Scottish Marine and Freshwater Science Volume 3 Number 6: Development of a GIS based Aquaculture Decision Support Tool (ADST) to determine the potential benthic impacts associated with the expansion of salmon farming in Scottish sea lochs

1. Introduction

In light of increasing demands for seafood around the world and reduced productivity in the fisheries sector, there will be an increasing requirement for aquaculture to address the short fall. Fisheries Commissioner, Dr Joe Borg, speaking at the 'European Aquaculture and its Opportunities for Development' conference in Brussels on 16 November 2007, commented:

"There is growing demand for seafood worldwide not only due to population growth but also because per capita consumption of seafood is expected to grow between now and 2030 by 50%", and further said that "...aquaculture appears to be the only viable option to meet this growing demand". Aquaculture now accounts for nearly 50% of the world's food fish; whereas in 1980 only 9% of the fish consumed by people came from aquaculture ( SSPO website, 2012). Also, one objective of the draft National Marine Plan ( MSS, 2012a) is to increase finfish production in Scotland by 50% on 2010 levels by 2020.

Understanding the environmental impacts of an expanding aquaculture industry will ensure that the industry is not unnecessarily restricted and environmental impacts are managed effectively. This paper presents a GIS based aquaculture decision support tool ( ADST) to assist Marine Scotland with planning the sustainable development of the aquaculture industry in Scotland. This tool has been developed alongside and incorporates elements of a project to develop ecological sustainability indicators for Scottish aquaculture (Greathead et al., 2012, in press). It will provide advice to better understand the distribution of and the potential limitations due to the deposition of organic matter onto the seabed from fin fish aquaculture, that could feed into the development of Marine Scotland's National Marine Plan. Present models predict the environmental impact from existing or intended sites; whereas this method allows the environmental impacts from any future potential sites to be predicted.

This GIS tool visualises the number and position of potential new aquaculture sites within sea lochs and the maximum biomass suitable for each site based on benthic impact by using a modified version of the Marine Scotland Science benthic impact ( MSS - BI) model. This has also been used to visualise the amount of suitable space available to aquaculture development and provide advice on the "spatial requirement" necessary to support the theoretical maximum biomass (B _TMAX) in a sea loch. Both these processes should assist in determining where there is scope for aquaculture development.

The final outcome, in conjunction with the indicators project, will be to obtain a definitive figure for the maximum potential biomass that can be produced at potential individual sites and in a sea loch as a whole, bearing in mind organic deposition and spatial factors.

Background

Aquaculture is a vital part of the Scottish economy and culture with its roots going back to the experimental beginnings in the 1960s, to the large commercial expansion from the 1980s to the present. Commercial salmon farming in Scotland has grown from about 5000 tonnes production per annum (t yr ^-1) in the 1980s to about 130 000 t yr ^-1 in 2008 (Marine Scotland Science ( MSS, 2008)). The average size of fish farm has also grown from about 85 t yr ^-1 in 1985 to over 490 t yr ^-1 in 2006. Operations are increasingly becoming concentrated on larger sites with over half the production coming from sites with a consented biomass of greater than 1000 tonnes during 2007 ( MSS, 2007).

This increase in production has not just been due to the expansion of the number and size of sites; but also is due to improvements in feed and feeding technology (Black, et al., 2008), husbandry techniques, the use of vaccines and immuno-stimulants (Bricknell and Dalmo, 2005 and Galeotti, 2007) and cage technology (James and Slaski 2009).

At present, further expansion of aquaculture could be difficult both financially and environmentally as the aquaculture industry is frequently challenged on its interactions with the marine environment and the degree to which current practices may be considered sustainable. To ensure that future growth in the aquaculture industry is acceptable, this growth will need to be shown to be sustainable. In addition, the aquaculture industry relies on the goods and services provided by the marine environment, such as the provision of clean oxygenated water and the dispersion of wastes. A sustainable industry will need to ensure that these goods and services are not compromised.

"Sustainable Development is the management and conservation of the natural resource base and the orientation of technological and institutional change in such a manner as to ensure the attainment and continued satisfaction of human needs for present and future generations. Such sustainable development (in the agriculture, forestry and fisheries sectors) conserves land, water, plant and animal resources, is environmentally non-degrading, technically appropriate, economically viable and socially acceptable" (Code of Conduct for Responsible Fisheries (Food and Agriculture Organisation of the United Nations, 1995)).

The aquaculture industry will always have a global footprint as it relies on globally sourced ingredients for feed, energy and chemicals and may, therefore, never be considered truly sustainable (Wurts, 2000). The task for Marine Scotland will be to ensure that the local and regional impacts of aquaculture are sustainable in the long term. This would require that irreversible damage to the ecosystem does not occur and that it does not impact on the capacity of the ecosystem to provide the goods and services needed not only for the aquaculture industry but also other users. Also, the long term health and functioning of all the biological components that make up the ecosystem should be protected. This project, as well as the indicators project (Greathead, et al., 2012), will concentrate on the environmental aspects of aquaculture, where it is assumed that environmental sustainability equates with a level of aquaculture that maintains good ecological status. However, social and economic sustainability should also be considered when planning for a sustainable aquaculture industry ( Figure 2.1) (Marine Conservation Society, 2007).

Figure 2.1: The three components necessary for sustainable aquaculture (Marine Conservation Society, 2007)

Recent developments in the aquaculture industry have not only allowed increased production, but have also reduced its environmental footprint. For example, improvements in feed and feeding technology have meant that less feed is required to produce the same amount of growth with less waste, which reduces the amount of waste food and faeces that reaches the sea-bed. The use of vaccinations has reduced the requirement for antibiotics to treat bacterial diseases such as furunculosis and "in-feed" immune system stimulants have helped reduce mortality due to viral diseases (Bricknell & Dalmo, 2005 and Galeotti, 2007). In 1993, only 5 500 tonnes of salmon were produced for every tonne of antimicrobials sold, in 2000 this figure increased to 67 000 tonnes of salmon for every tonne of antimicrobials (Marine Conservation Society ( MCS, 2007)).

Environmental guidelines currently place limits on allowable nutrient enrichment and benthic impact from fish farming. These impacts are regulated by the Scottish Environmental Protection Agency ( SEPA) by discharge consent licences, issued under the Water Environment (Controlled Activities) (Scotland) Regulations 2005 ( CAR). These environmental concerns ultimately limit the total production of farmed fish that can be accommodated in Scottish coastal waters. The Scottish Government Locational Guidelines ( MSS, 2012b) are published quarterly and provide advice on the level of development in 115 sea lochs around Scotland, based on the cumulative impacts of nutrient enhancement and benthic impact of all the fish farms with a discharge licence from SEPA in each sea loch. Predictive models are used to assign each sea loch an index of 'nutrient enhancement' and 'benthic impact'. These index values are then combined and the resulting value determines the Category of that sea loch, one, two or three (Gillibrand et al., 2002). A Category One sea loch is where new fish farming developments were unlikely to be acceptable due to an already high level of development. Category Two sea lochs are where the prospects for future developments were likely to be limited and Category Three sea lochs are where there appears to be greater potential for future development. These categories state a predicted level of impact given a certain level of development and do not reflect the actual environmental status of the sea lochs. The detailed methodology for the Scottish Government Locational Guidelines can be found in Gillibrand et al. (2002).

1.1 Benthic Impacts of Marine Aquaculture

Most of the studies into benthic impacts from aquaculture have concentrated on the local effects as fish farm wastes rarely travel more than a few kilometres from the site of origin. The cumulative effects of multiple areas of degraded seabed within a sea loch or regional sea area are poorly understood due to the complexity of seabed habitats and ecosystem processes. The accumulation of organic waste (fish faeces and uneaten fish feed) on the seabed results in localised deterioration of sea bed habitats and changes to benthic (seabed) communities, biodiversity and nutrient balances. The effects of this organic enrichment can be seen to be graduated away from the source of input ( i.e. the salmon cages), following the pattern first described by Pearson and Rosenberg (1978), which was based on studies of organic enrichment from paper mill effluent before aquaculture became an established industry. Pearson et al. (1986) also related the long term changes in the sedimentary benthos to organic inputs and long-term temperature changes. Organic enrichment of sediments is the most widely studied of the impacts from fish farming with documents by Brown et al. (1987) and Gowen and Bradbury (1987) being some of the first to describe in detail the effects of effluent from salmon farms on the benthos and environment of Scottish sea lochs and coastal waters respectively. The effects documented included highly reducing sediments close to the cages, reductions in dissolved oxygen in water overlying the sediment, changes in benthic fauna abundance and diversity and increased carbon content.

These studies have concentrated on the impacts on benthic macro-fauna; other studies that have concentrated on the meio-fauna have indicated that the meio-fauna may be less stressed by organic enrichment than the macro-fauna (Austen et al., 1989), although there is much variation in response, possibly due to other factors such as the degree of oxygenation of the sediments and the presence of toxic contaminants (Warwick, 1993). With regard to the geochemical and physio-chemical changes in sediments as a result of fish farm wastes, the most comprehensive study was made on behalf of the Technical Advisory Group of the British Columbia Ministry of Environment (Brooks, 2001).

Some habitats are more susceptible to benthic enrichment than others, for example Maerl beds (Hall-Spencer et al., 2006) and eel grass beds (Diaz-Almela et al., 2008). A large proportion of the species on rocky reefs are filter feeders and are, therefore, also susceptible to smothering from solid wastes.

Figure 2.1.2 describes the key impacts of marine fish farming and the link between these impacts and the amount of food that is put into the cage system and fish. Feed pellets were composed primarily of fish meal and fish oil, together with other minor components such as binders, added minerals and anti-oxidants. However, the fish products are now significantly substituted with vegetable derivatives. The fish obtain an insignificant proportion of their nutrition from natural food items present in the surrounding water, which means that the inputs and wastes from a fish farm can be accurately proportioned. Nitrogen supplied to the fish is held in the fish meal (protein) component of the feed. While a small proportion of the pellets may not be ingested and hence lost as waste feed, the bulk of the nitrogen is partitioned between increase in fish biomass, faeces and soluble excretion products. The amounts of nitrogen in these various compartments, per tonne of production of fish, can be calculated from simple mass balance models (Davies, 2000). The estimated value for waste input to the water column therefore takes account of both soluble and particulate organic nitrogen, and the nitrogen loss associated with waste feed.

Figure 2.1.2: Key impacts of marine fish farming originating from the feed input to the system; solid wastes are closely linked to benthic impact and soluble wastes are closely linked to water column impacts

Spatial Requirement Advice

There is a limited amount of space available for the development of aquaculture in Scotland. Therefore this advice has been developed to provide advice on the "spatial requirement" necessary to support the theoretical maximum potential biomass (B _TMAX) in a sea loch, and limitations for aquaculture with regard to available space within sea lochs.

1.2 Methods

To determine the spatial requirement for the maximum potential development of aquaculture in Scotland several components were required:

1. A standard value for production per km ².
2. The precautionary limits of benthic impact (BI _pa) and nutrient enhancement (ECE _pa) calculated in the indicators project (Greathead et al., 2012, in press).
3. The theoretical maximum biomass of farmed fish in each loch (i), such that the precautionary limits for nutrient enhancement (ECE _pa) and benthic impact (BI _pa) are not exceeded ( B _TMAX(i); t yr ^-1).
4. The surface area suitable for aquaculture for each sea loch ( A _{(15-70 m)} ; km ²).

The standard production value was calculated from annual fish farm production data and Crown Estate leased areas. Data from the Crown Estate on the current areas leased for fish farm development (Sept 2008) were collated for each sea loch and the total area of seabed within each sea loch that was leased to fish farms calculated ( A _L; km ²). The total surface area that would be suitable for aquaculture (surface area where depth 15 - 70 m, A _(15-70m); km ²) in each sea loch was also calculated from a previous MSS GIS project to re-digitise the sea lochs (Annex 1a and 1b). The shallower depth of 15 m was chosen as most pens at salmon aquaculture sites are 10 m deep, an extra 5 m was added to this to allow for adequate water circulation below the cages at all tide and wave heights. The deeper depth of 70 m was chosen as the maximum depth fish farms are currently moored at. There could be much debate about these values, however, it was felt that they represented best practice without being too restrictive.

Annual production data in each sea loch were collated to determine an average production value for each sea loch for each year from 2000-2008 ( P _(i); t yr ^-1) ( MSS, 2006-2008). Each sea loch has different production data for various reasons (water quality, husbandry/management techniques), therefore, the data for each sea loch were averaged using three different scenarios; 2004-2008 and 2005-2008 and 2006-2008. The values for 2004-2008 were used to determine a mean production value for all sea lochs ( ; t yr ^-1), as these data had the smallest standard error ( SE) and standard deviation ( SD) (157.18 and 1388.15 respectively). The complete production data can be seen in Annex 2a. The Crown Estate leased areas were also very variable therefore the total leased area values for each sea loch were taken for 2008 and the mean area ( ; km ²) for all the sea lochs with available data was calculated. The standard errors and standard deviations were also calculated for the values and recorded in Annex 2b. The and values were then used to calculate the standard value for production per km ² ( P _d; t km ^-2) as in Equation 3.1.1. Zero returns were ignored in the averaging process in this case as the aim was to produce an average figure for production for active sites in a sea loch.

3.1.1

The P _d value could be taken as a proxy for production efficiency (production per km ²), which can either be calculated as an industry wide standard as here, or as a loch specific value using loch specific and values ( Annex 4).

The maximum biomass of farmed fish in each loch (i), such that precautionary limits for nutrient enhancement and benthic impact were not exceeded ( B _TMAX(i); t yr ^-1), was calculated using a modified version of the Equilibrium Concentration Enhancement model (ECE.bas) outlined in Gillibrand et al. (2002). This model calculated the degree of nutrient enrichment in terms of concentration of nutrient nitrogen ( ECE, μmol N l ^-1) and the percentage of the low water area of a sea loch that is 'degraded' ( BI, % LW area), resulting from the total stocking biomass of fish in that sea loch.

To allow the ECE.bas code to perform iterative functions it was transcribed into Matlab and modified. This was verified to ensure that the two variations of the code gave identical results. The code was modified to allow iterative variation of the total farmed fish biomass in each loch until set limits were reached (ECE_MOD.m). Where a loch had multiple farm sites, the proportion of biomass in each farm was kept constant. The modified code calculated the maximum potential biomass in each sea loch, such that the precautionary levels of benthic impact and nutrient enhancement were not breached.

The maximum limit for nutrient loading (ECE _L) for Shetland and Mainland Scotland were 12.6 μmol l ^-1 and 8.7 μmol l ^-1 respectively, based on OSPAR guidelines. The precautionary limits for nutrient loading (ECE _pa) for Shetland and Mainland Scotland were 9.9 μmol l ^-1 and 6.4 μmol l ^-1 respectively, based on the standard deviations of the data (Greathead et al., 2012, in press). The critical limit for area of the loch floor classified as "degraded" (BI _L) was 10% of the suitable area and the precautionary limit (BI _pa) was 8%, based on the Joint Nature Conservation Committee's Guidance for undertaking habitat assessments. A carbon loading on the loch floor of greater than 7 kg m ^-2 yr ^-1 was regarded as degraded (Gillibrand et al., 2002; Greathead et al., 2012, in press). The minimum value of the two biomass values for each loch gave an indication of whether benthic impact or nutrient enrichment was the limiting factor; i.e. the threshold that was breached first as levels of fish farming were increased. This value was taken as the B _{TMAX (i)}. Full details of this methodology and how the thresholds were set are available in Greathead et al. (2012, in press).

The value for B _{TMAX (i)} (t yr ^-1) was divided by the standard production per km ² ( P _d, t km ^-2 yr ^-1) to provide the total area required ( A _REQ(i); km ²) to achieve the B _TMAX(i) ( Equation 3.1.2). This area was then converted to a percentage of the area suitable for aquaculture for each sea loch (% A _{(15-70 m)}).

3.1.2 A _{REQ (i)} = B _TMAX(i) ÷ P _d

Conversely the percentage of the B _TMAX that could be achieved in each sea loch, given P _d and A _(15-70m)(i), was also calculated ( Equation 3.1.3).

3.1.3 % B _TMAX = ((A _{(15-70 m)(i)} x P _d)/B _TMAX(i)) x 100

Data for 114 sea lochs were compiled. However, there were insufficient data to calculate sea loch specific production for 36 of these as either there were no production data returned from these sealochs in the MSS production surveys, or there were no leased area data recorded by the Crown Estate. Therefore, the standard average production value was calculated from 78 sea lochs. This calculated standard value was then used to calculate A _REQ for all sea lochs as values for B _TMAX and P _d were available for all sea lochs.

1.3 Results

The total average area of seabed within each sea loch that has been leased to fish farms in 2008.

= 0.304 km ² ± 0.037 (0.355)(93) (Value ± SE (SD)(N))

The average production value for all sea lochs for 2004-2008

= 1499.59 t yr ^-1 ± 157.18 (1388.15) (78)

The average value for production per km ²

P _d = 4929.323 ± 517.076 t km ^-2 yr ^-1

The total surface area that would be suitable for aquaculture in each sea loch as well as other sea loch attributes are shown in Annex 1a.

Figure 3.2.1: The relationships between the maximum biomass at the precautionary thresholds for nutrient enrichment (ECE _pa) and benthic impact (BI _pa) by sea loch. The smallest value in each loch is the limiting factor in each sea loch ( i.e. lochs 80, 38, 85 and 88 were limited by benthic impact). Only top 40 sea lochs ordered by ECE _pa are shown (see Annex 3 for data for all 114 sea lochs).

Figure 3.2.1 shows the maximum theoretical biomasses, calculated with the ECE_MOD.m model, that corresponded to precautionary levels of nutrient loading and benthic impact for the top 40 sea lochs with respect to nutrient loading (ECE _pa).

The full results for the 114 sea lochs used in these calculations can be seen in Annexes 3 and 4. Only six sea lochs were limited by benthic impact. There were not enough data to run ECE_MOD.m for 11 sea lochs, which meant that the B _TMAX in these sea lochs was based only on nutrient enrichment.

When the values for B _TMAX were compared with the biomass that is currently licensed for each sea loch (B _C), the B _TMAX was less than the licensed biomass in three sea lochs ( Annex 4). Therefore, if no other factors are taken into consideration, there is potential for the expansion of aquaculture in the majority of sealochs (97%). However, this does not take into consideration spatial issues or site specific benthic impacts.

Annex 4 also shows the total area required (A _REQ) to achieve the precautionary B _TMAX, for the 114 sea lochs used in these calculations and the percentage of the B _TMAX that could be achieved in each sea loch.

When described as percentages of the 'Low Water' area of each sea loch, these areas ranged from 107% of the loch area down to just 17%. Alternatively, when A _REQ was described as percentages of the surface area of each sea loch suitable for aquaculture (A _{(15-70 m )} ), these areas ranged from 120 857% of the loch area down to 48%, due to very small surface areas with depths between 15 and 70 m in some sea lochs. All of the sea lochs would require more than 45% of the A _(15-70m)(i) to achieve their maximum potential biomass. Only 27 of the 114 sea lochs (24%) would require < 100% of the A _{(15-70 m)} to achieve their maximum potential biomass, which means that the majority of the sealochs would require more than the available A _{(15-70 m)} to support the B _TMAX. Ten sea lochs (8.7%) had no areas between 15 m and 70 m depth and, therefore, could not support any of the B _TMAX.

Looking at these calculations from an alternative point of view; the percentage of the B _TMAX that could be achieved in each sea loch, given P _d and A _(15-70m)(i), ranged from 0% to 1313%. Only 26 of the 114 (23%) of the sea lochs were able to support 100% or more of the B _TMAX.

1.4 Discussion

There is a limited amount of space available for the development of aquaculture in Scottish sea lochs. Therefore, this advice has been developed to provide information on the limitations for aquaculture with regard to available space within sea lochs.

The spatial advice developed in this section will be a useful tool in planning for the future development of aquaculture in Scotland. The areas in sea lochs within the 15-70 m depth range and available to aquaculture development is finite, which means that the expansion of aquaculture should be concentrated where space is available. The calculations for the 'area required' advice showed that only 20% of the 114 sea lochs would require < 100% of the A _(15-70m) to achieve their maximum potential biomass. Therefore, it would appear that in the majority of the sea lochs (80%), due to the limited area with depths suitable for aquaculture, it would be extremely improbable that the maximum potential biomass could be reached.

This is confirmed by the percentage of the maximum potential biomass that each sea loch could support, where only 22% of sea lochs had the potential to produce 100% of their B _TMAX, with 8.7% not able to support any aquaculture production within the A _{(15-70 m)}.

So, although there is a high theoretical potential for increasing the biomass of fish produced in the majority (97%) of sea lochs, the potential level of aquaculture development in a sea loch is limited by a combination of suitable available area and production efficiency not necessarily by the environmental limits used in current environmental guidelines. For example if the P _d increased, the percentage of B _TMAX that could potentially be produced in a sea loch would also increase.

These results are theoretical, designed to be the first step in giving an overall picture of how feasible aquaculture development would be in certain areas, given a certain level of production and space available and do not consider the environmental impacts of the expansion of aquaculture. Environmental factors have been applied in Section 4. Spatial planning enforced by Local Authorities also will place restrictions and criteria on further aquaculture development in their area based on the requirements of the other users of the area.

Coastal areas are utilised for many other activities and so in some sea lochs there is extreme competition for space and the issue of visual amenity can be an important factor for the development of aquaculture sites. This highlights the need for spatial planning to assist with the prioritisation of activities. The responsibility for the planning and development of aquaculture has now been transferred to the Local Authorities. The Scottish Government have reviewed the various Aquaculture Framework Plans and similar planning documents prepared by local authorities within Scotland. For example the Loch Fyne ICZM plan (Argyle and Bute Council, 2009) indicated that there is limited scope for the expansion of aquaculture even though it could potentially produce 516 549 t of fish per year without breaching the nutrient enrichment and benthic impact thresholds; 99% of which would be able to be produced within the A _{(15-70 m)} ( Annex 4).

The Crown Estate leased area data needed to be evaluated with caution when used to calculate the standard production value (P _d), as leased areas varied considerably in size in relation to the actual size of the farm. This was highlighted by the high SE and SD values. Another possible figure would possibly be the 25m AZE (Allowable Zone of effect) area recorded by SEPA or actual farm sizes; however, these data are not available for each farm. Another option would be to use an area value calculated from the average stocking density; however, this would entail an unacceptable level of variability as both the stocking densities used by each farm and the size and depth of cages vary considerably. Variation in the production data was mitigated by using average values ( i.e. 2004-2008) with the smallest SE and SD values. Loch specific values for P _d could also be used for more accurate planning at the sea loch scale.

It is not just spatial competition issues that could affect the amount of production in a sea loch; logistics for the effective management of each site, such as distance for boats to travel from the shore base, road access, shore base availability and deep water access for supply boats will also affect how much of the potential biomass is feasible in each sea loch. These factors need to be assessed on a site by site and loch by loch basis and will also vary with the standard operating procedures, area management agreements and codes of conduct for each fish farm company.

GIS Based Aquaculture Decision Support Tool ( ADST)

The benthic impact of a fish farm is determined by the position of the fish farm in the sea loch. This GIS tool visualised the number and position of potential new aquaculture sites within each sea loch and the maximum biomass suitable for each site based on benthic impact. At the heart of this tool was a modified version of the Marine Scotland Science benthic impact ( MSS - BI.m) model. This will assist in the spatial visualisation of the positioning of potential new fish farm sites based on associated environmental impacts.

As shown in Section 3, the spatial aspect of aquaculture is important considering the increasing pressure on space from many other sectors, such as offshore energy production, tourism and leisure, pipelines, navigation and fishing. The development of marine spatial planning tools is ongoing in organisations, such as Marine Scotland, Local Authorities (Aquaculture Framework Plans) and Universities (Hunter, et al., 2006). Considering these issues, highlighting areas regionally and nationally, where aquaculture would be most suitable and areas where further development of aquaculture would be unsustainable would be a useful tool. This could in turn provide information on the total possible productivity of the aquaculture industry given certain criteria. This is envisaged to be a guidance tool not a tool for detailed placement of aquaculture sites. The GIS aspects of the ADST were undertaken by Seazone Solutions Ltd.

1.5 Methods

This project was divided into two phases: Phase 1, to determine areas that are potentially available for aquaculture within sea lochs; and Phase 2, to determine the benthic impact of any new aquaculture sites in each sea loch and the cumulative benthic impact from these. The final outcome, in conjunction with the indicators project, will be to obtain a definitive figure for the maximum potential biomass that can be produced at any potential new sites and in each sea loch as a whole bearing in mind environmental and spatial factors and which sites would be environmentally sustainable.

1.5.1 Phase 1

A bathymetric base layer was created in order to extract areas of bathymetric depth of between 15 m and 70 m for use in the project; bathymetric modelling using charted bathymetry data (best scale available UK Hydrographic Office S57 holdings, i.e. derived from charted sources at varying scales) was conducted using the modelling software package BathySIS.

Depth soundings, contours and areas of fixed water depth ( i.e. dredged areas) from Electronic Navigation Charts (of best scale) in the project areas of interest, were used as inputs into the GIS project.

Model results were exported as half degree grid cells; each grid had 1800x1800 cells with an approximate cell size of 30 m ² (depending on the latitude of the cell). The grid defined geographic space as an array of equally sized square grid points arranged in rows and columns. Each grid point stored a numeric value that represented a geographic attribute (such as elevation or surface slope) for that unit of space. Each grid cell is referenced by its x, y coordinate location.

Masking layers were then created (as ESRI Shapefiles) based on the following sets of features:

1. Protected Areas.
2. Port and Harbour exclusion zones with a 1000 m radial buffer around point features.
3. Existing fish and shellfish farms with a 1000 m buffer.
4. Existing Crown Estate aquaculture Leases.
5. Marine infrastructure e.g. wind farms and pipelines with a 1 km buffer for points and a 100 m buffer for linear features.
6. MOD exclusion zones ( PEXA Areas)
7. Activity and licence areas e.g. oil and gas, wind farm licence areas.
8. Transportation routes e.g. ferry routes with a 1000 m buffer.
9. All wrecks in Seazone list with a 1000 m buffer.
10. East coast from Duncansby Head to the English Border (due to presumption against aquaculture development).

Each masking layer was removed from the bathymetric base layer using a clipping tool. The resulting areas were called Potential Aquaculture Development Areas ( PADA).

The results of this exercise do not take into consideration other restrictions such as local amenity, wave and tidal restrictions, and environmental impact. The environmental restrictions regarding nutrient and benthic enrichment were assessed with new tools and existing models modified to accommodate iterative functions in Phase 2.

1.5.2 Phase 2

The number and size of farms that can be placed in a sea loch will be restricted by the physical space available, local amenity use and planning, the maximum potential biomass for each sea loch and the potential benthic impact at each site.

The PADA for individual sea lochs were calculated by filtering the final results polygons from Phase 1 to include only the polygons that intersected with the polygons of the sea lochs defined by the "Locational Guidelines" (Gillibrand et al., 2002), used throughout this project.

A combination of automated and manual processing was then applied to identify the potential maximum number of farms within the PADA of each sea loch using the following criteria:

1. The new site shapefiles were 1 km in diameter, so that there would be 1 km distance between centre points.
2. The centre points of all the new site shapefiles must be within the boundaries of the PADA in each sea loch.
3. The new site shapefiles do not overlap with the other aquaculture masking shapefiles ( e.g. existing aquaculture sites and CE lease areas).
4. The centre point of new sites may not overlap the other masking shapefiles.
5. The new tool will only apply to PADA polygons of >1 km ² within the defined sea lochs.

The values for PADA and number of sites (existing plus new) were entered into the attribute table for the 'sea loch' layer in the ADST GIS project.

Two processes were then applied to determine the maximum potential biomass that can be produced in a sea with regard to environmental factors and which sites would be environmentally sustainable. It was assumed that the biomass of fish held at a site would be sustainable if this did not result in the nutrient enrichment or benthic impact thresholds being breached.

1. Maximum potential biomass in each sea loch limited by SEPA modelling restrictions (B _SMAX(i)).

2. Maximum potential biomass in each sea loch limited by B _TMAX(i) and then modified by the benthic impact below the individual sites (BI _(s)) to determine a new maximum potential biomass based on C-flux below the individual sites (B _CMAX(i)).

Currently the Scottish Environmental Protection Agency ( SEPA) grants consents to fish farms to discharge wastes. SEPA use the "AutoDepomod" model to determine the potential amounts and distribution of wastes produced by a certain biomass of fish held on an individual fish farm. This model is currently validated up to 2500 t, which means that presently fish farm sites can only be licensed to hold up to 2500 t of fish at any one time. Therefore, the number of predicted new sites in a sea loch (n _P(i)) plus the number of existing sites (n _E(i)) were multiplied by 2500 t to determine the maximum biomass in each sea loch, taking these modelling restrictions into consideration (B _SMAX(i)) ( Equation 4.1.1).

4.1.1 B _{SMAX (i)} = ( n _P(i)+ n _E(i)) x 2500

The number and size of new sites that would be environmentally sustainable in each sea loch is dependant on the B _TMAX for the sea loch as a whole but also could be restricted by the benthic impact below the individual sites (BI _(s)). Benthic impact, measured by Carbon flux (C-flux, g m ^-2 yr ^-1) on the seabed, should not be so great as to produce conditions that would not support a community of bioturbating organisms. Carbon accumulations of 548 g m ^-2 yr ^-1 (Cromey et al., 1998) and 1498 g m ^-2 yr ^-1 (Eleftheriou, et al., 1982) have been shown to cause degraded benthic conditions. Black et al. (2008) states that carbon accumulation greater than 10 kg m ^-2 yr ^-1 would produce highly significant effects on the seabed. Therefore, in this study a C-flux level of 7 kg m ^-2 yr ^-1 has been taken to be the point at which a benthic community becomes degraded, but still able to support a viable bioturbating community of opportunistic deposit-feeding invertebrates (Gillibrand et al., 2002; Cromey et al., 2002; Hargrave et al., 2008).

The number and position of fish farm sites that would be environmentally sustainable in each sea loch with respect to C-flux under individual sites was then calculated. The MSS_BI.m model was modified to allow an iterative function to calculate the maximum biomass that could be placed on each of these new sites before the peak C-flux limit was breached below the cages. Set values for other parameters such as stocking density (20 kg m ^-3), cage net depth (10 m) and Food conversion ratio (1.17) were used. This model also calculated the area that was degraded beneath each fish farm. Therefore, a further dimension to this tool was to calculate the percentage degraded area (% A _LW) associated with the B _CMAX at each site and the total in each sea loch. The percentage of LW area was used in this case as the benthic impacts should be assessed for the whole sea loch not just the A _{(15-70 m)}.

The new sites generated by this process were added to the existing sites in each sea loch to generate a new input file for the MSS_BI.m model. The new sites were assigned 'x' values (distance from sea loch mouth (x, km)), a parameter within the model that allowed more biomass on sites nearer the mouth of the sea loch (Figure 4.1.1). This is because tidal currents are stronger near the sea loch mouth and therefore there would be greater potential for the dispersion of particulates within and out of the loch. A larger dispersion area means that although the impacted area would be larger the peak deposition would be less (Gillibrand et al., 2002).

The model was run to determine the maximum biomass that could be held at each site before the C-flux threshold of 7 kg m ^-2 yr ^-1 was breached. These values were summed for each sea loch to produce a new maximum potential biomass for each sea loch (B _CMAX).

All the new sites within each sea loch, their associated 'x' values, B _CMAX values and '% Degraded Areas' were entered into the attribute table for the 'New Farm Area' layer in the ADST GIS project. Colour grading was applied to the B _CMAX values in the 'New Farm Area' layer to reflect the level of benthic impact at each potential new site ( Figure 4.1.1).

The attribute table for the 'sea loch' layer was updated with the total B _CMAX for all sites in each sea loch and the final B _MAX, which was the most limiting value for the maximum potential biomass in each sea loch from all the calculations.

Colour grading was applied to the B _MAX values in the 'sea loch' layer to reflect the level of benthic impact in each sea loch.

Figure 4.1.1: Example of how the new tools in the ADST were used to identify the number of potential new sites within the bathymetric base layer of each sea loch; potential new sites are assigned 'x' values according to how close they are to the sea loch mouth ( e.g. 1-4 km), which would result in greater benthic impacts further from loch mouth

1.6 Results

1.6.1 Phase 1

Figure 4.2.1 below is an extract from the ADST from around the Isle of Mull on the west coast of Scotland. This shows the bathymetric base layer and some of the masking layers that represent potential restrictions to possible aquaculture development.

The outputs from Phase 1 of the ADST included: the area of the bathymetric base layer, the area of each of the masking layers and two values for the total area available to aquaculture ( PADA) ( Table 4.2.1). The initial area for the PADA ( PADA 1) included a masking area identified as MOD exclusion zones; these areas were extensive and were not necessarily exclusive to aquaculture, therefore, a second and final PADA ( PADA 2) area was calculated omitting these areas from the masking layers ( Annex 5). Figure 4.2.2 is an extract from the ADST that shows the distribution of the bathymetric base layer. Figures 4.2.3 (a) and Figure 4.2.3 (b) show the considerable difference in extent between PADA 1 and PADA 2.

1.6.2 Phase 2

Table 4.2.2 summarises the results from the first part of Phase 2 of the ADST and shows that nearly 3 000 km ² is potentially available to aquaculture within the sea lochs. However, of the 114 sea lochs included in this project only 33 had space available for new farms, the majority of which were on the Scottish mainland and Western Isles. Only two Voes in Shetland had space available for new farms (Ronas Voe and Selivoe). In total it would be spatially possible to place 389 new farm sites in these 33 sea lochs. The range in number of new sites within these sea lochs was between one and 51. The full list of sea lochs and allocation of new farm sites as well as existing sites can be found in Annex 5.

Annex 6 shows the results after applying the two processes to all the sites in each sea loch (existing and new) to determine the definitive figure for the maximum potential biomass that can be produced in a sea loch bearing in mind environmental and spatial factors and which sites would be environmentally sustainable. Six sea lochs were limited by the fact that at present, due to model validation issues, the maximum biomass on a site is 2500 t. However, this is a false restriction as the 'AutoDepoMod' model validation could be improved or a different model used.

Therefore, this resulted in a list of sea lochs where the total biomass was limited by three factors: the total theoretical biomass (B _TMAX); the C-flux modelling (B _CMAX) and the potential degraded area as a percentage of A _LW.

B _TMAX: 3 (2.6%) sea lochs were limited by the total theoretical biomass; the sum of the B _CMAX values of all the sites in Lochs Meanavagh, Seaforth and Sheilavaig was more that the B _TMAX in these sea lochs. These lochs were also limited by % A _LW degraded (see below).

B _CMAX: 111 (97%) of the sea lochs were limited by the deposition of carbon below the farm site. However, 11 of these were zero returns because there were no sites, new or existing to apply the MSS_BI.m model to.

% Degraded Area: None of the individual sites had B _CMAX values that produced degraded areas that were more than the BI _pa (8% of the A _LW). However, the total degraded areas in three of the 114 (2.6%) sea lochs were more than 8% of the A _LW. These three sea lochs were also actually limited by B _TMAX.

The limiting factors were highlighted in bold in Annex 6.

The application of these factors has resulted in a final definitive figure for the maximum potential biomass that can be produced in a sea loch bearing in mind both environmental and spatial factors (B _MAX). The total B _TMAX for all the lochs was 4,753,690 t yr ^-1 and the final total B _MAX was 1,146,892 t yr ^-1, which is approximately 24% of the B _TMAX.

The application of benthic models has, therefore, generally resulted in a reduction in the maximum potential biomass in the sea lochs compared to those calculated in Section 3 (B _TMAX), even where this was previously restricted by nutrient enhancement. This B _MAX value was recorded in the 'Sea Loch' layer of the ADST GIS project as well as in Annex 6.

Figure 4.2.4 shows an extract from Phase 2 of the ADST showing the bathymetric base layer and the distribution of the potential new aquaculture sites with colour grading applied to the B _MAX values in two lochs on the West coast of Scotland.

Figure 4.2.1: Extract from the ADST from the area around the Isle of Mull (West coast Scotland), showing the bathymetric base layer and some of the masking layers used; potential space for aquaculture development are indicated by the blue areas not covered by the masking layers

Table 4.2.1

Results from Phase 1 of the ADST.

Layer number	Layer Name	Area (km 2)
0	Bathymetric Base Layer	43 140.62
1	Protected Areas	11 249.37
2	Port and Harbour zones	1 964.72
3	Existing fish and shellfish sites	1 126.44
4	Existing Crown Estate fish farm lease areas	49.10
5	Marine Infrastructure	1 542.72
6	MOD exclusion areas	81 754.65
7	Licensed activity areas	10 575.30
8	Transport routes	4 257.93
9	Wrecks	5 010.17
PADA 1	Results layer including MOD exclusion areas	10 084.80
PADA 2	Results layer excluding MOD exclusion areas	31 977.04

Figure 4.2.2: Results from Phase 1 of the ADST; the bathymetric base layer of the ADST, surface area with depths 15-70 m.

 

Figure 4.2.3: Results from Phase 1 of the ADST: Comparison of the two PADA, (a) PADA 1: final results including the military practice zones in the masking layers and (b) PADA 2 excluding the military practice zones from the masking process.

Table 4.2.2

Results from Phase 2 of the ADST ( NW Mainland = Loch Duich to Loch Eriboll; SW Mainland = Loch Ryan to Loch Hourn; (% All Lochs, % Regional Lochs)).

Parameter	Result
Total PADA in sea lochs	2 735.95 km 2
Total number of new sites	389
Total number of sea lochs containing potential new sites	33 (29 %)
Number of sea lochs containing potential new sites in Shetland and Orkney (33)	2 (1.7 %, 6 %)
Number of sea lochs containing potential new sites in Western Isles (24)	7 (6 %, 29 %)
Number of sea lochs containing potential new sites in NW mainland and Skye (31)	12 (10 %, 39 %)
Number of sea lochs containing potential new sites in SW mainland and Mull (26)	12 (10 %, 46 %)
Range of number of new sites	0-51

Figure 4.2.4: Results from Phase 2 of the ADST: Extract from the ADST, showing an extract from the attribute table, the bathymetric base layer and potential new aquaculture sites in Lochs Dunvegan, Bracadale and Bay on the West coast of Scotland. The sites have been colour graded according to B _MAX values

1.7 Discussion

Section 3 provided information on the general capability of a sea loch to produce the theoretically maximum potential biomass with regard to nutrient enrichment and benthic impact. The majority of the B _TMAX values used in these calculations were limited by the nutrient enhancement in the water column, as only six sea lochs had B _TMAX values set by the benthic impact threshold (BI _pa). In this section a GIS tool was used to further define the limiting factors in sea lochs, with regard to benthic impacts.

At present models can only predict the environmental impacts of existing or planned new sites as the position of the site is a required parameter for the predictive models. The ADST provides the positions of possible new sites and, therefore, can clearly ascertain a definitive figure for the maximum potential biomass that can be produced at potential new sites and in a sea loch, bearing in mind environmental and spatial factors. This will assist with the future planning, regulation and licensing of a sustainable aquaculture industry in Scotland by prioritising farms for development by benthic impact ( i.e. highest biomass without breaching C-flux threshold).

In Phase 1, the PADA could vary in size and position depending on the depth parameters set for the bathymetric base layer and the masking layers parameters. For example, the masking layer for ports and harbours could be completely removed as not all ports and harbours exclude aquaculture. Also the masking layers for existing fish farm sites and leases could be modified as not all these farms are actually being used.

At present, the masking layers do not include meteorological and hydrological phenomena such as wind strength, wave height and currents. The tolerances of cage equipment to environmental stresses will limit where a fish farm could be placed without the risk of the cages collapsing; however, the technology is improving all the time with offshore submersible cages becoming a real possibility in the future (James and Slaski, 2009). These offshore areas also will require suitable impact models to be developed to regulate the amount of fish that can be held on these sites to ensure that there is limited risk of harm to the environment (Holmer, 2010).

There is the perception that in some areas, particularly Shetland, all the available suitable sites for aquaculture have already been developed has been shown to be accurate; only two of the 33 Voes in Shetland and Orkney had any potential space for aquaculture development. Other regions such as the South West of mainland Scotland were shown to have more scope with regard to the expansion of aquaculture in sea lochs. However, as highlighted in the discussion for Section 3, regional 'Aquaculture Framework Plans' may place the needs of other sea users over the expansion of aquaculture even within environmental boundaries.

Although Section 3 showed that spatial competition issues affect the potential amount of production that is theoretically possible in a sea loch, environmental limitations also apply. Environmental factors need to be assessed on a site by site and loch by loch basis and will also vary with the standard operating procedures and codes of conduct for each fish farm company.

The environmental restrictions regarding benthic enrichment were assessed in Phase 2 of the ADST and concentrated only on the PADA within sea lochs to rule out any uncertainty surrounding the PADA in offshore areas and to provide distinct quantifiable study areas.

Although some criteria were applied, the allocation of sites within the PADA was purely arbitrary and was not intended to show definitive positions of potential new sites. The site positioning was used only as a framework for the BI model, as the level of impact is dependant on the positioning of the sites. In future, more sophisticated 3D models will make this positioning even more important. The exact parameters used to determine the new site positions could also change, as will possibly the size and shape of the PADA, which will both influence the number and position of potential new farms.

The application of the benthic impact model to the potential new sites in the sea lochs shows that benthic impacts could be the key limiting factor for the further development of aquaculture in sea lochs, excluding any planning restrictions.

The impact of organic deposition on the sea bed varies considerably, due to the highly variable hydrographic conditions at the sea bed, the complicated processes of benthic-pelagic coupling, the variability in composition of the sea bed and the debate around the role of species richness and or biomass in ecosystem processes (Bolam et al., 2002). There are, however, some primary driving factors that need to be considered when assigning a threshold for organic deposition on the sea bed which include: maintaining the redox values and oxygen concentrations above levels that could cause anoxic conditions, maintaining some bioturbating organisms to ensure recovery of the seabed, and maintaining biodiversity to ensure that ecosystem processes are not inhibited (Nickell, et al., 2003; Widdicombe, et al., 2000).

Within sea lochs there are varying proportions of Priority Marine Features ( PMFs), such as Maerl beds that are particularly sensitive to sedimentation and organic enrichment (Hall-Spencer et al., 2006). In future the ADST could incorporate known areas of such sensitive habitats and remove them from the PADA. This will consequently reduce again the areas that are potentially available to aquaculture development and therefore the potential maximum biomass each sea loch could produce.

The results from Phase 1 showed that the area available to aquaculture within the sea lochs is limited; however, there are considerable areas that could be available to aquaculture out with the sea lochs. Presently the positioning of fish farm cages is limited by the technical and management challenges faced in more exposed areas. There is also a maximum current speed that the fish can tolerate (James and Slaski 2009). Figure 4.2.3 showed that the majority of the PADA are offshore and therefore more susceptible to the effects of wave height, tidal streams and currents. To exclude areas where the conditions are too extreme for cage farming will require modelling or more detailed investigation; e.g. the "Exposure algorithm" available from the Scottish Association for Marine Science ( SAMS). The threshold levels for cage integrity also vary depending on technical developments. These areas also do not take into consideration other restrictions such as local amenity and environmental impact. Local amenity issues should be covered within the Local Authorities "Aquaculture Framework Plans" or Marine Spatial Plans.

Specialised offshore cages are in development and include fully submersible cages; however, this technology poses different management and husbandry challenges and may not be suited to conditions in the majority of UK offshore waters (James and Slaski 2009).

To expand the ADST into coastal waters, distinct areas will need to be defined that are hydrographically or geographically distinct. This may require the use of a tidal excursion model to delineate areas, as used for the ISA management areas, or more complex hydrographic models. Although this report has presented certain factors to ascertain ecological sustainability other factors that are not covered here such as the effective treatment of sea-lice and the sustainability of the fish meal and oil that comprises the fish feed may eventually be more limiting than any of the factors presented.

The ADST could be further improved by incorporating exposure algorithms (including known cage/net tolerances), PMF mapping and developing a model that can run within the GIS, possibly by the use of a pseudo layer.

The ADST could be then be used within other Marine Spatial plans to highlight areas that could be suitable for aquaculture development and avoid compounding pressures across sectors, similar to the tool used by the Crown Estate (Marine Resource System, MaRS) to highlight areas that could be suitable for offshore wind energy production.