If you have a question about coyote scat, you might want to give Laura Prugh a call. In fact, if you have a crime scene involving a coyote -- a raided chicken coop, perhaps -- hand Prugh a well-preserved poop from the purported perpetrator and she can probably tell you which coyote did the deed.

Laura Prugh holds the evidence. Back in the lab, she can determine the original owner of this scat. (photo: Wendy Arjo)

Over the last few years, the researcher at the University of British Columbia has spent countless hours analyzing more than 1,000 coyote scats, and become expert at genotyping their DNA.

Crime investigators have been using this technique for some time, analyzing blood and hair samples for their genetic fingerprints, and some wildlife researchers now use hair samples to determine animal genotypes. But feces were not considered ideal for DNA work as they can degrade before they are analyzed.

Prugh did not set out to become an expert in this rarefied field. For her doctoral project in wildlife ecology, she wanted to figure out what coyotes in the Alaska Range were eating when the numbers of snowshoe hares -- their preferred prey -- were down. (See yourYukon #366)

She needed to know the number of coyotes in her large research area, and found that traditional research methods were not working well. "The genetics work started as a side project," she explained. "It is a pretty new technique, and there was a high chance that it would fail, so I did not want to rely on it totally."

"The main reason I did it is that coyotes are so difficult to count; they are probably about the cleverest animal out there."

To study a population of animals in a particular area, biologists typically count the number of tracks intersecting a set trail. Smaller animals, with smaller home ranges, can be live-trapped to estimate population size.

But with wide-ranging, trap-wary coyotes, these techniques don't work. For three winters, Prugh and an assistant patrolled 75 kilometres of snowmobile trail in the northern reaches of the Alaska Range. Counting the tracks intersecting the route helped indicate whether the number of coyotes was increasing or decreasing, but did not show the total number in the area.

"I was a little sceptical of the track counts as even the snow depth can affect how coyotes use the trails. The assumption that they will use the trails every year does not hold up," she said. "One winter when there was little snow, the coyotes abandoned the trails and trotted across the bare south-facing slopes. So the track counts were lower even though the number of coyotes was not."

While patrolling the trails, Prugh and her assistant regularly collected coyote scats as they planned to analyze them for dietary information. Other researchers had tried and failed to genotype feces, but -- after hearing of one successful project on coyote feces in the Los Angeles area -- Prugh decided to give the procedure a try.

As no commercial labs do this work, she learned what she could from other researchers and then spent several months refining her lab techniques -- learning to isolate and replicate the coyote DNA contained in the feces. While such work might sound unnervingly high tech and sophisticated, Prugh said much of the procedure is "basically like following a recipe."

Even a scat can be mined for its genetic fingerprint. (photo: Laura Prugh)

Once the DNA has been isolated, a contraption called a thermocycler does most of the work. It uses varying temperatures to set up a polymerase chain reaction in which the DNA is copied over and over again. "You put in the DNA and then a bunch of extra amino acids and enzymes which unzip the DNA helix -- rip it apart -- and then new chemicals bind to it and create new double helixes," she explained. "Technically you could do the same thing on a stove top."

When taking DNA from feces, error rates tend to be higher than with blood or tissue samples, so Prugh sampled each feces five times for its genetic fingerprint, wanting to ensure that her results were reliable.

"Even a low level of errors can lead to inflated population estimates because whenever you have an error you create a new individual, so an error rate as low as 5 percent can double a population estimate," she said.

This pilot project convinced her that she was getting accurate results. After securing funding to genotype all of the 1,237 scats in her collection, she determined that they had been deposited by 56 individual coyotes in the study area. Only 16 of them had been captured during her project; they were darted with tranquilizers from a helicopter.

These coyotes were fitted with radio collars so that their movements could be monitored from a plane equipped with a radio transceiver. Prugh says the extensive lab work required for DNA genotyping can be less expensive than air time, particularly in remote areas such as the one in which she was working.

Also the scats reveal much more than the total number of animals. They show what the animal has been eating, and the DNA analysis can determine family groups and who is breeding with whom.

"You can do all sorts of really neat things with the genetic data besides population estimates. For example, you can look at how genetically different populations are," she said.

As a side project, she also tested how accurate she and her assistant were in identifying the scats. While the classic coyote scat -- with its twisted form and tapering end -- is easily identifiable, more irregular ones can be confused with scat from wolves, lynx and foxes.

She tested not only their ability to identify the scat, but also how certain they were of the identification. To her surprise, she found that their identification skills were almost startlingly accurate.

"To me our certainty levels matched so well with the genetic work that it almost looked like fake data. I was really pleased that our ability to judge was so accurate."

For more information contact Laura Prugh at prugh@zoology.ubc.ca.