Federal marine regulators said the cause of a stranding of up to 200 melon-headed whales off Kaua'i during July 2004 Navy war games may never be unequivocally determined, but sonar use is a "plausible, if not likely, contributing factor."

The mass stranding of such whales was the largest ever recorded in Hawaiian waters.

The findings come at a time of increasing evidence that sonar can harm whales and other sea life, and a planned repeat this summer of the Rim of the Pacific war games.

The Navy for the first time in March applied for a federal permit for the "harassment" of marine mammals during Rimpac war games scheduled for about five weeks, beginning June 26. The move comes in response to legal and environmental pressures, but also with the Navy acknowledgement that data about possible sonar impact is increasing.

The report released yesterday by the National Oceanic and Atmospheric Administration's Fisheries Service was tempered by a caveat that there are many unknowns with marine mammal acoustics. Still, the report details the one in a growing list of strandings over the past 10 years that NOAA Fisheries said coincide with military mid-frequency sonar use.

Asked if the Hanalei Bay stranding leans less or more toward Navy sonar as the plausible cause, Brandon Southall, NOAA Fisheries acoustics program director, said, "It's plausible and likely, if not probable."

The agency said there were no obvious significant weather or oceanographic events, harmful algal blooms, or predator or prey events that could explain the animals' behavior.

Michael Jasny, a senior consultant with the Natural Resources Defense Council, said the "verbal gymnastics" used by NOAA Fisheries to explain the cause are at odds with the direction of the report. "The report points a finger right at sonar," Jasny said.

In October, the environmental organization sued the Navy, saying, "there is no dispute that the Navy's use of mid-frequency active sonar can kill, injure and disturb many species, including marine mammals."

Among the council's complaints was that the Navy regularly failed to comply with federal environmental law, including its failure to obtain for exercises like Rimpac the "incidental harassment authorizations" that are required by the Marine Mammal Protection Act.

Lt. William Marks, a Navy spokesman, yesterday said the Hanalei Bay stranding was not "conclusively" linked to sonar, and there is no evidence of sonar causing harm to any marine mammals in nearly four decades of biennial Rimpac exercises.

Marks also noted there was more than six hours between the last use of sonar on July 2, 2004, and the whales' arrival in the bay on July 3.

"Judging from those facts, and the conclusion from NOAA that it is not scientifically supportable that sonar was the exclusive cause of the stranding, it is extremely unlikely that sonar was the cause," Marks said.

Between 150 and 200 deep-water melon-headed whales congregated in the shallow waters of Hanalei Bay for more than 28 hours in what scientists described as odd behavior.

Attendees of a canoe blessing saw the whales enter the bay in a single wave formation at 7 a.m. on July 3. The whales were herded out to sea on July 4.

One animal, a calf, was found dead in the bay on July 5. A necropsy was performed, and although cause of death could not be definitively determined, NOAA Fisheries said maternal separation, poor nutritional condition, and dehydration contributed to the death.

Six Japanese and U.S. vessels transiting to the Pacific Missile Range Facility on July 2 intermittently transmitted active sonar over nine hours as they approached from the south.

Animals to the south and east of Kaua'i could have detected the active sonar on July 2 and reached Hanalei Bay before 7 a.m. on July 3, NOAA Fisheries said.

"However, data limitations regarding the position of the whales prior to their arrival in the bay, the magnitude of sonar exposure, behavioral responses of melon-headed whales to acoustic stimuli, and other possible relevant factors preclude a conclusive finding regarding the role of sonar in triggering this event," the report said.

A study published in 2003 in the journal Nature found that high-powered Navy sonar may give whales and other marine mammals a form of the bends, or decompression sickness, with bubbles forming in organ tissue.

Strandings that NOAA Fisheries said were "most likely" caused by sonar occurred in Greece in 1996, the Bahamas in 2000, Madeira in 2000 and Canary Islands in 2002.

Preventive measures are planned for Rimpac including a requirement that when marine mammals are detected within 1,000 meters of sonar, a ship will limit transmission levels to at least 6 decibels below normal operating levels.

Within 500 meters, sonar would be limited to 10 decibels below typical operation, and within 200 meters, it would be turned off, she said.

Jasny said the Navy can do more, and countries such as Australia turn off active sonar when marine mammals are detected within 4 kilometers of a ship's sonar.

Reach William Cole at wcole@honoluluadvertiser.com.