Sticklebacks: connect origin, genetics and husbandry

The threespine stickleback, Gasterosteus aculeatus, connects natural biological variation with experiments conducted under controlled conditions. That strength is lost when population origin and husbandry history disappear from the study record. Establishing a research colony therefore involves more than selecting an available fish species: facilities need to preserve the identities of populations, families and experimental groups. One stickleback population cannot automatically stand in for every other population.

Use diversity as part of the design

Marine and freshwater sticklebacks differ in morphology and other biological traits, making them useful for investigating adaptation. In their 2012 Nature study, Jones and colleagues combined a reference genome with sequences from twenty additional individuals representing marine and freshwater populations. They identified genomic regions repeatedly associated with divergence between these environments.

A central finding was the contribution of existing genetic variation to repeated adaptive change. Regulatory differences and chromosomal inversions were also implicated. This does not mean that every difference between two populations shares the same genetic basis or that a complex trait can always be reduced to a single locus.

The model is useful because it supports defined comparisons. A study of salinity adaptation, one of body form and one of behaviour may require different populations and different experimental conditions. Animal selection should follow the biological question rather than being determined solely by which stock happens to be available.

Keep population identity attached to the data

For each group, records should connect source population, generation in captivity, family or cross and transfer history. This is an operational recommendation derived from the comparative nature of the model. A label containing only the species name can erase the very factor the experiment was intended to investigate.

Genotype and phenotype should also remain distinct. External appearance is not a substitute for genetic identification, and an associated genetic variant does not describe every aspect of an animal’s performance. Images, measurements and samples should use stable identifiers rather than relying on the tank position at the time of collection.

Family structure and housing units need to be considered when interpreting replication. Several related fish housed together do not provide the same information as multiple families distributed across independent units. The level at which the study will draw conclusions should be established before fish are assigned to tanks, with statistical input appropriate to the design.

Build the facility around the selected stocks

The Stickleback Stock Center’s welcome guide, updated in March 2025, describes circulation, aeration, biological filtration, thermal control and water monitoring. It includes both individual tanks and larger connected systems. This practical experience helps facilities plan their equipment; it should not be represented as a universal welfare standard.

Shared water can make environmental control easier while also connecting groups epidemiologically. Independent systems may support isolation but require separate equipment checks and records. The chosen configuration should therefore appear in the biosecurity plan as well as in the experimental description.

Temperature and salinity need to be recorded as experienced conditions, including relevant variation, rather than simply as controller settings. Fish from a freshwater population should not automatically receive conditions copied from a different stock. The team must establish that the selected environment suits the animals and question, and specify how it will respond to system drift or equipment failure.

Match feeding to development and observed intake

The Stock Center’s feeding protocol uses newly hatched Artemia for young fish and describes different preparations for juveniles and adults. It emphasises food particle size, observation of consumption and removal of uneaten material. Its recipes document a particular facility’s practice; they are not comparative evidence that one formulation is best for every research programme.

Feed records should be detailed enough to interpret changes in growth or reproduction. Product, batch, particle size, schedule and ration changes can all be relevant. A fixed quantity delivered to a tank does not establish that every fish consumed the same amount, particularly if body size or access to food differs.

Planning a dependable live-food supply before hatching is also useful. A production interruption should not become an undocumented change in the experimental environment. Staff can distinguish planned feeding, actual distribution and observed consumption without claiming to have measured individual intake when only group-level observations were made.

Interpret behaviour with husbandry information

Behavioural endpoints should be defined for the study: feeding responses, activity, social interactions or use of space, for example. A difference between groups may be biologically meaningful, but it also needs an environmental and clinical interpretation. A sudden loss of activity should not automatically be accepted as the expected expression of a genetic difference.

The team should agree beforehand which signs prompt a system check, veterinary assessment or review of an animal’s participation. Stress-related observations and maintenance events should remain linked to the data. This can prevent an equipment problem from being mistaken for a population-specific effect.

Acclimation and handling history also belong in the experimental record. Their timing cannot be supplied by a universal rule independent of animal condition and study objectives. When welfare considerations require a change, that decision and its possible effects on comparison should be documented rather than hidden to preserve an apparently uniform protocol.

Recognise the limits of the evidence

The genomic study establishes the value of comparisons across populations and repeated environmental transitions. It does not provide a complete husbandry prescription. Conversely, a stock centre’s operational guidance is useful experience but does not test every combination of provenance, life stage and research endpoint.

A facility starting a colony can use these sources to prepare a justified local plan, identify unresolved questions and define what will be monitored. It should retain the option to revise that plan when animal observations or system performance show that an assumption was incorrect.

Vetofish can support the review of housing conditions, incoming-stock health and routine observations. The aim is to protect the animals while maintaining the traceability required to interpret an experiment: which fish were studied, under which conditions, and how far the findings can reasonably be applied.