By Sarah DeWeerdt
Three diagonal lines on Samuel Wasser's computer screen may help explain why orcas, or killer whales, are disappearing from Puget Sound.
The lines track how the whales' thyroid hormone levels decline over the course of a summer, as Chinook salmon become scarcer and the whales begin to go hungry. In lean years, some whales are pushed past the edge of survival. "The results have been really powerful already," after just three field seasons of study, says Wasser, head of the Center for Conservation Biology (CCB) at the UW. "I think we've shown that without question there's a very strong nutritional signal."
The study, helmed by Wasser's graduate student Katherine Ayres, features an unusual research assistant: a black Labrador retriever mix named Tucker is helping the team collect samples of whale feces, or scat. Tucker is one of 30 dogs in the CCB's conservation canine program helping sniff out scat from vulnerable species in far-flung corners of the globe as well as Puget Sound.
To find orca scat in the water, Tucker stands at the bow of the boat and sniffs the wind. When he catches a whiff, he becomes animated, straining over the edge of the boat. Then, the boat tacks back and forth as dog and humans play a collaborative, interspecies game of getting-warmer and getting-colder, racing to spot the sample and fish it from the water before it sinks. It's a complicated dance—"but it works fantastically," Wasser says. The dogs can find up to four times as many samples as humans can by following their own noses.
Wasser pioneered the use of physiological markers in scat as a UW graduate student, investigating reproductive cycles among baboons in Tanzania. Scat contains hormones, immune markers and toxins that provide a picture of an animal's health status. It holds traces of DNA that enable researchers to determine the animal's species, sex and even individual identity. Unlike a blood sample, scat can be collected without disturbing (or, in some cases, even seeing) the wild animal it came from. And scat is pretty much ubiquitous in the environment. "It's really quite an amazing tool," Wasser says.
Wasser's mountain-man beard hides rounded features—in his mid-50s, he's still baby-faced. He has a soft voice and, at first, an almost understated way of speaking. But get him talking about whale poop and he lights up. He is, in canine terms, keen.
After receiving his Ph.D. from the UW in 1981, Wasser continued to develop and apply scat-based physiological measures, first as a Smithsonian Institution researcher and later with a joint appointment at the Woodland Park Zoo and in the Department of Obstetrics and Gynecology at the UW School of Medicine. In 2001, he was awarded an endowed professorship in conservation biology and launched the Center for Conservation Biology, which aims to use non-invasive techniques like scat analysis to answer major conservation questions.
During the mid-'90s, as Wasser was getting more involved in conservation questions, he was also mulling over the problem of sampling bias in his studies. Among many species, for example, males use scat to mark their territory, leaving deposits on trails where they're easy for human researchers to find. Lactating females, by contrast, often hide their scat. How could Wasser ensure a representative sampling of, well, samples?
That's how the dogs got involved. "It came to me that if we could train narcotics dogs to instead find scat, that would be the perfect thing," Wasser recalls.
Good conservation dogs, like good narcotics dogs, are ones with a high play drive—"they are just insane about a tennis ball," as Wasser puts it. That quality makes these pups hard to keep as pets—the dogs in Wasser's program are all pound rescues—but it also means they work hard if they know they'll get to chase their ball around at the end. "So the beauty of this is that when these dogs are out in the wilderness looking for scat, they don't really care about the scat," whether it comes from a male or a female, an adult or a juvenile, or is hidden or out in the open. "They care about the ball," and will do whatever they can to find the scat and earn their reward.
Wasser first put the dogs to the test in a study of grizzly and black bears in Alberta, Canada. Studies of kit foxes in the Mojave Desert, weasel-like Pacific fishers in California and giant armadillos in Brazil, among other species, followed—all resounding successes. A single dog can be trained to find scat from up to 18 different wildlife species, and can smell a sample up to half a kilometer away on land and one nautical mile away on the water. Recently, four dogs worked in the tar sands of northern Alberta for 10 weeks and found 1,600 samples of caribou, moose, and wolf scat—all under at least two feet of snow. This summer, dogs will journey to Cambodia for a study of tigers, and Indonesia and Malaysia for a study of rhinoceroses.
When they're not working, the dogs live at a new state-of-the-art kennel facility in the UW's Pack Experimental Forest, with 4,300 acres of wilderness nearby for training, exercising and keeping the animals in top mental form. At the Pack Forest facility, Wasser's group is also working on teaching the dogs a new trick: distinguishing scat from different individual animals. Wasser got the idea for this project from a police-dog program in the Netherlands, in which dogs matched criminals to evidence from crime scenes. He thought that if dogs could distinguish individual humans by smell, perhaps they could also tell apart individual bears, caribou and so on. "Our dogs absolutely can do it," Wasser reports. The approach could help stretch conservation dollars by minimizing the number of samples that need to be subjected to expensive DNA analysis.
Meanwhile, Wasser and Ayres—along with Tucker—wait for the orcas to converge on their summer grounds in Puget Sound again. This year, Wasser wants to gain more insight into how lack of food interacts with stress from boat traffic and environmental toxins to put the whales at risk. To do that, he hopes to get Tucker out on the water for the entire field season. "If you're trying to partition these different factors," he says, "you really need a lot of samples."