Hungry scent detection dogs may be more prone to false alerts
Bottom line
Dogs trained for scent detection may be more likely to give false alerts when they're hungry, according to new research from the University of Waikato and Adelaide University. In the study, published in Learning and Motivation, six dogs completed scent-detection sessions before or after breakfast using an automated system. Their hit rate stayed high in both conditions, at about 90%, but false alarms increased when the dogs worked before eating, suggesting hunger affected decision-making more than odor detection itself. A related paper in Behavioural Processes found the dogs that hadn't eaten were more likely to whine and focus on the food dispenser, though they still stayed engaged with the task. (adelaide.edu.au)
Why it matters: For veterinary professionals working with detection, conservation, biosecurity, or sport dogs, the findings point to feeding schedule as a potentially modifiable performance variable. The researchers said the data don't support simply feeding every dog immediately before work, but they do suggest that meal timing, reinforcement strategy, and the cost of false positives should be considered together, especially in jobs where an incorrect alert can carry operational or clinical consequences. The study was small and laboratory-based, with six dogs and controlled odor mixtures, so translation to field settings will need more validation. (adelaide.edu.au)
What to watch: Whether follow-up studies in operational detection dogs confirm that adjusting pre-work feeding routines can reduce false positives without undermining motivation or stamina. (adelaide.edu.au)
Key facts
- Study type
- Laboratory scent-detection study
- Researchers
- University of Waikato and Adelaide University
- Sample size
- Six dogs
- Task
- Identify a target mixture of four odors
- Design
- Sessions before or after breakfast in an alternating design
- Main finding
- Hit rate stayed around 90%, but false alarms increased before breakfast
- Related paper
- Behavioural Processes
- Limitation
- Small sample and laboratory-based
A new study suggests that hungry scent detection dogs may be more prone to "cry wolf." Researchers from the University of Waikato and Adelaide University reported that dogs working for food rewards were significantly more likely to falsely indicate non-target odors when tested before breakfast than after eating, even though their ability to correctly identify the target scent remained high in both conditions. The work was published in Learning and Motivation in 2026, with a companion behavior paper published in Behavioural Processes. (adelaide.edu.au)
The study builds on earlier work often described as the "breakfast effect," which has suggested that dogs can perform some search tasks more accurately when they are less hungry. In this newer experiment, the investigators tested a more specific prediction: that hunger might not sharpen olfaction so much as shift response bias, making dogs more likely to treat similar smells as targets when food is more motivating. That distinction matters because it reframes the issue from sensory capacity to decision-making under reinforcement. (sciencedirect.com)
According to the university announcement and the journal summary, six dogs were trained to identify a target mixture of four odors using an automated scent-detection system. The dogs completed sessions either before or after their morning meal in an alternating design. Average target detection stayed around 90% in both conditions, but the dogs were less able to reject incorrect scents when hungry, producing more false alarms before breakfast. A separate analysis of the same cohort found that pre-breakfast dogs were more likely to whine and attend to the food dispenser, consistent with greater frustration or food focus, although they remained engaged in the task. (adelaide.edu.au)
Additional background from University of Waikato materials suggests the broader research program uses tightly controlled, automated scent-delivery methods designed to reduce unintentional cueing and standardize response collection. In related project documentation, dogs were screened for food motivation and familiarity with the laboratory setting, and the apparatus combined odor presentation with touchscreen responses and automated feeders. That controlled setup strengthens the internal validity of the findings, while also underscoring that operational conditions in field detection work are more complex than a lab task. (researchcommons.waikato.ac.nz)
The researchers' public comments were measured. Associate Professor Tim Edwards said hunger appeared to change "the likelihood of a dog saying a target scent is present," rather than simply making dogs better or worse at smelling, while co-author Dr Jade Fountain said false alarms can matter as much as misses in roles where people depend on dogs to distinguish one scent from another. The team also emphasized that the findings shouldn't be reduced to a simple recommendation to feed every working dog right before a job, because feeding strategy still has to account for workload, training method, individual variation, and reward structure. (adelaide.edu.au)
Why it matters: For veterinarians, behavior specialists, and professionals supporting working dogs, this study adds to the growing evidence that performance is shaped by physiology, motivation, and task design, not just olfactory ability. In practical terms, that could affect how teams think about pre-shift feeding, treat value, reinforcement schedules, and interpretation of borderline alerts. In disease detection, explosives screening, conservation work, or competitive scent sports, a false positive may trigger unnecessary follow-up, wasted resources, or missed trust in the dog's signal. The findings may also be relevant when advising pet parents involved in nose work or scent-based enrichment, particularly if they're using food rewards and noticing impulsive or overly broad indications. (adelaide.edu.au)
There are important limits. The sample size was small, the work was conducted in a laboratory, and the task involved trained responses to controlled odor mixtures rather than real-world search environments. That means the study is best read as evidence of a meaningful behavioral mechanism, not yet as a field-ready protocol change. Still, because feeding time is a manageable variable, the paper gives trainers, veterinarians, and working-dog programs a concrete factor to evaluate alongside health, stress, and reinforcement plans. (sciencedirect.com)
What to watch: The next step is likely operational research in larger cohorts of working dogs to test whether meal timing changes false-positive rates in real deployments, and whether any benefit varies by task type, reward system, or individual dog. (adelaide.edu.au)