How to Read Feed Protein Spec Sheets and COAs

Bottom line

Before comparing protein numbers, check that they use the same basis. You also need the same test method for a fair comparison. Ask for a certificate of analysis, or COA, for your lot.

Check whether protein includes water or uses dry matter, called DM. Published averages can guide you when comparing different feed materials. They cannot show that your lot meets the agreed specification.

What do the different documents tell you?

A specification sheet states the product requirements agreed before delivery. A COA gives test results for the named sample or lot.

An independent inspector checks only the work agreed in your contract. This may include sampling, document checks or selected laboratory tests.

Read these documents together and ask who collected the tested sample. You also need to know how that sample represents your lot.

Is the protein result DM or as-is?

Dry matter excludes water, while as-is includes the material’s water content. The term “as fed” also means that moisture is included.

The INRAE-CIRAD-AFZ fish meal table gives 92.2% dry matter as fed. Its protein 70% type contains 74.8% crude protein on a DM basis.

The conversion is 74.8 x 0.922 = about 69% as fed.

The table itself also gives 69.0% crude protein as fed. These are published averages. They are not results for your purchased lot.

For your conversion, use moisture and protein results from the same material.

What should you read beside crude protein?

Crude protein is a protein estimate calculated from measured nitrogen. It does not tell you how much protein animals can use.

Fat results depend on the method used to extract the fat. Ash is the mineral material remaining after the sample is burned.

Compare these results using the same units and treatment of moisture.

Meat and bone meal

For meat and bone meal, Kerr et al. reported these results from their study.

Reference	Crude protein, % DM
Study of 97 samples	Average 56.9; range 48.0 to 64.8

There is no public reference for fat and ash. Ask for the method and the agreed limit.

These results show how much protein varied among the tested samples. Your purchase still needs agreed limits and results for its own lot.

Poultry by-product meal

For poultry meal, INRAE-CIRAD-AFZ lists “processed animal proteins, poultry” in different protein classes.

Published class	Crude protein, % DM	Fat, % DM	Ash, % DM
Protein above 70%	82.2	12.1	5.5
Protein 60 to 70%	67.5	23.2	10.7

These table names may describe something different from your product name. Check the product definition before you compare the figures.

Kerr et al. found average crude protein of 63.5% DM in 16 samples. Values ranged from 55.0 to 71.5% crude protein DM.

Do not combine these study results and table classes into one specification.

Hydrolysed feather meal

For hydrolysed feather meal, hydrolysis means processing that breaks down the feather structure. High protein alone does not show that animals can digest it well.

INRAE-CIRAD-AFZ reference	Crude protein, % DM	Fat, % DM	Ash, % DM
Average	85.7	9.2	5.6

The table gives a protein range of 69.7 to 98.1% DM. Kerr et al. reported an average of 92.1% DM across 23 samples.

The two averages differ by about 6 points. So ask the supplier what the product contains.

Blood meal and haemoglobin powder

Blood meal reference values do not set acceptance limits for haemoglobin powder.

INRAE-CIRAD-AFZ reference	Crude protein, % DM	Fat, % DM	Ash, % DM
Blood meal average	93.9	2.0	3.3

Some studies report blood meal crude protein above 100% on a dry matter basis. Kerr et al. reported a value of 100.7% crude protein DM.

The calculation, nitrogen x 6.25, can give a value this high. It does not mean the sample has more protein than its own dry weight.

There is no public reference for haemoglobin powder. Ask for the method and the agreed limit.

Fish meal

For fish meal, check the named type before comparing its published composition.

INRAE-CIRAD-AFZ reference	Crude protein, % DM	Fat, % DM	Ash, % DM
Fish meal, protein 70% type	74.8	9.8	15.2

FAO also gives historical protein figures on an “as received” basis.

Fish meal type, FAO 1986	Crude protein, % as received
Herring type	About 72; usual range 68 to 74
South American anchovy type	About 65
White fish type	About 65

These old figures can guide comparisons between different fish meal types. You still need current results for the lot you are buying.

Total volatile nitrogen, or TVN, measures nitrogen compounds that increase as the raw material spoils. Histamine is another compound that can form as raw material breaks down.

There is no public reference for TVN, histamine, salt and sand. Ask for the method and the agreed limit.

Fish oil

There is no public reference for fish oil limits in feed. Ask for the method and the agreed limit.

Request these results with their units, methods and agreed acceptance limits:

• Free fatty acids: fatty acids released as the oil breaks down.
• Peroxide value: a measure of early products formed during oxidation.
• Anisidine value: a measure of later products formed during oxidation.
• TOTOX: a combined oxidation measure using peroxide and anisidine results.
• Antioxidant: its identity, amount added and records for the purchased lot.

Oxidation is a chemical change involving oxygen that can damage oil. Codex CXS 329-2017, section 1, applies to fish oils for human consumption. It covers fish oil used in food and in food supplements. Its limits are not acceptance limits for your feed oil.

Can a pepsin result show nutritional value?

Pepsin is an enzyme that breaks down protein during digestion. Laboratory tests use it to measure protein digestion under set conditions.

Moritz and Latshaw tested feather meal using 0.2% pepsin. Samples with higher pepsin results had less sulphur amino acid that birds could use. This means birds could use less of those sulphur-containing amino acids. Amino acids are the small parts that make up protein.

Hendriks et al. studied 94 meat and bone meals. Pepsin results did not match how well amino acids were digested in the gut. Ileal measurements assess digestion at the end of the small intestine. Their correlation, r = -0.06, was close to zero, meaning no clear link.

Kawauchi et al. tested meat and bone meal and poultry meal in dogs. Different pepsin strengths gave different results, and prediction was weak.

Eaksuree et al. measured feather meal pepsin digestibility in vitro, meaning laboratory testing. The control without enzyme treatment gave 70% in this pepsin test. The meal treated with a commercial enzyme gave 68%. The same meals were also fed to broiler chicks. The laboratory pepsin results agreed poorly with the live-bird results.

You should therefore avoid using pepsin results alone to judge nutritional value.

Which methods and lot details should you check?

Common test methods

These are methods used in published studies and reference programmes.

Test	Method reference and source
Crude protein	Combustion, also called Dumas; AOAC 990.03; Munoz et al. (2020)
Dry matter (moisture)	AOAC 930.15; Munoz et al. (2020)
Ash	AOAC 942.05; Munoz et al. (2020)
Pepsin digestibility	AOAC 971.09; AOCS fish meal reference programme; confirm pepsin concentration
Fat after acid treatment	Acid-hydrolysed ether extract; AOAC 954.02; Park and Adeola (2022)
Fat	Procedures with or without acid hydrolysis; Regulation (EC) No 152/2009, Annex III, Part H
TVN for fish meal	Regulation (EC) No 152/2009, Annex III, Part E

Acid hydrolysis means acid treatment before the fat is extracted. Check this detail before comparing fat results from different methods.

Before accepting a COA, check these details against your agreed specification:

• Laboratory: check its identity and accreditation scope for the relevant tests.
• Dates: compare production, sampling and analysis dates, and question any inconsistencies.
• Lot number: match it across the COA, packing list and shipping documents.
• Method: confirm the analytical reference and pepsin concentration where relevant.
• Basis and units: check that results and limits use matching bases and units.
• Authorisation: check who authorised the report and ask about missing details.

ILAC explains that accreditation bodies use ISO/IEC 17025 to assess laboratory competence.

Which warning signs need further questions?

Ask for more information when you see these document warning signs:

• Every result repeatedly equals its specification limit across different reported lots.
• Lot identifiers are missing or too general to identify your material.
• You cannot verify the laboratory or its relevant accreditation scope.
• The report omits analytical methods or the pepsin concentration used.
• Analysis dates do not fit the production or sampling dates.
• Protein, fat and ash seem inconsistent with the declared product.
• Only TVN or histamine appears, without a clear testing scope.
• Antioxidant details are missing for fatty meals or fish oils.

You need to investigate these signs before making an acceptance decision. None of these signs alone proves that a shipment fails.

MANTA view

Agree the test methods before you discuss the reported results. Compare unclear results with your specification and ask the laboratory for details.

Further reading: Feedipedia (INRAE, CIRAD, AFZ, FAO), not used as a source of numbers.