Climate change: evidence review in Agriculture, Forestry, Land Use, Waste

3 Potential wider impacts GHG mitigation in agriculture, land use, land use change and forestry

3.1 Qualitative evidence

Table 4 provides an overview of the wider impacts of the GHG mitigation options (detailed narratives can be found in Appendix A1 and Appendix A2). The scores show the direction and magnitude of impact (positive denoting favourable impact) and the colour scale provides an assessment of the robustness of the available scientific evidence (weak evidence refers to situations where there is limited availability of evidence and/or there are conflicting findings, while robust evidence refers to conclusive evidence). The majority of the WIs were positive or neutral, with also a high number of variable impacts (i.e. positive and negative impacts both possible), but there are no strongly negative impacts.

There is evidence on co-benefits potentially arising from all MOs. Multiple robust co-benefits are related to from on-farm renewable energy, precision farming, AD, agroforestry, optimal mineral N use, livestock health, reduced livestock product consumption, afforestation and peatland restoration, indicating the potential for delivering co-benefits in a range of policy areas. Strong and robust positive effects were found for AD on resource efficiency, low emission manure storage and application on NH ₃ emissions, reduced livestock product consumption on human health, afforestation on air quality and on flood management and peatland restoration on soil quality and biodiversity.

Adverse impacts were associated with eight MOs, though evidence on some of these was limited and therefore the impacts are uncertain. Negative impacts with moderate or robust evidence were found for on-farm renewables, AD, improving livestock health, reduced livestock product consumption, afforestation and peatland restoration. On-farm renewables can have a small unfavourable impact on land use by occupying areas could be used for other purposes. Anaerobic digesters produce air pollutants ( NO _x and PM) in the combustion process. Improving livestock health might negatively affect biodiversity if habitats are altered to reduce vector borne diseases (e.g. field drainage to reduce mud snail populations, which act as a vector for liver fluke) and also from certain medications released to the environment via livestock excreta. Reduced livestock product consumption might lead to increased pesticide use due to higher vegetable consumption. Afforestation might result in increased tick populations near grazing livestock, increasing the risk of tick-borne diseases. Finally, increased nitrogen and phosphorous leaching is possible in the first years of peatland restoration. Careful planning and implementation are needed to minimize these effects.

Several impacts were variable, calling for specific implementation to maximise the benefits while reducing adverse impacts. These variable effects were mostly associated with reduced livestock product consumption, afforestation, low emission storage and application of manure and peatland restoration. In most of these cases the reason behind the variable impact was that either the MO or the WI is an aggregation of varied technologies or impacts, respectively. For example low emission storage and application of manure includes various technologies related to manure storage and manure spreading; these technologies have different effects on the environment. In other cases the effects on a WI greatly depends on the particularities (e.g. location, species, management, ownership) of implementation, for example covering the digestate from AD can mitigate the otherwise increased NH ₃ emissions, and the location of afforestation and peatland restoration projects can define whether the cultural effect is positive or negative.

The most uncertain MOs (i.e. those MOs with the highest number of WIs supported only by weak or moderate evidence) were reduced livestock product consumption, livestock health and optimal soil pH, and, to a lower extent, low emission storage and application of manure and more legumes. On the other hand, WI's related to afforestation and optimal use of mineral nitrogen seemed to be the best explored. This is to be expected for the former three MOs, as research has relatively recently started focusing on their GHG effects (either globally, like reduced livestock product consumption and livestock health, or in the Scottish context, like optimal soil pH). Further research could help in closing these knowledge gaps. Highlighted areas are soil pH impacts on water quality, soil quality and biodiversity, the influence of improving livestock health on pesticides and human health, and the effects of reduced livestock product consumption on the structure of agricultural production with particular emphasis on soil quality, biodiversity, animal health and welfare, employment, social and cultural impacts.

Many MOs can have co-benefits in relation to air and water quality, resource efficiency and human health, and these co-benefits can be promoted by integrated approaches in these policy areas. The WIs that had the highest number of variable co-effects were soil quality, flood management and water use, household income and human health. Again, policy integration of these areas and GHG mitigation is key in maximising the net benefits.

The impact categories least affected by the MOs considered air quality other than NH ₃, NO _x or PM, cultural impacts and crop health (four to five MOs impacting on any one of them). However, the magnitude of these impacts emphasises the importance of integrated approaches. For example, the cultural impacts of afforestation or peatland restoration requires the consideration of both environmental and social aspects in planning and management, and the likely impact of agroforestry on crop health calls for developing capacity to incorporate crops pest and diseases assessment in local and regional decisions on agroforestry.

Four WIs were found to be the most uncertain (i.e. with the highest number of MOs with weak or moderate evidence on these impacts): household income, consumer and producer surplus, employment and cultural impacts. On average the environmental impacts were more robust, with the least uncertainty around NH ₃ and NO _x emissions and resource efficiency.

Table 4 Summary of the WIs of the GHG MOs

3.2 Quantitative aspects: models and tools and valuation

MO implementation typically involves trade-offs and synergies with other policy goals. Evidence is required to evaluate these and to identify how impacts are attributable to different policies. Modelling the potential impacts is an important part of such an exercise, along with establishing the monetary values of the WIs to serve as a common metric between them.

This section summarises the modelling capacity for capturing the particular effects of the individual MOs on the WIs and the available monetary values. Model suitability is summarised in Table 5 - Table 9. The list of the WIs where currently robust valuation is available is presented in Table 10, a full list with monetary values can be found in Table 59. More detailed model and monetary value descriptions are provided in Appendix A2 and A3, respectively.

There is widely applied modelling capacity for most of air and water quality aspects. UK-specific models are available to estimate most of the air and water quality effects of changes in farm management and changes in land use related to the MOs. No suitable models were found for some water quality aspects (e.g. heavy metal pollution effects of optimal soil pH and faecal microorganism effects of low emission manure storage and application). Monetary values (used by the UK Government) are available for the major air pollutants ( NH ₃, NO _x, PM), however, these only include health impacts of secondary PM formation, and do not account for other health impacts or any environmental impact (e.g. acidification, eutrophication). Some monetary values exist for water quality impacts from nitrogen pollution and general water quality status (the former is location specific). No monetary values were found for phosphorous pollution of water.

Soil quality modelling is overwhelmingly soil carbon modelling, since this is an important component of structure quality. Other aspects of soil quality are not normally included in the relevant models (for example physical and hydrologic), making it unfeasible to estimate the quantitative impacts of some MOs (Anaerobic digesters, More legumes, Manure storage and application) on Soil quality. The valuation of soil quality is possible through the impacts on agricultural productivity (i.e. using market values).

The expected impacts of MOs on flood management and water use can be quantitatively assessed with hydrological models. On the valuation side existing spatially explicit property damage values can be used to value flood risk.

Larger scale land use changes related to afforestation and reduced livestock product consumption can be predicted using models (e.g. econometric or agent based models), which capture the economic drivers (subsidies, markets) and the biophysical constraints of land use. No models were found to quantify the changes potentially induced by on-farm renewables and more legumes.

Existing models are capable of estimating the biodiversity impacts of MOs which result in land use change (reduced livestock product consumption, afforestation and peatland restoration), but farm management changes (related to MO1- MO9) cannot currently be assessed. Monetary values for biodiversity are based on the way habitat improvements change the status of charismatic and non-charismatic species.

No models or tools were found to quantify the WIs on Animal health and welfare and crop health. If quantitative estimates were available, the value of the animal and crop health effects could be captured by the production changes. Existing animal welfare monetary values relate to livestock systems (e.g. free range versus caged) and cannot be linked to welfare outcomes and therefore to the management changes implied by the MOs assessed in the report.

Economic models (e.g. computable general equilibrium ( CGE) models, Input-Output ( IO) models and Social Accounting Matrix ( SAM)) can quantify three WIs (Household income, Consumer and producer surplus and Employment and part of Resource efficiency). But these are more suited to assess larger scale impacts than those occurring at the farm scale.

For human health, dietary models are well-developed, as are those relating health to air and water quality related impacts. . But a number of more specific health effects cannot be currently modelled including zoonoses and antimicrobial resistance. There are existing estimates for the monetary value of some human health impacts.

Social and cultural impacts are difficult to quantify and no models or tools were found apart from those to quantify the recreational benefits of green space. Evidence is also limited on the monetary values of the cultural impact; existing values are based on improvements to habitats on 'sense of place'. Currently there is no evidence on the valuation of social impacts.

Table 5 Models for air quality assessment ( WIs 1-4, see model description in Appendix A2)

		WI1	WI2	WI3	WI4
		Air quality: NH3	Air quality: NOx	Air quality: PM	Air quality: Other
MO1	On-farm renewables	No impact expected	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM
MO2	Precision farming	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM Farmscoper	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM
MO3	Optimal soil pH	EMEP4UK GAINS/ UKIAM	No impact expected	No impact expected	No impact expected
MO4	Anaerobic digesters	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	No impact expected
MO5	Agroforestry	DNDC MODASS- THETIS EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	No impact expected
MO6	More legumes	EMEP4UK GAINS/ UKIAM Farmscoper	EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	No impact expected
MO7	Optimal mineral N use	EMEP4UK GAINS/ UKIAM Farmscoper	EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	No impact expected
MO8	Manure storage and application	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM Farmscoper	No impact expected	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM	EMEP4UK (regional) SCAIL (local) GAINS/ UKIAM
MO9	Livestock health	EMEP4UK GAINS/ UKIAM Farmscoper	No impact expected	No impact expected	No impact expected
MO1O	Reduced livestock product consumption	EMEP4UK GAINS/ UKIAM	No impact expected	No impact expected	No impact expected
MO11	Afforestation	FOREST- DNDC MODDAS- THETIS	EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM	EMEP4UK GAINS/ UKIAM
MO12	Peatland restoration	No impact expected	No impact expected	No models/tools found	No impact expected

Table 6 Models for water and soil quality assessment ( WIs 5-8, see model description in Appendix A2)

		WI5	WI6	WI7	WI8
		Water quality: Nitrogen leaching	Water quality: Phospho-rous	Water quality: other	Soil quality
MO1	On-farm renewables	No impact expected	No impact expected	No impact expected	Windfarm carbon calculator (wind turbines). CARBINE (biomass fuel crops)
MO2	Precision farming	LUCI, ADAS Wales, Farmscoper, NIRAMS	LUCI, ADAS Wales, Farmscoper	LUCI (sediment), ADAS Wales (pesticides), Farmscoper (pesticides)	Spacsys
MO3	Optimal soil pH	LUCI	LUCI	No models/tools found	Century
MO4	Anaerobic digesters	ADAS Wales, Farmscoper, LUCI, NIRAMS	LUCI, ADAS Wales, Farmscoper	No impact expected	No models/tools found
MO5	Agroforestry	DNDC, LUCI, NIRAMS	DNDC, LUCI, ADAS Wales	LUCI (sediment), ADAS Wales (pesticides), Farmscoper (pesticides)	DNDC, CARBINE (soil carbon stocks)
MO6	More legumes	Farmscoper, NIRAMS	No impact expected	No impact expected	No models/tools found
MO7	Optimal mineral N use	LUCI, ADAS Wales, Farmscoper, NIRAMS	LUCI, ADAS Wales, Farmscoper	No impact expected	No impact expected
MO8	Manure storage and application	ADAS Wales, Farmscoper, LUCI, NIRAMS	LUCI, ADAS Wales, Farmscoper	No models/tools found	No models/tools found
MO9	Livestock health	ADAS Wales, Farmscoper, LUCI	LUCI, ADAS Wales, Farmscoper	ADAS Wales (veterinary medicines)	No impact expected
MO10	Reduced livestock product consumption	LUCI, ADAS Wales, Farmscoper, NIRAMS	LUCI, ADAS Wales, Farmscoper	ADAS Wales (veterinary medicines)	DNDC, CARBINE (soil carbon stocks)
MO11	Afforestation	LUCI, NIRAMS	No impact expected	No models/tools found	CARBINE (soil carbon stocks)
MO12	Peatland restoration	LUCI	LUCI	No models/tools found	LULUCF Inventory (soil carbon stocks)

Table 7 Models for assessing flood management and water use, land cover and land use, biodiversity and animal health and welfare ( WIs 9-12, see model description in Appendix A2)

		WI9	WI10	WI11	WI12
		Flood management, water use	Land cover and land use	Biodiversity	Animal health and welfare
MO1	On-farm renewables	No impact expected	No models/tools found	No models/tools found	No impact expected
MO2	Precision farming	IHMS, SALTMED	No impact expected	No models/tools found	No models/tools found
MO3	Optimal soil pH	IHMS, SALTMED	No impact expected	No models/tools found	No models/tools found
MO4	Anaerobic digesters	No impact expected	No impact expected	No impact expected	No impact expected
MO5	Agroforestry	IHMS, SALTMED, LUCI	LULUCF Inventory	SNH's IHN, Eco-Serve GIS, InVEST, AgBioscape, LUCI	No models/tools found
MO6	More legumes	No impact expected	No models/tools found	AgBioscape, SRUC's Biodiv Calc, InVEST, Eco-Serve GIS	No impact expected
MO7	Optimal mineral N use	No impact expected	No impact expected	No impact expected	No impact expected
MO8	Manure storage and application	No impact expected	No impact expected	No impact expected	No models/tools found
MO9	Livestock health	No impact expected	No impact expected	No models/tools found	No models/tools found
MO10	Reduced livestock product consumption	IHMS, SALTMED	Spatial econometric and agent based models	SRUC's Biodiv Calc, AgBioscape	No models/tools found
MO11	Afforestation	IHMS	Spatial econometric and agent based models, LULUCF Inventory	SNH's IHN, Eco-Serve GIS, InVEST, AgBioscape	No models/tools found
MO12	Peatland restoration	IHMS, SALTMED	LULUCF Inventory	SNH's IHN, SRUC's Biodiv Calc, Eco-Serve GIS, InVEST	No models/tools found

Table 8 Models for assessing crop health, household income, consumer and producer surplus and employment ( WIs 13-16, see model description in Appendix A2)

		WI13	WI14	WI15	WI16
MO1	On-farm renewables	No impact expected	IO/ SAM, CGE	IO/ SAM, CGE	IO/ SAM, CGE
MO2	Precision farming	DSSAT/ APSIM	CGE	CGE	CGE
MO3	Optimal soil pH	DSSAT/ APSIM	CGE	CGE	No impact expected
MO4	Anaerobic digesters	No impact expected	IO/ SAM, CGE	IO/ SAM, CGE	IO/ SAM, CGE
MO5	Agroforestry	No models/tools found	No impact expected	No impact expected	No impact expected
MO6	More legumes	ROTOR, LUSO	No impact expected	No impact expected	No impact expected
MO7	Optimal mineral N use	No impact expected	No impact expected	No impact expected	No impact expected
MO8	Manure storage and application	No impact expected	CGE	No impact expected	CGE
MO9	Livestock health	No impact expected	No impact expected	No impact expected	No impact expected
MO10	Reduced livestock product consumption	No impact expected	IO/ SAM, CGE	IO/ SAM, CGE	IO/ SAM, CGE
MO11	Afforestation	No impact expected	IO/ SAM, CGE	IO/ SAM, CGE	IO/ SAM, CGE
MO12	Peatland restoration	No models/tools found	IO/ SAM, CGE	No impact expected	No impact expected

Table 9 Models for assessing resource efficiency, human health, social impacts and cultural impacts ( WIs 17-20, see model description in Appendix A2)

		WI17	WI18	WI19	WI20
		Resource efficiency	Human health	Social impacts	Cultural impacts
MO1	On-farm renewables	AgRECalc, AGRILCA	See air quality models	No models/tools found	No impact expected
MO2	Precision farming	AgRECalc, AGRILCA	See air and water quality models	No models/tools found	No impact expected
MO3	Optimal soil pH	No impact expected	No models/tools found	No impact expected	No impact expected
MO4	Anaerobic digesters	AgRECalc, AGRILCA	See air and water quality models	No models/tools found	No impact expected
MO5	Agroforestry	No impact expected	See air and water quality models	No impact expected	No models/tools found
MO6	More legumes	AgRECalc, AGRILCA	No impact expected	No impact expected	No impact expected
MO7	Optimal mineral N use	No impact expected	See air and water quality models	No impact expected	No impact expected
MO8	Manure storage and application	AgRECalc, AGRILCA	No models found for acid risk assessment; also see air and water quality models	No impact expected	No impact expected
MO9	Livestock health	AgRECalc, AGRILCA	No models found for zoonosis and antibiotic risk assessment; also see air and water quality models	No impact expected	No impact expected
MO10	Reduced livestock product consumption	IO/ SAM, CGE	DIETRON, PRIME	No models/tools found	ORVal
MO11	Afforestation	IO/ SAM, CGE	Effects of forest-related exercise on health: no models; also see air and water quality models	No models/tools found	ORVal
MO12	Peatland restoration	No impact expected	No models/tools found	No models/tools found	ORVal

Table 10 Robust monetary values of the wider impacts

	Wider impact	Included in the value	Reference
WI1	Air quality: NH 3	Cost of morbidity and mortality arising from secondary PM formation. Recommended use for UK national evaluation. 2015 prices.	Defra (2015)
WI2	Air quality: NO x	Cost of morbidity and mortality arising from secondary PM formation. Recommended use for UK national evaluation. 2015 prices.	Defra (2015)
WI3	Air quality: PM	Cost of morbidity and mortality from direct exposure and value of building soiling. Recommended use for UK national evaluation. 2015 prices.	Defra (2015)
WI4	Air quality: other: sulphur dioxide	Cost of morbidity and mortality from direct exposure, from secondary PM formation and value of building damage. Recommended use for UK national evaluation. 2015 prices.	Defra (2015)
WI9	Flood management	Estimated flood damage values are available in the SEPA Flood Risk Management Strategies.	SEPA (2015)
WI18	Human health	Impact on both life years and quality of life based on willingness to pay.	(Glover and Henderson 2010)

A1.1 Research gaps

The review revealed certain areas where the evidence about likely adverse impacts is not robust. Improving the evidence base in such cases can ensure that policies minimise these effects while maximising GHG benefits. WIs that can be either co-benefits or adverse impacts depending the way the MO is implemented also require further investigation to ensure that total benefits are maximised. However, research capacity in terms of modelling the WIs is not equally well developed for all MOs and WIs, as detailed in Section 3.2. Table 11 presents those MO- WI combinations in red where there is a highlighted research need but inadequate modelling capacity was found. This emphasizes the need for investment in further research and development of modelling capability. The four wider impacts most affected were soil quality, biodiversity, animal health and welfare and human health. Orange cells in the same table indicate those areas where the highlighted research need can be more readily answered by existing models. This mainly relates to three MOs: optimal soil pH, reduced livestock product consumption and afforestation.

The nature of greenhouse gas effect implies that GHG mitigation is not a spatial issue. However, most of the co-benefits and adverse effects are highly sensitive to the location of the land use or farm management change. To maximise the net benefits at regional or national level, spatially explicit integrative approaches are needed.

Furthermore, decision support tools which integrate the different environmental, economic and social aspects at a high level and offer standardised and more comprehensive appraisal could be useful tools for policy makers.

Table 11 Areas highlighted for further research