Reptiles are widely cast as more primitive and reflexive compared to warm-blooded animals. However, recent studies show they can learn, express core emotions like anxiety, and even seek out pleasurable experiences. Still, the extent to which reptiles experience emotions and moods hasn’t been directly tested. Cognitive bias research, which measures an animal’s behavioral response to ambiguity, has revealed the presence of mood states in birds and mammals. But these methods have yet to be used with reptiles.

Emotions are immediate but fleeting responses to specific stimuli, whereas moods are long-term background feelings detached from any specific triggers. Both are affective states viewed as evidence of sentience, which is increasingly guiding modern animal welfare practice and policy. If we’ve been underestimating reptiles’ capacity for sustained, emotional experiences, then this knowledge gap has real consequences for the millions of captive reptiles in zoos, labs, and the exotic pet trade.

This study represents the first attempt to provide empirical evidence of mood states in reptiles using cognitive bias research methods. By observing naturally occurring behavioral differences in 15 red-footed tortoises (Chelonoidis carbonaria) across three tests, researchers investigated how persistent moods may influence the reptiles’ decision-making.

Methodology

The experiment combined a cognitive bias task that measured each tortoise’s baseline mood (optimistic or pessimistic tendencies in expectations) with a novel object and a novel environment test, both of which measured anxiety. The researchers predicted that optimistic individuals would show less anxious behavior in response to novelty.

Pre-Test Training

Each tortoise was individually tested in custom arenas with a designated start box and five marked locations arranged in an arc radiating from the starting point. A small blue food bowl could be placed at any one of these marked locations.

Since cognitive bias tasks measure how animals respond to ambiguity relative to what they already know, the tortoises were first trained to distinguish between two fixed bowl locations: a “positive location” consistently rewarded with arugula and a never-rewarded “negative location.”

After a period of habituation where they could just wander around the arena and eat, tortoises had daily training sessions of nine trials with two minute breaks in between. In every trial, the food bowl was pseudo-randomly placed at either location, and the reptiles were given up to two minutes to approach it. A tortoise was considered trained when they reached the bowl faster in each of their last three positive trials than in each of their last three negative trials across six consecutive sessions (minimum of 36 trials). Trained animals moved onto the true testing phase.

Testing Mood And Anxiety

Over three days, the tortoises underwent nine cognitive bias testing trials per day. Three new bowl locations were introduced to the arena: a near-negative, middle, and near-positive location. None of these new, ambiguous locations were rewarded.

Each day included three trials with the bowl placed at one of the ambiguous locations, interspersed with six trials at the known positive and negative locations. Speed of approach indicated each tortoise’s expectations: faster approaches suggest optimism and slower ones suggest pessimism.

Then, within two weeks of completing the cognitive bias task, two anxiety tests were conducted:

• Novel object test: Each tortoise was placed in their familiar arena diagonally opposite from an unfamiliar bead coaster. The researchers recorded how long it took them to move and to approach the coaster, and measured how far they extended their head.
• Novel environment test: Floor and wall coverings with different textures, colors, and patterns were added to make the arena unfamiliar to the tortoises. Again, the researchers collected data on head extension and the time it took them to begin moving.

A tortoise’s level of confidence or anxiety can be determined by how far they extend their heads from their shells. However, a five-centimeter extension means something different for a small tortoise than a large one. Imaging software was used to account for the animals’ size differences, calibrating measurements relative to their shell size for more accurate comparisons between individuals.

Results And Limitations

All 15 tortoises were successfully trained and tested, though one was excluded from the novel object analysis after defecating in the arena (a possible stress response).

Analysis of approach times found significant overall differences in how quickly tortoises approached the five bowl locations:

• The slowest approaches (approximately 50 seconds) occurred at the negative location.
• Slightly faster approaches (around 35 seconds) occurred at the near-negative location.
• The fastest approaches (roughly 15 to 20 seconds) occurred at the positive, near-positive, and middle locations.

The researchers describe this as an optimistic bias. Rather than becoming more hesitant as the bowls got closer to the negative spot, the tortoises appeared to treat uncertain locations almost as favorably as the known reward zones. More optimistic individuals were also more confident in the anxiety tests: they extended their heads further and investigated the bead coaster quicker. These patterns, the study claims, suggest the tortoises were making decisions based on their underlying mood states.

Although the results offer compelling preliminary evidence of affective states in tortoises, the researchers acknowledge several limitations in their work. Only 15 individuals from one reptile species were tested, and the two anxiety tests didn’t correlate with one another — a tortoise’s behavior in one test didn’t predict how they would react in the other. This is a common construct validity issue for psychological studies attempting to measure abstract concepts like mood. It can be hard to tell if, for example, the behavior being measured is capturing anxiety or some other emotion.

Implications For Reptile Welfare

These findings require replication and larger sample pools before broad conclusions can be drawn. Still, they provide the first empirical evidence that reptiles can be tested using cognitive bias methods previously validated only in mammals and birds.

Current care standards for reptiles are almost entirely based on physical health. In countries like the U.S., reptiles are excluded from most federal animal welfare laws, leaving millions unprotected by frameworks that account for an animal’s mental well-being and enrichment. The researchers call for additional species to be assessed using their methods, such that habitats and care practices better reflect behavioral and emotional needs.

If reptiles can experience persistent affective states, even in limited contexts, then outdated assumptions about their emotional simplicity should be reconsidered. Advocates can use this research to support reform in how the welfare of these often overlooked animals is assessed, regulated, and understood.