Krillin’ it: the Antarctic Krill fishmeal fishery

Krillin’ it: the Antarctic Krill fishmeal fishery

Krill remains one of largest underexploited marine resources in the world.

Antarctic krill (Uwe Kils, 2011)

Krills are a group of various zooplankton that generally look like small shrimps. Krill and other zooplankton (free-floating organisms that don’t photosynthesize their own energy) float in the water column and represent a link in the food chain between primary producers (like phytoplankton) and larger animals ranging in size from herring to whales.

One of the largest populations of krill in the world, Antarctic krill, is in the southern ocean. It is here that it plays a role as a keystone species as the main species through which nutrients and energy are transferred to higher trophic levels. The harvesting of this krill is under strict

Antarctic krill habitat (NASA, 2005)

regulations to ensure there is enough of this krill for higher trophic level organisms, including penguins, whales, and fishes. Its biomass has been difficult to estimate, but the current estimate is near 400 million tonnes. Even with a generous precautionary limit for the ecosystem function, this could be a huge source for the world’s fishmeal and oil needs. However, the habitat of this krill is the sea ice, which is becoming less stable and krill may lose a substantial part of their habitat over the next 100 years.


The major argument for catching more krill as an input into feeds is that its demand as a direct human food is very limited. There have been some previous examples of it being processed into human food in Japan, but this was fairly limited in scope. In contrast, the former USSR in the 1980s, and now Norway currently have large krill fisheries (>100,000 tonnes annually) for the production of fishmeal. Krill are targeted by pelagic and bottom trawl fisheries that use a very fine mesh net. The organisms are so small that this is the only functional way to catch these organisms.

The impediment to these fisheries growth has been a factor of cost so far. Krill meal and oil are too expensive to use as the major constituent of feeds and thus must be supplemental for certain nutritional properties. However, as fish that can be used as food are re-directed more towards this purpose, the cost paid for these fish will rise. Eventually, krill will become a more cost effective feed ingredient, and we’ll be feeding more of our fish with krill. As this increasingly becomes the case, we’ll need to watch the development of these changing ecosystems to ensure we continue to fish with the whole ecosystem in mind.


Gears used to Fish for Feed

Gears used to Fish for Feed

There is a multitude of different fishing gears that are used to catch fish. These gears generally vary in the scale of fishing undertaken, and are dominated by a few categories of gear. These different gear types and categories range in their impact on the ecosystem’s habitat. In addition to my work on reduction fisheries, I have started a project on global fisheries gear use to assign all fisheries catches to a general gear type based on the fishery, country, species, and time period. This will give a great insight into the evolution of gear use over time and global trends associated with gear use; but, it has also made me think more about the effects of different gears.

The two main gears most important to dedicated reduction fisheries are pelagic trawls and purse seines. A pelagic trawl, or mid-water trawl, is a net towed behind a boat that is

Mid-water Trawl
Pelagic trawl (FAO 2016)

targeting species not on the ocean floor. This is in contrast to a bottom trawl that generally targets species that live on or near the ocean floor. In contrast to the bottom trawl, a pelagic trawl has a minimal impact on the ocean floor, as it is not meant to touch it. Gears like bottom trawl fisheries have a large impact on the surrounding ecosystem by catching non-target fish and changing the physical structure of the ocean floor ecosystem. As pelagic trawls don’t drag along the ocean floor, it also catches less of species that are not targeted, both fish and otherwise, a common complaint against bottom trawl fisheries.


Purse seines on the other hand operate by releasing a net off a fishing boat and then making a sharp turn in a circle to meet up with the net. In this way, the net

Purse seine diagram (AFMA 2015)

surrounds a targeted school of fish and then the ‘purse wire’ is drawn tight at the bottom
of the net like an old-styled purse with a drawstring. The fish can then be brought aboard from the contained net. Purse seines are frequently used to target fish for fishmeal like Peruvian anchoveta, Atlantic herring, and Chilean jack mackerel, but are also a major gear for tuna fisheries worldwide.


While these gears make up the majority of dedicated reduction fisheries, by-catch from bottom trawls (including shrimp trawls) are increasingly retained and turned into fishmeal. I have touched on this earlier in my post on sources of fishmeal and this gear

Bottom trawl drawing (Wikimedia 2005)

is important as both a source of by-catch and for ‘biomass fishing’. As bottom trawls are highly nonselective in the species they catch, this is problematic in monitoring the health of the individual fish populations. Furthermore, this method of fishing can have destructive impacts on the physical structure that is often built by slow growing organisms such as corals. This physical structure is important for the ecosystem as it creates a sheltered environment for small and juvenile fish, and other organisms.


Both pelagic trawls and purse seines are considered relatively ‘clean’ gears with little or no by-catch when compared to other gears. The other major difference in these gears is that they have relatively low fuel use per tonne of fish landed when compared to other gears like bottom trawls. This makes them ‘cleaner’ from a carbon footprint standpoint, as fuel use is a major contributor to the carbon footprint of fisheries, and to fishmeal and oil production. These dedicated reduction fisheries generally operate as a very low-impact fishing method when compared to other fishing gears. However, as more and more fish for fishmeal is coming from by-catch of bottom trawl fisheries and a shift toward biomass fishing, the general trend seems to be a shift from these cleaner methods to a more problematic method of fishing for fishmeal. As a whole this would signal a shift from lower impact fishmeals to higher impact fishmeals mainly based on the gear being used for catching the fish. This highlights the importance of gear when evaluating different forms of fishing for feed.

The other half of the pie: agricultural feed inputs

The other half of the pie: agricultural feed inputs

Most of the feed given to farmed salmon isn’t fish. In fact, fish inputs make up a smaller and smaller portion of most salmon feeds and in Norway are around 30%. While the focus of this blog is the use of fish as feed in livestock and aquaculture, it is important to consider other parts of the equation. In this post, I will cover what makes up the other portions of commercially produced fish feeds, namely land-based inputs. The bulk of many fish feeds today are agricultural inputs rather than fish inputs.

Norwegian Salmon Feed
Cashion et al. (2016) 

However, the impacts and concerns associated with agricultural inputs are quite different as compared to marine inputs. While I have discussed the impacts of the use of fish as feed instead of food in part, the use of non-fish inputs carry their own costs.

The food security concerns of using crops for feed are similar to using fish for feed. While animal protein and certain micronutrients are more important when considering fish inputs, the major other constituents form the bulk of calories consume globally.In general, the major inputs into commercially produced feeds are proteins from wheat, corn, and soy, and fats from soy and canola. Corn, wheat and rice are the three crops that contribute the most to humans’ caloric consumption at roughly 2/3 globally. These crops and their processed products are thus obviously important for global food consumption, but in a different way from fish products. Fish products are often consumed for their high-quality proteins and micronutrient content (such as Omega-3s in addition to other vitamins), rather than to form the bulk of the calories of one’s diet.

Brassica rapa
Canola is now a major fat source for aquaculture diets

To grow fish in aquaculture, a certain nutritional balance must be sought but there are also fundamental constraints about how that feed can be converted into biomass of the fish. In other words, there are limits to how much we can reduce the amount and energy content of feeds and still get out the same quantity of fish at the end of the day. Recent advancements have been made at reducing the ‘feed conversion ratio’ (e.g., the amount of feed given to the animal compared to the amount of animal food we get out) and at reducing the fish content of certain feeds, but these result in a shifting of the burdens associated with them. A reduction in fish content (be it meals or oils) is often paired with a substitution of agricultural inputs such as increased soy protein or canola oil. While this reduces demand on marine ecosystems in this form, it also increases the demand for these agricultural products. For the displaced environmental burdens, this is called ‘environmental problem shifting’ where pressure and impacts are shifted from one part of the system to another. While substituting fishmeal protein for soymeal protein can reduce pressure on marine ecosystems, the increased demand for soy globally is causing increased conversion of sensitive ecosystems into agricultural land for soybean production, such as in the extreme example of the clearing of the Amazon rainforest to grow soybeans.

However, if the feed conversion ratio can be improved, this means that less input is required to produce the same amount of desired end product. This reduces costs not only for the farmer, but environmental costs associated with producing these inputs in the first place. As fish inputs into aquafeeds are often the most costly, feed developers and farmers that produce their own feed want to minimize fish inputs while not compromising on the feed conversion ratio or the quality of the final product. When not enough fish inputs are included the result can be a slow growth rate and/or a low-quality product that might not appeal to consumer’s tastes. By reducing the length of time the species is farmed before it can be sold, fish inputs can improve the economic and environmental performance of aquaculture in some areas.

While many contest the use of fish as feed inputs in aquaculture, they should be used to maximize food security outcomes across these different methods of food production including capture fisheries, aquaculture, and terrestrial food production. When considering the reduction of fish inputs to aquafeeds, it must be balanced by knowing that if the same production is going to continue those other inputs need to come from somewhere. It is thus not enough to minimize fish inputs to fish farming, but we must seek to minimize the environmental, social, and economic costs associated with seafood production.

Primary production required for fisheries products

My first peer-reviewed article came out a few weeks ago and can be found here, or through my ResearchGate profile here. The article focuses on refining a measure of biotic impact as measured in life cycle assessment. Life cycle assessment attempts to quantify most environmental impacts of a product from ‘cradle-to-grave’. Thus, the greater consideration of biotic impacts is an important step forward for my focus area: fisheries. As we take fish out of the sea, we need to account for the impact on the ecosystem. This has been most commonly done through a measure of the primary production (or photosynthetic activity) that would have had to occur at the bottom of the food chain to produce these higher food chain items. The method used to date to quantify this primary production required (PPR) did not take into account differences in how energy transfers up the food chain in different ecosystems or the total level of primary production in that ecosystem, or some species-specific factors that I’ll get into later.2000px-ecological_pyramid-svg These ecosystem factors are important to consider as some ecosystems are less efficient at transferring photosynthetic energy up to higher trophic levels, and some ecosystems have lower amounts of primary production occurring annually. Taken together, this means some ecosystems will have less energy to begin with (in their annual energy budget if you will), and even less that will make it to higher trophic level species that we often harvest in fisheries for human consumption. From an energetic perspective, some fish represent a much greater proportion of that fisheries annual ‘energy budget’ than other fish.

I addressed this problem by trying to give better consideration to the factors mentioned above. First, I looked into how different ecosystems transfer energy and came across something in ecosystem modeling called transfer efficiency. After completing a biomass model of how an ecosystem functions, transfer efficiency can be one of the products based on how efficiently energy is transferring to higher trophic level organisms in the ecosystem. This ecosystem-specific transfer efficiency for somewhere like the North Sea will give an indication of energy transfers (for the North Sea it is more efficient that the global average of ~10-11%) in this specific ecosystem and make the results of the PPR equation more specific to that ecosystem. For the North Sea, the transfer efficiency is higher than the global average of ~10-11% meaning that based on this alone, harvesting fish biomass in the North Sea will require less base level photosynthetic activity than the global average.

Second, I included the total annual ecosystem primary production of the source

Chlorophyll concentration in the oceans

ecosystems I considered. I could thus relate the removal of fish biomass in the form of primary production to the total ecosystem’s primary production to give an indication of the scale of how much fish biomass we are removing in the context of total energy available in the ecosystem annually. Again, this will make the results more specific to the ecosystem we are considering through contextualizing the impact within the affected ecosystem.

Finally, I explicitly used species-specific energy and fishmeal and oil yields. Different species yield different amounts of fishmeal and especially fish oil depending on their characteristics (mainly the fat content of the fish at the time of processing). While the other improvements consider the ecosystem the species was sourced from, this focuses on the species itself making the results more specific to the species and its attributes.

While all of these factors may seem small, the continual ignoring of them meant that the results of the previous research often did not consider these factors to their full extent. Differences in the species and ecosystem factors mentioned above generate starkly different results. The figure below shows a difference of a factor of 20 for the same species harvested from different ecosystems. All other factors were the same, except the source ecosystem’s transfer efficiency and underlying primary production.

For me, I am a researcher so that the findings may be applied to move humanity towards a more sustainable future. If the previous method cannot distinguish between a 20x difference in ecological impact in one ecosystem compared to another, then the method must be improved. If we demand more from the companies that produce these feeds, we at least need to give them the proper tools to make these decisions. I know that many of these companies take their commitments seriously and have funded research that explores how they can improve their environmental performance. The fact remains that we have been using a system that does not consider all the factors it could, and my improvement to the PPR measure is one such example. There are other areas where we must continue to expand on our research, and this was a small area that I was lucky to be a part of working on.

Sources of Fishmeal

In my last post, I discussed the largest source of fishmeal and fish oil globally, reduction fisheries (if you didn’t get a chance to read it, catch up here!). Reduction fisheries have been the primary source of fishmeal and fish oil since at least the late 1950s. In this article I’m going to give an introduction to three (mainly) distinct other sources of fishmeal and fish oil. These are important to consider as not all sources of fishmeal are of an equal quality (for when the products are used, such as lower protein content or a lack of some important amino acids), or of equal environmental and ecological impact. This second point is much more nuanced and will be the subject of many future posts, but I’ll give an introduction to some of the considerations here. The four main distinct sources are: reduction fisheries, by-products of human consumption fisheries, by-catch of human consumption fisheries, and biomass fishing.

  1. Reduction Fisheries

Reduction fisheries are fisheries for the dedicated purpose of landing fish for the production of fishmeal and fish oil. They are the largest producers of fishmeal and fish oil globally (65% of global production according to the UN’s Food and Agriculture Organization; FAO 2014), and are also some of the largest fisheries globally including Peruvian anchoveta and Atlantic herring. These fisheries set out to target specific fish for their characteristics and the resulting quality of fishmeal and fish oil they can produce. As these fisheries generally catch only their target species (with small amounts of other by-catch), they can be managed as individually which distinguishes them from two other sources described below.

  1. By-products

Fishmeal is often made from the processing by-products of direct human consumption fisheries and, according to the Food and Agriculture Organization, accounted for 35% of fishmeal production in 2012. This is a huge increase from recent years and predictions are that fishmeal from by-products will surpass fishmeal from reduction fisheries over the next 50 years. These products come from using the parts of the fish that humans often do not eat. Once the desired part of the fish is removed (often the fillet, other muscle tissue, or roe for herring and sturgeon fisheries), the remaining portion (often called trimmings, carcasses, or by-products) is processed

Processing of hake leaves by-products that can be used for fishmeal

into fishmeal. The amount of fishmeal produced from these by-products (historically mainly derived from ‘whitefish’ including Atlantic cod, pollock, and hake) is less than the yield when whole fish from reduction fisheries are used, and the resulting fishmeal is often of a lower protein content and lower quality of amino acids present.


Recently, there has been a developing body of work to quantify the environmental and ecological impacts of fisheries and aquaculture such as in ecological footprint analysis or live cycle assessment. When accounting for the environmental and ecological impacts of using these by-products, there is a debate over whether these products are inherently ‘free’ because they are not the main purpose of these fisheries and they would theoretically go to ‘waste’ after. However, this ignores the physical reality that energy was expended by the fish (or other organism) to grow these parts that humans have decided are undesirable, and that the catching and transporting of this mass requires the same amount of energy regardless of the value to the fishers. The ‘by-products’ therefore cost energy to acquire, but also represent usable energy in the form of potential food or feed products. They therefore share the environmental and ecological costs associated with their source fisheries, which becomes important when comparing different sources of fishmeal and fish oil.

  1. By-Catch
Photo Credit: Patrik Henriksson (

By-catch is often caught along with the targeted species in fisheries, and the use of this catch is mixed. Some by-catch, often being low-value catch, is discarded back into the water with a fairly high likelihood of mortality. However, some by-catch is landed due to economic value or legal requirements. Norway, for instance, requires all by-catch to be landed in its no discard policy for its fisheries. Much of this by-catch that is not marketable for human consumption is sold to fishmeal factories. Under this system, fishers are incentivized to reduce their by-catch as it is less valuable than their target species, but they are also effectively subsidized for landing fish with no market value for human consumption. While this system reduces by-catch and increases transparency of by-catch and discards, the effect of harvesting individuals that may have survived being discarded has not been fully examined. The major source of fishmeal from by-catch is from shrimp fisheries in Southeast Asia and China where all the catch is landed but only a small fraction was targeted shrimp. The majority of the catch is sold for fishmeal production.



  1. Biomass fishing

The distinction between using by-catch for fishmeal production, and biomass fishing is the lack of a target species. Biomass fishing occurs mainly because the former target species (in highly non-selective fisheries like shrimp trawling) is now present in such a low abundance, that there is no incentive to sort through the catch for the former target species, and there is an option of selling all of the catch to fishmeal producers. This non-selective fishing thus becomes for ‘fish’ biomass as a source of protein. This raises concerns separate from reduction fisheries as reduction fisheries intend to target 1 or possibly 2 species whose populations can be managed in fisheries stock assessments. However, targeting all species that can be caught by a bottom trawl fishery means that no species or population is being managed individually and that some are likely to have their populations depleted at a faster rate than others. Many of these concerns apply to sourcing fishmeal from by-catch, as the non-target species populations are often not being managed.

While these forms of sourcing inputs for fishmeal production are fairly distinct, there is obviously some overlap. By-catch and biomass fishing sources both rely on non-targeted species to a certain extent, in that they are not the target species. Some consider by-catch to be the by-product of fishing where the method is not 100% selective for the target species, which encompasses most fishing methods (more to come on gear and gear selectivity in future posts!). While by-products have been playing a larger role in global fishmeal production, some of my current research will uncover more on the large role of by-catch and biomass fishing in Southeast Asia and China, which until now has not been distinguished completely from by-products. It is important for researchers to distinguish these different sources of fishmeal as they each have their own costs and benefits associated with them. This is part of my work, and what I will delve into deeper here.

Fishing for Feed explained Part 2: Reduction fisheries and their products

Fishing for Feed explained Part 2: Reduction fisheries and their products

In my first post, I wrote about the importance of fisheries for feed. In this article, I’m going to get into the main form that these fisheries occur and that is something called reduction fisheries. Reduction fisheries are fisheries for the purpose of the production of two products: fishmeal and fish oil. Fishmeal and fish oil are essentially the stabilized product form of fish protein and fish fats, respectively. They are called reduction fisheries because they ‘reduce’ whole fish to their component parts of fat and protein. This is the main form of fishing for feed, currently consisting of 75% of capture fisheries destined for non-human consumption uses (I’ll get into details on this in a later post). Reduction fisheries are those fisheries that are specifically for this product, and are not directed for human consumption.

As I mentioned, fishmeal and fish oil are the stable forms of these fish proteins and fats. This is important because fisheries are often seasonal depending on the abundance of the fish in the water, and this takes a seasonal product and turns it into something that is stable for a longer time. This occurs through processing of whole fish.

Fishmeal and fish oil are processed through essentially squeezing and cooking the fish (details can be found in this old UN Food and Agriculture Organization report). The squeezing presses the fats out the fish and is collected as fish oil, a valuable product for fish needing fatty diets (think salmon). There is a growing market for fish oil for human consumption, mainly in the form of Omega-3 supplements.

Fish oil Omega-3 supplements

Fishmeal on the other hand is the dried powder of fish once almost all the water has essentially been cooked out of it. Fishmeal is a complete protein (contains all amino acids) and thus is very effective as a supplement to livestock diets, and contains some amino acids that are rarely available from other sources. The quality of the protein in fishmeal is what makes it such an almost necessary ingredient to livestock production, and specifically aquaculture of piscivorous fish (fish-eating fish like salmon and tuna). However, even fish that are naturally herbivorous or omnivorous like various species of carps can benefit significantly in how fast they can be grown from supplements of fishmeal to their diets.


Separate from reduction fisheries, fishmeal and fish oil are often made from by-products of processing fish for human consumption (you don’t see a lot of fish carcasses in the grocery stores but you do see a lot of filets). This takes something that might be wasted and turns it into a product that is useful for livestock and aquaculture, although the protein quality is often less than that of fishmeal from reduction fisheries.


Schooling Pacific jack mackerel

Getting back to the fishery though, reduction fisheries are often (not always though!) targeting small schooling surface dwelling fish (AKA small pelagics). Common fish include various species of anchovies, sardines, herring, and menhadens. The schooling behaviour of these fish, and that they are often very numerous make these fisheries very efficient in getting them out of the water, and they can use gears that often don’t negatively affect other sea-life. The scale of these fisheries however can lead to overexploitation of the fish, which can lead to long-term collapses of population. In addition, these fish populations fluctuate with the productivity of the organisms below them on the food chain. These two factors can complicate management as its easy to fish a lot of these folks, and their population can fluctuate wildly based on climate and productivity of the ecosystem.

Reduction fisheries are the main vehicles for the production of fishmeal and fish oil, and fishmeal and fish oil are the main ways that fish are used for feed. This makes them very important to my research on how we use fish for feed and for food. They make up a huge portion of total capture fisheries, and are some of the largest fisheries in the world. And most importantly, if you eaten fish, pig, or chicken in the last few years, you’ve likely indirectly consumed some or a lot of these fish. Given that these fish are an important input into so many food production systems, we need to understand these fisheries impacts on their ecosystems, the people who fish, feed, and eat them, and the planet.