deep diving

Back in March 2000, the U.S. Navy was conducting a military sonar testing trial in the Bahamas. The tests coincided with a mass stranding of beaked whales -- 17 in all, 7 of which were dead, as a result of hypothermia -- on the shoreline. Coincidentally, a marine biologist named Kenneth Balkan, who'd formerly done acoustics research for the Navy himself, lived nearby, and was able to collect not just video footage, photographs and other documentation of the event, but also actual physical specimens from the dead whales. He sent the heads to the National Fisheries Service, which performed necropsies on the animals, and concluded that the dead whales had experienced acoustically-induced hemorrhages (fat emboli) around the ears. The scientists reasoned that the resulting disorientation may have led to the mass stranding.

What was surprising was that this sort of injury is indicative of decompression sickness -- what deep sea divers call "the bends," usually resulting when they surface too quickly. It's surprising to see this in beaked whales, because, well, they're pretty adapted to a deep sea environment. (Recent studies by researchers at the University of California, Santa Cruz, have suggested that such animals are protected during deep dives by elevated levels of oxygen-carrying proteins in their brains (globins), which protects them from hypoxia.) There have been other strandings before and since the Bahamas event all over the world, many appearing to coincide with military sonar tests as well. This has caused scientists and environmentalists alike to wonder if it's the mid-frequency sonar that's causing the strandings, either directly or indirectly. After all, sonar pulses have been used by whalers in the past to startle their prey and cause them to rise to the surface so they could be more easily harpooned.

Marine researchers have suspected since the mid 1990s that mid-frequency active sonar might be connected to the various strandings, but the 2000 Bahamas event was the first time scientists had been able to fully document a stranding event sufficiently to rule out any other possible cause. The evidence was so convincing that the Navy accepted fault for the incident in its official report -- although it also argues that this was a rare confluence of factors, and that those unique conditions don't apply to every sonar testing exercise. And there's the rub: environmental groups, bolstered by the growing body of scientific studies on this issue, don't think it's nearly such a rare confluence as the Navy suggests, and the tension between these two groups has given rise to some fairly contentious, high-profile lawsuits over the years. 

Fast forward almost eight years later, and the scientific debate is still going on. So are the lawsuits, mostly brought by the National Resource Defense Council (NRDC) and other interested parties. These are aimed not necessarily to stop military sonar testing altogether -- the group maintains they support the need to train sonar operators by conducting the tests -- but to enforce stricter mitigation measures, similar to those already in use by the oil and gas exploration industries, for example, which conducts high-energy sonic surveys to detect oil and gas deposits beneath the ocean floor.

From the Navy's perspective, sonar training exercises are critical to US national security interests: smaller diesel submarines can't be detected with passive sonar, for example, and over 40 nations deploy these kinds of subs. Ultimately, it's about balancing the interests of national security with environmental concerns. That makes it one of those complex issues about which reasonable people might reasonably disagree.

Legal stuff is always complicated -- just keeping track of the various filings, different cases, and decisions and appeals is practically a full-time job. I won't attempt to include an exhaustive list, but here's a handy recap of what's been going on with the most recent lawsuit, focusing on the Southern California region. It concerns a ban on the use of mid-frequency sonar in Navy training missions off the Southern California coast, until better safeguards are adopted to protect the animals. (Definition of terms, so we know what we're talking about here: Mid-frequency is in the 1 to 10 kHz range, and the systems used by the Navy are "pings" lasting as long as 103 seconds, at levels of around 235 decibels. The NRDC successfully sued to ban the Navy's use of low-frequency active sonar, in the 250-300 Hz range, in certain areas several years ago.) The California lawsuit was filed in March 2007 (the California Coastal Commission joined the NRDC in the suit), and US District Judge Florence-Marie Cooper issued a temporary restraining order against the Navy for lack of compliance with federal environmental laws -- based on the Navy's own estimates that the exercises would temporarily disrupt or overtly harm some 170,000 marine mammals.

The Navy appealed, and won a reprieve from the 9th Circuit Court, but that victory was short-lived. In November, a US appeals court (a second panel of three judges) upheld the ban, although the panel didn't buy the need for a broad absolute injunction for two years to avoid irreparable harm to the environment. So the judges  instructed Cooper to craft a "narrowly tailored" mitigation plan to meet legal environmental requirements. Cooper issued her decision yesterday.

The NRDC is asking that the training range be restricted, and for a safety zone of 1-1/3 miles where no sonar would be allowed whenever a marine mammal is in range. Geographic avoidance is one of the most effective mitigation strategies, so the NRDC is asking the Navy to avoid areas with a high abundance of sensitive species present for breeding, feeding, or other activities critical to their survival. For instance, the Channel Islands house a marine sanctuary that is home to some 36 species of marine mammals, five of which are on the endangered species list. And during the summers in Southern California, there's a large population of blue whales in the coastal waters.

The Navy, in contrast, wants a safety zone of roughly two-thirds of a mile. Navy spokespersons counter-argue that the agency already has some 29 procedures in place to avoid harming marine mammals, and adding even more would cripple the effectiveness of their training exercises (designed to train young sailors how to detect submarines), forcing them to shut down most of its Southern California range. It would also seriously undermine efforts to establish a similar sonar training range off the coast of North Carolina --a move that is already the subject of much opposition from environmental groups and a handful of elected officials.

Yesterday, Cooper sided with the NRDC. Again. Her ruling severely limits the Navy's use of mid-frequency sonar in its Southern California training range, banning the use of sonar within 12 nautical miles of the coast, and expanding the safety (or "shut down") zone from 1,100 yards to 2,200 yards. She also included a requirement that the Navy keep two trained (by the National Marine Fisheries Service) lookouts on hand during exercises, and conduct a 60-minute monitoring of the training area before any exercise. The Navy will in all likelihood appeal this decision. Who knows? The case could get all the way to the Supreme Court before it's done.

So far, the courts are finding the NRDC's arguments persuasive, but what about the actual underlying science? The NRDC believes there is "overwhelming" scientific evidence that sonar testing can be harmful to marine mammals like beaked whales and dolphins, and issued a report in 2005 (Sounding the Depths II: The Rising Toll of Sonar, Shipping, and Industrial Ocean Noise on Marine Life) outlining the problem and its proposed solutions for mitigation. Sonar related strandings have been confirmed off the Canary Islands and in Spain since 2000, and there have been similar incidents in Hawaii and Washington state where the direct link between the strandings and sonar is less clearly established.

I'm sympathetic to the marine mammals' plight, but "overwhelming" is a pretty strong term; in my experience, scientists tend to hedge a bit more when discussing the implications of their findings. And such is the case here. I've chatted (formally and informally) with lots of scientists involved with ocean acoustic research, some specifically with sonar, others studying how sound propagates in the marine environment (it's as complicated as how sound propagates in air, with lots of contributing factors), and the consensus seems to be that there is strong circumstantial evidence, but we're not quite to the point of reaching a broad scientific consensus as yet because there's so much we still do not know.  For starters, apparently, a definite link to sonar as a cause of a stranding can only be determined if the animal is discovered quite soon after death -- a rare occurrence.

That's why even those scientists who agree there's a detrimental effect, and are sympathetic to the NRDC's cause, tend to caution that a direct causal relationship has yet to be established for most stranding events, the Bahamas incident excepted. It's strong circumstancial evidence, to be sure, but the ocean is naturally a pretty noisy environment, thanks not just to sonar, but shipping noise (low-frequency rumblings from engines, propellers, etc. found on merchant ships), seismic testing, even communication among the whales themselves. What makes mid-frequency sonar (and to a lesser extent, low-frequency sonar) different from other kinds of sound and/or noise in the ocean?

Such impacts depend on a large number of variables: the loudness, frequency, and duration of the sound, for example, the hearing sensitivity, age, gender and behavior of the animals, even the specific circumstances under which they hear the sound -- i.e., are they hearing a sonar pulse for the first time, in which case it would be more startling. Scientists rely on complex models to estimate sound levels at different distances and depths from the sound source, and to then simulate the behavior of any animals that might be within range. Those models are pretty good, but really, to more accurately make predictions, scientists need better data on the animals' behavior. (There is a project to create an extensive online database of the diving, movement and acoustic behaviors of fish, marine mammals and sea turtles to aid the scientific community in determining the effects of sound on them, called the Marine Wildlife Behavior Database.) It's also tough to document everything going on in a given area and figure out which activities might be contributing to a stranding. Even with the 2000 Bahamas stranding, the circumstances are not completely understood: factors including ocean canyon topography, surface ducts, multiple ships, reverberation enhancement (resulting from things like the presence of steep canyon walls), and so forth that, taken together with mid-frequency sonar, initiates a chain of behavioral and physiological responses in beaked whales.

Equally unclear is how, exactly, sonar contributes to the symptoms observed in stranded animals.  There are two prevailing hypotheses. One is that the acoustical properties of sonar cause nitrogen to leak out of the whales' blood. However, a more likely hypothesis, according to scientists like NOAA's Brandon Southall, is that sonar causes a behavioral panic response in the whales. When resurfacing, scuba divers practice a five-meter safety stop before surfacing completely, to guard against the bends. Beaked whales appear to show similar behavior in their dive profiles, following deep dives with a series of shorter progressively shallower dives. If the animals panicked and altered this behavior in response to the sonar pings, the abrupt changes in pressure could give rise to effects such as nitrogen bubbling out of the blood. The problem with this hypothesis is that any loud noise from non-sonar sources should also be affecting the dive profiles, and therefore also contributing to what seems to be a growing number of stranding events.

It would be unfair to assume the Navy is completely unconcerned about the potentially negative effects of their training exercises on marine mammals. For instance, the Office of Naval Research has an aggressive research program in place to study the acoustic impacts on marine life (with plans to increase its budget for such research to $18 million this year); so do other federal agencies, including the NSF. Environmental groups like the NRDC tend to view in-house research as a bit suspect in terms of objectivity, because the Navy controls the results. But that, too, is a bit unfair, since we're talking about peer-reviewed academic research, for the most part. Even assuming the odd biased scientist, realistically, they can't all be dissembling. Call me an idealist, if you must, but I really believe most scientists are smart enough to take account of their own built-in biases.

Anyway, there are ongoing studies of whale ear anatomy being conducted by Darlene Ketten at the Woods Hole Oceanographic Institute (WHOI), for example, using advanced imaging techniques (CT scans and MRIs, mostly) to get a better idea of how whales hear, and what their exact frequency hearing range might be, although I'm puzzled as to how they manage to get a whale to hold still long enough for an MRI -- sedation? (Of course, I'm assuming a live whale.) Also, how do they get it to fit inside the big magnetic doughnut from hell? Inquiring minds need to know!

As for how whales specifically react to different kinds of sound, there's an ongoing, major Behavioral Response Study in the Bahamas going on right now with researchers from WHOI, NOAA Fisheries, the Navy and a half dozen academic institutions to measure the behavioral response of different cetaceans (beaked whales, pilot whales, etc.) to mid-frequency range sonar and sonar-like sounds. These are not creatures who can be easily brought into the lab, or be trained to perform in a certain way for experiments; scientists need to study the animals in situ. To do this, they're relying on some pretty neat technology, including digital acoustic recording tags (DTAGs). These are basically sophisticated acoustic sensors (hydrophones) coupled with 3D accelerometers (and magnetometers) mounted onto suction cups, capable of recording audio, pitch, roll, heading and depth, equipped with 6.6 gigabytes of memory. 

The tags are attached to the animals using cantilevered long carbon-fiber poles -- a tricky feat in itself, because whales generally don't want to stay still long enough to let you attach a suction cup to their skin -- and then they merrily go on their way. (One of the scientists, Peter Tayack, phrased it thusly in an article for Oceanus: "It's still a difficult process that requires luck, patience, decent weather, and some measure of fortitude. We find ourselves in tiny boats, trying to sneak up on large and often intractable wild animals to stick something on them with a long pole, during the small fraction of time they are at the surface. Any one of our 'subjects' could swim away from us or dive at any time. The work is exciting on many levels." Clearly, Tayack is a master of understatement.

The tags are recovered after recording is over, and the data is transferred to a computer, where it can be analyzed. So for the first time, scientists can go along with a whale as it dives, hear what the whale hears, and "observe" how it behaves in its natural environment. Among other things they've learned is that beaked whales are "the extreme breath-hold champions of all animals studied so far," according to WHOI engineer Mark Johnson, who developed the DTAGs several years ago. They even beat out sperm whales, which can dive for more than an hour to depths of more than 4000 feet (1200 meters), although they typically make 45-minute dives to around 2000 to 3820 feet.

The animals also exhibit interesting dive behavior. After a deep dive, the whales stay near the surface and make a series of shallow dives for an hour or so. No one is sure why they do this, but one hypothesis is that they have used up most of their oxygen stores before the dive is done and are thus forced to resort to anaerobic metabolism for the last small part of the dive. Since this gives rise to lactic acid buildup, the whales may need a rest period to process the buildup -- similar to what athletes experience after an intense workout, and rely on "cool down" periods to clear lactic acid from overworked muscles. The whales also ascend much more slowing than they should really need to after a deep dive -- again, something that puzzles researchers.

So the research is progressing, inch by inch, as science does. And the lawsuits are inching along as well.  While I support the need not to over-state the case and base such legal/policy decisions on sound science, I just hope that scientists don't take too long reaching a broad consensus on the issue; some of those marine mammal species at risk are on the endangered list, and might not have much time.

Comments

[Lab Lemming]

"Call me an idealist, if you must, but I really believe most scientists are smart enough to take account of their own built-in biases."

No, they are smart enough to cover them up and make themselves look impartial.

Call me a cynic, but can't they just ping the whales wearing trackers to see if they react?

[Robert Wright]

Your reference to lactic acid made me think of this:

http://www.nytimes.com/2006/05/16/health/nutrition/16run.html?_r=2&adxnnl=1&oref=slogin&adxnnlx=1199366030-9Y0baK2IEGCkPtCdzt/E3A&pagewanted=print&oref=slogin

[Axiom]

"The problem with this hypothesis is that any loud noise from non-sonar sources should also be affecting the dive profiles, and therefore also contributing to what seems to be a growing number of stranding events."

This remark is a bit of fore-shadowing as it turns out that the stick-on accelerometer / hydrophone behavioral study included a variety of sound sources including recorded mid-frequency sonar and orca calls. One of the interesting results (in the quite small sample set) was that both mid-frequency sonar and orca seemed to produce a resulting behavioral change (orca was a much stronger response) in the beaked whales. However, the response was to cut short the time on station at the bottom of the dive (when they're actually feeding) and make a lateral move, not a vertical one. The beaked whales went silent, moved laterally out of the area and, minutes later, began a controlled climb (definitely not a panic rise).

Small sample set caveats apply, of course (see the remarks in the article about slippery whales).

However, one point that you did not mention was that all this took place at AUTEC where the underwater tracking range hydrophones provide unique information on beaked whale vocalizations (among many other things, of course). One interesting indication resulting from that data is that the sightings of beaked whales (who tend to bob just below the surface, as you mentioned) do not match up with the vocalization data. Perhaps there are quite a few more whales than we thought....

[The Flying Trilobite]

Totally fascinating, thanks for this article.

I'm quite enjoying your writing. 'New 'round these parts. Felt the need for a cocktail. Maybe a Shirley Temple.