Shelter Dog Breed Identification: A High-Stakes Game

Any way that you look at it, the national animal shelter intake numbers are staggering. Each year, about 4 million dogs enter the shelter system in the U.S., with about 53-77% of those being strays. Dogs entering the system as strays have an unknown history and breed heritage, and short of using DNA testing, which is expensive and time consuming, breed ID is often left to shelter staff. Shelter workers use a variety of imperfect criteria to assign a breed to a particular dog, everything from weight, coat color, temperament, and more.

The accuracy of shelter worker breed assignments is all over the map, though controlled studies have tended to find it to be unreliable at best. One significant study found that shelter staff could match one breed within a dog’s profile in only seven of 20 subjects. Another study using videos for identification found that dog professionals’ ability to identify one breed in a dog’s DNA was, on average, fewer than 30%. For dogs like pit bulls, or those that fit a pit bull-type visual profile, who face breed-specific legal restrictions, this type of (mis)identification could be a matter of adoption or euthanasia, life or death.

The purpose of this study in particular was twofold: first, the authors wanted to “report the breed heritage of a large sample of mixed breed shelter dogs based on genomic breed testing.” Secondly, they wanted to measure just how much visual breed identification by shelter assess agreement of visual breed identification by shelter staff at one of these locations by comparing the primary and secondary breeds indicated by staff and those identified by DNA analysis. The researchers performed separate visual identification and DNA tests of over 450 dogs each at two different shelters to arrive at their results.

In these nearly 1,000 cases, DNA testing found only 4.9% of the dogs were identified as being purebred, standing in sharp contrast to the approximately 50% of dogs living in households who are purebred (or at least thought to be). When it came to staff identification of the breed, the staff assignments agreed with the most prevalent breed in the DNA tests 56.7% of the time. However, when it came to matching both primary and secondary breed, staff only matched on 10.4% of dogs. Perhaps most concerning, however, is that “33.3% of visual breed assignments by staff didn’t match any of the up to eight breeds indicated in the analysis.”

Why is this study significant and why should companion animal advocates pay attention? First of all, according to the authors’ knowledge, “this is the largest reporting of breed heritage in sheltering to-date.” That along is enough to pay attention. However, the greater results should give everyone pause. In addition to the above, this study found that pit bull-type dogs waited longer to be adopted than other dogs: in one of the shelters studied, the difference was 17.8 days, and in the other it was 23.2 days – almost double the time of other breeds.

Knowing that visual identification can be such an inaccurate process, shelters and other companion animal advocates may want to consider other ways of identifying the dogs in their care. Additionally, the study suggests that shelters could move away from the emphasis on breed in general, and focus instead on “communicating the morphology and behavior of the dogs in their care to best support matchmaking and adoption efforts.”