Impacts of Open Pen Freshwater Aquaculture Production on Wild Fisheries

2 Study methods

2.1 Impacts on wild populations

Review of published biological studies

Published scientific literature relevant to understanding the impacts of freshwater cage rearing of salmonids in the northern hemisphere was assembled and reviewed. In the first instance, the review encompasses the literature related to defining the general potential for impacts on local native fish populations, based on current scientific understanding of population interactions in freshwater ecosystems. Secondly, the review examines the published scientific evidence that impacts have actually occurred in respect of populations of Scottish salmonids. These two sets of evidence are considered separately.

The review first addresses whether there is a general case that impacts could occur from freshwater pen rearing and secondly the specific evidence of interactions between pen-reared salmonid fishes and wild salmon populations generally. A conceptual framework for interactions is set out using insights from the broader biological literature relating to interactions between cultured and wild fish in aquatic environments. It encompasses all likely sources of impact and pathways of interaction, including both direct genetic and indirect ecological impacts, and considers the available scientific evidence for the nature and importance of each framework element. This framework is used to structure the way in which the actual evidence for impacts is considered. The general observational or experimental evidence that potential interactions may in fact occur, focusing on studies of salmonid species is reviewed. However, where salmonid studies were limited or lacking, the review was extended to evidence from studies of other species. The objective of this part of the review was to identify those potential pathways for which there was actual evidence regarding the effects of pen rearing or of impacts.

A potential for impacts to occur does not mean that impacts have actually occurred in a specific context, only that they could. For impacts to occur, interacting elements must be spatially coincident. Thus the spatial co-incidence of freshwater cages and salmonid populations in Scotland is reviewed. The overlap in the distribution of freshwater pen sites with those of the six Scottish salmonid species is examined. In addition to spatial coincidence, the extent of any interaction will be conditioned by the absolute numbers of the interacting elements as well as their relative abundance. Thus the spatial coincidence of wild and farmed fish, as well as their absolute and relative numbers, needs to be considered in determining the actual potential for impact in a given context.

Towards addressing this issue, the latest information on the distribution of freshwater pen and wild salmonid species in Scotland is compiled. Additionally, numbers of reported escaped farmed salmon and rainbow trout were collated from MSS information and discussions were held with MSS in respect of the ability to accurately quantify number of wild salmonids in lochs ecosystems generally and the availability of useful information for lochs with freshwater pens considered.

Finally, the review considers published scientific papers as well as unpublished reports providing insight into the actual nature and extent of interactions and impacts in locations in Scotland where freshwater pen rearing and wild salmonid populations are coincident. The scope for comparing the status of salmonid populations in catchments with freshwater pens compared to those without is explored in respect of Atlantic salmon, the only species for which there is widespread data on numbers of juvenile or adult fish. The quality of the science and the evidentiary base, its comprehensiveness and the specific research gaps that might limit the ability to draw conclusions, were evaluated.

The search for relevant primary literature was carried out in the first instance using Google Scholar. Libraries such as that of Marine Scotland Science Freshwater Laboratory were visited to search for historical literature and general biological information relevant to the review. Additionally, Trust biologists working in the field were consulted as were individuals working in the farming sector. Finally, scientists working in the field of interactions between farmed and wild salmonids in Canada, Iceland, Norway, Ireland and the UK, the countries, where freshwater cage rearing has been carried out historically or is presently practised, were contacted.

Scale and location of industry

In order to assess the implications of freshwater aquaculture in relation to any possible interaction with wild freshwater species, it is essential to know in detail:

• Where each aquaculture unit is located, and what species it produces
• What type of unit: tank-based; pen-based
• What scale of production (either as annual production biomass or maximum permitted biomass on site)
• How aquaculture units are located with respect to salmonid fishery districts

Information sources have been as follows:

Marine Scotland Science Annual Production Survey ⁷ .

The most recent version of the survey to have been published is 2010 and provides good national level information on freshwater production, number of type of sites, volumes involved etc. Unfortunately spatial information is rather limited.

Public Database of Production Sites

Additional information was obtained about freshwater aquaculture units from the Scottish Government website ⁸ . The Excel file is useful because it provides:

• The name of the fish farming company and the name of the site
• The registration number for the site
• OS Grid reference
• An indication of the type of aquaculture unit: fresh or sea water; pen or tank/raceway/recirculation system
• Species of fish produced
• Various aspects of fish health status

The Excel spreadsheet requires significant manipulation in order to analyse it for particular types of species or production system, and it contains no information about the size or volumes of the fish holding units. It also does not contain any details about scale of annual production (in terms of number or biomass of fish) or about maximum holding capacity in terms of biomass.

Specific Information from MSS

The individual site key information was requested from MSS for the current and historic freshwater sites in Scotland.

The study team received information in the form of various spreadsheets and several shape files. These various files, together with the public database, have allowed the team to sort the various sites and generate a list of freshwater pen sites in terms of:

• Site number
• Site name
• Location
• Species
• Production method
• Holding volume
• Date of registration / becoming inactive / becoming deregistered

No information was provided about quantity of production from each site. Production levels for salmon smolt can be inferred by comparing water holding capacities given for each with the national average output per unit volume available from the MSS Annual Survey reports.

By comparison of the OS grid reference in these datasets with various on-line map resources, the team have been able to identify the freshwater body and river or catchment system within which each site is located, and therefore those where interactions with wild salmonids can potentially take place.

Fish Farming Companies

The production companies themselves are potentially important sources of information with respect to geographic presence and scale/type of operation in different regions of Scotland. They are also sources of key information for other parts of this study (see below). Four large companies, covering both salmon smolt and rainbow trout production, together with one smaller specialised smolt producer shared detailed information with the study team. Initial cost models derived from that information were then consulted upon with all fish farming companies in Scotland for whom there were contact details, and additional information was received and used to refine the models further.

2.2 Costs, benefits and financial implications of changes in production practices

2.2.1 Conceptual approach

Any change in policy regarding the reduction in use of freshwater pens potentially gives rise to a range of costs and benefits that will be felt variously by the private sector, public sector, angling sector and wider society from changes to the natural environment.

It is common practice to subject public sector investment or policy options to cost-benefit analysis, which aims to capture all of the costs and benefits of a project or policy over its lifetime to assess whether the investment is worthwhile to society overall. In the current debate there are four main sectors or groupings in society which may be affected by a change in policy regarding use of aquaculture pens in freshwater lochs: the aquaculture industry itself, the public sector, angling sector (anglers, river owners, River Boards and Trusts), and wider society who may benefit from amenity and conservation values of fresh water bodies and their wild fish populations. Set out below is the conceptual framework of current costs and benefits that may arise for each sector with and without freshwater lochs being used for production of fish.

Table 1: Conceptual framework of costs and benefits from use of freshwater bodies with and without rearing fish in pens

Use	With pens		Without pens
Sector	Costs	Benefits	Costs	Benefits
Aquaculture industry	Capital costs pens Operational costs pens	Production of smolt, trout	Capital costs of alternatives (tanks) Operational costs alternatives (tanks) / other smolt sourcing	Availability of smolt, trout
Public sector	Regulatory costs pens	None	Regulatory costs for alternatives	None
Angling	Angling costs	Angling (potentially reduced)	Angling costs	Angling (potentially improved)
Environment	Discharges to air, water, land. Potential conservation value loss Visual/amenity impacts from pens	Avoided cost of land-based alternatives / other smolt sourcing	Discharges to air, water, land. Visual/amenity impacts from land based / other sourcing	Avoided cost of loch-based alternatives. Potential conservation / amenity value enhancement

Potential for quantification

In many examples of policy change, the costs are fairly easy to determine in monetary terms but the benefits much less so. This particularly applies where the policy relates to potential improvements to the natural environment, as in this case. A change in production method has the potential to alter each of the costs and benefits within the framework above. The potential to measure these changes is as follows:

Aquaculture sector, costs: A change from current pen use, either to improve containment standards of pens or to remove them entirely and replace them with tanks on land, will create a new set of capital and operational costs. These can be estimated through the literature, interactions with industry and cost models based on first principles. Detailed methods are set out in Section 2.2.4

Aquaculture sector, benefits: Aquaculture sector benefits are considered to be neutral from a change in production method. The unit value of salmon smolt and trout is only marginally influenced, if at all, by the method of production. Benefit to industry will thus be the same provided the same volume of fish can be produced using new methods.

Public sector: The public sector incurs costs through regulating the industry. These may change somewhat in a newly structured freshwater industry, and there will be a cost burden with implementing any change. These will be relatively minor compared to industry costs and difficult to quantify and so will be considered as a simple semi-quantitative score between options for change.

Angling sector: Changes in costs and benefits relating to the angling sector could potentially arise if it was established that removing pen production would enhance wild stocks. Angler spend is the only means of quantifying economic activity in this sector and methods are available for estimating this in those catchments which overlap with pen production. Other non-use values of wild fish stocks are considered under environmental costs and benefits.

Environment: Changes in production method will cause a different set of environmental impacts. Movement of pen capacity to land-based tanks will have impacts on land use, resource use etc and these will be described for scenarios of change. Changes to scenic and amenity values are also considered under the heading "environment". Regarding wider ecosystem services, DEFRA has produced guidance on valuing changes to the natural environment likely to arise from policy options in terms of changes to ecosystem services that may arise ⁹ . This takes the basic position that ecosystem services should be considered in terms of benefit to humans, i.e. ecosystem services:

• Generate income and wellbeing
• Prevent damage that might otherwise inflict cost on society

The suggested approach valuing changes in ecosystem services is to:

1. Establish the environmental baseline

2. Identify the potential impacts of policy options

3. Quantify those impacts on specific ecosystem services

4. Assess the effects on human welfare

5. Value the changes in ecosystem services

The DEFRA guidance acknowledges the many difficulties in quantifying changes to ecosystem services and states that the approach is "purposefully introductory". Nevertheless the approach can help in identifying impacts from policy change, so is a prompt to ensure that all potential changes are listed and discussed qualitatively, if not quantified.

In 2011 DEFRA published a very wide ranging UK National Ecosystem Assessment. This assessed the state and value of the natural environment and ecosystem services for the whole of the UK ¹⁰ . Unfortunately there is little within the National Ecosystem Assessment that helps us with valuing the services provided by freshwater lochs in Scotland. The chapter on economic valuation discusses the various possible approaches to valuations and the "Stated Preference" method is given as probably the most applicable to valuation of such services as "water quality" or "species conservation", which are those of closest relevance to the current study. Undertaking a "Stated Preference" study of the public's views regarding wellbeing of wild salmonids in certain river systems would require a significant primary survey exercise and is outside the scope of this study.

The only valuation data within National Ecosystem Assessment provided in relation to fish populations rivers is in the Working Paper "Freshwater, Wetland and Floodplain Ecosystem Services" which provides some headline figures on angler spend in England and Wales in 2009, based on research by the Environment Agency. The nearest equivalent for Scotland thus appears to be the Radford report on economic impact of freshwater angling in Scotland, (already cited).

Given these constraints it appears to be impossible to monetarise any benefits to river ecosystem services that may arise from policy options regarding changes to freshwater pen production.

Within the conceptual framework, therefore, the only change that is possible or appropriate to monetarise is that for aquaculture industry costs.

2.2.2 Economic analysis

Policy options on the possible future constraint of freshwater pen use are roughly framed in the study brief as either no further expansion, or a programme of phasing out freshwater pen use.

To test the impact of changes, "Base" Scenarios, i.e. estimates of what would happen without any policy intervention, are constructed for estimates of industry demand for smolts and trout in future, using estimates from industry, government and the study team's own assessment. As well as a main future demand figure for smolt and trout as the base case, low and high projections of demand encompass plausible range of demand, to inform possible range of additional costs.

Three "High level" Scenarios regarding changes to the method of sourcing of smolt are examined, along with sub-scenarios considering various degrees of change within some of the "High Level" Scenarios. Costs which could arise in the various Scenarios of change are compared to the "Base" Scenario, i.e. they are those which will be additional to those which would occur anyway. As discussed above, benefits will not change from the Base Scenario for fish production and are otherwise uncertain for the other sectors. Thus, for comparison of Scenarios, additional costs only will be presented, rather than a conventional cost-benefit analysis. These will be expressed as Net Present Costs, which shows costs that may occur over an extended period in present terms, in line with normal public sector practice comparing policy options.

For context, the study also examines the likely losses which would occur if activities were to cease completely in extreme scenarios of change, i.e.

a) if freshwater pens were totally removed and their capacity not replaced with land based facilities

b) if freshwater pen use continued but was so damaging that all angling in the affected catchments were to cease.

For scenario a) the ongrowing units have no alternative uses and so the loss of supply of smolt which are currently supplied from pens would result in the loss of profit for around half of the output of the salmon production industry. About one third of trout production would be lost.

For scenario b) angler spend in all those catchments where cages exist would be lost, although some degree of substitution of activity could occur.

These extremes are very unlikely to occur and so are not subject to comparative analysis with the more plausible scenarios.

2.2.3 Financial analysis

A mandatory change in the methods of freshwater production will have various implications for industry which need to be considered beyond purely economic costs. Aquaculture businesses would need to raise capital to invest in new land based facilities and this is discussed. Also assessed is how new regulation may affect industry viability, as well as implications for the number of new large land-based sites which might be needed, as well as possible implications for employment.

2.2.4 Methods for quantification of costs of change to industry

The study used four main approaches to assessing costs:

• Direct requests to industry participants to share (on a strictly non-attributable basis) some key cost information.
• Peer-reviewed literature and 'grey' literature (where we consider the information sources to be reliable).
• Equipment and system suppliers, who are commonly able to discuss operating cost implications of the types of capital equipment they are providing to the sector, as well as capital costs themselves.
• Internal knowledge, based upon the team's own experience of building and operating tank-based aquaculture systems, and upon its knowledge of the biological needs of the species involved, and of the sources of up to date information about the cost of supplies required to meet those needs.

Direct Requests to Industry

The study team contacted many key industry figures, and requested non-attributable information about different aspects of costs involved in smolt and trout production. Feedback from industry was generally very helpful, and is reported in detail in Section 3. Details of all contacts are included in Appendix 2.

Literature

It should be noted that almost all available literature concerning aquaculture production costs, whether capital or operating costs, is of relevance to this study. Pen production is different from land-based production in many ways, but within each of those broad categories the principles of producing fish are generally very similar. The key issues are:

• Production of Biomass. The inputs and outputs for a particular type of fish production unit all relate to the active biomass on the farm. Smaller fish are generally metabolising food and growing faster than larger fish, but that difference is taken into account by the different stocking densities that prevail for each size range. Published capital and operating costs of land-based ongrowing farms can thus be used as guidance for possible costs for producing juvenile fish such as smolts, as long as production volumes are known and key operating features, such as stocking density, water flow rate and oxygenation are taken into account.
• Value of Biomass. Taking into account the point made above, it is relatively easy to see why land-based units, and particularly complex ones such as recirculation systems ( RAS), are likely to be more economically viable for juvenile fish such as salmon smolts - they are worth considerably more on a 'per kg' basis when sold into the market. One kilogram of salmon smolts is notionally worth £10, whereas one kilogram of adult salmon might be worth some £3.00 - £3.50, but the systems used to produce them have potentially very similar inputs in terms of costs. Recent experience with RAS ongrowing farms in the UK has demonstrated how difficult it is to keep production costs down to a level where the enterprise is viable, whether the species is tilapia ¹¹ , sea bass ¹² or barramundi. ¹³
• Species Irrelevance. Information about the economics of production of one type of finfish aquaculture species, in a particular type of production system, is of direct relevance to considerations about another type of species in a similar system. As long as basic biological differences such as growth rate, food type, food conversion and stocking density are taken into account, it is possible to gain useful insights from almost any published study. A good example would be to consider modern Danish 'model trout farms' as just another way of holding, growing and harvesting cold water salmonids - which could be trout, or which could be salmon smolts.
• Seawater and Freshwater. Whilst there are some very clear differences between operating characteristics of seawater and freshwater RAS units ( e.g. buffering capacity, waste particle buoyancy, effectiveness of foam fractionation), for more traditional flow-through land-based farms, and for pen-based farms, the differences between seawater and freshwater economics are not necessarily so marked. This is particularly true if modern freshwater pen units in Scotland are being engineered to the high technical standards ordinarily seen in marine pens. Marine land-based units always require pumped water supplies, whereas many freshwater land-based units do not. However, as long as the differences between specific systems are understood, published information about costs in marine systems can be informative when considering freshwater production economics.

For the reasons outlined above, this study has obtained published information about the economics of many types of finfish aquaculture. Assessment of the value of the literature and its specific applicability to the present study is presented in Section 3.2.

Equipment Suppliers

The study team was able to interview several suppliers of land-based aquaculture equipment at the Aquanor exhibition in Trondheim in August 2011 - specifically in relation to the present study. Several suppliers indicated they might be willing to supply additional written information to this study. These groups were contacted in November 2011, and information was requested. One of the main European suppliers of recirculation systems for smolt production shared a significant amount of information about capital and operating costs, and provided the study with several Excel models that allowed easy 'what-if' scenario testing.

Internal Knowledge

One of the study team members has been involved commercially in land-based aquaculture for many years, and has also studied and reported on the technical aspects and economics of such systems on many occasions ¹⁴ . The importance of this experience comes into play when interpreting and collating the various sources of information as described above in the detailed analyses in Section 3. It is important to stress that, with the exception of some industry-standard technical and biological knowledge, the analysis in Section 3 is completely based on new information obtained for this study.

Cost implications

The financial implications of additional costs from any change in sourcing or production on an industry wide basis have been examined through comparing these to the average profit both on a unit cost (per kg) and on a whole industry basis, so as to place additional costs in context. Industry profit margins have been determined from annual or quarterly accounts of the leading production companies for Scotland, which represent a large sample of the whole Scottish industry, from 2006 to 2011, also a smaller sample back to 2000. These have been applied to the total tonnage produced in Scotland from MSS surveys. Other qualitative implications which could arise from change in practices are assessed from discussions with industry and the team's own knowledge.