Greg Marshall: 20 Years of Putting Cameras on Animals
Marine biologist Greg Marshall was scuba diving in Belize in 1986 when he saw a small fish hitching a ride from a shark simply by sucking onto its back. Marshall was inspired by what that smaller fish would get to witness by being attached to the shark and began working on a camera system that could attach to an animal and record everything the animal saw. His invention, the Crittercam, has evolved much since it was first deployed in 1987, and nature film fans and researchers continue to be dazzled by what the cameras capture. Marshall will speak on Thursday at the State Theatre about his work with the Crittercam.
Rodney the emperor penguin, equipped with a National Geographic Crittercam, gets ready to dive beneath the Antarctic ice. More Crittercam images of penguins, lions and seals in the slideshow.
City Pages: Explain to me a little about how the Crittercam works.
Greg Marshall: There are two types: the marine and the terrestrial incarnation. Most of the work we've done to date has been with the marine system. It is a system that incorporates all the video, audio, and environmental and geospatial data recording in the system. So you have to recover the system in order to get the data back since we're not able to transmit under water. So what that means is that not only does the housing have to contain all these components, it has to be rugged enough to go down to a thousand meters deep, then after it comes off the animal float back to the surface for recovery. That means we have to make the system slightly positively buoyant, very streamlined so it slips through the water easily. Then when it's sitting at the surface, a radio antenna pops out and transmits a signal, and we can triangulate on that signal and go out and pick the system up where ever it happens to be floating. From that point you can download the video, which is nowadays on flash memory, not on videotape any longer. We pre-program the system to come off at a certain time or under specific environmental conditions. For example, we know that these systems are certified for 1,000 meters, so if the animal goes to 1,000 the on board computer detects that and sends a signal then to abort the mission and release the system. And most of the deployments with a suction cup attachment mechanism, so the release is mediated by opening a valve in the suction cup, so it floods with water and releases.
There are two versions of the terrestrial system. One is much akin to the marine system in that all the data is recorded and we recover it after the deployment. The other is a video transmission system that is real time, so we can be three miles away observing what the animal is doing in real time. Those systems are totally remotely controlled, so we can get the video and audio and data transmission from up to three miles, and from five miles we can totally remotely control and interrogate the system. So by the end of the deployment, not only have we turned the system on and off during the deployment to sample behavior over time because those systems can transmit 25 or so hours of video, also at the end, we send a signal to release the entire collar from the animal, it drops off and sends a signal to us to pick it up.
CP: The Crittercam can go as deep as 3,000 feet? Can you see anything that deep?
GM: We've resolved that in a different way. In some of the systems we have on board image intensification capabilities, so we can amplify the available light 50,000 times. Even in very low light we can resolve and image. In other cases we also include a series of headlights within the system so we can project light into the water to be able to see what the animals are seeing. In all cases we're careful to do whatever we can to not change the animals' behavior because, after all, we are doing this for research and if we're changing their behavior, what's the point? So we work hard to engineer problems like that out. In a case like this with a headlight, we've tested a number of different systems and we're pretty confident that the animals we're working with can't see the color of the spectrum of the light we're using. So we're using a near infrared light that they're probably not adapted to in any case because red light absorbs so quickly into the water. So we're not affecting their behavior.
CP: What other sort of atmosphere difficulties do you face?
GM: Well obviously pressure is a huge challenge and light is a challenge. But predominantly I think in all cases the biggest challenge is to ensure that whatever we're deploying on the animal is appropriate for the animal and the research question. And that's always a challenge because it's easier for us to design and build bigger systems, it's easier for us to design and build boxier systems that don't take streamlining into consideration. We spend a lot of time and energy engineering the systems to be appropriate for the animal and the question we're trying to answer.
CP: Can you reuse the Crittercam after your retrieve them?
GM: Yeah whenever we can. Absolutely. I would imagine we've reused each system ten to 20 times. Now with these new systems, I imagine we'll reuse them, I'm hoping we can reuse them 30-50 times because they're much smaller, more robust, and we've tried to make them as universal as possible.
CP: What sort of data, besides video and audio, are gathered?
GM: We have on board temperature transducers, pressure transducers, light level meters to measure ambient light, we're measuring the accelerometry of the animal in three axis, we're measuring magnetometry- the compass direction the animal is moving in- the speed of the animal. By virtue of all of those things, when we reconstruct all those data sets, we're able to get a very coherent sense of how the animals are moving through and using their environment.
CP: What animal that you have you not been able to study with a Crittercam that you would like to?
GM: To date the systems have been too large, in my mind, to work with dolphins. I think dolphin behavior in society that is just so interesting that it's going to be absolutely fascinating for the first time to be able to work with them with our newest generation system. These new systems are 2.25 inches in diameter, and now I think for the first time appropriate for an animal the size of a dolphin and as energetic as a dolphin.
CP: I read that you were experimenting with a Crittercam you can place on birds.
GM: Yeah, that's the smallest of the terrestrial incarnations. The cameras and transmitters are quite small. So with a battery that's appropriate for the bird to carry we can do effective bird deployments. The challenge is in finding an interesting research question, everything we do is based on a research question, that you can resolve in an hour or two of transmission. That's sort of the limitation of those systems, they are only transmitting as long as the battery that animal can carry lasts, and that's a fairly small battery load.
CP: You've been recording animals with Crittercams for 20 years now, does it ever get mundane or routine? Or do you find yourself constantly learning new things?
GM: It really is the latter. We're constantly improving our capabilities, we're constantly working with new animals, and we even learn from animals we've worked with over years. We did years of work with harbor seals in Nova Scotia off the coast of Canada, and every year we learn something new. We're amazed to see behaviors that we hadn't seen before and hadn't anticipated before. It really helped us construct a whole new sense of what they're doing during the mating season, how they're feeding, what they're feeding on, and where. To paraphrase a friend and colleague, the great thing about Crittercam is that it shows you things you don't know you don't know. And that proves to be the case with most of the animals we work with. We really don't know what these animals are doing out there. We're surprised all the time.
CP: That leads to my next question. Obviously, your cameras have revealed a lot about how animals live outside of human sight, and we know so much more because of Crittercams. But do you ever worry that the cameras spoil a little of the nature's mystery?
GM: I know, it's a great question and important issue. On the one hand, yes, I mean mystery is wonderful and the mysteries of these animals is part of what inspires me to explore them and their relationship their environment. On the other hand, and I think in today's world more importantly, their world is changing so much, and the stresses they feel in their environments are changing so rapidly that in order for us to effectively understand what their basic needs are, their basic biology, we've got to be out there trying to understand them better so we can affect better conservation measures. A large part of what we do is in fact conservation biology the object is to understand these animals better to affect more effective management of the resources that they need.
CP: How many animals on any given day are outfitted with cameras around the world?
GM: We're on the verge of making it a daily event. Historically we’ve had to build these things one-off, basically designing each system and building each system as a unique unit. So historically we've had in our possession only somewhere between five and maybe 10 at any given time. Within the next month or month and a half, we're going to completing our next generation system which can be mass produced for the first time. We expect in the next year to build let's say a hundred to start off with and in the next two years maybe 500 of them. So then I'll be able to give you a better answer because then we'll have 30 projects going on at any given time around the world and we'll be able to say 'yeah, there are 50 systems deployed and tomorrow there will be 70 systems deployed.'
CP: When editing penguin tape, it is really boring to watch the video when they're standing on land, since the camera is pointed straight up, or is there always something to see?
GM: Yeah, good question, it's interesting, we use a saltwater switch, so the system turns off when they're on land and back on when they're in water, that's all controlled by the computer.
See Greg Marshall will discuss his work and innovations with Crittercam at the State Theatre Thursday. Go to www.hennepinthredistrict.org or call 612.373.5600 for more information.