The Deep-See Peers into the Depths

In the ocean’s shadowy depths lies one of the Earth’s last frontiers: the ocean twilight zone. It’s a vast swath of water extending throughout the world’s oceans from 650 to 3,280 feet (200 to 1,000 meters) below the surface, and it abounds with life: small but fierce-looking fish, giant glowing jellies, and microscopic animals that feed marine life higher up the ocean’s food web.

This cold, dark, remote region of the ocean has remained largely unexplored, but a team of scientists and engineers from Woods Hole Oceanographic Institution have pioneered an ambitious new vehicle to blaze a trail into this ocean wilderness. Known as the Deep-See, it is a modern-day, subsea Conestoga wagon filled with a remarkable array of instruments designed to illuminate the ocean’s mysterious interior and reveal how many and what kinds of animals live there.

“To date, scientists have used several methods to explore the midwater depths, but each had its pros and cons,” said WHOI marine biologist Larry Madin. Acoustic sonars could detect masses of animals that reflected sound well, but they usually couldn’t distinguish individual species. Nets could bring back some intact animals, but they often squished the more gelatinous ones and missed those that were very small or could get out of the way. And cameras weren’t so effective at capturing images of animals that were sparsely distributed, moving, and often small or transparent.

“The aptly named Deep-See combines complementary imaging systems that promise to detect a broad range of organisms,” Madin said. It is towed behind a research ship from an electro-optical cable that can transmit power and data between the ship and the vehicle in real time. And it’s big, weighing 2,500 pounds (1,250 kilograms) and measuring about 16 feet (5 meters).

“It has plenty of room for all kinds of acoustic sensors and optical sensors—that’s another fancy word for cameras,” said WHOI acoustic oceanographer Andone Lavery, the lead scientist on the Deep-See project. “This particular combination has never been used before to study the twilight zone.”

The platform’s sophisticated acoustic and imaging systems include:

• broadband, split-beam sonars to detect, count, track, and identify animals
• a holographic laser-based camera to capture 3-D images of tiny plankton
• a specialized stereo camera-and-lighting system to photograph jellyfish and other large animals
• sensors to measure seawater properties, such as temperature, salinity, and dissolved oxygen
• a sampler to collect DNA signatures of ocean twilight zone animals

WHOI mechanical engineer Kaitlyn Tradd helped to design and build the vehicle in three sections, or modules: the forward optics module for the cameras and lights, the middle acoustics module for sonars, and the aft module—the tail—for hydrodynamic stability and space to hold additional sensors and equipment.

“The three separate modules also allow us flexibility when it comes to how we configure the vehicle for a given scientific objective,” Tradd said. “The modules easily bolt together, and new sections can be developed and added should the need arise.”

As novel as the Deep-See is, many of its cutting-edge systems are built on decades of technology development and basic science research by WHOI engineers and scientists.

The early days of broadband acoustics

Lavery began working with sonar systems when she was just out of graduate school, and she quickly discovered that using sound to identify and count animals in the ocean is a tricky business.

A typical echosounder or “fish finder” works something like an acoustic flashlight, transmitting a single-frequency beam of sound into the water below a ship. Sound waves reflect off fish and other organisms, creating an echo that a receiver on the echosounder can detect. Many common fish, with their gas-filled swim bladders, provide readily detectable targets.

But what scientists really want to be able to do, says Lavery, is to tell how big the target animals are and how many there are—in other words, does the returning echo represent a single large fish or dozens of tiny zooplankton?

“When you have a single sound frequency, it’s really hard to tell,” Lavery said. “Because there are lots of different combinations of organisms that can give you a similar echo.”

WHOI acoustical oceanographer Tim Stanton is all too familiar with that problem. He spent more than 20 years bouncing sound waves of all different frequencies off individual organisms in test tanks. Lavery joined his efforts when she first came to WHOI as a postdoctoral researcher.

“We put one organism at a time in the test tank,” Stanton said—from large fish all the way down to a tiny swimming snail the size of a head of pin. To keep the snail in front of the beam of sound, Stanton restrained it with an acoustically transparent tether: a human hair.

“We did this both on land and on the deck of a ship, collecting nothing but live, pristine organisms, and making these series of measurements,” Stanton said.

Through that painstaking process, Stanton and Lavery were able to identify each species’ unique acoustic “signature”—the strength of the sound waves bouncing back off an organism at various frequencies. These included high-frequency sound waves—the kind needed to detect smaller crustaceans such as copepods and krill.

In the early 2000s, Stanton and Lavery started testing their lab-based signatures in the open ocean, working with seagoing acoustic systems that could transmit and receive sound at not just a single frequency, but at several different ones, or across a whole spectrum at once. They showed that different sound waves returned from different organisms, proving that this so-called broadband approach could distinguish and count animals in the open ocean.

For them, it was like going from getting information from only one radio station, then from several, then from all the stations across the entire FM dial.

Building on the BIOMAPER

One predecessor to the Deep-See was a ship-towed vehicle developed at WHOI called the Bio-Optical Multifrequency Acoustical and Physical Enviromental Recorder, or BIOMAPER-II. It was equipped with transducers that transmitted sound at 43 kilohertz (kHz), 120 kHz, 200 kHz, 420 kHz, and 1,000 kHz.

“BIOMAPER-II had a lot of high-frequency acoustics on it,” said WHOI biologist Peter Wiebe, who led its development. “That meant it could detect not just fish, but tiny plankton that can only be detected at higher frequencies.”

However, BIOMAPER-II could only descend to 300 meters—not deep enough to be useful in the ocean twilight zone. In contrast, says Lavery, the Deep-See can descend to 2,000 meters and transmit sound across frequencies from 1 to 500 kHz.

“One of the big advantages of the Deep-See,” said WHOI scientist and engineer Dana Yoerger, “is that it puts high-frequency acoustics right down into the twilight zone. You can’t use high frequencies from a ship because they are quickly absorbed in seawater, long before they can reach the twilight zone.”

But having a small library of laboratory-derived acoustic signatures isn’t sufficient. The signatures for most already-identified twilight zone animals remain unknown, let alone for the species yet to be discovered. In addition, scientists need to ground-truth the acoustic data by seeing with their own eyes what the sonar systems are detecting. From a ship, without a submarine, they have only two options: cameras and nets.

The next generation of cameras

BIOMAPER-II had several bio-optical sensors and a video plankton recorder, or VPR—a kind of underwater microscope that could capture high-resolution images of tiny particles and plankton from 50 microns (0.002 inches) up to a few centimeters (about an inch and half) in size. The Deep-See improves on its predecessor with two camera systems capable of capturing images of the organisms detected by its acoustic arrays: one holographic, one stereo.

The first is a small-area, holographic camera system, developed by emeritus WHOI biologist Cabell Davis at the marine technology company he founded, Seascan. The camera system is analogous to the BIOMAPER-II’s VPR, “but it uses lasers to take detailed, 3-D images of tiny plankton in their natural environment without disturbing them,” said engineer Cliff Pontbriand, who worked on the camera at WHOI. To do that, the holographic system sends out a laser beam with a diameter of 1.5 inches (3 centimeters)—about seven times per second—from a transmitter on one side of the Deep-See’s front frame to a receiver 3.3 feet (1 meter) away on the other side of the frame.

The Deep-See’s stereo camera system also builds on an earlier technology, known as the Large Area Plankton Imaging System, or LAPIS, which Madin and colleagues developed more than a decade ago.

“The original LAPIS was really a proof-of-concept to provide images of larger organisms than the VPR could,” Madin said. Special strobe lights provided illumination for the LAPIS cameras, allowing them to “see” in dark ocean water down to 1,640 feet (500 meters) and capture low-resolution, black-and-white images of both opaque animals such as krill and transparent ones such as jellyfish and salps.

“The trick is in the lighting, which needs to be reflected for opaque targets but refracted—from beside or slightly behind—for transparent ones,” Madin said.

The Deep-See contains a next-generation LAPIS camera system. It images a 1-square-meter swath of water using a more versatile LED-based lighting array rather than power-hungry strobes, and it produces 24-megapixel images instead of 1-megapixel ones. The higher-resolution image quality makes it easier for scientists to identify twilight zone animals and even study some of their behavior.

Genetic evidence

To complement its acoustic arrays and camera systems, Deep-See carries a host of sensors that measure seawater characteristics, such as temperature, salinity, dissolved oxygen concentrations, and the amount of light available to marine plants.

In addition, a sampler aboard the Deep-See collects filtered seawater containing genetic material from organisms living in it. Using cutting-edge gene-sequencing technology, WHOI biologist Annette Govindarajan is analyzing the water for this environmental DNA, or eDNA, seeking genetic evidence of life.

“Environmental DNA will allow us to detect evidence of twilight zone animals, including those missed by other sampling methods,” Govindarajan said.

Even with its combination of acoustics, imaging, environmental sensors, and water sampling capabilities, the Deep-See will not tell researchers everything they want to know about the twilight zone. To really understand this little-known region of the ocean, scientists will need to combine data from the Deep-See with information gleaned from traditional net tows and gathered by new underwater robotic systems such as the Mesobot. Using a multifaceted approach, Lavery says, should make it possible to reveal more accurately the abundance and diversity of animals in the twilight zone and to understand their behavior. It will also help determine how that behavior affects the ocean’s chemistry, including the transfer of the greenhouse gas carbon dioxide from the atmosphere to the deep ocean, which has huge ramifications for Earth’s climate.

“No system is foolproof,” Lavery said. “But I think that with Deep-See’s combined capabilities, we can begin to get at some pretty important questions.”

The Deep-See had its first sea trials in August 2018, on WHOI’s first ocean twilight zone expedition. The nine-day cruise aboard the National Oceanic and Atmospheric Administration’s research vessel Henry B. Bigelow was a collaborative mission with NOAA’s Northeast Fisheries Science Center and the University of Connecticut. The Bigelow navigated beyond New England’s continental shelf to the deeper waters of the northwest Atlantic Ocean, where the vehicle’s unique combination of instruments collected more than 22 terabytes of data.

On its first foray into the twilight zone, the Deep-See has already challenged scientists’ previous understanding of life in the deep ocean. Earlier acoustic explorations suggested that twilight zone animals were concentrated in one or more dense layers. However, because most of these early acoustic systems operated at lower frequencies and were mounted on the ship’s hull, the sound scattering they detected was mainly from animals with internal gas bubbles, such as swim bladders in fish or gas-filled chambers in jellies that help them float. Many organisms, especially ones not containing gas bubbles, were acoustically invisible.

The more perceptive Deep-See was able to detect twilight zone animals with and without gas bubbles, spanning a diverse range of species—and found that they were spread throughout the twilight zone at all depths.

“That was really surprising,” Lavery said. “I’m eager to find out what the Deep-See will reveal to us next.”

Funding for the development of Deep-See came from the National Science Foundation.