We decided to represent Carbon in term of CO2 equivalent and analyse the change that could occur in four different scenario of land use change, especially considering the expansion of the arable land and verify whether the woodland expansion targets for 2020, could help to reduce the emission quantity.
In the baseline case, Carbon (kg C) stocks are calculated at the 25 x 25 m grid resolution and compartmentalised into below ground (soil) carbon and above ground (vegetation) carbon.
In addition CO2 flows (emissions) from livestock are also calculated.
Generating the Soil carbon map: The values used were the median of carbon stocks values for the first 40cm derived from the Scottish Soils Database. This contains the National Soil Inventory of Scotland (NSIS) profile samples collected on a regular 10 km grid (Lilly et al., 2010). Values for each pixel in Aberdeenshire using a hybrid GAM-geostatistical 3D model (Poggio and Gimona, 2014,).
Soil carbon values will change over a long period to reach a new equilibrium after the land cover changes from one use to another. For example, if there is a land use change resulting in net carbon loss this was assumed to take place over 50 years whilst a land use change resulting in carbon sequestration requires a longer duration, to reach equilibrium, of 100 years (NEA, 2011). Therefore, we calculated the net change (positive or negative) for each scenario for the duration between 2014 and 2050 (36 years) assuming a linear model, and assuming that change due to improved capability would occur on average after 18 years.
Generating the Vegetation carbon map: We used values for different vegetation based on the data in Cantarello E. et al., 2011.
Generating a livestock carbon emission map: Animals emit greenhouse gas (CO2 equivalent) through the digestive process. The number of livestock for each scenario was predicted using a GAM geostatistical model based on several covariates including land use type, slope, altitude and the two coordinate (X and Y) from 25 meters resolution grid. The model was trained on current census data. Any change in land use corresponds to a change in livestock quantity, and therefore in per capita emissions, a change in term of Carbon loss or gain.
Total Emissions: Changes in carbon stocks in all compartments are converted into CO2 equivalent. Emissions are also assumed to result from fertilisers (6 Kg of CO2eq per kg of N input) and their production. An extra 15% of emissions are assumed on agricultural land due to energy expenditure in cultivation.
We then calculated the total net flow of carbon in term of CO2 by adding changes in the soil organic carbon, the vegetation carbon and emissions carbon grids together. This was displayed for each sub-catchment in Aberdeenshire.
We used the InVEST model to Estimate Water Quality focussing on Nitrogen and Sediment Retention. This was only carried out for the Huntly Local Focus Area because a) modelling at the regional scale is currently too time consuming and b) we want to explore how stakeholders in the LFAs evaluate land use change when provided with information on the consequences for other services such as water quality. We used the literature to determine the efficiency that each land use type retains nitrogen and sediment. In the case of nitrogen, the predictions take into account the fact that crops receive more inorganic N but also retain it in the crop more effectively than woodland. Woodland only receives atmospheric N. Similarly for sediment retention, the model takes into account the land use type as well as slope. The map outputs therefore show the Nitrogen retention capability of each sub-catchment under each of the land configuration options generated by the tool.
Land cover around Huntly is dominated by Arable and Improved Grassland in the east and south, changing to heather and heather grassland and forest moving to the west. More than half of this study area lies inside Nitrate Vulnerable Zones. Therefore it is pertinent to explore the consequences for nutrient transport from prime land expansion.