National Geographic's Crittercam systems are research instruments worn by wild animals. The systems allow scientists to study the animals' behavior and ecology. Crittercams allow video and sound recording as well as environmental data collection (such as depth, temperature, and acceleration readings).

Marine Crittercams are small streamlined water- and pressure-proof systems that include recording equipment. Terrestrial Crittercams—built to be worn by land animals—are collar-mounted and do not record. Rather, they use radio waves to transmit video, sound, and other data to receiving and recording stations.

Crittercam systems are deployed on animals for research, to go where the animals go and see what they see, undisturbed by a human observer. They allow us, for the first time, to witness the lives of animals from the animals' own perspectives.

In 1986, while diving off the coast of Belize, marine biologist Greg Marshall encountered a shark. The shark approached him, then disappeared with three quick strokes of its tail.

Stuck to the shark's belly was a remora fish. As Greg watched the shark disappear, it occurred to him that if he deployed a camera in place of the remora, he could observe the shark's behavior unfold without disturbing the shark.

Inspired, Greg built the first prototype of what would come to be called Crittercam and deployed it on a sea turtle in 1987. The turtle behaved normally—the first indication that this type of system had potential as a research tool.

Since 1990 National Geographic Society funding and support has enabled Marshall and his team to continuously improve and diversify Crittercam systems.

Today the team designs and builds Crittercam in Marshall's Remote Imaging lab at National Geographic headquarters. There, engineers and technicians, led by Mehdi Bakhtiari, are on a mission to make Crittercam systems smaller and more powerful.

Over more than a decade, we have conducted Crittercam field studies with more than 40 species of sharks, sea turtles, whales, seals, and penguins. We recently began a terrestrial Crittercam research program, working with lions, hyenas, and grizzly bears.

Collaborating with biologists worldwide, the Remote Imaging team has made more than 400 deployments to tackle biological mysteries, including the following: How do humpback whales hunt together? Why do blue whales call? Where do Hawaiian monk seals find their food? How do harbor seals stake out territory? How do emperor penguins hunt? Where do sharks stalk their prey? When do leatherback turtles mate?

The method depends on the species. We deploy Crittercam systems in collaboration with scientists who are experts in their fields, working with them to adapt or develop attachment methods.

For whales, dolphins, and leatherback turtles, we developed special suction cups. With seals and hard-shelled turtles we use a small adhesive patch. Custom-tailored, backpack-like harnesses do the job for penguins, and a passive fin clamp keeps Crittercam swimming with sharks.

For land animals such as bears, lions, and hyenas, we've developed a Crittercam collar.

The purpose of using Crittercam systems is to record animal behavior that's not influenced or disturbed by the presence of a human. If Crittercam bothered the animal, we would not be able to record natural behavior.

We deploy as quickly and gently as possible, and most animals are back to diving, feeding, and interacting soon after being outfitted with a Crittercam. Some animals, especially female seals, show initial interest in the package they are carrying, but soon resume their day-to-day life in the wild.

In 400 or so deployments, we have lost 12 or 13 systems, making our retrieval rate about 97 percent.

We learn how marine animals use their habitat, where they feed, and how they interact with animals of their own species and of other species. We gain insight into how they communicate, how they mate, and how deep they dive (and why). We uncover why they may stay underwater for as long as they do, how they move, how they hunt and feed—virtually every aspect of life.

We're starting to gain some of these same insights for the land-based species we've begun working with.

If you just add up the cost of the parts, a marine Crittercam costs between U.S. $7,000 and $13,000. If you add in the research and development costs, they're almost priceless.

Every time we deploy a Crittercam, we're basically throwing thousands of dollars' worth of equipment into the ocean or the outback. In addition, marine Crittercams record all data onboard, so if we don't get the systems back, we don't get the video, the sound—nothing.

For marine animals, we can release Crittercam systems remotely by programming a computer inside the system to trigger the release mechanism at a certain time. The Crittercam then floats to the surface, where we can track and recover it using its onboard radio beacon.

For land animals, at any time we can send a signal that causes the Crittercam collar to detach from the animal.

To find Crittercams after they're released from animals, we use homing devices like ultrasonic signalers and radio transmitters, which enable us to track the systems.