Professor Braithwaite is a professor of fisheries and biology at Penn State University. She spoke with ACE Researcher Kieran Greig on February 20, 2018. This is a summary of their conversation.
How is the welfare of farmed fish usually measured?
There is no simple answer to this question, since different species have different indicators; nevertheless, there are a few general factors one can examine. One of these is health, which can be measured through disease rates, mortality rates, and body condition, among other things which demonstrate the general well-being of the fish. On top of this, behavior can be a useful measurement, insofar as one can determine whether a fish is swimming in a species-appropriate fashion, whether they are eating within their full area or only in certain parts, or whether they are schooling or swimming tightly.
Another parameter is the physical nature of the farm environment itself, with parameters such as water quality, oxygen saturation and—especially with certain species such as salmon—currents in the tank. One can also bring in health professionals to take specific measures and examine things like eye, skin, and gill condition, amongst other things. However, it is worth reiterating that these measures and analyses should be used differently for different species, and that different species have different health and environmental requirements. It should also be noted that what precisely is done, and how it is done, depends on who is involved and why.
Are there techniques that are more generally used or preferred?
Generally speaking, some of the most commonly used techniques are unsurprisingly those that are easier to measure or carry out, although there can be some contention about the reliability of some of these methods. Once again, it varies from species to species, but physiological and physical aspects are some of the most common. It is relatively easy for a farmer to measure water temperature or oxygen saturation. But if you have the right kind of operators, it is also possible to look at food consumption in order to assess changes in appetite, which tend to be good markers that something is wrong.
Body condition is a usually a good parameter, but is hampered by the fact that by the time the body condition of a fish has noticeably changed, it is often too late; this is normally seen at a late stage, suggesting that things have likely been going on for a while. This can be a serious problem for some measurement techniques, since they can only be effective if they are done before a certain point, which requires prediction. For example, if you’re concerned about the stress levels of fish, you won’t see the warning sign of increased cortisol unless you are already monitoring such hormonal changes, and by the time stress is more evident, it can become chronic and much more difficult to reverse or affect. Effective measurement techniques, therefore, should factor in early warning in order to avoid missing something. However, measurement can be complicated even in early stages, since one has to handle the fish in order to take a blood or cortisol sample. It is desirable to do so, but not simple.
What are some additional advantages or disadvantages of some of these commonly used techniques?
Some of the biggest advancements have happened in salmon and trout welfare because the European industry has been quite proactive. There is no single area that you can measure to gain an accurate understanding of welfare, so there needs to be a way to sort out the different variables; this has been done effectively in Norway, and by Professor of Aquatic Population Health and Welfare Jimmy Turnbull in Scotland. This kind of work involves looking at the index or indices for the particular sea cages or whatever set-up you are examining. You examine a number of factors—such as those mentioned above—and put them into a model and use it to determine the welfare status of the fish. These are some of the more powerful predictors of welfare status, but here are still some downsides. For one, you are measuring a huge number of things on a routine basis, meaning that there is a lot you need to get right. On top of this, if fish are physiologically stressed, this can affect the other measurements.
In terms of the weighting of these indicators, the group in Norway has also been able to evaluate the reliability of various measurements around many farms in order to develop weighted scores and practical norms. While the actual output of fish welfare evaluations can depend on what is being measured, they can still be described in terms of numerical scores—for instance, the Norwegians use at least two models which chart general fish welfare.
How would you describe the state of our knowledge about how best to measure farmed fish welfare?
Arguably, it is better than it was, and it has moved quite far along in the last ten years; however, there is still a long way to go. Although these models are a start, we still need several improvements; for instance, measurement processes are currently quite unwieldy to use, and many of the most important advancements in terrestrial farm animal welfare that have occured and are occuring relate to the simplification of processes for quick, accurate solutions from measurement. We might be asking too much of farmers at the moment, and this increases the risk that they might take shortcuts or do things at a lower quality.
Essentially, there are areas that are currently well-understood, but much work needs to be done to fill in both methodological and epistemological gaps. One significant problem at the moment is that when we talk about fish welfare we talk in terms of a hegemonic group, yet we really need to examine the specific needs of different species.
When reading relevant literature, should we be concerned about validity? Do some publications have vested funding interests or known biases?
There is not too much need for concern about bias, except perhaps in terms of what the research is actually looking at. There has been much success in terms of humane endpoints, and there is therefore plenty of literature on this issue. Conversely, some areas are very difficult to work on, which means there are fewer experiments of certain types published. There isn’t a particular bias based on the funding of corporate groups; however, there can be great differences between funding influences in the U.S. and in the E.U., although this is not particularly an issue for fish.
What welfare interventions do you think are the most meaningful?
This is something that is again species-dependent, and what we know about methods varies between different fish. Less is known about carp because research has so far been less detailed although there is some work going on, especially when it comes to humane killing. Knowledge of catfish species is moving a little more, but its progress remains somewhat slow. Most is known about trout and especially, salmon.
In terms of meaningful welfare methods, it really depends on what one is trying to measure. Humane endpoints remain low-hanging fruit in terms of welfare—much is known about farmed salmon welfare because a lot of work has been done on electrical stunning. However, although this is essential morally and for methodological understanding, it is just one stage in the animal’s life, and one where it is not always easy to see what has been done well. Nevertheless, some meaningful welfare methods might involve bringing other species’ endpoint research up to the level of salmon. In any case, it does not seem to be the case that there is enough knowledge on catfish or carp at the moment to say what is the most meaningful measure.
To what extent does the size and location of farmed fish populations around the world affect intervention aims and methods?
Welfare in Asian countries is behind other regions, in the same way that the U.S. has been about 10 years behind Europe. It is worth discussing which species are most numerous—catfish are without a doubt the most farmed fish in Asia and globally, with a large proportion coming out of Vietnam. As already discussed, a lot more is known about salmon and trout but there is a significant drop-off when it comes to lesser-farmed fish, with research and interventions in other areas being much less developed. I’ve looked at sturgeon and would like to learn more but that part of the industry is still somewhat behind the salmon and trout industries—salmonid farmers and companies have been very proactive compared to other parts of the industry.
To what extent can the knowledge we have about salmon and trout be transferred to other species like carp, catfish, and pangasius?
Some things transfer well, but others do not—there are a number of differences in the aforementioned indices and models in terms of their final results, but they are actually looking at similar things in the process. For example, they look at stocking density, daily mortality rates, water chemistry and temperature, the general health and condition of fish, and their appetites, among other things. These indicators give you information about how well the fish are in a general sense, and they are applicable across species. Although numbers vary, there are similar factors in measurement.
Many factors differ between species, especially when it comes to physiology. Fish biology also needs to be better understood. For example, catfish cope well at higher stocking densities than other species—while this might be unhealthy for other species, certain catfish can survive and even thrive in these conditions and should avoid lower stocking densities. The industry needs qualified biologists to come in and examine the indices and measure them against the physiology of the fish in order to come up with a weighting system for welfare factors. This doesn’t have to be a huge task, but it depends on what is required. It might take between six months and a year to get a list of important welfare considerations, but the actual testing might take around three years.
In terms of internal species differences—for example, for different species of carp—will there need to be different measures, or will they tend to be similar?
This also depends on the species, biology, and physiology of the fish. With carp, for instance, there is huge diversity; crucian carp can cope with oxygen-poor water for weeks on end if necessary due to their physiological adaptations. Other species of fish and other types of carp are more sensitive to this. Given that there is already a large range of species that we currently farm, there is a need to understand how each of them differ. We also need to be careful about introducing new species to farms since they can create even more complexity. There needs to be investment and care when it comes to gathering information and planning interventions relating to different species’ environments.
What are the most promising ways that fish farming practices could be changed in order to improve fish welfare—what would be some of the easiest things to do in order to lead to significant welfare improvements?
This has already been answered to some extent for salmon, but for other species there needs to be recognition that there is a need for welfare. Cultural, regional, and global differences have a huge impact, and farmers of other species and in different geographical region need to become more willing if we hope to see an industry change like we did with salmon. Salmon farming has seen a lot of progress largely because the industry was willing to make changes, and because they looked into humane harvesting early on. This is not currently the case for carp or catfish, suggesting that a lot more should be done for the more relatively novel species, and for species for which there is a lack of knowledge or research. Indeed, understanding species can lead to very large gains. For example, for salmon it has been found that percussive stunning can be particularly effective due to the fact that they have uniquely shaped heads which allow for the design of a precise apparatus. Yet for other species, other forms of stunning might work much better—for instance, electrical stunning seems to be better for carp and pangasius.
Is there adequate evidence for any other interventions for catfish, carp, or pangasius? For example, do we know about the effects of specific factors like stocking density and oxygen level?
At present, stocking density is a hard thing to deal with full-stop and needs improvements. One option could be to look at guidelines, but there are no straightforward answers and the factors are different for different species—for example, catfish can withstand higher stocking densities than some other species.
Despite the fact that these specific factors are not low-hanging fruit and are especially challenging, this is not to say that we shouldn’t look at them.
Apart from slaughter reform, are there any other reforms we could campaign for that would apply across most of the commonly-farmed species? What about for transportation?
There are certain things we can do. For example, we could examine what physiological traits do well in captivity and use selective breeding to increase these traits and harden fish tolerance. Where we have bred selectively—although often by accident—we have ended up with problems like aggression, so there is a need to pay some attention to these things in conjunction. Perhaps we could use modern techniques to elucidate and aid in developing changes in genetics.
In terms of transportation, salmon tend to be transferred at least twice in their lives, but this is not necessarily done for other species. I was on a fish farm in California about a year ago where I was surprised to find that the carp there were all transported live, because that was the way that consumers wanted them. This brought up some interesting issues. The transportation wasn’t particularly well regulated or monitored, and neither was their storage in actual stores which could last a significant amount of time.
Are some fish farming methods (raceway systems, nets in oceans, nets in ponds) better than others?
What makes a farming method good depends on what you are trying to focus on and improve. There are some interesting projects in Norway where they have decided to no longer farm the fish out at sea; rather, they use land-based hangars with what are essentially olympic-sized pools full of fish. The fish are placed in and moved through different parts of the pool depending on their size and species, and change whether they are in salt or fresh water depending on what stage of their lifespan they are in, or whether they have species-specific needs. This is mainly done to prevent disease, since infections such as sea lice can be a serious problem for welfare. These large, land-based tanks are doing a good job at maintaining fish health in this sense. They’re also especially good for large numbers of salmon and trout because it is easier to control their environment this way, and because the fish are taken out of harm’s way. They can be very expensive though, and there are different parasitic diseases for different species.
What is your opinion of the various welfare certification schemes for fish welfare like Global Gap and RSPCA Assured?
I’m not familiar with all of the standards available, but I’m aware of the RSPCA Assured labelling system and assessment scheme and I would certainly refer to it. It was created by people who have a strong grasp of fish welfare and it’s very well validated. Other labels might be similar, but people should first ensure their validation and source before using them.
What additional research do you think is most needed to decrease uncertainty about farmed fish welfare interventions?
One of the things that we rely on hugely in farmed animal welfare is an understanding of animal behavior. However, there is less of an opportunity for this when it comes to fish because they can often be in dark environments and/or living in large schools.
One of the ways that the chicken farming industry has looked at this is with the use of optic flow. Marian Dawkins’ work has contributed significantly to this: cameras were put into large chicken sheds in order to observe thousands of birds at a time. These optic cameras recorded the chickens and were then uploaded onto a computer program to examine their flow and movement with the scope of understanding their health and behavior. For example, if a bird or group of birds become lame, the flow changes and the program picks it up very quickly and much more effectively than other, more basic methods of observation.
Something very similar could be done for fish. Whether in a runway, a net, or deep water, it could be possible to view changes in movement related to fish health with cameras tracking regular movement patterns. For example, fish tend to avoid sick individuals and groups—this is something that the optic camera would be able to capture early on. On top of this, fish moving to warmer water might be an early sign of illness. This has been studied on a somewhat small scale in lab conditions, but could certainly be scaled up with funding.
Is there anything else you would like to mention?
Certain characteristics should be selected in breeding—one of the things we currently know is that when it comes to behavioral phenotypes, hardy animals that can cope with handling do much better since handling is a significant part of farming.
There is also a need to think more in terms of practice: there should be an examination of technologies which might be used to reduce farmed fish suffering. This applies especially to handling—with salmon, there has been a net benefit from reducing handling through creative uses of technology to move fish around. There is also a somewhat underappreciated need to provide fish with autonomy in their environment; if they can choose their feeding times, for instance, then there seems to be less food waste. Technology could also be used to trigger food release when the fish are hungry, giving them control over portions and sustenance. This has the added benefit of providing stimulation and variation in their environments, which is not discussed a lot, but is still very important for welfare.