Marine researcher Kelly Newman wanted to learn if transient killer whales were dining on the fur seals that abound in the waters off Alaska's Pribilof Islands. With limits on what her eyes could contribute to her research, Newman turned to her ears.Using a sophisticated hydrophone, the graduate student last summer captured the distinctive sound of orcas killing -- pouncing on seals or ramming them like freight trains, often at night, then calling in their companions to share in feeding.Newman says her work, so far just a pilot project, affirms that visual observations are limited in determining behavior of marine mammals, especially if the animals don't want to be seen."You may see a killer whale, and you may see a fur seal, and you may see something going and maybe think there might be predation, but you can't be sure," she said.Newman is pursuing a doctorate in marine biology at the University of Alaska Fairbanks. The waters off the coast of Alaska have been a focus for researchers because of changes brought on by warming, such as diminished sea ice, and population drops in pinnipeds such as fur seals and Stellar sea lions.Like dozens of other scientists focused on wildlife research in the vast North Pacific Ocean, Newman faces formidable physical challenges before she collects an iota of data for analysis.It costs $1,000 to fly round trip from Fairbanks to the Pribilofs, a group of four islands in the Bering Sea 300 miles from the mainland. St. Paul, the largest island, has a village of the same name with about 500 residents, and St. George Island, 47 miles to the south, has about 125. Otherwise, the most famous residents are the 210 species of birds that nest in the islands' cliffs and attract birders from around the world.The remote outposts also are the breeding grounds for more than half the world's northern fur seals. Males arrive at rookeries in late spring and stake out territories. Breeding females arrive in mid-June, give birth a few days later and begin nursing. The nursing period lasts through October, and mothers split their time between the shore with their lone pups and the sea, where they forage.A fully mature male weighs up to 600 pounds and a mature female weighs 90 to 110 pounds, but in the open sea neither is a match for killer whales, which can reach 27 feet and weigh 10 tons. Resident killer whales eat fish and generally are seen in certain areas for much of the year. Transient killer whales feed on marine mammals and roam over wider areas. Each pod within the groups appears to have its own unique, recognizable type of calls.The National Marine Fisheries Service declared the stock of northern fur seals to be depleted -- below its optimal sustainable population -- in 1988 and the Pribilof population has continued to decline. Between 1998 and 2004, estimated pup production declined 6.2 percent annually on St. Paul and 4.5 percent on St. George.The most recent estimate for the number of fur seals in the Eastern Pacific stock, which includes Bogoslof Island in the Aleutians, is 721,935, according to the National Marine Fisheries Service. Federal biologists estimate the worldwide population at 1.1 million, with the Pribilofs accounting for 55 percent, down from 74 percent in 1992.No one knows why the population is down, but killer whales are one of the suspects. So are limits on food that could be tied to global warming or commercial fishing, said Alan Springer, Newman's adviser, who has conducted bird and fur seal research for years in the Pribilofs.Newman's study is not aimed at answering the question of fur seal decline. Rather, her goal is to add to the body of knowledge by studying the habits and group composition of transient whales that visit the islands and the dynamics of killer whale predation.Her pursuit of a doctorate carries challenges not faced by English majors. She received $10,000 from the Alaska Sea Grant program; nearly half paid for the lease and training on an "autonomous acoustic recording device," otherwise known as a pop-up recorder, developed by the Cornell University Bioacoustics Research Program for use on the ocean floor. It captures marine sound, and when its battery runs low or its data disk is filled, an operator sends a signal to "burn" a cable holding it in place and the device pops to the surface.Marine research on the high seas commonly involves the use of big, expensive boats, Springer said. Acoustic documentation can provide 24-hour surveillance with less risk and expense.In July, Newman placed the device three miles offshore in an area where orcas would have access to seals departing from shore from four or five rookeries. A colleague retrieved the recorder 22 days later and returned it to Cornell, where the sound data was downloaded and sent to Newman in September.She didn't know what she had, but was pleasantly surprised."When I opened that first file and got sound right away, I said, 'This has just got to be luck,' " she said. "I was expecting to get nothing. We weren't sure it was going to work."Instead, she collected whale sounds on 19 of the 20 days the hydrophone recorded data. Soon she was listening to the cold, efficient killing of fur seals by orcas, commonly in pods of three to five animals."Lots of work has been done on transient killer whale behavior," she said. "The accepted theory is that they are very quiet before a hunt and will use a lot of vocalization and noise during the feeding event."The recordings belie the violence in the water. From silence, the sensitive recorder picks up a thump, a pause, another thump, then whale calls, sort of a high-pitch warble. It's the sound of a silent killer whale ramming a seal, flopping on it or even throwing it in the water, then a summons to eat.The black-and-white hunters were mostly quiet from early evening until sunset but active afterward, which may correspond to the time when seals leave rookeries to catch pollock, herring, capelin or squid."The most common call time to catch them was 2 a.m.," Newman said of the orcas.Nocturnal activity, she said, is essentially nonobservable by other techniques.In coming months, Newman will continue to review data and try to identify call types. She wants to connect calls to particular pods and correlate sound with visual observations recorded by residents. She wants to learn if the orcas are making calls for reasons other than feeding, and, when they're silent, whether they are hunting or simply in another area.Next summer, she hopes to collect more data, this time with several recorders.