How to properly calculate Fish-In Fish-Out (FIFO) ratios

Thanks to regular reader and commenter Dev for suggesting this for our library. We have added it to the “Feed” section.

Critics of aquaculture often cite that it takes X pounds of feed to grow one pound of fish and that this is unsustainable.

However, the calculation they often use to come up with this figure is wrong, because their calculations fail to account for the fact that the same fish is used to make both fishmeal and oil;how moisture is a factor;  and also that trimmings from other fisheries are used to make fishmeal and oil.

Here is a more detailed explanation of how FIFO is actually calculated, and how the ratio is actually far lower than critics say.

The ratio for salmon, according to the FIFO calculations, is 1.68 to 1.


4 thoughts on “How to properly calculate Fish-In Fish-Out (FIFO) ratios”

  1. I think we are on the same page (I used numbers for “salmonids” which includes trout for total production and for the total amount of meal and oil used salmon aquaculture). The Atlantic salmon numbers are in agreement with mine if I subtract the trout data. If we go with your numbers then, according to The State of World Fisheries and Aquaculture 2012, FAO in 2009 17.9 million tonnes (Mt)of fish were caught in the global reduction fishery and that created 5.74 Mt of fishmeal and 1.07 Mt of fish oil. Salmon farming used 13.7% of the total fishmeal produced and 36.6% of the total fish oil produced. Maybe we can also agree that the world production of salmon (excluding trout) in 2009 was 1.465 Mt. (it’s the number used in the iffo paper we are directed to).
    Perhaps where we don’t agree is on the assumption that fish oil and fish meal are somehow fungible. It’s true, forage fish provide both oil and meal. And it’s also true that fish feed includes meal and oil. But I feel uncomfortable with just adding the two together as is done in the numerator of the iffo equation,

    FIFO Ratio = (Level of fishmeal in the diet + Level of fish oil in the diet/Yield of fishmeal from wild fish + Yield of fish oil from wild fish) X FCR

    It seems to me, for example, that 99 tonnes of fishmeal plus 1 tonne of fishoil is different from 99 tonnes of fishoil plus 1 tonne of fishmeal; fishmeal was historically a by-product of oil production and used for fertilizer; fishoil has always been dear. The impact on the limited world supply of omega 3 would certainly be radically different with the two cases. Yet here each gives the same result (100 tonnes). This new way of measuring FIFO is missing something.
    Let’s see if we can figure it out. We have the happy (and unusual) situation where we agree on enough of the numbers to get pretty close to an answer. So let’s try a little math. And lest we forget, “The SAME fish that made the oil also made the meal.” Correct me when I trip up.
    If we look at just the oil side of the equation, to reiterate, 36.6% of the total oil production or 391,620 tonnes went to produce 1.465 Mt of salmon. Since it took 17.9 Mt of fish to produce 1.07 Mt of oil, it must have taken 6.55 Mt of forage fish to produce the oil to produce 1.465 Mt of salmon, a ratio of 4.5:1.
    “Wait a minute.” I hear the protestations. “What about the fishmeal that’s left over?”
    Indeed, the same 6.55 Mt of wild fish used to produce the oil also produced 2.1 Mt of fishmeal (since globally, 17.9 Mt produced 5.74 Mt of fishmeal). Since only 0.786 Mt of fishmeal was needed for salmon there must be 1.31 Mt of fishmeal in excess to what was used. We can’t just ignore that.
    But clearly, as we already pointed out, we can’t count this fishmeal as equivalent to fishoil either; if we did, this excess 1.31 Mt would amount to a quantity that is more than the total world supply of oil, certainly absurd. What to do?
    For me the solution is simple. Really, I’m only concerned with the efficient use of that part of the ocean resource that is truly limited. I don’t think I’m alone. It is in fact what’s driving the growth in aquaculture, namely, “the absence of cost-effective alternative sources of dietary lipids rich in long-chain highly unsaturated fatty acids (HUFAs), including eicosapentaenoic acid (20:5n-3) and docosahexaenoic acid (22:6n-3).” (UN paper referred to earlier) Reasonably, I would concern myself with that.
    This important factor can be easily included in the calculation. The quantity of oil in the left-over fishmeal is 111,972 tonnes (fishmeal is 8.55% oil). Since we didn’t really use it, let’s subtract it from the amount of oil used in our original determination. To produce 1.465 Mt of salmon, then, requires only 391,620 – 111,972 = 279,648 tonnes of oil. Now the amount of forage fish required (at 16.7:1) to make this net amount of oil for salmon aquaculture is 4.670 Mt tonnes. FIFO = fish-in/fish-out = 4.789 Mt / 1.465 Mt = 3.27:1
    Here we have a rough idea of how much omega-3 (oil is 20% omega-3) is removed from the ocean to produce salmon. In an ideal world, that is.
    When you work out FIFO from feed composition as we did here, there are a lot of important things that are missed. A quick, back of the envelope calculation demonstrates this. The total oil used by the salmon farming industry is 0.4545 Mt. (0.392 Mt of fishoil plus 0.0629 Mt that came with the fishmeal). Forage fish (using iffo numbers) are 8.55% oil. Therefore 5.316 Mt of forage fish were used to produce 1.465 Mt of salmon. FIFO is 3.6:1.
    Why is this number higher?
    It’s partially because this simpler way of calculating FIFO starts with fish in the ocean and ends with salmon produced; all the “local” variables that intermediate when you make the calculation using feed composition: added fishoil and fishmeal derived from by-products, fish that are fed but don’t make it to market because of disease or escapes, countries that have different FCR’s, sporadic higher fish oil diets for hatcheries or for final conditioning, during smoltification and/or early rearing, etc.–non of these alter this result. It’s also closer to what we really want to know.
    I understand both these methods scant the value of the protein in the left-over fishmeal, but that’s better than ignoring the special value of the omega-3 altogether. The iffo formula is dangerous to efforts for global food security because by equating protein and omega-3 to obtain a low FIFO it gives false assurance.

    1. We think we are in agreement here.

      You said:

      Really, I’m only concerned with the efficient use of that part of the ocean resource that is truly limited. I don’t think I’m alone. It is in fact what’s driving the growth in aquaculture, namely, “the absence of cost-effective alternative sources of dietary lipids rich in long-chain highly unsaturated fatty acids (HUFAs), including eicosapentaenoic acid (20:5n-3) and docosahexaenoic acid (22:6n-3).” (UN paper referred to earlier) Reasonably, I would concern myself with that.

      We totally agree!

      Responsible salmon farmers have seen this coming for a while and in the last decade the amount of fish oil and meal in the feed has declined dramatically. But there’s that bottleneck where we have to find ways to reduce our dependency on wild fish, but still grow fish that have a healthy amount of omega-3 in their flesh. Plant and algae-based oils may hold the answer, but we’re still investigating the possibilities.

      Your calculations look right. Working it out from the UN FAO numbers, the 3.6:1 ratio looks accurate. So IFFO’s 1.68:1 ratio seems off. It also doesn’t agree with some other information IFFO has published showing a FIFO similar to the one you calculate here. Perhaps they are subtracting the byproduct (22% of the feed) from the total amount of fish used, which looks like what they are doing in this chart:

      IFFO FIFO calculation

      The math seems to have been puzzling to the feed companies too. EWOS has figured out a different calculation, which basically compares the nutritional value of the fish used in salmon feed with the nutritional value of salmon. This more accurately tracks how fish oil, the precious part, is used and seems to show that because of the way salmon convert feed and grow, they actually create more nutritional value for humans than what the salmon consume. Take a look. MPDR is Marine Protein Dependency Ratio and MODR is Marine Oil Dependency Ratio:

      EWOS nutrient ratios

      You can see the whole document in context with the formulas here, it’s on page 9.

  2. So how does iffo “actually” calculate the fish-in/fish-out ratio. In 6 pages of math they carefully add into the mix just the right amount of production of shrimp (using the the left over meal produced from the fish caught to make the oil for the salmon aquaculture) to obtain a reasonable sounding FIFO of 1.68:1. Had they used a species that uses even less meal, say tilapia, they might have achieved an even better value, probably around 0.5:1.
    If the FIFO ratio really is 1.68:1, then for a world production of 1.9 Mt of salmonids (including trout) the amount of forage fish used would be 3.192 Mt. Yet the salmonid farming industry, we are told by UN and even iffo, uses about 11.5 Mt of forage fish.
    What are the salmon farms doing with over 8 Mt of forage fish if they’re not feeding it to their salmon?

    1. We’re not sure where you are getting your numbers. Could you provide a link?

      According to the latest United Nations Food and Agriculture Organization report on fisheries and aquaculture, in 2009 (the latest year for which complete data is available) 17.9 million tonnes of fish were caught in the global reduction fishery (fish caught to render into fishmeal and oil). That created 5.74 million tonnes of fishmeal and 1.07 million tonnes of fish oil. That was used in a wide variety of applications from poultry feed to fish feed. Salmon farming used 13.7% of the total fishmeal produced and 36.6% of the total fish oil produced.

      That would mean salmon farms used 786,380 tonnes of fish meal and 391,620 tonnes of fish oil in 2009.

      The important thing to remember here is that the fish meal and oil come from the SAME FISH. You will get more meal than oil from a fish (roughly 1.73 times more) but salmon feed uses less oil than meal (about 25% of the feed is fishmeal and about 15% is fish oil, 1.67 times more meal than oil), so it almost balances out.

      Another important thing to remember is that wild fish aren’t the only sources of fishmeal and oil used by aquaculture. According to the FAO report, 6 million tonnes of trimmings (byproduct from other fisheries) were used to create fishmeal and oil.

      Another thing you need to take into account is the FCR, or Feed Conversion Ratio, which is a measure of how many kilos of feed (which includes marine ingredients and plant ingredients) it takes to grow one kilo of fish. This number factors into the calculations and varies from farm to farm because of ocean conditions, fish health, etc. On average in B.C. though, it’s about 1.2, rarely more than 1.3. Some farms have achieved 1.1 and lower.

      Apologies if it’s a bit confusing, the math can get complicated because of all the local variables that need to be accounted for. But what all this means is that once you account for the fact that meal and oil come from the same fish, account for the FCR, and account for the actual percentages of marine ingredients in aquaculture feed, is that it works out to a ratio of fish in (wild fish) to fish out (farmed salmon produced) of less than 2:1. The FIFO number you quote is 1.68:1 which is probably a global average.

      Here in B.C. our number is much lower. In fact, the most recent numbers comparing the amount of wild fish used in feed to the amount of farmed salmon produced is about 1.24. That means for every kilo of farmed salmon (live weight) produced, it took 1.24 kilos of wild fish (live weight).

      That’s an incredibly efficient use of resources.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s