Every large sirenian once grazed on algae along the shores i of the cold North Pacific Ocean. Shipwrecked with Vitus Bering’s expedition on Bering Island, the larger of the Kommandorskiye (Commander) Islands, George Wilhelm Steller (b. 1709-d. 1746), the only biologist to observe the species alive before its extinction ca. 1768, observed and described the sea cow and recorded his observations on external and internal anatomy, natural history, and hunting as seen during his enforced stay on the island between November 1741 and August 1742.
I. Characteristics and Systematic Relationships
Two other common names for Steller’s sea cow are the German “Borkentier,” referring to its rough bark-like hide, and the Russian “kapustnik” (cabbage eater), which Steller attributed to inhabitants of Kamchatka, where dead sea cows reportedly washed ashore after storms. Because Steller’s posthumously published observations (1751) predated Linnean taxonomy, the sea cow was given its formal scientific name by later authors on the basis of Steller’s description (no type specimen has ever been designated). Among numerous synonyms, the generic and species names now recognized as correct are Hydrodamalis (Retzius, 1794) and Hydrodamalis gigas (Zimmerman, 1780). The latter combination was authored by Palmer (1895). The generic synonym used most commonly in the 19th century was Rytina (Illiger, 1811) (improperly emended to Rhytina). For full synonymy and comprehensive references, see Domning (1978, 1996).
Hydrodamalis gigas was a very large toothless sirenian. Sire-nians are wholly aquatic, primarily vegetarian, mammals. H. gigas was the only sirenian adapted to cool-temperate or cold climates and an algal diet. It differed from all other mammals in feeding predominantly or exclusively on algae and in having a manus composed only of carpals and metacarpals, with the phalanges vestigial or absent. Its pectoral limb had a densely bristled blunt termination that Steller described as claw-like or hoof-like (Fig. 1). It differed from its predecessor (Hydrodamalis cuestae) in lacking teeth and from most fossil and living sirenians other than manatees in having a reduced deflection of the rostrum from the occlusal plane (Fig. 2). This characteristic correlated with feeding in the water column above the bottom. Its uniquely large size may have been a factor in its ability to tolerate cold. Surface water temperatures around the Kommandorskiyes range from 0°C in winter to 10°C in summer.
Like other sirenians, sea cows had paired nostrils, abdominal testes, a pair of axillary mammae, no hindlimbs, horny plates on the occlusal surfaces of the rostrum and mandibular symphysis for macerating vegetation, and ribs and other bones that were swollen (pachyostotic) and completely compact (osteosclerotic). It swam by dorsoventral undulation of the body and its horizontally expanded triangular caudal fluke, resembling those of cetaceans. See Domning (1978, 1994) for more detailed morphological diagnoses.
The order Sirenia is agreed to be monophyletic and has been placed in the superorder Tethytheria, which includes the Proboscidea (elephants) as well as an extinct group of marine herbivores, the Desmostylia. The cladistic analysis by Domning (1994), based on morphological characters of skull, mandible, and dentition, supports the conclusion that H. gigas was the last of two species in the genus Hydrodamalis, sharing the subfamily Hydrodamalinae with its ancestral genus Dusisiren. The Hydrodamalinae, Dugonginae, and Halitheriinae constitute the family Dugongidae. The Sirenia also include three other families: the extant Trichechidae (manatees) and the extinct Prorastomidae and Protosirenidae.
II. Distribution
Historically, the sea cow was known only from Bering and Medney (Copper) Islands at the western terminus of the Aleutians, 200 miles to the west of Attu Island and 150 miles east of the Kamchatka Peninsula. Pliocene and Pleistocene fossils demonstrate that the distribution originally extended around the North Pacific from Baja California, Mexico (30°N), along the coast of North America to the Aleutian chain and south in the western Pacific to Honshu, Japan (37C>N). No evidence has been adduced for geographic variation. The northernmost Pacific was likely marginal habitat as members of the Kommandorskiye population may have been stunted. Adults at Bering Island had a body length ca. 750 cm (25 ft.) and a weight of 4500-5900 kg (10,000-13,000 lbs.). Fossil material from farther south suggests maximum body lengths may have reached 9-10 m.
Figure 1 A reconstruction of Steller’s sea cow by Pieter Arend Folkens. Outstanding features are the relatively small head, bulky body, rough skin, ivhale-like flukes, and blunt unflipper-like pectoral appendages.
III. Natural History and Behavior
Bering Island is approximately 50 miles in length and Med-ney Island 35 miles long. Both are narrow, high, and rock)’. Sea cows foraged on softer parts of large marine algae (“kelp”) on open coasts in rocky subtidal and intertidal waters. Steller referred to four types of macroalgae (all still undescribed in his day) as preferred foods; these have been variously identified but probably include such forms as Agarum spp., Maria prae-longa, Halosaccion g}andiforme, Laminaria saccharina, Nereo-cyctis luet.keana, and ThalassiophyUum clathnis. Tougher steins and holdfasts were not eaten, washing up in heaps on the shore where sea cows fed. Seagrasses have also been suggested as elements of the diet, but this is doubtful. Of the two available forms, Phyllospadix spp. are among the toughest seagrasses and Zostera marina grows on soft bottoms in protected waters. In any case, H. gigas was probably dependent on a small number of forage species. Feeding along the shoreline, with backs above water, the sea cows were not observed to submerge; they moved about using their flukes and maintaining position on the rocks with their bristle-tipped pectoral limbs. Algae growing on rock faces were bitten off or pawed free with the forelimbs, and the softer parts were separated by the bristly lips as if “cut off with a dull knife.” If the animals were indeed unable to sub merge, they could access food only to depths of about a meter below low tide level. Steller wrote that they fed “in herds.” keeping the young in the center of the group. He did not specify herd size but referred to family groups as consisting of a pair with their offspring. In addition to rocky shorelines, Steller stated that sea cows were “fond of shallow sandy places along the seashore, but they like especially to live around the mouths of river and creeks, for they love fresh running water.”
Sea cows had no fear of humans. Feeding with the head submerged and half of the back above water they were easily approached and could be touched or speared from the rocks. One method of capture was for a swimmer or a boat crew to approach the intended victim with a large hook attached to a line to the shore. When the hook had been driven into the animals body, a shore crew would attempt to drag it toward the beach. Steller reported that sea cows made no sounds even when wounded by hunters. Of particular interest are the descriptions of both helping behavior and evident pair bonding in this context. Steller reported that while an animal was being dragged ashore other sea cows would attempt to dislodge the hook and/or break the rope. When one female had been dragged ashore, the accompanying male kept station offshore at the site for at least 2 days.
On the basis of his observation of “family groups” and the evidence of strong pair bonds, Steller suspected that the animals were monogamous. Mating apparently involved only a pair, with the male following or herding the female during prolonged “amorous preludes.” Copulation involved mutual clasping with the pectorals, the female in an inverted position and the male above.
Figure 2 Skulk of the two recent dygongids drawn to the same scale. The Hydrodamalis gigas skull (left) shows the lack of teeth and the relatively undeflected rostnim compared with the downwardly deflected rostrum of Dugong dugon (right). The dugong’s tusks and molar teeth are indicated by arrows. The sea cow’s head was relatively small compared with its body and that of the dugong relatively large so that the size of the two skulls does not reflect relative body size.
Steller’s views as to the annual reproductive cycle present some enigmas. He observed mating behavior and copulation in the early spring, and deduced that gestation was more than 12 months, but stated that young were born “at any time of the year, but most frequently in autumn.” Fall births would seem likely to put the young at a disadvantage, especially in view of his observations that “during the winter they are often suffocated by the ice that floats about the shore” and “in the winter the animals become so thin that, besides the bones of the spine, all the ribs show.” As Steller did not arrive on the island until late November and left the following August, he could not have had direct knowledge of a concentration of births in the fall. He never saw more than one very small calf with a female, implying that births were single.
As an algivore, Hijdrodamalis may have been dependent on an ecosystem in which the carnivorous sea otter (Enhydra lutris) suppressed invertebrate herbivory in shallow coastal waters. Like the sea cow, the sea otter is confined to the shallows within 1 km of shore where it forages for urchins and other invertebrate herbivores, usually to a maximum depth of about 130 ft. (40 m). Within this zone, otters keep sea urchins and other invertebrate algivores in check. Hydrodamalis was dependent on fast-growing kelps on rock faces and on the kelp canopy in deeper waters. Wliere sea otters are extirpated, grazing by urchins results in “kelp barrens.” Thus sea cows depended on sea otter predation that maintained the shallow-water kelp community.
Steller reported both external and internal parasites. The crustaceans that he described as infesting the sea cows’ skin have been interpreted as cyamids or caprellids, but identification will probably remain uncertain. The only specimens purported to exist are ones attached to alleged sea cow skin fragments, but these latter are likely to pertain instead to whales. “White worms half a foot long,” observed by Steller in the stomach and intestine of the sea cow, were probably ascarid nematodes; both these and the ectoparasites probably went extinct with the sea cow itself. Hydrodamalis, at least the young, may occasionally have been prey for killer whales (Orcinus orca) and large sharks.
IV. Evolutionary Origins
Hydrodamalis evolved from tropical and subtropical, seagrass-eating dugongid ancestors that shifted from bottom feeding to surface feeding and from a seagrass in warm protected waters to an algal diet and a high-energy, low-temperature environment. Seagrasses, which have a large part of their biomass below ground in the form of roots and rhizomatous storage organs, are the preferred forage of the extant dugong (Diigong dugon) and apparently of the ancestral dugongid line. Kelps have no below-ground biomass. Their softer, more edible, growing portions (the bulk of nutritious biomass of the taller kelps) are at or near the surface, suspended by floats. Fossils record this transition from bottom rooting to surface foraging. Hydrodamalis descended from the widely distributed tropical genus Metaxytherium, a small-tusked, seagrass-eating Miocene halitherine dugongid. At the time a diversity of herbivorous marine mammals foraging along the east-central Pacific shorelines was presumably supported by a diverse flora of tropical and subtropical seagrasses, as well as algae (perhaps including kelps in cooler, more exposed microhabitats). The sister group and apparent immediate ancestor of hydrodamalines was the species M. arctodites, living in southern California and Baja California by about 14-15 Ma. Probably already present were the earliest members of its descendant genus Dusisiren, as well as a dugongine, Dioplotherium allisoni. These three sym-patric sirenians, representing the three dugongid subfamilies, probably occupied distinct feeding niches. The large-tusked Dioplotherium with its strongly deflected rostrum was likely a bottom feeder, rooting for rhizomes of the larger seagrasses. The small-tusked Metaxytherium probably depended on the smaller rhizomes and leaves of seagrasses. Dusisiren and later hydrodamalines, in contrast, progressively gave up bottom and rhizome feeding, specializing on kelps growing higher in the water column, reducing their rostral deflections, and losing their tusks. Also present with these Miocene sirenians were three or more genera of desmostylians, hippopotamus-like bottom feeders, presumably also herbivorous and feeding on intertidal and subtidal seagrasses and kelps.
As the climate cooled after the Middle Miocene, and tectonic uplift rendered the western North American coastline more emergent (replacing protected embayments with more exposed, higher-energy habitats), the cold-water plants increased their dominance in the marine flora, and the tropical seagrasses finally disappeared altogether. With them went Dioplotherium, Metaxytherium, and all the desmostylians. By the Late Miocene, Dusisiren, having earlier thrown in its lot with the kelps, had increased its cold tolerance and its morphological specialization for kelp eating (passing through the successive evolutionary stages D.jordani and D. dewana) and had extended its range to the northwestern Pacific. By the end of the Miocene it had evolved into Hydrodamalis ctiestae, which gave rise in the Pleistocene to H. gigas.
The prehistoric distribution of sea cows and sea otters (E. lutris) was coterminous. Anderson (1995) speculated on the evolution of a sea cow-sea otter relationship and sea cow morphology as follows: “As the hydrodamaline range shifted northwards, the adoption of a kelp diet may have contributed (along with the thermoregulatory demands of a cooling environment) to selection for large adult body size, high birth weight, prolonged parental care, and low reproductive rate. At the same time, the more effective chemical defenses of kelps growing below the sea otter foraging zone might have favored the sea cow’s specialization for surface feeding and its loss of diving ability.”
V. Extinction
The vulnerability of the sea cows in their last stronghold made it possible for the weak and scurvy-ridden castaways of Bering’s expedition to secure an abundant supply of food and to escape to the mainland in the summer of 1742. Having found ways to capture the sea cows, the survivors were able to divert manpower to salvage materials from the wreck of their ship and build a smaller vessel in which to reach Kamchatka. That voyage, made possible by sea cow vulnerability, carried with it the news of fortunes to be made in hunting sea otters and fur seals. A fur rush followed, fueled by sea cow meat. The first hunting expedition wintered on Bering Island in 1743-1744. By 1763 several parties had spent up to 9 months on the islands, living almost exclusively on sea cows and salting down barrels of meat to provision the 2- to 3-year expeditions to the Aleutian chain and the north Asian and North American coasts in search of furs. Sea cow hides were also used to make large skin boats (baidarkas or umiaks). In 1754-1755 a fur-hunting expedition was forced to winter on Bering Island because the sea cows had been extirpated on Medney Island. The last specimen on Bering Island was reported killed in 1766. Steller’s biographer, Leonard Stejneger, summarized accounts of the unrestrained killing by the fur hunters and attributed the extinction of the sea cows to ruthless slaughter.
These magnificent animals would have been at least as vulnerable to healthy and skilled Pleistocene hunters along mainland shores as they were to the weakened Russians on the Kom-mandorskiyes. Domning (1978) has suggested that human hunting may have extirpated sea cows from areas within reach of aboriginal hunters, leaving only the population around the remote Kommandorskiyes. Over most of its range the sea cow may have been the only marine mammal to succumb to “Pleistocene overkill.” The final drama played out on Bering and Medney Islands suggests a complementary and more complex and instructive story. Anderson (1995), following a suggestion by Delphine Haley, proposed that the final extinction of the sea cow resulted not solely from ruthless harvesting, but from a cascade of events beginning with extirpation of the local sea otter population around the islands in the first rush for furs. Decimation of the otters in all probability triggered a sea urchin population explosion and the disappearance of chemically undefended shallow water kelps that were the sea cows’ main food supply. Invasion of the shallow waters by chemically defended deep-water kelps left the hunted remnant of the sea cow population with kelp that was likely toxic. Anderson proposed that this may have been a reenactment of events when North Pacific coastlines were first colonized by humans in the waning Pleistocene.
