Fish feed formulation is arithmetic, not a recipe. You pick a protein target for the species and the stage, work out what share of each ingredient gets you there, then multiply it back to check that it does. Uganda's own aquaculture manual reproduces thirteen published formulations for African catfish alone, so copying somebody else's list of proportions is the one method that cannot be relied on.
Formulating Fish Feed to a Specification, Not Copying a Recipe
Look at what those thirteen catfish diets have in common, which is almost nothing. Fish meal runs from 2 percent of the mix to 55 percent. One of them is built from exactly two ingredients, a shrimp meal at 34 percent and bran at 66 percent, and still reaches a stated 40 percent crude protein. Stated protein across the set runs from about 34 to 55 percent depending on the stage the diet is for. They were formulated in different places with different things on sale, and every one of them is a valid answer to the same question.
That is the whole argument for learning the calculation. A recipe is the output of somebody else's ingredient market on a particular day. What travels is the method: a target, the composition of what you can actually buy, a share for each, and a check.
Before any of that, the target. What tilapia, African catfish and mirror carp need from a diet is laid out on the fish feed requirements page, together with how to tell whether a bought bag delivers it. Ugandan guidance puts the tilapia requirement at 28 to 30 percent crude protein depending on pond fertility and African catfish at 32 percent.
What Protein Level a Ugandan Fish Feed Has to Reach
The ministry's manual states a tilapia protein figure in three places and the three look different until you see what each is doing. The requirement table gives 28 to 30 percent. The feeding charts pair a 36 percent feed with small fish and step down to 32 percent above roughly 80 g. The formulation appendix lists four complete tilapia diets that land between 29.9 and 31.0 percent.
These answer three different questions. The 28 to 30 is what the fish needs, in a pond whose plankton is supplying part of the diet. The 36 and 32 are grades a manufactured feed is sold at, pitched high so that a small ration still delivers what the fish needs. The 29.9 to 31.0 is where a complete diet actually lands once somebody has formulated one. Aim a farm mix at the top of the requirement rather than the bottom, because a pond's natural productivity is the thing you have least control over.
Protein is also not a single nutrient. Fish need a particular balance of amino acids, and plant ingredients are commonly short of lysine, methionine and tryptophan, which is why a diet that swaps out all its animal protein needs those added back synthetically. FAO's own guidance is to check arginine, lysine, methionine and tryptophan first, because if those four are covered the other six are most likely above requirement too. With an unconventional protein source, check all ten.
Ingredients in a Farm Made Fish Feed and What Each One Does
Every ingredient in a mix should be there for one stated job, and it should be the cheapest thing that does that job.
Protein. Fishmeal is the best all round source of protein and of the amino acids fish need, and it is the expensive line. It can be partly replaced by animal by products, oilseed meals and cakes, legumes, cereal by products and aquatic plants, all of which fall short on one amino acid or another, which is why partly is the operative word.
Energy. Cereals and grains carry the energy. Fish digest them badly, because plant material comes wrapped in thick cell walls, complex carbohydrate molecules and fibre, so grain destined for a fish feed is cooked to make it palatable and digestible. Skipping that step turns an energy ingredient into a filler that fouls the pond.
Fat. Oil seed cakes are the usual on farm source. Fish oil works and is expensive. At high fat inclusion the problem becomes oxidation, which turns the fat rancid and the feed useless.
Additives. Optional, small, and added at the maker's stated rate. Mineral and vitamin premixes come from veterinary pharmaceutical outlets and water stable formulations exist for aquatic feeds. Binders hold the pellet together. Mycotoxin binders and mould inhibitors sit under anti fungal agents, with preservatives, antioxidants and attractants beside them. Medicines on prescription can be dispensed through feed, and everything that follows from that, including the point at which fish can no longer be sold, belongs to a vet and to the product label rather than to a page like this one.
Using Pearson's Square to Mix Fish Feed to a Protein Target
The square balances two ingredients against one protein target, and Ugandan extension guidance carries a worked example. A farmer wants 100 kg of 32 percent crude protein feed from soya bean meal at 45 percent and ground maize at 10 percent.
Step three is where the ministry's written instructions stop short. They say to subtract the protein percentage from the figure in the centre, and they never say that each answer crosses to the other corner. The diagram beside the text is drawn correctly, so the published answer is right, but a reader working from the prose alone will assign the two numbers the wrong way round. FAO's own formulation chapter states the crosswise step outright, which is how you can tell this is a gap in the wording rather than a matter of opinion.
Checking the Mix Back Before You Grind Any Feed
This is the habit that makes the whole method safe, and it takes half a minute.
Run the check on any formulation you are handed, including one from a book. Two Tier 1 sources used for this page each carry a worked example whose own numbers contradict its stated figure: FAO's chapter announces a 25 percent carp diet and then computes and describes a 27 percent one, with every figure in the example reconciling to 27. The ministry's feed conversion box has a defect of the same shape. Neither source is unreliable, and both show why you re derive an example instead of copying its conclusion.
Formulating With More Than Two Feed Ingredients
Most real mixes use four or five. The square still works, through a grouping step: sort the ingredients into basal feeds, meaning under 20 percent crude protein, and protein supplements, meaning over 20 percent. Average the protein within each group, put the two averages into the square against your target, and that gives you the share of the mix belonging to each group. Split each group's share among its own members, then check the whole thing back the same way.
Two limits are worth knowing before you start. The square cannot solve protein and energy at the same time; balance protein first, then adjust energy with a high energy ingredient, which is cheaper than protein anyway. And reserve room in the hundred for the premix before you allocate anything else, since a formulation that already adds to 100 has nowhere to put it.
Choosing between two ingredients means comparing them on cost per unit of protein rather than cost per kilogramme. A cheaper sack with half the protein is the more expensive protein, and that comparison has to be redone every time prices move.
Why the Same Recipe Gives a Different Feed Every Batch
Here is the reason a recipe is a weak thing to own. East African fishmeal is mostly sun dried mukene, and its crude protein moves between about 40 and 60 percent with season, fishing location and handling. FAO supplies the mechanism: lipid in the fish rises before spawning and falls after, and as it moves the protein, ash and carbohydrate shares move with it.
Now apply that to the farm mix Ugandan extension material itself prints, which is 60 percent maize bran, 15 percent fish meal, 15 percent sunflower cake and 10 percent cassava flour. Fish meal is 15 percent of the weight, so at 40 percent protein it contributes 6.0 points to the finished feed and at 60 percent it contributes 9.0. That is a three percentage point swing in the finished feed from one ingredient, with the recipe unchanged. On a target near 30, three points is the difference between a feed that works and one that quietly does not.
Brans move too. A Kenyan rice bran that used to run around 10 percent crude protein fell to 3 to 6 percent after processing moved to other operators, with milled husk being blended in; wheat bran from industrial processors holds a steadier 14 to 17 percent by comparison. The practical rule falls straight out: formulate from the figure on the sack or from an analysis of the batch you have, not from a table, and run the square again whenever your supplier changes. What Ugandan maize bran, sunflower cake and cassava flour actually contain is a question this page cannot answer with a verified local figure, which is another reason the number on your own sack is the one that counts. The maize bran page covers that ingredient in its own right.
Antinutritional Factors and the Processing a Farm Feed Needs
Plant ingredients carry compounds that block digestion or bind nutrients: alkaloids, phenols, tannins, saponins and lignin, plus protease inhibitors, glucosinolates, gossypol, phytoestrogens and lectins. Their load rises as you raise the inclusion of the ingredient carrying them, so a mix that leans hard on one leaf meal or one oilcake accumulates the problem along with the protein.
Heat is doing most of the work in that list. Grinding and extrusion both degrade many antinutritional factors, which is part of why a steam extruded floating pellet converts better than a sinking one: the process breaks down starch and antinutritional factors together.
Inclusion Ceilings in Fish Feed and What Happens Above Them
Substitution has a ceiling, and the research on sub Saharan aquaculture feeds puts numbers on it. Above roughly 50 percent dietary inclusion, several alternative ingredients start to depress growth, feed utilisation and fish health rather than just sitting there. Microbial protein above 30 percent inclusion can slow growth. In trials, African catfish fed above 50 percent moringa leaf meal as a fish meal replacement showed altered liver and intestine tissue, and tilapia the same above 60 percent moringa protein.
The economics sit on the other side of the same line. Replacing fishmeal protein with plant or microbial protein cut feed costs by 15 to 30 percent where inclusion stayed below 40 percent. So the saving is real and it is available inside the safe band. Pushing past the band to save more is where the saving turns into a slower cycle.
Mixing, Pelleting and Drying Fish Feed on the Farm
Weigh out the measured quantities. Add any premix at the manufacturer's rate. Mix until the whole batch is uniform, with a spade for a small quantity or a mixer for a larger one, because an unevenly mixed batch means some fish get the premix and some get none.
From there the mixture goes one of two ways. Cooked and fed as a dough, which is fed immediately, or pelleted and dried before it goes into store. Wet feed is never stored. Binders are what hold a pellet together long enough to be eaten, and the specification they exist to meet is the one a good manufactured particle hits: uniform texture and size, no fines, and water stability of at least thirty minutes. Wheat gluten, cassava flour and wheat flour are on the ministry's own binder list, which is convenient, since cassava flour is already in most Ugandan farm mixes as an energy ingredient.
Then the paperwork, which is short and pays for itself. Give every batch you make a batch number and record what went into it, so that if something goes wrong with a pond you can trace it to a mix rather than guess. Use the oldest batch first. If you sell pellets, they have to be cooled, properly dried, packed in clean bags and labelled with the batch number. Storage after that is its own subject, covered on the feed storage page.
What the Ministry's Own Fish Feed Formulation Tables Show
The appendix of Uganda's aquaculture manual reproduces a set of FAO tilapia diets by life stage, and it repays reading closely, partly because of what changes across the columns and partly because of one thing the table gets wrong.
| Ingredient | Early fry | Fingerling | Pond grower |
|---|---|---|---|
| Cassava starch | 15% | 0 | 0 |
| Cassava meal | 0 | 23% | 22% |
| Coconut meal | 0 | 0 | 30% |
| Rice bran | 30% | 15% | 0 |
| Soybean meal | 0 | 30% | 25% |
| Fish meal | 47% | 25% | 20% |
| Fish oil | 5% | 4% | 0 |
| Dicalcium phosphate | 1% | 1% | 1% |
| Premix | 2% | 2% | 2% |
A fourth column for a cage grower sits between the last two and is close to the fingerling diet, at 20 percent rice bran and 25 percent soybean meal. Every column adds to exactly 100, which is the first thing to check on any published formulation.
The pattern across the table is the interesting part. Fish meal falls from 47 percent to 20 percent as the fish grows, and the fish oil disappears entirely in the pond diet, while the protein the diet delivers barely moves.
| Measure | Early fry | Pond grower |
|---|---|---|
| Crude protein | 30.0% | 29.9% |
| Crude lipid | 10.0% | 4.1% |
| Crude fibre | 3.8% | 6.0% |
| Gross energy | 2,800 kcal | 2,500 kcal |
Same protein, half the fat, more fibre, less energy, and a quarter of the fish meal. That is what a pond does for you: the plankton the pond is producing covers the difference, which is also why the requirement table gives tilapia protein as a range that depends on pond fertility.
Now the defect. In the published table the first proximate row is labelled Dry matter and carries 8.3, 9.0, 9.0 and 9.1 across the four diets. Those are moisture figures, and the proof is in the table itself: in each column that row plus crude protein, lipid, ash, fibre and nitrogen free extract adds to exactly 100.0. Read as printed, the early fry diet would be 8.3 percent dry matter, which is to say over 91 percent water, which no dry feed is. One wrong heading in a table whose every column otherwise reconciles, and the way to catch it is addition.
Where a Farm Made Fish Feed Should Not Be Used
Making your own has real advantages: ordinary farm implements will do, local by products get used, and dry supplementary feeds keep longer than wet ones. The costs are also stated plainly in the ministry's own comparison. A farm made feed carries a higher risk of fouling the water and a higher feed conversion ratio, meaning more kilogrammes of feed per kilogramme of fish. The ministry's own comparison puts it at about 5 kg of a farm mix per kilogramme of fish against about 1.8 kg of a complete pellet, with maize bran fed alone needing about 9 kg. Whether that trade pays depends on the price gap between the options on the day, which is arithmetic you have to run yourself.
And there is one flat exclusion. Supplementary and farm made feeds are not recommended for tanks, cages or intensive culture systems. Those systems have no natural food to fall back on and no capacity to absorb the waste, so a feed that is approximately right fails quickly. If you are running catfish in tanks or tilapia in cages, a complete manufactured diet is the only workable choice.
It is worth knowing why so many Ugandan farmers reach for the mixer in the first place. Uganda's twelve established feed mills cannot meet demand, with a reported maximum of about five tonnes a week each, and the local product has been described in the literature as carrying rough pellets, an imbalanced nutrient profile and poor digestibility, coming from unreliable electricity, a lack of modern processing and packaging, poor transport and storage, outdated formulation methods, formulations that are not species specific, and cheap ingredients used to hold the price down. A word of caution on the national numbers, though: two peer reviewed reviews of this same sector give local output figures that cannot both be true, one putting factory production at around 75,000 tonnes a year and the other implying something on the order of 3,000. Neither is a safe basis for a plan.
Where to Check a Fish Feed Formulation Before You Mix
Before a first batch, get the protein figure for each ingredient off the sack or from a laboratory, run the square, check it back, and take the result to your district fisheries officer to look over. Ingredient prices and what is available locally move through the year, so compare on cost per unit of protein each time rather than reusing last season's decision; the supplier directory shows what is being stocked near you, the feed requirement calculator turns a formulation into a quantity, and the feeding guide covers what to do with it once it is mixed. Anything you add that is a medicine follows the label and a vet, not a formulation table.
Fish Feed Formulation Questions From Ugandan Farms
Is making my own fish feed cheaper than buying pellets? Sometimes, and the sum is not the one most people run. Manufactured feed has been reported at roughly 3,500 to 3,800 shillings a kilogramme for sinking pellet and about 5,000 for imported floating feed. Against that, the ministry's comparison needs about 1.8 kg of complete pellet per kilogramme of fish and about 5 kg of a farm mix, so the farm mix has to cost under roughly a third per kilogramme before it wins on the fish. Work it out at the prices in front of you; the cost of fish farming page sets out the rest of the budget.
What protein should I formulate to? Around 30 percent for tilapia in a fertile pond, 32 percent for African catfish, and higher for fry of either species. Aim at the top of the requirement rather than the bottom, because a farm mix has more variability in it than a manufactured feed and you want the margin on the safe side.
Can I just use maize bran on its own? It is what many Ugandan farms do, and it costs about nine kilogrammes of bran per kilogramme of fish against under two for a complete pellet. Bran alone is an energy ingredient with a little protein attached, so the fish are growing on the pond's plankton with a supplement, rather than on a feed.
How do I know my mix actually hit the protein I aimed for? Multiply each ingredient's share of the mix by its own crude protein, add the results, and compare with the target. If you never do that step, you will not find out for a couple of months, and by then the feed conversion ratio has already told you in the most expensive way available.
Do I need a vitamin and mineral premix? Yes, and leave space for it in the hundred before you allocate the other ingredients. Premixes come from veterinary pharmaceutical outlets and go in at the rate the maker states. Water stable formulations exist for aquatic feeds, which matters, since a premix that leaches out of the pellet before the fish eats it has done nothing.
Why do my pellets fall apart in the water? Binder, grind and drying, in that order. A pellet has to hold together for at least thirty minutes to count as stable, and that needs a binder such as cassava flour, ingredients ground fine enough to knit, and proper drying before storage.
Can I replace all the fishmeal with plant ingredients? Not all of it. Plant proteins are short of lysine, methionine and tryptophan, and several alternative ingredients start depressing growth and fish health above roughly half the diet. Staying under about 40 percent replacement is where the cost saving has been measured without a performance penalty.