Killifish: standardise husbandry before the study

The African turquoise killifish, Nothobranchius furzeri, offers researchers a relatively short timescale for investigating vertebrate ageing. That advantage makes husbandry records especially valuable: a feeding change, an undocumented incubation history or a tank incident may alter the animals’ trajectory before the experimental intervention begins. Research facilities need a clear account of those conditions to decide what a difference between groups actually means.

Start with cohort identity, not the species label

A tank labelled “killifish” does not identify an experimental population adequately. Species, strain, generation, sex and previous housing belong in the study record. Age also needs an operational definition. Egg collection and hatching are separate events, and their dates cannot simply be substituted for one another when reporting an animal’s history.

Embryonic diapause is particularly relevant in this annual species. Nath and colleagues’ husbandry protocol covers the management of embryos as well as hatching and subsequent care. It provides a practical framework for maintaining a colony, rather than evidence that every variation between facilities has no biological consequence. Recording incubation history helps researchers recognise a potential cohort difference before assigning it to the intervention under investigation.

An effective cohort record connects parental origin, egg collection, incubation conditions, hatching and housing changes. That identity should follow animals through transfers. If groups are combined, document the event explicitly. A future analyst should be able to reconstruct which animals shared an earlier environment, rather than relying on the arrangement of tanks at the end of the study.

Treat a diet change as an experimental event

Food affects more than whether fish grow. In their 2020 study, Žák and colleagues compared MZCS 222 killifish receiving frozen chironomid larvae with fish weaned onto a commercial pelleted diet. Several growth and reproductive measures were comparable. However, pellet-fed animals had more visceral fat, and some liver findings differed with diet and sex. Similar external performance therefore did not mean an identical physiological baseline.

The study supports the feasibility of a formulated feed under its tested conditions. It does not establish interchangeability across all strains, housing arrangements or lifespan experiments. In particular, liver histology needs to be interpreted against an appropriate dietary background. A facility adopting another feed should define what it needs to validate for its own research endpoints rather than assuming that acceptance of pellets answers every nutritional question.

Record product identity, batch, storage, particle size, feeding schedule and the transition procedure. An emergency replacement belongs in the incident log even when fish readily accept it. Uneaten food, feeding refusals and unequal access can reveal problems that a record of the amount offered will miss. If ration size changes with fish size, set out the adjustment rule so that operators apply it consistently.

Measure the environment experienced by the fish

A system setpoint is useful, but it is not a complete description of conditions in every tank. Temperature, light cycle, water quality and flow should be documented at a scale appropriate to the study. A monitoring plan can include representative checks across rack positions and operating periods where differences are plausible. The objective is to identify meaningful variation, not to accumulate readings without a decision attached to them.

Life stage matters. Nath and colleagues describe husbandry adjustments through development, including the need for young fish to reach food without unsuitable flow carrying them away. Adult male aggression requires deliberate housing arrangements and observation. Covers also help prevent jumping. These routine details connect animal welfare directly to the reliability of the research environment.

When equipment fails or maintenance changes a tank’s conditions, record affected groups and the likely duration. A disturbance limited to one part of a rack can become confounded with treatment allocation. Prompt reporting gives the scientific team a chance to apply its predefined approach to unexpected events. Omitting an incident from the record does not remove its possible biological effect.

Make behavioural observations reproducible

Quiet fish are not necessarily unstressed fish. Feeding access, aggression, swimming and body condition need to be assessed together and in context. A short ethogram can help staff describe the same observations consistently. Its usefulness depends on clear, observable definitions and a stable recording method, rather than on the number of behaviours it contains.

Alert criteria and humane endpoints should be agreed with the designated veterinarian and project leadership before the experiment begins. In an ageing study, deterioration associated with a husbandry problem should not automatically be classified as normal ageing. Necessary care takes priority; documenting the intervention allows the resulting observations to be interpreted appropriately afterwards.

A preparatory recording period is often useful. Ask two operators to assess the same defined situations and compare their descriptions. Do they recognise a feeding refusal in the same way? Can they distinguish an aggressive encounter from routine proximity? Do they know whom to contact and how quickly? Resolving those practical questions before primary data collection reduces avoidable ambiguity in the study record.

Agree what standardisation means for this experiment

Standardisation is an explicit set of suitable conditions, supported by records of what actually happened. It is not a requirement to copy every detail from a published facility. The two sources discussed here serve different purposes: one provides a husbandry protocol, while the other evaluates a dietary transition in a defined experimental setting. Their recommendations and findings should be transferred with those limits intact.

Before starting, bring the cohort identity record, feeding procedure, water monitoring plan and observation sheet together. Discuss the experimental unit and the handling of unexpected events with the person responsible for analysis. Fish sharing a tank may also share influences that matter to the outcome; whether they represent independent replicates depends on the design and question being tested.

This preparation also makes later protocol changes easier to assess. State which changes require a fresh baseline or a separate cohort and who makes that decision. Where a deviation is unavoidable, retain enough information to describe it clearly in the methods and limitations. A consistent record is more useful than a nominal standard that gradually diverges from everyday practice.

Build an interpretable history of the animals

A well-run colony gives researchers an account of what their animals experienced, not merely a list of equipment settings. Vetofish can support water surveillance, health assessment and the design of practical observation records in coordination with the designated veterinarian and scientific team. Addressing husbandry factors before the study starts helps the team distinguish an experimental finding from a change in the conditions that produced it.