Dogs Detecting Scat

Detection dogs are vital partners in conservation work, and many have been trained to find scat from specific species such as jaguars and ocelots. Their ability to sniff out scat assists researchers who want to estimate population levels, to track where various individuals of endangered species are living and to learn what they are eating. Advances in genetics allow these researchers to determine the sex of the animal who left the scat and even to link scat to specific individuals. Working dogs make it possible to learn this information in a non-invasive way that does not risk harming animals and that does not require actual sightings of these elusive species.

When dogs find the wrong kind of scat and alert their handlers as though it is the target scat, it costs researchers time and money because that sample will be processed, wasting research resources. (Luckily, it doesn’t ruin the research because DNA tests on the sample will reveal the error and it will be excluded from the study.)

Errors in which dogs alert to scat of the wrong species are usually assumed to be due to incomplete training of the dog or of the human handler. However, a new study (“How Behavior of Nontarget Species Affects Perceived Accuracy of Scat Detection Dog Surveys”)indicates that this blame may be misplaced. In fact, there are three possible explanations of the apparent mistakes made by dogs, and they relate to the ecological complexity of the environment and the behavior of various nontarget species. Other species are contaminating the scat. The result is that genetic analyses in the lab find that the DNA in the scat is not from the species being studied.

The genetic profile of the scat is changed by nontarget species in at least three ways. Species other than the one of interest may urinate on the scat of interest, causing confusion in the lab analysis of the genetics. Nontarget species may carry the target scat in their mouths, causing a similar contamination of the sample that makes it appear that the detection dog has made an error. A third way that contamination of the sample can occur due to natural behavior in the field is coprophagy. If an individual of another species eats the scat of the target species, the scat of the animal who consumed it will still contain the odors of the target species, which is why a detection dog will alert to it. Urinating on scat, eating it and holding it in the mouth to move it are all natural behaviors of a number of species of wild animals. It is especially common in species who use urine and feces to mark territories—behavior that is widespread in both canines and felines.

In a series of studies, researchers tested an experienced detection dog’s response to scat that was contaminated by urine, by saliva or by having been ingested by a different species. There were also tests of the dog’s response to scat with prey remains typical of the target species. They determined that dogs alerted to scat of the target species even when it had been urinated on by a nontarget species. They also found that genetic analyses of such scats sometimes misidentified the species that left the scat because of the urine on it. The dog also identified scat that came from a nontarget species that had ingested scat from the target species. Such samples were also identified in the lab as “mistakes” by the detection dog when in fact the olfactory profile the dog was trained to find was present. An observational study of coyotes allowed access to cougar scat revealed mouthing of samples by a coyote, a racoon and an opossum—and that can cause genetic contamination. Several coyotes and one fox were observed urinating on the scat samples, which can also change the genetic profile of the material.

Detection dogs are rewarded for finding the proper olfactory profile regardless of the genetic profile of the sample and regardless of whether other olfactory profiles are mixed in with the samples. The rate of scat samples that are not what researchers want despite correct identification of odor profiles by the dog vary with the environment. In Argentina where dogs are used to find jaguar scat, there are not a lot of other animals interacting with the scat, and “mistakes” are uncommon. However, studies of cougars in North America are often compromised by the presence of a large number of coyotes. Where domestic dogs are prevalent, studies are often hampered by high rates of urine marking on scat.

Even though some apparent errors by dogs may be due to contamination, that does not mean that there are no canine mistakes. It’s important to make sure that dogs and handlers are properly trained, as the increased use of detection dogs in conservation has unfortunately led to such high demand for dogs that there has been a lowering of standards and the use of some inadequately trained teams. Even with the best trained dogs and people, research projects can achieve more accurate results by collecting only the freshest scats, as those are less likely to have been contaminated.

It’s a challenge to determine if problematic samples are due to dog or handler error or to natural contamination of the samples in the field. This research shows that blaming the dog (or the handler) for any error is the real mistake.