How Toothed Whales Feed by Sean Chamberlin

Whereas mysticetes feed on plankton, the odonotocetes prey on nekton, primarily large crustaceans (shrimp or decapods), squid and fish. Some species, like the killer whales include sharks, other marine mammals and seabirds in their diet. Like baleen whales, toothed whales exhibit feeding morphologies and behaviors that suit their particular habitat and prey choice. They also appear capable of learning new feeding strategies and modifying behavior to overcome shifts in prey type and abundance. However, the diversity of odontocetes and their more active predatory lifestyle make them more difficult to study. A great deal remains to be learned about their feeding behavior.

Because teeth play a prominent role in the feeding of toothed whales, a brief discussion of tooth morphology and number is warranted. Unlike humans, odontocetes develop one set of teeth for their entire life. They also tend to be similar in shape and size—mostly peg-shaped like the rounded cap on the end of some ballpoint pens—and they tend to exist singly with open roots and spaces between them. Scientists who study odontocete teeth typically classify them based in the number that occur on the upper and lower jaw of the animal. Thus, a dolphin with 30 teeth on the upper jaw and 30 teeth on the lower jaw would be indicated as 30/30. Many odontocetes have a greatly reduced number of teeth or no teeth on the upper jaw. The Risso’s dolphin has two tooth on the lower jaw (0/2). The sperm whale has no teeth on the upper jaw and 25 teeth on the lower jaw (0/25). Porpoises (Phocoenidae) typically have even numbers of teeth on both jaw (from 15/15 to 30/30) while dolphins vary widely (up to 65/58 in the spotted dolphin). The shape of the teeth is one of the few ways to distinguish between porpoises and dolphins: in porpoises, the teeth are spadelike, in the dolphins they are conical.

Somewhat surprisingly (at least to these authors), the teeth of odontocetes are better adapted for grasping and holding than tearing and slashing (like shark teeth). The notable exception are killer whales whose teeth along the upper and lower jaw interlock when the jaw is closed and whose shape enables them to rip and tear the blubber of marine mammals on whom some killer whales feed. However, their teeth do not allow them to chew so pieces of flesh must be swallowed whole where they are processed in their complex of multiple stomachs.

One possible explanation for the reduced number of teeth involves the type of prey and perhaps the swimming dynamics of odontocetes. The beaked whales and sperm whales—those species with the fewest teeth—feed largely on squid. These species are well known for suction feeding in which the animal generates a flow of water through its mouth by expanding its oral cavity. As the animal approaches its food, the suction mechanism creates a flow of water that sucks the prey into the mouth or, at the very least, prevents the bow wave of the swimming animal from pushing the food away. A wide number of aquatic animals, including sharks, ray-finned fishes, turtles and salamanders display this type of mechanism and its presence may be convergent (having evolved independently) in these species as a solution to feeding in water (see Summers et al., 1998). Suction feeding has been hypothesized for beaked whales, harbor porpoises, the narwhal, sperm whales and other dolphins (see Berta and Sumich 1999 for discussion and references) and may be generalized for feeding on squid and some fishes. The arrangement of the teeth permits water to flow in a manner that enhances suction feeding or that simply permits the animal to close its jaw without squirting the prey from it. In this way, the teeth may act like the bars of a cage to trap the prey (e.g. Norris and Møhl 1983).

Another possible explanation for the homogeneity and spacing of teeth in some odontocetes concerns their mode of finding food through echolocation. Some scientists have proposed that teeth spacing and jaw geometry play a role in the reception of sound during echolocation (e.g., Goodson and Klinowska 1990). Numerous studies have demonstrated that the lower jaw acts as a receiver for sound during the echolocation of dolphins (see references in Aroyan 2001). The role of the teeth in dental reception of sound is less definitive, although studies support at least some aspects of this hypothesis (e.g., Potter and Taylor 2001). The general idea involves sensing the very small time differences in the received acoustic signal at each tooth and processing that signal into information about the reflected object. (For more information, see spotlight on echolocation.) The absence of teeth in the upper jaw of some cetaceans and the reduction of teeth in others allows the possibility that teeth may be important for a function other than feeding. Is it possible that echolocation has driven the evolution of teeth in cetaceans?

In addition to finding food, sound production in cetaceans has been hypothesized as a means for catching food by stunning or disorienting prey, a type of feeding called acoustic predation (e.g. Martin et al 2001). The sperm whale, in particular, with its reduced and occasionally deformed lower jaw, has long been a puzzle for its mode of feeding. Thomas Beale, a British surgeon who spent time with the South Sea whaling industry in the 1800s, noted in his oft-quoted booklet that “the teeth of the Sperm Whale are merely organs of prehension, they can be of no use for mastication, and, consequently, we find that the fish, &c. which he occasionally vomits, present no marks of having undergone that process” (Beale 1835). Two separate whaling expeditions in the 1950s found entire giant squid—at least 34 feet long—in the stomachs of captured sperm whales (see Ellis 2001). Even live squid have been found in the stomachs or observed swimming from the stomachs of lanced sperm whales (e.g. Norris and Møhl 1983). A. A. Berzin, who published an extensive study of the sperm whale in 1971, observed several sperm whales whose stomachs were full or near full despite the fact that their jaws were “turned aside already at the very base, or turned inside out with the tooth row outward, or else rolled up in a ring” rendering them useless for even holding prey. Berzin offered the idea that sperm whales use acoustic stunning to subdue their prey. The intense and pulsed nature of sperm whale sounds may be consistent with that hypothesis and explain the presence of live prey in the stomachs of sperm whales.

Based on these observations, Ken Norris and Bertel Møhl (1983) hypothesized that “some odontocete cetaceans may emit sounds so intense that they prey is debilitated and capture made easier.” The acoustic predation hypothesis has at least two fundamental and simultaneous requirements: 1) that cetaceans can produce sounds of an intensity and frequency capable of disorienting or stunning prey; and 2) that prey hearing and/or tissues respond to or are negatively affected by sounds in the intensity and frequency emitted by the cetacean. Citing their own observations and other studies, Norris and Møhl provided strong evidence that both of these conditions could be met. More recent studies support the ability of cetaceans to produce broad bandwidth and intense sounds (e.g. Au and Wursig 2004 and references therein). The debilitating effects of sound on anchovies and salmon have been demonstrated (e.g. Marten et al., 2001) as well as the ability of numerous species, including chad, herring and tuna to “hear” echolocation signals (Mann et al., 1998; Wilson and Dill 2002; Finneran 1999).

Of course, direct observation of a cetacean sound resulting in a stunned and consumed prey would also support this hypothesis. Towards this goal, researchers working in a fjord in Norway simultaneously deployed a video camera and a hydrophone—whose output was directly recorded on the videotape—to observe killer whales feeding on herring. Like humpback whales and other cetaceans, killer whales have been observed to engage in cooperative hunting, groups of individuals working together to locate, herd and consume prey. One type of cooperative behavior observed in killer whales is carousel feeding, where killer whales work together to encircle and dive beneath their prey to drive them into a ball (a dense school) at the surface. The carousel feeding observed in this study, while complex, involved two phases: 1) school (or prey) herding, in which the whales circle their prey and use acoustic signals (vocalization) to manipulate the prey into a ball and possibly to communicate and coordinate the hunt; and 2) feeding, during which the whales employ tail slaps—literally slapping the fish with their tail—to stun their prey and lunging behavior—accelerating rapidly—to maintain the ball of fish. Tail slaps that resulted in a stunned herring generated a loud noise whereas tail slaps that missed fell silent. Based on these observations, physical contact delivers the blow that impairs the herring and the loud “bang” results from that contact. In other words, acoustic predation does not operate, at least in this instance. While it is possible that previous observations of loud noises during feeding and the appearance of stunned fish (not observed simultaneously) could be explained by the physical contact hypothesis, too few observations have been made to determine whether the acoustic predation hypothesis should be rejected (see Domenici et al., 2000).

A possible alternative to acoustic predation is acoustic manipulation (e.g., Janik 2000). On the basis of a study of low frequency bray calls—whose sound resembles the bray of a donkey—by wild bottlenose dolphins (Tursiops truncatus), Janik concluded that dolphins may manipulate the behavior of their prey using these sounds. Bray calls have been associated with energetic social interactions in addition to feeding (Brito et al., 2000) so their specificity for feeding may not be general. However, prey manipulation using sound may be consistent with carousel feeding in killer whales and dolphins. Observations on captive bottlenose dolphins suggest that they may be able to disorient their prey with click trains (Hult 1982). The detection of dolphin sounds by American shad (Plachta and Popper 2003; Mann et al., 1998) and herring (Wilson and Dill 2002) may be exploitable for herding these fish during feeding. This hypothesis may also explain why bray calls attract other dolphins (i.e., to assist in prey herding) and possibly yellowfin tuna (e.g., Schaefer and Oliver 1998; Finneran et al., 1998), who often feed alongside dolphins. Acoustic manipulation of prey may represent one of a spectrum of possible uses of sound during feeding. The evidence suggests that cetaceans may find, identify, manipulate and even possibly disorient or stun their prey using a broad spectrum of sound frequencies, types and intensities.

Killer whales typically feed on fish but in some regions of the world, they prey upon marine mammals, including seals, sea lions, sea otters, dolphins, porpoises, and gray and blue whales, among others. Populations that feed on mammals represent a distinct ecotype, a sub-population of a species associated with a particular ecological role or habitat. In killer whales, these ecotypes take the form of distinct pods, groups of individuals that form strong associations with each other (see also cetacean behavior and social organization). By using photographic and other techniques that allow scientists to identify and track individuals in pods, scientists have been able to discern the behavior and social systems of these social groups, especially in the Pacific Northwest and the coastal waters of British Columbia. In these waters, three distinct types of populations are recognized: 1) resident pods, who typically form stable, multi-generational, matriarchic societies of females and their offspring, males and female; 2) transient pods, made up of loose associations of mother and offspring or several females, and occasionally adult males; and 3) offshore pods, about whom less is known except that they travel along the Pacific U.S. coast from Alaska to California (Hoelzel at al., 1995). Pods do not appear to mix and may be genetically distinct. In addition, pods maintain distinct dialects, a pattern of vocalizations that is unique (but changeable) to a particular pod. Relevant to our discussion here, resident populations appear to strictly eat fish while transient populations consume strictly marine mammals. Because of their vocalizations (see below), offshore populations were hypothesized to be fish eaters. That hypothesis that appears to be supported by observations of fish in the stomachs of two offshore killer whales in Alaska (Heise et al., 2003).

Transient killer whale pods that feed on marine mammals employ stealth or sneak attack feeding, a behavior in which echolocation and other acoustic signals are minimized, presumably to avoid detection by their acoustically sensitive and fast-swimming prey. These killer whales produce short and isolated clicks in unpredictable patterns that may be masked (made less conspicuous) by background noises, a phenomenon known as acoustic crypticity (Leonard et al., 1996). These silent types likely rely on hearing to find prey and produce sounds only when “killing or eating prey” (Ford et al., 2000). Feeding on mammals requires a different set of challenges than feeding on fish, reflected in the often brutal behavior of killer whales engaged in a hunt. Killer whales may ram their prey or strike them with their tail flukes repeatedly in an attempt to exhaust or drown the victim. Cooperative hunts on gray whales take on particularly gruesome aspects as a pod may pursue a mother and calf for hours, using a variety of techniques, such as ramming or biting at their fins and flukes, to separate the pair. Mother gray whales have been observed to roll on their backs to provide a safe temporary platform for the calf but this behavior only delays the inevitable. Once the calf is separated from her mother, it is quickly subdued, drowned and eaten (Ternullo and Black 2002). Killer whales have been observed to attack a 60-foot blue whales and jump on the back of the victim to slow it down or drown it. They have also been observed to attack and kill a great white shark (e.g. http://www.seaworld.org). In some instances, killer whales will only disable a prey to demonstrate a feeding technique to a juvenile, an example of social learning. The tossing of prey observed in Patagonia populations who prey on elephant seal pups or Alaska populations who prey on porpoises may be related to social transmission of information. Such displays often end with the juvenile killing the victim.

Clearly, odontocetes display a wide range of complex behaviors associated with feeding. As a result, their impact on marine ecosystems is difficult to assess quantitatively. Nonetheless, as an apex predator, odontocetes likely play a large role in regulating the abundance and structure of marine food webs. Aggregations of seabirds and other species (including yellowfin tuna, dolphins and sea lions) with humpback whale feeding “frenzies” serve as one example of the way in which cetacean behavior drives mass consumption of prey (e.g. Black 1999). Some scientists have suggested that switching of prey by killer whales from Stellar sea lions to sea otters may be responsible for the dramatic decline in Alaskan sea otter populations since the 1990s (Estes et al., 1998). Declines in Stellar sea lion populations since the 1980s, either as a result of overfishing, killer whale predation, or other causes (see NRC Report 2003), have presumably caused killer whales to seek alternative prey like the sea otter. While limited direct evidence exists, calculations based on (among other factors) the dietary requirements of killer whales and the presumed number of transient, mammal-eating killer whales present in the Aleutian Islands suggest that killer whale predation could be responsible for the observed sea otter declines (Springer et al., 2003). These scientists hypothesize that shifts in prey selection from whales to smaller marine mammals may be linked to declines in whale populations as a result of whaling prior to World War II. They note that a similar link between declining southern elephant seal populations and killer whale predation as a result of whaling was proposed in the Southern Ocean in the 1970s (e.g., Barrat and Mougin 1978).

The effects of whaling may have been global in extent and may persist to this day. The role of behavior is often not considered in models of marine ecosystems. Studies on Galapagos sperm whale populations following the cessation of whaling in 1981 demonstrated that these whale populations shifted their habitat further east in response to whaling and may not return to their original location (Whitehead et al., 1997). Thus, like shifts in killer whale behavior, the behavior of other cetaceans may alter marine ecosystems, at least on a local scale. Further research on the ecological role of whales may provide a better understanding of the role of animal behavior in the structure and dynamics of marine ecosystems.

Note: Potter and Taylor note that “it seems strange Tursiops truncatus [bottlenose dolphin] does not also have some flat teeth for grinding, as other predators do, to make digestion easier. There appears to be some cost to being a predatory homodont, and this suggests there must be some compensatory benefit.” These authors imply that dental reception or dental focusing of acoustic waves to improve echolocation is that benefit.