Amoebic gill disease: monitor before losses

Amoebic gill disease (AGD) does not begin when Atlantic salmon stop feeding. Neoparamoeba perurans can colonise gills while production indicators remain only subtly changed. Waiting for a clear decline in appetite, growth or survival therefore means allowing respiratory function to deteriorate before acting. A useful programme connects stock observation, standardised gill scoring, laboratory confirmation and the environmental history of the site.

One gill disease, several layers of evidence

AGD is associated with N. perurans on the gills. Gross lesions often appear as pale, thickened patches, but their distribution differs between arches and individual fish. Microscopically, epithelial hyperplasia and lamellar fusion reduce the surface available for gas exchange. Fish then need greater ventilatory effort to maintain oxygen transfer.

No single measurement captures that progression. Gross scoring describes accessible lesions, PCR provides evidence of the amoeba and its relative load in the specimen, and histology shows the distribution of tissue damage. None of them alone measures the respiratory capacity of the population or defines the urgency of treatment.

Observable signs are non-specific: faster opercular movements, use of highly oxygenated areas, reduced activity and declining feed intake. Similar changes can accompany other gill pathogens, jellyfish exposure, hypoxia, recent treatments or poor water quality. The investigation therefore needs to preserve a broad differential diagnosis.

What recent monitoring studies add

A field study during a natural Tasmanian outbreak compared salmon at low and high AGD prevalence and after freshwater treatment. Gross gill scores were associated with histopathology, yet substantial variation remained among fish, particularly after treatment. A population average can therefore conceal very different individuals, and an immediate gross improvement does not demonstrate complete tissue recovery.

A 2026 proof-of-concept study measured respiratory frequency by camera during an experimental challenge. Ventilation increased alongside gill scores, histological lesions and PCR findings. The design, however, used few experimental units. The authors appropriately frame respiratory monitoring as hypothesis-generating rather than as a universal decision threshold ready for every commercial farm.

Video monitoring may nevertheless provide an early, continuous signal without handling fish. Its best current role is to trigger focused investigation. A persistent change in respiratory frequency can prompt gill examination before growth and survival decline, while confirmation still depends on clinical and laboratory evidence.

Make gill scores comparable

Scoring only supports decisions when teams examine the same structures in the same way. A written plan should specify the number of fish, arches inspected, scoring scale, trained observers, sampling frequency and capture conditions. Reference images and periodic calibration sessions reduce observer drift.

The sample should represent the population, not merely fish that are easiest to catch or already severely affected. Record scores with temperature, salinity, dissolved oxygen, feed intake, mortality, handling and recent treatments. The time series is more informative than a single score.

When the trajectory is unusual, gill swabs or small biopsies can add molecular evidence. Downes and colleagues found that gill swabs paired with validated assays could support non-destructive detection. Sampling technique, preservation, laboratory method and interpretation must still be agreed in advance. A positive molecular result does not measure lesion extent or prove that one organism explains all respiratory compromise.

Separate detection from disease

The amoeba’s presence should be interpreted with lesions, population dynamics and other possible gill insults. A weak PCR signal in clinically normal fish does not carry the same meaning as concurrent increases in gill score, ventilation and mortality. Conversely, extensive lesions with a negative PCR result warrant review of sample quality and investigation of alternative causes.

Treatment decisions depend on local rules, farm history, lesion severity, temperature, salinity, welfare constraints and operational capacity. Freshwater bathing is used in some salmon industries, but efficacy, feasibility and welfare impact cannot be copied from one site to another. Any procedure must account for crowding, handling, water quality, recovery and reinfection risk.

Post-treatment assessment should extend beyond the following day. Tissue lesions can persist while parasite load changes. Follow-up points need to use the same indicators and be timed to the production system and expected disease dynamics.

A three-level operating plan

At stock level, teams observe ventilation, distribution, appetite and mortality while checking water parameters. At a frequency set by risk, they perform standardised gill scoring on a representative sample. When trends change or a consequential intervention is considered, veterinary examination, histology and an appropriate molecular assay are combined.

This system needs locally derived alerts rather than a number copied from one paper. Cameras can supply a continuous respiratory indicator, but interpretation must be calibrated for fish size, light, current, temperature and other drivers of ventilation. Data become actionable only when the protocol states who reviews them, how an alert is confirmed and which investigation follows.

Conclusion

Early AGD surveillance means connecting complementary observations. Gross scoring describes visible damage, PCR detects the agent in the specimen, histology characterises tissue response, and respiratory monitoring may reveal a continuous shift. The greatest practical gain comes from defining sampling, alerts and responsibilities before the high-risk period.

Vetofish can help design that programme, train teams in gill scoring, coordinate diagnostic sampling and interpret results alongside husbandry data so that decisions match the actual risk at the site.