac{{\left( {{B_{0 - 5}} - {C_{0 - 5}}} \right)}}{{{{\left( {1 + r} \right)}^5}}} + {\text{ }}\frac{{\left( {{B_{5 - 15}} - {C_{5 - 15}}} \right)}}{{{{\left( {1 + r} \right)}^{15}}}} \nonumber\\ & \quad + {\text{ }}\frac{{\left( {{B_{15 - 25}} - {C_{15 - 25}}} \right)}}{{{{\left( {1 + r} \right)}^{25}}}}\end{align} \tag{ 2 }$$

where B₀₋₅ is the benefit received from forestry (i.e. the income from carbon payments) in years 0–5, B₅₋₁₅ the benefit received in years 5–15, and B₁₅₋₂₅ the benefit received in 15–25 yr. B was based on carbon price, which we resampled across 10 000 iterations, under a uniform distribution bounded by the lowest price in the current UK carbon market (∼US$4 per tCO₂) and the social price of carbon estimated by the UK government (US$68 t⁻¹ CO₂eq). C is the costs associated with carbon payments, allocated to 5 or 10 yr verification intervals. Following the WCC guidelines, no net carbon sequestration was assumed on pasture prior to conversion.

Total costs were divided by forest area, which was resampled under a uniform distribution 10 000 times, since a farmer may decide not to turn their entire farm to forest. Forest size was assumed to range from 5 to 125 ha, since forests <5 ha do not qualify as 'standard' in the WCC and >125 ha is the largest farm considered by Redman (2017), representing an upper limit to forest conversion. r is the discount rate. In both forest scenarios, it was assumed that the landowner prepares documentation for the WCC rather than contracting a project developer and, as such, no additional cost has been applied.

The NPV of sheep pasture increases with lamb productivity (figure 1(a)). When the farmer and spouse are paid for their labour, farmers normally make a net loss with only the most productive farms breaking even. Upland farms make smaller losses (median: $6016 ha⁻¹) than lowland farms (median: $8010 ha⁻¹) because of their lower costs (figures 1(c) and (d)). However, upland farms are also less productive (median: 364 vs. 563 kg ha⁻¹).

Zoom In Zoom Out Reset image size

When labour is unpaid, sheep farming can generate net profit (figure 1(b)), although the median remains negative ($586 ha⁻¹). The difference between upland and lowland farms narrows, with median losses of $1150 ha⁻¹ for upland farms and $26 ha⁻¹ in lowlands, making lowland farms more profitable (figures 1(e) and (f)).

Varying discount rate produces the same patterns in NPVs. However, because future costs are reduced, the lowest NPVs are elevated as discount rate increases (figure S1 (available online at stacks.iop.org/ERL/15/104090/mmedia)).

The NPV of forests receiving carbon payments was calculated in relation to carbon market price and forest size (figure 2). The size effect arises because the verification costs of the WCC are fixed regardless of forest size. For a naturally regenerating deciduous woodland, at a low carbon price ($4 per tCO₂), forests larger than ∼25 ha make a profit, but all forests become profitable at higher carbon prices (≥$20 per tCO₂) (figure 2(a)). Deciduous woodland is projected to sequester a net total of 85 tCO₂eq ha⁻¹ over 25 yr. For conifer forests, changing carbon price has less effect on NPV (figure 2(b)). The largest conifer forests (>100 ha) break even at ∼$14 per tCO₂, although even under the highest carbon prices, profit is minimal. Carbon sequestration is also markedly lower (7 tCO₂eq ha⁻¹) than deciduous woodland over 25 yr.

Zoom In Zoom Out Reset image size

For naturally regenerated deciduous woodland, increasing discount rates lowers the NPV. Higher discount rates reduce the gradient of NPV over carbon price: with a discount rate of 10% under the highest current carbon market price ($20 per tCO₂), NPV only reaches ∼$170 ha⁻¹ (figure S2).

Planted forests were not profitable within the 25 yr time-horizon, except under the highest carbon prices, breaking even at a carbon price of $55 per tCO₂eq (figure 3). However, net carbon sequestration was high, at 147 tCO₂eq ha⁻¹.

Zoom In Zoom Out Reset image size

Varying discount rates had the same effect as seen in the natural regeneration; increasing discount rate reduces the effect of carbon price and the range in NPV (figure S3).

Sheep farming without subsidies is only profitable if farmer and spouse labour are unpaid (figure 1), even then, only the most productive farms break even (figures 1 and 4). In contrast, farmers can generate profits without subsidies by claiming payments for the carbon sequestered in deciduous forests that have naturally colonised former pasture land (figures 2 and 4). This is because of the minimal costs involved, particularly for forests larger than 10 ha when the carbon market price is high, and is consistent with previous studies showing that the natural regeneration of forest can be cost-effective (Macmillan et al 1998, Gilroy et al 2014, Evans et al 2015). However, slow carbon sequestration by naturally regenerated scots pine woodland means that this land-use conversion is only profitable when allowed (and is possible without deer fencing) across a large area and under high carbon prices (figures 2 and 4).

Zoom In Zoom Out Reset image size

Without financial aid, planting woodland to claim carbon payments is not financially viable within a 25 yr time frame under current market prices ($4-20 per tCO₂eq) because of high establishment costs, requiring a price of ∼$55 for the largest forests to become financially viable (figure 3). Though this exceeds the current market price of carbon, it is comparable to estimates of the social price of carbon. The social value of carbon is the economic cost generated by the additional emission of one tCO₂ (Nordhause 2017), and represents the market price that society should be theoretically willing to pay. Estimates range by three orders of magnitude, with the UK government valuing the social cost at $68 t⁻¹ CO₂eq (2009 prices) (Valatin 2011), which exceeds the break-even value for large planted forests.

Compared to a naturally regenerating woodland, plantations sequester more carbon because higher yield classes and tree densities can be achieved. Consequently, NPV increases more rapidly with increasing carbon price, making plantations a more favourable climate change mitigation strategy. If a minimal disturbance scenario is adopted and the ground is prepared by hand, the net carbon sequestration and therefore NPV of plantations increases further. Moreover, planting trees produces timber of a higher quality, giving greater potential for the wood to substitute more GHG intensive materials or fuels, further enhancing the climate change mitigation potential (Cannell 2003, Morison et al 2012). We do not account for timber revenue however, if managed effectively, this can generate future economic return (Heaton et al 1999, Nijnik et al 2013, Hardaker 2018) and could generate different outcomes between plantations and natural regeneration over the longer-term. Carbon payments could be implemented as a means of revenue for farmers awaiting forest establishment and timber returns.

Plantations become more profitable if tree planting is subsidised. In England, the government-funded Woodland Carbon Fund (WCF) grant currently covers 80% of the establishment cost (Forestry Commission 2018b). Applying this grant allows larger forests to profit at carbon prices ∼$12−14/tCO₂eq, and NPV to increase as high as $500 ha⁻¹ under the highest current market price for carbon (figure S4). The benefits of forestry are diminished when subject to higher discount rates. However, the market price of carbon is expected to increase over time as restrictions on emissions are tightened (Valatin 2011), countering the effects of discounting. In addition, payments for carbon are often received upfront (Vicky West, Forestry Commission, pers. comm.), meaning they would not be subject to de-valuation over time.

5.1.1. Where would pasture conversion to forest be feasible?

Planting exotic conifers such as Sitka spruce would generate a higher revenue from carbon payments because they are over three-times more productive than native deciduous trees (Read et al 2009). However, maximising carbon sequestration and associated payments are unlikely to be the sole objectives of tree planting, and both pasture and native woodland provide uncosted additional ecosystem services (table 2). Conversely, plantations of exotic tree species are considered ecologically and environmentally damaging (Brockerhoff et al 2008, Bunce et al 2014).

Pasture is managed primarily for the provision of forage for grazing animals and this often comes at the cost of other ecosystem services. For example, fertiliser application and increased stocking densities can reduce biodiversity (Petit and Elbersen 2006, Firbank et al 2008, Emmerson et al 2016). Carbon storage is the only service explicitly considered in this study, but WCC projects provide multiple co-benefits, including recreation, water pollution regulation and flood control (eftec 2015; table 2). Consequently, the social and environmental benefits of forests in Great Britain was estimated at £2 billion a year (currently ∼US $2.5 billion; table 3; eftec 2015, ONS 2017), although this may not scale linearly with increased forest area. Regardless of their theoretical value, the relevance of ecosystem services depends on which ones, and by how much, governments choose to support. We may see subsidies re-structured to pay 'public money for public goods' (e.g. DEFRA, 2018c) but there will inevitably be trade-offs among services (Rodríguez et al 2006).

5.2.1. Potential barriers to land conversion