Pick from the Past
Natural History, March 1991

The Sounds of Seal Society

THROUGHOUT the history of Antarctic exploration, investigators have commented on the intriguing underwater sounds made by seals. In 1820, James Weddell, a British sealer who made the first detailed maps of the sea region of Antarctica later named for him, discovered a “mermaid... making a musical noise.”

Weddell seal

Photo: © istockphoto.com/Brian Lee
This seal species now bears Weddell’s name as part of its own, Leptonychotes weddelli. In 1905, surgeon, artist, and zoologist Edward Wilson, who accompanied Robert Falcon Scott on his first attempt to reach the South Pole, heard Weddell seals “produce a variety of musical noises, both laryngeal and oesophageal.” The sounds were striking to the explorers because they broke the overwhelming silence that surrounds the frozen sea. Often the only other sounds heard by explorers were the creaking noises of the moving ice. This eerie accompaniment to the seal sounds may have contributed to the image of an antarctic mermaid.

There are four species of antarctic seals: Weddell, crabeater, leopard, and Ross. When researchers drop a hydrophone into the water to eavesdrop, they often are hit with a cacophony that has been described as everything from jungle noises to sirens. The sounds may represent many seals of one species calling at the same time, as well as more than one species. These seals use overlapping areas of antarctic waters: some inhabiting solid shorefast ice; and some, floating chunks of pack ice. The number of vocalizations in the seals’ repertoires, the types and quality of the sounds they produce, and the context of their calls vary among the four species. Each also has a different social system, position on the food chain, breeding habit, and preferred ice habitat. One of my long-term research objectives has been to understand the ecology of these seals and to evaluate the adaptive significance of their various vocalizations.

From 1976 to 1981, I had the opportunity to spend most austral springs (October through December) studying the vocal behavior of antarctic seals. I worked in conjunction with Donald Siniff and his associates from the University of Minnesota, who were conducting long-term population studies, and with researchers from Hubbs Marine Research Institute at Sea World in San Diego. I collected recordings near two U. S. research stations in the Antarctic: McMurdo, which lies south of New Zealand, and Palmer, which is located south of Tierra del Fuego.

Studies near Palmer Peninsula were conducted from a sailing ship working its way through pack ice. Lying in my bunk at night, I could often hear the underwater vocalizations of leopard seals through the wooden hull. With his years of antarctic experience, our captain, Peiter Lenie, could maneuver the ship through floating ice and deposit scientists and equipment on floes without even waking nearby sleeping seals. While my shipmates tagged, weighed, and put radio transmitters on seals, I dropped a hydrophone over the edge of the ice floe and made tape recordings of underwater sounds. But there was one major problem: ship noise masked my recordings. For one brief moment, I considered asking the captain if he could turn off the ship’s engines, refrigerators, and generators so I could make quiet recordings. The only realistic alternative was for the ship to deposit me on an ice floe and then move far enough away so that my hydrophone could no longer detect the ship’s noise.

This technique did provide quiet recordings of crabeater seals, leopard seals, and killer whales, but the experience was often harrowing. The best recordings were made when the ship was out of sight and when leopard seals and killer whales were swimming nearby. On more than one occasion, I was stalked by leopard seals. These twelve-foot-long predators think nothing of jumping onto ice floes to seize penguins; then, with a shake of the head, they skin their prey, toss the feathered pelt into the water, and devour the carcass. They also eat newly weaned crabeater seal pups, as well as fish, seabirds, and krill. Knowing the leopard seal’s hunting methods,

Leopard seal

Photo: © istockphoto.com/Jan Will
I was very disconcerted to see one circling the floe I was on, periodically lifting its massive head out of the water to see if I was worth the effort of hauling out to catch. Pods of killer whales also patrolled the pack ice. After that first season, I decided that if I wanted to write my thesis, I had to find a better way to record.

There was a better way, a much better way, called sonobuoys. These electronic devices typically are dropped from military aircraft onto the ocean surface to listen for ships or submarines and then telemeter underwater sounds back to a plane or ship. Safely aboard ship, I dropped sonobuoys in the water, and as the ship moved away from the site, I could listen to seals as I sipped coffee on the ship’s bridge and record during all times of the day and night. It was wonderful.

On the other side of Antarctica, at McMurdo Sound, I conducted a series of concentrated studies on the most vocal of antarctic pinnipeds, Weddell seals. Field logistics at McMurdo were very different from those around Palmer Peninsula because these seals lived near shorefast ice. We lived in a ten-by-twenty-foot wooden hut about 100 feet from one colony and traveled back and forth among seven others. Each austral spring, Weddell seals return to traditional fast-ice sites in the sound to give birth to pups and to mate. They congregate in groups of up to 100 seals in sites next to long cracks in the ice. Females haul out on the ice to give birth, nurse their pups intensely, and fast for about two months. During this time, male Weddell seals establish underwater territories along cracks, caused by tidal rise and fall, running through the colony.

When pups are weaned, mothers go into the water and mate in the underwater territories below the pupping colony. As the fast ice breaks up, Weddell seals leave to feed at sea, mostly on fish and invertebrates. The various social interactions going on in the colony during this time (mother! pup, male/male, female/male) require an elaborate vocal repertoire. Mothers and their pups “bawl” and “cry,” usually on the ice, to locate one another and minimize aggressive interactions with other mothers and pups. During dives, Weddell seals’ lungs are collapsed and their mouths are closed, yet somehow the animals still vocalize. Thirty-four types of underwater calls are associated with territorial defense, aggression, submission, and some as yet unknown functions.

Some underwater vocalizations of Weddell seals are as loud as jet engines. And since sound travels farther in the ocean than in air and propagates especially well under polar fast ice (the thick ice cover dampens water movement, which could attenuate sound travel), seals can be heard underwater when none are visible for miles around. Making loud sounds that travel for miles may enable seals in different breeding colonies to communicate. In addition, a loud voice is undoubtedly useful when the seals are hauled out on the ice, for Antarctica is a windy place, with wind speeds sometimes in excess of sixty knots, effectively drowning out all but the loudest of sounds.

Words alone cannot adequately describe the underwater sounds of the Weddell seal, but I will try. Males have eleven types of frequency-modulated, descending “trills,” which can last up to seventy-three seconds. These trills are used by dominant males in advertising and defending their territories. Both sexes produce aggressive sounds in a series that sometimes slows down,


 or speeds up in repetition rate,


“Chirps” also occur in accelerating and retarding series, but their context seems to be submissive. Underwater fights are accompanied by aggressive chugs and responding chirps. The function of the “knock,” a common call that sounds like someone rapping on the door five times in a row, is unknown. Other calls of unknown role are a seventy-five second “mew” and a cheeplike sound that descends in frequency in a steplike manner:

“Seitz, Seitz, Seitz
      Seitz, Seitz, Seitz,
           Seitz, Seitz, Seitz.”

One of the most interesting characteristics of the Weddell seal’s repertoire is the use of short sounds as punctuation. For example, an especially loud, long trill may be preceded by a short, ascending whistle, seemingly to warn the listener to pay attention to an upcoming threat. A series of accelerated, aggressive chugs most often winds up with a “grr-gulp” sound that emphasizes the message. A threatening, guttural “glug” is preceded by a warning whistle and ends with two exclamations: “pop” and “crack.”

For three breeding seasons, I systematically sampled underwater vocalizations by setting up an automated cassette recorder programmed to turn on for two minutes every hour, night and day. As I had expected, with the approach of mating time and heightened territorial defense, the Weddell seals vocalized more and more often, up to seventy-five times per minute. Conversely, after the pups were weaned and mating was accomplished, the number of sounds decreased.

I didn’t expect, however, what happened on December 10, 1976. On that day, my automated equipment suddenly stopped detecting any Weddell seal sounds. Fearing that my first year of thesis data would be incomplete, I thoroughly checked all the electronic equipment but could find no problems. The silence lasted for a couple of anxious days. Then the recorder system started picking up sounds again, but different ones this time. I recognized these new vocalizations as coming from leopard seals. Also about this time, we saw and made recordings of killer whales near the edge of the fast ice. Sometimes they came close enough that I could feel the spray of their blows on my face. The new arrivals had come, I believe, in hopes of getting something to eat, for just about this time-as pups are being weaned and their mothers are returning to the water-the fast ice often breaks up, giving the predators easier access to the seals. I saw the same pattern in 1977 and 1979, when almost to the day, Weddell seals again dramatically reduced their underwater vocalizations and leopard seals and killer whales were sighted in the area. My observations convinced me that the sudden silence of these highly vocal seals was aimed at avoiding detection.

Since my 1976 to 1979 recordings of Weddell seals in McMurdo Sound, Ian Stirling and I have examined recordings made in other locations around Antarctica: Palmer Peninsula, Davis Station, and Cape Hallett. There were some basic similarities, in that all had trills, chugs, and chirps, but each site had calls heard only at that location. We wondered how this geographical variation comes about.

At weaning, pups stop their “bawling” puppy cries and begin to practice adult. vocalizations. We don’t know how much of the adult repertoire is innate rather than learned, but some learning certainly appears to go on. One day, I watched as a near-weaning male pup comically tried to make the trill of an adult male. The pup could not control the frequency or duration of the sound. Instead of a smooth, descending sweep, the best he could do was a warbling, intermittent sound that occasionally “cracked” like the voice of an adolescent boy. I believe that as pups lie on the ice in a colony and listen to the sounds produced by territorial males in the water below, they learn the appropriate adult repertoire. This learning presents a mechanism for geographical isolation and change in the repertoire. Suppose an adult male developed a new sound that was somehow more attractive to females or that helped him defend a territory. The pups would then be exposed to this new sound (since only successful territorial males hang around the colony) and could incorporate it into their developing repertoire. Each year, I would hear a few isolated sounds that did not fit into the colony’s existing call categories, which suggests that Weddell seals do “experiment” vocally.

Much less is known about the three species of antarctic seals that I studied around Palmer Peninsula, but what we have learned so far is suggestive of a link between complexity of vocalizations and mating systems. The most abundant antarctic seal is the crabeater, which eats mostly krill. A sole crabeater male hauls out on a floe with a single female in the spring about the time her pup, conceived some nine months before, is born. Crabeater trios are seen floating on their own “family” floes; in tight ice pack, many such families may be found. The female tolerates the male, even though he probably did not father her pup, because of the additional protection he provides her pup from leopard seals. The advantage for the male is that he will be present when the pup is weaned and the female becomes receptive to mating again. In this social system, there is no need for sounds to attract a mate or defend a territory, and being vocal might just attract predators, so natural selection has favored a very simple repertoire: both sexes emit a single groaning sound underwater. I like to call their rather boring, repeated vocalization, the role of which is unknown, the monotony of monogamy.

The leopard seal is probably promiscuous; males and females are rarely seen together and most likely meet only to mate. The males do not maintain a territory, so they have no need for aggressive vocalizations and should generally remain quiet to avoid being detected by their prey. The main function of vocalizations would most likely be to attract a mate. This species has four, rather lyrical vocalizations, somewhat like someone strumming a finger down the teeth of a comb, and to my human ears at least, their “comb calls” do seem to have an enticing quality. Their vocal repertoire, then, appears to be a compromise between opposing selection pressures, resulting in a few more types of sounds than the monogamous crabeater has.

Very little is known about the ecology of the seal that bears the name of James Ross, British captain of the Terror and Erebus (ships that in 1841 explored the area of Antarctica now called the Ross Sea). The Ross seal eats a mixed diet of cephalopods, fish, and krill. It is often called the singing seal because it characteristically lies on the ice with its head up and mouth open. When I have tried to record a Ross seal in this “singing” posture, however, I have never detected a single sound. I think that the posture is a way for the seal to display its distinctively striped throat and that the open mouth is a threat, perhaps to other Ross seals or to predatory leopard seals. A handful of underwater vocalizations have been recorded, but just how many different sounds this species makes is not yet known. There have simply been too few observations of this seal in the wild, although during one of our cruises our group did document—for the first time—the birth of a pup. We know that males and females do not form bonds in the spring, as crabeaters do; nor do they congregate in large colonies, as Weddell seals do. They are probably promiscuous, like the leopard seals. I expect the Ross seal will turn out to have an intermediate number of vocalizations used for intraspecific communication, but not so many that they will attract predators. I hope to return to the Antarctic someday and, with luck, perhaps help unravel the secrets of this rarest seal of the south polar seas.

Copyright © Natural History Magazine, Inc.

Return to Web Site Archive, Picks from the Past