Navy sonar: The dolphin experiment

The first day of the ONR (Office of Naval Research) conference was absolutely chock full, 7am till 11pm at night. About 20 of us scientists, plus the 3 Navy guys who run the marine mammal program (and also a curious NOAA person sitting in a corner "spying", as he put it). The bulk of the day was a series of 20-minute talks. Basically, everybody at the workshop got up and summarized what they've been working on recently, one at a time. (My own talk was on right whales and also on my recent sea turtle stress research, which uses some techniques applicable to marine mammals.) Since the 20 people in the room are basically everybody who's studying stress in marine mammals, worldwide, what the Navy guys were getting - and what they wanted - was a great summary of the state of the art in marine mammal stress research.

Despite a severe case of jet lag and sleep deprivation, I can safely say that this was the single most interesting day of scientific talks that I can ever remember attending. (At most scientific conferences I start to space out about halfway through the day. But today I was riveted, right to the last talk.)

So, next I'll post a few summaries of the ongoing research projects that people were describing. First up: Dorian Houser's dolphin experiment

******
Talk #1 - Stress indicators in dolphins
Of all the experiments described today, this one was the one that will most directly answer the question of: Does sonar atually stress ut a dolphin? Dorian, the fellow giving this talk, has what some (including me) would consider to be a dream job: working with the 70 tame, trained bottlenose dolphins kept by Navy dolphin trainers at their facility in San Diego. "We've got known life histories, they're trained, they'll wear harnesses, we can get voluntary blood samples." he said. For a physiologist, that is a dream situation. (The blood samples are taken from the tail - the dolphins know they will get a fish if they turn upside down and put their tails in the trainer's hands).

Yes, the Navy has 70 bottlenose dolphins. They also have about 30 sea lions. They used to have pilot whales and killer whales and belugas. They've had marine mammals for a long time, originally to study their swimming mechanics and streamlining (for torpedo design). Then later they realized dolphins could "assist the Navy with various activities", as Dorian rather mysteriously put it. I've read a few books about this, including one written by a former Navy dolphin trainer and the "activities" are just what you would guess: dolphins can find things on the ocean bottom, and can scope things out. (They don't use dolphins for offensive warfare - because dolphins can't distinguish friend from foe! They're just trained to find and tag things.) Mostly they are trained to search for ocean mines and swimmers (downed friendlies, enemy spies). My general impression from reading the books is that the Navy seems to treat their animals well. If they didn't, the animals would just take off, since many of the animals work in open ocean, in completely free situations and could leave at any time.

The dolphin training at the Navy, like dolphin training anywhere, works entirely by positive reinforcement. A dolphin will do just about anything for a fish. Dorian got onto an interesting tangent at one point about how many fish a dolphin needs per day, and how many fish he is allowed to use to "pay" the dolphins to participate in the sonar experiment. The classic dolphin-training approach is to use a small percentage of the dolphin's daily fish needs for that day's training activities - but only a small percentage, meaning, the animal won't starve if it choose not to participate. Typically, if the dolphin doesn't choose to participate in the activity, they get all of their remaining allotment of fish anyway at the end of the day.

Dorian said that you can gauge how much a dolphin likes or dislikes a certain activity by how many fish it demands to do the activity. (or, to put it more precisely, what % of the animal's baseline food intake comes from training.) Some activities just cost a lot of fish. Apparently dolphins have a finely calibrated sense of economics, and fish is the currency of choice!

Interestingly, sonar does not "cost" much to dolphins - most dolphins are willing to sit and listen to a sonar ping for relatively few fish, and don't have to be very hungry to be willing to take part in a sonar experiment.

So here's Dorian's study design. He hasn't done the experiment yet - it's all designed and ready to go. As you're reading this, think about: What do you think of the study design? Could anything be improved? Are there any flaws?

He's got 30 dolphins and 15 sea lions lined up for this experiment. Every animal is already trained on an "ABA task", which means that they know that they are supposed to start at a certain position in the pool (position A), swim to position B and touch a paddle, and then swim back to A.

In the middle of this task, Dorian is going to play a very brief, fraction-of-a-second, very loud, sonar sound at them. PING! He's testing several different decibel levels, ranging from loud to louder to extremely loud. Each animal will hear just 1 of the decibel levels. And then he'll watch, basically, for signs of distraction and annoyance. Stress, in other words..

So how is he planning to measure that ever-so-amorphous concept of "stress,", in dolphins and sea lions? Here's what he's planning to look at:

1. Distractability. Do they complete the ABA task in the normal amount of time? Might they abandon the task entirely?

2. Do they change their behavior in any other way? "We'll especially be looking for chuffing and tail slaps," Dorian said. Apparently that's what a dolphin does when it's annoyed or agitated. (I wanted to ask what "chuffing" is, but didn't get a chance.) Dorian's got cameras mounted all around and above the pool, plus a microphone running above water - there will be dolphin trainers will be stationed around the pool, calling out verbally when the dolphins dive or do any behaviors that the cameras might miss. There will also be underwater microphones (hydrophones) to pick up any vocalizations the dolphins make themselves. I was impressed at the pool set-up (he showed us photos) and I was also grateful that I am not going to be the person who has to go through all the recordings!

3. Heart rate. There will be ECG heart rate monitors strapped right on to the dolphins! (just like we use in physiology lab at UP - except the 3 electrodes are stuck on with suction cups.). When you are stressed or scared, of course, your heart rate speeds up. The same thing happens in any mammal. It's the hormone epinephrine - also known as adrenalin- that's doing that. BUT, there's a big complication in marine mammals: heart rate plummets when they dive. It can drop to shockingly low levels in deep divers on a long dive, as low as five beats per minute in certain species. Then it speeds up to above normal when they first surface. Dolphins are only "surface divers," meaning that they don't go deep, but, even so, their heart rate definitely drops a bit. So, Dorian's going to have watch when the animals are diving, how long they've been under, whether or not they've just surfaced, and he'll have to factor all that in to the analysis of the heart rate data.

4. Stress hormones. He's planning to take blood samples (every sample paid for with a fish or two, of course) at several times: several days prior to the event (to establish that animal's baseline), immediately after the event, and one week later. He'll assay these samples for epinephrine, glucocorticoids, and several other measures too.

What do you think? What would you do differently? A bigger sample size would always be nice, of course, but thirty dolphins is the best sample size I've ever heard of for a controlled experiment with captive marine mammals.

I love this study. It's perfect. It's exactly what we need! The rest of us can just go home! In fact, I want to go to San Diego to watch! Trained dolphins and sea lions... That he can get blood samples from... I couldn't get over it. I kept muttering to myself "Thirty trained dolphins wearing harnesses." (It's the contrast to the impossibility of getting a blood sample from a baleen whale that was so entrancing. I work on baleen whale physiology, which is pretty much an exercise in frustration from start to finish. It's like trying to study Martians. You can't capture them, you can't even see them, and sometimes you wonder if they even exist at all. And a baleen whale is the only type of animal, on the entire planet, that cannot be captured alive.)

I can see one thing that could be added to the experiment, though, which is to grab fecal samples too! In many marine mammals, such as the baleen whales, we can only get fecals, never blood. (And sometimes small changes in stress hormones show up only in the feces and are not detectable in blood.) If we're going to want to extrapolate the results of this dolphin experiment to other species... well, it would be great this experiment included fecal hormone!

So later that evening, when we're all at dinner and the wine is flowing freely, I lean over to Dorian and say "You must get poops! You have to get poops. Please get poops from your dolphins!" We get into a discussion about feasibility: what consistency are the dolphin poops, anyway? What do they look like? How "scoopable" are they? I'm shouting all this across Sam Wasser, who is sitting between us (Sam being the guy who invented the poop hormone assay technique, and also the guy who found a way to get DNA out of elephant ivory, but that's another story.) So Sam chimes in with a great design for a poop-scooping Pepsi bottle device that he's been using with free-swimming killer whales in Puget Sound. That's it. That'll work great.

Dorian then says to Sam & me, "Hey, you want to do anything with these dolphins? We've got 70. They wear harnesses and everything. Whatever you want to do. Let me know." SEVENTY TRAINED DOLPHINS! Are you kidding? We could test all our fecal hormones - see if they correlate to blood - test the new thyroid hormone assay - do an ACTH challenge - I've immediately got a dozen ideas. Here's hoping.

I'm muttering "Seventy trained dolphins wearing harnesses," under my breath for the rest of the evening.