How we prove a fish collagen lot is fish collagen
A fish collagen lot has to clear two separate identity questions: is the powder collagen, and is that collagen from fish. Hydroxyproline and the amino acid profile answer the first; only a species test or a documented species chain answers the second. The molecular weight distribution answers a third: whether the collagen has actually been cut into peptides or is still gelatin waiting to set.
Hydroxyproline proves collagen and only a species test proves fish, so a fish collagen file needs both.
The four layers of identity
Identity is not one test. A certificate that answers one layer and implies the others is where most identity failures begin.
| Layer | Question | Evidence | What it cannot show |
|---|---|---|---|
| Composition | Is the material collagen? | Hydroxyproline; amino acid profile rich in glycine, proline and hydroxyproline | Which species or tissue |
| Size | Is it hydrolysed? | SEC distribution; fraction below 3,000 Da | Species; whether the collagen is intact |
| Species | Is it fish? | Species-specific PCR on raw tissue, or a validated method on the powder; documented species chain | Which tissue; whether the tissue tested is the material in the drum |
| Chain | Which fish, which plant? | Landing or harvest record, process record, lot traceability | Analytical proof on its own |
The layers are cumulative. A lot with the right hydroxyproline and the wrong species passes the first test and fails the product; a lot with the right species and a gelatin curve passes the species test and fails the process.
Hydroxyproline against the protein figure
Hydroxyproline is the amino acid that makes collagen recognisable. It is formed by hydroxylation of proline after the protein chain is built, and it is present at only trace levels in most non-collagen proteins. A specification of 8.0–12.0% hydroxyproline by HPLC, read against a protein figure of ≥90.0%, describes a material whose hydroxyproline is roughly 8.9–13.3% of the protein mass. That ratio is the first identity check, and it costs nothing extra because both numbers are already on the certificate.
The ratio catches two failures. The first is a powder that meets the protein figure with a protein that is not collagen: soy, whey, pea or a hydrolysed cereal will lift the nitrogen without lifting the hydroxyproline. The second is nitrogen spiking: urea, ammonium salts or another nitrogen source can push Kjeldahl upward while the hydroxyproline stays where the collagen put it. Neither failure is visible from the protein line alone.
The ratio does not identify the species. Bovine, porcine and fish collagen all contain hydroxyproline, and their windows overlap; fish material often sits toward the lower half of the window, but a fish lot at 12% is not thereby mislabelled. That is why the species layer exists and why the hydroxyproline number cannot replace it.
HPLC and colorimetric methods do not return identical hydroxyproline values, so the method belongs beside the number.
The SEC peptide profile
Size-exclusion chromatography separates the peptides by hydrodynamic volume and reports a distribution, not a single number. The specification requires ≥90% of the mass below 3,000 Da, and a useful certificate reports the measured fractions: below 500 Da, 500–1,000 Da, 1,000–3,000 Da and above 3,000 Da. The average molecular weight is a derived figure and hides the shape of the curve; two powders with the same average can behave differently in a drink.
The distribution is also the fastest check on gelatin. Gelatin is collagen that has been cooked but not hydrolysed to peptides, and it carries a large fraction above 3,000 Da. A powder blended with gelatin to lift the protein figure will show a high-molecular-weight shoulder that a properly hydrolysed lot does not have, and it will also gel a solution on cooling. Over-hydrolysis shows the opposite pattern: a large fraction below 500 Da, more free amino acids, and a taste that turns bitter or broth-like.
SEC is a relative method: the result depends on the column, the calibration standards and the detection wavelength. The certificate should name the method, and a cut-off quoted without it is not comparable between suppliers.
What diffraction and infrared do not add
X-ray diffraction and infrared spectroscopy are the standard identity tools for a crystalline ingredient, where they separate polymorphs and hydrates. A spray-dried collagen peptide is amorphous, so diffraction returns a broad halo and adds nothing to the species question. Infrared shows the amide bands of a protein, which is consistent with collagen but not specific to fish. Neither method belongs in a fish collagen file as proof of source; they belong to the mineral salts and the crystalline amino acids.
The adulteration routes, and the test that finds each
Collagen carries a defined amino acid signature, and a powder mixed into it usually leaves a trace. The routes below are the ones a specification and a testing plan should anticipate, because each leaves a different fingerprint.
| Route | What it changes | Test that finds it |
|---|---|---|
| Bovine or porcine collagen | Species, not the collagen chemistry | Species-specific PCR; declared species chain |
| Gelatin | Molecular weight curve; gelling behaviour | SEC distribution above 3,000 Da; gel test |
| Other protein | Protein figure up, hydroxyproline down | Hydroxyproline-to-protein ratio; amino acid profile |
| Nitrogen spiking | Kjeldahl up with no peptide mass | Ratio; total nitrogen against amino acid nitrogen |
| Cheaper fish species | Species claim, not identity as collagen | Species-specific PCR; traceability record |
No single test covers the table. The combination of hydroxyproline, the amino acid profile, SEC and a species method is what makes a lot defensible, and the panel is small against the risk it covers.
Writing the specification
- “Fish collagen hydrolysate; species, tissue and country of origin named.”
- “Protein ≥90.0% dry basis, Kjeldahl with N × 5.55.”
- “Hydroxyproline 8.0–12.0% by HPLC.”
- “Molecular weight: ≥90% below 3,000 Da by SEC; fractions reported.”
- “Species identity supported by a species declaration and, on request, a PCR result.”
- “Allergen: fish declared; CAS 9007-34-5 is an identifier, not a species statement.”
The CAS line matters. 9007-34-5 is the generic collagen number, identical for fish skin, fish scale and bovine hide; a specification that leans on it has said nothing about the species.
Five questions for a supplier
- Which species and which tissue is this lot made from, and is the species on the COA?
- Which nitrogen factor and which basis were used for the protein figure, and by which method was hydroxyproline measured?
- Can you send the full SEC distribution with the method and the fraction bins, not only the cut-off?
- What evidence supports the species claim: a PCR result, a raw-material species declaration, or both?
- What is the fish allergen statement, and does the lot number on the drum match the COA and the retained sample?
The answers belong in the file next to the certificate. The COA page explains how the numbers are read, the grade comparison explains what changes with the tissue, and the compliance page sets out the allergen and market requirements. The HPLC method note describes the chromatographic side of the work.
Related documents
Send the specification you intend to use — grade, quantity and destination market — and we will reply with the batch COA, the technical data sheet and a quotation.
Request a quotation