Gas bubble disease: trace the system fault

Microbubbles in an aquarium are not always a cosmetic water-quality issue. When total dissolved gas pressure exceeds ambient pressure, bubbles can form in fish blood and tissues. This physical, non-infectious condition is known as gas bubble disease. A published outbreak in captive golden trevally shows why routine oxygen readings may miss the hazard. The investigation must connect what is happening to the animals with total gas pressure and the real operating state of the life-support system.

The risk can be deceptive. Temperature, pH, nitrogen compounds and dissolved oxygen may all appear acceptable while an air leak, pump or pressure change maintains supersaturation. Empirical treatment is therefore not the first answer. The immediate priorities are to identify the physical source, reduce exposure safely and document lesions with veterinary input.

Clinical signs point to a process, not a diagnosis

Supersaturated water drives gas into the body across the gills. Gas nuclei can then grow inside vessels or surrounding tissues. Intravascular bubbles may obstruct circulation, while extravascular gas can separate or compress adjacent structures. The gills are central because they exchange oxygen and carbon dioxide while also contributing to ion balance, acid-base regulation and nitrogenous waste excretion.

Observable signs are not uniformly specific. Abnormal swimming, loss of equilibrium, lethargy, reduced response to stimuli, darkening or repeated collisions with a wall should trigger investigation, but none confirms the disease alone. Visible bubbles in fins, gills, the lateral line, oral mucosa, periocular tissues, cornea or aqueous humour strengthen the suspicion. Exophthalmos, visual impairment and mortality can occur in severe cases.

A differential diagnosis remains necessary. Trauma, infection, toxins, inadequate oxygenation, buoyancy disorders and other systemic conditions may produce overlapping signs. Conversely, the absence of obvious external bubbles does not exclude branchial emboli or internal tissue damage.

What the golden trevally outbreak revealed

Hong and colleagues described 53 golden trevally (Gnathanodon speciosus) transferred to a new exhibit at a South Korean aquarium. About one month later, the first fish became lethargic and swam abnormally. Between July and September 2023, 24 individuals, or 45.3% of the group, developed signs. Bubbles were visible in the cornea, aqueous humour and oral mucosa; some fish lost equilibrium or struck the tank wall.

This was a case report, not a controlled trial. Detailed pathology compared only one symptomatic fish with one fish without apparent signs. In the affected animal, histopathology showed gas spaces in the hepatopancreas, gills, heart, muscle and blood vessels. Bacterial cultures did not identify an infectious cause. The observations supported mechanical gas injury, but they did not establish a species-specific tolerance threshold or a risk estimate transferable to every aquarium.

Routine measurements looked reassuring. Dissolved oxygen ranged from 7.74 to 8.13 mg/L, equivalent to 90.22–94.77% oxygen saturation. The team eventually identified persistent microbubbles and a plumbing problem. When the skimmer water level fell, a bypass pipe could draw air towards the circulation pump. The pump fragmented that air, helping small bubbles persist within the system. After the bypass was removed, no further gas-bubble-disease mortality was reported during several months of monitoring.

That sequence is a coherent causal investigation, not experimental proof. Its practical value lies in combining pathology, hydraulics and the response to a targeted engineering correction.

Why a normal oxygen value is not enough

An oxygen probe measures the oxygen component, not the sum of the partial pressures of all dissolved gases. Water can therefore meet an aquarium’s oxygen target while nitrogen or total gas pressure creates a harmful gradient. Visible microbubbles do not quantify that gradient either.

The relevant measurement is total gas pressure relative to local barometric pressure. Interpretation should account for instrument performance, depth, temperature, salinity, exposure duration and gas composition. Hydrostatic pressure increases with depth and can partially compensate for supersaturation. A fish confined near the surface may therefore have a different effective exposure from one able to move deeper.

A 2026 experiment exposed 1,440 Atlantic salmon, brown trout, rainbow trout and European minnows at several life stages to 100–120% total dissolved gas. Responses differed substantially between species and stages. Acute disease occurred in salmonid parr at 115–120%, with gill emboli and subcutaneous emphysema during the stronger exposures. Larvae, fry and minnows followed different patterns. Those values describe specific experimental conditions; they are not universal action limits for diverse aquarium collections.

A structured investigation for the first hours

Teams can organise the initial response around five practical questions.

1. Which animals are affected? Record species, size, life stage, occupied depth, tank, onset time, behaviour, visible lesions and mortality. Video swimming and photograph eyes, fins and gills when this can be done without adding handling stress.
2. What changed in the system? Look for recent transfers, restart or maintenance work, falling water levels, valve changes, pump replacement, rapid heating, air aspiration, cavitation or pressurised incoming water.
3. Where should measurements be made? Compare incoming water, pump discharge, degassing zones, surface and depth. Record total gas pressure, barometric pressure, temperature, salinity and oxygen together with time and equipment status.
4. Does the physical scene make sense? Check for water clouded by microbubbles, unusual flow, pump noise and interfaces that may draw air. Direct inspection should complement alarms, not replace them.
5. What veterinary confirmation is required? The veterinarian determines whether gill examination, sampling, imaging, necropsy or histology is justified while keeping competing diagnoses open.

Moving a symptomatic fish does not remove a common waterborne exposure. Transfer decisions should balance handling stress against the benefit of genuinely safe water. A treatment tank connected to the same system will not interrupt exposure.

Correct the source and prove recovery

The engineering response depends on the installation. It may involve controlled shutdown or bypass of faulty equipment, elimination of an air intake, restoration of an operating water level, verified degassing or supply from a tested water source. An improvised correction must not compromise oxygenation, temperature, salinity or biological filtration.

Improvement needs evidence. Repeat measurements at the same locations, test stability at different pump regimes, monitor the animals and document whether new cases stop. Restoring acceptable total gas pressure does not guarantee that all lesions will reverse. In the 2026 study, recovery trials were limited to selected groups and showed only partial recovery after sublethal exposure.

The incident review should then become prevention: an up-to-date hydraulic diagram, documented minimum operating levels, defined measurement points, tests after maintenance, meaningful alarms and a response procedure shared by aquarists, life-support technicians and veterinarians.

Conclusion

Gas bubble disease is an environmental system failure expressed through the animals. Ocular, branchial or behavioural changes may guide the investigation, but absent external bubbles do not rule out internal injury. A normal oxygen measurement cannot replace total gas pressure. Experimental evidence also confirms that vulnerability varies with species, stage, size, exposure duration and depth, so no isolated percentage should become a universal aquarium rule.

Vetofish can help public aquariums build an investigation plan, coordinate veterinary examination and diagnostics, review life-support critical points and formalise post-correction monitoring. The aim is to locate the physical cause, reduce exposure and turn an incident into verifiable prevention.