Farmed fish advocacy in Denmark

Top opportunities for improving the lives of fish

We conducted this research on behalf of Dyrenes Alliance. More information about this association's work for animals is available here on their website.

EXECUTIVE SUMMARY

In this report, we identify the top opportunities for improving the lives and welfare of farmed fish in Denmark.

To improve the lives of fish, it is important to focus on the asks that will bring about the biggest improvements in the lives of fish for a given investment of effort. One part of this is reaching as many fish as possible. By far, the groups of farmed fish with the highest numbers in Denmark are a) fish living on freshwater farms, and b) juvenile fish. Therefore, improving the lives of these two groups of farmed fish is a priority. Almost all farmed fish in Denmark are rainbow trout.

With this in mind, we have identified two main priority asks.

1. Welfare reform for freshwater trout farms. This ask would aim to improve the welfare of trout on freshwater farms. In the Danish trout industry, the freshwater farm stage is home to 105 million individual trout every year. So, this ask could secure comprehensive welfare improvements for an immense number of fish. This ask would probably involve legislative lobbying to establish new regulations, and the first step would be to visit farms and obtain data on a collection of welfare indicators.

2. Lobby for inquiry into juvenile trout welfare. This ask would involve lobbying the government to conduct an inquiry into the welfare of juvenile fish in Denmark. This would, in turn, get the industry to release more information on juvenile fish welfare and monitor the welfare of juvenile fish in the future. Every year, the industry produces roughly 148 million fingerlings and even more fry than this. The animal advocacy movement has recently turned its attention to juvenile fish, such as the work of the Norwegian Animal Protection Alliance, though campaigns are hindered by a lack of an evidence base. As a result, this ask could not only improve the lives of an immense number of fish in Denmark but also support future campaigns around the world.

There are a few smaller asks that would also be overall positive for the lives of fish, though the total benefit would be lower. These could be run as smaller campaigns, perhaps to secure some victories while the main campaigns are being run. These asks are: 1) increase the uptake of humane trout slaughter by farms; 2) encourage retailers to choose certification schemes that are focused on welfare; and 3) ban the live export of trout.

We also provide a detailed analysis of the context of Danish fish farming and detailed information about the industry, which will assist animal advocacy organisations in any campaign on fish welfare.

LEGAL AND MARKET CONTEXT

Legal context

The key legislation and regulations governing fish farming and animal welfare are:

• Danish Animal Welfare Act (2021). This act covers fish and states that animals must be treated in a way that minimises their suffering, anxiety, and lasting disadvantage.

• Dambrugsbekendtgørelsen (Farming Order, order 1567), last amended on 07/12/2016 (1). This order establishes the environmental regulations for freshwater fish farms, including emissions of nitrogen and other pollutants.

• Bekendtgørelse om saltvandsbaseret fiskeopdræt (1990) (Regulation relative to marine aquaculture) (2). This legislation is designed to limit pollution from saltwater aquaculture. It states that the construction of sea or the enlargement of existing farms requires prior approval and that with each permit maximum allowances of nitrogen and phosphorus emissions will be laid down.

The key government bodies responsible for fish farming and animal welfare are:

• The municipalities are responsible for environmental approvals of fish farms, with responsibilities laid out by the Dambrugsbekendtgørelsen. The vast majority of fish farms are located in the Jutland-Funen watershed and the Zealand watershed (3), so the most relevant municipalities are those associated with those two watersheds.

• Ministry of Environment and the Environmental Protection Agency

• Danish Animal Ethics Council

Denmark has in general relatively more progressive legislation on animal welfare by global standards. They received a rating of “B” from the World Animal Protection Index (4). Of note, they go beyond minimum EU requirements in some areas, such as in requiring non-fish vertebrates to be stunned before slaughter, without exception for religion (5). Their animal welfare act also acknowledges the sentience of animals, though it does not specify the implications of this, which may mean that it is mainly symbolic (5).

While not directly relevant to work on fish welfare, these factors suggest to us a relatively more ambitious work may be possible in Denmark, since the government has shown itself willing to consider animal welfare.

Notably, the Danish Animal Welfare Act does extend to all animals, including invertebrates, though comments by the Danish Animal Ethics Council suggest that the intention behind this is that animals will receive different levels or different kinds of consideration according to their distinct biology (6). We interpret this as an implication that in practice fish and especially invertebrates will receive considerably less moral consideration.

In addition to animal welfare legislation, animal advocacy efforts to be able to use environmental protection legislation to their advantage, since environmental and animal welfare concerns may overlap or converge in some cases. Concerning the protection of the environment, we found the Danish government to be remarkably motivated to act. This may be particularly true because of the disproportionately large amount of agriculture that is carried out in Denmark, leading to large nitrogen emissions. The Danish government’s concerns with the environmental impact of agriculture (and aquaculture) centres around these nitrogen emissions.

Fish welfare policies in Denmark

There is a limited official effort to improve fish welfare in Denmark. However, there are some encouraging signs that fish welfare is becoming more of a concern for the government and other stakeholders. Specifically, Denmark's strategy for the aquaculture sector for 2022-2027 recognises fish welfare as a research priority.

Also, scientists have recently created a Nordic Network for Fish Welfare (currently without an online presence). This Network contains experts from fish physiology, genetics, and other scientific disciplines in Norway, Finland, Sweden, and Denmark. Scientists at the National Institute of Aquatic Resources in the Technical University of Denmark are currently working to obtain funding to begin monitoring the current status of fish welfare in Denmark.

Dansk Akvakultur, an industry body, does have a statement that generally supports fish welfare, though the statement does not provide any concrete or specific commitments (7).

Retailers

Most of the trout produced in Denmark is exported (Figure 1). The biggest importing country is Germany, followed by Poland, Finland, and Vietnam. Most of the exports are within Europe, with a small proportion going to Asia, and negligible proportions going to Oceania, the Americas, and Africa. Therefore, when it comes to retail sales of Danish trout, the two most important countries are Germany and Denmark.

Germany has a relatively dispersed retail sector. The top five retailers are Edeka (20% market share of the food retail sector), Schwarz-Group (which owns Lidl and Kaufland) (14%), Rewe (13%), Aldi (containing both Aldi Nord and Aldi Süd) (10%), and Metro (10%) (8).

In contrast, Denmark has a concentrated retail sector. The top four retailers are Coop (38% market share of the supermarket sector), Dansk Supermarked (32%), Dagrofa (16%), and Reitan (9%) (9). The level of concentration is similar to countries like Australia, where the large market powers held by retailers helped to increase the market share of higher-welfare meat products (10).

This suggests that retailers could be a valuable ally in a fish welfare campaign, though perhaps more so in Denmark than in Germany.

Figure 1: The rough percentage of Danish trout production consumed in each country (A) and continent (B). Note that all percentages are rough approximations only, and the vertical axes have different scales. Most trout is exported, and only about 15% of trout is consumed domestically. Germany is, by far, the biggest importing country of Danish trout, followed by Poland, Finland, and Vietnam. Only a small percentage is exported outside of Asia.

Certification schemes

It is unclear exactly what proportion of Denmark's trout is certified by private certification schemes - but we have a few clues.

In Denmark, Aquaculture Stewardship Council (ASC) lists 31 freshwater trout farms, across 12 large companies, that are currently certified by ASC in Denmark (11). We have production data for 17 of these 31 farms that we obtained from municipalities (see Appendix). These 17 farms together produced 9,040 t of trout in 2021, which is 29% of the volume of trout produced in Denmark. If you correct for the fact that our data is incomplete, we would expect roughly 50% of the volume of trout in Denmark to be covered. Most of the farms in the dataset are large farms.

Therefore, we would expect the proportion of ASC-certified Danish trout to be around 50%, or perhaps a little bit lower or higher. The certified production comes from 31 farms, most of which are relatively large.

There may also be certification schemes other than ASC that are active in Denmark. The two most important countries for retail sales of Danish trout are Germany and Denmark.

• It appears that much of the fish sold in Germany are, or will soon be, farmed under certification schemes. See, for example, the commitments made by Edeka, Lidl Kaufland, Rewe, Aldi Nord, and Metro. Around 40% of the trout produced in Denmark is sold in Germany. Therefore, we would expect much of this 40% to be certified.

• Likewise, it appears that most of the fish and crustaceans sold in Denmark are, or will soon be, farmed under certification schemes. See, for example, the commitments made by Coop, Dansk Supermarked, and Aldi. About 10-20% of trout produced in Denmark is consumed domestically, and we would therefore expect most of this 10-20% to be certified.

Not all of the certification schemes mentioned by these retailers are strong for fish welfare. It may be relatively straightforward to encourage retailers to shift towards welfare-focused schemes (see below, 'Smaller Campaigns').

Imported fish

Although Denmark does import farmed fish and crustacean products from other countries, we believe that it is better to focus advocacy efforts on fish produced inside Denmark. The reason is that most fish and crustaceans sold in Denmark are farmed under certification schemes, which limits the opportunities for potential campaigns that would aim to improve the welfare of these imported animals - certification campaigns are a key way to improve the lives of imported animals, while there are many more options available to advocacy organisations to improve the lives of domestically farmed animals. Of the farmed fish and crustaceans imported by Denmark, the most substantial are over 100,000 t of farmed salmon each year (from Norway and other European countries), and just under 100,000 t of farmed shrimp each year (from Vietnam, Ecuador, India, and numerous other countries) (12).

THE FISH FARMING INDUSTRY IN DENMARK

Industry structure

The structure of the Denmark fish farming industry is depicted in Figure 2. Fish farming takes place in three main stages:

• The fry and fingerling stage involves breeding, hatching eggs, and growing the juvenile fish to fingerlings of around 50 g each. Most of these fish are transferred to on-growing, though a small number are used to stock recreational fisheries or wild fish populations.

• The on-growing stage involves growing the fish to around 250 - 350 g on freshwater farms. A small number are subsequently grown further, to around 3 - 4 kg, on sea farms.

• The harvest stage involves slaughter and sale. Around 100 million fish are harvested from freshwater farms each year, and around 3 million from sea farms. Stunning and slaughter are not regulated, but the industry voluntarily stuns 70% of fish before slaughter, so stunning and slaughter may also be less of a priority. Around 80 - 90% of fish are exported, mostly to other European countries and to Asia. Only around 10 - 20% of fish are consumed domestically. Of the freshwater harvest, around 30 million fish are exported live.

The stages with the highest numbers of fish each year are, by far, the fry/fingerling stage and the freshwater on-growing stage (Figure 3A). There are far fewer fish who experience marine on-growing, live export, and stocking fisheries or wild fish populations. This is also true if you account for the duration of time spent in each stage (Figure 3B). This suggests that it may be wise to focus fish advocacy campaigns on the fry/fingerling stage and the freshwater on-growing stage.

As of 2020, there are 156 trout farms in Denmark, and these produced 42,805 tonnes of trout. The number of farms has decreased over time, but the volume of fish produced has increased (Figure 4). Of the 42,805 tonnes of trout produced in 2020, around 71% was produced for consumption (the other 29% being fingerling/fry production, production for wild releases, etc).

There are three main types of trout farms in Denmark:

• Traditional freshwater farms. In 2020, there were 104 traditional freshwater farms in operation, producing about 36% of trout by volume (13).

• Model freshwater farms. These farms are partially recirculating and are adapted from traditional freshwater farms with the aim of reducing the emissions of nitrogens and other pollutants (14). In 2020, there were 33 model freshwater farms in operation, producing about 38% of trout by volume. There are two types of model farms, Model 1 and Model 3, which are differentiated by their water use. For detailed descriptions and photographs of model farms, see Jokumsen and Svendsen (14).

• Sea farms. These farms are located in the ocean and are used to grow trout transferred from freshwater farms to larger sizes. In 2020, there were 19 sea farms in operation, producing about 26% of trout by volume. Since trout are grown to much larger sizes on sea farms, this 26% represents far fewer than 26% of individual fish (see above).

Figure 2: Structure of the trout farming industry in Denmark. Where production weights in tonnes are given, numbers refer to 2020 data. The numbers of individuals harvested are approximate calculations. Data derived from (12–19).

Figure 3: Number of individual animals (A) and animal-years (B) present each year in each stage of the farming industry in Denmark. Animal-years are a way of accounting for the duration of each stage; animal-years = individual animals * duration of the stage. Numbers are approximate calculations. Fry/fingerling numbers assume 1% mortality each month, based on a study in Norway (20), but we suspect mortality could be several times higher, and therefore our estimates of fry/fingerling number would also be higher. Data derived from (12–19).

Figure 4: Number of trout farms (top) and volume of trout harvested (bottom) in Denmark over time (13,21,22). 1974-88 data are approximate. 1974-1992 data are all freshwater farms - a very small number of these (a few to a dozen) may be for eels.

Most trout are produced by a small number of large-scale farms

In Denmark, most of the freshwater trout is produced by a relatively small number of large-scale, technologically advanced farms. This is very optimistic and encouraging as it means that a campaign that focuses on just a handful of farms can nevertheless improve the welfare of a disproportionately large number of fish.

To illustrate, in 2011, the 10 largest enterprises accounted for 56% of production of freshwater trout in Denmark (a single enterprise may control multiple farms) (23). Since the industry has continued to become consolidated between 2011 and today, we expect that the 10 largest enterprises would, by now, account for greater than 56% of production. Thus, a campaign that focuses on these 10 enterprises would have a relatively large impact in terms of the number of fish whose welfare would be improved.

Likewise, in the data we obtained through freedom-of-information requests, about 12% of farms accounted for 50% of production in 2021 (Figure 5). Again, this shows how production is concentrated in a relatively small number of farms.

The freshwater trout industry has consolidated over recent decades with the average production per enterprise around three times larger than it was in 1990. Over this period, farms have become larger and more technologically advanced while many smaller farms have shut down (23). Most of the increase in average production volume has taken place in model farms with traditional farms only seeing a slight increase in average production volume (23).

On average, model farms tend to produce more trout than traditional farms. In 2020, the 104 traditional farms had an average production of 149 tonnes (90 tonnes for consumption); the 17 Model 1 farms had an average production of 307 tonnes (211 tonnes for consumption); and the 16 Model 3 farms had an average production of 680 tonnes (506 tonnes for consumption). Therefore, it might make sense for fish advocacy campaigns to focus on the largest Model 1 and Model 3 farms.

Figure 5: The volume of rainbow trout produced by farms in Denmark in 2021. In this graph, the farms are ordered along the horizontal axis from lowest production (left) to highest production (right). There are a small number of farms with very high production volumes (right) that, together, account for a disproportionately high amount of the production of trout in Denmark. Note that this data covers 75% of farms in Denmark, so it is not complete. Data source: Freedom-of-information requests from municipalities (see Appendix).

Most trout are produced in a small number of municipalities

Some municipalities produce more trout than others. Table 1 contains the production of freshwater rainbow trout in Denmark by region. According to our freedom-of-information requests, the municipalities that produced over 1,000 t of rainbow trout in 2021 are Ringkøbing-Skjern, Holstebro, VejenIkast-Brande, Varde, Vejle, and Haderslev. All of these municipalities are in the regions of Central Jutland and Southern Denmark.

We are missing data from the following municipalities: Ærø, Billund, Esbjerg, Faaborg-Midtfyn, Hedensted, Herning, Jammerbugt, Mariagerfjord, Nyborg, Odense, Samsø, and Vesthimmerland. It is possible that any of these municipalities could also be large producers of freshwater rainbow trout.

There is definitely no freshwater trout production in the following regions: Aarhus, Assens, Brønderslev, Fanø, Fredericia, Frederikshavn, Horsens, Kerteminde, Læsø, Langeland, Middelfart, Morsø, Norddjurs, Nordfyn, Odder, Skanderborg, Skive, Sønderborg, Struer, Svendborg, Syddjurs, Thisted, and Tønder.

Table 1: Production of freshwater rainbow trout by municipality in Denmark in 2021. Note that this data covers 75% of farms in Denmark, so it is not complete and there may be some important municipalities that are missing from this table. Data source: Freedom-of-information requests from municipalities (see Appendix).

Municipality	Region	Production (t)
Ringkøbing-Skjern	Central Jutland	3,976
Holstebro	Central Jutland	3,097
Vejen	Southern Denmark	2,165
Ikast-Brande	Central Jutland	1,732
Varde	Southern Denmark	1,250
Vejle	Southern Denmark	1,172
Haderslev	Central Jutland	1,118
Viborg	Central Jutland	826
Rebild	North Jutland	763
Favrskov	Central Jutland	572
Silkeborg	Southern Denmark	517
Aalborg	North Jutland	383
Lemvig	Central Jutland	382
Aabenraa	Southern Denmark	290
Hjørring	North Jutland	232
Randers	Central Jutland	15

Economics and finances

The average output and profit of each type of trout farm in Denmark are given in Table 2. All four types of farms have, on average, positive operating profits. Traditional farms tend to produce fewer fish per farm and have a lower operating profit. Model farms and sea farms tend to produce more fish per farm and have a higher operating profit.

The components of the costs of production for each type of trout farm is given in Figure 6. For freshwater farms, the biggest cost is feed. A sizable fraction of the cost is also made up by stocking, wages, and rent and services (meaning property rent, equipment rent, oxygen, and veterinary services). In comparison, sea farms pay more for stocking - presumably because they buy relatively large fish from freshwater farms, rather than fingerlings - and for landing and distribution.

One economic study found that consumers of Danish trout are quite responsive to prices - consumers buy less Danish trout if prices are higher (24). So, finding ways to increase the costs of production could be another way to reduce the scale of the Danish fish farming industry, though this is not a focus of our report.

Table 2: Output and profit of each type of trout farm in Denmark. Numbers given are average per farm for 2020 (13).

	Average traditional farms	Average model type 1 farm	Average model type 3 farm	Average sea farms
Fish for consumption, tonnes	90	211	506	506
Other production, tonnes	59	96	174	89
Eyed eggs, tonnes	3050	0	81	0
Gross output, 1000 DKK	4198	7510	14985	21905
Costs, 1000 DKK	3750	6937	14141	19760
Operating profit, 1000 DKK	448	573	844	2145

Figure 6: Components of the operating costs of each type of trout farm in Denmark. Percentages given are average per farm for 2020 (13). The biggest sub-components of 'rent and services' are property rent, equipment rent, oxygen, and veterinary services.

TOP CAMPAIGN OPPORTUNITIES

Summary of campaign opportunities

We have identified two high-impact, priority campaigns for improving the lives of farmed fish in Denmark. These are summarised in this table and discussed in further detail below.

Table 3: The two priority campaigns for improving the lives of farmed fish in Denmark.

Priority 1: Welfare reforms for freshwater farms

Background

This ask would aim to improve the welfare of trout on freshwater farms. In the Danish trout industry, the freshwater on-growing stage is one of the two most important life-stages in terms of numbers of trout per year (the other being the fry/fingerling stage) (Figure 2). As such, this ask could secure comprehensive welfare improvements for an immense number of individual fish.

When it comes to campaigning for this ask, we believe that the strongest option is legislative lobbying. Legislative lobbying involves working with policymakers, such as politicians and ministry employees and politicians, to establish new, legally-binding welfare standards. This type of campaign can be cost-effective, as it is possible to communicate with policymakers and establish impactful policies with a relatively small team. Furthermore, producers that adopt voluntary standards often have an incentive to advocate for those standards to become legally binding (10,25) - since there is already a lot of interest in voluntary standards in Denmark (e.g. ASC), it is possible that there is already some sufficient industry support for stronger legal regulations.

There are two other options for campaigning for this ask, though we consider them less suited to this particular ask in Denmark:

• Producer outreach, which involves collaborating with producers to encourage them to adopt higher-welfare practices. For example, this could involve developing a new industry code of practice with the cooperation of the industry. This approach may not be well-suited to Denmark - it has tended to work best in contexts where farmers are being inefficient in some way (e.g. Indian carp farms), which is not the case in Denmark. In Denmark, this type of campaign has one advantage: most trout are produced by only a small number of enterprises, so targeting just the biggest few producers could benefit a large number of animals (see above, 'Most trout are produced by a small number of farms'). Also, these large farms would have greater financial and technical ability to implement new practices. The challenge with producer outreach would be getting industry stakeholders to agree on meaningful welfare improvements that improve on existing standards (e.g. ASC standards).

• Corporate campaigns, which involve campaigning for retailers (e.g. supermarkets) and food service providers (e.g. restaurants and hotels) to only sell food produced to high welfare standards. This type of campaign depends on leveraging public opinion to place pressure on retailers, who in turn choose to supply only higher-welfare products. The biggest weakness of this approach is that it would probably conflict with the Better Chicken Commitment (BCC) campaign in Denmark - since the BCC involves corporate campaigns, this would mean that the BCC and this campaign share many of the same target retailers. Retailers may be less keen to meet with animal advocacy groups about multiple welfare asks at the same time and will struggle with the workload to transition multiple supply chains at the same time. On the other hand, retailers of Danish trout have already demonstrated a willingness to sell products of higher ethical standards - this might mean that further corporate campaigns targeting retailers of Danish trout are tractable. Certification schemes provide an existing collection of higher-welfare practices that are bundled into a collection. Many major retailers of Danish trout already appear to have adopted certification schemes, and about half of Danish trout production is certified by ASC (see above, 'Certification schemes'). We do believe that there is room for encouraging retailers to select more welfare-friendly certification schemes (see above, 'Smaller campaigns').

These campaign options are not mutually exclusive, and it is possible to combine them into a larger strategy - for example, producer outreach could be used to secure buy-in from a few large farms, which could then improve the likelihood of success in making the welfare improvements legally binding across Denmark via legislative lobbying.

Notably, scientists at the National Institute of Aquatic Resources in the Technical University of Denmark are currently working to obtain funding to begin monitoring the current status of fish welfare in Denmark. These scientists could be valuable partners for this campaign.

Remaining uncertainty 1: Welfare issues experienced by trout in Denmark

The impact of this ask strongly depends on the degree to which trout farmed in Denmark are currently experiencing welfare issues, and which specific welfare issues those are. Currently, there is very limited information about the status of farmed fish welfare in Denmark.

How could this uncertainty be resolved? Fortunately, there is a tool that can be used to gather information about the welfare of trout farmed in Denmark. This is the FISHWELL handbook on welfare indicators for farmed rainbow trout, freely available here. This handbook identifies a number of welfare indicators that can be collected when visiting fish farms. Below, we summarise the indicators from the handbook that will be most useful for assessing fish welfare on visits to trout farms in Denmark. This can be used as a guide if you visit trout farms in Denmark, and Animal Ask is happy to provide advice and guidance for conducting these visits.

The welfare indicators are divided into different categories depending on how easy they are to obtain and understand:

• 1 = can be used on the farm

• 2 = can be used on the farm but needs expertise, requires further data analysis and/or special equipment

• 3 = can be sampled on the farm but must be analysed in a laboratory

• 4 = requires an extended period of analysis in a laboratory

Here, we focus on level 1 and a few level 2 indicators, as these can be used on the farm with minimal expertise and no particular special equipment or data analysis. If you want further detail on these indicators, they are discussed in greater detail from page 152 onwards in the handbook. We focus only on indicators that: a) are non-invasive and do not require fish handling, b) are relevant to freshwater flow-through systems, rather than sea farms, and c) are relevant on-farm. We chose to limit our attention to on-farm welfare indicators as most specific procedures would be quick and, therefore, limited in their impact on overall fish welfare relative to longer-lasting, on-farm conditions. There may be an exception if the procedure causes particularly intense suffering - this may be the case for slaughter, but we already have a decent understanding of humane and non-humane slaughter of trout in Denmark, and this is discussed throughout our report. If you're interested in indicators for specific procedures, the handbook discusses indicators from page 202 onwards for crowding, pumping, slaughter, euthanasia, medical treatments, anaesthesia, vaccination, and transport.

The indicators are further divided into three groups: environmental indicators, which address the rearing environment; group-based indicators, which address a farm's fish population as a whole; and individual animal-based indicators, which involve looking at individual fish.

In the table below, the ideal values apply to all main life stages (sac fry, fingerlings, and on-growing fish) unless otherwise indicated.

If you visited a few fish farms in Denmark (particularly model farms with large production volumes) and collected this information, we would be able to see whether these indicators are typically within the ideal range. If so, then freshwater fish welfare might be decent on Danish farms, but if the indicators are typically outside of the ideal ranges, then this would suggest that there are major welfare problems that need to be addressed.

Table 4: Indicators that can be easily and non-invasively used to assess the welfare of trout on farms in Denmark (26).

Priority 2: Lobby for inquiry into hatchling welfare

Background

This ask would involve lobbying the government to conduct an inquiry into the welfare of hatchlings in the country. They would in turn get industry to release more information on the subject and monitor the welfare of hatchlings more significantly going forward.

There has been little attention paid to the welfare of juvenile fish. This is unfortunate because mortality rates are much higher in juveniles than in adult fish. In many cases it may be cheaper for industry to pay for additional fry and simply tolerate high mortality rates, rather than pay for welfare measures. For example, in Norway it seems to be common practice to rear 25% more fish than the hatchery has capacity for in the smolt stage, with the expectation that enough fish will die before reaching that stage because of high mortality rates (20).

One possible approach to encourage hatcheries to be concerned about mortality rate is to try and get subsidies implemented that are conditional on lowering mortality rates among larval fish. This might spur hatcheries to implement significant welfare reforms, whereas currently they had very little economic incentive to do this. Penalties for high mortality rates would be even better from the standpoint of limiting the profitability of the industry, but they would also be harder to implement because of opposition from the industry.

Animal advocacy organisations have been largely silent on the specific issue of juvenile fish mortality, contributing to a lack of public pressure for industry to reform. This situation is exacerbated by a lack of research on the subject making it more difficult to improve welfare even if it were made a priority. Denmark has no information about fish mortality in aquaculture across the lifespan. This ask would begin to tackle this issue asking for the information needed to inform work on the issue.

However, the situation is beginning to improve with some recent research on the subject. In recent years, Norwegian Animal Protection Alliance has commissioned several reports regarding juvenile salmon in Norwegian aquaculture (20,27). The campaign might broadly replicate the work of the Norwegian Animal Protection Alliance, except in Denmark.

Charity Entrepreneurship is also notable as one of the other few organisations considering researching work on this subject. Charity Entrepreneurship (28) considered incubating a new charity focused on milkfish welfare in the Philippines. As part of their research report on the subject, they considered whether that charity should focus specifically on juvenile welfare in that species. Though they concluded that the charity should be incubated, they did not think that there was a strong reason for it to focus on the welfare of juveniles over the welfare of adults. Nevertheless, they recommended further research on the subject. Their analysis of this issue drew primarily upon various EU countries, especially Norway.

Despite their conclusion in that case, our analysis in the case of Denmark suggests that focusing on juvenile fish welfare may be a particularly strong ask in Denmark. In Denmark, trout typically spend around 2 ½ months growing from fry to 5 g fingerlings followed by another 2 ½ months growing from 5 g fingerlings to 50 g fingerlings. This means they spend a total of around five months in a juvenile stage. This is followed by a period of 8 to 9 months when they are kept as adults in other freshwater or saltwater facilities.

This means that in terms of life years, each fish that survives to the end of their life spends longer as an adult. However, the number of trout surviving until they are 50 g fingerlings versus the number who survive long enough to be slaughtered is approximately 3 to 1. This overall suggests that the bulk of life years experienced by fish and agriculture in Denmark are experienced by juvenile fish. This is especially true given that our estimates of the population of juvenile fish in Denmark agriculture assume a 0% mortality rate before growth to 50 g. This is certainly an underestimate, especially as this is the stage with the highest mortality rates.

The following table reviews general evidence on hatchling welfare issues in aquaculture among a range of species. However, this table does not include welfare information specific to trout in Denmark because we are unable to find that information publicly accessible. Because of this lack of data, we think that beginning work on the subject with an inquiry in order to gather this data is a particularly strong approach. (Some of the indicators described above also apply to juvenile welfare.)

Table 5: Welfare issues facing juvenile fish. The poor state of the evidence for juvenile fish welfare has made it difficult for the animal advocacy movement to make progress on this important topic.

In addition to the value of his ask in informing future work on a very neglected group of animals in Denmark, we think this ask could have great value by helping inform work on this issue globally. We see the work of Norwegian Animal Protection Alliance as having had this effect, and we are keen to see others following them in advancing this cause.

There are many uncertainties surrounding the welfare of juvenile trout in Denmark and how to improve their welfare. This is why we believe that a research-based approach starting with the government inquiry to gather necessary data is the most appropriate path forward. We therefore do not consider the lack of evidence to be a strong reason against this ask, precisely because this ask aims to gather that evidence. We also find that the preliminary evidence suggests that work on the subject in the future will be high-impact.

This campaign could proceed by contacting and perhaps lobbying the Ministry of Food, Agriculture and Fisheries of Denmark, or their sub-agency, The Danish Fisheries Agency. The minister responsible for this, Rasmus Prehn, could also be contacted as part of this. Another approach could be to look for a grant to pay for this work to be conducted privately, as the Norwegian Animal Protection Alliance has done.

As mentioned above, scientists at the National Institute of Aquatic Resources in the Technical University of Denmark are currently working to obtain funding to begin monitoring the current status of fish welfare in Denmark. These scientists may represent valuable partners for this ask, whether they a) simply offer technical or political support or b) actually carry out the research or inquiry themselves.

Remaining uncertainty 1: Is it physically possible to improve juvenile fish survival?

The consideration that we think is most likely to be crucial in determining the value of future work to improve juvenile welfare is how tractable it is. Trout follow a reproductive strategy of having large numbers of offspring with very low chances of survival in most cases. The evolutionary goal of this reproductive strategy is to have fish make it to adulthood despite the high mortality rates. It is therefore possible that the higher mortality rate is, in a sense, built in, and may therefore be difficult to improve on. This is the most important key uncertainty for this ask.

China and Holzman (31) may provide an explanation for these higher mortality rates. They observed that mortality rates among juvenile fish were high even in ideal circumstances of the physics of the interaction of water and very small fish. During a critical period typically lasting for their first 7 to 8 days of life, larval fish have a very difficult time feeding because of the interference of water, even when prey are plentiful. They call this phenomenon “hydrodynamic starvation”.

These larval fish feed by suctioning food into their mouths, which is generating an inward flow of water to capture prey by rapidly expanding the mouth cavity. This process is disrupted by the hydrodynamics of water at that size, resulting in an estimated 40% loss in energy invested in opening the mouth, as well as other disruptions to this process leading them to repeatedly lose prey. As they become larger, this progressively ceases to be a problem.

The implications of this are that work on the welfare of juvenile fish will either have to solve this problem, or will have to make progress on other welfare issues while leaving hydrodynamic starvation, most likely the largest juvenile fish welfare problem, unsolved. The potential to improve on the extremely large mortality rates is probably the most significant draw to working on the welfare of juvenile fish, so not touching on this issue is less than satisfactory and will limit the potential of any welfare improvements to reduce the mortality among juvenile fish.

Three potential ways of improving on this are (32):

1. Supplying them with food that is easier to catch and ingest

2. Reducing the viscosity of the water they are kept in

3. Somehow ensure that they are born at larger sizes

We are skeptical of the tractability of any of these methods. However, this seems to be an unexplored area for welfare improvements, so we think that more research in this area to see if any of these methods or a new method could be made to work.

In terms of 1, we found some claims that larval trout in Denmark are currently being fed with compound food, rather than live prey such as rotifers and Artemia (brine shrimp) (14,33). This is atypical for most fish species, as most larval fish in aquaculture are fed with live food (34). Though we found some mention of the production of live food for fry in Denmark (35), consultation with an expert confirmed to us that compound food is used for larval trout in Denmark. This suggests that hydrodynamic starvation is less of an issue for trout in Denmark.

We remain uncertain about the significance of hydrodynamic starvation in this context. More research needs to be done on hydrodynamic starvation in aquaculture, and research on trout in particular would be very valuable. Before this is done we remain uncertain about how large of a role hydrodynamic starvation plays in Danish trout aquaculture and this should be considered speculative. All things considered, we do think that there is likely to be tractable work on the subject that could be enabled by an inquiry into hatchling welfare.

SMALLER CAMPAIGNS

Beyond our top priorities, we have identified a couple of smaller campaigns. These may be slightly easier to achieve than the priority campaigns. While these smaller campaigns should not be viewed as a substitute for the priority campaigns, they could help to build support for fish welfare while a large campaign is underway. The impact from these campaigns is lower than the impact from the priority campaigns, but these smaller campaigns still represent meaningful improvements in the lives of farmed fish in Denmark.

Small campaign 1: Humane slaughter of trout

This ask would involve getting 100% of trout in Denmark to be stunned before slaughter. As of 2017, about 70% of trout slaughtered in Denmark are stunned before slaughter, and about 30% are slaughtered using asphyxia in ice (38). Asphyxia in ice is horrible for the trouts' welfare and causes extreme suffering, whereas effective stunning before slaughter can avert much of this extreme suffering (38).

This ask benefits a sizeable minority of fish (30% of trout farmed in Denmark). This ask would only benefit these fish for a small percentage of their life, i.e. during slaughter, which lasts minutes. However, the suffering experienced during slaughter is extreme. There are certain philosophical views that propose that preventing this extreme slaughter is a high priority (39,40). As such, some thoughtful advocates believe that preventing the extreme, short suffering at slaughter could be overall better for fish than preventing the low-intensity, long suffering during the rest of these animals' lives (39). While this ask only benefits 30% of trout slaughtered in Denmark for a short period, the ask could nevertheless bring a disproportionately high benefit for the lives of these trout - we believe that this ask is a suitable small campaign for this reason.

We think that this could be a good campaign for legislative lobbying. Other organisations around the world have had success with legislative lobbying for humane fish slaughter. Also, since 70% of trout slaughtered in Denmark are already stunned (and this is probably mostly large farms), it is possible that the aquaculture industry will express minimal resistance to this campaign. In fact, farms that already stun the fish may even be incentivised to support this campaign - a new legal requirement to stun fish before slaughter will impose costs on farms that do not stun fish, which benefits competing farms that do stun fish.

A survey by the European Commission reports that the 30% of trout who are slaughtered via asphyxia are slaughtered on-farm. In contrast, the 70% of trout who are stunned before slaughter are slaughtered at dedicated slaughterhouses.

One way to begin working on this ask would be to contact farms that slaughter trout via asphyxia and ice and ask the farm owners about their barriers to implementing humane slaughter. As a general rule, smaller fish farms find it more difficult than larger farms to afford an investment in stunning equipment. In early 2022, Animal Ask surveyed manufacturers of humane fish slaughter equipment in Europe. An electrical stunning machine (e.g. manufactured by Ace Aquatec) costs around 100,000 EUR (740,000 DKK). The necessary pump (e.g. manufactured by FAIVRE or VAKI) and dewatering unit (e.g. manufactured by IRAS) might add an additional 50,000 EUR (370,000 DKK). Please do not hesitate to contact Animal Ask for more details on these prices if you are interested in campaigning for humane fish slaughter.

For farms that cannot afford to purchase this equipment on their own, creative solutions exist (38). Some countries have mobile slaughterhouses which travel between farms to stun and slaughter fish. Some countries also have cooperatively owned slaughterhouses that allow multiple small farms to combine money to invest together in the stunning equipment. It may also be possible to campaign for government subsidies for farms that want to purchase or rent slaughter equipment. Speaking with trout producers in Denmark might help to clarify which of these, or other, strategies would help farms to adopt stunning before slaughter.

Small campaign 2: Encourage retailers to choose welfare-focused certification schemes

Most retailers in Denmark and Germany have committed to ensuring that the fish products they supply will be mostly or completely certified by private certification schemes. This seems to have been partially motivated by a campaign by WWF. However, the policies adopted by these retailers - and, therefore, the choice of certification scheme - are focused on sustainability, not welfare. Many of the policies adopted by the retailers are also vague as to the specific certification schemes that will be used. About half of Danish trout production is certified by Aquaculture Stewardship Council.

It may be possible to meet with these retailers and encourage them to pick certification schemes that are strong on both sustainability and fish welfare. This could secure some relatively easy wins without a big campaign - it could be possible to make the retailers' choice of certification schemes more welfare-friendly simply by sharing resources like Aquatic Life Institute's benchmark report. This campaign would be small, involving just a few meetings with retailers, but could nevertheless secure some early victories and improve the lives of many fish.

The biggest concern is that this campaign would require meetings between the animal advocacy movement (or your organisation specifically) and Denmark's major retailers. Animal advocacy organisations are already working hard to convince Denmark's major retailers to adopt the Better Chicken Commitment. As such, this ask may conflict with the Better Chicken Commitment campaign in Denmark. Therefore, we encourage you to try this small campaign if you think you can secure commitments without too much effort, but keep in mind that this campaign may also come at the cost of making the Better Chicken Commitment campaign more difficult.

Small campaign 3: Ban live export of trout

About one-third of trout produced in Denmark are exported live. This ask would involve campaigning to ban the live export of fish from Denmark. We believe that this ask could form a suitable small campaign which, although it brings a relatively small-scale benefit, can avoid a period of relatively intense suffering (transport) and has very few risks. We think that this campaign may also be tractable to achieve as it would probably represent only a minor disruption to supply chains (i.e. exporting dead fish rather than live fish).

All of the trout exported live from Denmark are destined for other countries in Europe. The biggest importing countries of live trout from Denmark are Germany (67% of Denmark's live trout exports by weight) and Poland (19%) (12).

We estimate that about 31 million individual trout are exported live from Denmark each year. This is a sizable minority of the number of trout farmed in Denmark each year - compare the 31 million individuals to, say, the 105 million who experience on-growing on freshwater farms each year (Figure 3A).

However, the journey duration is only a very small proportion of the trouts' lives. The exact journey duration is unclear, but we expect it would be within the vicinity of about 12 hours. This is a typical long journey duration for live animal transport in the EU, and it takes about 8 hours to drive from Jylland to Germany. This would indicate that only around 400,000 animal-years are experienced by fish during live export from Denmark each year. This is a very small number and almost negligible compared to other stages of production - compare the 400,000 animal-years to, say, the 79 million animal years experienced during on-growing on freshwater farms each year (Figure 3B).

We have not found any evidence on the welfare conditions experienced by fish exported live from Denmark. However, in general, fish can suffer from a range of welfare problems during live transport which relate to stocking density, handling, water quality, water movement, loading and unloading, physical damage, air exposure, and food deprivation (38,41–43). Handling and transport commonly cause acute and chronic stress with transported fish showing physiological stress responses (42,43).

Banning live export from Denmark would mean that 31 million trout are slaughtered domestically, rather than being exported live, each year. This would avert the welfare problems associated with live transport of these trout.

Our main concern about this ask is the tractability. We do not know why the industry exports trout live - there may be a price premium in the destination countries due to the freshness of live-exported trout. If the industry is strongly incentivised to continue this practice, then this ask might be difficult to achieve. Due to this uncertainty, we encourage you to keep in mind the tractability during the early stages of this campaign - if it appears that this campaign would require a large amount of effort to succeed, then it might make sense to instead swap to one of the other smaller campaigns we discuss in this article.

There are very few risks with banning the live export of trout. The biggest risk involves the method by which these trout would be slaughtered domestically if live export were banned. As of 2017, about 70% of trout slaughtered in Denmark are stunned before slaughter, and about 30% are slaughtered using asphyxia in ice (see below) (38). Asphyxia in ice is horrible for the trouts' welfare, whereas effective stunning before slaughter can avert much of the extreme suffering caused by slaughter. In Poland (about 19% of trout), it appears that all rainbow trout are slaughtered via asphyxia in ice - in this case, being slaughtered in Denmark would represent an improvement. We are unsure how rainbow trout in Germany are slaughtered - other species (e.g. carp) do appear to be stunned before slaughter (38). If most trout imported by Germany are stunned before slaughter, then being slaughtered in Denmark instead could be a bad thing. This would also depend on how Danish companies respond to a ban on live export - the fish could be transported to slaughterhouses before being killed and exported. Since fish slaughterhouses in Denmark do generally own stunning equipment (38), this would mitigate the risk that fish would be slaughtered via asphyxia in ice.

Please note that Animal Ask has a forthcoming report that goes into more detail on animal transport asks with particular sections on both transport in the EU and live fish transport.

OTHER CAMPAIGNS WE CONSIDERED

We conducted an in-depth analysis of three other asks: shut down sea farms; nitrogen regulations to restrict fish farms; and welfare requirements for fully recirculated aquaculture facilities (FREA).

Although all three of these asks initially seemed promising, they turned out to have major problems or risks that led us to conclude that they should not be recommended.

Not recommended: Shut down sea farms

Trying to shut down the fish farms in the ocean in Denmark. Either by trying to change the law so that all the fish farms have to close, or to target them individually with focus on the ones who perhaps have the worst impact on the environment.

Summary of impact for this ask

Background

Sea farms produced 11,291 t of trout in 2020. Sea farms slaughter trout at approximately 3-4 kg. This means that, in 2020, sea farms slaughtered roughly 3.2 million individual trout. On sea farms, the on-growing period on sea farms lasts for approximately 8 months. This means that, in 2020, there were roughly 2.2 million animal-years experienced by trout on sea farms. Sea farms account for only a minority of the number of fish and number of animal-years farmed in Denmark (Figure 2).

Preventing these 3.2 million individual trout from experiencing this suffering at sea would reduce the scale of the trout industry in Denmark. If this also meant that fewer trout were brought into existence, this would also be a good thing.

The biggest question is how shutting down sea farms would affect the demand of other fish products.

Remaining uncertainty 1: How will consumers switch to other fish?

Every ask that aims to reduce the production of animals will, to some extent, cause demand to switch to other products. In this case, it is possible that switching to other products will actually cause more animals to be farmed overall.

The evidence suggests that consumers would switch from large-size trout to salmon. Consumers would probably not switch from large-size trout to small-size trout as large- and small-size trout are perceived as different products. On the fish market, consumers' substitution between different fish products appears to be based on characteristics like flesh and processing rather than the species of the fish (24). Sea farms produce large-size trout with pink flesh, and these large-size trout are generally used for smoking or sold fresh (18). In contrast, freshwater trout farms produce portion trout with mostly white flesh, and these portion trout can be sold whole, sold as fillets, or used for smoking (18,44). (Note that most of Denmark's trout are exported to other countries, like Germany and other countries in Europe and Asia.) Shutting down sea farms would reduce the amount of large-size trout on the market. Since portion trout and large-size trout are considered different products, we expect that consumers would not substitute toward portion trout. Rather, we expect that consumers could substitute toward products like salmon, which are a similar product to large-size trout (pink flesh, large-sized, and sold whole or smoked) (45). It has been shown in both the German and Japanese fish markets that large-size trout compete with salmon (44,45). We spoke to one expert who also believed that consumers would substitute from large-size trout to salmon, not to small-size trout.

We can model hypothetical scenarios. In scenario 1, we can imagine that consumers switch entirely from large-size trout to farmed salmon (e.g. from Norway rather than Denmark). In scenario 2, we can imagine that consumers switch entirely from large-size trout to small-size trout farmed in Denmark.

In these scenarios, there are two main metrics that we can use:

• Number of individual fish farmed per year. Fewer animals being farmed is a good thing. The number of individual fish is calculated as the production volume divided by the slaughter weight.

• Number of animal-years farmed per year. Animal-years account for the duration that the animals spend on farms. Fewer animal-years being farmed is a good thing. The number of animal-years is calculated as the number of individuals multiplied by the lifespan.

Our two hypothetical scenarios are given in Table 6. The table shows the current trout production on sea farms, as well as the number of fish and animal-years that would be required to meet the same level of production from either farmed salmon or small-size trout.

According to this simple analysis, if consumers switch from large-size trout to salmon, then this ask could be overall good for animals. In practice, the overall impact of the ask would also depend on how the welfare conditions differ between the three scenarios, as well as other factors like survival rates, the use of fish meal and fish oil, and the use of cleaner fish. We do not account for these factors, so our hypothetical scenarios should be considered as illustrative examples only. If consumers switch from large-size trout to small-size trout, then this ask is probably bad for animals.

Remaining uncertainty 2: Is the government open to this ask?

The second key uncertainty is whether the government would be open to this ask. The government has previously implemented policies that limit or prevent the establishment of new sea farms (15), but shutting down existing farms is a different issue. After all, shutting down all sea farms in Denmark would involve shutting down 19 businesses, which is politically very different to limiting the establishment of new farms. It is unclear whether this is politically feasible or even legal.

One way to address this uncertainty could be to establish a subsidy scheme where the government pays farmers to voluntarily shut down their farms. This has been used for freshwater farms (3).

Table 6: Current trout production on sea farms (left column), compared to two hypothetical scenarios: if consumers switch entirely to farmed salmon (middle column) or small-size trout (right column). We think that switching to farmed salmon is the likely scenario. There are several important factors that we did not consider in this table, like the relative welfare conditions of different species, relative survival rates, the use of fish meal, and the use of cleaner fish - these would all affect the overall verdict.

Not recommended: Nitrogen regulations to restrict fish farms

Making regulations in the use of nitrogen that are allowed in different areas of Denmark so that new fish farms can’t open and some of the current ones have to shut down.

Summary of impact for this ask

Background

We focus the analysis of this ask on freshwater fish farms - these farms account for the vast majority of fish farmed in Denmark, and the regulations for nitrogen emissions are slightly different for sea farms. However, many of our conclusions may also hold for sea farms.

For freshwater fish farms in Denmark, there is no particular limit to the level of fish that a farm can produce. However, all model trout farms in Denmark are subject to discharge controls (46). Each farm has a specific limit to the amount of nitrogen and phosphorus that they can release. These limits stop farms from producing more fish. Specifically, the release of nitrogen is the most important factor that limits the expansion of production (46,47).

When it comes to nitrogen regulations for fish farms in Denmark, there are two main mechanisms used to impose these regulations: nitrogen emission limits, and feed quotas. Each farm is governed by one of these (not both). Nitrogen emission limits were introduced as an alternative to feed quotas in 2012 (48).

For any given freshwater trout farm, the specific regulations that must be obeyed by that particular farm are laid down in an environmental approval granted by the municipality. This approval is, in turn, based on the Dambrugsbekendtgørelsen (Farming Order, order 1567), last amended in 07/12/2016 (1). A farm's environmental approval is renewed every 10 years.

The number of fish farms in Denmark has decreased quite substantially over time (Figure 3). Therefore, we do not believe that this ask would be important in stopping new farms from being established.

Most farms are not using their full nitrogen limits

Figure 8 shows that most farms are not using their full nitrogen quotas. Each data point is a farm with the colour showing whether the farm is governed by a discharge limit (red) or a feed quota. The horizontal axis shows each farm's production in tonnes with smaller farms being towards the left of the graph and larger farms being towards the right of the graph. The vertical axis shows the percentage of the nitrogen limit that each farm is using. Farms that are on (or above) the dashed grey line are using their full nitrogen limit, while farms that are below the dashed grey line are not using their full nitrogen limit.

Most farms are not using their full nitrogen limit. Notably, most larger farms are governed by discharge limits, and none of these large farms are using their full nitrogen limits.

This strongly suggests that reducing nitrogen limits would not cause a substantial reduction in the scale of trout production in Denmark.

Figure 8: This graph shows the percentage of each farm's specific nitrogen limit that is being used. Most farms are not using their full nitrogen limit, and this is particularly true for large farms. The one outlying farm (using over 200% of its nitrogen limit) is, we were told by the municipality, currently installing a new nitrogen filter. Note that this data covers 75% of farms in Denmark, so it is not complete. Data source: Freedom-of-information requests from municipalities (see Appendix).

Remaining uncertainty 1: Could this campaign succeed politically?

Even setting aside the question of nitrogen limits, there are two more key uncertainties facing this ask. Either of these uncertainties could severely reduce the amount of good that this ask achieves.

The first key uncertainty is whether a campaign could succeed in changing the law to make nitrogen regulations more strict. In other words, is the campaign tractable?

The environmental regulations placed on fish farms are considered relatively strict given that the aquaculture sector is quite small. The land-based aquaculture sector contributes to the country's waters less than 3% of the overall nitrogen emissions (47). This might mean that there is limited political will to place stricter regulations on aquaculture's nitrogen emissions in Denmark.

The most relevant nitrogen limits seem to be the municipality-set, farm-specific limits (see below). These are the limits on which we should focus our attention. Municipalities set nitrogen emissions limits for individual farms based on the conditions set out by the Dambrugsbekendtgørelsen (see above). This is the ordinance that would need to be amended to lower the limits for nitrogen emissions.

Municipalities impose two types of nitrogen emissions limits on farms: total emission limits, and BAT (Best Available Technology) limits. Farms need to obey both sets of limits.

For a particular trout farm, the upper limit of nitrogen emissions is calculated by multiplying together the four following numbers:

• Ftill which is the volume of feed originally permitted to be used by the farm under the 1989 executive order on freshwater pond farming (no. 224 of 5 April 1989). As a rough upper limit, Ftill would equal 1.2 multiplied by the fish farm's annual production of fish. There are a few additional criteria in the 1989 executive order that can reduce Ftill for particular farms.

• PTN which is the contribution of total nitrogen per tonne of feed. PTN is set at 56 kg.

• 100% - RN where RN equals the fish farm's total degree of purification. RN is set at 50% for nitrogen.

• 1.86. This is a conversion factor used to switch from feed quota regulation to emission control regulation.

As well as the total emissions limit, fish farms are also subject to BAT (Best Available Technology) standard requirements. The BAT requirements are designed to make sure that farmers are being efficient in using their allowed levels of emissions (i.e. producing a sufficient volume of fish for a particular level of emissions) (49). The BAT requirements give a maximum nitrogen emission per tonne of fish. This means that making the BAT requirements stricter would not reduce the total amount of fish that a farm can produce.

As such, to make nitrogen regulations stricter, the Dambrugsbekendtgørelsen would need to be amended to reduce the total nitrogen emission limits.

These nitrogen emission limits were first set in 2012, and they were revised only subtly in 2016. For this reason, it is difficult to say whether these limits could be reduced in future amendments of the Dambrugsbekendtgørelsen through lobbying efforts.

Remaining uncertainty 2: Will nitrogen restrictions decrease production?

The second key uncertainty is whether further nitrogen restrictions would cause production to decrease, assuming that nitrogen regulations are indeed limiting production.

Historically, nitrogen emissions from trout farms in Denmark have decreased over time (Figure 9, below). Despite this, trout production has increased over time (Figure 3, above). This suggests that stricter nitrogen restrictions do not necessarily cause production to be reduced. Furthermore, this also suggests that nitrogen regulations may become stricter in the future even without lobbying for this ask.

It is possible that farmers would find some other way to meet the new regulations without decreasing production, such as by developing new technologies. On one hand, nitrogen regulations are widely cited as the main factor limiting trout production in Denmark (46,47). On the other hand, the government has expressed interest in developing new systems, such as fully recirculated aquaculture facilities (FREA) (see below). These could increase trout production while also reducing nitrogen emissions (14). So, it is not guaranteed that lower nitrogen limits would cause production to decrease. It is possible that the industry would simply turn its attention to finding other ways to reduce nitrogen emissions, such as new water purification systems (50) or mostly closed production systems like FREA (though one estimate suggests that the nitrogen emissions of FREA are similar to existing model trout farms (14)).

Figure 9: Nitrogen emissions from aquaculture in Denmark over time (51).

Not recommended: Ensure fully recirculated aquaculture facilities (Danish acronym FREA) contain welfare requirements

Summary of impact for this ask

Background

RAS (Recirculation Aquaculure System) is a form of aquaculture that reuses water through enhanced filtration, reducing the need for an influx of fresh water from another source, such as a stream or spring. While full recirculation of this kind has not yet been implemented for trout in the country, partial recirculation systems are used (model 1 and model 3 farms), and there are plans by government and industry for this to become the future of aquaculture production in the country.

The only FREA (Danish acronym for fully recirculating aquaculture system) currently operating in Denmark are for the farming of eels and salmon (interview proceedings). This takes place at a more limited scale (2,000 tonnes of eels per year) than the farming of trout (17).

The following table (Table 7) shows the meaning of different levels of recirculation. In this diagram, model 3 farms correspond to “RAS low level” and FREA would roughly correspond to “RAS super intensive”. Meanwhile, model 1 farms have a lower level of recirculation, around 70% (14).

Table 7: Different levels of water consumption in different levels of recirculation (52).

Of these systems, traditional trout farming operations are still the most common with recirculation farms making up only a relatively small proportion of operations. In 2020, there were 33 recirculation farms (model 1 and model 3), compared to 104 traditional freshwater farms (see above, "The fish farming industry in Denmark"). However, partial recirculation farms are more productive per farm, so they still make up just over a third of total production numbers (Figure 4).

However, the greatest promise of FREA for the Danish government and for industry are the environmental benefits, the most significant of which is the reduction in nitrogen emissions. However, Jokumsen and Svendsen (14) note that the theoretical emissions from FREA systems would be 38 kg of nitrogen per tonne of trout produced, which is only a small reduction in emissions from the 40 kg of nitrogen per tonne of trout produced on current model 3 farms. This would suggest that there is not a strong environmental reason for switching to FREA.

Welfare impacts

Concerning RAS systems in general, Fish Welfare Initiative (53) writes that “Any risks to fish welfare must be identified before production starts, and then risk mitigation plans must be incorporated into operational procedures: the number and species of fish to be produced, staff training, backup systems for emergencies, proper monitoring and operational routines, proper disinfestation procedures, and alarm technology all must be considered”. This highlights the importance of working on welfare reforms in these systems in advance.

The highest priority welfare improvement for FREA is most likely ensuring adequate biosecurity measures. When biosecurity measures are carried out effectively in RAS, the risk of disease is much lower. However, if biosecurity measures fail and pathogens enter the system, the spread of the disease can be rapid because of the high stocking densities. The risk of this happening is somewhat lower compared with RAS systems in FREA where a smaller amount of external water is taken in, eliminating one potential source of pathogens. It is therefore particularly important for FREA to build biosecurity measures into the design of the whole system, rather than adding them afterwards as an afterthought. When biosecurity measures fail, pathogens can spread rapidly because of the high stocking densities in RAS and because of the difficulty of treating disease outbreaks without disrupting the biofilter. This makes it particularly appropriate as a welfare measure to stress in advance (54).

Jokumsen and Svendsen (14) list some of these critical biosecurity precautions for FREA farms:

• “The physical construction of the FREA facility shall prevent any intrusion of vectors of pathogens

• Prevent any contamination of water intake (i.e., shielding against any vectors of fish pathogens)

• Secure against intrusion of animals and birds

• Optimum bio security when introducing new material (e.g., disinfected eyed eggs, as fish may be carriers of pathogens). Preferably the FREA should have its own brood stock. 40

• Strict hygienic procedures for admittance to the FREA (e.g., lock for bath/wash and change to clothes and footwear exclusively for internal use). External visitors should be avoided

• Prevention of dissemination between the separate sections of the FREA facility (e.g., separate equipment, ketches, etc.)

• Strict hygienic procedures (disinfection) for lorries delivering feed and for feed storage

• Strict hygienic procedures (disinfection) for lorries collecting fish (i.e., certified disinfection) Strict hygienic procedures for disposal of dead fish, eggs, etc.”

Interim the general welfare concerns in RAS systems, please see our forthcoming short report evaluating welfare issues in RAS and evaluating if RAS is overall worse for fish welfare. Overall, it is also uncertain if RAS represents a detriment to welfare (as with most forms of intensification in livestock farming) or if it represents an improvement (because the system allows for more tight control of welfare relevant measures).

Remaining uncertainty 1: Future uptake of FREA

A significant uncertainty with this ask is the degree to which, in line with government and industry intentions, FREA will see widespread adoption in the future. There are other possibilities such as a continuation of the status quo, other technologies, or a significant

shrinking of the industry. One expert who interviewed thought it might not make economic sense to switch to full recirculation technology over the partial recirculation technology already in use.

So far, the use of recirculation technology (model 1 and model 3 farms) has been growing in Denmark. Between 2005 and 2011, 30% of Danish trout aquaculture production moved from traditional aquaculture operations to partial recirculation farms (23). Though this is not full recirculation technology, it still suggests that the industry is moving in the direction of recirculation.

However, one reason that FREA might not be adopted on a wide scale in Denmark is because, while it is more environmentally friendly in terms of nitrogen emissions, greenhouse gas emissions can sometimes be considerably larger. Tilman and Clark (55) finds that some recirculation systems have substantially higher greenhouse gas emissions than traditional aquaculture, mainly stemming from increased electricity usage. This may mean that if the acute nitrogen pollution problem in Denmark can be solved in other ways, this trade-off may no longer be seen as desirable. However, an easier solution would be for farms to simply adopt the practices of those farms with lower emissions. If this means that this concern is unlikely to pose a serious obstacle to the adoption of recirculation systems.

Remaining uncertainty 2: Fish welfare in FREA

Even concerning RAS in general, there is limited information to be confident in welfare improvement recommendations. Concerning FREA, there is again less information on welfare improvements.

From our research, it looks as though FREA is broadly similar to RAS in which welfare improvements are most important. It appears as though FREA resembles an extreme version of RAS, and that welfare benefits and issues are magnified in FREA compared to other RAS systems. This suggests that the same general precautions are required in both RAS and FREA systems, limiting the effectiveness of this ask again.

Ultimately, we do not believe there is enough information to effectively inform a campaign on this ask. Though more information can be gathered through expert interviews and through talking with the company, FREA Solutions, this ask does not seem strong enough for it to be worth taking this step, especially given Key uncertainty 1.

APPENDIX: FREEDOM-OF-INFORMATION REQUESTS

In addition to the secondary research we conducted for this report, we also submitted requests for information regarding freshwater farms from Denmark's local municipalities. We contacted all municipalities in the regions of North Jutland, Central Jutland, and Southern Denmark, which makes 52 municipalities in total. Most municipalities sent us a reply. Many of these municipalities reported having no freshwater farms.

We received data from 16 municipalities. The combined data covers 103 farms, which is 75% of the 137 freshwater farms in Denmark in 2021. The combined data covers 18,491 t of production of freshwater rainbow trout, which is 59% of the 31,514 t of freshwater rainbow trout produced in Denmark in 2021.

For each farm, we asked for data about: the farm's nitrogen discharge limit OR feed quota; the farm's actual nitrogen emissions; the farm's actual feed consumption; the farm's production volume; and whether the farm is a traditional or model farm.

Some of the data is summarised in Figure 5 (production levels of each farm), Figure 8 (nitrogen limits), and Table 1 (production by municipality). We are also happy to make the full data available to animal advocacy organisations upon request.

REFERENCES

1. Dambrugsbekendtgørelsen: Bekendtgørelse om miljøgodkendelse og samtidig sagsbehandling af ferskvandsdambrug [The Farming Order: Order on environmental approval and simultaneous case processing of freshwater pond farming] [Internet]. BEK nr 1567 af 07/12/2016 Jul 12, 2016. Available from: https://www.retsinformation.dk/eli/lta/2012/130

2. Denmark. Bekendtgørelse om saltvandsbaseret fiskeopdræt [Internet]. Retsinformation. 1990 [cited 2023 Jan 10]. Available from: https://www.retsinformation.dk/eli/lta/1990/640

3. Miljøministeriet. Forslag til vandområdeplanerne 2021-2027 [Proposal for the water area plans 2021-2027] [Internet]. Miljøministeriet; 2021. Available from: https://mim.dk/media/226716/vandomraadeplanerne-2021-2027.pdf

4. WAP. Denmark [Internet]. World Animal Protection. [cited 2022 Dec 27]. Available from: https://api.worldanimalprotection.org/country/denmark

5. Denmark. Danish Animal Welfare Act [Internet]. 2021. Available from: https://www.globalanimallaw.org/database/national/denmark/

6. DAEC. Statement on the New Animal Welfare Act. The Danish Animal Ethics Council ; 2016.

7. Dansk Akvakultur. Dansk Akvakultur holdning til fiskevelfærd [Internet]. Dansk Akvakultu; 2021. Available from: https://danskakvakultur.dk/wp-content/uploads/2021/06/Fiskevelfaerd_Statement_181211_Final.pdf

8. Ruhm E, Rehder LE. Germany: Retail Foods [Internet]. USDA Foreign Agricultural Service; 2017. Report No.: GAIN Report Number: GM17025. Available from: https://apps.fas.usda.gov/newgainapi/api/report/downloadreportbyfilename?filename=Retail%20Foods_Berlin_Germany_8-7-2017.pdf

9. Ganthier J. Seafood Export Profiles Denmark Sector Report [Internet]. Sea Fish Industry Authority, UK Trade & Investment; 2014. Available from: https://www.seafish.org/document/?ufprt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

10. Carey R, Parker C, Scrinis G. How free is sow stall free? Incremental regulatory reform and industry co‐optation of activism. Law Policy. 2020 Jul;42(3):284–309.

11. Aquaculture Stewardship Council. Find a Farm [Internet]. Aquaculture Stewardship Council. 2022 [cited 2022 Dec 22]. Available from: https://www.asc-aqua.org/find-a-farm/

12. United Nations Comtrade. UN Comtrade [Internet]. 2022. Available from: https://comtradeplus.un.org/

13. Statistics Denmark. StatBank [Internet]. 2022. Available from: https://www.statbank.dk/statbank5a/selectvarval/define.asp?PLanguage=1&subword=tabsel&MainTable=AKREGN&PXSId=186536&tablestyle=&ST=SD&buttons=0

14. Jokumsen A, Svendsen LM. Farming of freshwater rainbow trout in Denmark. 2010 [cited 2022 Nov 15]; Available from: https://www.semanticscholar.org/paper/789680f53946be4f510659c25871bce0d643d4e9

15. Pedersen PB. Monitoring and regulation of marine aquaculture in Denmark. J Appl Ichthyol. 2000 Jul;16(4-5):144–7.

16. Pedersen K, Skall HF, Lassen-Nielsen AM, Nielsen TF, Henriksen NH, Olesen NJ. Surveillance of health status on eight marine rainbow trout, Oncorhynchus mykiss (Walbaum), farms in Denmark in 2006. J Fish Dis. 2008 Sep;31(9):659–67.

17. Larsen K. National Aquaculture Sector Overview: Denmark [Internet]. Fisheries and Aquaculture Division, Food and Agriculture Organization of the United Nations. 2022. Available from: https://www.fao.org/fishery/en/countrysector/dk/en?lang=en

18. European Commission. Case study: Portion trout in the EU: Price structure in the supply chain, focus on Germany, Italy and Poland. European Market Observatory for Fisheries and Aquaculture Products (EUMOFA); 2021.

19. Skov C, Berg S, Eigaard OR, Jessen TK, Skov PV. Danish fisheries and aquaculture: Past, present, and future. Fisheries. 2020 Jan;45(1):33–41.

20. Brit Tørud, Britt Bang Jensen, Siri Gåsnes, Synne Grønbech, Kristine Gismervik. Animal Welfare in Fish Hatcheries. Norwegian Veterinary Institute; 2019.

21. Semrau J, Ortega Gras JJ. Fisheries in Denmark [Internet]. European Parliament; 2013. Available from: https://www.europarl.europa.eu/RegData/etudes/etudes/join/2013/513972/IPOL-PECH_ET(2013)513972_EN.pdf

22. Iversen TM. Fish farming in Denmark: Environmental impact of regulative legislation. Water Sci Technol. 1995 Jan 1;31(10):73–84.

23. Nielsen R, Asche F, Nielsen M. Restructuring European freshwater aquaculture from family‐owned to large‐scale firms–lessons from Danish aquaculture. Aquac Res [Internet]. 2016; Available from: https://onlinelibrary.wiley.com/doi/abs/10.1111/are.12836

24. Nielsen M, Jensen F, Setälä J, Virtanen J. Causality in demand: a co-integrated demand system for trout in Germany. Appl Econ. 2011 Mar 1;43(7):797–809.

25. Veissier I, Butterworth A, Bock B, Roe E. European approaches to ensure good animal welfare. Appl Anim Behav Sci. 2008 Oct 1;113(4):279–97.

26. Noble C, Gismervik K, Iversen MH, Kolarevic J. Welfare Indicators for farmed rainbow trout: tools for assessing fish welfare [Internet]. nofima.brage.unit.no; 2020 [cited 2022 Aug 2]. Available from: https://nofima.brage.unit.no/nofima-xmlui/bitstream/handle/11250/2660614/Welfare-Indicators-for-farmed-rainbow-trout-Noble-et-al.-2020.pdf?sequence=1

27. Tørud B, Størkersen K. Work with Fish Welfare in Hatcheries. Norwegian Veterinary Institute; 2021.

28. Cox V. Milkfish Welfare in the Philippines. Charity Entrepreneurship; 2022.

29. Gåsnes SK, Oliveira VHS, Gismervik K, Ahimbisibwe A, Tørud B, Jensen BB. Mortality patterns during the freshwater production phase of salmonids in Norway. J Fish Dis. 2021 Dec;44(12):2083–96.

30. Gesto M, Jokumsen A. Effects of simple shelters on growth performance and welfare of rainbow trout juveniles. Aquaculture. 2022 Mar 30;551:737930.

31. China V, Holzman R. Hydrodynamic starvation in first-feeding larval fishes. Proc Natl Acad Sci U S A. 2014 Jun 3;111(22):8083–8.

32. Schuster R. Why do 99% of baby fish die? [Internet]. Haaretz. 2014 [cited 2023 Jan 26]. Available from: https://www.haaretz.com/life/2014-07-03/ty-article/.premium/why-do-99-of-baby-fish-die/0000017f-dbbf-d3a5-af7f-fbbf0be00000

33. FAO. EIFAC Workshop on Mass Rearing of Fry and Fingerlings of Fresh Water Fishes/Papers [Internet]. FAO. 1979 [cited 2023 Feb 24]. Available from: https://www.fao.org/fishery/docs/CDrom/aquaculture/a0844t/docrep/009/AE993E/AE993E14.htm

34. Sales J. First feeding of freshwater fish larvae with live feed versus compound diets: a meta-analysis. Aquac Int. 2011 Dec 1;19(6):1217–28.

35. The Fish Site. BioMar ramps up hatchery feed R&D [Internet]. The Fish Site. 2023 [cited 2023 Feb 24]. Available from: https://thefishsite.com/articles/biomar-ramps-up-hatchery-feed-r-d

36. Schukraft J. Invertebrate sentience: a useful empirical resource [Internet]. Rethink Priorities. 2019 [cited 2022 Apr 18]. Available from: https://rethinkpriorities.org/publications/invertebrate-sentience-useful-empirical-resource

37. Romero Waldhorn D, Autric E. Shrimp production: Understanding the scope of the problem [Internet]. Center for Open Science; 2022 Dec [cited 2023 Feb 24]. Report No.: b8n3t. Available from: https://ideas.repec.org/p/osf/osfxxx/b8n3t.html

38. European Commission. Welfare of farmed fish: Common practices during transport and at slaughter [Internet]. European Commission; 2017. Available from: https://publications.europa.eu/resource/cellar/facddd32-cda6-11e7-a5d5-01aa75ed71a1.0001.01/DOC_1

39. McAuliffe W, Shriver A. The Relative Importance of the Severity and Duration of Pain [Internet]. Effective Altruism Forum. 2022. Available from: https://forum.effectivealtruism.org/posts/7KJZWLMepdZ4mHvxS/the-relative-importance-of-the-severity-and-duration-of-pain

40. Gómez-Emilsson A. Logarithmic Scales of Pleasure and Pain [Internet]. Qualia Research Institute. 2019. Available from: https://qri.org/blog/log-scales

41. Branson EJ, editor. Fish Welfare. Chichester, England: Wiley-Blackwell; 2007. 316 p.

42. Yang Y, Wang T, Phillips CJC, Shao Q, Narayan E, Descovich K. Knowledge of, and Attitudes towards, Live Fish Transport among Aquaculture Industry Stakeholders in China: A Qualitative Study. Animals (Basel) [Internet]. 2021 Sep 13;11(9). Available from: http://dx.doi.org/10.3390/ani11092678

43. Sampaio FDF, Freire CA. An overview of stress physiology of fish transport: changes in water quality as a function of transport duration. Fish Fish . 2016 Dec;17(4):1055–72.

44. Nielsen M, Setälä J, Laitinen J, Saarni K, Virtanen J, Honkanen A. Market Integration of Farmed Trout in Germany. Mar Resour Econ. 2007 Jan 1;22(2):195–213.

45. Asche F, Bjorndal T. The Economics of Salmon Aquaculture. John Wiley & Sons; 2011. 248 p.

46. Dalsgaard J, von Ahnen M, Pedersen PB. Nutrient removal in a slow-flowing constructed wetland treating aquaculture effluent. Aquac Environ Interact. 2021 Sep 9;13:363–76.

47. Nielsen R, Andersen JL, Bogetoft P. Dynamic Reallocation of Marketable Nitrogen Emission Permits in Danish Freshwater Aquaculture. Mar Resour Econ. 2014 Sep 1;29(3):219–39.

48. Plesner LJ. State of play - Aquaculture and its legislation in Denmark. In 2020. Available from: https://submariner-network.eu/images/Denmark.pdf

49. Environmental Protection Agency. Comparison of legal regulation and technology level requirements, for aquaculture facilities producing rainbow trout in freshwater, in selected European countries [Internet]. Ministry of Environment and Food of Denmark; 2016. Available from: https://www.submariner-network.eu/images/BAT_report_trout_FINAL_Oct16.pdf

50. Nielsen R. GREEN AND TECHNICAL EFFICIENT GROWTH IN DANISH FRESH WATER AQUACULTURE. Aquacult Econ Manage. 2011 Oct 1;15(4):262–77.

51. Miljøstyrelsen. Punktkilder 2018 [Point Sources 2018] [Internet]. 2019. Available from: https://mst.dk/service/publikationer/publikationsarkiv/2019/dec/punktkilder-2018/

52. Bregnballe J. A Guide to Recirculation Aquaculture. FAO; 2015.

53. Billington T, Cerqueira M. Fish Welfare Improvements in Aquaculture. Fish Welfare Initiative; 2020.

54. Ebeling JM, Timmons MB. Recirculating aquaculture. 2010; Available from: https://biofile.moy.su/_ld/3/330_RECIRCULATING_A.doc

55. Tilman D, Clark M. Global diets link environmental sustainability and human health. Nature. 2014 Nov 27;515(7528):518–22.