A few months ago I read an article about the Miami Heat basketball organization using Covid sniffing dogs to admit fans into the arena and was blown away. They can't actually detect the virus, but they can pick up the chemical differences in the composition of breaths exhaled between healthy and infected individuals (Dorman 2021). I've always heard about dogs being used to detect drugs and track fugitives, but the ability to detect a virus by sniffing a person's breath is just on a whole other level. I started thinking about possible applications and the idea of using dogs to detect pollutants in the environment came across my mind.
While researching the capabilities of these sniffing dogs, I searched for any examples or projects that involved using sniffing dogs as pollutant detectors and I came across an EPA proof of concept from 2003. The idea was to train sniffing dogs to be able to detect various environmental contaminants that range from house molds to illegal pesticides (NERLScience 2003). This article was very compelling, but upon further research the idea never came to fruition. In the coming paragraphs, I'm going to discuss the potential pros and cons of using these dogs as surveyors for toxins and why I think this should be incorporated into ecotoxicological efforts.
What makes pollutant sniffing dogs a better option for detecting contaminants than traditional sampling methods? Sniffing dogs can save time and money in the field. Imagine a scenario where your goal is to pinpoint the origin of some point-source pollution in a large swathe of land. Sampling the entire area is expensive and time consuming. You decide to divide it into plots and sample random locations until you get closer and closer to the origin of the pollutant and you eventually find it. How would this situation look if you had used pollutant sniffing dogs? You could release the dog(s) into the area and, similar to a blood hound sniffing out a perp, let them sample continuously until they reach the source of the pollutant. In Australia, sniffing dogs have been used to seek out and detect leeks in natural gas pipelines (NERLScience 2003). Additionally, the incorporation of these dogs into ecotoxicological efforts isn't very daunting. The infrastructure and knowledge of training these dogs is already there, it's just a matter of training them to detect the desired contaminant. Also, these dogs are able to be trained to detect a new chemical/toxin in at most 2 weeks, presenting the opportunity for one dog to be swift and dynamic in its use in the field. Finally, and most importantly, working with these sniffing dogs would make an ecotoxicologist's job 1000% more awesome.
There have been concerns in the efficacy of these dogs in their current field of use in the police force. In one study on the accuracy of these dogs in the field, researchers found that the dogs were accurate 87.7% of the time, but gave a false positive in 5.3% of trials, and in 7% of trials they failed to find the target substance (Jezierski et al. 2014). While these numbers aren't perfect, I think using these dogs at a minimum as an initial sampling option would be beneficial. Additionally, using multiple dogs decreases those chances of false positives and negative responses significantly. The same study references the fact that there can be biases imprinted into the dogs from the trainer through training. Essentially, if the dogs aren't trained correctly, then they will not perform correctly. This would certainly be an issue going forward, but not one that is unsolvable. Developing specific training regimes for pollution sniffing dogs would be key in ensuring successful and accurate results later in the field.
Overall, I think that these dogs are a valuable asset that is essentially untapped in the US. Implementing them in the field would yield quicker and cheaper results. There seem to be multiple applications for these animals that are not yet mainstream. I think that one main driver for the lack of use of sniffing dogs and other animals for various tasks like detecting diseases and other important jobs is the hesitance in accepting results that are derived from an animal. For instance, scientists trained pigeons to read CT scans of the human brain and detect brain cancer. When compared to actual doctors, the pigeons ended up being more accurate. Yet, a large majority of people would be very unlikely to let a pigeon determine the results of their CT scan. What do you think? If you think these animals would be a great addition to the ecotoxicolgy field, what do you think hinders their integration?
Dorman D. 2021. Are COVID-Sniffing Dogs the New Tool in Helping Detect the Virus? https://cvm.ncsu.edu/are-covid-sniffing-dogs-the-new-tool-in-helping-detect-the-virus/
NERLScience. 2003. Pollution detecting dogs: Proof of concept. https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=56111&Lab=NERL
Jezierski T. et al. 2014. Efficacy of drug detection by fully-trained police dogs caries by breed, training level, type of drug search and search environment. Forensic Sci Int. 237:112-8.
s://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=56111&Lab=NERL#:~:text=Dogs%20are%20regularly%20used%20in,to%20detect%20termites%20in%20houses.
https://www.glendaleca.gov/government/departments/police-department/k9-unit/faq#:~:text=How%20much%20do%20the%20dogs,and%20detection%20is%20another%20%2411%2C000.
This is such an interesting post, and something I have never considered. I wonder- are there any potential health risks for the dogs made to inhale contaminants? I am imagining that rounds of training would frequently expose the animal to chemicals, as well as when they are preforming the actual environmental surveys. I was also curious as to what percentage is still detectable by canines- would very weak concentrations have lower success rates?
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteI love this idea! I feel like most people don't have faith in a living organism being able to detect anything with certainty which is one of the reasons that bioindicators are not as widely used. A lot of people don't even trust their doctor to give them correct test results. Also with a dog, the question becomes will the dog always tell that they sense a chemical? I think this problem could be easily solved with training though. I am curious about how many chemicals can one dog be trained to detect and how long does it actually take for a dog to complete a search. Also, if no dog detects a chemical, should samples from the site still be taken?
DeleteNote: I did not mean to delete my comment, but I realized that I few spelling errors that I needed to fix.
This is a super cool concept. I do wonder how this would work given there are so often a large combination of contaminants in a given system. I would be curious to know how many contaminants one dog could detect and if the presence of other contaminants could throw their detection off. Could you possibly train dogs to detect common combinations of contaminants? This would be an interesting concept, I wonder what the costs are to train each dog as well.
ReplyDelete