Natural Scent Detection Capabilities in Dogs and Wolves

A red and white hound nose to the ground trailing a scent

Polgár Z, Kinnunen M, Újváry D, Miklósi Á, Gácsi M. A Test of Canine Olfactory Capacity: Comparing Various Dog Breeds and Wolves in a Natural Detection Task. PLoS One. 2016 May 6;11(5):e0154087. doi: 10.1371/journal.pone.0154087. PMID: 27152412; PMCID: PMC4859551.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4859551

Summary by: Charlie Clarke

Background

Domestic dogs and wolves rely on their noses for access to food, scenting each other, and generally navigate the world. These skills have been harnessed for a number of scenting tasks such as detection, search and rescue, and discrimination. Certain breeds have been bred specifically to excel at scenting. Trained bloodhounds were able to trail the scent of a human 48 hours old with a 96% success rate. These abilities are not limited to only bloodhounds, police organizations often use German Shepherds, Labradors, and Belgian Shepherds. This study sought to assess whether the scenting ability of dogs and wolves changes across breed. This is incredibly difficult to study because most odor detection success depends on the training. An experienced trainer may be able to achieve a higher success rate than an inexperienced trainer. Additionally some dogs have such a high drive for work that you can accidentally end up measuring the drive of the dog as opposed to their scenting capabilities. The researchers set up a measurement system that first would create a practical detection task that could assess olfaction of differently trained dogs and second would quantify relative differences in olfactory sensitivity by testing with decreasing amounts of odor. 

Olfaction Test Setup

First the dogs in the study were split up into 4 groups, scenting breeds, non-scenting breeds, short-nosed breeds, and wolves. The test was an outdoor setup to accommodate the wolves. Four ceramic pots, with a hole in the bottom were placed upside down on grass. A small plastic container with a lid was placed inside each pot. Three pots did not contain any odor while one pot had the raw meat bait placed in it. After 1 minute of placement of the meat, the dogs were indicated to start searching. Each dog was tested at all five difficulty levels and their indication responses were recorded. If the dog indicated correctly they were allowed to eat the meat within the pot. If they false alerted or failed to detect odor that was the recorded threshold for that dog. 

Results

In the natural detection test, all of the dogs and wolves were able to detect levels 1 through 4 better than chance. At level 5 non-scent breeds and short nosed breeds were not able to detect better than chance while scent breeds and wolves could detect better than chance. 

After running the test once, the dogs and wolves were tested again. The dogs did not improve their performance on the second test, while the wolves did statistically improve. Their average success rate increased from ~65 to 85%. The retested wolves improved mostly at the higher levels. 

Discussion/Summary

This natural detection test demonstrates that there is a difference in sniffing capacity based on breed groups. It will take more research to determine how significant these results are. The hardest level of this test showed that non-scenting breeds were slightly better than chance, brachycephalic breeds were even with random chance, and the scenting breeds were much more capable than random chance. This test was less reliable for the wolves as they were generally less motivated than the dogs and improved significantly with repetition. Interestingly, the dogs were not able to improve on their second pass, indicating that they were sufficiently motivated to accomplish the task. This protocol could be utilized to test puppies for detection work at different stages in their career, allowing the trainers to career change puppies that are underperforming earlier. This could also indicate that scenting breeds have a greater capacity for olfaction than we typically give them credit for, and that they may be better positioned for detection work despite being seen as more difficult to train. Future research investigating the intersection between trainability and olfaction sensitivity, could prove beneficial for organizations dedicated to training detection dogs as well as to pet owners interested in scent sports.

This work by the Functional Dog Collaborative is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.