The brine shrimp toxicity project
For a school science fair, Kyle, 17, proposes to expose 240 brine shrimp (Artemia) to four concentrations of a household bleach-based cleaner plus a control, recording mortality at 24 and 48 hours, and repeating the test three times. The biology teacher supports it as a good example of dose-response. The school has no formal animal ethics policy; brine shrimp are not covered by animal welfare legislation, and the cleaner is already sold in supermarkets with hazard warnings, with published toxicity data available online. Kyle has a budget of 30 euro and needs results in two weeks. The deputy head asks the ethics department whether the project should be approved, whether the number of animals matters, and whether there is a difference between testing on brine shrimp and testing on mice.
1. Apply the three Rs and a cost-benefit assessment to Kyle's proposal.
Read the worked answer
Replacement. Could the educational aim be met without animals? Yes, substantially: published toxicity data exist for the product, and a literature analysis, a database exercise, or a simulation would teach dose-response just as well. Replacement is not absolute - hands-on work motivates students - but the least sentient option that achieves the aim should be chosen.
Reduction. 240 animals across three replicates and four concentrations plus control implies no power calculation and no sequential design. Fewer concentrations, a pilot, sequential testing that stops when an effect is clear, or sharing data with another class would all reduce numbers. The onus is on the student to justify the number, not on the committee to prove it excessive.
Refinement. Death is the most severe possible endpoint. Sub-lethal endpoints (immobility, impaired swimming), humane endpoints, shorter exposure, careful handling, appropriate temperature and water quality, and disposal would all reduce suffering. Kyle should be supervised by someone competent to recognise distress.
Cost-benefit. Harms: killing or damaging up to 240 animals, if they can be harmed at all. Benefits: a school science-fair exhibit, in a case where the answer is already known and published. The benefit is modest and largely pedagogical, and it can be obtained by replacement. On that basis the proposal as written does not pass the cost-benefit test; a redesigned version could.
2. Does the capacity of brine shrimp to feel pain matter morally, and how does speciesism bear on the question?
Read the worked answer
Why sentience is the relevant criterion. Moral status is generally grounded in the capacity to have experiences that matter to the being - pain, distress, perhaps fear. Species membership by itself is not a reason to discount suffering: that is the core of Singer's argument against speciesism, the view that we favour members of our own species simply because they are our species.
The empirical complication. Brine shrimp are small arthropods, and evidence of nociception and integrated pain experience in them is limited and contested. It is reasonable to hold genuine uncertainty. Two mistakes are available: assuming that because they are small and unlike us they cannot suffer; and assuming that any nervous system entails suffering. The proportionate response to uncertainty is precaution - avoid killing large numbers where the purpose does not require it.
Speciesism properly understood. Anti-speciesism does not mean all animals count equally. It means that if a being can suffer, that suffering counts, and that equal suffering should be weighed equally unless there is a relevant difference. The relevant difference between a shrimp and a mouse is not species, but the likelihood and richness of suffering, which is an evidential question. This justifies much stricter controls on mouse research without licensing indifference to shrimp.
The residual point. Even where suffering is doubtful, deliberately killing 240 animals for a demonstration that teaches nothing new still requires justification, because respect for living things is a virtue the school is trying to cultivate, and the students are watching what it permits.
3. State conditions under which the project would be acceptable, and one version that would not be acceptable even with supervision.
Read the worked answer
Conditions for acceptability. (1) Replacement is attempted first and the residual need for live animals is argued in writing. (2) Numbers are justified by a power calculation or a sequential design, and are the minimum needed. (3) A sub-lethal or humane endpoint replaces mortality. (4) Exposure times and concentrations are limited to what is necessary, with the lowest number of concentrations that answers the question. (5) A named, competent supervisor oversees handling, husbandry and disposal, and can stop the work. (6) The project is approved by a designated staff member with a written record of the reasoning. (7) Students are briefed on the ethics, not only the method, and report on the ethical reasoning as part of the project.
A version that would not be acceptable. Repeating an already-published acute toxicity test on a supermarket product, at increasing concentrations chosen to produce visible mortality because that makes a more dramatic display, with 240 animals per replicate and no reduction, on the grounds that 'they are only shrimp'. Even with perfect supervision, the harms are real if the animals can suffer, the scientific and educational benefit is duplicative, and the exercise teaches that small, unfamiliar lives may be expended for spectacle. Supervision addresses welfare, not necessity - and necessity is where the proposal fails.
The general principle. Animal ethics review asks not only 'was it done kindly?' but 'did it need to be done at all?'. Kyle's project needs redesign, not merely a supervisor.