In April 1917 UC-30 was heading back to Heligoland. She was a minelaying submarine of the Imperial German Navy, fifty-one metres long. In March she had sailed for Ireland under orders to lay mines, but an engine failure forced her to turn back with the job undone. Off Jutland she struck a mine in a British field and sank. The authors of the study date the sinking to 20 or 21 April. The wreck lies at about twenty-four metres, some ten kilometres off the Danish coast. On board, by their estimate, between 3,120 and 4,812 kilograms of explosive remained.

In the spring of 2022 divers of the Royal Danish Navy went down to the wreck with steel cages full of blue mussels, Mytilus edulis. They tied them to three mine wells in the stern, right above the exposed explosive. Next to each mussel cage they set a second one holding passive samplers, discs that capture the TNT dissolved in the water flowing through them. The cages went down on 30 March and came up on 5 July, fourteen weeks later. A mussel filters water to feed and keeps what it finds there. In the laboratory, with algae that drive it to filter at full speed, a five-centimetre mussel clears almost five litres of water an hour, a seven-centimetre one almost ten. That is the size used in the cages, and the measurements hold between 11 and 15 degrees. In the sea a mussel often filters less. That is why it has served for decades as a sentinel of pollution.

The study appeared in Marine Pollution Bulletin. The first author is Romina Marietta Schuster of the Alfred Wegener Institute in Bremerhaven, and there are eleven authors in all. It has been online since 17 April 2026, and in August a press release from Aarhus University brought it back into the news. The UC-30 mussels were compared with those placed on two other North Sea wrecks. The first is the light cruiser SMS Mainz, sunk on 28 August 1914 in the Battle of Heligoland Bight, at about thirty metres. The second is KW58 Hendericus, a Dutch herring drifter seized by the Kriegsmarine to carry munitions and sunk in Belgian waters on 27 November 1944. Mines and mortar rounds can still be seen on her deck.
In the water, TNT turned up only at UC-30. The highest value was 72.6 nanograms per litre, and a nanogram is a billionth of a gram. Alongside the TNT there were up to 40.7 nanograms per litre of 4-aminodinitrotoluene and 62.7 of 2-aminodinitrotoluene. These two compounds, known as 4-ADNT and 2-ADNT, form when bacteria or light turn one of the molecule’s three nitro groups into an amino group. In the sediment inside the mine wells TNT averaged 3.11 micrograms per kilogram, on the surrounding seabed 0.01. At the Mainz nothing was found, in water or sediment. At KW58 the sediment next to the anti-tank mines on deck held 151.6 micrograms of TNT per kilogram, but the water, sampled fifty metres from the wreck, was clean.

In the mussels the signal was sharper. Of the 59 analysed from UC-30, 56 carried TNT or its derivatives. TNT averaged 1.4 nanograms per gram of dry weight, 4-ADNT 15.6 and 2-ADNT 15.2. Each derivative exceeded its parent compound tenfold. At KW58 the values stayed below one nanogram per gram, at the Mainz they were zero. No mussel died, in any cage.

TNT residues in mussels (sum of mean TNT, 4-ADNT and 2-ADNT, in nanograms per gram of dry weight) and glycogen in the digestive gland (percentage of stained area, mean and standard deviation) at the three wrecks studied. Chart produced by ScienceOnline from data in Schuster et al., Marine Pollution Bulletin 2026, doi 10.1016/j.marpolbul.2026.119738.

The authors looked for damage in the digestive gland, the organ where the mussel breaks down foreign substances. They measured lipofuscin, a pigment that settles in the lysosomes, the cell’s waste sacs, when membranes are oxidised. In the UC-30 mussels it covered 18.34 per cent of the tissue examined, at the Mainz 6.80. Neutral lipids, fats that a stressed cell stores instead of using, rose from 1.68 to 2.72 per cent. Glycogen, the mollusc’s energy reserve, fell from 4.93 per cent at the Mainz to 2.87 at UC-30 and 2.67 at KW58. Catalase, an enzyme that disposes of the hydrogen peroxide produced by oxidative stress, was more active in the Mainz mussels than in those of the other two wrecks. All the differences were statistically significant, with a probability of chance below one in a thousand.

The authors draw a clear picture from this. Mussels exposed to TNT spend their reserves on defence, and the defences slowly give way. They saw it at very low concentrations. Acute toxicity of TNT to marine organisms begins between 0.5 and 10 milligrams per litre. The peak measured in the water at UC-30 is about seven thousand times below the lower threshold.
The study’s weak point is the comparison. The planned reference was Borkum Riffgrund, a nature reserve far from any wreck. The cage placed there was lost. The Mainz stands in as reference instead, because no TNT was found there. But the Mainz is another wreck, in another stretch of sea, at another depth. The KW58 mussels came from the pylon of a Belgian wind farm, those for UC-30 and the Mainz from the island of Sylt. The UC-30 cages were tied to the hull, the others sat on frames lowered to the seabed. Every wreck, the authors themselves write, releases other things too, from the lead and copper of antifouling paints to coal tar and lubricating oils. None of these was measured. With a single wreck per condition, the study shows that the UC-30 mussels were in worse condition than the Mainz mussels, but it cannot show that the cause is TNT and TNT alone. It is a strong association, consistent with the same group’s laboratory experiments, but it is not firm proof.

TNT is toxic to humans too. The authors recall that after the First World War workers in TNT factories fell ill with gastritis, dermatitis, aplastic anaemia and toxic hepatitis. TNT enters mainly through the skin, and the liver turns it into the same 2-ADNT and 4-ADNT found in the mussels. In the North Sea, according to the North Sea Wrecks project led by the German Maritime Museum, there are at least 680 wrecks with a military past, at least 250 of them in Danish waters. In German waters of the North Sea and the Baltic about 1.6 million tonnes of conventional munitions remain, most of it dumped deliberately after 1945. Between 1945 and 2014, contact with munitions in those two seas killed at least 418 people and injured at least 680, according to a 2015 count the study reports.

The UC-30 mines are in direct contact with the water. Their metal casings, the authors write, are largely or entirely eaten away by corrosion. The research was funded by the European Union through the Interreg North Sea programme, and the authors declare no competing interests. They call for long-term monitoring and a risk assessment wreck by wreck. The study does not say how much TNT is left in UC-30, or how long it will take for all of it to get out.

References
Schuster Romina Marietta, Binder Franziska Isabell, Bünning Lillian Tabea Hannah, Strehse Jennifer Susanne, Andresen Katrine Juul, De Rijcke Maarten, Van Haelst Sven, Wichert Uwe, Grassel Philipp, Maser Edmund, Brenner Matthias, «Environmental risks from world war shipwrecks: Field-based biomarker evidence from caged mussels in the North Sea», Marine Pollution Bulletin, 229, 119738, online 17 April 2026, August 2026 issue, open access CC BY 4.0. doi: 10.1016/j.marpolbul.2026.119738
Schuster Romina Marietta, Binder Franziska Isabell, Marx Ute, De Rijcke Maarten, Andresen Katrine Juul, Van Haelst Sven, Brenner Matthias, Möller Lars, «A multi-biomarker approach using caged blue mussel, Mytilus edulis, on three World War shipwrecks in the North Sea», PANGAEA, 13 August 2025, CC BY 4.0. doi: 10.1594/PANGAEA.984199
Bünning Lillian Tabea Hannah, Strehse Jennifer Susanne, Scheer Lars Christian, Schuster Romina Marietta, Marx Ute, De Rijcke Maarten, Van Haelst Sven, Möller Lars, Andresen Katrine Juul, Brenner Matthias, Maser Edmund, «Chemical analysis of explosive chemicals in water, sediments, and mussels (Mytilus spp.) from three shipwrecks from World War I and II in the North Sea», PANGAEA, 18 February 2026, CC BY 4.0. doi: 10.1594/PANGAEA.991466
Riisgård Hans Ulrik, Larsen Poul S., Pleissner Daniel, «Allometric equations for maximum filtration rate in blue mussels Mytilus edulis and importance of condition index», Helgoland Marine Research, 68, 193–198, 3 January 2014. doi: 10.1007/s10152-013-0377-9
Previous articles on ScienceOnline
Donati G., «The sea as a library», Marine Pharmacology & Toxicology · 1, 14 September 2026
Donati G., «The sea hare that made nothing», Marine Pharmacology & Toxicology · 2, 15 September 2026
Donati G., «Pelagia noctiluca: what to do, what not to do», Marine Pharmacology & Toxicology · 3, 21 September 2026