BOSTON (AP) -- An ocean experiment that was accidentally conducted amid the shipping silence after Sept. 11 has shown the first link between underwater noise and stress in whales, researchers reported Wednesday.

The analysis indicated that a drop in a stress-related hormone found in the right whales was tied to a dip in ocean noise that followed a near-standstill in ship traffic, due to security concerns following the attacks.

The work indicates whales and other sea life that use sound to communicate and travel can be harmed by the noise. That could prompt more research and eventually influence future ocean traffic and development, said New England Aquarium scientist Rosalind Rolland, the report's lead author.

"This is definitely a very important piece in the puzzle that lends credence to the idea that, yes, we potentially have a problem out there and we need to learn a lot more about it," Rolland said.

The report combined data from two unrelated experiments in Canada's Bay of Fundy that happened to be occurring simultaneously. One involved acoustic recordings of right whales; the other the collection of whale feces samples, which contained stress-indicating hormones.

It wasn't until 2009 that Rolland realized the information existed for the analysis, published Wednesday in the British journal Proceedings of The Royal Society B.

"Here is the first solid piece of evidence that says there's a link between noise level and stress," said Christopher Clark, director of the bioacoustics research program at the Cornell Lab of Ornithology, who was not a paper co-author. Clark noted stress has long been tied to longevity, reproduction, disease and other key health indicators in whales.

There's no international standard for what ocean noise levels should be, and it's been tough to get at what kinds of problems it causes, Rolland said.

The use of military sonar at sea has been one source of tension between governments and conservationists, who claim that such sounds kill whales and other marine life.

The Bay of Fundy is bordered primarily by the Canadian provinces of New Brunswick and Nova Scotia. Rolland was there in September 2001, taking right whale fecal samples in the midst of a study on the health and reproduction of the endangered North Atlantic right whale.

She remembered getting word at the waterfront of the terror attacks, then seeing her crew in tears as they watched the coverage. It was a brilliant day, and after a while, the crew decided to go on with their work, as a measure of defiance and also because the bay was "calming for the soul," Rolland said.

"It's like our cathedral," she said. "It's a beautiful place."

That day and those following were like a primal ocean scene, Rolland said. "There was nobody out there except for us and the whales."

Around the same time, another researcher, Susan Parks, was getting acoustic recordings on mothers and their calves for research on the social behavior of the whales.

The data didn't come together until late 2009, when Rolland started researching stress and underwater noise to prepare for a workshop organized by the Office of Naval Research. She realized Parks had four days of sound recordings from the bay, two days before and two days after Sept. 11, and she had five years of data on stress hormone levels for the whales that included that time.

A hunch, and then quick analysis by Rolland, showed a possible correlation between a drop in sound and the drop in whale stress hormone levels. The naval office eventually agreed to fund the work that led to Wednesday's paper, she said.

The more rigorous analysis showed a significant decrease in background noise in the bay post-Sept. 11, including a drop in the low frequency sounds that ships emit and which the whales use to communicate.

Scientists compared the stress hormone levels found in the whale feces during the five-year period and found them to be markedly lower only during the time when ship traffic was down immediately after Sept. 11.

Rolland said caveats come with any accidental study. A planned study would have had more acoustic and hormone data. This study obviously can't be repeated. And it's also unclear how much chronic stress from noise the whales can take before the population is affected, largely because it's impossible to conduct controlled experiments on 50 ton animals.

But even with the caveats, Rolland said, "It's pretty good evidence. We have no other explanation for these findings."

© 2012 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed. Learn more about our Privacy Policy and Terms of Use.