Probably everyone is familiar with the walrus, but did you know that it generates a vacuum in its mouth to suck clams out of their shells? And this is just one example of suction feeding, the feeding mode typically used by bony fish…

Feeding is, of course, essential for animals to survive. There are several different feeding modes, but amongst the teleosts, suction feeding is the norm. Here, the fish shapes (and often protrudes) its mouth and then rapidly expands its oral cavity via the action of the hyobranchial apparatus, thus creating a negative pressure. When the mouth is opened, water rushes in, drawing in the prey item from a distance. This mode of feeding, which is only effective in an aquatic environment, has evolved independently in several other groups. Not only in the sharks, which are usually thought of in terms of oceanic chain saws, but also more sedately in salamanders, the tadpole of a frog, a number of reptiles and even some aquatic mammals. The basic motor pattern is similar across these animals, maybe as a result of hydrodynamic constraints acting on aquatic prey capture, but certain features can differ significantly.

Suction feeding in sharks and rays

Within the sharks and rays, suction feeding has arisen a number of times. In the group known as the carpet sharks (Orectolobiformes), which includes the bamboo sharks (Hemiscylliidae) and the nurse shark (Ginglymostomatidae), suction feeding is the rule (one exception is the filter-feeding whale shark Rhincodon typus). It shows a number of striking convergences with the teleosts, employing very much the same functional principles (e.g. a circular and terminal mouth with aperture control and a very rapid mandibular depression), but based on a markedly different anatomy. In the nurse shark (Ginglymostoma cirratum), the more vigorous of the suction activities are accompanied by audible popping sounds, which were speculated to be the result of cavitation (explosive collapse of air bubbles). If this is correct, this is one of the few examples of this type of high-energy physics in biology, linking it convergently to alpheid and mantid shrimps.

Other suction feeders amongst the sharks are the bullhead sharks (Heterodontiformes) and the chain catshark (Scyliorhinus retifer). The latter is a rather unusual example, as it belongs to a family dominated by ram feeders, the Scyliorhinidae, and shows no distinct anatomical adaptations to suction feeding. However, the chain catshark is a deep-water species that captures prey on the sea floor (as opposed to the ram-feeding scyliorhinids that hunt in the water column), and here suction feeding is particularly effective. It has also evolved in another group of bottom-dwelling cartilaginous fish, the rays, where it seems to be the primitive feeding mechanism. As Philip Motta and colleagues note: “Therefore, specialisation for inertial suction feeding including rapid jaw opening, formation of a round, somewhat terminal mouth, prominent labial cartilages, a dentition reduced in size, and generation of large subambient suction pressures has apparently evolved independently in conjunction with a benthic lifestyle in several elasmobranch lineages” (Motta et al. 2002, Copeia, vol. 1, p. 36).

Suction feeding in amphibians

Whilst adult frogs typically gulp prey or use highly protrusible tongues (a form of feeding that has arisen convergently in such groups as the chameleons), the tadpoles usually feed by grazing the substrate of algae and similar growth. However, the dwarf African clawed frog (Hymenochirus boettgeri) has independently evolved a system of suction feeding that is strikingly similar to that employed by teleosts. It uses effectively identical mechanisms of protruding the mouth rapidly and then enlarging the mouth cavity by head movements, elevating the cranial region and hyobranchial depression. Not surprisingly, this tadpole is a predator and has conspicuously large forward-pointing eyes.

The aquatic larvae of salamanders also employ suction feeding, having arrived at this mechanism independently. In the majority of species, the larvae metamorphose into terrestrial adults and undergo major changes in their feeding apparatus, with the tongue becoming an essential part. Some species, however, remain fully aquatic throughout their life, and here the adults feed by suction as well (e.g. in the neotenic axolotl Ambystoma mexicanum and the common mudpuppy Necturus maculosus). Two of these species, the hellbender (Cryptobranchus alleganiensis) and the Japanese giant salamander (Andrias japonicus), which are both large and rather primitive, are particularly remarkable. Their feeding apparatus is unique among suction feeders in structure and behaviour, because it allows for asymmetric suction feeding that involves asymmetric jaw movements – they can essentially open only one side of their mouth. However, the basic motor pattern is similar to that recorded in other suction-feeding vertebrates.

Suction feeding in reptiles

One example of suction feeding in reptiles occurs in some aquatic turtles, such as the freshwater species mata mata (Chelus fimbriatus), chicken turtle (Deirochelys reticularia) and narrow-bridged musk turtle (Claudius angustatus). This is particularly interesting, as aquatic turtles are probably descended from a terrestrial ancestor and have thus modified their anatomy from one employed for terrestrial feeding. Another example is the Australian snake-necked turtle Chelodina longicollis. Snake-necked turtles capture agile prey, employing a combination of fast neck extension, bucco-pharyngo-oesophageal expansion and retraction of the head. Therefore, as Johan van Damme and Peter Aerts note, “in spite of largely different structural solutions, optimal feeding conditions as deduced for suction in feeding fishes are also employed by Chelodina. This further promotes the assumption that hydrodynamics constrain evolutive solutions for aquatic feeding” (van Damme & Aerts 1997, Journal of Morphology, vol. 233, p. 113), making this an excellent example of convergence. Interestingly, suction feeding has also recently been reported from a turtle that had been assumed to be fully terrestrial, the Indochinese box turtle (Cuora galbinifrons). This is intriguing, because suction feeding and terrestrial feeding require a rather different anatomy. The structure of the box turtle’s feeding apparatus basically reflects a compromise between the two modes – while the shape of the palate and the tongue design resemble that in terrestrial feeders, the construction of the hyoid complex is more consistent with aquatic feeders. However, suction feeding is only used for the transport of small prey items in water, whereas larger objects are moved by the tongue, indicating that the generated suction forces might not be strong enough.

Another potential instance of suction feeding in reptiles involves a rather extraordinary group of Triassic marine reptiles with exceptionally long necks, the protorosaurs. It has been speculated that one species (Dinocephalosaurus orientalis) also fed by suction. The negative pressure could have been generated by a rapid straightening of the neck and simultaneous expansion of the cervical ribs, which would have resulted in an increase in the oesophageal volume. However, this idea has been challenged.

Suction feeding in aquatic mammals

In mammals, suction feeding has evolved independently in several lineages of pinnipeds and cetaceans. The mammalian master of suction feeding is probably the walrus (Odobenus rosmarus). Walruses can generate strong negative pressures in their oral cavity, supported by a short, wide muzzle, a piston-like tongue and a uniquely vaulted palate, to suck bivalves (especially clams) out of their shells. For excavating their benthic prey, they employ the opposite of suction, hydraulic jetting, where they force water out of their mouth. A robust, strongly fused mandible helps to counteract the forces exerted during feeding. Bearded seals (Erignathus barbatus), which feed on diverse benthic prey such as crustaceans, molluscs and worms, possess suction capabilities similar to that of the walrus. In a series of striking parallels to other suction feeders, they form a circular mouth, with the aperture controlled by the lateral lips, combined with a rapid depression of the tongue area. Furthermore, ringed seals (Pusa hispida), crabeater seals (Lobodon carcinophagus) and harp seals (Pagophilus groenlandicus) are also presumed to feed by suction.

Among the cetaceans, some toothed whales (Odontoceti) employ this feeding mode. Many odontocetes use suction to some extent, but only some are considered true suction feeders. These include sperm whales (Physeteridae and Kogiidae), beaked whales (Ziphiidae), the narwhal (Monodon monoceros), the beluga (Delphinapterus leucas), some delphinids – pilot whales (Globicephala) and Risso’s dolphin (Grampus griseus) – and possibly porpoises (Phocoenidae). They are adapted to suction feeding by having a reduced number of teeth as well as blunt heads and mandibles. The bony hyoid apparatus is particularly important for generating suction, along with the associated musculature and ventral throat grooves. It has been suggested that true suction feeding evolved once in cetaceans, early in their evolutionary history. Intriguingly, one species of baleen whale, a group of filter feeders, might also use suction feeding. The gray whale (Eschrichtius robustus) has been observed to dive to the ocean floor, draw in invertebrate prey and force out water and sediment through the baleen. However, other researchers have described different feeding techniques for this species, and gray whales possess only some of the osteological characters typical of suction feeders.