The Challenges of Managing Water for Wetland Ecology, Flood Mitigation and Agriculture in the Upper Lunan Water, an Intensive Arable Catchment in Scotland

The “Water for All” project has aimed to develop a multi-disciplinary science case for adaptive management through a Payment for Ecosystem Services (PES) scheme in the Lunan Water, a lowland catchment in Scotland. Management needs for high and low flows, standing water levels and flow routing to/from high nature value mesotrophic wetlands were appraised. A key part has been the development of a real time hydrological-hydraulic model of the upper Lunan Water as an aid to management and scenario analysis especially with respect to existing and potential hydraulic structures. This provides better working knowledge and forecast-based simulations of high or low flow situations for catchment management planners, farmers, riparian owners and other local stakeholders. Engagement with local users and residents has included surveying Willingness To Pay (WTP) for hydraulic management as a function of governance mechanisms, development of a catchment management group , and interviews and workshops with riparian and other land-users. The work has highlighted the joys and sorrows of seeking to develop a PES approach and lessons to be learnt in project management, promotion of multiple benefits, catchment-scale water governance and the vices and virtues of “benign neglect”.


Introduction
Watershed or catchment management encompasses a complex tapestry of multistakeholder interests and water rights, cultural and historical norms, uncertainty of open system processes, and economic and environmental change [1,2]. Moreover, the technical challenges to delivering across multiple objectives are significant. These include management of hydraulic structures, calibration of hydrological models of acceptable accuracy and precision, the need for real time or forecasted data streams and understanding of impacts of water management actions on ecosystems. A range of approaches has developed in recent years that aim to provide an integrated, consultative approach to natural resource management, among which

Environmental Management
Payment for Ecosystem Services (PES) has emerged as a promising approach [3,4]. PES schemes have been envisaged as an efficient way to ensure the provision of local public goods, such as water [5,6]. Such schemes have the potential to increase the availability of ecosystem services for beneficiaries, whilst compensating the providers of such services. Examples of successful PES schemes with a focus on water include the Vittel water catchment area in France [7] and the "Upstream Thinking" scheme in SW England [5]. However, some aspects of catchment water management, such as the widely distributed and ill-defined benefits, and legacy issues, can undermine local participation and willingness to pay [3,7].
This chapter outlines the evolution of an engineered water management strategy ("Water for All") for the Lunan Water catchment area, Angus, Scotland (see Figure 1). The project originally envisioned alleviation of winter flooding in the upper catchment, summer low flows and water shortages in the lower catchment, and nutrient/sediment pollution in wetland areas, to be managed as a PES scheme through local management and funding. The key engineered element was the introduction of remotely managed hydraulic controls (e.g. a tilting weir) to deliver the envisioned benefits.
Integrated analysis was made up of three main components: • Hydrological analysis focused on deriving rainfall-runoff relationships for inflowing streams and a hydraulic model of the main channels and standing waters. This allowed simulation of upstream water levels and partitioning of flows as a function of existing or proposed management of hydraulic structures. It made use of long-term monitoring and survey data and the hydraulic modelling tool HECRAS 5.0.7; • Ecological analysis considered the vegetation characteristics of the wetland using the UK National Vegetation Classification scheme [8], and the likely  mixing behaviour of water from the river with other sources contributing to the wetlands, using End Member Mixing Analysis [9]; • Governance and stakeholder analysis focused on regular engagement through a stakeholder group (Lunan Catchment Management Group, which was set up at the start of the project in 2017), a survey of Willingness to Pay (WTP) for the tilting weir option, and interviews discussing the uncertainties and governance gaps that might impede implementation.

Ecosystem services provision potentially influenced by water management
At the start of the "Water for All" project, a workshop was run which compared the Lunan Water with the River Leven, another lowland catchment in eastern Scotland. Eight main elements of Ecosystem Services Provision that could be affected by water management in the two catchments were identified. These are summarised in Table 1. Of these, those marked as having a high potential for influence in the Lunan Water became the focus of attention for that catchment. A separate project [10] explored the potential of flushing for control of algal blooms in Loch Leven, on the River Leven, using the Lake algal model PROTECH [11].
The key risk factors considered for improved management were: • Flood risk mitigation. A tilting weir, as a lateral structure on the common lade, could be used to modify flows from the upstream lakes to achieve a lower base level in the lakes prior to winter high flows, and provide managed release. Dredging of the mill lade might also be used to achieve this, but with less flexibility and greater potential for adverse impact; • Risk of eutrophication of floodplain wetlands. Phosphorus, nitrogen and sediment discharge from the lakes displayed strong annual variation. Inflow of sediment/nutrient-rich waters can lead to loss of biodiversity in wetlands [12]. A tilting weir, as a lateral structure on the common lade, could help divert these sediment/nutrient rich waters away from floodplain wetlands at key times;

Influences of hydraulic management on Ecosystem Service Provision
• Risk of low flows. This could cause damage to ecological status of the river and economic impact to downstream irrigators. An in-line tilting weir at the lake outlet to the common lade could act as a penning structure to facilitate retention of water in early summer in the lakes, to provide water for abstraction and maintain low flows. This element of the analysis was considered in detail elsewhere [13].

Hydrological analysis
The surface water sub-catchment areas for the upper Lunan Water have been separated, for water balance purposes, into the areas used for hydraulic modelling, which are shown in Figure 2. The areas of each of these sub-catchments were used to scale measured and modelled flows from monitored sub-catchments where water flow and water quality monitoring occurred. These are the Balgavies Burn sub-catchment, at Westerton (area = 4.4 km 2 ), which generates real time stage and flow data and the Baldardo Burn sub-catchment at Wemyss (area = 2.4 km 2 ), which generates water level data.
The "Water for All" project focused on the outlet zone of Balgavies Loch (Figure 3), where the Lunan Water discharges into a partially confined common channel (lade). This lade controls water delivery to a mill or returning to the river, controlled by an existing engineered gated weir, and water also flows from the river into to its lateral floodplain wetlands (Chapel Mires) via a non-engineered spillage zone, which replaced a now-blocked engineered spillway in the 1970's. The ecological value of these wetlands may be vulnerable due to ingress of sediment and nutrients from the river [12,14]. Hydro-ecological assessment involved developing a model of the upper Lunan Water and the operation of this outlet zone. The aim of the modelling is to generate a time series of historic, real-time and forecast based surface water and ground water flows and water levels in the upper Lunan Water catchment, whose area is defined by the surface water outlet at Milldens bridge (Grid Reference NO 354526 750566). These could then be used to provide triggers or other guidance for hydraulic management. This model is now running in a real time and forecasting mode. 1 The hydrological-hydraulic model has the capability to simulate water levels in the upper Lunan Water lakes as illustrated in Figure 4, and flow routing between the Common Lade and the Lunan Water.   The model calibration provided the basis for some level of certainty about the impact of existing and potential hydraulic management. The scenario analysis pointed to the rather low impact of installation of a tilting weir on upstream water levels, whether this were to be installed either above or below the confluence of the Balgavies Burn and the Common Lade. However, it also pointed to a larger impact on water levels if dredging/vegetation management or other interventions that affect the Manning roughness coefficient downstream of Balgavies Loch are undertaken, in conjunction with tilting weir installation. It also shows that flow routing could be significantly impacted by a tilting weir, or by reinstatement of the blocked spillway downstream of the current spillway (see Figure 3), especially if combined with local dredging and/or vegetation management.

Ecological analysis
Rescobie and Balgavies Lochs, covering 1.78 km 2 , support over 60 species of breeding birds and with their surrounds form a Site of Special Scientific Interest (SSSI). There are many aquatic species such as Menyanthes trifoliata (Bogbean) and Utricularia australis agg. (Bladderwort) that occur in shallow water and any significant change in water levels could affect them. In the area south of the Lunan Water (Chapel Mires -see Figures 5 and 6) there is a complex mosaic of open water, willow scrub and sedge-dominated fen vegetation including Carex rostrata-Calliergon mire, Filipendula ulmaria -Angelica sylvetris mire, Phalaris swamp, Carex rostrata swamp and Carex rostrata-Equisetum fluviatile sub-community occupying the lower lying areas. This has led to this area also being included in the Rescobie and Balgavies Lochs SSSI. The Nationally Scarce Cicuta virosa (Cowbane) and Lysimachia thyrsiflora (Tufted Loosestrife), could be threatened by changes in water levels or the input of sediment/nutrients as shown by [12].
Water sampling around Chapel Mires compared with the outlet of the Lunan Water at Bagavies Loch, and End Member Mixing Analysis (EMMA) confirmed the vegetation analysis and helped confirm the need to manage pollutant rich water inflow, especially if storm waters threaten the southern and eastern parts of the wetland, which are more pristine (see Figure 7).   The following list summarises some key findings of this engagement process:

Aerial view of chapel mires (red line shows SSSI boundary) and gradient of vegetation from (top right) Phalaris arundincea (reed canary-grass) and Sparganium erectum (branched bur-reed) rich (sediment tolerant) to (bottom right) Carex rostrata (bottle sedge) and Menyanthes trifoliata (Bogbean) rich (sediment intolerant) associations with distance from the Lunan water (located on northern boundary).
1. It is challenging to demonstrate technical feasibility The development and validation of the model to assess the impact of management scenarios on water levels and flow routing was a complex and time-consuming process. The project was set up with both technical and social science goals, which needed to run in parallel, yet each of which had uncertain outcomes;

Predicted benefits are quite thinly spread across users
The multiple benefits of the scheme for flood risk, low flow risk, and ecological conservation was seen as a selling point for a Payment for Ecosystem Services scheme, but in practice it diluted the beneficial impact for each stakeholder and made the technical justification of the scheme based on robust evidence, more challenging; 3. The strongest concerns are for long-term management and legal issues The catchment has historically supported a series of working water mills for grain processing, and there is still historical memory of water disputes arising from these. While none of these mills are now operating commercially, there is still a recognition that active water management has the potential for dispute and legal challenge. A "benign neglect" approach, based on passive management, is therefore favoured by some. Governance systems to deal with such issues are generally weak in Scotland; 4. Among those in favour of "Water for All", there were no clear champions The multiple benefits aspects of the proposed scheme has meant that it has proved difficult to identify a single agency with sufficient interest to promote proposals. This has changed recently, with a more unified strategy to promote the benefits of the scheme for wetland ecological conservation; 5. The need to dedicate time and energy to pursuing approval for installation and management after installation The costs of demonstrating the proof-of-concept elements of the research project have detracted from the pursuit of management and governance planning post-installation and have led to adoption of a proposal that is more passive (reinstatement of blocked spillway) rather than active (management of a remotely operated tilting weir);

Insufficient or doubtful benefit to stakeholders
The need for co-ordinated long-term planning and the existence of likely tradeoffs (e.g. between up-and downstream interests; between wetland ecology and farming; between fisheries and irrigators; around issues of access to the environment) has led to some stakeholders considering the work required to

Deliberation and adaptation of objectives
In any integrated water management project of this kind, it is also vital to respond adaptively to stakeholder issues and barriers to implementation [15]. This project was one of a group of case studies that considered adaptive catchment management across Scotland 2 .
Having considered the concerns listed above, a modified proposal was tabled to the Lunan Catchment Management Group and to riparian owners. This focused on re-instating a blocked spillway downstream of the existing spillway where the river and Common Lade separate (CMS in Figures 3 and 9). This blockage had been installed in the 1970's, to enhance the flow of water to the mill lade to feed a recently reinstated historic water mill. Modelling of scenarios of hydraulic management showed that this approach would provide one of the key benefits of the original scheme, namely the protection of Chapel Mires wetland ecology from sediment from Newmills Burn during peak runoff events. It would have a more passive approach to management and be in keeping with current river restoration policies aiming at returning to natural flow regimes, where possible. The modified proposals were agreed in principle by the local interest groups, opening the way for solicitation of funding for the re-instatement works required.

Conclusions
In any project of this kind, it is vital to respond adaptively to stakeholder issues and barriers to implementation. This chapter highlights key elements of an "action research" project to underpin Scottish Government science policy for sustainable management of water resources in Scotland. The paradigm of Payment for Ecosystem Services underpinned the approach taken and the aspiration was to see a pilot PES scheme developed and assessed within the original 5-year time. However, the combined challenges of demonstrating the technical benefits, adaptive response to stakeholder concerns and lack of governance structures for water resources