Quantifying impacts of peatland-based windfarm development on aquatic carbon and nutrient exports

Many onshore windfarms in Scotland are being built on peatlands. As a consequence, the impacts of development activities on the peatland and downstream environments are an important emerging issue. For example, a reduction of the quantity of carbon (C) stored in the peatland and increased phosphorus (P) and nitrogen (N) concentrations in streams may cause changes in the peatland carbon sink strength and aquatic carbon emissions, and exceedance of EU Water Framework Directive (WFD) P threshold values, respectively. To further assess the impacts from peatland-based windfarm development, the following four research questions were posed: 1. Which specific development activities have the greatest impact on concentrations of C, N and P in streams? How long does it take for C, N and P concentrations to return to baseline conditions? 2. Which specific windfarm development activities caused the greatest aquatic export of carbon and nutrients? 3. What are the controls on concentration and export of aquatic C, N and P in streams draining peatland developments? 4. How effective are site-derived adsorbent materials and an iron ochre at preventing phosphorus from entering upland streams on peatland? Presented in this thesis, are the results from 3-weekly spatial streamwater sampling of 18 sub-catchments (ranging 0.03 - 12 km2) – from October 2011 to March 2013 – which covers the majority of the peatland-based development and one year beyond, of the Whitelee windfarm Phase 2 extension 16 km south of Glasgow, Scotland. Dissolved and particulate organic carbon (DOC/POC), soluble reactive phosphorus (SRP) and total oxidised nitrogen (TON) concentrations were measured and exports were calculated from stream gauging. The spatial catchment controls on stream water chemistry were investigated through a multiple linear regression analysis. Fraction of sub-catchment area that was subject to forest felling was found to be significantly positively correlated with concentrations and exports of DOC, POC, SRP and TON; with TON specifically by felling that had occurred more than one year previously. Density of tracks / cable trenches / turbines, and fraction non-forested area in general exerted a negative control on exports. Evidence of SRP attenuation processes in addition to dilution were found in some streams. While neither river sediment nor track or cable trench gravels were very effective at removing SRP, from the results of batch adsorption experiments, a sample of mineral subsoil was. SRP was attenuated in areas where there was a greater proportion of peaty gleys, in the riparian zone along the main stream, and the outcome of the batch adsorption experiments points to attenuation of SRP onto mineral subsoils, where present. Comparing long-term (seven year) time series of the original Whitelee windfarm directly north of the site, with the extension site (this study) it was possible to estimate the time required for recovery to pre-development concentrations. Mean annual [DOC] increased to 31.4 mg l-1 during development of the extension site. While it took five years for the DOC to recover at the original site, in the year following development at the extension site, mean annual [DOC] was still above the baseline. Mean annual [POC] increased to a maximum of 2.6 mg l-1 during development of the extension site, and as with the original site it returned to the baseline the year following development. The mean annual [SRP] increased to a maximum of 50.5 μg l-1 during development, leading to a reduction in water quality from ‘good’ to ‘moderate’ based on comparison with Environmental Quality Standards (EQS). It took four years for the SRP to recover at the original site, and mean annual [SRP] was still above baseline at the extension site in the year following development. The streamwater C, N and P has not been measured as close to, and for such a range of, windfarm development activities. This research has demonstrated links between windfarm development, forest felling in particular, and increased concentrations and exports of DOC, POC, SRP and TON in streams. While the increases were localised within the site itself, investigation of attenuation processes point to peaty gley soils along the main stream, and tracks, cable trenches and turbines – and their associated settling lagoons and silt fences – as potentially mitigating the increased dissolved carbon and nutrients. Outcomes of this research are suggested improvements to the guidance for developing peatlands, especially with regards to monitoring streamwater carbon for a sufficient period post-development, and for the removal of brash on peat soils. Further research could investigate the use of adsorption materials to mitigate phosphorus mobilisation from brash sources to streams, over a time scale of at least two years at the field scale. Laboratory-based environmental fate studies of the increased streamwater carbon could investigate the rates of CO2 efflux with time and with varying concentrations of nutrients, which would help to improve the Scottish Government's carbon calculator estimates of carbon loss from peatland-based windfarm development.