Fish necropsy: protect diagnostic value

A dead fish can still answer important questions for the rest of an aquarium collection, but its tissues start changing immediately. Autolysis alters cells and organ architecture, while bacterial multiplication or contamination can blur microbiological findings. The operational question is therefore not simply whether to perform a necropsy. It is how to preserve enough reliable information for the veterinarian and diagnostic laboratory to distinguish disease, husbandry-related injury and post-mortem artefact.

Gill tissue may change within minutes

An experimental study in the Journal of Fish Diseases evaluated post-mortem histological change in farmed Atlantic salmon (Salmo salar). Fish were kept at room temperature (21°C), refrigerated at 4°C or frozen at −20°C, then examined at 0, 1, 4, 24 and 48 hours. The researchers also sampled gills at short intervals during the first 45 minutes at room temperature.

Gill changes were already visible after 10 minutes. Pyloric caeca and pancreatic tissue were also among the fastest tissues to deteriorate. Moderate to severe autolysis generally appeared from four hours, while severe changes affected several organs by 24 hours. Heart and skeletal muscle remained more resistant than gills and digestive tissues.

Cooling slowed some early changes but did not make delay harmless. Freezing did not stop autolysis and introduced artefacts that could complicate microscopic interpretation. This does not mean that freezing is wrong for every purpose: some molecular tests require frozen material. It means that freezing the whole fish by default may sacrifice histology. The receiving laboratory should specify the correct container, temperature and timing for each planned test.

The necropsy plan starts before a death

The most effective improvement is made before an emergency. A public aquarium should know who authorises examination, which cases have priority, where clean equipment and containers are stored, what exhibit data must accompany the animal, and which laboratory will receive the samples. The plan should separate an expected isolated death from a possible collection-level event, such as rising mortality, similar findings across species, a recent water-quality incident, an animal introduction or a technical failure.

Case information begins at the exhibit. Useful records include species and individual identity, size, sex when known, time found, last observation alive, clinical signs, treatment, feeding, recent movements and water parameters. Clear photographs of the animal, lesions and relevant surroundings add context. Without a timeline, even a technically correct laboratory result may be difficult to relate to the event.

Freshly dead animals usually offer more information than those recovered after an unknown interval. Where clinical circumstances and veterinary judgement justify it, a moribund animal examined immediately after appropriate euthanasia may provide better diagnostic material. FELASA–AALAS recommendations for laboratory fish caution that fish found dead are poor histology candidates because autolysis is prompt. Such fish may still support PCR testing, but post-mortem bacterial overgrowth must be considered. These research-facility recommendations are not universal rules for every aquarium species; they demonstrate why time of death and sample history belong in the interpretation.

Sampling order protects the evidence

A diagnostic necropsy is not just a dissection. Its order depends on the clinical question and tests requested. Samples requiring aseptic collection, particularly for bacteriology and some virology workflows, need to be obtained before opening contaminated surfaces or reusing instruments compromises them. Fresh mounts for parasite examination and cytology also lose information quickly.

Histology samples should then represent the correct anatomical sites, with tissue thickness, orientation and fixative ratio agreed with the laboratory. A 2023 paper describes standardised locations and orientations for around 40 organs and tissues in rainbow trout weighing 300–2,000 g. The work was designed for ecotoxicology rather than diagnosis across aquarium collections, but its general lesson is useful: consistent sampling reduces variation caused by anatomy, sectioning angle and tissue selection.

Molecular, microbiological and histological samples are not interchangeable. One piece of tissue cannot simultaneously sit in fixative, bacterial transport medium and an RNA-preserving workflow. Containers should be laid out and labelled before the animal is opened. If tissue is scarce, tests should be prioritised against the diagnostic question instead of collecting a little of everything without a clear purpose.

Taxon-specific anatomy changes priorities

Generic checklists can miss high-value sites in unusual taxa. A 2024 protocol for wild and aquarium-housed sharks and rays gives early priority to the chondrocranium, including endolymphatic pores and ducts, inner ears, brain and olfactory structures. The method was developed across 189 elasmobranchs from 18 species. It does not replace a whole-body examination, but it shows that leaving the head until last may reduce diagnostic yield in these animals.

The same principle applies to very small fish, syngnathids, flatfish and species with fragile organs. Available tissue volume, anatomical access, likely diseases and speed of autolysis all change the plan. A short taxon-specific sheet for the collection’s priority groups is safer than one identical checklist for every fish.

Five actions for an aquarium team

Public aquariums can strengthen post-mortem investigation through five practical measures:

1. define when a necropsy and a collection-level alert are triggered;
2. maintain a kit that separates clean instruments, containers, fixative, labels and personal protective equipment;
3. agree tissue, volume, temperature and transport requirements with the laboratory for each test method;
4. train keepers to document the animal and preserve the timeline without performing unauthorised procedures; and
5. connect the necropsy report with exhibit data, related cases and corrective actions.

These measures do not guarantee a cause of death. Gross lesions may be non-specific, a positive PCR result does not always prove disease, and a negative culture does not exclude an agent when collection or transport was suboptimal. Interpretation needs the combined picture: epidemiology, clinical signs, water quality, pathology, microbiology and molecular testing.

Limits of the evidence

The post-mortem timing study used one species under controlled storage conditions. Its minute- and hour-based observations should not be converted into universal acceptance limits for tropical marine fish, large elasmobranchs or chilled-water species. The trout sampling paper was developed for ecotoxicology, and the FELASA–AALAS recommendations focus on research colonies. The elasmobranch protocol is deliberately anatomy-specific.

Together, these sources support a principle rather than a single recipe: collect perishable information early, document the delay, preserve separate material for different methods, and use species-specific anatomy. Each aquarium should validate its procedure with its veterinary and laboratory partners.

Conclusion

Diagnostic value starts declining in the first minutes after death, but not at the same rate in every tissue or for every test. Gills and some digestive tissues require especially rapid attention. Refrigeration buys limited time; it does not replace preparation, and freezing can damage histological interpretation. A written plan, deliberate sampling order and prior laboratory coordination turn an isolated mortality into more reliable evidence for collection health.

How Vetofish can help

Vetofish can help public aquariums design or review mortality procedures, set taxon-specific sampling priorities, train teams, organise traceability and interpret clinical, environmental and laboratory findings together. The resulting workflow can be adapted to each collection’s species, facilities and risk profile without promising a diagnosis from any single test.