ELMA, Wash. -- Everything was going as planned, until elk began to die.
A helicopter swooped in over the herd, and a marksman leaned out, firing tranquilizer darts at elk fleeing across the meadow. Animals hit with darts soon began to weave and stumble, watched by workers waiting to move in once they dropped.
The capture, conducted in March to relocate wild elk from a burgeoning population in western Washington to a failing herd 80 miles away, looked as if it would go without a hitch, wildlife officials recall.
Then one of the tranquilized elk bolted into a rain-swollen creek. As narcotics coursed through its body, the normally strong swimmer flailed about, drowning as its head slipped beneath the rushing water.
A second elk ran into the woods, staggering as the drugs took effect. Pitching down, the animal buried its nostrils into soft earth and lay there, unable to move, until it suffocated.
Two other elk collapsed in the open, but by the time they were blindfolded and strapped onto pallets for transport, they were having trouble breathing. Their systems were shutting down from the trauma of the chase and the drugs, and they died within minutes.
It's a side of wildlife research and management seldom seen by the public: A surprising number of wild creatures are inadvertently killed or injured by the scientists entrusted to protect them.
Wildlife biologists do more these days than venture into nature with notepads and binoculars. Animals are chased, darted, netted, drugged, tagged, banded and radio-collared. They are cut open and implanted with satellite transmitters. They are caught in leghold traps, then released and caught again.
Decisions about how much death, injury and harassment to allow in the pursuit of science are left largely to the wildlife managers and researchers themselves.
"Anytime we lose an animal, it's not a good day," said Jack Smith, who ran the elk capture for the Washington Department of Fish and Wildlife. "But if you don't go into an operation expecting that you're going to lose a few animals, then you're not being properly realistic."
Marc Bekoff disagrees. An animal-behavior expert who has written books on the intelligence and emotional life of animals, he believes researchers have obligations to the animals they study.
"The methods you use are set up by the philosophy you have. If you think it's OK to let individual animals die, it can be a self-fulfilling prophecy."
Wildlife researchers often underestimate the effect they have on their subjects, said Bekoff, who teaches biology at the University of Colorado.
"Wild animals are living on the edge," he said. "They're trying to get food and water, they're trying to survive. They can't tolerate stress."
Wildlife researchers say some mortality is inevitable.
The day that four elk died in Washington, Smith's team successfully captured 20 others, for a mortality rate of 17 percent. Overall, the five-day relocation effort killed eight elk but successfully moved 81, for a 9 percent mortality rate.
"We would definitely do it again," Smith said. "We were within what we consider acceptable limits."
John Pierce, chief scientist for the wildlife division of the Washington Department of Fish and Wildlife, said researchers have grown more concerned about individual animals during the past 30 years.
"In this day and age, it's a given that animal-welfare concerns are a major part of our plans," Pierce said. "It's a matter of ongoing attention, discussion and improvement."