Duckweed Animal Feed: Species, Limits and EU Rules
Duckweed as animal feed suits ruminants best, hits a physical ceiling in poultry and pigs, and in the EU its legal status turns on the water it grew on.
Most pages about duckweed animal feed open with one number: crude protein, somewhere around 28 percent of dry matter. It is a real number. It is also the least useful one in the dataset, because the samples behind it run from 15.5 percent to 35.6 percent, and a formulator who buys the average gets whichever end of that range happened to arrive.
So this article is organised the other way round. First what the spread is and why it exists, then what each class of animal does with it, then the question a European buyer meets before any of that matters: what water did it grow on.
The average is the least useful number
The reference table for dried duckweed pools 18 samples for crude protein and 17 for ash. Crude protein averages 27.8 percent of dry matter with a range of 15.5 to 35.6, and ash averages 18.9 percent with a range of 3.8 to 35.6. Neutral detergent fibre sits at 40.3 percent, crude fibre at 13.1.
Read the ash column again. A tenfold spread in the mineral fraction is not measurement noise. It is the plant recording what was dissolved in its water.
The cleanest demonstration comes from a study that fed three real batches side by side to Ross 308 broilers. A mix of Lemna minuta and Lemna minor came in at 17.5 percent crude protein, Spirodela polyrhiza at 24.6 percent, Lemna obscura at 37.0 percent. Tannins were 2943, 2890 and 303 mg per kg respectively. Standardised ileal digestibility of methionine ran from 49.9 percent in the first batch to 90.4 percent in the third. Phosphorus digestibility ran from 50.8 to 78.9 percent.
Those are three different ingredients wearing one common name. The authors’ conclusion is the operative one for anybody buying: a defined and stable biomass composition, matched to the animal, is the precondition for using duckweed at all. That is a cultivation problem before it is a nutrition problem, which is why duckweed cultivation and feed formulation are not separable questions.
Ruminants get the most out of it
If you rank the animal classes by how much duckweed they will take, ruminants win, and it is not close.
Recent work fed Lemna gibba to fattening lambs, replacing 50, 75 and 100 percent of the alfalfa hay in a diet where alfalfa made up 40 percent. Up to 40 percent of the diet, meaning alfalfa replaced entirely, did not harm digestibility or fermentation and in the balanced rations improved them. Dry matter and fibre digestibility were highest in the isonitrogenous and isocaloric rations containing 40 percent duckweed. The same paper puts duckweed dry matter production at 10 to 30 tonnes per hectare per year against 8 to 12 for alfalfa.
That is the optimistic end, and it is in vitro. A more conservative reading comes from gas production work on a total mixed ration at a 70:30 roughage to concentrate ratio, which concluded that 10 percent replacement of the concentrate mixture was the level with no adverse effect and found metabolisable energy falling significantly as inclusion rose, down to 5.15 MJ/kg at 30 percent duckweed.
Both results can be true. One replaces forage, the other replaces concentrate, and duckweed is much better at the first job than the second. Its pooled metabolisable energy for ruminants is 11.7 MJ/kg of dry matter, respectable as a forage substitute and thin as a concentrate. A duckweed cattle feed decision turns on which of those two things you are actually asking it to do.
Poultry and pigs hit a physical ceiling before a nutritional one
Move to monogastrics and the ranking inverts, for a reason that has nothing to do with amino acids.
Duckweed holds water. In the three-batch broiler study, water binding capacity of the control diet was 119.6 percent; at the highest duckweed share it reached 364.9 percent. Average daily feed intake fell from 154.2 grams per day on the control to 74.9 grams per day. The bird stops eating before the protein content becomes relevant.
The trial record follows from that. A University of Zimbabwe finisher trial with 160 male broilers at 0, 10, 20 and 30 percent duckweed found intake, weight gain and feed conversion all declining with inclusion, and 10 percent was the level that matched the control. Energy explains the rest: apparent metabolisable energy for poultry is 4.9 MJ/kg of dry matter, against 11.7 for ruminants. The full picture for birds is in the separate piece on duckweed as chicken feed.
Pigs are the interesting middle case
Pigs sit between the two, and the best evidence is a 120-day trial with 24 fattening pigs at Cantho University. Ensiled cassava root with fresh duckweed supplying about 15 percent of total protein produced no significant difference in daily gain or feed conversion against the conventional rice by-product control, and significantly thinner backfat at 2.09 against 2.95 cm. The same paper cites earlier work in which fresh duckweed supplied 45 percent of total diet dry matter at 83 percent dry matter digestibility, where the duckweed protein content was over 30 percent.
Two conditions carried those results. The duckweed was fed fresh rather than as a dried meal, and it came from a high-protein batch. The same discussion notes the counter-example: a low-protein high-fibre duckweed meal at only 10 percent of the diet cut pig growth by 30 percent over 40 days. Pooled nitrogen digestibility for the growing pig is 64.0 percent, which is the number to plan around when the batch is unknown.
Fish are the fourth class and behave differently again, which is why duckweed in fish feed is treated separately.
Methionine decides the premix
Every animal class converges on the same gap. In the pooled amino acid panel, methionine is 1.8 percent of protein, against lysine at 4.0 and leucine at 7.5. The Belgian pilot below ran a full essential amino acid panel and flagged histidine and methionine as potentially limiting for pigs and cattle while rating the total profile suitable for broilers.
Assume the premix carries the methionine. And read any duckweed amino acid profile you are quoted next to the digestibility figure for that same harvest, because a high total at 49.9 percent digestibility is a worse ingredient than a lower total at 90.4 percent.
In Europe, the water decides whether it is feed at all
Here is where the smallholder guides stop and a European feed buyer’s problem starts.
Duckweed grows fastest on nutrient-rich water, and the cheapest nutrient-rich water on a farm is diluted manure. That is exactly what the best outdoor feed-crop evidence used. A pilot at a pig farm in Pittem, Flanders, grew Lemna x japonica clone 6580 on the diluted liquid fraction of pig manure plus nitrification and denitrification effluent, producing 8 tonnes of dry biomass and 2.8 tonnes of protein per hectare per year, and then assessed the harvest as a feed ingredient.
Now put that next to the law. Annex III of Regulation (EC) No 767/2009 lists among prohibited materials faeces, urine and separated digestive tract content, irrespective of any form of treatment or admixture. That provision names the manure, not a plant grown in water derived from it, and those are genuinely different things. But the distance between them is a judgement an auditor makes, not a settled classification you can point at, and the burden of evidence falls on whoever placed the material on the market.
Two more rules bear on the same batch. Directive 2002/32/EC sets maximum limits for arsenic, lead, mercury and cadmium and prohibits diluting a contaminated feed material to bring it under a limit, which matters because duckweed concentrates what is dissolved around it. And Regulation (EU) No 68/2013 makes the Catalogue of feed materials voluntary to use while reserving its names, so a description has to match the entry it claims and the operator has to satisfy Article 4 of 767/2009 regardless.
None of this makes duckweed an illegal feed. It makes provenance the first line of the specification rather than the last.
Why SERAPH separates the streams before there is a harvest
SERAPH runs two permanently firewalled streams. Production-grade biomass for food, feed and high-value products is one. Non-food remediation biomass is the other, and it never crosses into either. The separation starts at the first sensor reading rather than at the point of sale.
That is a company policy, not a regulation, and it is deliberately stricter than what the published research permits. The reason is the paragraph above. Containment and traceability are cheap to design into a system and impossible to retrofit into a harvest that has already been mixed, and the moment a buyer has to prove which water a batch grew on, an answer that was never recorded is the same as no answer.
Where SERAPH is in that work is stated plainly: TRL 3, proof of concept, with outdoor field validation still ahead. Behind it sits a knowledge base of 256 indexed duckweed papers mapped across 31 research areas in 5 domains, which is how thin coverage in this literature becomes visible instead of assumed.
The measurement layer is further along. SERAPH Vision reconstructs a cultivation canopy in 3D from a single capture, using 170,000 surface points drawn from 5.9 million captured, computes canopy volume, and already tells duckweed species apart on a real tray, frame by frame. Growth and stress reading are still in development. Given that the species is the single variable most responsible for the composition spread this article opened with, knowing what is actually growing is not an accessory to the feed question. It is the measurement the whole thing rests on.
The specification to send a supplier
Five asks, in order of how much they change the answer.
Species, not common name. Lemna obscura and Lemna minuta are both duckweed and they are not the same purchase.
Crude protein and ash on that harvest, from that batch. A figure quoted from a review tells you what the genus can do.
The growing medium, in writing, with documentation. Treat it as a compliance question first and an agronomic one second.
A contaminant panel against the Directive 2002/32 limits, because the plant’s efficiency at uptake is the same property that makes it useful in remediation.
Season length. The Flanders pilot reported growth ceasing after roughly 100 to 120 days as pH and electrical conductivity rose in the medium. A ration you formulated in July needs an ingredient in November, so the harvesting equipment and the season have to be planned as one thing.
Get those five and the inclusion rate mostly answers itself. Skip them and you are formulating against an average, which is the one number in the whole dataset that describes nothing you can actually buy. That is the same trap that catches most duckweed protein comparisons.
FAQ
What animals can eat duckweed?
Ruminants, pigs, poultry and fish all appear in the published feeding record. The order matters more than the list: ruminants take the highest share of the diet, pigs sit in the middle and do best on fresh high-protein material, and poultry are limited by intake rather than by nutrition.
Is duckweed a legal animal feed in the EU?
Plant biomass as such is not prohibited. Manure and its derivatives are, contaminant limits apply to the finished material, and the operator carries the burden of showing the batch is safe and correctly described. In practice the first question is what water it grew on, and the answer has to be documented rather than asserted.
Is duckweed cheaper than soybean meal?
It has not been demonstrated as a like-for-like replacement anywhere in the record. The case for it is local supply and nutrient recovery, plus a yield per hectare that compares well with alfalfa, not a lower cost per unit of digestible protein.
Can you feed duckweed fresh, or does it have to be dried?
Both appear. The strongest pig results used fresh duckweed, and one trial ran it at 45 percent of diet dry matter that way. The poultry evidence is largely on dried meal of known origin. Fresh removes a drying cost and adds a shelf-life and hygiene constraint.
Does SERAPH supply duckweed biomass for feed?
No. Remediation-grade biomass never enters food or feed here, and the two streams are separated from the first sensor reading. SERAPH supplies the cultivation system and the know-how rather than biomass, and it is at proof of concept stage with outdoor validation still ahead.