Cephalopod sentience: reassess research protocols

A single label is no longer adequate for cephalopod welfare. A 2026 assessment in Biological Reviews found strong evidence of sentience in octopuses and cuttlefish, substantial evidence in squid, and far less evidence in nautiluses. This uneven record should not lead facilities to treat all species as interchangeable, or to confuse a lack of data with evidence that a capacity is absent. The operational implication for research institutions is clear: species, procedure, welfare indicators, humane endpoints and pain management must be connected to evidence that has actually been established.

Eight criteria organise the evidence

Schnell and colleagues reviewed more than 120 studies against eight criteria. Four address the nervous system: nociceptors, integrative brain regions, connections between nociceptors and those regions, and modulation by anaesthetic or analgesic compounds. Four concern behaviour: motivational trade-offs, avoidance learning, wound-directed protective behaviour and seeking a location associated with relief.

Octopuses and cuttlefish were rated with high or very high confidence on six of the eight criteria. Squid reached that level on five. Nautiluses reached it on only one. The authors make an important qualification: whenever a taxon did not meet a criterion, this reflected insufficient evidence rather than strong evidence that the capacity was absent. The scores map the current literature; they are not a final ranking of sensitivity.

The framework is neither a diagnostic test for an individual animal nor an independent regulatory standard. It helps distinguish converging neural and behavioural evidence from a simple reflex response. It also requires octopuses, cuttlefish, squid and nautiluses to be considered separately rather than treating findings in one group as automatically transferable to another.

Pain is more than arm withdrawal

An experiment published by Crook in 2021 provides one influential part of this evidence. Octopuses exposed to a noxious stimulus subsequently avoided the chamber associated with that experience. After local anaesthesia provided relief, they instead preferred the place associated with relief. Wound-directed grooming and prolonged neural responses added to the pattern. Together, the results are consistent with an affective component of pain rather than an automatic withdrawal response alone.

They do not allow staff to assign a precise subjective state to every colour, posture or activity change. A single sign may reflect pain, fear, exploration, altered water conditions, sleep state or normal species-specific behaviour. Monitoring should therefore combine activity, mantle ventilation, feeding, shelter use, posture, coloration, attention to a body area and change over time. Baseline observations must be collected before a procedure so that each animal can be compared with its own normal range.

An ethogram is valuable only when it supports a decision. Teams should define who observes, when and for how long, under which lighting conditions, and which rule triggers reassessment, treatment, procedure termination or humane killing. Video may improve consistency, but recording does not replace a predefined interpretation and response plan.

Analgesia remains a major evidence gap

Recognising the risk of pain does not mean that a validated drug protocol is available. A 2026 study evaluated lidocaine, bupivacaine and extended-release bupivacaine in 83 juvenile Octopus bimaculoides used in an arm-regeneration model. Postoperative changes in posture and gait resolved within 72 hours, but none of the tested treatments demonstrated sufficient local analgesia. Lesions were also observed around injection sites.

The study does not show that all analgesia is ineffective in octopuses. It shows that these products, delivery routes, doses and this model cannot be promoted as a universal solution. Transferring a vertebrate protocol without validation, or treating immobility as evidence that pain is absent, risks both animal welfare and scientific validity.

Before an invasive procedure, the project file should identify why the intervention is necessary, which non-invasive alternatives were considered, the selected anaesthetic approach, the analgesic strategy, monitoring parameters, observation period and rescue plan. Where no validated analgesic regimen exists, uncertainty should be reflected in severity assessment and drive smaller tissue impacts, earlier endpoints and specialist veterinary review. It should not be converted into false reassurance.

European law is a baseline, not a complete protocol

Directive 2010/63/EU includes all living cephalopods within the scope of legislation protecting animals used for scientific purposes. It applies replacement, reduction and refinement, project evaluation, severity classification, appropriate anaesthesia and analgesia, and early humane endpoints where possible. It does not, however, provide one monitoring or pain-relief protocol suitable for more than 800 species.

Delegated Directive (EU) 2024/1262 updated accommodation and care requirements, including separate tables for cuttlefish, sepiolids, squid and octopuses. A corrigendum published in June 2026 replaced the relevant table. EU facilities should therefore check the consolidated legislation and national implementation rather than relying on an old husbandry sheet. Regulatory minimum dimensions do not replace an assessment of functional space, shelters, visual contact, water movement, substrate and escape behaviour for the species concerned.

Protection outside the European Union remains inconsistent. International collaborations need to distinguish the legal minimum at the study site, funder or journal requirements, and the level of refinement supported by current science. The absence of species-specific local rules does not demonstrate that a practice is harmless.

Five links every protocol should make

A practical review begins with an inventory of the species and life stages actually held. For every procedure, facilities can then connect five elements: individual baseline, expected harms, observable indicators, planned interventions and humane endpoint. Scoring systems should be tested between observers and revised as new evidence becomes available.

Invasive procedures deserve particular scrutiny. Restraint, implantation, injection, sampling, surgery and recovery should not be hidden within one broad category called “handling”. Each phase creates different risks and needs an explicit timeline. Routine husbandry also matters: shelter, light, the ability to avoid visual exposure, water quality and enrichment influence both recovery and the expression of altered states.

The 2026 assessment therefore supports an evidence-based precautionary approach. It does not make all cephalopods equivalent, and it does not close the gaps around nautiluses, understudied species or medicines. Its main practical effect is to make uncertainty visible and manageable within the protocol.

Vetofish can help research facilities conduct veterinary reviews of cephalopod protocols, build observation tools, examine humane endpoints and design proportionate anaesthetic and analgesic monitoring. The purpose is to document and refine decisions, without claiming pharmacological validation that the literature has not yet provided.