Zebrafish enrichment without losing control

Adding an artificial plant, shelter or substrate image to a zebrafish tank seems straightforward. Yet each item may alter behaviour, access to resources, cleaning and experimental outcomes. An open 2023 study shows why stocking density and enrichment cannot be managed as separate questions. It supports testing standardised, observable and biosecurity-compatible options rather than treating bare tanks and complex décor as the only choices.

What the experiment compared

Researchers reared AB-strain zebrafish in 9.5-litre tanks at 1, 3 or 6 fish per litre. Four environments were compared: bare tanks, a gravel image, an artificial plant, or both the image and plant. Three replicate tanks were used per condition, for 48 groups in total.

Agonistic behaviour was observed twice weekly for nine weeks. The team also measured water-borne cortisol after a group-level stressor, tested individual fish in a novel tank and analysed brain monoamines. This mix of measures matters because no single marker summarises welfare.

The clearest result concerned density. Fish kept at 1 fish/L were more aggressive than those at 3 or 6 fish/L and released more cortisol after water level was lowered. At 6 fish/L, growth was lower and more variable. Within the study conditions, the authors therefore considered the likely optimum to lie between 3 and 6 fish/L.

Enrichment did not produce one uniform benefit across every measure and did not clearly alter the individual novel-tank test. This does not prove that enrichment is useless or that one plant suits every facility. Its effects depend on design, group, density and outcome.

Define the function before selecting an object

Enrichment should answer a stated need: provide retreat, structure space, encourage exploration, diversify flow or support feeding behaviour. Without a functional objective, staff cannot decide whether an item improved the situation.

Spontaneous preference alone is not proof of lasting benefit. An attractive resource can be monopolised by dominant fish, increase encounters or hide a compromised animal. Observation must therefore cover the whole group and the distribution of access, not only average use.

For an initial trial, choose a simple reversible modification. An external substrate image adds visual structure without placing material in the water. A smooth three-dimensional object may provide shelter, but it must be sized to prevent entrapment and positioned so it does not create a hydraulic dead zone.

Make hygiene a design requirement

Sanitary compatibility should be assessed before purchase. Materials must tolerate the facility’s validated cleaning and disinfectant process without known leaching under the intended conditions. Porous surfaces, hard-to-disassemble joints, sealed cavities and sharp edges increase the risks of biofilm, retained organic matter or injury.

Each object type needs a written procedure covering low-stress removal, cleaning, inspection, disinfection, rinsing, drying and discard criteria. If items move between tanks, their pathway must follow unit health status. Tank- or rack-dedicated enrichment limits cross-transfer.

Visibility remains essential. Staff must be able to count fish, observe feeding and rapidly detect abnormal swimming, loss of condition or mortality. Décor that prevents reliable surveillance is not a controlled refinement. One solution is a structured zone paired with an open area and a clear observation angle.

Protect scientific value

The rearing environment is an experimental variable. Protocols and publications should describe enrichment as precisely as density, photoperiod or feed. Introducing objects to a colony without recording date, cohort and line can create unexplained heterogeneity.

A pilot should define measures before the change: aggression, shoal dispersion, zone use, feeding, growth, mortality and water quality. It should last beyond the initial novelty response and include several tanks, because the tank is the relevant experimental unit for many group outcomes.

The Sen Sarma study should not become a universal prescription for 3–6 fish/L. Working volume, flow, age, strain, sex, feeding and scientific purpose can change the response. The important lesson is that very low density is not automatically favourable and that density, enrichment and growth should be evaluated together.

Run a reversible pilot

Start with a limited number of racks and one standardised enrichment type. Before, during and after the pilot, trained staff use the same ethogram and record maintenance events. Water quality and cleaning time are monitored to identify hidden costs.

Predefined stopping criteria may include sustained aggression, injury, unequal feed access, water-quality deterioration, unreliable counting or disinfection failure. Conversely, distributed use, better recovery after routine intervention and preserved surveillance may support gradual expansion.

The final decision should involve husbandry staff, the designated veterinarian, the animal-welfare body and scientific leads. It must also define exceptions for quarantine, infection studies, behavioural phenotyping or projects requiring a specified environment.

Standardise what matters

Standardisation does not require every tank to be bare. It requires the relevant environmental features to be specified and consistently applied. Record object identity, material, position, replacement interval and cleaning method. Photograph the standard arrangement and document any deviation.

If enrichment could influence a measured phenotype, incorporate it into study design rather than removing it by default. Randomisation, blocking and transparent reporting can preserve inference. The welfare cost of a barren environment and the scientific cost of unrecorded complexity are both avoidable when facility and project teams plan together.

Conclusion

Zebrafish enrichment is a refinement process, not decoration. Available evidence shows context- and density-dependent effects, with possible but non-uniform benefits. A robust option has a defined function, remains cleanable and observable, is tested across replicate tanks and is reported as an experimental variable.

Vetofish can support aquatic facilities with risk assessment, indicator selection, pilot protocol design and joint evaluation of welfare, biosecurity and reproducibility.