Sea Mammals: Carl E. Abegglen* U. S. Fish and It'ildlife Service, Division of Resources and I\'ildlife Research, Anchorage, Alaska Population

The nanrine mammal resources nenr Amchitkn Island consist from near extinction at the start of the twentieth century. of sea otters, harbor seals, and Steller sea 1io11s as Conservation measures, national nnd i~ttemationnl, haue perntnnent residents, northern fur seals that migrate been many, some even hauh~gbeen started in Russian times. througla Aleutian passes, and wholes nnd porpoises in the The crucial and finally strccessftrl ones are the Fur Seal surrouttdiftg seas. Archaeological and historic data on Treaty of 1911 and tlte Exectrtiue Order of 1913, which nni~nnl populations indicate that the species present tlten estnblislted what is now known as the Aleutian Islands were the same as those present today nnd dentoxstrate tlre National Il'ildlfe Refuge. The marine m?nnral populations contii~uedimportawe that sea mammals haue played in tlre (whales excluded) around Amchitka and in the western island's history. Sen otter observations nnd surueys made Aleutian Islands are h good condition. front 1935 to 1974 document the recovery of this species

ARCHAEOLOGICAL INDICATIONS OF suggest that these grooved teeth were used for SEA MARIMALS personal decoration-as pendants for nose orna- nlents. The prehistoric people of Amchitka, in collimon Desautels et al. iuiearthed fireplaces associated wit11 the historic Aleuts, had a maritime economy with large cut wvl~alcbones. The close association and were dependent on the sea for the bulk of suggested to them that these may have been used their existence. These people also applied their as supporting beams for structures. None of these skills in ocean fishing to marine mammal hunting. bones were found in an upright position. They \\.ere expert sea~nenand traveled in groups or The animal remains from the six sites dug by family units from one islalid to another without difficulty. The practices reqi~iredfor existence on Desautels et al. in 1969 consisted of sea mammal, Amchitka and in the Aleutians iin general changed fish, and bird bones as ~vellas shells of marine very little during several millenia of human occupa- invertebrates. Site RAT 31 yielded more than

~~~ 11.000 mammal bones. Except for some human 11011..~ elements, the re~naillder\\,ere from \\,hales, pin- From their 1971 dig on Amchitka, Cook, nipeds, and sea otters. The few whale bo~les Dixon, and Holmes (1972) report the recoveiy of represented both large and small cetaceans. Species numerous sea mammal bones, which were ap- identified i~lcludedsea otters, harbor seals, Steller parently used as a\vls, casting lance heads, picks, sea lions, and a few northern fur seals. About 50% and for decoration. Among these were eight of the sea otter bones were from juveniles. From notched or grooved teeth, including four harbor the material collected, it appears that \vhole sea seal canines, one sea lion postcatline tooth, and otters were brought back to the village sites but three harbor seal incisors. Desautels et al. (1970) pi~inipedswere butchered else~vhere. Percentages recovered 25 grooved teeth from Amchitka excava- of sea mammals found in site RAT 31 generally tions in 1969. One of these \\.as perforated as well reflect historical records concerning Aleut food as grooved. Clark (1968, cited in Cook et al., 1972) preferences. Desautels et d. repolt Hrdlitka (1945) did not report finding grooved sea mammal teeth as having said that sea otters, which were available on ICodiak; however, Spaulding (1962, cited in near the shore, \\,ere killed in large numhers and Cook et al., 1972) mentioned sea lion teeth that that the Aleuts also killed many yo~ulgharbor \\rere grooved on one end from Agattu. Cook et al. seals. These statements are in general agreement with the numbers and species of the animal remains from site RAT 31. Far fewer sea lion *Now retired, at Olympia, \\'ashington. bones were found co~nparedxvitli the number of sea otter and harbor seal bones. The s~llallnumber valuable fur-bea~inganimals served as the catalyst of fur seal renlains is consistent with that species for this exploration. A Russian vessel is reported to being a migrant without established rookeries in have sailed to the Arctic Ocean from IVitus Bering sailing northeasterly fishhooks, wedges, flakers, a\vls, clubs, and oma- from ICommander Islands, and Bering died on the island wllich r\ltasov and his Tartars first secured, towards the that now bears his name. The return of his close of the seventeenth cento~y,on the Kamchatkan lieutenant, Chirikov, from the same area \\'as of coast,-had it not been for this incentive the exciting, particular importance in turning attention to those pushing, aggressive, indomitable search made by the shores. Chirikov's crew brought hack many valu- Russian 'Promishlyneks' rvottld never have been under- taken. Indeed, for that matter, much of the glory able furs, including the skins of 900 sea otters. r\.hicl~ old Titus [sic] Bering is enveloped with, as a Rumors of the richness of this nexvly discovered discoverer, was not due to his love for geography or country kindled a spirit of enterprise in the hydrography, but it was the direct stimulation of fur merchants of Siberia. Tales of Beling's companions hunters for a rich return. They backed him; they fitted oot Itis small, miserable vessels, which in light of the and crew, \\,hen they retur~ledin 1741, did even present Itour, make his soyages fairly fabulous, \\!hen Inore to increase the desire of merchants to profit the rickety, 'ram-shackle).' construction of his rough from trade in sea otter skins, and, during the Amoor-built shallops is understood. . . . relnaining years of the eighteenth century, It rvas not, therefore, until the Russians opened up merchants and ad\lenturers made numerous the trade, swiftly s~~pplementedby the third voyage of voyages eastward from Russia. Captain Cook and the aroused attention of the IIudso~l Bay Company, which speedily began to search the The first fur hunter to undertake the daring sea coasts of British Columbia and Oregon in those early voyage in search of sea otters was Elnilian Basov, a day-it \\.as not until this action \\,as taken, toward Cossack sergeant from Ieamchatka. IIe made four the close of the seventeenth centuly and the beginning voyages but only reached Bering and Copper of the eighteenth, that the sea-otter became knolvn, Islands. Later trips by others led lo the discovery first to the courts and then to the nobility of the civilized world . . . . of new islands and to surveys of the Alaska During the first few years after discovery the mainland. numbers of sea-otters taken all along the Aleutian In 1763 Shelikhov, a prominent Siberian mer- chain, and down along the ~vholenorthwest coast as far chant, led an expedition to Bering Island, as the southern boundary of Oregon, were very great, Unalaska, and Icodiak. He spent 4 years on these and, compared with what are noxu captured, seem perfectly fabulous. For instance, \rre are told when the islands alcl established his olvn trading company, Pribylov Islands \sere first disco~rered, t\vo sailors, which later formed the nucleus of the Russian- Lukannoo and Keikov, killed at St. Paul's Island during American Company. In 1790 Shelikhov employed the first year's occupation, 5,000, but the next year Alexander tlndreevich Baranof to organize and they secured less than 1,000 and six yearsafternot a manage the affairs of his company. Baranof used single sea-otter reappeared, and none have been there since. his talents to develop and increase hunting and \Vhen Shellikov's party first visited Cook's Inlet trading throughout the Aleutian Islands and main- they secured 3,000; during the second year 2,000; in land Alaska. the third season only 800, and in the succeeding year Foreign ships also visited these shores. Bcrkh they obtained 600, and finally, in 1812, less than 100, (1823, translated in Ricks, 1963, p. 38) says that and since then not one-tenth of that number, although I am told, at the date of this writing, that during the Shelikhov complained in 1787 to Govcrnor Jacobi past two years more than 500 sea otters annually have about foreigners on American shores belonging to been taken on the coasts of Cook's Inlet . . . . Russia:

Eighteenth Century explorations throughout In August of last year during my sojourn in the North Pacific arca opened up nelv regions to Petropavlovsk Ilarbor, I learned from Ji'illiam Peters of colonization and exploitation. Sea otters and other the [British East] Indian Company that a ship of their Sen ~bln?nntnls:Resotirces a11d Po/~~clotion 495 company had visited our boundaries in 1.atitude 30' on board. Sea otter hunting \\,as a cooperative activity the Northwestern Coast of America in 1785 for trading requiring from 4 to 20 hidarkas, each carrying one purposes. I do not knori. whether or not they had or two hunters. The boats \vould start out in a permission from the Russian government. They then,- selves stated that they bartered more than 800 sea wide semicircle, keeping 50 to 100 yards apart. otters and a considerable number of for-bearing land The sea otter comes to the surface to breathe at animals in a short time. It follows from the above least every 5 min; his head sometimes remains stated facts that the great profits which sl~ouldbelong visible only a few seconds. \\'hen an otter came up to the Russian citizen arc being usurped by people of \vithin casting distance of one of the hunters, he other nations \vho do not o\sn the adjoining land and have no rights in this sea. \vould throxv his dart. The otter was not likely to be killed outright by the sinall bone point of the In 1799 the Russian-American Company \\,as dart, but its mo\~ements\\,ere impeded by the dart, granted a monopoly on hunting in the Aleutian and it \\.as soon dispatched (Veniaminov, 1840, Islands area. In 1811 the Russians, under Baranof, Vol. 11, pp. 342-344). established themselves on the California coast. \\'hales \irere hunted in an entirely different Baranof was relieved of his post as manager of the manner. Instead of being surrounded, they were North American colonies in 1818. After that killed tvith poison lances. I-Iunters approached Russia's fortunes in the New \\lorld began to hales carefully from the rear, cast their spears, decline. Sickness, mismanagement, and competi- aid rapidly retreated. The lance head with its tion from the Hudson Bay Company were primary poison blade became detached from the shaft and factors in the decreasing income derived from the remained in the \vhale's flesh, causing the whale to fur trade. die after about 3 days. If the hunter was fortunate, In 1864 delegates of the Russia1 government the nrhale \vould drift ashore (Veniaminov, 1840, began to negotiate with the United States for the Vol. 11, pp. 132.134). sale of Russia's possessions in North America. Congress approved the treaty, transferring the territory in 1867, and the llext year appropriated SEA MAMMAL RESOURCES the $7,200,000, which was Russia's price for Alaska. Sea Otter (Er~l~ydralutris) \\'hen the Russians reached the Aleutians in the The only immediate government action after 1740s, practically every island Isas inhabited. the cession of Alaska to the United States in 1867 tlgattu \\.as reported to have had 31 villages and was to establish a contract with the Alaska Unalaska 24, with numerous settlements on other Coinmercial Conlpany for the regulated exploita- large islands. By 1831 there \\.ere only 10 villages tion of fur seals. It was only 17 years later that a on Unalaska. Only 15 of the other islands were local government was established in Alaska, and inhabited, and the total population is estimated to there was essentially no regulation of fur hunting have been less than 2000. and trading in the Aleutian Islands. As a result, by Aleut villages \Irere on the seacoast; the in- the time of the Fur Seal Treaty of 1911, it seemed teriors of the islands \Irere completely unoccupied that the sea otter population had been estermi- and seldom visited. No people were more de- nated and the ftir seal population \\,as in danger of pendent on the sea than the Aleuts. The land provided only a few of their needs-stones for extermination. The treaty devotes only 1 section knives and other implements, grass for ~veaving, out of 27 to the sea otter; all the rest, including the and a few plaitts for food. Everything else came \,cry title of the treaty, arc concerned ~vithfur from the sea. For most of their food they seals. depended on sea mammals, fish, sea birds, sea The beginnings of contemporary sea otter urchins, and mollusks. Their clothing \\,as nlade studies by Alnericans were informal. There were from the skins of the sea mammals and birds and intermittent almost casual pnblished references to their boats from driftxvood and skin. Implements, sea otters bet\\recn the 1911 treaty and 1935; these weapons, and household utensils Isrere made of \\,ere primarily notations that a few sea otter skins bone or drift\\rood, and material for their houses had been confiscated or surrendered and sub- seq~~entlysold at public auction. \\,as drift\vood and whale bones (I\,IcCartney, Chap. 5, this volume). Su~lreys. 1935. The persistence of reports of The Aleuts \Irere skillful sea hunters. In their sea otter observations sent to the Bureau of single- or double-hatched bidarkas (light skin Fisheries ultin~ately had its effect. Three memo- boats), they made long coastal voyages and often randa in the files of the Anchorage office of the ventured far from shore in pursuit of sea otters, U. S. Fish and \\'ildlife Service (F\\'S) arc worthy seals, sea lions, and e\.en \\,hales. Their \\,capons or mention. The first was prepared June 21, 1935, \\,ere light darts and spears cast with a thro\ving by Lieutenant Hoxvard B. Hutcllinson, U. S. NI\~, for the Commander, Aleutian Islands Survey said it was planned in due course to send a copy to this Expedition. Entitled "A Report on the Existeuce Bureau for appropriate consideration. As I remember, the report showed that about 600 sea otters were of Sea Otter in the Rat Islands," it says: observed at Amchitka Island this summer. Captain 1. The Biological Survey of the Department of the Canaga said that another report, a copy of which he Interior [Agriculture] requested the Navy Department also had planned to have sent to this office, referred to in 1932 to have the Alaskan Su~veyExpedition of that about 1,000 sea otters at another place. Captain year investigate and verify the existence of the sea Canaga expressed apprehension lest Japanese might otter (Enltydra lutris) which \\.as reported to exist in capture numbers of these valuable animals for their the western islands of the Aleutian group. The animal pelts. He suggested need of adequate protective mea- was not identified in that year by any member of the sures. I suggested to Captain Canaga that the Nary expedition, either at Attu or Kiska, although the chief Department might be able to cooperate in this rvork in of the Attu Natives assured the commander of the connection with expeditions to the region. I further expedition that there were a few at Agattu, which suggested that as a feature of such cooperative effort, would soon he gone because the Japanese came the Navy Department might consider the establishment frequently to take them. 2. On hIay 1933 [1935?] it of a radio station in the region and thus, in addition to was determined definitely that the sea otter does exist the Bureau of Fisheries employees, there would he the in the Rat Island group. On that date, the reporting radio station personnel. officer [Hutchinson] xvhile ashore on the northwestern end of Amcl~itkalsland at the survey station, BIRD, 1936. 0. J. hdurie and C. S. \\rilliams were saw eight (8) adult animals and three (3) of the young assigned to make au inventory of resources in the or pups moving about in the kelp beds which surround Aleutian Islands \\'ildlife Reservation, as it \\'as the station. !\'it11 the aid of good binoculars, two then called. This inventory was initiated primarily animals were watched at play in the clear water under because of the govenlment's responsibilities re- the sheer bluff of the station. From this favorable point of vantage it was seen that the animals were sea lating to the blue fox industry in the Aleutiaus aud otter and not seals nor sea lion. There is no doubt that not from concern for sea otters. klurie (1959, the animal in question is a sea otter. p. 285) says: The memorandum continues with numbers of In spite of occasional poaching, in 1936 we found animals seen and finally an estimate that the sea substantial sea otter populations in several places otter population around Amchitka numbers 1000 throughout the Aleutian chain, and we made a con- selvative estimate of 2,000. hlost heartening of all, adults and 500 pups. they upere extending their range, not only in the The second memorandum was prepared by Aleutians, hut also along the Alaska Peninsula. How- \\lard T. Bo~ver,Chief, Division of Alaska Fisheries, ever on our last visit to Sanak Islands, the sea otters for a Mr. Jackson on Aug. 19, 1935. It told of a had not reappeared, although at one time this area mas recent telegram from the Commander of the Bering one of the best sea otter hunting territories (since our visit, five sea otters have been seen). Sea Patrol Force of the Coast Guard and gave an estimate of 1000 to 6000 sea otters around The data obtained in 1936 indicatcd that there Amchitka Island and indicated that they were were still a few otters remaining about isolated spreading to other nearby islands. Bower em- islands, and infrequently they visited bays of the phasized the necessity of protective measures so North Pacific. The most southeasterly point at that this gain and reestablishment of sea otters ~vhichthey \\,ere reported \\'as h,lontagtie Island in could be maintained and expanded. Bower sug- Prince \\'illiam Sound. To the east, trappers and gested a conference to be attended by representa- local residents said sea otters were present near tives of the Coast Guard, the Navy, the Bureau of Shuyak Island and Latax Rocks, just north of Biological Survey, and the Coast and Geodetic Afoglak Islaud in the Icodiak group. From Sutwik Survey to meet during the Isinter to determine the Islaud uear the lane of Alaskan shipping, they were best means of consert~ing and protecting the also reported. Eleven sea otters were seen off Aleutian Islands' sea otter herd and to formulate Simeonof Island in the Shumagins. The rocky plans that could be put into effect for the next Sanak group was reported to harbor a few, whereas season. illformation obtaiued from Mr. Homer Jelvell, The third memorandum was prepared by Alaskan Game \\larden, indicated that they oc- \\lard T. Bower aud sent to Comn~issio~lerBell of curred 011 the Alaskan Gulf side of the Alaskan the Bureau of Fisheries. Bo~ver's tnemorandum peninsula at Aiugnak columns, the southeast por- says in part: tion of Cape ICumlik, and at Kujulik Bay. This morning Captain B. L. Canaga, Room 2058, In the Aleutian Islands proper, natives said that Navy Department, Telephone Branch 214, called at otters occurred occasionally about the follo~viug this office and said that a confidential report had been islands: Seguam, Little Tauaga, Kanaga, Iciska, received from an officer attached to the Navy f'.xpedi- tion to the Aleutian Islands this summer in regard to Semichi, Agattu, aud Attu. The survey party sea otters in that region. Captain Canaga permitted me visited all these islands except Semichi and Agattu to read the report and then took it away with him. He but observed no sea otters. However, the party did Sea ~l.lam?ttals:Resoiwces and Popt~latio~~497 encounter the anilnals at Umak, Ulak, ICavalga, A partial count of sea otters around Amchitka Ogliuga, Amchitka, and Tanaga Islands. numbered 1321 animals; this was estrapolated to At no st of these islands, only a few sea otters arrive at an estimate of 1761 animals for the entire were seen, the greatest number at any one island island. being 48. No\vhere were they abundant, 10 per 1939. Reference to the Amchitka sea otter mile representing the greatest concentration ob- investigation was given coverage again in the 1939 served. The islands of Amchitka, Ogliuga, and Alaska Fisheries and Fur Seal Industries Report Tanaga appeared to be the most productive; a most (Bower, 1941, p. 159) with the following state- liberal estimate of the number of otters occurring ment: "The substation \\~hichwas established on about the three being 700. The small concentra- Amchitka Islaild in 1937 for sea-otter investiga- tions near other islands might raise the total tions and patrol, and xvhich had been in operation number of sea otters in North America to 2000. To each summer since that time was maintained on a add a note of contrast, in the early 1800s, one year-round basis during 1939-40." The sea otter island, Unalaska, produced 1000 of the animals count this year was 1355 a~~imals;the total- al111ually. population estimate, includi~~gpups, \\,as 1870 In the 1936 Alaska Fisheries and Fur Seal (Loy, 1940). Lensink (1962) also reports an earlicr I~ldustriesReport, Bower (1937) noted that five count tl~atyear of 1030. Coast Guard cutters were assigned to patrol the 1940. J. B. Arkmgan and Grant Ritter, sea North Pacific Ocean and Bering Sea for the otter wardens on Amchitka for the Bureau of protection of fur seals and sea otters. Lieutenant Fisheries, estinlated the Amchitka sea otter popula- Commander S. P. Swicegood, U. S. Coast Guard, tion to be 1650, 69 being pups, from a census commanding officer of the Cl~elatl,reported a made in July and August (U. S. Fish and \\'ildlife count of 814 sea otters on Amchitka (Swicegood, Service, 1940, unpublished report). 1936). 1943. The first known account of an aerial 1937. Comments on sea otters this year by survey for sea otters is recorded in a letter dated Bower (1938) \\'ere: September 28 from Frank L. Beals, Refuge A substation on one of the western >\leutian h,lanager, Aleutian Islands National \Vildlife Islands was established for the expansion of sea otter Refuge, ICodiak, to Frank Dufresne, Regional investigations and patrol.. . . Two trips [by the Director, Fish and \\lildlife Senrice, Juneau (U. S. Pengtcin] were made to the western Aleutians-one in July and one in Septetnber-in connection with the Fish and Wildlife Service, 1943, unpublished re- sea otter patrol. port). Accotnpanying the letter is an outline chart of Ainchitka and Rat Islands; numbers of otters In this year Lo)' and Friden (1937) counted seen are noted on the chart at the point of 1321 otters on Amchitka, almost all of them on obser\ration. The sea otter count on June 24 for the Pacific coast. Amchitka was 3417. 1938. Substantially more information ap- 1 Beals was very active in the Aleutians peared for this year in the report by Bower (1940, before 1942 and made conspicuous efforts to p. 161). A two-paragraph description with the continue this activity during the period of military heading "Substation for sea-otter patrol" reads: operations. 13s letter is reproduced in its entirety: The Burean's work for the furtherance of sea- The following report is made of an attempt to otter investigations and patrol was continued. At the estimate the sea otter population of Amchitka and Rat substation established on Amchitka Island in 1937, 4 Islands, i\laska, through the use of aeroplanes. additional overnight cabins, 8 X 10 feet, were built on While engaged- in regular patrol of the Aleutian the south side of the Island at 8-mile intervals. At the Islands National \Yildlife Refuge it became my good camp site in Constantine Harbor, a 10 X 16-foot fortune to make the acquaintance of a Squadron of po\rverhouse was constructed and a 5-horsepower U. S. Navy Obseivation-Scout Pilots, then stationed on engine was installed for use as a porrer plant. T\vo r\mchitka Island. At their suggestion and with the radio masts were erected on concrete bases, and radio permission and sanction of the Commander of the antennae were strung leading to a receiving set and a U. S. Army Force and the Captain of U. S. Navy transmitter. Buildings that had been erected in the Facilities for A~nchitkaIsland I was taken on flights previous year were given a coat of paint. around Amchitka, Rat, Segula, and Little Sitkin Islands The number of sea otters counted in the rricinity of for the purpose of observing sea otter and estimating ~\mchitkaIsland in 1938 was considerably less than the their number and general distribution. estimate for the previous year. Whether the difference U. S. Navy Pilot G. T. Joynt, a former biology was due to an error in the count or to a change in the student already actively interested in sea otter, had on habits and distribution of the animals was not known. prerrious occasions observed their general location in It is anticipated that further light will be thrown upon this area. To Mr. Joynt belongs most of the credit for this problem by investigations in succeeding years. what success we had in checking their numbers in the Anlcliitka-Rat Island area. The other pilots \\-ere for making an accurate check on the distribution and willing and helpful with their time and personal approximate number of sea otter in the Aleutians. observations but it was the old biology student who Personal observations from the air were made in smelled them out and maneuvered his plane in position the vicinity of Amchitka, Rat, Segula, Little Sitkin, for effective obseivation. Attu (h'lassacre and southeastern part only), and The type of plane used was ideally suited for this Semidi [Semichi] island groups. The only s~~ccessful work as it throttled down to fly safely and with flights \\*ere at Amchitka and Rat Islands. Poor stability at close to one hundred and thirty miles per visibility due to fog plus limited time for choosing hour. Good visibility from an open cockpit can be had good flying weather resulted in nothing of value being if desired by simply cranking back the protective learned regarding sea otter about the other islands. co~\.lingand banking slightly when over the object to Two trips were made from Amcllitka to Davidoff and be observed. Khvostoff islands but each time fog whipped in and Our method of procedure rvas to fly around the forced a hasty about face and luirried trip home. The island a number of times to become familiar rvith planes were equipped with wheels and it was not landmarks and principal concentrations. Then over the considered prudent to tempt fate too far. plane's interphones to communicate our opinions and These same pilots have made many scouting flights reach an agreement on each pod of sea otter observed around Semisopochnoi Island and report that they and plot it on an outline chart of the island. An average have never seen sea otter in that r,icinit\r. Scattered size pod was selected and its actual number of animals individuals and small groups have been observed in the counted as closely and accurately as possible. This was kelp beds around Davidoff and Khvostoff islands which then used as a yardstick in estimating the number of lie between Segula and Little Sitkin. None of the otter in each of the other groups encountered. Obsen~ation-Scootpilots at Attu island were familiar The results of our sunfey are not offered as \\!it11 sea otter or knoxv where or ho\v to look for them anything more than an approximation. I believe it rvas and say they have not observed the animals as I altvays our tendency to see and estimate more animals described them. A civilian boat captain claimed to than actually were in the groups. Because of the sea know what he was talking about and maintains that otters peculiar and cl~aracteristicposition in the water, they are often seen in the kelp off i\lassacre Bay and that is of floating on its back with the txvo extremities around tlie Se~niili [Semichi] Island gsoups. In an of its body prominently displayed, head and forefeet at attempt to check up on sea otter in the Sernidi one end, hind flippers and tail at the other end and its [Semichi] groups I encountered difficulty with midsection mostly submerged and hidden from view, General Copeland ~vhowas in command of the base one is inclined to see t\vo animals where there is really there and rvas 'evacuated;' the general does not like only one and come out on the plus side in the count. civilians, particularly 'D- Game Wardens.' One of Offsetting this sonlewl~atis the fact that Ire did not the pilots from Amchitka reports having seen 'a count or include singles and pairs in the survey and thousand or more' in groups of islands around Ogliula they represent an appreciable number of otter rr.11en and Skagul. A civilian boat captain and several dif- tlie entire island is considered. ferent Army officers told of ha\*ing seen them in the The majority of groups observed were in the vicinity of Kagalaska strait and Adak island but I offshore margin of the kelp fringe that rings the talked with \('=It Kemsack, a civilian guide for the islands. Singles and pairs, etc., rvere noted to keep Alaska Defense caminand \'ho has trapped on Adak closer inshore and often right in the middle of thick for the past several years and has patroted Adak island masses of kelp. Only rarely were sen otter observed in and adjoining waters for the Army and Navy since its open water more than t\vo nliles from shore or half a occupation by our armed forces, and in his opinion mile from protective kelp patches. Mr. Joynt and his there are no more otter there than in past years before friends report that it is rare indeed that they have seen milita~yactivity in the western Aleutians. an otter as far as five miles offshore but that they have obsenred an occasional single animal idling along on its 1945. A partial count of 365 animals was back as far as 10 miles from the nearest land. made from a limited Itumber of checkpoints on I personally experienced difficulty in picking out Amchitka (U. S. Fish a11d \\'ildlife Sen'icc, 1953, young sea otter I-iding on their mothers midsection but unpublished report). Pilot Joynt, rvhose eyes are more alert and better trained, claims to have seen many of the little fellows. 1949. Aerial surveys of inshore waters of IIe tliougl~tas many as one juvenile to each ten or Arnchitka and Rat Islands sho\ved 1321 sea otters trve1r.e adults. I x\.ould not want to offer an estimate of around the t\vo islands (Jones, 1951). Lertsi~lk yomig sea otter around i\mchitka and Rat Islands but judging from personal observations made from shore (1962) and ILenyon (1969) report that 1087 of through high power field glasses, they do not appear to these were around Amchitka and the remaining be less than 1 remember having observed on past visits 234 around Rat Island. to Amchitka. It is interesting to note that the sea otters reaction 1953. The Quarterly Narrative Report for the to aeroplanes is entirely negative. They seem to accept period Septe~nber-December 1953, Aleutian the plane as something natural and quite ordinary. On Islands National \\'ildlife Refuge, is quoted (U. S. several occasions we came down on them with engine Fish and \\'ildlife Service, 1953, unpublished re- roaring and approached as close as fifty feet above them. They only trvitched their \vhiskers and went port): right on with their loafing. One or two of them might III 1873 (39 [38?] years before the sea otter submerge but not 514th any display of great fright or season \\.as closed in 191 1) the Atka Natives took 129 concern. It seems to me that the plane is an ideal tool sea otter on their hunting grounds (Buldir Island on the Sen ~llaiilnlnls:Resources nild Poptilatioil 477 west to Islands of Four hlts. on the cast). The average 1968. Froin 1968 through 1974, helicopters annual take from 1873 to 1896 was 96 animals. Today, \\?crc used for sea otter surveys on Amchitka (1953), 82 [80?] years later we sal\*agcd 50 pelts of sea Island, and the counts became so numerous that I otter from the island of A~nchitkaalone. have sumniarized them in Table 1. 1954. A sea otter survey was conducted by On Nov. 7, 1967, a small conference !\.as held 1.cnsink on April 26 \\it11 tlie follolving results: in Columbus, Ohio, attended by representatives of "Observations were made from a Navy UF (Alba- the Bureau of Sport Fisheries and Wildlife, the tross) bet\veen 1130 and 1500 hours on April 26. Alaska Department of Fish and Game, the Battelle Seas \\'ere flat, sky partly overcast to clear but Colulnbus Laboratories, aid the Atomic Energy visibility \\,as escellent. Transects covered the Commission, to establish acceptable methods of Delarof Islands, the hidrcanof Islands, and Gareloi assessing any immediate and long-range effects of Island." IIc reported 320 animals in the Delarof underground nuclear tests on the Amchitka sea Islands, none at Gareloi, and 684 in the hidreanof otter population. One of the conference's four Islands for a total of aboot 1000. "General recommendations was to conduct annual aerial impressions \\'ere that Tmaga and ICanaga have surveys of sea otters around the island (Spencer, otter populations fully as dense as that of Am- 1969). Later Spencer determined the follolviiig chitka. Otter were foiund on all of the Dclarofs, criteria to be "minimally acceptablc for helicopter indicating a good population although only onc sulveys:" large pod was seen, that being on Amatignak. l.LtTind velocity less than 15 to 17 knots, Characteristics of otter distribution on Amchitka if preferably \\,esterly. applied to anilllals 011 tlie Delarofs indicate far 2. No heavy surf (frequently a residual condi- more otter are present than were actually counted" tion fro111 past storms, even \\.hen tlie present ~viiid (U. S. Fish and IVildIife Service, 1954, unpublished velocity is lo\\'). report). 3. \'isibility of iiiore than 5 miles. 4. Overcast skies to eliminate sun glare and 1956. Between May 10 and August 25, reflection. Leiisink (1962) completed surveys 011 about half 5. AIinimum ceiling of 500 ft. the Amchitka shoreline with binocitlars and a 6. No rain. telescope. In the first survey 1604 sea otters \irere counted, and in the second 2568, including 384 \\'it11 the use of thcse criteria, seven counts were pups. Spot-checks on unsur\,eyed parts of tlie made in August and September. The totals listed in island indicated that the survey figures coulcl be Table 1 are the actual counts of otters seen extrapolated to the entire island, indicating (from \\,itbout any correction for percentage seen. T~vo the second count) a total population of 5637 snnreys were also made at Rat Island, one on animals. September 24 yielded 431 sea otters and oiie on October 6 yielded 456 sea otters. 1959. ICenyon and Spencer (1960) made 1969. A combination of aerial photography aerial counts of sea otters in the western !\leutian and visual obscr\ration from a helicopter was used Islands in May and counted 9507 animals. Tlie to derive sea otter counts this year. Stephan (1971) Amchitka sea otter count on Aslay 19 \\,as 1560. reports a maximum count of 2354 otters 011 1962. On March 29 and April 5-10, 1Cen)~on September 2. Later studies of the photographic (1969) and others (U. S. Fish and \\'ildlife Selvice, record corrected these results, and the date of the 1962, unpublished report) lnacle an aerial survey of highest survey yielded the three co~uitslistecl in sea otters from ICiska Island to Alnak Island in the Table 1. Aleutian Islands, in the Sanak Islands, Saidinan 1970. Sul-veys conditcted after 1969 returned Reefs, and Shumagin Islands, and in areas along the to visual observations made from a helicopter, south coast of the Alaska peninsula. A total of following the basic criteria established in 1968. 10,364 sea otters \\,as observed. Obser\ration condi- Five full counts and one partial count \\.ere nlade tions \crere poor in the Rat and Andreanof Islands, in September, as listed in Table 1 (U. S. 1;ish and but observation conditions east of the t\ndreanof \\'ildlife Sewice, 1970, unpublished report). Islands were generally excellent. 1971. In this ycar counts were made of the 1965. t\n aerial suivey of sea otters fro111 Cold particular segnlent of coastline near the proposed Bay to Attu Island was made diuring April and Cannikin test; tlie results of this cndeavor are Alay. A total of 12,687 sea otters was counted. Tlie reported by Estes (Chap. 21, this volume). Bad count at Amchitka 011 May 2 was 1144 (I

Table 1-Snalmary of A~ncllitkaSea Otter Counts

Date hlethod Count Ohserver Co~n~nents

Surface Hutchinson Lensink says 1000 Surface Smicegood Lensink says 804 Surface Loy and Friden Surface Loy and Friden Extrapolated to 1761 Surface Lov and Hewitt Surface LO; Surface hlanzan and Ritter Estimate Airplane eats Lensink says 3420 Surface Gray Partial count Airplane Jones Binoculars Lensink Extrapolated to 3525 Binoculars Lensink Extrapolated to 5637 Ailplane Kenyon and Spencer Extrapolated to 2080 Airplane Kenyon and Spencer Extrapolated to 1520 Helicopter Sow1 Helicopter Spencer and Cater Helicopter Spencer and Cater Helicopter Spencer and Cater Helicopter Spencer and Cater llelicopter Cater Helicopter Cater I'hoto+llelicopter Stephan and Mercier Photo+helicopter Stephan and Alercier Photo+helicopter Stephan and Alercier Helicopter Cater Partial count Helicopter Cater Helicopter Cater Helicopter Cater Helicopter Cater Helicopter Cater Extrapolated to 3133 Helicopter Cater Helicopter >\begglen, Estes, and Schneider Helicopter Abegglen and Estes Helicopter )\begglen and Estes Helicopter Abegglen and Estes Helicopter Abegglen and Estes Helicopter Abegglen and Schneider Helicopter >\begglen and Schneider

made. A sulvey on October 2 gave a count of 3241 Populatioll Status. Lensink (1962) and Ken- sea otters (U. S. Fish and C\fildlife Senrice, 1971, yon (1969) have analyzed the population history unpublished report). of these islands and in particular of Amchitka where most of the popul&on counts were made. 1972. Surveys in 1972 were cooperative ef- forts. Representatives of the Bureau of Sport I

Removals fi'oin Transplants to Prince Otller William Southeastern British Year Atnchitka Aleutians Sound Atta Pribilofs Alaska Goln~iibia Washingtoe Oregon Zoos Harvests

*Sources: Kenyon, 1969; Vanya, 1970; K. B. Schneider, personal commn~unication.

artificially from the population and transplatltecl or shipped to the National Zoological Park, \\'ashing- ha~~estedis other loss, in capture or for use in ton, D. C. All three otters died within a week of scientific experiments or \vhate\'er.) A chrono- arrival. logical rksiunk of some of the details of these 1955. Sea otters were transplanted from Am- transplants to natural-habitat areas and to zoos chitka to the Pribilof Islands by a chartered fishing follo\\.s. vessel, the Paragon, betxveen hiarch 28 and April 4. 1951. An attempt to transplant sea otters this The 19 suwivors of the 31 sea otters aboard the year failed because of lack of basic kno\cdedge ship were released at Otter Island, but ICenyon about the require~~lentsof the animals. Thirty-five (1969) is reasonably certain that none sulvived. animals that had been captured on t\mchitka Three died within minutes after being placed in Island died in captivity before the transplant could sea\vatcr. take place. T~vosea otters \\,ere transported from Am- chitka to Seattle by air and were released in the 1954. A press release from Headquarters, \\foodland Park Zoo on October 10. One died Seventeenth Nwal District, ICodiak, dated hIay 4 within a felv days. The second, a fctuale named gave the follo\\'ing caption to an accompanyillg Suzy, lived until Oct. 27, 1961. photograph: Tile first sea otter to be successfully transported 1956. On April 10 five sea otters captured on from its native habitat of Amcllitka Island in the Amchitka Island \\.ere transported by ship and Aleutians is taken ashore at the U. S. Naval Station, released on Bird Island in Attu's R'lassacre Bay by Adnk, Alaska on April 20, 1954 from the seagoing tug R. D. Jones (U. S. Fish and \\'ildlife Service, 1956, USS TILLAMOOK. Lt. Francis Bean, USN, of Seattle, unpublished report). Jones notes in his report that \\'ashingtoa, Commanding Officer of the TI1,I.A- he kno\vs one of the five otters died after rhe 1\1001<, supelvises unloading operations as hlr. Rob- ert D. Jones, JI., Manager, Aleutian Islands National release. \\'ildlife Refuge and Espedition Leader, and crew T~vosea otters from Amchitka Isla~ldwere rneniber take the first cage ashore. placed in the \\foodland Park Zoo in Seattle July 30 (Kenyon, 1969). Onc of the otters died 16 This sea otter \\.as one of four captured in k1arch days later, after appearing to be in excellent and April on Amchitka and tra~tsportedto Ad& to condition on arrival. be held there until the end of h4ay. One of the animals died on May 5. The remaining three sea 1957. An attempt to transplant eight sea otters were transported by na\,al vessel to Seattle otters from Amchitka Island to St. Paul Islantl was to be placed in the \\foodland Park Zoo on June 1. made on December 11; bad weather caused an en They remained there until July, \\!hen they were air route delay, and six otters died that night. A storm Sen M~IIIIJIR~S:Reso~~rces nlfd Popftlntio~r 503 canccled the flight to St. Paul Island, and the two Reef and Cape Arago and 24 otters at Port Orford remaining otters were taken to Seattle and placed \\rere made on Jnne 24. in the \\'oodland Park Zoo on December 14. One of the t\so otters died December 16; the other 1972. The Alaska Department of Fish and lived until September 22, 1958 (Icenyon, 1969; Game, \vith cooperation from the Fisheries Ice- Kenyon and Spencer, 1960). search Boartl of Caiada aud the British Columbia Fish and \\'ildlife Department, transp1;mted 46 sea 1959. Seven sea otters were transplanted from otters fro111 Prince \\'illiam Sonnd to Checleset Bay, Amchitka Island to St. Paul Island on Ahy 20. Vancouver Island, on July 19. In addition, 3 were 'l'hese anilnals arri\zcd at St. Paul Island in excellcnt given to the Vancouver, British Columbia, public condition after about a 3.5-hr flight from Am- aquarium. chitka (ICcnyon and Spencer, 1960). Because of its history of near extermination, 1965. Iuld 10 to Yakutat Bay. throughout the \vestern Aleutians. 1968. The Alaska Department of Fish and Game, \\,it11 cooperation from the Atomic Energ Northern Fur Seal (Cnllor/fi~~ftsflvsi~lns) Commission and the Bureau of Sport Fisheries and The Pribilof Islands \Irere discoveretl in 1786 \\lildlife, transplanted 55 otters from Amchitka to and 1787 by Gerassim Pribilof, Russian navigator St. George Island in the Pribilofs; 301 otters were ru~dfur trader. This \\.as thc rcsult or Inore than 18 transplanted from ilmchitka to six locations in years of search for the fur seal brceding colonies. southeast Alaska. Sinls (1906) estimated the herd size at the time of 1969. 'The Alaska Department of Fish and discovery to be 2,500,000 animals. Beginning at Game, with cooperatioll from the Atomic Energy that time and contini~ingfor the next 150 years, Commission, transplanted 120 otters: 58 to Khaz therc \\'ere periods of extrenle esploitation and Bay in southeast Alaska; 29 to Pt.Grenville, periods of no exploitation. In 1911, .ivhen the \\'ashington; 29 to Vancouver Island, British United States assumed control of the sealing Columbia; and 4 to the Taconla aquariiun to form industry, the herd size was estimated to be 21 study colony. 215,000 seals. The large number of fur seals moving north and 1970. The Alaska Department of Fish and south through the passes in the Aleutian Islands Game transplanted 14 sea otters from P~incc ;uld east and west along the chain during migration \\'illiam Sound to Checleset Bay, Vancouver Island, periods made appearance of the fur seal not British Columbia, in July. Onc of the four taken to unusual. Tacoma in 1969 was transferred to the Vancouver, British Columnbia, public aquarium. The Alaska The importance of the fur seal in the Aleutians Department of Fish and Game and the Atoniic is its ninnbers as a nligra~ltand not as a resident Energy Commission cooperated in the transpla~lt species. of 62 sea otters from Amchitka on July 18-19, 1970: 31 went to the State of \\'ashington (1 died Steller Sea Lion (Ero~~elopinsj~tbntn) and 30 \\'ere released near Lapush) and 31 went to This spccies, named after Georg \\'ilhelm Oregon (2 died shortly after arrival and 29 \\,ere Stcller, was prominent among the sea mammals released near Port Orford). encoiuntcred by the earlicst explorers in the Aleu- 1971. The Alaska Department of Fish and tians. The sea lion was also a victin~of the ruthless Game, tile Atomic Energy Commission, and the killing of sea ln'anlmals by Russians and Americans, Oregon Gamc Conlmissioll cooperated in the trans- and, by the end of the 1800s, relatisely few plant of 64 sea otters from Anlchitka to the remained. Sea lions xvcrc common in the Aleutians, Oregon coast. Releases of 40 ottcrs near Simpson ruld today they 11a\~returncd to ~vhathas bccn estimated as close to their original numbers (ICen- grounds and rookeries on and acljacent to the yon, 1971). island (U. S. Fish and \'\'ildlife Sewice, 1962, Scammon (1874) describes the use of sea lions unpublished report). by the Aleuts: 1965. The April 18-May 9 aerial census The dead animals are then skinned, and their covering the Aleutian Islallds and part of the hides packed in tiers until fermented sufficiently to Alaskan peninsula from Cold Bay to Attu showed a start the hair, when they are stretched on fiames to dry, and eventually to become tlte covering or planking count of 63,933 sea lions. Of this number, 710 for the Aleutian haidarkas and baidaras. The fat is \Irere counted at Amchitka (U. S. Fish and \\'ildlife taken off and used for fuel, or the oil is rendered to Service, 1965, unpublished report). burn in their lamps. The flesh is cut in thin pieces from Thorsteinson, Nelson, and La11 (1961) reported the carcass, laid in the open air to dry, then tanned, on an experimental harvest of sea lions in 1959 to and worked into water-proof clothing. The stomach is emptied of its contents, turned inside out, then determine the feasibility of a commercial venture. inflated and dried for oil bottles, or it is used as a Five sea lion rookeries supplied 616 sea lions; 464 receptacle for the preserved meat; and what remains of \\'ere processed into 200 tons of ground meat and 9 the once formidable and curious animal is only a tons of whole liver. There were many technological mutilated skeleton. problems, but the autl~orsdetermined that sea A systematic use of aerial photography for the lions could be taken in sufficient numbers to be evaluation of sea lion populatiolls in Alaska gave profitable. I\.lathisen (1959) data for his count of 149,264 Private commericd hunters, regulated and animals. An illcomplete count for the Aleutiw monitored by the Alaska Depxtment of Fish aud Islands derived from his data was 73,090 animals. Game, conducted an annual harvest of Steller sea h~Iathisen poiuted out that the counts were of lion pup skins from 1963 through 1972. The total necessity minimum figures because ol~lyanimals harvest was 45,178 skins, with rookeries in the that \irere on lai~dwere counted; aniinals feeding or eastern Aleutians supplying 12,405 of these. None away from the rookeries and hauling grounds were were taken from the western Aleutians. not coltl~ted.Later ailalysis by Mathisen aiid Lopp (1963) increased the Aleutian Islands sea lion Harbor Sed (Pl~ocavit~tlinn) population to about 83,000 animals. ICenyon and Rice (1961) estimated the sea lion The harbor, or hair, seal occurs throughout population of all Aleutian Islands aud areas near the Aleutian Islands, lising in protected bays and the end of the Alaska peninsula to be about river mouths and hauling out on sinall offshore 100,000 animals. This figure \vas derived from the islands and exposed reefs (Itenyon and Scheffer, first complete aerial sui~~eyof the Aleutian Islands. (1955). h'lurie (1959) did not findit abundant in the The estimated number of sea lions for Amchitka Aleutians during his 1936 expedition, but histori- \\,as 1250 n.itli the foll~~vi~~pdistribution: cal and archaeological evidence indicates that this species was common. It is eomlnon today. Bird Rock 50 Very little research bas been carried out on Chitka Point Islet 50 harbor seals esclusively in the Aleutians. Aerial Ivakin Point 300 surveys for sea otters hwe at various times in- East Cape 200 cluded harbor seals and sea lions as additional St. h,Iakarius Islet 50 species observed, but geuerally harbor seals have Southwest islets and rocks 600 been neglected. The Alaska Department of Fish The Aleutiatl Islands appear to be the center of and Game has had active research and harvest abundance for sea lions and the point from \vhich programs inlocations other than the Aleutians, and seasonal mo\~ementsoccur. the bulk of seal skins bountied by Alaskan hunters Ite~lyon's aerial census of sea lions from the came from more accessible areas. Islauds of Four Moulltains to Sandman Reefs, A harbor seal population is difficult to count hlar. 3-4, 1960, \\.as 55,325 animals (U. S. Fish because the seals are scattered, breeding in small and I\'ildlife Service, 1960, unpublished report). colonies rather than large rookeries as do fur seals and sea lions. A population estimate for Amchitka 1962. A sea lion census was inade on Island in 1971 \\.as 900 to 1000 seals (C. Hardy, i\,larch 29 and April 5-10 in coinlection \\'it11 the U. S. Fish and \Vildlife Service, persoilal com- sca otter census. A total of 54,704 animals was munication). The results of three harbor seal coul~ted (U. S. I.'isli and I\'ildlife Service, 1962, counts are listed belo\\'. ~tnpublished report). Robert D. Jones made a census of Steller sea liotls at Ainchitka by dory. He 1960. An aerial couut of harbor seals from counted a total of 3584 animals at thc hauling the Islands of Four Mountains to Sandman Reefs, Sea ~llantmnls:Resources uizd Poptrlatio~l 505 March 3-4, gave a count of 3591 animals (U. S. south of the Aleutians. It is one of the larger Fish and \\lildlife Service, 1960, unpublished re- cetaceans and is important to the \vhalitlg industry. port). Little Piked \\'hale (Balaei~optera acutoro- 1962. An aerial count of l~a~borseals in strata). The little piked whale is also k11ow11 as the conjunction with a sea otter census of A'larch 29 nlinke whale and is distributed widely in the North and April 5-10 gave a count of 6000 plus animals Pacific. (U. S. Fish and \\'ildlife Service, 1962, unpublished Baird's Beaked \\'hale (Berardizts bairdi). The report). Baird's beaked whale is distributed throughout the 1965. An aerial count of harbor seals con- Bering Sea in the North Pacific Ocean. It has been ducted in colljunction wit11 the April 18-klay 9 halvested by the Japanese \vltaling fleet. sea otter survey gave a count of 4868 animals from Cuvier's Beaked \lae (Zipl'ius caui- Cold Bay to Attu Island (U. S. Fish and \Vildlife Service, ~~n~ublisbedreport, 1965). rostris). The range of the Cuvier's beaked \\,hale includes the waters of the North Pacific Ocean and Cetaceans the Bering Sea. It does not go into the Arctic Ocean. Statistics on hoxv many of different species of Stejneger's Beaked \\'hale (~llesoplodortstejize- cetaceans have been taken annually in recent years geri). The Stejneger's beaked ~vhale is fou~td in and near the Aleutian Islands are given by througl~outthe Aleutians and in the Bering Sea. Its h,Ierrell (Chap. 15, this volume). The follo\ving distribntion has been cocretated with the ctistribu- species* have been observed or harvested in the tion and migration patterns of Pacific salmon. area. Beluga \\'hale (Delf,ltii~aptentsletrcas). Belugas Blue \\%ale (Balaenoptem atusculus). Blue range on both sides of the Bering Strait aud do\\m \vl~alesare generally found in three summer loca- into Bristol Bay on the American side. This species tions: (1) the eastern Gulf of Alaska, (2) the area is uot kllo\vn to migrate south of the Aleutian south of the eastern Aleutians, and (3) the area Islands alld is only a rare visitor to the Aleutians. from the western Aleutians to ICanlchatka. Doi, Nemoto, and Ohsumi (1967) calculated the wm- Gray \\'hale (Escltricl~tius robustus). The gray mer population of blue whales to be about 1420, a \vl~alepasses through the Aleutians en route to its drop of just over 1000 animals from a postwar summer feeding grounds in the Bering Sea and the population of 2430. Rice (1971) interprets these Arctic Ocean and again en route to its breeding data as an indication that blue \vhales were never grounds in the bays ru~dlagoons of Baja California, abundant in the castern North Pacific and that Ivlexico. The gray \vhale is colnpletely protected there has been little decrease in their populatioll from commercial whaling efforts. size. Blue whales 11a~'ebeen con~pletelyprotected Humpback \\'hale (~llegaptera itouae- beginning with the 1966 season (International angliae). FIumpback xvhales are known to occur Ccmmission on \\'haling, 1967). fairly close to shore in the eastern Aleutians. Finback \Vhale (Balaeitopteru pl~ysahcs). The Pacific Right \Vhale (Balaeita glacialis). Kice summer range of finback \vhales includes the (1971) says that the "ICodiak groutld," \vhich immediate offshore waters of lands around the ellcompassed thc entire Gulf of Alaska from North Pacific Ocean. The increased pressure of Vancouver Island to the eastern Aleutians, was pelagic whaling on this species has reduced this renowned in the nineteenth century as one of the stock to well below the maximum sustainable yield best areas for lli~ntingright \\,hales during the (Ohsumi, Shimadzu, and Doi, 1971). summer. A fe\s could be foulld in the southern Sei \\'hale (Bnlaei~optera borealis). The sei Beling Sea and all across the North Pacific at that whale distribution in the summertime is similar to season. This species is protected by the Interna- that for the finback whale except that sei \\.hales tional \\'haling Commission. rarely go north of the Aleutian Islands. Bowhead \\'hale (Balaena ?itysticetus). The Sperm \\'hale (P11)rseter catodoit). The sperm borvheact whale is extremely rare in the Aleutians. \shale is \videly distributed and is found north and Bowheads spend the winter in the loose southern edge of pack icc, ~vhichusually extends across the central Bering Sea from Igray whale by Rice and \\'olman (1971) revised the nomenclature for species is protected by the International \\haling this species; the revised version is used bere. Commission. Pacific \Vhitesidcd Dolphin (IZngei~orl~y~zchtrs There has always been a question whether this ~bli~~ridei~s).Tlie Pacific \vhitcsided whale is animal had ever occupied the Aleutian Islands. wiclely distributed and very common in the North Stejneger (1883, p. 84) said \\'osaessenski had obtained a rib of a sea coxv from Attu Island, and, in Pacific Ocean. conversation, Stejneger expressed the belief that sea Iiiller \\fltale (Orciilz~sorca). Tllc killer whale co\v remains might be fo~~ndon >\gatto Island. Goode et al. (1884, p. 136) wrote as follon.s is the most \vidcly distributed liiarilic niannnal in concerning this find: '\\'osnessenski found a rib of the tlie world and is commonly seen in tlie Aleutians. unimd on Attu, the last island of the archipelago, but as Brandt suggests, it may have been derived from a Pilot Whale (Globicephnln scntntnof~i). Dis- Rkytinn nrasl~ed thither by the waves. hlr. Lacien tribution of the pilot ~vlialeis considered to be the Turner kindly inforn~edme that an aged ivo,naa stated nortliem waters of the Pacific Ocean, including tlie that Rhytina had been seen at Attu by her father, bnt Aleutian Islands and south to latitude 36" (Nislii- sucll testimony is, perhaps, not altogether satisfactory.' tvaki, 1967). Thus, we may have some evidence that the sea cow may have occurred on the rvesternnlost Aleutian Harbor Porpoise (Plfocoeitn pl~ocoet~n).hmlurie Islands, and it would be extremely interesting to have (1959) saw two harbor porpoises on May 23, 1937, irlentification of bones from old Aleut village sites. To in the Slnniiagin Islancls, and, on Aug. 13, 1937, lie date, studies of such midden material have not revealed the presence of sea cow remains, and, on the whole, it sax\' three in the harbor at Atka Island. is likely that this animal never inllnbited these islands Dall Porpoise (Pl~ocoer~oidesrlnlli). Tlic Ddl except as an accidental straggler. porpoise is widely distributed in the North Pacific Recent studies on the archaeology of Anichitka Ocean. (Dcsautels et al., 1970; Cook et al., 1972) bear out hIurie's statement that midden tilaterial has not yct Pacific \ilalrus (Odobelncs rostitnrus) verified tlie presence of sea co\vs at those particular Tlic Pacific walrus is associated with the Aleutian locations. Islands because of the proximity of its ~vintering However, a Pleistocene fIydroda~~~aliswas dis. habitat, tlie Bristol Bay area. There arc records of co\cred on Anichitka in 1969 from interglacial \valruses being killed south of the Alaska peninsula, beach sand and gravel (Card, Lewis, and \Yhitmore, iulcl Tur~ier (1886, as cited by Alurie, 1959) 1972). The Pleistocene find was identifieel as recorded tlie kill of a 2-year-old Inale walrus at H)~drodai~zcrlisRetzius, 1794. Fragrnelits of the Attu Island in September 1880. hlore recently, in bones and associated pecten shells were submitted 1960 Jones (U. S. Fish and \\'ildlife Service, 1960, for uranium serics dating, and this indicated their unpublished report) found the intact carcass of a agc to be about 135,000 years. young rc,alrus ncar South Bight on Amchitka Berzin, Tikhomirov, and Troinin (1963, cited Islatid. These animals lilust be regarded as strays; by Nisliiwaki, 1967) reported that sis large sen the Aleutian Islands are not a usual part of the ani~nalswere seen by nlen on tlie whale catcher \\.alrus range. The rare occurrence of walruses in B~drni~near Cape [Point] Navarin on the Siberian the Aleutians is reflected in the absence of identi- coast in July 1962. It was presumed that tbesc fiable walrus midden remains from Aleutian animals were Steller's sea cows. archaeological investigations. Doctors Colin and Kate Bertram (1964), having read tlie account by Belzin et al., speculated (in Steller Sea Co\v (Hydroda~rialisgigns) English) on the survival of sea cows, citing the snnrival of fur seals and sea otters as eviclence of h,Iurie (1959) says: this possibility, as ~vellas the remoteness and rarely Our knorvledge of the sea cow depends mainly on visitecl nature of tlie Poitit Navarin area. However, the accomrt of Steller, who, in the disastrous \\.inter the very identification of these aninials as sea corvs when Bering's expedition was wrecked on Bering Island has bcen questioned by Hepter (1965). after discovery of Alaska in 1741, had alnple oppor- tunity to study this ani~nalat first hand. The sea cox\, furnished food for Bering's party, as \uell as for other CONSERVATION kIEASURES expeditions that used the Commander Islands as a starting point for Alaska. It \\,as exterminated by The earliest esploitation of Alaskan fur bearers 1768.* by thc Russians and by poaching Americans and Englislnnen can only be described as ruthless arid *Sea co\\,s were very nlnlerous about the coast of ignorant. Even today tlie early opinio~l occ;~. Ramtshatka, and the i\leutan islands, at the tin~e\vhen they were first discovered; but the last of this species \\.as killed sionally crops up that thc aninials \\,ere disappear- in 1768 on Bering's island, and none have been ever seen ing, not because they were being overhunted, but since (Satter, 1802, p. 181). because they were nio\)ing away to avoid hunting Sen ~lln~?i~~taIs:Resozirces n11d Pop~ilatio~i 507 pressures (Fedorova, 1973, p. 105). Yet, \\,hen the The Tribunal of Arbitration at Paris made Russian go\rcrnment finally interested itself in certain a\vards, and the 53rd Congress acted these f;~rawayterritories, it was partly on the basis on tlpr. 6, 1894, to give effect to these of regnlating and even consenring the fur resources. awards. Congress, on Dec. 29, 1897, acted to Thus the Russian-American Company leas formed prohibit the killing of fur seals in the waters in 1799 by a forced merger of existing comptnies of the North Pacific Ocean. and given a monopoly to eliminate competitioll The Act of Apr. 21,1910, prohibited the killing and consequent ~\~astefulnessas x\~ellas to increase of sea otters or fur seals within the liniits of the imperial revenues and to make contact with the Alaska or in the waters of Alaska. natises more humane. In 1805, Rezanov, a noble- 011 July 7, 1911, the governments of the man from the court, made a personal inspection United States, Great Britain (on behalf of trip to Russian America and, among other things, Canada), Japan, and Russia signed a Conven- put a stop to the indiscriminate killing of fur seals tion for the Preservation and Protection of (Bancroft, 1886, p. 446). In 1833, Governor Fur Seals (and sea otter) (37 Stat. 1542, hluraviev establisl~edseasons and quotas and areas Treaty Series 564). This convention, com- wliere hunting \\.as forbiciden (Bergsland, 1959, monly kno~vnas tlie Fur Seal Treaty of p. 75; Fedorova, 1973, p. 189). 1911, prohibited pelagic sealing :uld sea otter killing in high seas waters north of the 30th The purchase of Alaska by the United States in parallel of north latitude. It also prohibited 1867 was in sonle respects a setback. Congress did the importation of sea otter skins into tlie concen~itself with the fur seal population by acts four treaty member countries unless the skins of 1868 and 1870, under which a monopoly on had been anthenticated as to legality of taking fur seals at tlie Pribilof Islands was leased to origin. The treaty was termi~~ateclwhen the Alaska Commercial Company. There was no control over the taking of any other fur-bearing Japan abrogated it on Oct. 23, 1940, ef- animal. Eve11 the number of fur seals continued to fective Oct. 23, 1941. The existence of the Aleutian Islands National decline. It was the actions of 1911 that tunled tlie \i1ildlife Refuge stems from Executive Order tide: total protection for the sea otter, a ball 011 1733, dated h.lar.3, 1913. Its full text is pelagic sealing, and a closely regulated harvest of given in Appendix t\ of Chap. 6 of this the fur seals by the government itself. To this has been added recently the State of Alaska's action in volume. trallsplanting sea otters back into portions of their A Convention for the Regulation of \\'haling former range wliere they had disappeared. These \\?as concluded at Geneva on Sept. 24, 1931, measures have been outstandingly successful. Both and canle into force on June 17, 1932. The populations, fur seals and sea otters, arc healthy convention and the international agreements and viable by any standard. following it are the basis for the present It~ternationalCollvention for the Regulation This section lists the succession of pertinent of \\'haling. In making this agreement, the federal la\vs and orders relating to the conserva- various ~vorld go\lernments signing it pro- tion, protection, and management of sea niamnials claililed that they recognized the interests of in the Aleutian area. Also included are treaties and nations of the world in safeguarding for international agreements to {vliich the United future generations the great natural resources States is a party. These statutes and agreements are represented by the \\,hale stocks. They the basis for the status of lnarine nlammals in the recognized tliat the history of \vhaling had Aleutian Islands today. been one of fishing one area aftcr another and of one species after another to such a Acts of 1868 and 1879 applying to tlie harvest degree that it had hecolne necessary to of fnr seals in the Pribilof Islands. protect all species of whales from further A treaty between the United States and Great overfishing. Britain (acting for Canada) was concluded on They further recognized that the \\,hale Feb. 29, 1892, at \\'ashington, D. C., for the stocks \sere capable of nati~ralincreases if purpose of submitting to arbitration certain ~vhalingwas properly regulated. Increases in questions concerning the preser\'ation of f~n the size of whale stocks \\rould permit in- seals. creases in the number of whales that could be On Feb. 21, 1892, the 5211d Congress acted to captured without endangering these tiatu,.al extend to the North Pacific Ocean the resources. Realizing tliat it \\,as in the co111- provisions of earlier statutes protecting fur mon interest to achieve the optimum level of seals and other fur-bearing animals. whale stocks as rapidly as possible by con- fining \vhding operations to those species mru~agementof the fisheries and wildlife of \t.hich \\.ere best able to sustain exploitation, Alaska to the State of Alaska. It also a system of international regulations for authorized the transfer to the state of real whale fisheries rvas established to ensure and personal property of the United States effective conservation and development. The \\~hichis in Alaska and specifically used for International Convention with Schedule of the conservation mcl protection of the \\'haling Regulations, signed at \\lashington fisheries and \vildlife of Alaska. 011 Dec. 2, 1946, entered into force for the Executive Order 10857 terminated federal Unitecl States on Nov. 10, 1948. A 1956 functions in Alaska and transferred property protocol esteaded the application of the held by the United States to the State of convention to helicopters and other aircraft Alaska. and included provisions for inspection. The The Fur Seal Act of 1966 (16 USC 1151-1187; protocol entered into force on klay 4, 1959. P.L. 89-702) repeals the Fur Seal Act of The International \\'haling Convention is 1944. It continues the responsibility of the not generally believed to be as succcssflrl as Secreta~yof the Interior for conser\ration and its high principles \\.oulcl indicate, but this is harvestitlg of North Pacific fur seals and for too sensitive a matter of current history to go administration of the Pribilof Islands reserva- into here. tion for various purposes, including conserva- The lengthy history of the Interl~ational tion of fur seals and other wildlife. It tt'haling Con\~entionis being repeated by the contains provisions specifically protecting sea 1954 International Convention for the Pre- otters. It authorizes contracts for research on vention of Pollution of the Sea by Oil. This convention entered into force for the United fur seals and sea otters with persons and States on Dec. 8, 1961, and sets forth mea- agencies and provides for payment to the sures to prevent pollution of the sea by oil State of Alaska pursuant to the Alaska Statehood Act of 70% of the net proceeds discharged from ships. A resolution on marine pollution adoptecl on Oct. 2 1, 1969, from the sales of fur seal skins taken in harvest programs. It also provides for pay- by the United Nations General hsselnbly ments to Alaska of 70% of the net proceeds ackno~vledged a significant contribu tion by from sea otters taken by the Secretary on the oil to pollution of the sea. It noted also high seas or ~vithin the Aleutian Islands effects that might arise from the exploration National \\'ildlife Refuge. and exploitation of resources of the sea bed and ocean floor. For that and other reasons, The Endangered Species Conservation Act of the Assetnbly convened in 1973 an interna- 1969 (P.L. 91-135) provided broad authority tional conference on marine pollution for the and policy guidance for a comprehensive purpose of preparing a suitable international program for the conservation, restoration, sgreelnent for placing restraints on the and propagation of native fish and wildlife contamination of the sea, land, aud air by threatened with extinction. The Endangered ships, vessels, and other equipnlent operating Species Act of 1973 has replaced it. in the marine environment. In establishing the Marine Mammal Protection The Fish and t\'ildlife Act of 1956 (16 USC Act of 1972 (16 USC 1371-1384, 7 42 a-754) establishecl a comprehensive 1401-1407; P.L. 92-522), Congress found national fish and wildlife policy. It estab- that certain species and population stocks of lished the present U. S. Fish ruld t\'ildlife marine mammals are or may be in danger of Sersice and directed the Secretary of the extinction or depletion as a result of man's Interior to provide continuing research, es- activities. These should not be permitted to tension, and information senices aud to take diminish beyond the point at which they any necessary steps to develop, manage, and cease to be a significant functioning element conserve fishery ant1 wildlife resources. in the ecosystems of tvhich they are a part In 1957, Canada, Japan, Russia, and the United and should not be permitted to diminish States executed a11 intcrilil conrrention on below their optimum sustainable population. conscr\ration of North Pacific fur seals. This Further measures should be taken to re- convention was anended by protocol on plenish any species or population stock that Oct. 8, 1963. has already diminished below that popula- The Alaska Statehood Act of July 7, 1958 (48 tion. In particular, efforts should be made to USC Chap. 2-Alaska, Sec. 6; P.L. 85-508) protect the rookeries, mating groi~nds,and authorized the transfer of administration and areas of similar significance of each species of Sea ~Va~nnzals:Resources and Populatio~l 509 marine maminal from the adverse effects of 1939, pp. 83.168, Department of Commerce, \\'ashing- man's actions. ton. -, 1938, Alaska Fishery and Fur Seal Industries in 1937. Beginning on Dec. 21, 1972, there was an Appendix to the Report of the Commissioner of indefinite complete ban on the taking of Fisheries, pp. 125-150, Department of Commerce, marine mammals and on the importation into \\'ashington. the United States of marine mammais and -, 1937, Alaska Fishery and Fur Seal Industries in 1936. marine mammal products. The Act allo!vs the >\ppendix to the Report of the Commissioner of Fisheries, pp. 326-347, Department of Commerce, Secretary of the Interior to issue permits for Washington. scientific research and public displays. It also Clark, D., 1968, Koniag Prehistory, Ph. D. Dissertation, allolvs the taking of marine mammals by University of \\'isconsin, hladison (not seen by Alaskan Indians, Aleuts, and Eskimos for !\begglen). subsistence purposes or for creating and Cobb, J. N., 1907, The Fisheries of Alaskain 1906, Depart- selling authentic native articles and clothing ment of Commerce and Labor, Bureau of Fisheries when not accomplished in a wasteful tnalmer. Document No. 618. The Act has jurisdiction over walruses, sea -, 1906, The Commercial Fisheries of Alaska in 1905, Department of Commerce and Labor, Bureau of Fish- otters, polar bears, manatees, seals, ~vhales, eries Document No. 603. sea lions, and porpoises. Cook, J. P., E. J. Dixon, and C.E. Holtnes, 1972, The Endangered Species Act of 1973 (16 USC Archaeological Report, Site 49-Rat-62, Amchitka Is- 1531-1543; P.L. 93-205) extends the princi- land, Alaska, USAEC Report HN-20-1045, University ple of protection of endangered species to of Alaska. plants as \vet1 as to fish and wildlife. A new Desautels, R. J., A. J. hIcCurdy, J. D. Flynn, and R. R. category of threatened species has been Ellis, 1970, Archaeological Report, Amchitka Island, established. The definition of fish and wild- Alaska, 1969-1970, USAEC Report TID-25481. Doi, T., T. Nemoto, and S. Ohsumi, 1967, hlemorandum on life is broadened to include any mammal, Results of Japanese Stock Assessment of Ii'hales in the fish, bird, amphibian, reptile, mollusk, Nortl~Pacific, International Commission on \\'haling, crustacean, arthropod, or other invertebrate. Report of the Comn~ission,pp. 111-115, London. The Secretaly of the Interior has been Douglas, J., n.d., Fur Trade Returns, Columbia District and directed to establish a list of endangered and New Caldonia 1825-1857, Archives of British Colum- bia, Victoria. Ms. 408 (not seen by t\begglen). threatened species and to establisl~reyla- Elliott, H. \\I., 1887, The Sea-Otter Fishery, in The tioils for the protection of these species. This Fisisheries I~~destriesof the Uniterl States, U. S. Commis- legislation also provides for cooperation with sion of Fish and Fisheries, Section 1'. History and and assistance to the individual states and h,lethods Text, Vol 11, pp. 483-491, G. Brown Goode foreign go\rernments. Except for certain and Associates. limited purposes and under regulations, the Fedorova, S. G., 1973, The Russian Populntio~~in Alaska importation or exportation of endangered or nrzd Califomin, Lnte 18th Centary-1867 (translated threatened fish and wildlife is prohibited. and edited from the 1971 Russian Original by R. I\. Pierce and I\. S. Donnelly), Limestone Press, Kingston, Ontario (hiaterials for the Study of Alaska History, No. 4). REFERENCES Gard, L. hI., G. E. Lewis, and F. C. \\'hitmore, 1972, Steller's Sea Cow in Pleistocene Interglacial Beach Bancroft, 11. !I., 1886, History of Alaskn, 1730-1885, Deposits on Amchitka, Aleutian Islands, Ceol. Soc. A. L. Bancroft & Co., San Francisco, Calif. Atner., Bull., 83: 867-870. Bergstand, K., 1959, t\lettt Dialects of i\tka and Attu, Golder, F. A,, 1914, Russian Expa~tsion on the Pacific: Tm~ls.Anter. Phil. Soc., 49(3): 1-128. 1641-1850 (reprinted 1960 by Peter Smith, Glouces- Berkh, V. N., 1823, A Chronological History of the ter, hlass.). Discovery of tlte Aleutian Islnnds. Translated by R. A. Goode, G. B., et al., 1884, The Fisheries and Fishery Pierce, Limestone Press, Kingston, Ontario, 1974. (hla- Industries of the United States, Sec. 1, Part 1, 47th terials for the Study of Alaska History, No.5.) Congress, 1st Session, Senate, Mist. Doc. 124, Washing- Bertram, Colin, and Kate Bertram, 1964, Does the "Ex- ton. tinct" Sea Colv Survive? A'ew Scientist, 24(415): 313. Heptner, V. G., 1965, hiore About Steller's Sea Cow, Berzin, A,, E. A. Tikhomirov, and 1'. I. Troinin, 1963,1\'as Priroda, 54(7): 91-94 (in Russian). Steller's Sea Coxv Exterminated? Prirodn, 52: 73-75 (in HrdliEka, A,, 1945, The Aleutian nnd Commander Islnnds Russian) (not seen by Abegglen). nud Their Inhabitants, The \\'istar Institute of Anatomy Bower, I\'.T., 1941, Alaska Fishery and Fur Seal Industry and Biology, Philadelphia. in 1939. Administrative Report No. 40. Appendix 11 to International Commission on ]\'haling, 1967, Seventeenth the Report of the Commissioner of Fisheries for Fiscal Report of the Commission, London. Year 1939, pp. 97-184, Department of Commerce, Jones, R. D., 1951, Present Status of the Sea Otter in I\'ashington. Alaska, \\'ildlife hlanagement Institute, in Tronsnctiotts -, 1940, r\laska Fishery and Fur Seat Industries in 1938. of the 16th i\'orth A inericaiz IYildlife nnd A'atirrol Administrative Report No. 36. Appendix I1 to the Resources Cottference, pp. 376-383, \\'ildlife Manage- Report of the Commissioner of Fisheries for Fiscal Year ment Institute, \\lashington, D. C. Kenyon, K. I\'., 1971, Status of Marine i\lam~nalsin the Rice, D. I\'., 1971, l\'bales and \\'hale Research in the Eastern North Pacific Ocean, U. S. Department of the Eastern North Pacific, in Proceedings of the Conference Interior 102 Statement. on Biology of \\'hales, Luray, Virginia, June 1971, U. S. -, 1969, Tl~eSea Otter in fhe Eastern Pacific Ocean, International Biological Program. U. S. Department of t11e Interior, Bureau of Sport - ,and V. B. Scheffer, 1968, I\ List of the Marine Fisheries and \\'ildlife, North American Fauna, No. 68. b'lammals for the \\'orld, Special Scientific Keport- -,and D. I\'. Rice, 1961, Abundance and Distribution of Fisheries No. 579, U. S. Fish and \\'ildlife Sewice. the Stellet. Sea Lion, J. ~~Inmrnnlog)~,42(2): 223-234. -,and A. A. \\'olman, 1971, The Life History n~td ,andV. B. Scheffer, 1955, 'The Seals, Sea Lions, and Ecology of the Crny ll'l~nle(Escltriclitius robustus), The Sea Otter of the Pacific Coast, Circular No. 32, U. S. American Society of hlammalogists, Special Publicalion Fish and \\'ildlife Service. No. 3. ,andD. I,. Spencer, 1960, Sea Otter Population and Ricks, h'I. B., 1963, l%e Earliest History of Alnskn, Cook Transplant Studies in Alaska, 1959, Special Scientific Inlet Ilistorical Society, Anchorage. Report, \\'ildlife No. 48, U. S. Fish and \\'ildlife Service. Rocquefeuil, C., 1823, Journal d'un Voyage Autour du Khlebnikov, I<., 1835, Biography of Alesnnder Andreiuitch blonde, Pendant les Annees 1816, 1817, 1818 et 1819, Bamnoff: Gouer,tor-General of the Russian Coloi~iesit1 par M. Camille de Roquefeuil, Pontl~ieu,Paris (not seen Anzericn, Navy Printing Office, St. Petersburg (in Rus- by ?\begglen). sian). (Excerpts translated by George Guilsher. Other- Sauer, hi., 1802, An Accorcnt of n Geograp/ticnl and wise not seen by Abegglen.) (Also translated by M. 13. Astronon~ical Expeditio,~ to the A'orthern Ports of Ricks in Tile Earliest Ilistory of Alaska, Cook Inlet Russia (reprinted 1972 by Ricl~mondPublislting Co., Historical Society, Ancl~orage,1963. Also translated by Ltd., Richmond, Surrey, England). R. A. Pierce, 1,inrestone Press, Kingston, Ontario, Scammon, C. I 1874, Tlre i\hrine hla~n,,tnls of the 1973.) Northzoesterrr Const of North Americn (new introduc- irkw wood, J. U. (Comp.), 1975, Bioenvironmental and tion by \I. B. Scheffer. Republication of original work Hydrologic Studies, Amchitka Island, Alaska. Fall 1974 published in 1874), Dover Publications, Inc., New Task Force Report, ERDA Report Bh,lI-171.158, Bat- York, 1968. telle hlemorial Institute. Sims, E. \ 1906, Report on the /flaskan Fur Sen1 Lensink, C. J., 1962, The History and Status of Sea Otters Fisheries, House Document 251, 59th Congress, 2nd in Alaska, Ph. D. Thesis, Purdi~e University, Xerox Session, U. S. Government Printing Office, I\'ashington, University hIicrofilms, Ann Arbor, hlich. D. C. Loy, C. L., 1940, Sea Otter Report of Amchitka Island, Spat~lding,A. C., 1962, Archneological Itrvestigations on r\laska, July 10, 1939 to January 15, 1940. Un- Agnttt~,Aletrtiaa Islaitds, Anthropology Papers No. 18, published Report in F\VS files. hluseum of Anthropology, University of hliclligan (not -, and 0. A. Friden, 1937, Sea Otter Sl~~veyof Amchitka seen by Abegglen). Island, Alaska, July ll to September I, 1937. U11- Spencer, D. L., 1969, Amchitka Bioenvironmental Program published Report in F\\'S files. Interim Report, Sea Otter Sutveys, Amchitka Island, hlarsh, hl. C., and J. N. Cobb, 1911,The Fisheries of Alaska Alaska, /\ugust 27-October 6, 1968, USAEC Report in 1910, Department of Commcrce and Labor, Butcau BMI-171-120, Battelle h4emorial Institute. of Fisheries Document No. 746. Stejneger, L., 1883, Contributions to the History of the -,and J. N. Cobb, 1910, The Fisheries of Alaska in Commander Islands. No. 1-Notes on the Natural 1909, Bureau of Fisheries Document 730. Ilistory, Including Descriptions of New Cetaceans, Proc. -,and J. N. Cobb, 1909, The Fisheries of Alaska in U. S. i\'afl. Alus., 6(4): 58-89 (not seen by Abegglen). 1908, Bureau of Fisheries Document 645. Stephan, J. G., 1971, iliocnvirontnental Surveillance ,andJ. N. Cobb, 1908, The Fisheries of i\laska in Employing I'lrotogrammetric Techniques, BioScience, 1907, Bureau of fisheries Document 632. 21(12): 667-682. hlatbisen, 0. A., 1959, Studies on Steller Sea Lion Sturgis, I\'., n.d. (circum 1799), Fur Traders in thc North (E. jubntn) in Alaska, in Tra,tsactio,ts of tlre 24th Pacific, Archives of British Columbia, Victoria, Pac. Ms. American il'ildlife and Natural Resources Conference, K. 33 (not seen by Abegglen). pp. 345-356, \\'ildlife Management Institute, \Vashing- Swicegood, Lt. Coindr. S. P., 1936, i\mchitka Sea Otter ton, D. C. Survey Expedition. Unpublished memo from Com- -,and R. J. Lopp, 1963, Pliotographic Census of the manding Officer, U. S. Coast Guard Cutter Phelnu, 31 Steller Sea Lion Herds in ~\laska,1956-1958. Special August 1936, (in U. S. Fish and \\'ildlife Sen7ice files). Scientific Ke~ort-Fisheries No. 42. U. S. Fish and Thorsteinson, F. V., R. \\'. Nelson, and D. F. Lall, 1961, \\'ildlife Service. Experimental Hawest of the Steller Sea Lion in Alaskan blurie. 0. 1.. 1959. Foto~o of the Aleution Isln~~dsmid Waters, Special Scientific Report-Fisheries No. 371, Alaska Eeninsu~a;U. S. Fisi and \\'ildlife Service, North U. S. Fish and \\'ildlife Service. r\mericnn Fauna, No. 61. Turner, L. hi., 1886, Coatribtctior~sto the i\'afrtml IIislory Nishiwaki, hl., 1967, Distribution and hfigration of hlarine of Alarkn. Arctic Series Publication No.2. Signal hlammals in the North I'acific Area, Bulletin No. 1, Service, U.S. hny, \\'ashiugtoo (not seen by Ocean Research Institute, University of 'Tokyo. Abegglen). Ohsumi, S., Y. Shimadzn, and T. Dai, 1971, The Seventh hlemorandum on the Results of Japanese Stock Assess- I'anya, John, 1970, hlemo to James A. Ilazper, Director, ment of \\'hales in the North Pacific, lnte~lational Division of Game, Juneau. Subject: Sea Otter Trans- Commission on \\'haling, 21st Report of the Commis- plant. Alaska Department of Fish and Game, June 9, sion, London. 1970. Petroff, I., 1884, Alaska: Its I'opulation, Industries, and \'eniaminov, I., 1840, votes err llre lslnnds of fhe U>~nlaska Resources, in Tenfb Census of the Unifed States, District (tra~>slationfrom the original Russian, Human Val. \'Ill, U. S. Department of the Interior. Relations *\yea File, New Ilaven). Population Estimates James A. Estes * Arizona Cooperative Wildlife Research Unit, and University of Arizona, Tucson, Arizona Feeding Behavior ofSea Otters

Feedi~rgbehnuior and the nbu~rdaitce of sen otters were been submerged and co~rsequentl)~not counted as the studied at Anzclritkn Isla~rd, Alaska, betrueen 1Vouember aircraft passed ouerlread. In 1972 the Amchitka Islnnd sea 1970 nnd Septet>rber 1972. Observations of feeding sen otter populatio~i was estimated at about 6400 aai~~mls, otters su,~gest,to differences between sexes with regnrd lo npproxintately 40% of wlrich were itt the Bering Sea axd the atnount of tinre spent rliuing or on tire surface. 60% of wlrich mere in the Pacific Ocentt. This distribution Appnreittly, rliuing and surface titites both bzcrense br corresponds with tlte distribtttion of sen otter feeding deeper zuater, but tile percentages of time spent on tlte I~abifat.Assumi~tg that sen otters live withi~ttlre 30-fathom surfice ntrd submerged remain npproxintately cotrstont (55-111)depth cotitoar, I estinrnte on island-wide sea otter tlrrotiglrout the day. No differences in either tltepatter~tor density of 59 nttimnls per square tau tical mile of hnbitat tlre percentage of tinte spent feeding, restitrg, or grooming (17 per square kilometer). This esfintnte may be sepnrated were obserued bet~uee~rsummer and winter. into 68 a~timalseer sqtrare ntile (20 per squnre kilometer) in Sl~ore-bnsedcotr~rts of sea otters are cotrsistentl)~higher tire Bering Sea and 54 animals per square mile (16 per tlrnrr nerial courr Is froin helicopters. Tltk prob"b1y is square kilometer) in the Pacific Ocean. I belieue that seueral ottribr~tnbleprimarily to tlre itrcreosed amon?rt of tinte tlrat earlier estimates of tlte number of sea otters at Amcl~itkn can be spent uiezui~zgan area from shore-based stntiotts and Island were biased lolu. thus tlre detection of attimalr tltat otherwise would hnue

The sea otter (E71lrydrn hctris) cvol\~edas an integral published which set forth the results of his o1vn part of the near-shore ecosysteln in the North research, \vhich spanned more than a decade, and Pacific Ocean. Du~ingthe 150 years from the mid- brought up to date virtually everything kno\vn 1700s to the beginning of the twenticth century, about the species. Subsequently it has been the this species was driven to the verge of cxtinctio~l nlost germane reference for several investigations, by \vhite fur traders. Exploitatio~t of sea otter including studies sponsored by the U. S. Atomic populations continued until 1911, at !\~hiclt time Energy Corn~nissionon the sea otter population at the animals \\,ere protected by an international Amchitka Island. convention bet!veen the United States, Russia, Ihed over about one- efforts have produced a confusing array of results third to one-half its original range (Kenyon, 1969). and probably so~neinvalid suggestions of carrent The fil-st estensive account of the sea otter, by populatio~ltrends. Barabash-Nikiforo\r, was published in 1947. In Distribution, fenlale reproduction, and meta- 1969 a monogrwph on the sea otter by I

This chapter considers population numbers and from the approximate position of the animal as feeding behavior of sea otters at Amchitka Island. 1 plotted on a depth contour map. assume that the population is fluctuating through 3. Food items captured. time about some point of stable equilibrium. This 4. Weather conditions. assumption is supported primarily by conclusions The lniui~llumsample size for estimating diving made by Kenyon (1969) and is confir~nedby my and surface times per individual \\.as determined by olvn observatio~lson natural nlortality (Estes and incorporating the sample sariance amoug diving Smith, 1973). I also assume that competition for times (sZ), obtained from a pilot study, and the food is limiting further population growth. The desired confidelice inte~valof mean diving times amount of time budgeted by the population for (2b) into the formula: tl = sz tZ/bz, where t = foraging activity should reflect this limitation. ta/2," (a is 0.05 and v is degrees of freedom determilled by the sample size of the pilot study). MATERIALS AND METHODS From this formula (Steel and Tonic, 1960) I determined the number of observations (u) re- Feeding Behavior of Sea Otters quired for a particular confidence interval (2b) at the 0.05 level of statistical significance. Prelimiuary I observed feeding and diving behasior of sea estimates of mean diving and surface times \sere otters from Nosember 1970 until Septcmbcr 1972. gathered from ttso feeding sea otters. The means AcIost of this work was done during January and and variances of dising and surface times recordecl February 1971 aud during July and August 1972. for these two animals are sho\vn below: During early phases of the study, I randomly selected observation areas in some of the more reulote coastal regiotis of Anlchitka Island. Inclem- Diving Surface cnt \scather and logistic difficulties soon proved this method of selection impractical and forced thc X,sec s2 X, sec s2 use of only those areas \\~hichwere easily accessible Sea otter No. 1 by road. For this reason ICirilof Point and Constan- (female with large pop) 36 106 40 44 tine Point \Irere nlost frequently used during later Sea otter No. 2 phases of the study. Large ntunbers of sea otters (lone female) 22 91 19 10 rvere always present and obsersable at these areas. Most of observations were made along the Bering Sea coast of Amchitka Island because viewing coliclitions \sere geiterally superior there as Because of the nonhomogeneoos sample vari- compared to coastal areas of the Pacific Ocean. ance between the diving times of thc two animals, the larger sample sariance (i.e., 106) was used to The primary objectives of behasioral studies calculate n. I arbitrarily defined the desired confi- were to determine (1) the amount of time feeding sea otters spend submerged, (2) the diulnal pattern dence interval at k10 sec, and thus n = 4.8. A of feeding and other actisity, and (3) seasonal minimum of five obsersations of dising and surface times rvere take11 from each animal; more obsersa- sariation in population feeding behavior. I recorded dising and surface times of feeding tions \sere taken ~shenpossible. Betrseen observa- sea otters. \Vhen possible, the following informa- tions food items were identified with binoculars or tion also was noted during each obsenration: a spotting scope. Food items were simply classified as fish or invertebrate unless they could be more 1. The sex and age of the animal (i.e., male, female, and female nit11 pup). Sex was determined specifically identified. Diurnal patterns in the feeding behavior of sea by the presence of a penis or teats.* All animals otters also were estimated from the same study were classified as either adults or pups. areas at Amchitka Isla~~d.I originally attempted to 2. The approximate water depth in which the observe continuously a single individual animal animal was feeding. \\later depth Isas estimated through long periods of time. This technique was impractical and was soon abandoned. I subse- *Sex determination of sea otters in the field was more quently noted the actisity of all sea otters within difficult than I had originally snspected. After examining the viexving range of 10x binoculars at one-half several subadult males that were held in captivity, I hour intersals and during daylight hours. Sea otter concluded that definite sex determination of anything but activity was divided into the follo\sing categories: adults at close range is not possible. Even at short ranges sexually identifying characteristics of subadult animals are (1) feeding, (2) grooming, and (3) resting. Follo\\r- often not apparent (i.e., the male's penis or the female's ing some experience sea otter activity could be teats). ,\dolt females \vith pups are obvious at long ranges. categorized after only several seconds of observa- Populrrtio?~Estbnntes atid Feeding Beliauior of Sen Otters 513 tion of a single animal. Animals whose behavior NUMBER OF OBSERVATIONS could not be classified or \\'hose bchavior did not 10987654321 12345618910 fall into one of the above categories were not 500 tabulated. hlost of the sea otters I observed could 4 7 (NO.) be classified easily into one of the three behavioral 6:oo +12 136 categories. 7:OO '56 162 The follo\ving criteria were used to categorize 8:w '72 9:oo 23, 149 feeding sea otters: 3 61 1. Possession of a food item. 1o:oo 188 u ;; : t 78 193 2. The exhibition of a characteristic chewing ZI1:oo C 7;; 1132 ~novementin ~vhichthe head moves sharply up and 98 1102 g 12:oo 118; 1100 down. Q ll0t 186 a 1:oo Z 158 3. Rolling in the water to \\rash accirmulated 1131 8118 2 2:oo 141, ,142 food scraps from the fur. "7 139 1 1141 4. The close accompaniment of gulls in search 0 3:OO (65 k 1163 z 162 1 1119 of discarded food scraps. 5 4:OO $601 4 99 m 1121 1130 5. Repeated diving. 5:OO 55 1 ,134 177 Sea otters vigorously cleaning their fur were 6:OO 395 -43 categorized as grooming. Often this involved re- 7:OO 149 1104 peated rolling or violent splaslling in an apparent 8:OO 1111 1 114 attempt to force air into the pelage. At long 9:OO 108 distances grooming behavior occasionally may be -43 1O:W -36 mistaken for feeding. Resting sea otters lie motionless on their backs WINTER I SUMMER in the water or "haul out" on land.* Frequently resting animals congregate into large closely asso- Fig. 1-Numbers of behavioral observations and ciated groups, although solitary resting individuals total numbers of animals observed. Bar lenzth- indi- cates the number of days for which observations were are not uncommon. made at a particular time and season. Numbers at end I observed sea otter activity during the summer of bars indicate the total number of animals obsetved. (July and August) and winter (January and Feb- ruary) seasons. The number of obse~vatioiisand the total nuinber of animals observed during each large numbers of sea otters as compared to other time period, in summer and winter, are shown in areas around Amchitka, Areas C and D were Fig. 1. selected without prior kiio\vledge of either the population structures or the relative abundances of Techniques for Estimating the Abundance animals that occupied the areas. \Vithin each study of Sea Otters area a number of shore-based viewing stations were selected from which animals could be counted with Sea otters !\,ere counted from shore-based binoculars and spotting scopes. The helicopter stations with lox binoculars and a 15 to 60x transported obse~-\fersto these viewing stations. variable-powercd spotting scope and from air by Immediately preceding or folloxving a shore-based observers who were seatcd in an Alouette I1 heli- count in a given area, sea otters were counted from copter. The boundaries of shore-based counting the air. Aerial counts originally were made by a areas are given in Fig. 2. Two study areas each mere single observer who was seatcd beside the pilot. selected on the Pacific Ocean and Bering Sea sides However, later efforts showed that two observers of Amchitka Island (labeled A tlirough D in counting animals on either side of a line bisecting Fig. 2). Area A is located adjacent to the Cannikin the longitudinal axis of the aircraft provided higher site along the Bering Sea coast.+ Area B was numbeis, and this technique was subsequently selected with prior knowledge of the sea otter used. Aerial counts with two observers were also population structure (Kenyon, 1969; personal ob- made for the entire island. servation) and the belief that it contained relatively Animals were counted from shore at each viewing station within a given study area. \\'hen people and equipment were available, counters *"Hauled out" is commonly used to describe marine began at either end of the study area and worked mammals resting out of the \vater. tCannikin is the code name of an underground nuclear toward the center. After several visits to each study detonation by the U. S. Atomic Ene1.g~Commission in area, obser\.ers became familiar with confines of 1971. viewing stations and were able to avoid overlapping 514 Estes

BERING SEA

PACIFIC OCEAN

Fig. 2-Map of A~nchitkaIsland showing tile location of shore-based counting areas. counts of contiguous stations. All obsen~ablesea right front of the aircraft. Obselvers were locatect otters \vere counted by scanniug the vielvillg at the left-front and right-rear seats, and the stations from border to border. Some diviilg recorder sat in the left-rear seat. A zigzag pattern allimals as \\'ell as others in obscure locatioils on \\?as flown through kelp beds or sltallo\cr areas the surface inevitably were nlissed. Ho\vever, virtu- extending far offshore. The same pilot \\,as used ally no animals \\,ere counted txvice during a scan throughout the aerial coiintiitg operation, and of the vie\ving station. At \fiewing station bound- flight patterns remained relatively coilstant after aries, errors were introduced either by couilting the first several co~mts.\\re frequently fle\v several one animal twice or by not including it in either miles offshore at the east and west ends of station. I believe that the mean of these errors Amchitka and at several other places on the Pacific tends toward zero because of the large number of Ocean coast because these areas are sltallo\\~erthan observatiolls and the effort that \\'as made to coul~t those along the Bering Sea coast. only those aniinals \\'ithi11 the predescribcd bound- A subjective classificatioll of sea state and aries. Kelp beds or offshore rocks \irere used as xveather collditioils was assigned at the bcgirtiliilg markers or reference points between vie.iving of each couinting operation. These conditions stations. frequently changed slightly during the census or \\re made aerial counts arouild the entire coast between areas aro~undthe island. of Amchitka Island or along large coastal segments of the islaud bet\\reen Jutle 1971 and September 1972. A number of people assisted with these Vierving ooerations. devendinr- on availability. An addi- classification Description of conrlitions tional person recorded the number of sea otters observed in 45 segments into which the Alnchitka 1 Ocean surface glassy. Air clear. Sea otters can be seen at maximum range of 10x coastline \\,as divided clurillg all island-wide counts binoculars. Viewing conditions excellent. (Fig. 3). 2 \\'indselocitp 5 to 10 knots (2.6 to 5.1 The helicopter \vas flolen counterclockwise mlsec). Light riffle on ocean snrface. around Amchitka at an altitude of approximately Sea calm. Sea otters far offshore may not be seen. 150 ft (45 m) and sufficiel~tlyclose to shore that 3 \\'ater choppy. !\'bite caps pxesent. the near-shore observer could easily see animals 4 Heavy seas. onshore. The helicopter pilot was located at the Poptiintion Esti~natesnttd Feediitg Behavior of Sea Otters 51 5

Fig. 3-Amchitka Island showing the shoreline segments uscd in island-\\videaerial surveys.

Aerial counts \\,ere made only when the vie\\,- times of sea otters feeding in water estimated to be ing classification was 1 or 2. Reliable shore-based 0 to 5 fathoms (0 to 9 111) in depth were 47 and 41 counts could only be made n41en viewing condi- sec, respectively. The mean diving and surface tions were nearly perfect (i.e., classificatio~l1). times of sea otters feeding in water estitnated to be Viewing classificatio11s 3 and 4 are equally inade- between 5 and 15 fathoms (9 to 27 m) in depth quate for both shore-based and aerial counting. were 83 and 84 sec, respectively. Feeding sea otters spend 52% of the time submerged and 48% of the time on the surface. RESULTS I attempted to obtain diving and surface times from sea otters feeding in water deeper than 15 Feeding Behavior fathoms (27 m). Unfortunately tlte aninlals took The diving and surface times of feeding sea alarm before I could get close enough to obtain otters at Amchitka Isla~ld are summarized in any data. Therefore most of my observations on Table 1. The sex of the animals, the approximate feeding sea otters are from areas less than 5 depth of tvater in \vhich they \\.ere feeding, and the fathoms (9 1n) in depth and where they could be food items they were eating are also includecl. \vatclted from shore. No differences between sexes in either diving The available data are insufficient to indicate or surface times of feeding sea otters are apparent differe~lcesin diving behavior between sea otters from these data. Female sea otters averaged 50 sec feeding on fish or invertebrates. Fisll is an impor- per dive and 43 sec per surface period. Males tant sea otter food item in some areas (Lensink, averaged 50 sec per dive and 41 sec per surface 1962; Kenyon, 1969; Burgncr and Nakatani, period. 1972). My observations of feeding sea otters Feeding in deeper water apparet~tlycauses an indicate that fish are take11 opportunistically md increase of both diving and surface time intenrals, that individuals cannot be classified exclusi\~elyas although this conclusion mnust remain tentative either fish or invertebrate feeders. because of the small amount of data from animals Diurnal patterns in the activity of sea otters arc feeding in deep \\rater. The mean diving and surface summarized in Fig. 4. No differences in feeding, Table 1-hlean Diving and Surface Times, Estimated Water Depth, and Forage lten~sof 25 Sea Otters Observed Feedine at An~cllitkaIsland

Status of hlean diving hlean surface \Yater anin~al time, sec timc, sec depth, fathoms Forage items

Female Ilnknown Female Unknorvn Female Unknown Female Unknown Female Unknorvn Female Unknown Female Unknown Fernale Invertebrates Female Fish and invertebrates Female Fish Female lnvertebrates Female lnvertebrates Male Invertebrates Male Invertebrates Male Invertebrates Male Invertebrates Female with pup Unknown Female with pup Invertebrates Female with pup Invertebrates Female with pup Invertebrates Unknown Unknown Unkno~vn Invertebrates Unknown Invertebrates Unknown Invertebrates Unknown Invertebrates Unknowl Unknown - ?Female -XMale -YO to 5 fathoms -x5 to 15 fathoms xTotal

resting, or grooming behavior are evident betxveen 3:30 p.m. Grooming activity remains relatively summer and ~vinter.These seasons represent ex- constant throitghout the day during both xvinter tremes in daylight and \\reather. Therefore diurnal and summer. patterns in sea otter activity probably are relatively Ac1y observations on feeding sea otters indicate constant throughoi~tthe year, at least regarding the that about 52% of their time is spent submerged. amounts of time budgeted for feeding, grooming, This information, in conjunction with the diur~~al and resting. feeding patteru, can be used to calculate the Two peaks in feeding activity, with concurre~lt expected perceutage of the population that is declines in resting activity, are evident during submerged during daylight hours. Let ai equal the sunlmer days. A peak in feeding occurs about percentage of the population feeding at time ti. 8: 00 a.m. and another peak occurs about Grooming and resting sea otters spend 100% of the 5:30 p.m. Bering Standard Time (subtract 1 hr for time on the surface and therefore need not be local su11 time). During winter it is dark at these considered. Thus 0.52ai gives an estimate of the times. Ho\\~ever,a decline in feeding activity from percentage of the sea otter population (excluding the morni~lgpeak is apparent. Diurnal patterns of small pnps) that is submerged at time ti. resting activity of sea otters are imrersely related to The expected percentage of the population patterns of feeding activity. The highest percent- that is submerged as a function of time of day is ages of resting animals were during early morning sho\vn in Fig. 5. The increased percentage of the and late evening. An extended peak in resting populatioil that is submerged during morni~lgand activity also occurs between about 9:30 a.m. and evening compared with the middle of the day Popnlatio~aEstivlates aild FeedilfgBehauior of Sea Otters 51 7

c +0 u 80 0 + 60 m W a 40

20

0 4:30 6:OO 830 1O:oO 12:OO 2:OO 4:W 6:OO 8:OO 1O:OO BERING STANDARD TlME WINTER

BERING STANDARD TlME SUMMER

Fig. 4-Diurnal and seasonal patterns in feeding, grooming, and resting acti\,ity of sea otters at Amchitka Island. corresponds \sit11 increased feeding activity. Diur- observed during the \\,inter of 1971, 1095 (or nal patterns during summer and winter are gen- about 62%) were feeding. Sixty-five percent (1910) erally consistent, aitd the percentages of the of 2918 sea otters obsenred duri~tgthe summer of population that arc submerged are nearly equal 1972 \sere feeding. during both seasons. The sea otter population at Attu Isla~tdwas ICenyon (1969) presented data collected at observed during a 1-day visit to Cltichagof Harbor tlmchitka Island ~vhich indicate that during the on June 22, 1972. This populatiolt probably con- month of August sea ottcrs spend 55% of the time tained fewer than 100 aitiinals (ICenyon, 19G9; feeding. data are in general agreement, al- personal obsen~atioit).Several sea ottcrs observed though they sho\v a somcx\~hathigher percentage of at Attu \\,ere feeding on ~nitssels(~l.[odiollis sp.), time spent feeding. Of 1758 sea otters that I false jingle shells (Porlocles~nussp.), and sea urchins 518 Estes

-July-Augult ---- January-February

5%30 0 LU em w

5:OO 7:w 9:OO 11:OO 1:W 3:OO 500 7:W 9:OO BERING STANDARD TIME

Fig. 5-Diurnal and seasonal patterns of the estimated percentage of sea otters tlmt are sttbmerged at Alnchitka Island.

(Strongylocer~trot~~s/~ol)~a~a~~tI~z~s). One sea otter parameters [Var (X); Var (Y); Cov (X,Y)] , which \\.as observed pounding two mussels together on its in turn is possible only when repetitive observa- chest. This beha\lior dso is common to sea otters tions of Xi and Vi are available at some assumed off the coast of California and is used to break fixed value of the independent variable gi. How- open food items with exoskeletons that are too ever, if a linear model, such as Vij = p + XiP + eij, is thick for the animals to open with their teeth. Sea fit to the data of this study, there appears to be no otters at Atnchitka rarely pound food, apparently means of separately estimating p, 0, and the because few large inrrertebrates are available owing variances of X and Y. For this reason classical to intense food use there. regression procedures in which the Xi are consid- I observed no sexual segregation in the sea ered fised will be used. otter population at Attu. This contrasts \\'it11 the Model 1 assumes that the rcgression of shore- situatiotl at Amchitka where sexual segregation is hased to aerial sea otter cou~~tspasses through the well definecl. origin. Intuitively, this is a logical assumption because one x\rould expect shore-based and aerial Estimates of Sea Otter Abtuldance counts in a give11 area to be related by a simple ratio. However, at A~ncliitkaIsland there are large The results of island-\\ride aerial sca otter numbers of sea otters in all sufficiently large counts are sho~vi~in Table 2. The results of coastal segments, ild therefore the regression simultaneous shore-based and aerial sea otter counts fttnction near the X aiid Y axes is only conceptual. \vhicli \$,ere obtained from thc study areas shotvn in Realistically, shore-based counts and aerial counts Fig. 2 are given in Table 3. \\ithiu relatively large study areas [i.e., about 10 I estimated the number of sea otters at miles (16 km) of coastline] are both much greater Amchitka by considering sitnultaneously obtained than zero, and a best-fit line through these points shore-based and aerial counts and island-\vide aerial may \\'ell depart significantly from the origin. counts. I assinne that shore-based counts accu- By rnodel I, rately represent the abmtdance of sea ottels within localized study areas. A least-squares technique was used to estimate the sea otter population munber (Yo) at Amchitka where Yi = ith shore-based sea otter count (esti- Island because, if certain assuinptions are satisfied, mated population number within study it provides an unbiased esti~nateof the paratneter area) (vector). Tn'o linear models for regressing shorc- Xi = it11 aerial sea otter count urithin the based connts from aerial counts were investigated. study area Strictly speaking, shore-based cou~lts (Yi) and PI = slope of the rcgrcssion equetion aerial connts (Xi), respecti\,ely, are random vari- E, = error ables representing the ith simultaneous obselva- tion. A technique for analyzing linear models \\,hen For an estinlate of the total number of sea X and 1' both arc ra~idom(Acton, 1959) is based otters at Amchitka (Yo), i! is sufficient to substi- on the estimation of any t\vo of the follorving tute Xo into Eq. 1 to give Yo = >;,,PI, where Sois Popnlatiorr Esti~~?ates(III~ Feedirty Bel~nuiorof Sea Otteu 519

Tablc 2-Numbers of Sea Otters Coontrd \\'itl~inAreas Indicated h Fig. 3 Daring Isla~~d.\~'i~lcAerial Surveys

8/23/72 8/25/72 Proportion Proporti081 Proportion Subttnil of of of or Pacific No. subtotal No. subtotal No. subtotal No. subtotal

2 1 Subtotal Percent of total

6/5/72 8/23/72 8/25/72 9/1/72 9/11/72 Proportion Proportion Proportion Proportion Subunit of of of of of Bering No. s~~btotnl No. sabtolal No subtotal No. s!aBlotal No. No. subtotal

101 27 46 50 51 41 23 35 88 90 53 46 22 26 39 39 53 98 7 1 llOt complete 1 935 Percent of total 'l'olal 520 Estes

Table 3-Compxison of SLore-Based and Aerial Sea Otter perimeter. \Yitlt this method the line is not being Counts from Selected Study Areas Acourld A~nchitkaIsland* extrapolated beyond the clata but rather is the sum of i~iterpolatedvalues. The analysis of variatlce or Shore-based Viewing Sl~ore Study to aerial model I1 is given in Tablc 4. Elelncnts of the Date conditio~~sbased Aerial areat coitllt ratio parameter vector (i.e., p and 0,) are estimated by the normal eqaations. To test thc significa~lccof p, 5/20/72$ I1 (rain) 192 50 111: 3.848 I set up two hypotheses: Ha, y = 0, and Ha, y f 0. 5/21/72 I1 (clear) 641 191 D 3.365 By I-I. [total st1111 .of squares - (sum of squares 5/25/72 I1 346 172 A 2.01 e2) 5/25/72 11-111 420 215 D 1.95 P2 + sum of squares ez)]/(mean square is 5/26/72 11 343 185 C 1.85 distributed as F1 ,8. The level of significance \\'as arbitrarily chosen at 0.05; thus P(Fl .8 5.32) = 5/27/72 I1 382 221 A 1.73 > 5/29/72 I 602 186 D 3.24 0.05. Because 253 (the observed value of F) 9 5/29/72 1-11 452 202 I\ 2.24 5.32, Ho \vas rejected, and inode1 11 \\,as used to 6/07/72 I 450 288 A 1.56 estituate the island-\vide population number. Fro111 8/24/72 I 521 297 A 1.75 model 11, Yo= 1581 + (1.20)(4042) = 6432 sea 8/24/72 1 797 436 D 1.83 otters (an arithmetic error has heen poitlted out to 8/31/72 I1 501 348 I\ 1.44 llle in my earlier estilnate of 6918 sea otters). hleao 1.96~0.17n=10 Under stated assumptions of the model, this statistic is a11 unbiased estimate of the llumber of "No data were obtained from area B. sea otters at Amchitka. A 95% confidence illterval +See Fig. 2 far location of there study areas. around this estimate is 5226 to 7638 animals. $Only segments 1 to 4 counted. §Not used io later analysis. Figure 6 sho\vs 30- and 50-fathom (55- and '31-m) depth contours around Amchitka. \\'itbin the 30-fathom depth contour, these are 38 scguare miles (or 35% of the total area) on the Bering Sea the island-\vide aerial count (4042 sea otters). The sicle of Amchitka and 72 square miles (or 65% of 4042 sea otter count was the highest island-reidc the total area) on the Pacific Ocean side (ICenyon, aerial count. I believe this count is lnost nearly 1969, p. 150). Data from island-\vide popillation representative of aerial coultts made \vithin study surveys suggest that a greater proportion of the sea iNeas under near ideal vielving conditions. \\'it11 the otter population lives on the Pacific Ocean side standard least-squares technique of parameter esti- than on the Bering Sea side of the island. On mation (i.e., by evaluating the normal equations), Aug. 23, 1972, x\dlen viewing cotlditions mere 0, = 1.82. A11 analysis of variance of model I is co~nparableon both sides of Amchitka for count- given in Table 4. ing sea otters, 40% of the allimals were observed on kIodel I1 assumes that the regression line is the Bering Sea side and 60% were observed on the linear but not collstrailled to pass through the Pacific Ocean side. On Sept. 11, 1972, the day of origin. By nlorlel TI, the highest aerial count of the entire island, 44% of the animals were seen on the Bering Sea side and 56% \Irere seen on the Pacific Ocean side. Thus the distributio~l of the population bct\veen the two \\,here y is the Y intcrcept, 0, is the slope of the regression equation in modelII, and E, is the resitlual error in motlel 11. Table 4-Analysis of Variance of &lode1I and hlodel I1 Equation 2 is w estimator of shore-based sea Regressiors of Shore-Based vs. Aerial Sea Otter Counts otter cou~lts (or the estimated 11uniber of sea Degrees otters) u,itltiu the ith study area. If y proves Source of freedom Sum of squares Near square significant, the tii must be sumlned around the cntirc island rather than simply substituting So for Model I Xi (as in model I) because y must be accoutltcd for on an island-\vide basis. Thus Due to 6, 1 2384250 Due to el -9 -102298 11366 'Total 10 2486548

Alodcl I1 Due to (2.h) 2 2412847 \vhcrc yI = y, = . . . = p, and 11 is the number of Due to €2 8 73701 9213 segmcnts of coastline (\\~hicllare of similar length - Total 10 2486548 to the stucly areas) contained ~vithinthe island Po/)tclation Estin~atesand Fee~li~lgBel~auior of Sen Otters 521

BERING SEA

Pig. 6-The 30- and 50-fathom depth contours at Amchitka Island, sicks of Amchitka is similar to thc distribution of animals per square mile (16 per square kilometer) available habitat. of habitat in the Pacific Ocean. On the basis of ICenyon's (1969) discussion and Variable weather conditions and variations in my own observations, to estimate roughly the sea the distribution of sea ottcrs arc probably the two ottcr density at Amchitka, I assumed that most of most important factors influencing the numbers of the sca otters are unable to dise deeper than 30 animals counted during aerial surveys. I previously fathoms (55 111). I also ass~unedthat the entire area mentioned that flat glassy cahn seas and high cloitd within a particular depth contour is usable sea cover provide the most ideal counting conditions. ottcr habitat and of equal value to the population. That the distribution of anim;tls is also important is Txvo factors \vhich also may bc important but evident from thc results of t\vo island-wide connts. \\rhicl~I ha\ze chosen not to consider in thc interest On Aug. 25, 1972, we counted 3524 sea otters of simplicity (and lack of information) arc varia- (island \vide) with near ideal vie\sing conditions. tion in productivity as a function of depth and On Sept. 11, 1972, \ire observed 4042 animals with heterogeneity of substrate type ancl configuration. vie~ving conditions of light surface chop and I csti~natc that 6432 sea otters live at occasional rain. On August 25 sea ottcrs were Amchitka, 60% of \vhich are on the Pacific Ocean scattered throughout the kelp beds and offsl~ore side end 40% of \vhich are 011 the Bering Sea side areas \\'hereas on Septenlbcr 11 large numbers of of the islantl. By assutning that these animals occur otters were bunched at the outer edges of kelp within the 30-fathom (55-111) depth contour, I beds. Even though superior vielving conditiot~s calculated an island-\\ride sea ottcr density of 59 existed on August 25, a greater island-!vide sea animals per square nautical mile (17 per squ;s.e otter count \\.as obtained on September 11 bccause kilometer) of habitat. This estimate nlay be dividcd the aniinals occurred in groups and thus were easier into 68 animals per squarc mile (20 per square to connt than on August 25 when they \\.ere nlore kilometer) of habitat in the Bering Sea ant1 54 dispersect. DISCUSSION congregate into groups at traditionally used areas. If these areas are not included in the obsewation Population Time Budget areas, a bias will exist toxvard feeding activity. Conversely, when poor vic\ving conditions exist, a Prolonged rough seas certainly cause many sea bias to\vard resting activity ~voulclbe obsenred in otters to scek shelter near sltore (Lcnsirtk, 1962; i111 area where sca ottcrs are resting. For this reason ICenyon, 1969). Ilowever, I lime observed individ- the technique is valid only on days \\,hen visibility uals feetling in cstremely rough water, and I also is such that a sufficiently large (representative) have notcd tltat the number of animals ltauleci out expanse of ocean can be observed. R4y opinion is or concentrated in sltcltered areas during severe that this generally should include at least several storms never approaches the numbcr of anin~als miles of coastline. Furthermore, home range, terri- knotvn to inltaliit tl~oscsame areas. I therefore toriality, individual movements, and factors that conclude that a substantial portion of the sea otter relate to spatial strategies of sea otters are not \\,ell population at Amcltitka is able to tnaintain itself in understood. In view of thcse unkno~vns, any the open sea despite inclement \veatlter. This supposition tliat a representative segment of the conclusion is supported by the observation that population is being observed should, at the very certain sea otter populations in other geographical Icast, be considered \\.it11 skepticism. areas (i.e., north of Unimak lsland in southcrn Bristol Bay) rarely, if ever, comc ashore In the results section I ntentioned a small (I<. Schneider, personal communicatio:~). No\vever, discrepancy bet\uecn my data ancl data collected the severe storms common to the Aleutian Islands by ICcnyon (1969) on the percentage of titile sea make observational work difficult, and iutdetected otters budget for feeding. Differences bet\veen my changes in population beltavior may accompany data and ICcnyon's may be the result of different changes in the weather. On calm days at least, I study tcchniques or real cltangcs in feeding activity have seen sea otters at distances of about 2 miles since 1955 \\?hen his data were collected. As I froln shore. suggested earlier, tlte ciilferenccs that ltly data indicate between summcr and \\inter feeding activ- The similar diurnal tittle patterns and pcrcent- ity probably do not represent changes in popula- ages of feeding activity during siunmcr and winter tion foraging strategies. (Fig. 4) suggest either that sea ottcrs feed during hours of darkncss in tvinter or that less food is In contrast to Amchitka, tlte sea otter popula- consu~ttedby tlte populatiolt during tvinter than tion at Attu is far bclo\v equilibrium density. This during sinnmer. Because of tlteir high and contin- probably resalts in an abuudant food supply for those animals living at Attu. Sea otters obse~-\recl ual nutritional requirements (ICenyon, 1969; Mol~ison,Rosenmann, and Estes, 1974; Morrison, feeding at Attu \\!cre all near shore [I estimated 200 yards (183 m) or lcss], ancl they moved only Kosenmann, and Estes, Chap. 23, this volume), I small ciistanccs while feeding. Feeding sea otters at believe that decreased food consumption over long tlmchitka frequently arc scattered to ovcr a ntile pcriods of time is unlikely. ICenyon (1969) pre- front shore, anti oftell they move considerable sented evidence that sea otters liavc xvell-developed distances during a feeding period. These behavioral tactile and olfactory senses, and thus it is not differences are probably related to an abuntlant inconceivable that they can feed in the absencc of near-sl~oi-efood supply at Attu. Conversely, the sea light. Ho\vever, no data are available which provc tltat sea otters are nocturnal feedcrs. Therefore tlte otter food supply at Amcltitka probably is being significance of Fig. 4 rcntains unresolved. maximally exploited. Sea otters must therefore search more and dive to their l~lasintu~ncapabil- The apparent increase in grooming during ities to obtain sufficient food under thcsc circum- winter over stunmer (Fig. 4) probably is not real stances. This probably results in greater amo~untsof but probably is an artifact of limited offshore time spent in foraging activity by sea ottcrs at visibility during \vinter months. At long distances Amchitka compared ~vitltthose at Attu. Altltouglt grooming animals arc difficult to distinguish from I have no quantitative data from Attu to compare those \vItich are fccding. Tlterefore a sntaller \\,it11 the diurnal feeding pattern at Amchitka, ntost percentage of the population tvas identified as of the animals I saw at Attu were resting. grooming during wulttmer months. A liniitatio~i of the observational technique Sea Otter Popnlation Estimates \\~lticltI ha\-e used is that a superficial kno\vledge of A the structure and behavior of the local sea otter For a paralnctcr estinlatc 0 to bc of practical pol~ulationis required before an assumption can be utility, it should be unbiascd and relatively prccisc. made tltat the area is representative of the entire By definition, a paramcter estintate is iunbiascd if popi>lation. Fur example, resting sea ottels tend to tlte expected value of tltat estimate E(0) is equal to Estimates atrd Feediq Belravior of Sen Olters 523 the parameter 8 for all 0. In this study 8 is the total (46 m) altitude and at a masimiun~speed of about number of sea otters at Amchitka Isli~nd.Bias is 70 knots (130 km/hr). In 1970, 3927 sea otters the rliffercnce between the espected value of the were counted during island-wide censuses. From parameter estimate Aand the true value of the this count the population at Amchitka was esti- parameter [i.e., E(0 8); Hoel, 19711. Direct mated at slightly over 4000 animals (U. S. Fish and population counts are almost always biased lo\\' as \\'ildlife Se~vice,unpublished data). estimates of population numbers because some Stepha11 (1971) used oblique infrared color individuals arc invariably missed. photography to estimate the number of sea ottcrs An estimate should also be relatively precise. at Amchitka. He pltotogsaphed population cotlcen- From a set of estimators of 8, the estimate that has trations \vith a llelicopter-mounted camera and minimum variance is referred to as the best visually counted additional scattercd a~imals.This estimate. Frequently the estimate with minimum technique led hiin to esti~natc that Amchitka variance is not unbiased and thns perlvaps not best Island contained . . . "its many but not more than in terms of study objectives. Therefore, \vhen one 2800 sea ottcrs" during 1969. selects an esti~nationtechnique, the tolerability of The primary advantage of fixed-wingetl aircraft bias and lack of precision should be carefully (airplanes) as census vehicles is that large areas ca11 consideretl. Obviously estimates with large and be covered in relatively short periods of time. unknol\.n bias are only \,aluable as population Considering the hiundrecls of miles of shoreline indexes and are relatively ~vorthlessas estimates of inhabited by sea otters, from the western Aleutian population numbers. Con\,erscly, highly variable Islands to Prince William Sound, this technique was estimates, even though they may be unbiased, are most practical for early U. S. Fish wd \Vildlife of little value it~~lessa large nun~berof repetitions Sewice studies. The prinvary contribution of these are available. Furthermore the variance of an studies \\.as the establishment of tre~ldsin sea otter estimate can be determined through repetition of distribution and density over the entire Alaskan the sample whereas bias may remain obscure unless range of the species. Rclany animals undoubtedly specific techniqt~esare rlcveloped to evaluate its were missed because of litnitations imposed by the magnitude and direction. For this reason ma11y large rapidly flying aircraft. investigators it1 wildlife biology are intuitively Itenyon (1969) concluded from his studies that Inore aware of variance than bias. Thus the the majority of sea otters live within the 20-fathom concepts of bias and precision are critical, and they (37-in) depth contour, and rarely they are found in must be considered carefully \vhen evaluating waters as deep as 30 fathoms (55 111). Although population estimates. 1 will consider the sea otter there are occasional reports of sea otters diving in population at Amchitka with these concepts in deeper areas (It. Schneider, personal comtniunica- mind. tion), Itenyon's conclusion defines the area within The first large-scale studies of sea otters in w11ich counting efforts should be concentrated, Alaska were done by the U. S. Fish and \lfildlife This conclusion also implies that coastal areas Service. Usually they censused sea otters fro111 a where the subtidal platfor111descends gradually !\,ill DC-3 aircraft flying at about 120 knots (222 be Inore difficult to census than areas where the km/hr) between altitudes of 200 and 400 ft (61 sea otter habitat is narrower. and 122 m). Two obse~versin the cockpit countecl I believe that tl~cnumber of sea otters esti- sea otters. From these aerial coiunts the nnnlber of mated from this study was biased low. In view of sea otters at Amchitka was estimated to be fronl teel~~~iqi~esand the scope of these investigations, about 1500 to 2500 animals between the years no data were collected svith tvhich to estimate the 1949 and 1965 (Kenyon, 1969). magnitude of bias. Additionally, the relative preci- On the basis of aerial counts and counts made sion of these estimates is unkno~vnbecause the from headlands ~vithbinoculars and a spotting censuses were not replicated over sufficiently short scope, Lensink (1962) estitnated that between time periods. 4000 and 7000 sea otters lived at A~nchitkain Data obtained from U. S. Fish and \Yildlife 1956. These estimates were tlismissecl by Kenyon Scn~icehelicopter censuses \vhich began in 1968 (1969) and apparently were not considered seri- are su~nmarizedin Table 5. Only maximum counts ously in subsequent work. are included for each sulvey period. Helicopters \irere first used as vehicles from \\'atson, Jolly, and Graham (1969) shoxved that \vhich to count sea otters by the U. S. Fish and helicopters are significantly better \vhicles than \\'iltllife Service in 1968. The technique and prelim- airplanes from tvhich to census terrestrial ma~nmal inary results were discussed by Spencer (1969). A populations. The prinvary advantages of helicopters single obsenper seated beside the pilot counted sea over airplanes as vehicles for counti~lgsea otters are otters as the aircraft was floxml at about 150 ft gseater mancu\~er;ibilityand slower flying speed. Table 5-Summary of Aerial and Photogrammcuic Sea Otter Counts of An~chitkaIsland Made Between 1968 and 1972

hlaxirnum count 1968 1969 1970 1971 1972

Bureau of Sport Fisllcries and I$'ildlife* 2302(7)t 3927(5%)t 3241(l)t Battellc Biernorial Institute$ 2354 2773 University of Arizona*

*Visual counts. tNumber in pnrcntheaes is the numbel-of counts, $Visual and pl~otographiccounts.

Thos kelp beds, offsliore islands, and otlier\vise Sea otters also tend to congregatc into large i~laccessiblc areas can be tl~orot~ghlysearched. groups at tiines and to scatter in a more random Consequently estimates of sea ottcr il~uinbersat spatial distribution at other tiines. Clumped distri- Amchitka Island during these studies are almost butions of sea otters tend to increase the size of twice as large as earlier estimates obtained with thc total counts. airpliules. A summary of the sea ottcr counts that were Because of the time inte~valbetween Ildition~l are distinctly u~lprecise (i.e., the counts, ivllen disadvantages. The cost is very high. A large repeated, arc highly variable). This high variability amount of film \vould be required to cover all the apparently is correlated primarily rvith variable sea otter habitat het\\~censhore and the 30-fathom viewing conditions and an unpreclictable spatial (55.111) depth contour, even along relatively sillall distribution of sea otters. Poor \ie\ving conditions espailses of coastline. Species are also difficult to are caused by inclcnlent \\.cather, which prevails identify from photographs (i.e., sea otters may be almost contini~ouslyin the Aleutiau Islands. Strong- confused ~vithharbor seals). \\incls may blo~va; any time of the year, and fog With a spotting scope from shore-based sta- and rain are common during calmer months. Sea tions, Lensink (1962) coiuuted about 1.75 times ottcrs are most easily seen \\,hen seas are glassy inore sea otters i11 a given area than he was able to calm and the air is clear with a high i111brokcn sce through binoculars. This suggests that many overcast. Such icleal vieu'ing conditions arc never aitinials at Amchitka Island are distributed far con~~noilbut prevail most frec~uentlyduring late offshore. Lensink's estimate of the number of sea hshy through early June and late August through ottcrs at Amchitka Island has been criticized on latc September or early October. the basis of overlap betn'een contiguous co~utting Estiltlates and Feeditlg Behauio~of' Sea Otters 525 areas and localized movement of animals. The U. S. submerged at any given time. Animals which are on Fish and \\'ildlifc Sei~riceobservers believed that the surface but which sinlply are not seen contrib- thcy \irere seeing 50 to 75% of the sea otter utc further to lot\' bias. Thus I believe even the population during counts from airplanes and that highest acrial connts are biased 10x1, by 30% or Lensink's estimates \\,ere biased high (ICenyon, more. Shore-based observations on days \vhen 1969). I believe that Lensink's technique, although surface visibility is very good provide a crudc not statistically designed to provide an unbiased lneans by which to account for this bias. estimate, applies a logical estimate of correctioil for bias. The lack of replication precludes any comment on precision. There are several disadvantages to the shore- based counting technique. During 1.ensink's study This work was supported by contract the lack of a helicopter necessitated travel by foot AT(21-6)-520 to the University of Arizona. I thank between counting areas. This was a time-consuming C. Abegglen, R. Anthony, S. Brown, G. V. Byrd, process which limited the amount of coastline that J. Faro, R. Glinski, C. Hardy, K. Iles and hers and Availability of Ringed Seals in the Eastern Procedures of Statistics, hlcGraw-IIill Book Company, Canadian Arctic. Arctic. 14: 162-175. New York. hlorrison, P., M. ~osenmalk,and J. A. Estes, 1974, hfetab- Stephan, J. G., 1971, Bioenvironmental ~u~veillance~m- olisnl and ~h~~~~~~~~,f~ti~~in tl,e seaotter, p,Lysiol, ploying Photogrammetric Techniques, BioScience, 21: --Zool.. 47:- 218.229.- - - - .. . - - . 677-682. Spencer, D. L., 19G9, Amcllitka Bioenvironmental Program. Sea Otter Sorveys, Amchitka Island, Alaska, \\'atson, R. M., G. M. Jolly, and A. D. Graham, 1969,T\vo August 27-October 6, 1968, USAEC Report BhI1- Experimental Censuses, Enst Afr. Aqic. For. J., Special 171-120, Battelle hlemorial Institute. Issue Vol. XXXI\': 60-62. Ecological John F. Pr~Zrnisano' Fisheries Research Institute, University Interactions of \Vashinpton Ju,mes A. Estest Cooperative \l~ilcIlifeResearcli Unit, University Involving the of Arizona Sea Otter

A cooipariro~~of westen! Ale~rtia?~Isla,~ds with and zuitf~out intertidal cow?nu,iity. The remoual of sea otters causes sea otters shows this species to be important to the increased herbiuory anrl ultimately results in destructio?~of organizntion of the earsb bore conltnu~tifies.Sea otters n~acropkyteassociations. Our obseruntions suggest that a control herbiuorous invertebrate pop~rlationsand indirectly dense sea otter population illdirectly supports fa~cnaasso- affect waue exposure and the compositiolt of the rocky ckted with mncrophyteprin~aryproductiuity.

One of the most striking features of the nearshore (1) the intertidal area, or that part of the shore community at Amchitka Island is the dense popu- alter~latelyflooded by water and exposed to air by lation (approximately 6400 animals) of sea otters, tlie rhythmic movements of the tides; and (2) the E~zlzyclrn lirtris (Estes and Smith, 1973; Estes, subtidal area, or that area from lowest loxver lo~v Chap. 21, this volunie). The estimated density of 17 water to tlie 55-m (180-ft) depth contour. sea otters per square kilotlleter (59 animals per squae nautical mile) of feeding habitat, waters 55 FACTORS CONTROLLING THE STRUCTURE 111 (180 ft) or less in depth (Kenyon, 1969), is one OF MARINE COMh,IUNITIES of the highest concentratio~lsof this predator ever Island Biogeography. To properly understand recorded (Estes and Paln~isano, 1974; Estes, the marine communit)~of Amchitka, me must Chap. 21, this volume; Lolvry and Pearsc, 1973; realize that the study of island com~nunities Iextinctions caused by grass beds as a result of overgrazing by dense stochastic events (ic., scverc \\,cather and food populations of sea urchins has been observed over a shortage) and by coactive forces (competition and wide geographical range. Hi~nmclmanand Steele predation). Extensive predation by dense popula- (1971) believe that sea urchins are responsible for tions of sea otters nlay lead to local extinction of the scarcity of uoncorallinc algae in many coastal prey species and associated co~nmunitics.Competi- areas of Ne~vfoundland.Icain and Jones (1965) tion from dense populations of sea otters may stated that grazing by sea urchins determines the cause other species to become locally extinct. vertical distribution of algae on the Isle of h,lan. \\'here sca ottcrs themselves become locally ex- INear Islands, organisms and com- :ad (1963) found that the distribution of dense sea Ecological I?~tevactious111uolub1g the Sen Otter 529 urchin populations and abundant algal gro\vtl~Tsere lease from grazing pressure pertnits an increase in almost mutually exclusive in the sublittoral area of the size of nearshore and interticla1 kelp beds and the northeastern Atlantic Ocean. Dayton, Rosen- associated connnunities (Jones and ICain, 1967; thal, and A,lahan (1973) reported a reduced macro- Paine and Vadas, 1969). Thus we hypothesize that phyte association due to sea urchin grazing in the sea otters are responsible for the extensive subtidal Chilean archipelago. The reduction of benthic and intertidal kelp beds and for the lack of dense macrophyte populations by sea urchin grazing populations of certain benthic in\~crtebratesin the along the coast of California has been described by subtidal and intertidal con~n~unitiesat Amchitka. North (1965) and in the Gulf of Mexico by Camp, A corollary to this hypothesis is that a small Cobb, and Van Breedveld (1973). The removal of population or the absence of sea otters enables the sea urchins by experimental tnanipulation (Paine sea urchin population to increase, which in turn and Vadas, 1969; lShemya for 1 \\reek each in September 1971 and mann, and Estes, 1974; h,lorrison, Rosenmann, and July 1972, at Attu for 4 days in July 1972, and at Estzs, Chap.23, this volume). In vie\\, of the Adak for 9 days in September 1972. An estimate preceding discussion and of the abiundance of sea of sea otter population at each island studied is otters at Amchitka and collsidering that the aver- given in Table 2. All these islands except Adak age sea otter weighs about 23 kg (50 lb) (I

Barnacle and ~n~tsselpopula- Sea urchin Sea otter Benthic algal tions 03% inter- Island population Size Pop~~lation beds tidal bench

Anicllitka Dense Small Sparse Extensive and little Sparse grazed Adak Dense Small Sparse Extensive and little Sparse g~azed Attu Sparse Large Dense Sparse and heavily Dense grazed Shemya None Large Dense Sparse and heavily Dense crazed

BERING SEA

Fig. 1-hlap of western Aleutian Islands shoxving the Near and Rat Island groups.

The landward edge of the bench is almost always than are bays, harbors, and areas tvith a wide adjacent to cobble or sand beaches, but at times it bench. Offshore kelp beds also protect the shore- terminates against rock faces. The bench is not line from wave action, as Moore (1958), Lewis continuous aroui~dthe island but is interrupted by (1964), and MacGinitie and I\IacGiiiitie (1968) cobble or sand beaches. The elevation of the bench have observed elsewhere. These more-protected is approximately incan sea level (Powers, Coats, areas have more silt deposition than do areas and Nelson, 19GO). The bench is best developed exposed to waves. In general, the intertidal bench along the Pacific Ocean side of the island where it on the Bering Sea is narrower (Fig. 4) and tltus is 100 in (330 ft) wide in places (Fig. 3). The more exposed than the intertidal bench on the bench is flat, porous, and soft enough in most Pacific Ocean, even though the Pacific Ocean is places (except at headlands) to fasten nails with a more turbulent than the Bering Sea (Gard, Chap. 4, hammer and even to be pulled apart by hand. this volume). Channels, tide pools, boulders, and sea stacks Tidal range is approximately +2.3 to -0.7 in (Lebednik and Palmisano, Chap. 17, this volume) (+7.5 to -2.3 ft). Atean tide level is approximately are not abundant but do occur. +0.5 m (+1.6 ft). Spring tides are diurilal (one high Headlands and areas with narrolv or no inter- and one lolv per day). Neap tides are inixetl tidal benches are more exposed to wave action semidiurnal (two highs of unequal levels and two ..I 'IU!O~ 30[!.1!>1 'PU~ISI 'am3 1!.1L3 'keg pus 'ur?~ll~c[dopa~'sn3n.g) uo!Sa.l le.rol)gp!ru (2) atenb~je pue ap!s ma30 3!j!3e~ alp uo a~o3yana 'a8rr!.r~ le~o,)!leldns (I) :SUO!~~.Iaaq 0111: pap!n!p pue 'ILI!~~a8ue~ aU!x *'i(eg sn!reqeIQ 1s apew a.1a.u Iraq sty earl? Isp!l.ta)n! ayA '.ra]upl pue rrwnlnu II! saipuls an!sua)rrI .pue[s! arp 30 uo!,.rod lsesqlnos %qS!u 1e pue mwurns pue Supds u! 1q8!1dsp Krr!.rup ay, uo pa13npoo3 a.ral~sa!pli~s .$no JO JSOJ,~ .1n3~osap!$ hi07 .(Aep .iad spat {enbatrn jo s,\ro~

oo!lsnmsqo lEUOSlad B ax 8CLV 6961 'uoduax L 18~1 9EET 6961 '~'J'(~JX 21 &ID& 0922 6961 'uo6uax 6 1622 81LI 6961 'uo.(xrax $ L66 GGE 6961 'uoLuax I > 08 6961 'uoduax o o 6961 'UOLU~X o o 6961 'UOLUJW 0 0 (wz!qeq lo z1~[9LSZ) gepv

~~61'41!"'~PU" sax" 11 ZFP9 ZPOP FLGI 'ql!WS pUL'SalS3 51 Lt.54 TPZE ELGI '~II!"sPU~ salsa $1 9PLP ELLZ 6~61'%I~!WSPUE salsa I I GZOP PPFZ ELGI 'ql!ms pue salsa 01 OPGE ZOEZ 6961 '~06u~x f OZPI PPI I 6961 'uodcra~ 9 0802 0951 6961 'uoL1Ti P 6PPI LSOI 6961 "JO~~UJX ZT 95% LIP£ 6961 'uoduax 6 18EF 45EI 6961 'uoAuax 6 ZOFE IZEI 6961 '"0Luax 8 OOIE PIS (i"i!q*l~JO cvILL) =)~?!YJU~Y

uo!,ehiarqo ,soor>a,j 0 0 uo!leUasqo lsuor~a,j 0 0 6961 '~~Auax GN UN 6961 'u~.(uax o 0 6961 'uoh~lW 0 0 rhataqs 532 Pabnisano and Estes

17SDOOE Bird 6ird cape Rock rg

BERING SEA

0- 5 km

PACIFIC OCEAN

St. hlakariur Point

Fig. 2-Map of Amchitka Island, Alaska.

Fig. 3-\Vide intertidal rock bench at low tide south of Rifle Range Point on the Pacific Ocean coast of Amchitka island, Alaska (Aug. 7, 1971). Ecological I~zteractio~lsI~tuoluiicg tlce Sea Otter 533

Pig. 4-Narrow intertidal rock bench at low tide south of Square Bay on the Bering Sea coast of Amchitka Island, Alaska (Aug. 22,1971).

Crown Reefer Point

BERING SEA

0 1 2 3 k",

Kirilof point lvakin Paint

PACIFIC OCEAN

Fig. 5-Map of southeastern portion of Amchitka Island, Alaska, showing study areas referred to in this chapter. SHEMYA ISLAND

Fig. 6-Map of Shemya Island, Alaska, showing study areas referred to in this chapter. factors at these islands because they were visited intertidal area contained rock benches and cobble only once or t\vice for short periods of time. beaches similar to those at Amchitka. The well- Ho\vever, conditions observed at these islands developed bench at Massacre Bay was the primary (climate, water temperature, salinity, and seasonal study area (some gra\!cl areas occurred on this development of invertebrate atld algal communi- bench). Murder Point and Chichagof Harbor were ties) were comparable to those at Amchitka during also studied. similar times of the year. Adak is similar to Attu in size, and its coastline Shcmya is significantly smaller than the other is also generally inaccessible because of the lack of three islands. A road adjacent to the shore circles roads. Its many bays and harbors are protected the island and provides complete access to the from the open sea. Only a small portion of the slloreline. Tile Bering Sea side has a continuous northeastern corner of the island was visited (Fig. sand beach. The Pacific Ocean side has a rocky 8). Here the intertidal area is predominantly steep shoreline composed mostly of boulders and cobble. cobble beaches. There are also sand beaches and areas of large boulders. Adak does not have There are well-developed benches 011 the northeast portion of the island at Bomb Point and Urchin well-developed intertidal benches, although some Point (these names ~vereassigned by the authors), narrow benches [lo to 15 m (33 to 50 ft) wide] at \vhich were tile primary study areas (Fig. 6). The Kuluk Bay and at Cape Adagdak were studied. A be~lchesare similar to those at Amchitka except large number of boulders and small sea stacks made that they have more and larger tide pools. the bench surface irregular. Because of these features and the vertical nralls adjacent to the Attu is larger than Amchitka. The intertidal bench, Adak has more vertical space in the area is generally inaccessible because of the lack of intertidal area than the other islands studied. Zeto roads, and thus only a small portion of the eastern Point, Finger Bay, the breakwater at Sweeper end of the island was visited (Fig. 7). Here the Cove, and Clam Lagoon were also studied. Ecological I~rteractio~~sIi~uolui?ig the Sen Otter 535

172'40' E

BERING SEA

ATTU ISLAND

Fig. 7-hlap of Attu Island, Alaska, showing study areas referred to in this chapter.

METHODS nearshore communities at islands where sea otter densities have recently undergone large changes Rationale (Amchitka and Adak); and (3) observing, manipu- Islands often may seive as natural experiments lating, and experimenting with orgallisms (primar- for testing biological hypotheses (Carlquist, 1965; ily invertebrates and kelp) that we suspected of A,IacArthur and Wilson, 1967; MacArthur, 1972). being affected by sea otter predation. The advantage of tltis method is that certain Sea otters have been near equilibrium density variations among islands may be considered experi- in the Rat Island group of the Aleutian archipelago mental treatments, and thus results can be ob- (Fig. 1) for about the last 20 to 30 years, follo\ving tained immediately. The disadvantage is that geo- almost complete annihilation by Russian and graphically isolated islands are allopatric; thus American fur traders during the eighteenth and ul~controlledextraneous variation may creep into nineteenth centuries (Kenyon, 1969; Estes, itlterpretations of results more than in well- Chap. 21, this volume). Conversely, the sea otter cotltrolled experiments. population bas not reestablished itself in the Near The direct manipulation of sea otter popnla- Islands (Fig. 1). The once abundant sea otter tion densities in the western Aleutian Islands (by population there also was extirpated by over- introduciilg or removing a~limals) would have exploitation, but immigrants from the densely revealed their role in structuriilg nearshore com- populated Rat Islands were uilable to reach the munities. Such a study was not possible owing to Near Islands, ~vl~ichare approximately 330 km resource and time limitations. We attempted, how- (200 miles) to the west northwest of Amchitka and ever, to determine the role of the sea otter are separated from the Rat Islands by \vide deep indirectly by (1) comparing nearsbore communities oceanic passes. Since 1959 there have been scat- of nearby islands sinlilar in most respects except tered reports of sea otters in the Near Islands for sea otter densities; (2) comparing in time (Kenyon, 1969), although no major population Fig. 8-hlap of hdak Island, Alaska, showing study areas referred to in this chapter.

reestablishtllent has occurred. Thus the presence inside Y16-, jh-, or 1.111~ sample frames and and absence of sea otters in tlie Rat and Near returning them to our laboratory for measurements Islands, respectively, senre as our natural experi- or by measuring these parameters in situ. Sample ment. After careful consideration and consultation, plots were selected by tossing a frame arbitrarily we believe that other factors, such as climate and into selected algal associations and by placing a tides, ~vhichare capable of dynamically influencing frame at random intervals along either side of a nearshore marine community structure, are suf- rando~ntransect which was perpendicular to shore ficiently si~iularbetween tlie two island groups that and which transversed the entire association of they need not be seriously considered, and thus we intertidal algae (O'Clair and Chew, 1971). Food should be able to determine tlie actual effect of sea habits of intertidal invertebrates were determined otter pretlation on the nearshore community sim- by observing fecdiug animals or by examining gut ply by comparing the Rat and Near Islands. contents of collected animals. We collected addi- tional information during "beach walks" along Intertidal Community many kilometers of shoreland during daylight low tides. Most of our observations were restricted to the midlittoral area [+0.8 to 0 ni (+2.6 to 0 ft)] and Grazing Studies. To test the effect of sea sublittoral fringe [0 to 0.7 m (0 to 2.3ft)] urchin grazing on intertidal algae at Amchitka (i.e., of rock bc~lchesbecanse these areas provided the to si~llulate lo\\' sea otter densities), we placed greatest biotic diversity (\\'einmann, 1969; O'Clair, twenty 19-mm-diameter sea urchins (Strotl- Chap. 18, this volume). g~docentrotuspolyaca~~thnts) from the subtidal area We determined the size and density of plants at Constantine Harbor (Fig. 5) in each of four and aniinals either by re~liovingall nlaterials from 0 .5-m2 plastic (Vexar) mesh cages (mesh, Ecological I~zteractiortsI~zuolui~tg the Sea Otter 537 3 by 4 mm) with alumi~lum mesh tops (mesh, fresh seawater, seawater and four isopods (Idotea 13 by 13 mm). T\vo cages were permanently se- wos~zesse~ukii),or sealvater and four amphipods cured (nailed) in areas xvith a canopy of either (Parallorchestes ocltote~~sis).The 12 cups \\.ere 100% La~niitarialo~tgipes or Hedo/~I~yIl~r~~zsessile in kept otctdoors and \\,ere observed se\rcral times a iVIakarius Bay (Fig. 5) in September 1971. A day. The water was changed after each observation, control cage ~vithoutsea urchins was also provided and the condition of the algae was recorded. This for each algal type, and the percentage of algal experiment proceecied for 2 days and was then cover in all cages was recorded bimonthly from repeated with another goup of similar plants and December 1971 to June 1972. animals. To determine the develop~iientof marine algae at Shemya in the absence of grazing by sea urchins Growth and Feeding Studies. The small diam- (i, to simulate high sea otter densities), in eter of sea urchins in the intertidal area at September 1971 we removed all sea urchins from Amchitka could be due to exte~lsivepredation or two tide pools (6 by 3.3 by 1.5 m and 3.5 by 1 by to genetic limitation. To determine \vl~ichfactor is 0.8 m) that contained only coralliile algae and most important, in May 1971 we transported five l'ltalassioplzyllzt~~zclatlzrzts. T~voadjacent tide pools sea urchins (?,15.2 mm in diameter) to the marine rvith sea urchins were selected as controls. In July laboratories at Friday Harbor, Washington, and 1973 the species and percentage of cover of algae placed them in a permanently filled continuous- and the number of sea urchins were recorded for flow seawater table. Water temperature ranged all four tide pools. from 4 to 15'C during the experimental period. We studied the effect of limpets (Collisella The water table was periodically supplied with pelta) grazing on intertidal algae by using fifteen brown algae, mostly ~Vereocystis ltcetkea~ta and 133-cm2 bottomless plastic dog dishes (Dayton, Costaria costata. The diameters of the sea urchins 1970) ~vithaluminum mesh tops as inclusio~land \irere measured bimonthly. This growth experiment exclusion cages. Three cages each were secured in was terminated in June 1972 after all five sea 1-m2 areas that had recently been denuded of L. urchins had died, apparently because of lo\\, lo~zgipes,11. sessile, and I~alosaccio~~glalartdiforlar~ze at dissolved oxygen and lo\\' salinity. h,Iakarius Bay and of Coralli~tauar~couuerie~tsis and Intertidal sea stars and snails prey on barnacles /llaria crisfm at Duck Cove (Fig. 5) in July 1971. and mussels (Connell, 1972; Spight, 1972). The TIVO cages in each area contained three small size of these predators on the benches at 30-mm-long limpets, and the third cage sersed as a Amchitka may be due to low densities of barnacles control. T\vo weeks before the experiment, the and mussels or to genetic limitations. To test this, study areas \\rere cleared of plants and animals with in A4ay 1971 \ire transported 20 sea stars [Lep- a shovel, putty knife, and \\'ire brush ,uld by bunling tasterias alaske~tsis(%, 14.9-mm arm radius; 3, 1.5 g twice with a 1 : 1 mixture of fuel oil and unleaded live \\'eight)] to Friday Harbor. Their water table gas. Species composition and the percentage of was periodically supplied with opened live clams cover of algae in all dishes were recorded bi- (Protothaca stanzittea), mussels (~Vlytilusedulis) (4 monthly until August 1973. to 8 cm long), and live barnacles (Bala~zusglalz- \\re selected four areas (50 to 150 cm2) of bare dtcla). The weight and arm radius were determined rock in the ~nidlittoralregion at hflakarius Bay, bimonthly until all 20 sea stars died (June 1972). each of nrhich contained one or two limpets (20 to Twenty predacious snails (Tltnk li~larza; %, 18 mm 40 mm long). Each area \\.as marked wit11 a nail total length) were tra~isportedfrom the benches at aid rope. Begiiltii~lg in July 1971, limpets were Amchitka to the Department of Zoology, Univer- removed and subsequently kept from two areas, sity of Washington, in July 1971 \\,here they were and the other two areas served as controls. Algal placed in liter beakers containing seawater and cover was recorded bimonthly in all four areas several live barnacles (B, glartdztla) at a controlled until August 1973. temperature of 10°C. This experiment lvas termi- Because the dense staudi~igcrop of algae at nated after 1 month (R. T. Paine, personal com- Amchitka is not noticeably grazed by epibenthic munication). herbivores, \vc wondered if this abundant resource h~lussel (1V1. ecltrlis) densities may be lo\\' at is used by other herbi\~oresafter it is dislodged Amchitka because of predation or inadequate from the substrate. To determine if drift kelp food. To determine this, in April 1972 we trans- \vould be catell by small intertidal crustaceans ported about 50 subtidal mussels (I\.[. edulis; %, (amphipods and isopods), we collected and cut 17 mm long) from Puget Sound to Amchitka. into I-cm2 pieces L. lo~~gipes,II. sessile, A. cris/~n, Twelve mussels were placed on each of threc and '4. fistttlosa. Sis pieces of each alga were 1-cm-diameter polyethylene ropes, and four mus- placed in 240-ml cups containing either filtered sels were placed in each of three 700-ml perforated 538 Palntisarao ar~dEstes polyurethane bottles. The ropes and bottles were the subtidal region in Puget Sound in secured dog suspended from the dock in Constantine Harbor dishes in the Laminnria and Halosaccion zones at \vith the mussels just below the xater surface at h,Iakarius Bay to test the effects of siltation on the 0.7-m (-2.3-ft) tidal level. The mussels were mussel survival. The mussels appeared to be at- measured every four inonths until August 1973. tached to the substratum wlvllen the dishes werc Selective grazing by sea urchins on a competi- removed on Dec. 19, 1972. The condition of the tively dominant intertidal alga could affect com- mussels was observed in April and August 1973. munity composition by reducing competition for Settling Studies. The absence of appreciable space ai~dthus permitting more algal species to barnacle (Bala~zusspp.) and mussel (IM.edulis) sets coexist. To determine if sea urchins at Amchitka on the benches at Amchitka may be due to the feed selectively, at Constantine Harbor we con- absence of larvae or to the absence of available ducted food-preferencc tests in a 1.5-m(5-ft)- substrate. To determine if barnacle or mussel laivae diameter by 30-cm(l2-in.)-deep plastic wading \vould set if space were provided, in December pool supplied with constantly flowing fresh sea- 1971 we nailed one each of five 1-m-long stainless- water. A weighted lvooden "Y" maze divided the steel wires, each holding five 15-cm(6-in.)-long pool into three equal sections. Five 19-mm-diam- oyster shell halves (Crassostrea gigas) from Puget eter sea urchins from the intertidal area at Am- Sound, in the same five 1-m2 denuded areas used cllitka were placed in one branch of the maze; T. in the limpet grazing studies. The shells as well as clatl~r~is(a subtidal kelp) was placed in another; the denuded rock areas were checked bimonthly and either I,. lortgipes, A. crispa, or H. sessile (all until August 1973 for settling animals. lo\ver intertidal kelp) was placed in the third. Tltnlassiopltyllto,~clatltrtrs was used in all tests, and Determination of Wave Exposure. Because ure one of the other three kelps \\,as used during two believe that the degree of exposure affects com- consecutive days. Fresh kelp was provided daily. munity structure, we wanted an independent quan- At 0800 and 2000 each day, the location and titative measure that would enable us to compare activity of each sea urchin was recorded. The sea the relative exposure of our study locations. urchins were replaced to their original location \Veinman11 (1968) observed that the width of L. (twice daily), and the position of the t~vokelp lorzgipes blades reflected relative exposure (blades species \\.as exchanged. Each test ran for six were wide in protected areas and narrow in ex- consecutive days and then was repeated with five posed areas). Therefore we used the blade width of different sea urchins. T\vo similar tests were carried this kelp as an indicator of exposure. out at Amchitka with two groups of 75-mm-diam- eter sea urchins from the intertidal region at Shemya. Subtidal communities were observed at Attu and Shemya Islands in the Near Island goup, at Siltation Studies. Silt accumulates in pro- Amchitka Island in the Rat Island group, and at tected areas and is capable of smothering barnacles Adak Island in the Andreanof Island group. Quali- and mussels. We conducted experiments at Am- tative observations over wider areas in these islands chitka to determine the intertidal areas of silt indicate that our study areas are representative of accumulation and the effects of siltation on mussel respective island groups. survival. Time limitations and logistic problems allowed To determine areas of silt accumulation in the only a single dive* on June 21, 1972, at Murder intertidal region, we nailed t~voplastic dishpans Point, Attu Island (Fig. 7). \Ve also made nine dives [lo cn~(4 in.) high and 560 cm2 in area] in each during August and September 1971 and June 1972 of the same five 1-,n2 delluded areas used in the at three different locations along the coast at limpet grazing experiments. These pans were Shemya Island. All these study areas werc chosen placed over areas that contained 100% cover of primarily because of accessibility in view of the young algae (either LIlva Iactticn, fiicus distichlis, limited time available for work in the Near Islands. H. gland$or~r~e,H. sessile, or A. cris[~a) in July We studied four subtidal areas at Amchitka 1972. For water movement and light penetration, between March 1971 and September 1972. Forty- the pan tops were removed and four equally spaced six dives were made during the observation 4- by 2-cm openings were placed 2 cm above the period.* All study areas \\,ere selected from thc bottom. The areas covered by the pans were ex- Bering Sea side of Anlchitka Island for the amined periodically until the pans \\.ere removed in follotvi~lgreasons: August 1973. On Dec. 15, 1972, we placed two groups of 15 mussels (dl. edi~lis,20 to 60 mm long) each from *Refers to number of dives made by a "dive team." Ecological Irrteractio~zsI?tvolviirg the Sea Otter 539

1. The Bering Sea was logistically more conve- biomass to the epibenthic macrophyte association nient because support facilities were located at in the subtidal area of the western Aleutian Islands. Constantine I-Iarbor (Fig. 5). 2. Weather conditions are generally more favor- We estimated the coverage of each species or able on the Bering Sea than on the Pacific Ocean. species goup at 3-m (10-ft) depth inte~vals,begin- Prevailing tvesterly winds cause the Pacific Ocean ning at a depth of 3 m (10 ft) and starting from a to be rough, whereas the Bering Sea is compara- point arbitrarily selected along the shore. Vcgeta- tively more calm. tion coverage was estimated to a depth of about 3. The underwater slope is steeper along the 27 m (90 ft) (when possible), at which point the Bering Sea coast than it is along the Pacific Ocean dive~sswam palallel to the shore for a sufficient coast of Amchitka (see the Geological Su~veymap distance to avoid ove~lapwith the area observed in the pocket at the back of this volume); thus a. during descent. They then made similar coverage diver can observe a relatively geatcr depth range estimatcs during ascent. without excessive swimming or boat work. \Vhen depths to coverage estimates were as- 4. Several dives made in the Pacific Ocean in signed, tidal fluctuatio~ls were not considered. the vicinity of Makarius Bay (Fig. 5) confirm that Water depth was determinecl by a standard oil- the community structure in the Pacific Ocean is filled depth gage, which is probably not accurate to qualitatively similar to that in the Bering Sea. more than a few meters. Tidal fluctuations at Amchitka ale not great. The tidal range is 1.2 m Bat Island and ICirilof Rocks (Fig. 5) are (4 ft) at Sweeper Cove oil Adak Island (Coast and heavily exposed duling storms off the Bering Sea. Geodetic Survey, 1968), and the tidal range at Constantine Point is protected from the direct Amchitka Island is close to this value. force of storm surf and is apparently only moder- Data on vegetation covcrage were t~ansformed ately exposed. Kirilof Point is well protected from to Arcsin square-root percent values to approxi- heavy seas by Icirilof Rocks and the confines of mate a normal distribution (Ostle, 1963). The Constantine Harbor. Our observations of the rocky mean, standard deviation, aud 95% confidence subtidal commuuities at Amchitka therefore are intelval were calculated within the transformation Inore extensive and from a wider range of exposure for each species or species group at each depth for of subtidal conln~unitiesthan at the Near Islands. data collected at Icirilof Point. The calculatio~~s We observed depths of subtidal epibenthic then were retransformed to percentages from communities to 25 m (80 ft) in the Near Islands Arcsin square-root percent values. Only the means and to about 40 m (130 ft) at Amchitka.* were determined for data collected at the remain- The percentage of cover and vertical distri- ing three study areas because of small sample sizes bution of the following epibenthic algal groups and at each depth. species were documented. 1. Lamittarin lo~lgi/)es. Sea urchin (S. polyacntithzrs) densities and size 2. The digitate Lnmi~zuria life form. This in- class distributions were estimated according to the cludes L. groe~rlattdica, L, de~ttigern, and L. technique desc~ibcdby Barr (1971). Briefly this yezoeirsis. involves random placement of a 0.25-in2 quadrat 3. Agnrrznz cri6rosttrt~. on the ocean floor and counting the number of sea 4. 'I%alnssiophyllto~rclatlrrns. urchins within a quadrat. The animals were taken 5. Des~narestinsp. to the laboratory where their test diameters were 6. Foliose Rhodophyta. measured. The data used to estimate size class 7. Total ~nacrophytic vegetation-The pre- distributions and the maximum density of sea dominantly subtidal (digitate) species of Latttittaria urchins at Amchitka arc from Barr (1971). Addi- were grouped because they are not individually tionally, we measured sea urchin densities at 3-~n recognizable except on close examination. Al- depth intelvals between the depths of 3 and 20 m though I,. yeroetlsis is identifiable on close in situ (10 ant1 66 ft) at each of the four A~nchitkastudy examination, L. groetrla~zdica and L. dentigeru areas and at depths of 3, 10, and 25 m (10, 33, and require stipe section for species identification 82 ft) at Slle~nyaIsland. (Druehl, 1968). Otherwise e;rch of the preceding Sea ttrchin wet weight (blotted dry) and test species or groups of species is easily recognized. diameter were measured from 110 individuals Together these species contribute tnost of thc througl~out the naturally occu~ringsize spectrum at Shemya and Attu. We fit the linear model yi = p + 0 (xij) + eij to the data by the mctllod of *Altbough at Amchitka we dived to 40 m (130 ft) on least squares, where y is the log sea urchin mass in one occasion, most work rvas conducted between the grams, x is the log sea urchin diameter in milli- surface and 27 m (90 ft). meters, p is -7.875, ,tnd 0 is 2.992. A corrciation 540 Palnlisnlzo adEstes

LOWherbivore Extensive rblzb;$tim of Reduction densities (sea little grazed in tlie size urchinr) marine algal invertebrate$ of piodatorr beds

Moderate density of Common Eiderr Reduction of Of snail il~ellr High density High of horbivoraur predacious fish cmstaceani famllhipods. Roducfion of iropodr. and hermit crab niyridr) populalionn

Abundance of harbor seals

SEA OTTER POPULATION

grazed

Harbor seals reduced

Fig. 9-Diagram of the interactions within nearshore communities with and without sea otter populations. See Discussion for complete details. coefficieilt of 0.992 indicates a very good linear fit principal factor responsible for altering the com- of the log-log transformation. Parametric inte~val position of these marine communities (Fig. 9). estimates and tests of l~ypothesisarc given in the Furthermore, the coiltents of Aleut kitchen mid- appendix. The variaitce of (e) is quite small, and dens at Amchitka indicate that sea urchins were therefore simple least-squares estimators of fl and /3 larger centuries ago than today (Desautels et a]., are rather precise. \\'ith the use of this model as our 1970) tvhen, Kenyon (1969) speculates, there were cstimator, biomass was superimposed on each size fc\ver sea otters because of intensive native hunting class distribution of sea urchins. pressure. That Desautels also found sea otter bones in these middens supports this speculation. DESCRIPTION OF COhlhJUNITIES Intertidal Corn~nu~iities Community Differences Amchitka. A complete inat of virtually un- The most prominent biological features of tlte grazed algae occupies intertidal benches, channels, marine commiuitity studicd at Amchitka are the and tide pools at Amchitka (Fig. 10). La~nimaria extensive little grazed subtidal and intertidal beds lo~~gq~escoinpletely covers the lowest portion of of ~nari~tealgae and the sparse population of the intertidal region (sublittoral fringe). Hedo- certain benthic in\~crtebrates(Table 1).In contrast, ~ILYEIZL~Zsessile predontiltates in the lower mid- the prominent biological features of the marine littoral region in protected areas and on the communities studied at Attu and Sltemya are the landward side of wide benches. Airria crispcc sparse heavily grazed offshore aitd intertidal beds occirpies this portion of the lower midlittoral of marine algae and the dense population of large region in exposed areas and 011 narrow benches. benthic invertebrates (Table 1). The marked de- Halosaccio~zglandifor111e, Iridaen conzztcopiae, and cline in the population and size of sea urchins that Fnclcs disticlbzts cover the higher zones of the occurred concurrently \sit11 an increase in the ~nidlittoral region (see Lebednik and Palmisano, population of sen otters at Adak and timchitka Chap. 17, this volluine). during the previous 15 years (Kenyon, 1969) There is little evidence of grazing on larger suggests that sea otter prcdation has been the plants in the kelp comlnunities at Amchitka. Ecological Itcteractions I~evolui?tgthe Sea Otter 541

Fig. 10-Continuous mat of virtually ungrazed algae on intertidal rock benches at Amchitka Island, Alaska, at lo\\, tide (Aug. 7, 1971).

Grazed kelp is recognized by its serrated edges. brates is given by O'Clair in Chap. 18, this volume). This characteristic mark made by sea urchins and However, sessile filter-feeding invertebrates (barna- other benthic herbivores, such as chitons, was cles and mussels) are Inore abutldant in areas noticed in less than 1% of the attached kelp exposed to a higlter degree of wave action, e.g., examined. Over 39% of young kelp [under 20 cm Banjo Point (Fig. 5), than in protected locations, (8 in.) lo~tg]were grazed (Table 3); ho\\~e.e\.er,the e.g., klakarius Bay (Table 4). size of the serrations and the abundance of small No sea urchins mere observed grazing intact ainphipods and isopods suggest that these small yegetation on the bench, in cha~lnels(Fig. ll), or crustaceans may have grazed the smaller kelps. In tide pools. Ho\\.ever, sea urchins did graze algal I~lvcrtebratcs are itlconspicuous on the intcr- debris and detritus in these locations. Sea urchin tidal benches at Amchitka (Fig. 10). Barnacles guts contained ~nacrophyticalgae and diatoms. (Bala?tus glat~dula and B. carioslrs) and mussels (~\'lytilztsedulis) (sessile filter-feeding invertebrates) Adak. Altllough the algal community of Adak and sea urchins (S. pol)~aca~~tltzts),chitons covers the intertidal rock bench and is ungrazed, it (Katliarit~attozicata), and limpets (C. peltn) (motile is not as extensive as the one at Amcllitka. The herbivorous invertebrates) arc small and scarce majority of the community is composed of L. (Tables 4 md 5). Some invertebrates (barnacles, longipes and A. crispa. IIedoplcyllt~~~tsessile is not limpets, and snails) are covered by silt or over- abundant. Tl~alassioplzyllut~cclathrvs and Cyma- grown by algae in areas protected from \vave there triplicata, normally subtidal kelp at Am- action. Even the most abundant invertebrates chitka, are present in the intertidal area at ICuluk [polychaetes, certain crustaceans (amphipods and Bay. isopods), sea anemones, sea cucumbers, and mol- At Adak, as well as at Amchitka, the invcrte- lusks (clams and snails)] are i~lcollspicuousbecause brates on the rock bench are small, scarce, aud of their small size and occurrence under algal mats, incotlspicuous. There are, however, extensive mus- in algal holdfasts, under rocks, and on the bottoms sel (I\{. ed111is) beds and dense populations of of pools and chatn~els.Only small littorine s~lailsin barnacles (B. gla~cd~tlaand B. cariosns) on vertical the mid and upper intertidal areas are conspicuous. .\\'ails, on sea stacks, and on the sides of large Therefore in most areas intertidal algae appeared to boulders. Dense populations of large predacious be the sole inhabitants of the intertidal bench (a suails (Tlzais li~,~n)were observed feeding in these lllore complete description of intertidal inverte- areas (Table 6). The breakwater protecting 542 Palt,lisnno and Estes

Table 3-Percentage of Intertidal Raadom 111dividualKelp Plants Grazed and Random 0.25-m2 Quadrats Containing Grased Kelp, Amcbitka and Shernya Islands, Alaska, 1971-1973

Amcbitka Shemya Sample Tide Sample Tide Item size BEIIC~I pools Channels size Bench pools Channels

Random plants Laini~iarinlongipes >200 mm long 16,000 <1 < 1 < 1 1600 ND* loot 100'1 Alnrio crispa >200 mm long 3,000 200 mm long 4,200 < 1 < 1

Table 4-&lea11 Numbers Per Square Meter of Selected I~~tcrtidalInvertebratcs at Amcliitka and Shemya Islands, Alaska, 1968-1973*

Shemya$ Amchitka Tidc- hlakarius Balijo All vool Chnnnel Invertebrate Bay ~oik areast Ber~cli walls bottom

Sea urchin (Shongylocentrotl~s polyacanthus) Limpet (Colliselln peltn) Chiton (Knthnri~iatunicntn) Barnacles (Bolanus glnndaln) (B. cnriosus) hlussel (~llytilusedvlis) Snail (Tl~aislimn) Hermit crab (Pagvrirs hirsutitrsci~lvs)

*Preliminary data from transect-line ?16-mZ plots [Amchitka, N = 32; Shemya, N = 23 (O'Clair, in prepxation)] a~dfrom randomly selected '/4-m2 plots [Amchitka, N = 171; Shemya, N = 9 (see Estes and Palmisano, 1974)l. tRifle Range Point, Duck Cove, Makarius Bay, and Banjo Point. f Urcbin Point.

Sweeper Cove contained a dense population of sea Shelnya. Algae are heavily grazed in the lower urchins, (S. polyncantlz~rs)among the large boul- intertidal area (Fig. 12), in channels (Fig. 13), and ders (Table 5). This is the only area along the shore in tide pools (Fig. 14) at Shemya, atlcl there is a ~vhere live sea urchins were seen. Hermit crabs definite browse line at the lolv water level (Fig. (Pagurus sp.) also occur at Adak. klussels, clams 12). Channels and tide pools beneath the low (Protothaca sp.), and cockles (Clinocardit~?~t,zut- water level are devoid of most species of algae. tnlli) were obse~~,edin Clam Lagoon. Law~inarinlo~~gi/)es does not form a complete mat Ecologicnl I?rternctio~tsI71~0Ivi?rg tlre Sen Otter 543

Table 5-Sizes of Selected Invcrtebratc Herbivores from the IVestern Aleutian Islands* - Area x R N SD CI

Sca urchins Amchitka Intertidal bench 18.9 4 to 45t 224 5.7 f11.2 Subtidal area Constantine Dock* 30.9 17 to 56 345 6.7 f13.1 10 to 20 m$ 16.5 3 to 34 872 5.8 f11.4 45 to 80 mn 41.1 26 to 80 207 10.4 k20.4 Shemya Intertidal bench 30.4 4 to 74 289 17.9 t35.1 Hedophyllum zone 22.8 4 to 70 153 16.0 f31.4 Ln~ninariazone 37.6 10 to 74 136 16.8 f32.9 Tide-pool walls 52.2 25 to 70 130 7.7 fl5.1 Channel bottom 53.4 22 to 81 140 17.4 f34.1 Attu Intertidal bench 52.8 28 to 99 14 ND ND Channel bottom 87.5 78 to 108 13 ND ND Adak Breakwater** 28.4 10 to 50 121 8.6 f16.9 Growth experimenttt May 1971 15.2 12 to 18 5 2.4 f6.7 May 1972 36.2 34 to 38 5 2.0 f5.6 hlidden remains$ $ 69.7 21 to 106 1403 9.8 f19.2

Limpets Amchitka Intertidal bench 27.8 18 to 51 60 1.8 f3.6 Adak Intertidal bench 25.1 10 to 42 75 1.5 f2.9 Shemya Tide-pool walls ND 19 to 67 200 ND ND

Chitons Shemya Intertidal bench 67.2 23 to 88 77 18.1 f36.0

*%, average individual size in millimeters (test diameter for sea urchins, total length of longest axis for others); R, range of individual sizes in millimeters; N, number of animals measured; SD, standard deviation; CI, 95% confidence interval; ND, no data. tOnly six sea urchins were larger than 29 mm in diameter. tCollected beneath dock with scuba 15 to 6 m (16 to 20 ft\, deer]... $~arr,1971. llCollected 45.50. and 80 m (148. 164. and 262 ft). deep.. bv bottom dredae. **C~,llcctcdat luw tide betwccxt ll~crocks of the Stvc~.pcl.Cove bre.~kwste~.. ttA~nchitk:tintrrtidal rc:~t~rcl~itls feel ,at 1:riday Ililrbo~~I.abor:tfories for I year. SSDesautels et al., 1970.

in the sublittoral fringe. Instead there are areas of occupies areas of several square meters in this zone only L, longipes stipes and holdfasts or of bare (Fig. 15). All L. loltgipes, A. crisf~n,and H. sessile rock and areas containing T. clrthrus and L. overhanging channels are grazed. Eighty percent of groeirlnridica. (T. clnthrirs and L. groe~rln~idicc~occur all 0.25-111' L. lo~rgipes plots and 75% of all II. only subtidally or in tide pools on Amchitka.) sessile plots snmpled contained grazed plants Hedophyll~rnzsessile is heavily gazed in the lower (Table 3). The only plants not heavily grazed were midlittoral region and is absent or heavily grazed in coralline algae and T clatlbrus, which arc usually channels and tide pools. The upper midlittoral present in channels, tide pools, and the lower region and the landward end of the bench contain intertidal areas. H. glandiforme, I. comz~copine,andl? distichzrs and Dense populations of large sea urchins (S. are not heavily grazed. Tlzalnssioplbyllzrm clntlbrirs polyc~caibtl~r~s)(Pig. 16) and large chitons (K. 544 Pal~nisanoand Estes I I

Fig, 11-Ungrazed Laminaria longipes on wall and bottom of an intertidal channel at Amchitka island, Alaska, at lo\s tide (Aug. 7, 1971).

Table 6-Sizes of Intertidal Predacious Sea Stars and Snails in the \\'esten~Aleutian Islands*

Size IYeight - - Area x R N SD CI x R N SD CI

Amchitka 14.9 3 to 44 1144 6.0 f11.8 1.5 0.1 to 13.6 571 1.5 +2.9 Shemya 25.6 6 to 55 41 11.3 t22.7 5.3 0.1to25.8 33 4.2 f8.6 Growth exwerimentt

i'ltmis limn Amchitka Intertidal 20.9 10 to 35 98 8.5 f 16.9 hIakarius Bay 18.1 15 to 20 8 2.0 * 4.7 Duck Cove 17.2 10 to 27 57 6.9 f 13.8 Banjo Point 29.6 25 to 35 23 8.0 f16.5 Constantine Harbor 24.1 14 to 31 10 5.2 +11.8 Adak Kuluk Bay 22.6 11 to 40 330 8.0 f15.7 Attu Cbichagof Harbor 42.8 34 to 50 10 6.8 +15.4

*?, average individual size in millimeters (arm radius for sea stars, total length of shell for snails) or live weight in grams; R, range of individual sizes in millimeters or live weight in grams; N, number of animals measured;SD, standard deviation; CI, 95% confidence interval. tAmchitka iutertidal sea stars fed at Friday Harbor Laboratories for 1 year. Ecological Interactions IILVOIV~IE~the Sea Otter 545

Fig. 12-Heavily grazed algae on intertidal rock benches at Shemya Island, Alaska, at lo\\. tide. Note sea urchins in foreground and barnacles (white) in background (Sept. 2, 1971).

Fig. 13-Grazed Laminaria lo~igipeson wall and bottom of an intertidal channel at Shemya Island, i\laska, at low tide. Note density of sea urchins in channel (Sept. 2, 1971). Pig. 14-Tide-pool wall and bottom devoid of algae at Shemya Island, Alaska, at low tide. Note sea urchins beneath rrrater surface and limpets (white owls) on tide-pool wall (Sept. 2, 1971).

Fig. 15-Large area of T/~alassiophyll~cntclathras in Laminnria zone at Shemya Island, Alaska, at lo\%'tide (June29, 1972). Ecological Interactions Inuoluiitg the Sea Otter 547

Fig. 16-Dense concentrations of large (F, about 70 mm in diameter) sea urchins in lower intertidal zone at Shernya Island, Alaska, at low tide. Note large sea star (arm radius, 55 mm) to right of center (Sept. 2, 1971). tzcnicata) and extensive beds of n~ussels(IU. ednllis) analyses of over 100 sea urchins (range of 10 to and barnacles (B. gla.lnndula and B. ca~ios~cs) 75 nnn in diameter) showed that kelp was most (Table 4) occur on the intertidal benches at abundant followed by Ulun sp. and diatoms. She~nya.Sea urchin density and size are greater in Leptasterias alaslte~zsisand T. linza were observed channels and tide pools than in the lower and preying on mussels and barnacles. One L. alasker~sis higher intertidal regions (Tables 4 and 5). Chitons was obse17)edscavenging a sea urchin. are present only in the Herlopl~yllt~n~and Larni- mria zones. Balarlus gla?~rlulais most abundant in Atto. The algal associatioil at Attu is heavily the upper midlittoral region (Fig. l2), but B. grazed in the loxver intertidal area, in channels, and carios~tsis most abundant in the sublittoral fringe in tide pools and is very similar to the association and lolver midlittoral region where it occurs ~vith at Shemya. TIzalassiopl~yllzc~iiclatl~ncs occurs in the L. lorzgipes (Fig. 17) and H. sessile. R'lnssel beds are lower intertidal area, in tide pools, and in channels most abundant oil the seaward edge of the bench and is virtually ungrazed. in the high midlittoral region. Predacious sea stars The invertebrate association at Attu also is (Leptaste~fasalaskensis) are larger here than at similar to the association at Shcmya. Dense popula- Amchitka (Table 6). Hermit crabs (Pagu~nssp.) are tions of sea urchins (S. polyacaz~tl~us),chitons (K. abundant on the bench, in tide pools, and in tu~zicata),~nussels (M, edulis), and barnacles (B. channels in the upper midlittoral region. Amphi- glandttla and B, cariostts) occur on the bench. pods and isopods are not as abt~ndanthere as at Densities of hermit crabs (Pagrc~~tssp.) and crusta- Amchitka. Fewer than 20 individual amphipods ccans (amphipods and isopods) also are similar to and isopods were present under small rocks at low those at Shemya. Littleneck clams (Prototlraca sp.) tide. Limpets (C. pelta) are large and abundant on occur in gravel beds in the ~nidlittoralregion, and tide-pool ~valls and in the lower intertidal area the large chiton (C~yptocl~itor~stelleii) was noticed (Tables 4 and 5 and Fig. 14). in the sublittoral fringe. Sea urchins were observed grazing pri~narilyL. The food habits of sea urchins and predacious lor~gipes,If. sessile, A. oispn, and Ulva sp. Only sea stars and s~~ailswere similar to those at rarely were they obse~x~edgrazing T. clnthrus. Gut Shemya. 548 Pal~nisanoand Estes

Fig. 17-Abundant Bnlanus cnriosur in Laminoria zone at Shemya Island, Alaska, at low tide [ruler is 30.5 cm (12 in.) long] (June 28, 1972).

Sea urchins at Massacre Bay and Rlurder Point distribution) at Adak appears to be comparable to on Attu are larger than those at Shemya (Table 5). that at Amchitka. He made only a single dive at Sea urchins in Chichagof Harbor appeared to be Adak in 1972 [at 7, 13, 20, 27, and 34 m (23, 43, smaller (although no measurements were made) 66, 89, aild 112 ft) depths], but he made 62 dives than those at the other two study areas at Attu. at Amchitka and 4 dives at Shemya.

Subtidal Commi~nities Shemya and Attu. Subtidal macrophytes are generally absent from solid-rock substrate it1 the Amchitka and Adak. The subtidal macro- Near Islands (Figs. 19 and 21). They occur only in phytic associatio~lat Amchitka extends froin the small isolated patches, frequently located on sub- intertidal area and covers most of the solid-rock marine pinnacles, and often in a state of destruc- substrate to depths of 20 to 25 m (66 to 82 ft) tion from sea urchin grazing. (Figs. 18 to 20). h'lajor contributors to this associa- Sea urchin densities at Shemya and Attu tion are Alatia fistulosn, Lamitlaria lotzgipes, L. decrease with increasing \\rater depth (Fig. 19). groenla~~dica,L. yeroensis, L. delztigera, Agarz~n~ Immediately belolv mean water level, sea urchin cribros~r~n,Tl~alassiophylltcnz clathrus, Desmarestia populations frequently co~npletelycover the entire sp., and various Rhodophyta. Lanlinaria and rocky substrate. Agarln~zare the most abundant forms. Sea urchin populatio~ldensities increase with Vegetation. No data on vegctation coverage increasing depth at Amchitka (Fig. 19). At the were collected at the Near Islands. Subtidal macro- sublittoral fringe, sea urchins are seen infrequently phytes are absent from most areas at the Near unless they arc associated with algal holdfasts or Islands, and thus a coverage of zero percent is protcctive cracks and crevices in the substrate. inferred from our general obser\~ations(Fig. 19). Openly exposed sea urchins are rarely seen in water At Amchitka L. ~072gipes extends from the shallo~\~erthan 10 nl (33 ft). Below about 15 to 20 sublittoral fringe to a depth of 3 to 10 m (10 to 33 m (50 to 66 ft), lliglt-density sea urchin popula- ft). This species possesses a stipe that alloxvs it tions are openly exposed in some areas. flexihility in an area ~vherewave action is maxi- Phillip A. Lebcdnik (personal communication) mum. The association of three species (L. reported that the rocky subtidal com~nunity(algal groertla~zdica,L. yeroe~tsis,and L. de~ztigera)begins associations and sea urchin size, density, and at mean water level and extends out to a depth of Ecological Interactio~~sIt~uolui~~g the Sea Otter 549

CONSTANTINE POINT KlRlLOF ROCKS KlRlLOF POINT

2 .s ? 9 .? k -5 $ .!

DEPTH, rn Fig. 18-Vegetation coverage as a function of depth at the four Amchitka Island, Alaska, stndy areas. 550 Pahnisaizo nttd Estes

SEA URCHIN DENSITY, number of individuals pcr 0.25 m2 (SURFACE) 0

Fig. 19-Total macrophyte coverage and sea urchin density as a function of depth at Amchitka and Shemya Islands, Alaska.

Pig. 20-Dense ungrazed subtidal kelp beds in nearshore community of Amchitka Island, Alaska (Lami~tarialo:tgiDes in foreground and Alnria fistulosa in background) (Aug. 5, 1972). Ecological Ittternctiolts I~tvolvblgtlte Sea Otter 551

Fig. 21-Nearshore community of Shemya Island, Alaska, sho~ringa virtually complete absence of subtidal kelp. Note light-colored rock bottom uune 30, 1972).

23 to 25 m (75 to 82 ft). This association of population at Amchitka is shown in Fig.22a. species is dominant bet\veen about 3 and 12 m (10 These data indicate that the maximum test diain- and 40 ft). Agarun~criDroslr111, \\rhich begins at a eter of subtidal sea urchins at Amchitka Island is depth of about 6 m (20 ft) and extends out to a about 35 mm. Biomass distribution (as a function depth of at least 23 to 25 m (75 to 82 ft), is ~llost of size class) is also given in Fig. 22a. Biomass abiundant between 16 and 20 m (52 and 66 ft). density probably is more indicative than numerical Tlmlassiophyllu1tt clatl~rtrsand Destnarestia sp. are density of resource exploitation by various size oxlly relatively minor contributors to the sublit- classes. toral macrophyte association and seldom cover The size class distribution of a typical high- more than 10% of the rocky substrate. Tltakas- density sea urchin population at Shemya, from siuphylla~t~ciatl~rits ranges in depth between 3 and water about 2 to 3 m (7 to 10 ft) in depth and 20 m (10 and 66 ft) and is most abundant at a immediately adjacent to the intertidal area, is depth of about 10 to 12 m (33 to 40 ft). shown in Fig. 22b. Apparently two peaks are in Des~~tarestiasp. also ranges in depth between 3 and this distribution, one at 10 to 15 mm in diameter 20 in (10 and 66 ft) but sho\\~sno distinct pattern and another at 60 to 65 inn1 in diameter. The of maximum abundance. Foliose Rhodophyta oc- maximum size of sea urchins from this sample is cur from the sublittotal fringe to beyond 25 111 (82 about 87 mm, although indi\~idualsover 100 lnrn in ft) in depth. Near the surface Rhodophyta grow diameter were found \vhile we \\,ere specifically beneath the Lanti~za~facanopy as \veil as searching for large animals. The biomass of sea epiphytically on the Phaeophyta. Abundant space urchins also is shown in Fig. 22b. A total biomass is available on the rock substrate in the shallow density of 12,328 g/m2 \\.as esti~nated for the subtidal region because blades of L. goe~zlarldica, Shemya subtidal sea i~rcltiilpopulation in areas of L. detttigera, and L. yezoe~lsisare extended a ineter high population density. The hionlass of the or more toward the surface on thick inflexible Amchitka sea urchin popillatioil in areas of maxi- stipes. Rhodophyta are abundant continuously mull1 density \vas estimated at 1496 g/m2. Of total from the surface to beyond 25 m (82 ft), perhaps sea urchin bionlass density at Shcmya, 11,672 showing a slight increase in abunda~cewith in- g/lnZ (95%) is contribittcd by that scgment of the creased depth (see Chap. 17, this voltun~e, for popi~iationthat is larger than the largest sea urchin distribution of A. fistulosa). observed from tlte Amchitka samples. Sea Urchin Populations. The size class dis- At Shemya, at depths of 10 to 15 m (33 to 50 tribution of a typical high-density sea urchin ft), the peak of small animals in the size class distribution is reduced, and the peak of larger tide pools, and around and uuder rocks. Here many individuals is absent. The biomass density also is species of abundant amphipods and isopods (often reduced to 824 g/m2. At these depths 65% of the hundreds of individuals per squarc ineter) and sea biomass is contributed by animals larger than the urchins feed on this algal debris and detritus. niaximum size at Amchitka (Fig. 22c). About 5000 Emperor Geese (Pltilncte ca~~agica) At deptbs of 20 to 25 in (66 to 82 ft), spend the winter at Amchitka (\\'illiamson, populatioi~densities are further reduced, aud there Emison, and \Vhite, 1971) \vhere they feed in the are 110 apparent peaks in the size class distribution. intcrtidal area during low tide. Field observatio~ls Here the biomass density is reduced to 208 g/m2, and stomach analyses from six adult birds shotsed 39% of xvhich is corrtributed by animals larger than that Uiua sp. and small red algae (Pol;olzyrn sp. ant1 the maximum size at Amchitka (Fig. 22d). Rlrody?,te?tia sp.), but not kelp, \\'ere eaten (Palmisano, 1975). Similar findings were made by RESULTS \\'illianison, Emison, aild \\'bite (1972). Intertidal Community Growth and Feeding Stucies. Five intertidal sea urchins from A~ncl~itkadoubled in size within 1 Grazing Studies. At Amchitka sea urchin graz- year after they were transferred to the marine ing experiments in L. loltgipes and H. sessile zones laboratories at Friday Harbor and kept in a water denio~~stratedthat dense populations of sea urchins table contailling a constant supply of kelp cottld reduce algal cover. Twenty 19-mm-diameter (Table 5). sea urchins grazed 80% of all encaged pla~~tsaud The food-preference tests conducted at reduced algal co\ler by 60 to 70% within 9 months Amchitka shox\red that sea urchins from Amchitka whereas coi~trolcages maiiltai~led100% algal cover. iuld Sheinya prefer L. lo~tgipes,A. crisp, and H. At Shemya both tide pools cleared of sea sessile to T. clatl~rrts. urchins in September 1971 had a 90% coxrcr of Mussels [I\{. eedlis (Z, 18 n~mlong)] from Ulua lactt~cnby June 1972. The larger tide pool Puget Sound that had been transferred to Am- contained three II. sessile plants and five large sea chitka tripled their size (%, 58mm long) in urchins (about 75 mm in diameter). The smaller 18 n~onths. tide pool co~ltai~ledno sea urchins. There were no At Friday Harbor, intertidal sea stars from noticeable changes in algal species or in percentage Amchitka, xvhich were fed ope~tedlive clams, of cover in the two control tide pools from \vhich mussels, and barnacles for 1 year, increased more sea urchins had not bee11 removed. than one and one-half times in size and more than Limpet graziug experiments co~lductedin dog four times in weight (Table 6). Field observatio~ls dishes on nelvly denuded rock surfaces in the at Amchitka on 35 feeding sea stars (L. alaske~uis) midlittoral region at Duck Cove aild in the midlit- showed that small unattached prey, such as crusta- toral region and sublittoral fringe at i\,Iakarius Bay ceans (amphipods and isopods) and mollusks showed that after 1 year dishes ~vithoutlimpets (clams and snails), comprised over 90% of their had 100% algal cover (mostly U. lnctrlca and H. diet (by number). gla~zdifor~te)\\'hereas dishes with three 30-mm- Tlra~ilimn from irttertidal benclics at Amchitka long limpets had bettveen 2 and 10% cover of the preyed on live barnacles at the University of same algal species. Within 1 year after limpets bad \\'ashington in Seattle. Ficld obsei-\'ations on 106 been removed from the Hnlosaccio~~zone at feeding s~lails(T. li7na) sho\vcd that small littorille h,Iakarius Bay, previously grazed (bare) areas be- snails (Littoriua) co~nprisedover 93% of their diet came completely occupied by small red algae (by number). Tltais li~~tnalso preyed on barnacles, (Rlto~l)~?~~etrinsp. and I. cor~~ticopicre).Similar arcas \vl~erebarnacles were abundant in the iiltcrtidal \vhcre limpets had not been removed remained region (e.g., Co~lstantineHarbor), and on l~lussels bare. that had been trailsferred from Puget Sound to the Pieces of subtidal and intcrtidal drift algae intertidal region at Makarius Bay. placed in small cups containing seawater and either ttmphipods or isopods were noticeably grazed after Siltation Studies. Of 10 dishpans placed in the 2 days. These algae remained intact 1v11en placed in intertidal area in July 1972, only those enclosing cups containi~igonly filtered seawater. Hedopl~yllum and Hnlosaccio~t (in high and pro- Field observations sho\ved that drift kelp on tected areas) contained silt. The silt averaged 8 cm the shoreliile (composed mostly of subtidal species, (3 in.) in depth, and there were no sessile a~~ilnals such as il. fistttlosa) provides food and shclter for or ~iiacro~hyticplants inside the pans. Pans enclos- a~~lphipods(Orchestin) and kelp flies (Coelopidae). ing L. lo?tgipes, C. ua~tcouuerietrsis,and A. crispn 11s kelp decon~posesit accumulates aroluld algal (in lolv and exposed areas) contained healthy holdfas~sin calm areas, in small depressioi~sand attached algae but no measurable silt. Ecological Inteluctio~~s1i~volvi)lg tlze Sea Otter 553 554 Paln~isai~oand Estes

Silt 1 to 10 cm (0.4 to 4 in.) deep covers large each adult averaged one dive and five sea urchins areas of bays and landlvard edges of wide intertidal every 2 min. At this rate one sea otter \\rould eat benches. No sessile invertebrates or macrophytes 150 sea urchins per hour. were observed in these areas, although the occur- During one week in September 1972, sea otters rence of crustose coralline algae beneath the silt were seen daily at higlt tide within 1 m (3 ft) of the suggests that the silt cover is absent during part of break~vater at Sweeper Cove, Adak, where the the year. (See Lebednik and Palmisano, Chap. 17, water depth was less than 3 m (10 ft). These this volume.) animals predominantly fed on sea urchins that Silt also accu~nulatcsaround algal holdfasts and \\,ere estimated to be 20 to 50 lnln in diameter. rocks in sheltered intertidal areas and in higher One adult sea otter was observed for 30 min, intertidal zones. Little or no silt accu~nulatesat during ~vhichtime it averaged one dive and four sea exposed sites or in lo\\rer intertidal zones. urchins per minute. At other areas at Adak, sea Thirty mussels from Puget Sound (10 to 60 otters were observed feeding on itnidentified food mm long) that had been transplanted to the items, although their scats along the rock and Halosacciou zone (high intertidal region) at cobble shoreline provide evidence that crab, clam, i\4akarius Bay in December 1972 were covered \\pith and nn~sselsare consiuned. silt and were dead 1\4ien esa~ninedin April 1973. A Approximately 25 sea otters were seen at similar group of nut~ssels transplanted to the Chichagof Harbor on Attu. Several of these otters La~i~i~~ariuzone (lo\\, intertidal region) were alive were obse~vedfeeding on sea urchins and mussels and free of silt in April and in i\uwst 1973. ~vithin100 m (325 ft) of shore. No other sea otters were seen at Attu. 011 subsequent visits to Attu in Settling Studies. No baruacles or n~usselsset 1975 and 1976, more than 250 and 340 otters, on oyster shells or in the 1-m2 denuded rock respectively, were counted. surfaccs at Duck Cove or ivIakarius Bay. Ho~vever,a \Ve saw no sea otters during either trip to few barnacles (B. carioslrs) set on the mussels, Shemya. No otters \Irere observed in 1975 or 1976 ropes, and bottles suspended from the dock in at Sllenlya. Constantine Harbor. Also a few barnacles set at Rifle Range Point ill three 0.25-m2 areas denuded Blade \\'idti1 of Latnina~iulotrgi$es. The blade of C. va~~co~tverie~~sis. ~vidthof ungrazed L. loilgipes was greater in pro- tected areas than in exposed areas at each of the Sea Otter Predation. On se\leral occasions four islands studied (Table 7). small groups (t\vo to six) of sea otters were observed feeding on iutertidal benches (sublittoral fringe) during high tide at Makarius Bay, Rifle DISCUSSION Range Point, Duck Cove, and Square Bay. Data collected on three adult sea otters (one with a pup) The direct and indirect consequences of sea at Makarius Bay during 30-mi11sampling periods on otter predation on intertidal communities and the five consecutive days in August 1973 showed that effect of predation by sea otters on subtidal

Table 7-hlean Blade IVidth (mm) of Ungrazed Intertidal Lnt?litmrin longi>es in Areas Snbjectively Considered Exposed or Protected (from Wave Sltock) in the IVesterrl Ale~ttianIslands, Summer 1972* Exposed Protected - - Island N x. R SD C1 I\' x R SD 61

Aincl~itka 100 34.3t 20 to 45 7.3 t14.4 100 91.1$ 50 to 140 18.8 +37.2 Adak 100 35.4s 22 to 46 6.9 f13.7 100 88.2n 48 to 140 16.9 f33.5 Sbemya 100 24.6** 15 to 38 6.1 t12.1 100 45.8tt 28 to GO 13.3 i26.3 Attu 100 23.355 15 to 38 5.2 f10.3 100 88.27n 51 to 138 18.0 f35.6

*N, nun~bermeasured; S, average individual size in millimeters; R, range of individual sizes in millimeters; SD, standard deviation; GI, 95% confidence intend. +Rifle Range Point. $Makarius Bay. $Cape Adagdak. nKuluk Bay. **Urchin Point. ttThere were no protected intertidal benches at Shemya; this value came from L. lo,rgiprs plants in a drainage channel near the shorervard portion of the Urchin Point bench. s Shlurder Point. BnMassacre Bay. Ecological Interactioits Involuiitg the Sea Otter 555 communities are discussed, and this information is Category I. Sea otters consume 20 to 23% of integrated with tl~eoreticalconcepts of community their body weight in food daily (Kenyon, 1969). organization in an attempt to elucidate the role of Stomach analyses of (Kenyon, 1969; Lensink, sca otters in the nearshore com~nunityat Am- 1962; Burgner and Nakatani, 1972), field observa- chitka. tions of (Kenyon, 1969; Estcs and Smith, 1973; personal observations), and feeding experiments Intertidal Community with captive sea otters, all at Amchitka, show that sea otters are opportunistic predators on benthic Predation can influence the structure of inter- intertidal invertebrates (sea urchins, chitons, mils- tidal co~n~niunitiesby preventing competitive exclu- sels, limpets, etc.) and bottom-dwelling fishes. This sion or by directly affecting local extinction suggests that the dense population of sea otters at (Connell, 1961a; Paine, 1966). Wave shock Amchitka preys heavily on known food items and (Kicketts and Calvin, 1968) influences con~n~unity is responsible for scarce populations of such structure in a similar way in addition to reducing intertidal animals as the sea urchin. silt accumulation and pro\riding moisture to high High densities of intertidal sea urchins and intertidal zones. Our study suggests that sea otters mussels at Amchitka in the 1930s and at Adak in can influence both of these factors and therefore the 1950s, when sea otters were less abundant on are an important factor in determining co~nmunity both islands (I

r- or I<-Selectioii? The concept of r- and K-selection was conceived originally by i\~IacArthur and \\'ilson (1967) to explain reproductive strate- gies of species that are closely associated eco- Fig. 23-Sea urchins massing about a strand of logically. I11 an uncro\vded environment, selection Alnria firtrclosn that has drifted into a large tide pool preference is estended to~vard r-strategists or in lower intertidal area at Shemya Island, Alaska toward species with the highest intrillsic rate of (Sept. 2, 1971). nati~ral increase. The strategy is to convert a maxinlum anlount of food into ne1v members of fistnlosa that had drifted into a large tide pool of the population. In a crolvded environment, i.e., the lower intertidal area at Shemya. Also, ~vhile one in which popi~lations exist near equilibrium diving on Attu we encountered several groups of density, I<-strategists are favored. Efficiency of sea urchins massed together in the for111 of a food use is important in mailltaining large popula- sphere. \\'e separated one of these spheres and tions, and wasteful exploitation of food resources, found a small piece of TItalassiopl~yllzrn~blade near such as for the production of large numbers of its center. offspring, is selectively unfavorable (Crow and Displacement of nutrients by Ivave action from ICimura, 1970). Pirulka (1970) sunl~narizedsolne of the large sta~ldingcrop of intertidal macrophytes the correlates of r- and I<-selection, a few of which may also be important to sea urchins in the shallow are given in Table 8. subtidal area at Attu and Shemya. I11 fact, appar- From our prior discussion and from the infor- ently there is almost no source of llutricnt produc- lnation given in Table 8, sea urchins at Amchitka tion sympatric ivith subtidal sea urchin popula- appear as r-strategists whereas those at Attu and tions. The expected distribution of nutrient input Shemya appear as K-strategists. This is unlikely (drifting algae from other areas ancl nutrie~lts because we are dealing with a single species. Thus xeashed from the intertidal area) corresponds we expect these populations to be similar geneti- closely with observed distribution of sea urchins. cally and either r- or K-strategists but certainly not I11 direct contrast to the distribution of sea both. urchins at Shemya and Attu, sea urchins at \\re propose that, whereas sea urchins at Attu Amchitka are most severely limited in the im- and Shemya fulfill many requirements of I<- Ecologicnl Ijtternctiot~sIt~uoluittg the Sen Otter 563

Table 8-Sosle Correlates of r- and K-Selection* Attu and Shemya. This species feeds largely on c~astacean invertebrates, which, in turn, are as- Item r-selection I<-selection sociated with kelp beds (see earlier discussion). hlortality Often catastrophic, Density dependent. Higher trophic forms at Amchitka, which are density independent. dependent on fish for some part of their food, Survivorship Often type Ill Usually types I and probably either use different foraging strategies or (Deevey, 1947). II (Deevey, 1947). mai~ltaitlsmaller populations at Attu and Shemya. Population Variable in time, non- Fairly consistent in Sea otters coilsulne greater amoullts of fish at size equilibrium; usually time, equilibrium; well below caving at or near carrying Amchitka than in areas of low sea otter population capacity of environ- capacity of the density (Lensink, 1962). Harbor seals (Pltocn U~~IL- ment; unsaturated environment; litin) are less common at Attu md Shemya that1 at communities or saturated com- Amchitka where groups of over 100 com~nonly portions thereof; munities. haul out at numerous meas dong the coast ecological vacuunls. Iutraspecific Variable, often lax. Usually keen. (Kenyon and Icing, 1965). Harbor seals feed and inter- predominantly on nearshore fishes and cepha- specific lopods at Amchitka (Wilke, 1957; Kenyon, 1965). competition The Bald Eagle (Hnlineetits lettcocepl~abts), altltough abundant in the Rat Islands, is absent in *From E. R. Pianka, On r- and K-Selection, Americon Naturalist, 104: 593 (1970). the Near Islands (Sekora, 1973; personal observa- tions). Bald Eagles in the Aleutians are largely strategists, the species has evolved through r- dependent on marine productivity. Fish, marille selection. Thus the tendency toward maximum mantnals, and marine birds have been found to energy exploitation by w u~~controlledr-strategist constitute the majority of their diet on Amchitka is illcon~patiblewith maintenance of the energy (White, Emison, and \\'illiamson, 1971). Also, drift resource (e.g., macrophytes are destroyed). algal stipes are major constitnents of Bald Eagle nests at Amchitka Island (persotla1 observation). Sea Otter and Productivity of the Nearshore Tbc absence of eagles in the Near Islands may Con~monity. Benthic macrophytes are of con- therefore be related indirectly to reduced mac- siderable importance to nearshore producti\ity in rophyte productivity. temperate waters. Blinks (1955) stated that, The Common Eider (Sottlnteria tt~ollissii~ra)is although littoral marine algae occupy only a by far the most common duck in saltivater areas of narrow coastal zone as compared with the elltire Attu Islaud (Near Islands) (Byrd, 1973). This ocean, their productivity may be several orders of species probably owes its high density to the sea mag~litudegreater per comparable wit than the urchins (Byrd, 1973). Althongh the Common Eider open sea exccpt in areas of up~velling. n'lann is present at Amchitka, it is not nearly so abundant (1972b; 1973) found that forests of Latttit~n,.inand there as at Attu, probably reflecting the decreasetl rlgnntm, which characterize the subtidal area of availability of sea urchins at Amchitka Islzu~d. temperate oceans, contribute considerably to the Sea otters also may contribute to the growth of total productivity of coastal waters in the North kelp beds (and associated species of animals) along Atlantic Ocean. He attributed this to the attach- much of the west coast of North America, h,lcLeru~ ment of plants on the rocky substratum and thus (1962) reported that sea otters in California to an unusually good supply of nutrients circulated completely remove large sea urchins (S. fin11- by tides and currents. Significantly the range of ciscnttzts) from areas by predation, permitting feeding habitat of sea otters coincides with, and lusuria~lt development of the Nereoc)lstis- extends beyond, the area of epibenthic algae. As Pterygophorn association. North (1965) attributed such the sea ottef may indirectly maintain the kelp recent kelp bed macrocy cyst is) reduction to sea beds (see earlier discussion). urchin graziug and further speculated that an Some anilmd species present at Amchitka but increase in ~Vlncrocystis deusity in the h,lonterey scarce or missing at Attu and Shemya may depend area resulted from sea urchin removal by sea otters. on the highly productive macrophytic algae as a h4iller and Geibel (1973), holvcver, poiut out that nutritional base. Those animal species whose food recent changes in distribution and abundance of webs originate from macrophyte productivity cer- ~Vlacrocystis and ~\'ereoc)*stis in many areas are tainly could be adversely affected by its removal related to factors other than sea otter prcd a t'ion. (Paine, 1966). Thus, although sea otters clearly reduce benthic The rock greenling (Hexngmttttnos lago- invertebrate populations in these areas of the cepl~nlcts),abundant and commonly encountered California coast, other interactions are not well while diving at Amchitka, is infrequently seen at understood. 564 Pal~r~isartoand Estes The Aleut has inhabited the eastern Aleutian reducing wave shock aud increasing siltation rates, Islands for at least 8400 years (Laughlin, 1972) \ve fitrthcr conclitde that sea otters are indirectly and Amchitka for at least 2500 years (Desautels et respollsible for the sparse populatiotls of barnacles al., 1970; A,IcCartney, Chap. 5, this volome). and tnussels and for the small size and loxv number Laughlin (1972) believed that the Aleut influenced of predacious snails and sea stars and the small the sea otter by direct exploitation through hunt- llumber of hermit crabs. ing and competition for food resources by It is possible that physical factors, such as gathering sea urchins. \\re doubt that the Aleut was intertidal benches, coastline features, and climate, able to compete with sea otters for food, although or other biological factors, such as other predators Aleuts collected sea urchins from the iutertidal and and competition resource shortages, are partly or very shallow subtidal zones. I-Iowever, Aleuts cer- fully responsible for the structure of the nearshore tainly were capable of reducing sea otters by coln~lluilityat Amchitka. Horvever, comparisons of hunting, as is evident from the fact that sea otters \vestem Aleutian nearshore communities, similar in were driven nearly to esti~lctio~lby Aleuts after most aspects except for densities of sea otter they were enslaved by Russian fur traders. Further- populations, and the results of field and laboratory more, evidence from middens suggest that Aleuts esperiulents strongly indicate that the sea otter is depleted sea otter populations long before the the keystone species at Amchitka that is primarily arrival of white men. Ho\ve\'er, from the stand- responsible for the observed structure of the point of evolutionary time, the Aleut is recent. nearshore community. Thus his importance as a natural predator in the nearshore commiu~ityis not yet clear. Some of the preceding discussio~lis speculative. Ho~vever, the sea otter quite obviously is an important species in determining structures and Research was supported by AEC contracts dynamic relationships with nearshore commutlities, AT(26-1)-520 and AT(26-1)-171 through subcon- thus fitting the concept of Paiue (1969) of a tract from Battelle Memorial Institute, Columbus, keystone species. Until the recent resurgence of sea Ohio. \\re are indebted to K. Chew, R. Nakatani, otters on the Pacific coast, biologists \\'ere unaware and N. Smith for guidance, and to S. Browtl, of the i~lfluellceof sea otters on intertidal and R. Glinski, P. Lebednik, C. O'Clair, C. Simenstad, subtidal communities within the original range. N. Smith, and G. Tutmark for field assistance. Our studies indicate that the sea otter was a \\'e thank P. Dayton and R. Paine for helpful com- significant factor in the e\~olutionof the nearshore ments in preparing the manuscript, J. Jackson for ecosystem of the North Pacific Ocean. assistance with logistic problems, and R. Femald for providing facilities at Friday Harbor Labora- tories. The U. S. Air Force, U. S. Coast Guard, aud CONCLUSIONS U. S. Navy provided access to their facilities at Shemya, Attu, and Adak Islailds, respectively. That sea otters are capable of reducing both Final drafts \\,ere prepared by the Fisheries Re- the size and llumber of sea urchins is indicated by search Institute and the Auke Bay Fisheries Labo- their voracious appetite, analyses of their stomach ratory. All photogl-aphs are by J. F. Palmisano. contents, and feeding observatiolls of wild and captive animals. The overwl~elmingevidence in the literature and the results of graziug experiments conducted at Amchitka and Shemya clearly APPENDIX: PARA\IETER INTERVAL ESTIMA- demonstrate that marine vegetation flourishes in TION AND TESTS OF SIGNIFICANCE IN THE the absence, but is destroyed in the presence, of LINEAR MODEL yi = p + P(xi) + ei dense populations of sea urchins. From further ohservatio~lsof similar areas in the western Aleu- Variance (4)+ 1.454 + 2 tians with aud \\.ithout sea otters (both in time and Variance (p) = 1.174 space), we conclude that in thc nearshore conl- 95% collfidellce intelval (p) = -8.114 to mu~lity at Amchitka sea otters are directly -7.636 responsible for the reduction of the size aud 95% confidence interval (0) = 2.924 to 3.059 nttmber of sea urchins \vithin their feeding range t-test of hypothesis (p = 0) t = 65.307* with and indirectly responsible for the resultant 11 1 degrees of freedom development of estensive algal beds and associated t-test of hypothesis (6 = 0) t = 87.310* with communities. Because these kelp beds can alter 11 1 degrees of freedom physical couditiolts in thc intertidal community by Analysis of variance for linear regression Ecological Interactio~lsI~tuolui?tg the Sea Otter 565

Dayton, P. K., 1973, Dispersion, Dispersal, and Persistence Source of Degrees of Sum of Mean of the Annual Intertidal Alga Postelsin pa1,rmlformis variation freedom squares square F Ruprecht, Ecology, 54: 433-438. , 1971, Competition, Disturbance, and Community Due to 1 289.226 289.226 7623.1* Organization: The Provision and Subsequent Utilization regression of Space in a Rocky Intertidal Community, Ecol. Residual -111 4.211 0.0379 itlonogr., 41: 351-389. Total 112 293.437 -, 1970, Competition, Predation, and Community Struc- ture: The t\llocation and Subsequent Utilization of *Significant at 0.01 level. Space in a Rocky Intertidal Community, Ph. D.Thesis, University of Washington. REFERENCES -, and R. I<. Hessler, 1972, Role of Biological Distur- bance in blaintaining Diversity in the Deep Sea, Addicott, J. F., 1974, Predation and Prey Community Deep-Sea Res., 19: 199-208. Structure: An Experimental Study of the Effect of - R. J. Rosenthal, and L. C. Mahan, 1973, Kelp Afosquito Larvae on the Protozoan Communities of Communities in the Chilean Archipelago: R/V Hero Pitcher Plants, Ecology, 55: 475-492. Cruise 72-5,Antnrct. J. U. S., 8: 34-35. Aleem, A. A,, 1956, Quantitative Underwater Study of Deevey, E. S., 1947, Life Tables for Natural Populations of Benthic Communities Inhabiting Kelp Beds off Cali- Animals, Quart. Rev. Biol., 22: 283-314. fornia, Science, 123: 183. Desautels, R. J., A. J. AfcCurdy, J. D. Flynn, and R. R. Barr, L., 1971, Studies of Populations of Sea Urchins, Ellis, 1970, Archaeological Report, Amchitka Island, Stron.olocentrotus SD.. in Relation to Undereround Alaska, 1969-1970, Archaeological Research, Inc., Nucleyr Testing at Alkhitka Island, Alaska, Bioscience, Costa Mesa, California. 21: 614-617. Diamond, J. hI., 1973, Distributional Ecology of New Blinks, L. R., 1955, Photosynthesis and Productivity of Guinea Birds, Science, 179: 759-769. Littoral hlarine Algae, J. dlnr. Rex, 14: 363-373. -, 1969, Avifaunal Equilibria and Species Turnovcr Rates Bur~ner.- R. L.. and K. E. Nakatani. 1972. Amchitka Bio- environmental Program. Research Program on Maine on the Channel Islands of California, Proc. Not. Acnd. Ecology, Amchitka Island, Alaska, Annual Progress Sci., 64: 57-63. Report, July 1, 1970-June 30, 1971, USAEC Report Doty, M. S., 1946, Critical Tide Factors that Are Correlated BhfI-171.144, Battelle hIemorial Institute. with the Vertical Distribution of Alarine Algae and Byrd, G. V., 1973, Assistant Refuge Alanager, U. S. Fish Other Organisms Along the Pacific Coast, Ecology, 27: and i\'ildlife Service, personal communication (from 315-328. observations made at ~ttuIsland between May 25 and D~ehl,L. D., 1968, Taxonomy and Distribution of June 1, 1973). Northeast Pacific Species of Lantinarin, Can. J. Bot., 46: 539-547. cam^.. . D. K.. S. P. Cobb. and "1. F. Van Breedveld. 1973. Overgrazing of Seagrasses by a Regular Urchin, Ehert, E. E., 1968, California Sea Otter Census and Habitat I,y techinus uariegatus, Bioscience, 23(1): 37-38. Survey, Underwater ~Vntar.,5: 20-23. Carlquist, S., 1973, Island Biology, Columbia University Estes..- I. A,.. and "1. F. Palmisano. 1974. Sea Otters: Their Press, New York. Role in Structuring Nearshore Communities, Science, -, 1965, Island Life: A Natural Histor), of the Islattds of 185: 1058-1060. the IVorld, Natural Histoly Press, Garden City, New -, and N. S. Smith, 1973, Amchitka Bioenvironmental York. Program, Research on the Sea Otter, Amchitka Island, Connell, J. H., 1972, Community Interactions on Marine Alaska. Final Report, Oct. 1, 1970-Dec. 31, 1972, Rocky Intertidal Shores, Annu. Reu. Ecol. Syst., 3: USAEC Report NVO-520.1, University of Arizona. 169-192. Fischer, A. G., 1960, Latitudinal Variations in Organic -, 1970, A Predator-Prey System in the h,larine Intertidal Diversity,Euolutio?t, 14: 64-81. Region. I. Balantrs glandula and Several Predatory Harper, J. L., 1969, The Role of Predation in Vegetational Species of Thais, Ecol. dlonogr., 40: 49-78. Diversity, in Diversity and Stability in Ecological -, 1961a, The Influence of Interspecific Competition and Systems, Report of Symposium held hiay 26-28, 1969, Other Factors on the Distribution of the Barnacle Brookhaven Symposia in Biology, Number 22, USAEC Chtkamnlus stellatus, Ecology, 42: 710-723. Report BNL-50175, pp. 48-62. -, 1961h, Effects of Competition, Predation by Thais Haxo, F. T., and L. R. Blinks, 1950, Photosynthetic Action lapillas, and Other Factors on Natural Populationsof the Spectra of Marine Algae, J. Gen. Plzysiol., 33: 389-422. Barnacle Balanus bolanoides, Ecol. AIonofr., 31: 61-104. Hedgpeth, J. iV., 1957, dlnrine Biogeography, U. S. Geolog- Crisp, D. J., and H. Barnes, 1954, The Orientation and ical Society, Memoir 67(1), pp. 359-382. Distrihution of Barnacles at Settlement with Particular Hewatt, 1V. G., 1937, Ecological Studies on Selected RIarine Reference to Surface Contour, J. Anim. Ecol., 23: Intertidal Conlmunities of Alonterey Bay, California, 142-162. Amer. dIidlnnd Nntur., 18: 161-206. Crossett, R. N., 1967, The Vertical Distribution of At- Himmelman, J. H., and D. H. Steele, 1971, Foods and tached Marine Algae in Malta, J. Ecol., 55: 361-371. Predators of the Green Sea Urchins Stro~~gylocentrotus Crow, J. F., and AI. Rimura, 1970, An httroduction to dro6acltiensis in Newfoundland \\'aters, dlar. Biol., 9: Populatio,t Ge~teticsTfteory, Harper & Row, Publishers, 315-322. New York. Ilutchinson, G. E., 1953, The Concept of Pattern in Dawson, E. Y., M. Neushul, and R. ivildman, 1960, Ecology,Proc. Nat. Acad. Sci., 105: 1-12. Seaweeds Associated witb Kelp Beds Along Southern Isakson, J. S., C. A. Simenstad, and R. I.. Burgner, 1971, Califo~niaand Northwestern hlexico, Pac. ~Vaturalist, Fish Communities and Food Chains in the Amchitka 14: 1-81. Area, BioScience, 21: 666-670. Jones, N. S., 1948, Observations and Experiments on the Zonation and Biomass of Seaweeds, Marine Biol., 12: Biology of Pntella vulgata at Port St. Mary, Isle of Man, 1.10. Proc. Tram. Liverpool Biol. Soc., 56: 60-77. -, 197213, Ecological Energetics of the Seaweed Zone in a -, and J. M. Kain, 1967, Suhtidal Algal Colonization Marine Bay on the Atlantic Coast of Canada. 11. Following the Removal of Echinus, Helgolander IYiss. Productivity of the Seaweeds, Afan'ne Biol., 14: dfeeresunters., 15: 460.466. 199-209. Jorde, I., and N. Klavestad, 1963, The Natural History of hlay, R. H., 1973, Stability and Complexity in Model the Hardangerfjord. 4. The Benthonic Algal Vegetation, Ecosystems, Princeton University Press, Princeton, N. J. Sarsia, 9: 1-99. McLean, J. H., 1962, Sublittoral Ecology of Kelp Beds of Kain, J. M., 1966, The Role of Light in the Ecology of the Open Coast Area Near Carmel, California, Biol. Lantinaria hyperborea, in Light as an Ecological BUN., 122: 95-114. Factor, British Ecological Society Symposium No. 6, Miller, D. J., and J. J. Geibel, 1973, Summary of the Blue Richard Bainbridge et al. (Eds.), pp. 319-334, Blackwell Rockfish and Lingcod Life Histories; a Reef Ecology Scientific Publications Ltd., Oxford. Study; and Giant Kelp Experiments in Monterey Bay, -, and N. S. Jones, 1965, Algal Colonization After California, California Department of Fish and Game, Removal of Echinus (abstr.), in 5th International Fish Bulletin 158. Seaweed Symposium, Dalhousie University, Halifax, -, J. J. Geibel, and J. L. Hook, 1974, Results of the 1972 Nova Scotia. Skindiving Assessment Survey, Pismo Beach to Oregon, Kenyon, K. \V., 1969, The Sea Otter in the Enstern Pacific California Department of Fish and Game, Alarine Ocean, U. S. Department of the Interior, Bureau of Resources Technical Report No. 20. Sport Fisheries and TVildlife, North American Fauna, Moore, H. B., 1958, Afarine Ecology, John Wiley & Sons, No. 68. Ltd., London. , 1965, Food of Harbor Seals at Amchitka Island, Morrison, P., M. Rosenmann, and J. A. Estes, 1974, Alaska, J. ~lfammalogy,46: 103-104. bletabolism and Thermoregulation in the Sea Otter, -, and J. G. King, 1965, Aerial Survey of Sea Otters, Physiol. Zool., 47: 218-229. Otber Marine hlan~malsand Birds, Alaska Peninsula and Nelson-Smith, A,, 1968, Biological Consequences of Oil Neutim Islands, 19 Apdl to 9 May, 1965, U. S. Fish Pollution and Shore Cleansing, in The Biological Effects and \Yildlife Service, Bareau of Sport Fisheries and of Oil Pollution on Littoral Communities, Proceedings \Vildlife. of a Symposium held at the Orielton Field Centre, Kitching, J. A,, 1941, Studies in Sublittoral Ecology. Ill. Pembroke, \Vales, on Feb. 17-19, 1968, J. D. Carthy Laminaria Forest on the West Coast of Scotland; A and D. R. Arthur (Eds.), Supplement to Volume 2 of Study of Zonation in Relation to \Vave Action and Field Studies Council. Illumit~ation,Biol. BILM., 80: 324-337. Neushol, M., 1967, Studies of Subtidal Marine Vegetation -, and P. J. Ehliog, 1961, The Ecology of Lough Ine XI, in \Vestern \irashington, Ecology, 48: 83-94. Tbe Control of Algae by Paracentrotus lividus -, 1965, Scuba Diving Studies of the Vertical Distribo- (Echinoidea), J. Aninz. Ecol., 30: 373-383. tion of Benthic Marine Plants. Botanica Gotho- Laughlin, \V. S., 1972, Ecology and Population Structure in burgensiaII1, Marine Biology, Vo1. 5, J. D. Costlow the Arctic, in Tlze Strt~ctureof Ifu~nnnPopulations, (Ed.), Proceedings of the Fifth Interdisciplinary Confer- ence on Marine Biology, pp. 161-176, Gordon & Breach G. A. Harrison and A. 1. Bavce. (Eds.). . .. DD... 379-392. Oxford University Press. Science Publishers, Inc., New York. Lensink. C., 1962, The History and Status of Sea Otters in Newell, R. C., 1970, Biology of Intertidal Ani~nals,Ameri- Alaska, PII. ~.~hesis,~urdue University, Xerox Univer- can Elsevier Publishing Co., Inc., New York. sity Microfilms, Ann Arbor, Ivlich. North, \V. J., 1965, An Ecological Study of the Point Loma Kelp Bed, San Diego County, State \Vater Control Lewis, J. R., 1964, Tlze Ecology of Rocky Sllores, English Universities Press, Ltd., London. Board, Sacramento, Calif. Limbaugh, C., 1955, Fish Life in the Kelp Beds and the O'Clair, C. E., and K. K. Chew, 1971, Transect Studies of Effects of Harvesting. University of California Institute Littoral Macrofauna, Amchitka Island, Alaska, Bio- of Marine Research, Institute of Marine Research Science, 21: 661-665. Reference 55-9, pp. 1-158. Odemar, M. \V., and K. C. Wilson, 1969, Population Studies of Sea Otters in California, California Department of Lowry, L. F., and J. S. Pearse, 1973, Abalones and Sea Fish and Game, Marine Research Report, Reference No. Urchins in an Area Inhabited by Sea Otters, Afarine 70-3.~~ ~ Biol., 23: 213-219. Ostle, B., 1963, Statistics in Research, Iowa State Univer- MacArthur, R. H., 1972, Geographical Ecology: Patterns in sity Press, Ames. the Distribzction of Sbecies. Haroer & Row. Poblishers. <. , . Paine, R. T., 1971, A Short-Term Expcrimental Investiga- Inc., New York. tion of Resource Partitioning in a New Zealand Rocky and E. 0. \Vilson. 1967. The Theorv of Island -. , . Intertidal Habitat, Ecology, 52: 1096-1106. Biogeography, Princeton University Press, Princeton, -, 1969, A Note on Trophic Complexity and Community N. J. Stability, An~er.Aratrrr., 103: 91-93. -, and E. 0. Wilson, 1963, An Equilibrium Theo~yof -, 1966, Food \Veb Complexity and Species Diversity, Insular Zoogeography, Evolution, 17: 373-387. Amer. Natur., 100: 65-75. MacGinitie, G. E., and N. MacGinitie, 1968, Nal~cral -, and R. L. \'adas, 1969, The Effects of Grazing by Sea History of Afarine Anin~als, IvlcGrarv-Hill Book Com- Urchins Strotlgyloce~ftrotus spp., on Benthic Algal pany, London. Populations, Lhnnol. Oceattogr., 14: 710-719. Mann, K. H., 1973, Seaweeds: Their Productivity and I'atmisano, J. F., 1975, Sea Otter Predation: Its Role in Strategy for Growth, Science, 182: 975-981. Rocky Intertidal Community Structure at Amchitka -, 1972a, Ecological Energetics of the Seaweed Zone in a and Other Aleutian Islands, Ph. D. Thesis, University of Marine Bay on the Atlantic Coast of Canada. I. Washington. :ological I~~teractioitsI?~uolvi~~g the Sea Otter 567

Pianka, E. R., 1970, On r- and K-Selection, Amer. Natur., Sutherland, J. P., 1974, Multiple Stable Points in Natural 104: 592-597. Communities,Amer. A'ntur., 108: 859-873. Porter, J. N., 1972, Predation by Acnntlrnrter and Its Effect U. S. Department of Commerce, Coast and Geodetic on Coral Species Diversity, Amer. Nntur., 106: Survey, 1968, TSde Tables, West Comt North and South 487-492. America, 1969, Superintendent of Documents, U. S. Government Printing Office, \\'asltington, D. C. Po\rers, H. A., R. R. Coats, and \\'.H. Nelson, 1960, \'adas, R. L., 1968, The Ecology of Agarum and the Kelp Geology arid Submnrit~ePl~ysiogeogroplry of Amchitka Bed Community, Ph. D. Thesis, University of JVashing- Island, Alaska, U. S. Geological Survey, Bulletin ton. 1028-P, pp. 521-554. Vance, R. R., 1973, On Reproductive Strategies in hfarine Preston, F. W., 1962, The Canonical Distribution of Benthic Invertebrates, Amer. Natur., 107: 339-352. Commonness and Rarity: Part I, Ecology, 43: 185-215. N'alker, F. T., 1947, Sublittoral Seaweed Sunreg, J. Ecol., Quast, J. C., 1968, Observations on the Food of Kelp Bed 35: 166-185. Fishes, California Department of Fish and Game, Fish. -, and \\I. D. Richardson, 1955, An Ecological Investiga- Btrll., 139: 109-142. tion of Lambtarin cloustot~iEDA.1 (L. hyperborea Fosl.) Ricketts, 15. \\I., and J. Calvin, 1968, Between Pacific Tides, Around Scotland, J. Ecol., 43: 26-38. 4th ed., revised by J. \\'. Hedgpeth, Stanford University Weinmann, F. C., 1969, Amchitka Bioenvironmental Pro- Press, Stanford, Calif. gram. Aspects of Benthic 'Iarine Algal Ecology at Sanders, H. L., 1968, hlarine Benthic Diversity: A Compar- Amchitka Island, Alaska, USAEC Report Bhll-171-115, ative Study,Amer. Natur., 102: 243-282. Battelle b'le~norialInstitute. - 1968, Aspects of Benthic Marine Algal Ecology at Schoener, T. 1971, Theory of Feeding Strategies, An~zu. I\'., Amchitka Island, Alaska, hI. S. Thesis, University of Rev. Ecol. Syst., 2: 369-404. JVashington. Sekora, P., 1973, Aleutian Islands Wilderness Study Report. White, C. hl., \\I. B. Emison, and F. S. L. Williamson, 1971, A Study Conducted as a Partial Requirement of the Dynamics of Raptor Populations on Amchitka Island, \\'ilderness Act by the Bureau of Sport Fisheries and Alaska,BioScie~tce,21: 623-626. \\'ildlife. Wild, P. \V., and J. A. Ames, 1974, A Report on the Sea Shotwell, J. A,, 1950, The Vertical Zonation of Acnlnen, Otter Etihydra lutris I., in California, California Depart- the Limpet, Ecology, 31: 647-649. ment of Fish and Game, Marine Resources Technical Simherloff, D. S., and E. 0. \\'ilson, 1970, Experimental Report No. 20. Zoogeography of Islands, A Two Year Record of N'ilke, F., 1957, Food of Sea Otters and Harbor Seals at Colonization, Ecology, 51: 934-937. Amchitka Island,]. Il'ildlife djanage., 21: 241-242. Simenstad, C. A,, 1971, Tl~eFeeding Ecology of the Rock Williantson, F. S. I.., Mr. B. Emison, and C. M. Ii'hite, 1972, Greenling, Hexagrammos lagocepl~alus,in the Inshore Amchitka Bioenvironmental Program. Studies of the Waters of Amchitka Island, Alaska, M. S. Thesis, Uni- Avifauna on Amchitka Island, Alaska. Annual Progress versity of \Vashington. Report, July 1970-June 1971, USAEC Report BMI- Spight, T. M., 1972, Patterns of Change in Adjacent 171-134, Batteile Memolial Institute. Populations of an Intertidal Snail, Thais la?~~ellosa, -, W. B. Emison, and C. M. \\'bite, 1971, Amchitka Ph. D. Thesis, University of Washington. Bioenvironmental Program. Studies of the Avifauna on Strathmann, R., 1974, The Spread of Sibling Larvae of Amchitka Island, Alaska. Annual Progress Report, July Sedentary Marine Invertebrates, Amer. Nohr., 108: 1969-June 1970, USAEC Report BhlI-171-131, Bat- 29-44. telle hlemorial Institute. This page intentionally left blank

Sea Otter P. R. Morrison M. Rosenrnann"' Metabolism and Institute ot Arctic .i01001, University of Alaska, Heat Economy ;":"~.k"~~~+ Arizor~aCooperative Wildlife Research Unit, University of Arizona, Tucson, Arizona

In ten 13- to 33-kg sea otters cnptured and maititaitied at hi-' "C-',was greoter for these mammals than for terrestrial Antckitka Islnnd, a basal metabolic level of 0.72 cm3 0, arctic manr~nals,but it only increased twofold on itnmer- g' hr-' was foutld in 17 ualtres measured ouer thermonee- sion in water. Il'ater infiltration into the fur aboue 25°C trnl zones of -20 to 21PC i~tair and 20 to 33°C in wnter. allowed suruiual nt even 33°C. Body tenrperature was Aueroge ualues of 1.0 and 1.2 cmVO,g' hi' nad regulated under neutral conditions at 38.1 r 0.34"C (mea~t rnaxitnutn unlues of 1.9 nitd 2.6 cfn' 0, g-' hr-' in air and standard deuiation), under colrl conditions at wnter, respectiuely, confirm the higlt food demand of 36.7 F 0.45"C, nnd under warm conditions at one-fifth to one-fourth their body weight in fish daily. 38.5 r 0.31" C with 44.1°C ns n single lethal unlue. rlIi~~i~numthermal co~tdtrcta,tcein air, <0.012 cm' 0, g'

The sea otter (Enhyjectswere enclosed directly were lost in two such early experi~nents(one each in a large water bath (60 by 60 by 120 cm) with in water and air), apparently from shock since the air space above connected to the manometer there was no other demonstrable cause. No unusual control (Fig. 1). One external circt~lationthrough a excitelnent or depression was observed, but the barium hydroxide lime canister removed C02 from animals suddenly stopped breathing. Thereafter no the air phase and a second through particle aind problems were encoiunterecl in some two dozen charcoal filteis renno\fed sediment aind osganics expcri~nentsunder conditions ranging from water from dne water phase. The temperature of the at 33°C to the equivalent of air below -70°C for latter circulation was monitored, aind heat was durations up to 8 hr. Forceful handling was neces- added or relnoved to maintain the desired T,. sary, especially in measuring body temperature, Although the volitme of the system was greatly but onr overall view of the sea otter as run increased, the iincoinpressibility of the aqueous experimental animal \\.as very favorable. phase allowed the animals to dive and swim freely Aletabolic measurements used a closed-circuit \vithout comproinising the manometric control. automatic manometric systcin that recorded the Body temperature (Tb) was measured before and tinne rcqttired to coinsunne successive aliquots of after nletabolism runs with a thermistor probe with oxygen (A,lorrison, 1951). In the basic arrangement 8- to 12-in. rectal insertions.

MANOMETER

THERMOSTAT

ICE OR HOT WATER

Fig. 1-Diagram of experimental arrangement for metabolic measurements in water. Sea Otter Illetaboliss~and Heat Ecotiomy 571 RESULTS AND DISCUSSION whose weight was 10 kg, was only 0.9 met. Of particular interest were values on river otters Basal hletabolism and the Thermoneutral Zone (Lzctra lzctm) weighing 7 to 10 kg which ranged B,letabolism was measured in air in 12 experi- rather widely from 1.4 to 2.0 niet. Accordingly, ments at Ta ranging from 23 to -20°C. Individual the B3ustelidae as a group do not sho~uan elevated experiments reflected bouts of activity with in- h,lb. tervening quiet periods (Fig. 2) which provided IVIien other marine ~iiaminalsare compared, we acceptable values for hrlb in 10 cases (averaging the do find a correspondence. Irving (1969) sum- three lowest points in each). These values showed marized earlier data showing high A'lb values of 1.7 no trend with Ta over the range of -19 to 21°C, to 2.1 met in young seals and 2.3 to 2.9 niet in which is thus \vithin the thermoneutral zone porpoises, and these values have been confirmed in (Fig. 3). The average of these values, 0.72 t; 0.06 recent studies on both groups (Han~ptonet al., cm3 0, g' h~l,lay far above the standard 1971; kliller et al., 1973; Iverson and I

The Atlb. of 0.67 to 0.72 cm3 0,- c'- lil-', with Another ~iotableexample of elevated metab- consistent values measured over a wide rauge of Ta olism is seeu in that very small mammal, the and in two media, appears to be a very reliable long-tailed shrew (Sores ci~~ereus),for ~vhichwe value, and yet it is 2.5-fold higher than the hIst for earlier proposed an explanation that may also a ~nammalof this size (i.e., at 2.5 met). To assess apply to the sea otter (A,Iorrison and Pearson, the implications of this difference, we must corn- 1946; Morrison et al., 1959). Although the two pae other species. The h'lustelidae as a family with species differ in mass by some 5000-fold and in their vigor and carnivorous mode do show a rauge metabolic rate and feeding cycle length by some 10- of B,lb. However, in the IVisconsin least weasel fold, they nay both he thought of as "can~ivorous (~l.lt~stelarixosa), for which more data are on hand, grazers" who take food frolii a restricted local the minimutn values were less than 1.2 met, and, in rmge as needed. Since the diet is largely protein, it the ferret (11.lustela putorizrs) and tayra (Eira must exceed the daily caloric requirement by the bnrbara), values belotv Mst were observed (h.10~- amo~untwasted in nonproductive meta b o1' IC con- rison and Ryser, 1976). Recently Iverson (1972) versions [Specific Dyiia~nicAction (SDA)] . Thus, reported values on a series of European mustelids out of an average daily consumpti011 of 1.6 cm3 ranging in size froin the weasel (~\.lzcstelanivalis) (at 0, g1 lif' (G met), one-fourth, or 1.5 niet, might near 100 g) to the wolverine (Gulo gulo) (at constitute SDA (Brody, 1945). If this SDA were 15 kg). Values in the latter species, \vhich is spread throughout the day, it would raise the compxable to our sea otters in weight, were 1.3 to minimum rate to the observed h,lb of 2.5 met. In 1.4 met and that in the badger (fileles meles), other words, \ire would postulate that the SDA is not expended in a metabolic bulge follo\\ring the meal as in other carnivores but rather is rationed *The term "standard metabolism" l~sefullydesignates out over the day. \\re Iiave no information as to the mean basal level for mammals of different size ho\v this might be effected. Delayed absorption according to the function hf/i\'= 3.8iV-0.2', or from the gut would meet the requirements but is I\! = 3.8W0"1,where M is in cm3 O,/hr and iV is ingrams (Brody, 1945). Basal or other metabolic values are con- hardly compatible ~vith observations of food, veniently expressed as multiples of this standard rate including undigested flesh passing within 3 hr (designated "met"), a unit that is thus independent of size. (Stullken and Kirkpatrick, 1955). HOUR OF DAY

Fig. 2-Representative metabolic experiment in air showing activity cycles and stimulation by Me-0, (4 : 1) (9 No. 9, 13.6 kg at -12°C).

T,. "C

Fig. 3-Sea otter minimum metabolic values in air, He-O,, and water. Symbols, average of three lowest values; points, lowest value. Horizontal solid line, all values (3 min) in air plus three resting in water; horizontal dashed line, six lou*est values in air and water; horizontal dotted line, standard metabolism for an werage 18-kg mammal. Diagonal lines show elevated metabolism in air (o), He-0, (a), and \\rater (e) in 13- to 16-kg animals. Sen Otter ~Vletabolistnand Heat Econonty 573

It does seem that each of these species has a voracious appetite and sttbstantial maintetia~~ce special problem in thermoregulation, the shre\v requirement. Thus Iienyon (1969) measured a because of its very small size and the sea otter consumption of 20 to 23% of the body weight of because of its cold aquatic habitat. Indeed, were it fish (0.89 kcallg) per day. This is equisaletlt to not for their high hslb, their thern~oneutralzones some 1.5 to 1.8 cm3 0, g' hf' , a value that xvould be much niore limited with critical tempera- corresponds ~\rellto our average of 1.4 c1n3 0, g' tures above their nor~nalT, rather than belo\\.. The hl-' for three anilnals s\\~imniingat 5°C \\'hen one use of SDA, representing heat that tuust be considers tlie constraint of our experimental produced and disposed of in any event, would format and tlie wastage of food during tank allo\\. thern~oregt~lationby physical means rather feeding.. than by chemical regulation.

Average Metabolism and Daily Energy Reqiiireinents Our greatest cold stress, lS°C in He-02, did not esceed the sea otter's metabolic capacity. 'fiis The h*I,, in air during ten 3- to 5-hr esperi- pelhaps is not surprising since in subsequent meuts in rather close confinement (Fig. 4) ranged esperi~nentswit11 mink (Alztsteln visoti), a much from 20 to 54% above the h.Ib (average was 0.99 smaller mustelid, we did not reach the limit of cold cm3 0, g' hr-I ). In water, \vhich allolved more tolerance until 60°Cin He-02. Although the sea freedom of movement, the average rate in 13 luns otters \\.ere rather closely confined in their metab- was higher, ranging from 25 to 125% above the h,lb olisnl chambers, sonle individuals sho\\.ed re- (average \\.as 1.21 cm3 0, " hf' ), and this niay nlarkable vigor in attacking the cage or tank lid to represent a lower lnaintel~anceie\lel for sea otters briefly sustain h,l,,,, as high as 2.8 and 2.6 cm3 in captivity. Sea otters are known for their 0, g' hf' in air and water, respectively.

1.6

1.2 t Average H20

- C -L 0.8 0 w 5

0.4 ......

0

T,. 'C

Pig. 4-Sea otter average metabolic values in air (o), Ile-0, (a), and water (0). Slopes of diagonals represent thermal conductances in cm3 0, g-' hr' "C'. iaax.' represents fur infiltrated rvith water. This A#lmas, of 10 met is comparable to that Table 2-Conductance Ratios in Sea Otters* found in mink (unpublished observations with I<. Hc Doi and L. Peyton) but is higher than values in Effect of medium Cmi,,,/c$,,, = > 1.9 H 0 air rodents, which range from 5 to 8 met (Rosenmann Cmf,./Cmin.' 2.1 and Morrison, 1974). However, in dogs, a species H, 0 air ~max,lCmax.= 1.3 comparable in size to the sea otter, A,ImaS. values air near 25 met have been consistently reported in Range of adjustment C$~,,/C~~,,,= >4.8 HO NO treadmill experiments (Young et al., 1959). This Cm;,./C,"I,, =>3.4 difference may relate to a more orderly mode of H 0 cm'ax,!/c~;,,= >2.O life in the sea otter !vithout requirements for H 0 air extreme exertion. Overall range Cm~,;/Cmin.=27.0 cH*O /c$,,= >13.8 max.' Thermal Condnctatlce *Coding as in Table 1. Follo~ving Ne\vton's law of cooling, the quotient of metabolism and the ambient-to-deep- body temperature differential (Tb-T,) has been used to describe an overall "thermal conductance." by some 10-fold. Thus the sea otter is maintained Table 1 gives such values under different condi- at close to its hflb even in icy water. Similarly in air tions in the sea otter. Since the lolvest T, (-19°C) the critical temperature lay below -20°C; so actual did not raise the metabolism above AcIb (i.e., was environmental conditions 011 land (Armstrong, above the critical temperature), we can only say Chap. 4, this volume) should never require more than Mb for maintenance. Table 1-Thermal bletabolic 1ncrelnents In contrast to air and water, the increased (Conductance) in Sea Otters conductivity of He-0, provided conditions below thermal neutrality (critical Ta of 5°C) so that a CriticalT,, Conductance, valid estimate of CInin,He = 0.023 cm3 O2 Mode* " C cm3 0,g' hi' 'C' hrl "C1 for 12- to 14-kg individuals could be - made for this medium (Fig. 4). This facilitation of Cai! <-I9 <0.012 mm. heat flow by He-Oa in the sea otters (>1.9-fold) was slightly greater than the 1.7-fold in the mink (unpublished obselvations with I<. Doi and L. Peyton). At higher Ta thermal conductance becomes H, 0 limiting for the dissipation of heat rather than for Cmax. 25 0.084 CH*O >0.165 its retention. Values for CnIa,, are estimated from max.' > 33 average metabolism for each T,. In the yarmest *Superscripts indicate medium; subscripts equilibrium experiment (T, of 1g0C), Cz,:,, was indicate minimum or maximuln limiting 0.057 cm3 O2 g1 hrl "GI . In a nonequ~l~b~ium conductance; mas.' designates conditions after experiment at T, of 22.5"C, which resulted in heat infiltration of fur with water. death, a slightly higher value of 0.070 cm3 0, tNonequilibrium experiment corrected for g1 heat storage. l11~' "C1 was calculated with correction for heat storage. In warm water (>28OC) the auimals showed a progressi\~eloss of buoyancy until only that the minimtun conductance in air (c$,,) was the head stayed above the surface, which indicated <0.012 cm3 O2 g1 11~'OC1 , a value that repre- that water had penetrated the fur. This afforded no sents only fair insulation as compared to northern inconvenience to the animals in which relaxed terrestrial mammals that bear longer fur. The sea behavior and e\.en sleep was observed and clearly otter's advantage, of course, lies in the resistance of facilitated heat transfer during the exposure. Under H 0 its pelt to penetration by water (Table 2), which these conditions Cmg,; was 0.165 cm3 O2 g' thus retains its integrity during immersion and hrl OC1 whereas at T, of 25°C with buoyance N active s\vimming. Thus ~2,:was <0.025 cm3 O2 retained, 0 was only 0.084 cm3 0, g-' %-I hr-' OC1 , although we suspect that the Mb of hi1 "C1. 0.67 cm3 O2 g-I hrl might have maintained the As overall conductance is increased at higher ani~nalat O°C (c::: of 0.018). By comparison, Ta, heat must be lost from the large flat paws that we have found that immersion of the sno\vshoe have an admirable conformation for such dispersal hare (Lepus attrericnz~ils)increased its conductance (Table 2). If the integrity of the fur is maintained, Sea Otter Metabolisnt and Heat Economy 575

more than two-thirds of the heat load must go Table 3-Body Ten~peraturesin Sea Otters According to through the paws in water at 26'C and perhaps Previous Condition four-fifths in air at 22°C. At 28°C and above, Holding pool Metabolism cl~amber survival appears only possible because of the Tb,* - infiltration of water illto the fur with a twofold Condition SD (No.) "C SI) (No.) Condition further increase in conductance, the highest values 44.1 (1) Air 23 obsei~~edbeing 7 times the minimum value in water Viacage 0.45(6)1. 38.5 0.13i7jz Air 14- 19 aud 14 times that in air. \\'e have subsequently (dried) H,O 25- 33 obseived a similar tolerance in young harbor seals via box' 0.37(10)8 38.3 (Miller et al., 1976). Direct 0.42(7) 38.1 0.25(6) Air -1 - 12 H,05-25 36.7 0.45(8)$ /\ir <-lo Body Temperature *These averages exclude animal No. 7, a grizzled female The rauge of values found in some 50 nleasure- estimated as the oldest of the group at 6 to 7 years. All her ments, including exposure to heat and cold and Tb values lay below the range of the other seven individuals several conditions of confinement, is shown in and >40 belovr their mean. tconfinement in screen holding cage with fan to dly Fig. 5. A meail Tb of 38.1°C was measured in fur for 20 to 30 min; significant vs. direct at 5% level. animals removed directly from the holding pool or $Significant vs. direct at 1% level. confinement in wooden weighing box for 10 to 15 min; not significant vs, direct.

after coilfi~iementduring a metabolic experiment at neutral Ta (Table 3). Si~iceboth circumstailces iiiclude some undefined activity and since these values represent daytime levels in a diurnal species, the basal level may be lower. This average value of 38.1°C under neutral and quiet conditioils cor- responds to that of 37.9"C by Stullken aud Kirkpatrick (1955) on two Nembutalized captives. These values fall below resting levels for smaller mustelids (i\fustela sp.) but are comparable to those from some geuera (~\.leplzitis and Taxidea) (Morrison and Ryser, 1976). This also correspoilds to an overall average of 37.8"C for "all" mammals (h'Iorrisoi1 and Ryser, 1952), although a trend toward slightly lo~verTb in aquatic mammals as compared to their terrestrial relatives has been noted (A~lorrison,1962).Stullkeil and Icirkpatrick's values on six iildividuals killed in the field varied widely from 35.0 to 40.S°C aild may reflect some range of prior activity. However, three middle values (37.5 to 37.8OC) may rcpreseut natural inactivity. Confinemei~tbefore a metabolism experiment resulted in small (0.4OC) Tb increases (Table 3), as did warm exposure during n~etabolismexperiments (25 to 33°C in water and 12 to 19°C in air). Iixving C and K~og's(1954) value on a captured sea otter is 36110 I I I 20 30 the same as our average of 38.5OC for coilfined BODY WEIGHT, kg (excited) subjects or oites uuder xvann conditions. The regularity of these values (Table 3) suggests a Fig. 5-Body temperatures in sea otters. I'revious confine- rather effective coutrol and cooling inechai~ismfor ment in metabolic chamber under different thermal condi- heat regulation. tions: Cold, Neutral, Ii'arm, and Lethal. Symbols represent animals measured directly from the pool (o), after transport The one experi~llentcarried to lethal hyper- inside holding box (u), and after drying by fan inside therinia demonstraled the wide therilial m;~rginof holding cage (0).(See Table 3.) 6OC mailable to this species. hloreover, the devel- 576 i\forrison, Rose~~~~la~~zt,and Estes oping stages and even the final critical Tb level did of one-fifth to one-fourth their body weight in fish not appear to impair the animal since full vigor was daily. Thertnal conductance in air, <0.012 cm3 O2 maintained \\,it11 hf,,,,,, > 8 met. This indi\ridual g-' hy1 "C' , was greater for these ma~innalsthan maintained a state of escitetnent tl~rougliouttlie for terrestrial arctic mammals, but it only increased 2-hr exposure at 22.5'C, \vith an average metabolic t\\rofold on iliimersion in water. \\'ater infiltration rate or 1.85 cm3 O2 g1 hr1. If all the heat into the fur at higher temperatures allo\ved survi\'al produced (-3.5 cmW2/g) had bee11 retained, the at even 33OC, perhaps significant in the far temperature \vould have increased some 20°C southern distribution of the species. Body tempera- rather than 5'C; so three-fourths of the heat load ture was regulated under neutral conditions at \\,as successfully dissipated. Clearly the avoidance 38.1 i 0.34"C (mean + standard deviation), under of nlaladaptive behavior (vigorous activity) \vould cold cotlditions at 36.7 & 0.45"C, and under warm have allo\ved sul\~i\ral, and another individual ac- conditions at 38.5 & 0.31 \vith 44.1°C as a single cepted a 3-lir exposure at 19'C with an average letlial value. These observations provide new bio- rate of 1.10 cm3 O2 " 1lr1 axld no Tb increase energetic insights and may allow improved tech- beyond the normal range. niques for holding and shipping these remarkable The Tb response to cold sho~vedgreater dis- creatures both for transplantation to new habitats placement averaging --1.4"C with considerable and as favored attractions in exhibition of animals. scatter. The conditions at -5 to -20°C in IIe-0, were severe, being roughly equi\ralent to -40 to -70°C in air. Ho\vever, the low Tb values appeared to be maintained, with no correlation with dura- This study was supported in part by National tion of exposure or with some range of Ta. Neither Institutes of Health grant GI\.[-10402. did the metabolism \ralues sho\v any progressive decline. Tlius, although not so precise, regulation REFERENCES to cold was very effective even at Ta below -40°C. Certainly these individuals as maintained under Brody, S., 1945, Bioenergetics and Growth, Reinhold optinla1 conditions did not co~lfirmStullken and Publisllil~gCorporation, New York. Kirkpatrick's (1955) impressioli of the sea otter as Hall, E. R., and K. R. Kelson, 1959, lllarnntnls of North a poor thermoregulator. America, The Ronald Press Company, Nerv I'ork. FIampton, I. F. G., G. C. \\'bittow, J. Szekerczes, and R. The tolerance of very warm water was an Rutherford, 1971, Heat Transfer and Body Tempera- unexpected finding in vie\\' of the normal habitat ture in the Atlantic Bottlenose Dolphin, lirrsiops of these animals, which must only rarely exceed Iruncntus, I~tt.J. Biometeorol., 15: 247-253. 5°C. Still tlie range of this species (as E~l/zyd~a Inring, I,., 1969, Tenlperature Regulation in Marine hlam- 11~t~istlereis?) (Hall and Kelson, 1959) estended as mals, in Biology of dfnrine illanrmolr, pp. 147-174, H. T. Anderson (Ed.), Academic Press, Inc., New York. far south as Baja California, and so it is of much -,and J. Krog, 1954, Body Temperature of Arctic and interest that even the northern populations have Subarctic Birds and Blammals, J. Appl. Plzys., 6: the pl~ysiologicalflexibility to tolerate the waters 667-680. of a s~n-\\'armedsubtropical lagoon. Iverson, J. A,, 1972, Basal Energy Metabolism of hlustelids, J. Comp. Pl&ysiol.,81: 341-344. -, and J. Krog, 1973, Heat Production and Body Surface hrea in Seals and Sea Otters, r\'orm. J. Zool., 21: 5 1-54. SUMMARY Kenyon, K., 1969, The Sen Otter in tlte Eastern Pacific The sea otter has a unique configuration for Ocea,~, U. S. Dcpal-tment of the Interior, Bureau of Sport Fisheries and Wildlife, North American Fauna, ther~noregulation in terms of its size, aquatic No. 68. mode, and habitat in the cold but constant Bering Miller, L. K., M. Rosenmann, and P. R. hIorrison, 1976, Sea. In ten 13- to 33-kg individuals captured and Oxygen Uptake and Temperature Regulation of Ne\\.- tnaintained at Amchitka Island under near natural born Harbor Seals (Phoca uitulina richnrdi) in Water, conditiotls at Constantine Harbor, a basal meta- Comp. Biochem. Physiol., A, 54: 105-107. bolic level of 0.72 cm3 O2 9' hr-' \\'as fou11d in Morrison, P. R., 1962, Body Telnperatures in Some Australian Alammals, 111: Cetacea (hlegaptera), Biol. 17 values measured over thermoneutral zones of Bull., 123: 154-169. -20 to 21°C in air and 20 to 33'C in \\rater. This -, 1951, An Automatic hlanon~etricKespirometer, Keu. level, more than 2.5 times the standard (average Sci. Instrum., 22: 254-267. basal) level for a mammal of this size, is unique -,and 0. P. Pearson, 1946, The Aletaholism of a alllong mustelids but conforms to the pattern seen Small hlammal, Science, 104: 287-289. in tiiarine manmals froln other orders. Average -,hi. Rosenmann, and J. A. Estes, 1974, A'letabolism and Thermoregulation in the Sea Otter, Physiol. Zool., 47: values of 1.0 to 1.2 c1n3 0, llrl , lnaxinii~m 218-229. valucs of 1.9 and 2.6 cm3 0, K' hi' m' air and -,and F. A. Ryser, 1976, Body Temperature in water, respectively, confirnl the high food denland Alustclids, in preparation. Sea Otter Metabolis~na~zcl Heat Ecorto?n)l 577

-, and P'. A. Ryser, 1952, Weight and Body Temperature S~nallHomeotherms by He-0, , Amer. J. Pltysiol., 226: in hlammals, Scietice, 116: 231-232. 490-495. -, F. A. Ryser, and A. R. Dawe, 1959, Studies on the Sttillken, D. E., and C. hI. Kirkpatrick, 1955, Physiological Phvsiolow of the Masked Shrew. Sores cinereus. Investigation of Captisity hlortality in the Sea Otter ~hjsiol.yool., 32: 256.271. (Enhydm lutris), Trans. ,\Forth Amer. Ilrildlife ~\'olur. Oritsland, N., and K. Ronald, Energetics of the Free Diving Res. Confi, 20: 476.494, Harp Seal (Pagophilss groenlandicus), Rapp. P.- I< Rdu,~.Cons. Int. Expl. ~Mer.,169: 451-454. Young, D. R., R. hfosher, P. Erve, and H. Spector, 1959, Rosenmann, I. and ~P.R. Morrison, 1974, hlaximum Energy hfetaholism and Gas Exchange During Treadmill Oxygei~Consutnption and Heat Loss Facilitation in Running in Dogs,]. Appl. Pl~ysiol.,14: 834-838. This page intentionally left blank