Geographic Setting M. I,. Merritt Sandia Laboratories, Albuquerque, New Mexico
Amchitka Island lies at latitude 51.5ON and longi- map, the 1 : 250,000 scale Rat Islands Sheet. This tude 17g0E. It is one of the Rat Island Group of sheet is the result of a reconnaissance survey during the Aleutian Islands, Alaska, which comprise the the wartime years and is not satisfactory for more emergent part of a long submarine ridge connecting than the most general use. Another regional map North America and Asia and separating the Bering generally available is the Sectional Aeronautical Sea fro111 the North Pacific Ocean (Fig. 1). Am- Chalt of the \\'estern Aleutian Islands issued by the chitka is allnost the souther~llnostAleutian Island, National Oceanic and Atmospheric Administratioll only nearby Amatignak being farther south. It is National Ocean Survey in 1 : 500,000 scale. thus almost the southernmost point in Alaska. The bathymetry of the surrounding seas is Amchitka is 2160 km (1340 statute miles) shown in the Environmental Science Sewices Ad- west-south~vest of Anchorage, Alaska, and 4000 mi~listration (ESSA)/Coast and Geodetic Survey km (2500 miles) west-north~vestof Seattle, \\'ash. map sheet Kiska 1910l\i-l of the 1 : 400,000 series It is almost half the way from North America to "Bathymetry for the Aleutian Arc," reprinted Asia, being 1400 km (870 miles) east of Petro- from ESSA I\,lonograph No. 3. This map sheet pavlovsk, Icamchatka, U.S.S.R., and 1230 km includes the whole of the Rat Islands from 175 to (765 miles) west of the tip of the Alaskan main- 180"E longitude. Inshore bathymetry is best land at False Pass. sl~o\vnin a set of four maps prepared by Doering- Atuchitka and all the Aleutian Islands except a feld, Amuedo, and Ivey of Denver, Colo. (now fen, of the esternmost ones are in the Aleutian Amuedo and Ivey, Consulting Geologists). These Islands National I\'ildlife Refuge. All the investiga- maps are the East Cape, iklakarius Bay, Chitka tions reported in this volume except for a few (e.g., Cove, and Bird Cape sections, A~nchitka.Copyright sea otter and oceanogaphic studies) were carried of these four maps remains rvith Amuedo and Ivey. out on Amchitka or in the nearby sea. Two Coast and Geodetic Survey Navigational Charts cover Amchitka and vicinity. These are MAPS Chart No. 8864, Rat Islands (Semisopochnoi to Buldir Islands) at 1 : 300,000 scale; and Harbor The U. S. Army Map Service has prepared good Chart No. 9123, Constantine Harbor, Amchitka, at topographic maps of the island in 1 : 25,000 scale 1 : 10,000 scale. as follo\vs: Rat Islands sheets 2123 I\' N\\', 2023 I Finally, the U. S. Geological Survey has pre- NE, 2023 I N\\', 2024 I1 S\\', 2024 111 SE, 2024 I11 pared a new Amchitka Island Quadrangle at NE, 2024 111 N\\', 1924 11 NE, in series Q801 [also 1 : 100,000 scale for this volume. It has contour designated, in the same order: Rat Islands (A-2) intervals on land of 100 ft and at sea of 25 ft to a I,(A-3) NE, (A-3) N\\r, (B-3) S,(B-4) SE, depth of 300 ft and 100 ft below that. This map is (0.4) NE, (B-4) N\\', and (B-5) NE]. These maps based on the U. S. Army hrfap Service and the were prepared from aerial photographs taken in Amuedo and Ivey maps. A copy of this map is in September 1948 and field surveys of 1949 and the pocket at the back of this volume. have a contour interval of 50 ft. As part of the Atomic Energy Commission (AEC) program on Amchitka, neur aerial photographs were taken in POPULATION 1968. From these photographs forty-one 1 : 6,000 scale maps covering the island with a contour The region around Amchitka is sparsely inter\ral of 10 ft were prepared. These maps are no populated. In the whole Aleutian Islands Census longer available. Division, which includes part of the Aleutian The region is sho~vnin one generally available Peninsula, there were only 8057 people in 1970 U. S. Geological Survey topographic quadrangle (Bureau of the Census, 1971). All communities
Geogvnphic Setting 3 listed by that census are sliotvn in Fig. 2. The inland. A,Iuch of the island is fringed by a wave-cut nearest populated places to Amchitka were then bench of rock as much as 100 m (330 ft) wide and still are the Adak Naval Station [300 km (190 (Fig. 4). There is only one harbor on the island, miles) to the east] \\,it11 a 1970 population of 2249 Constantine Harbor on tlie Bering Sea side near and the Sliemya Air Force Base [370 km the east end, and it is so exposed that, during (230 miles) to tlie \vest] with a population of storms from the northeast, ships must leave the 1131. Not reported by the census is a small U. S. dock. South Bight on tlie Pacific Ocean side and Coast Guard contingent on Attu just west of Chitka Cove on the Bering Sea side arc sometimes Sliemya. The nearest nonmilitary point of habita- used as shelters from storms (Coast Pilot, 1954). tion is the village of Atka [450 k~n(270 miles) east] with a population of 88, all Aleuts except Tlie eastern third of the island is gently rolling the schoolteacl~er and his wife. Almost equally country characterized by many shallow ponds with distant [about 800 km (500 miles)] to the east and little or no drainage connections and streams half west are tlic tlleut village of Nikolski on U~nnak hiddcn by the seemingly ubiquitous cover of Island, population 57, and klednyi (Copper) Island vegetation (Fig. 5). In the higher central portion of in the Russian Komandorskie Islands, \vhich the island, this gives way to areas of more integrated drainage and greater wind erosion, with Hrdlic'ka (1945, p. 397) reported had a populatio~i of about 150 in 1938. Amchitka itself has not had fewer lakes and with patches of bare bedrock a permanent resident population since 1849 (Dall, rubble on ridges over 60 nl (200 ft) high caused by 1877), although it has been internlittently oc- diurnal freeze-thaw and high winds (Fig. 6). \\lest of Chitka Point the island becomes quite moun- cupied several times since then, most noticeably by tainous, reaching a maximum elevation of 354 m the military in \\'orld \\Tar I1 and for a short nrhile thereafter, rluring tlie Long Shot test, and from (1160 ft) (Fig. 7). Here vegetation is quite sparse 1967 to 1973 by the AEC (Chap. 6, this volume). except it1 protected areas, sucli as stream valleys. Finally, tlie \vesternmost 5 km (3 miles) or so is a The Aleutians lie nearly on a great circle from windswept rocky and barren plateau about 240 m Anchorage, Seattle, or San Francisco, Calif., to (800 ft) high. Japan; so daily flights and frequent ship traffic pass by within a few hundred miles. In addition, the As a result of the military occ~tpationin \\'orld seas to both sides of the Aleutians are heavily \Var I1 and the more recent AEC occupation, there fished at some seasons of the year (Chap. 15, this are numerous signs of the presence of man. About volume). Japanese or Russian tra\vlers can occa- 2000 Pacific huts, \vhicli look like and are com- sionally be seen from the island. mo. ly called Quonset huts, remain from the military days. These arc scattered seemingly at random ill the area near Constantine Harbor, half TOPOGRAPHY AND APPEARANCE buried by turf berms against tlie elements and many lying \\re11 away from the remaining dirt Anichitka lies entirely ~vithin the Alaska- roads (Fig. 8, see also Chap. 6, Fig. 7, this volume). Aleutian physiographic province (\\'ahrhaftig, Tlie military built a jeep trail that became known 1965) and, biologically speaking, is an archetype of as Infantry Road along the crest of the island from a maritime tundra regime. * South Bight on tlie cast end to tlie beach near Bird Amchitka is 65 knl (40 miles) long and 2 to Rock at the west end. Tlie AEC improved this trail 7 knl (1 to 4 miles) wide and has an area of about into an all-weather road easily driven by passenger 30,000 ha (74,000 acres). As \vitli other Aleutian Is- sedan, although it must be plowed clear of snow at lands, Amchitka has a rugged coastline witli sea times during the 5%'wtcr. cliffs and grassy slopes LIP to 30 m (100 ft) high The military built three airplane runxvays. The nearly completcly surrounding the island (Fig. 3). original one, Fox Run\vay at the head of Constan- Sand beaches on tlie island are so fc~vthat their tine Harbor, \vas trenched before they left and is rarity occasions mention of the fact in some place- unusable. Baker Run\vay, which is 10,000 ft long names. Only a few of the beaches, sucli as the one by 300 ft \vide and xvhich has numerous paved at the head of Constantine Harbor and the one at taxi\vays and hardstands, is tlie main runway. It tlie nortli~vest end of the island bet\veen t\lcut has been resurfaced by the AEC. During the AEC Point and Bird Cape, give reasonably easy access program it \\,as equipped witli full Ground Con- trolled Approach (GCA) instrumentation for land- *A helicopter pilot on the program remarked how ing in poor weather; this has been removed, but the similar Amchitka looked to the Outer Hebrides of Scotland strip retilains an excellent landing place if the where he had spent his boyl~uud. weather permits a visual approach. The third
Geographic Setting 5
Fig. 3-Coastline near Square Bay. Fig. 4-Intertidal bench in Square Bay.
Fig. 5-Coastline between Banjo and Crown Reefer Fig. 6-Patches of rubble on ridges inland west of Points. Sea Otter Paint.
Fig. 7-Mountains near Topside Creek looking east. Fig. 8-Huts left from the military occupation south of Constantine Harbor. runway, Charlie Runway, has not been maintained STUDY AREAS and was used only once during AEC occupation. The port facilities in Constantine Harbor were The three underground nuclear explosions at originally of wood but have been completely Amchitka were Long Shot in 1965, Milrow in rebuilt of steel piling and timber decking. They 1969, and Cannikin in 1971. Long Shot was sl~ouldremain in good condition for a long time sponsored by the military; its minimal biological (Pig. 9). program did not have a well-defined study area. h,lilrow and Cannikin were at the so-called B and C During the AEC occupation a main camp to sites as shown in Fig. 11. Milrow was on the Pacific house 750 people was built on a hardstand off the Ocean side of the island divide and Cannikin on the southwest end of Baker Riinwa)~(Fig. 10). At the Bering Sea side. Both are in the eastern third of the end of 1975, only three barracks and the kitchen- mess hall building remained of this. For a short island; Cannikin, at an elevation of 63 m (208 ft), time there was a sinall camp at Topside, half~vay was at the eastern edge of the region \vIlere bare rubble patches begin to appear. through the mountains, and for a number of years The areas of most intensive study on Aflilrow there was a 200-man camp at the northwest end of and Cannikin were those within 5 km (3 miles) the island; both of these have been closed down, from the sites. The island is so long that more and all buildings have been removed. distant areas could be and were used as control areas not expected to be affected by the tests. Access to these areas Teas good where the roads existed. Access to offroad points was by foot, boat, helicopter, or Snow-Trac, a light-weight tracked vehicle.
TIME ZONES
Amchitka is nearly at the 180th meridian of longitude but keeps Bering Standard and Bering Daylight Tiines (165"\\' and 150°\\' meridian times). These times are related to Universal Tiine [Greenrvich i\lean Time (GMT)] as follo\\~s:BST = GMT - 11 hr; BDT = GMT - 10 ltr; and local sun time = GAflT- 12 hr 4 min, approximately.
PLACE-NAMES Fig. 9-Port facilities in Constantine Harbor. Amchitka has been knolvn by \vestern man since the 1750s, but few of the early names survive. Bergsland (1959) has reported some of the native Aleut place-names for features on Amchitka as they were kno~vnin 1950 (Table 1). Certain of these appear to be of Russian or modern origin; for instance, the Russian word for harbor, gavan' (raeaHb), is used; and thename numbered 1023 in Table 1, which I have been unable to correlate \\,it11 any known feature, appears to be anecdotal. The \vord Amchitka itself, ho~vever,is a genuine native name reported from early in Russian tinles (Orth, 1967). Place-names used by project peoplc \\.ere to a large extent new names, invented ad lloc for the inany features that had to be kept track of and for which no previous names appeared to exist. Names given on the U. S. Army Map Service maps were used, but, in the process, few other names used Fig. 10-Main AEC camp. earlier by the military or by the refuge manage- Geographic Settiftg 7 8 hlewitt Table I-Aleut Place-Nanles on Amcltitka*
Bergsland's Aleut name Desc~.iptionor translation Modern name listing no~nber
nmciglar Amchitka Amchitka Island a~ncigla-qnl:aga- The east side of Amchitka n~neiglmn qakayari lanri,~is The islets at the east side of Amchitka Sea Otter Rocks laingas it&. Ivory bay South Bight sdali- azogis Sea lion rocks off Ivakin Point agzrlsr A stay \\'hale Cove amciglnm ga.t,a,in. Amchitka harbor Constantine Harbor ga'rnno-ya.m The point at the harbor Kirilof Point kijtlam ads. Kirilof bay Kirilof Bay ln,ms~ilrtrr The big inshore island Bat Island ga~ana-silaca The other side of the harbor Makarius Bay lflnnstrr The insltore island Rocks off St. Makarii~sPoint lijiali-ya.ra Terenty's Point Rifle Range Point lanani ala The middle of the island Duck Cove (?) yani adlogis The long reefs Not identified si.duj Irasa. Cedor's skerry njaaasiin haso Afanasy's skerry Not identified j!c.j hasn. Rolf's skel~y hanilgzrr A village and a stay Site 49-RAT-66 in Cyril Cove isuris yam Hear seals point Crown Reefer Point nna.vgir cirn~mrsga. There is a creek Falls Creek cirsnglar Partly shallorv Limpet Creek toalag A group of sea lion rocks Column Rocks cinzig The two tracks Colun~nRidge alag hnlar IIas wind from two directions Chitka Cove nlag hala-lannsa. The inshore island at alag hatar tinanas tala' Cookers' lagoon Chapel Cove alnciglaai na'n'r agio. The western end of Amchitka 1n.gslar Thirsty Aleut Point solanas Bird's place or lying towards tlte front Bird Rocks
*Data taken from K. Bergsland, Aleut Dialects of Atka and Attu, Tra~ts.Amer. Phil. Soc., 49(3): 40-41 (1959). Phonological symbols used are explained by Bergsland on p. 8 of the journal, whence this table of "segmental phonemes":
Vowels long Glides aspirated Nasals aspirated Liquids aspirated Spirants aspirated Stops and affricates
ment survived. Table 2 lists those place-names for accepted. Nevertheless, these unofficial names fcaturcs other than lakes or streams commonly served an essential purpose. used during tlle program and hence elsewhere in Table 3 gives place-names of lakes and streams. this volume. The Geological Survey map in the Almost all of these are unofficial names. The two pocket at tlle back of this volume shows tlle biological projects most concerned with lakes and location of most of these. Table 2 also indicates streams, the project on limnology of the Battelle ~vhich of these place-names predate the AEC Columbits Laboratories and the project on fresh- occupation and which ones have been submitted to water ebology of the Utah State University (re- and have received acceptance by the U. S. Board ported in Chaps. 13 and 14, respectively, this on Geographic Natnes and hence are fnlly ac- volume), each created an accounting system to ceptable and official. Not all names used were designate lakes and streams. Their designations are submitted, and not all that were siibmitted were cross-referenced in Table 3. Geograpliic Setting 9 Table 2-Place-Names on Amchitka Other Than Lakes and Streams
Bergsland's Name Status First use* listing Comment
Aleut Point 1935 USN Amchitka (Tablc continues on page 10.) Table 2-(Continued) Bergsland's Name Status First use* listing Comment Rim Point St. Makarius Bay Makarius Bay St. Makarius Islands 1019 Loran Island, off St. Makarius Point St. Makarius Point USN Sand Beacl~Cove Sandy Cove Sea Otter Point Sea Otter Rock AMS 1010 Signal Cove Fox Cove South Bight ACP 1011 Square Bay Variant: Cyril Cove Square Bluff USN Stone Beach Cove Topside Camp AMS Trident Rocks RM Column Rocks Vista Island RM WE Site RM Whale Cove t 1013 White House Cove t 1957 AMS Windy Island t 1960 USGS *USN, place-names given by a navy expedition; AMS, place-names that first appear on an Army Map Service map; IRN, place-names given by the Imperial Russian Navv: RM. olace-names a~oearinefirst in ,. . . . . v ~~~ Refuge ~aha~er'sreports;and ~CP,plsce-name; appearing first in the Alaskan Coast Pilot. (Data taken from D. J. Orth, Dictiotrary of Alaska Place Names, U. S. Geological Survey, Professional Paper 567, 1967.), fPlace-names that appear on Army Map Seivice maps. t Place-names newly accepted by the Board on Geographic Names. Table 3-Amchitka Place.Names of Lakes and Streams USU* BCLt Coordinatest Big Lake Bridge Creek Burr House Creek Cannikin Lake Chapel Cove Creek Clam Lake Clear Pond Clevenger Creek Clevenger Lake Clevenger Lake Outlet Clover Lake Deep Lake Duck Cove Creek Duck Pond Emerald Lake Falls Creek Fault Lake Fumarole Creek Fox Runway Creek GE Lake Grauman Lake Quonset Lake Heart Lake 516981 Island Pond Pumphouse Lake Jones Lake 576965 Limpet Creek 367099 Geographic Setting 11 Table 3-(Continued) USU* BCLt Coorrlinatest Little Lake Long Lake Long Shot Lake Long Shot Pond Lower Twin Lake Mason Lake Midden Creek ,Mud Pond Pratfall Lake Pumphouse Lake Island Pond Quonset Lake Range Pond Rifle Pond Rifle Range Creek Clevenger Creek Rifle Range Ponds Range, Rifle, and nearby ponds Roller Coaster Creek Round Lake Sewer Pond Long Lake Signal Cove Creek (N&S) Two streams Silver Salmon Creek (outlet) Silver Salmon Lake Pistol Lake (military name) Sixty Four Lake Strange Pond Teal Creek Tent Lake Tern Lake Tin Lake Topside Creek Twin Lakes Upper and Lower Twin Lakes Ultra Creek Upper Twin Lake Weed Lake White Alice Creek White House Cove Creek Wild Lake Quonset Lake *Notation for this lake or stream used by freshwater ecologists from Utah State University. ?Notation for this lake used by limnologists from Battelle Columbus Laboratories. $Coordinates are based on the 1000 meter Universal Transverse Mercator Grid, zone 60. First three digits are the east coordinate and the second three the north coordinate to the nearest 100 meters. Thus 303148 means E6E00, N 5714800. 5 Name newly accepted by the Board on Geographic Names. q Name appearing on Army Map Service maps REFERENCES Dall, I\'. H., 1877, On the Succession in the Shell-Heaps of the Aleutian Islands, in Tribes of the Extreme North- Bergsland, K., 1959, Aleut Dialects of Atka and Attu, west, 1V.H. Dl. Furuhelm, and George Gihbs, Trotts. Atner. Plnil. Soc., 49(3): 1.128. Contributions to North American Ethnology, Vol. 1, Bureau of the Census, 1971, U. S. Ce~tsusof Population: pp. 41-91, U. S. Geographical and Geological Suruey. 1970. Number of It~habitants-Alaska, Final Report IIrdliZka, A,, 1945, Tlce Ale18tin11nnd Cotnnzonder Islnnrls PC(1)-AS, Superintendent of Documents, U. S. Govern- a~tdTlzeirInhabitants, The lvistar Institute of Anatomy ment Printing Office, \\'ashington, D. C. and Biology, Philadelphia. Coast Pilot, 1954, United States Coast Pilot, Alaska. Cape Orth, D. J., 1967, Dictionary of Alaska PlaceNames, U. S. Spencer to Arctic Ocenn, sixth (1954) edition, U. S. Geological Survey, Professional Paper 567. Department of Commerce, Coast and Geodetic Survey, Wahrhaftig, Clyde, 1965, Pltysiograplric Diuisio~fsof Alaskn, Serial No. 779. U. S. Geological Survey, Professional Paper 482. This page intentionally left blank Geologic History Leonard M. Garcl, Jr. U. S. Geological Survey, Denver, Colorado Atnchitka Island is one of the most southerly islnnrls hl the atpresevt and trees mere common. This uolcanic complex is Aleutian Islafrd nrc. Itr genernl tertns its geologic history is probnbly geaetical1)l related to one of the dlioceae platons. pobably represetttatiue of ntost of the Aletctin~iIslands. Dtcrhtg Pliocene time minor shallow basaltic and horn- Three distb~ctuolcnnic episodes hove bee11 recognized and blende ondesitic it~trirsionsoccurred. In enrly Pleistocene dated by fossil evidence nnd by radiometric dating. The time the island wos still aboue sen leuel, the clitnnte was oldest of these is early Tertiary, probably Eocette, and is somewhat warmer tl~artnow, and trees were pwsertt. During represented by altered basic to intermediate subntarirze glacial and interglacial episodes, sea leuel fluctuaterl. Durbtg lauas a~tduolcanic debris. A'est, coarse turbidites nttd other glocinl periods terraces were carved as much as 91 rn below sediments of basaltic debris shed from t~enrbyvolcanoes sea leuel, and deposits of two episodes of gl-lnciatioir are were deposited on the sea floor during late Eocene or early preserved. Durhtg interglncial thnes fliglrts of marine ter- Oligocene time. After n period of uplift, tilting, faulting. races were cttt nboue present sea leuel, atrd 111nritiedeposits erosion, aud ittvarion by small dioritic plutons in mid- on the islnnd are co~~elatedwith two of them. Tel~sio?~al Tertiary time, o subnerial volcanic conzples of andesitic fnulth~g continued into IVisconsinat~ tin~e;however, the composition was constructed during the i\fiocene. At that island appears to hnue been tectonicnlly stable during tlte time the isla~dprobably stood higl~erabove sea leuel than post ZOO0 to 4000 years. ~Iultidisciplinary studies by the U. S. Geological range; the Aleutian Islands are the exposed peaks Sul~,ey(USGS) conducted on and near Amchitka of this range. The Aleutian arc is one of many Island from 1964 to 1972 have contributed to a island arc systetns that border the Pacific Ocean better understanding of the geologic, hydrologic, basin. and tectonic environment of the island. These Fornvation of the Aleutian trench began during investigations are the ouly comprehensive earth- the Tertiary Period. The central and western parts science studies made in the Aleutians aud consti- of the trench are believed to have developed in the tute the first intensive geologic study of the area easlier half of the Tertiary (Hamilton, 1967), while since reconnaissance mapping by the USGS during the eastern end may not have been formed until the 1940s and 1950s (e.g., Powers, Coats, and Pliocene time (von Huene and Shor, 1969). els son, 1960). A geologic map of A~nchitkaIsland Unpublished marine geophysical data supplied is being compiled at a scale ~f 1 : 48,000. by the Lamont-Doherty Geological Obselvatory This chapter su~n~narizesthe geologic history suggest that the arcuate Bowers Ridge lying north of Amchitka Island as determined from these of Amchitka Island may be part of an older studies and provides a brief description of the inactive strilctural arc. Accordingly, Amchitka lies Aleutian Island arc. near the intersection of the Aleutian Ridge and the Bowers Ridge arcs (Anderson, 1971) (Fig. 1). The Aleutian Ridge can be divided into a series GEOLOGIC SETTING of physiographically similar blocks (Fig. 1) sepa- The Aleutian Are rated by transverse canyons of probable structural origin (Anderson, 1971). Each block has a silnilar Amchitka Island lies in the western part of the cross section: north and south slopes wid1 a wide Aleutian Island arc-one of tlle most striking flattened crest topped by peaks, sonle of which rise physiographic features on the face of the earth. above the ocean surface as islands. The south slope The arc consists of a curving submarine trench as in many, but not all, places flattens to a broad deep as 7600 m, which extends 3200 km from the terrace that conlmonly has closed depressions on Gulf of Alaska across the north Pacific to it. Each of these physiographic blocks appears to I EXPLANATION >- (L Banjo Point Formation 2 (L +W Amchitko Formation ~ok,pihw /avos and breccias of Kirilof Point Too, older breccios I Contact Fault Bar ond boll on downthrown side .c Drill-hole sile A A' TC F D TC C Tb 8 Tb Tb level Seo level , , .. . - ...... I500 M 4 -1500 M TOO' ~ok' Tool Too' \~ok 0 I0 20 NAUTICAL MILES 0- 10 20 KILOMETERS " Pig. 2-Geologic sketch ]nap and cross section of Atnchitka Island. tensional stress. According to A~idcrson (1971, slo\\4y southward from a relatively fixed source of p 1, these tensional features must reflect \rolcanism and is being rolled under itself along the 'I. . . igneous-related tectonisin including disteiision zone of subduction (Anderson, 1971, p. 17). Rc- of the surficial skin over rising and spreading cent geologic mapping on Amchitka (Fig. 2) has epizonal plutons, solcano-tectonic subsidence, and supplelnented and modified the original study by rifting. . . ." Most of the faults crossing Amcliitka Po\vers, Coats, and Nelson (1960). The new work trend east-northeast aud probably first moved in inclitded geologic mapping, isotopic dating, deep late Oligocene or h,Iiocene time before eruption of drilling, geophysical surveys, and hydrologic inves- the Miocene Cliitka Point lavas. Alo\~eineiit has tigations. contiuued on some faults into late Pleistocene This current work (Carr, Quinli\,an, and Gard, time; however, tliere has been no significant fault 1970; Carr et al., 1971) has defined and dated movement on Amchitka in Holocene time. rocks of three general episodes of volcanism: the Geologic mapping in the Aleutiaus indicates tlmchitka Formation, consisting of breccias of that the oldest rocks lie on the southern part of the principally intcrmediate composition, monolitho- ridge and the youngest occur in a narrow zone of logic glassy breccias and pillow lavas, and local active volcanoes along the northern edge. klelting tuffaceous beds nlostly of submarine origin, which and mixiup of hyperfusible materials along the is overlain by the Banjo Point Formation, consist- subduction zone beneath the arc pro\,ided the ing of breccias aud minor pillo\v lavas that are volcanic material tliat forms the ridge (Coats, mostly of basaltic coniposition, which, in turn, is 1962; Anderson, 1971). The north-to-south pro- overlain nonconformably by the Chitka Point gression from younger to older rocks aloiig the Fonnation, consisting of subaerid lava flows, flow ridge suggests tliat the ridge may hc migrating breccias, pyroclastic rocks, and minor tnffaceous sediments and co~~glomerates,all of andesitic com- part by a lo\\, saddle, Fumarole Valley, little nlore position. than 100 m above sea level. Intense hydrothermal Faulting began in mid-Tertiary time and con- activity in this area altered the rocks, allo~ving tinued intermittently along certain faults into late them to be easily eroded. Solifluction is exccptiotl- Pleistocene time. X~Iost faults exposed or inter- ally active there today. preted on Amchitka strike generally northeast- Northwest of the mountain segment, a high ward, dip steeply northwest, and tend to repeat the partly dissected plateau, ranging in altitude from stratigraphic section. Althouglt they are normal 214 to 275 m, extends to Square Bluff. Except for faults, slickenside directions indicate that there has this bluff, which is underlain by andesitic lava been some component of horizontal movement flolvs, the surface of the plateau is composed of along many of the faults. deeply weathered bydrothennally altered pyro- clastic rocks. Scattered well-rounded cobbles on Square Bluff (Powers, Coats, and Nelson, 1960, p. 525) are probably either remnants of conglomer- Amchitka Island, in the west central part of the ate of the Chitka Point Formation or, more likely, Aleutian arc, lies just west of the 180° meridian remnants of ont~vashfrom glaciers covering the and just north of 51°N latitude. With the excep- plateau. Andesite erratics as large as 1.5 m in tion of Amatignak Island, about 100 km to the diameter scattered on the plateau surface indicate east, it is the most southerly of the Aleutian that ice moved north\vestward from the mountain Islands. segment. Amchitka is about 65 km long and 5 to 8 km There is no reason to believe that the high wide. It is roughly boot shaped and probably owes plateau was formed by marine planation. Except its shape, in part, to faulting. Except for the for the rounded cobbles 011 Square Bluff, evidence southeastern foot, it trends northwest, paralleli~lg is lacking which would suggest that relative sea the trend of the Aleutian Ridge in the Rat block. level during either the Pliocene or the Pleistocene The island is paralleled on the north by the steep Epochs was ever 214 to 275 nl higher than at slope of the Rat Islands escarpment and on the present. Indeed, a subaerial basalt flow at Cyril south by the Alei~tianSlope, \vhich descends to the Cove, presumed to be of Pliocene age, suggests that trench. The Rat Islands escarpment descends relative Pliocene sea level was at or below present steeply from a submerged Pleistocene terrace at sea level. Altiplanation offers a simpler explana- -91 to -100 m to a maximum depth of -1500 m tion for the origin of the surface in attributing it to about 5 km northeast of the island. The escarp- solifluction and related mass movement, ~vhich ment bounds the south side of an area of graben- tend to produce flat surfaces at high elevations and like closed basins on the sea bottom north of latitudes. Amchitka. A smoothly rounded surface south of Chitka The southeastern foot end of the islaud paral- Point lies at the same altitude, 152 to 214 m, as lels and is delineated by east-northeast-tret~ding Chitka Point Bluff and a small flattish area north normal faults that form grabens at Constantine of Windy Island (Fig. 3). The origin of this surface Harbor and at South Bight. is probably also altiplanation. \Veakly cemented The island can be divided iuto various land- iron-stained unfossiliferous gravel and sand, which forms: a mountain segment, a high plateau, the crop out on the surface south of Chitka Point, I Chitka Point segment, lower plateaus, and an believe to be part of the Chitka Point Formation intertidal bench (Powers, Coats, and Nelson, rather than a littoral-zone deposit as suggested by 1960). Powers, Coats, and Nelson (1960, p. 541). The The moin~tainsegment, an erosional remnant mountain segment stands above the high plateau of the Chitka Point volcauo, lies between Chitka and Chitka Point segments because, in the moun- Point and Windy Island. It is composed of a tain segment, the Chitka Point Formation is sinuous ridge that rises as high as 366 1n above sea con~posedmainly of lava flows containing only level and roughly parallels the island axis. Extend- subordinate pyroclastic rocks, whereas the areas ing north and south from this ridge are several steep that have been beveled by altiplanation are mainly U-shaped valleys separated by prominent ridges. pyroclastic rocks with subordinate lava flows. Some of these ridges represent reversal of topogra- The lower plateaus, especially prominent on phy; they are formed of valley-filling lava flows the eastern half of the island (Figs. 3 and 4), are a that resisted erosion better than the itnderlying flight of Pleistocene interglacial terraces discussed pyroclastic rocks of the Chitka Point Formation. later in this chapter. Several landslides interrupt To the southeast an outlier of the mountain these interglacial marine terraces along the Bering segment, 300 m high, is separated from the main coast. The largest of these lies just east of Chitka Geologic Histoly 17 *-a :!2 z rn m Y p 3 m 0 -2.c 9 2 & $8Y 57rn - .Q" -'j ... +.9 < 2-0 2 2 3 0, u ln 3 4 8 3 - -c .-g: +; Egw, - e 0 ,o.*w2 .. 0 *- rr 0 a ::$ " S@E: L %mo u0 2%: c O lo; - ma0 -m bb2$ 0 G VER SALMON LAKE - T 0 1 2 MILES -0- 1 2 3 KILOMETERS Solid lines indicate bases of. abandoned sea cliffs. dashed where indistinct or absent. Altitudes converted from feet to nearest whole meter. I I I Fig. 4-Map showing distribution of marine terraces, eastern end of Amchitka Island. Altitudes from Holmes and Narver, Inc., topogaphic map, 1 : 6000, contour interval, 10 ft. Geologic History 19 Pig. 5-Landslide near Chitka Point on Bering coast. The slide, about 1 km2 in area, is in conglomerate, tuffs, and breccias of the Chitka Point k'onnation. Note disruption of interglacial marine terrace on the left side of the picture. Point and is about 1 km2 in area (Fig. 5). Its the sea cliffs (Pourers, Coats, and Nelson, 1960) unmodified appearance suggests that it may be and of abrasion by stones on the ramp betureen the quite young. All have developed in poorly consoli- actively retreating cliff aid the bench rather than dated conglomerates and laliaric breccias of the of direct \\,axre action. The frost~rivellrubble is Chitka Point Formation and could have been removed by storm \va\res. Streruns cascading over lubricated by sea-level fluctuations during the the sea cliff locally have prevented freezing and Pleistocene or merely by seismic shaking of satu- delayed the cliff retreat (Fig. 9). rated rocks. A terrace is cu~~entlybeing developed at about - meall sea level around most of the islancl perimeter. *A turbidite is a rock or sediment layer deposited from This intertidal bench is well developed on fine- a gravity-influenced bottom-flowing current laden \\*it11 grained sedimentary rocks but poorly developed on suspended sediment, \\'I>ich movcs swiftly down a sub- coarse-grained sedimentary rocks and igneous aqueous slope and spreads horizontally on the ocea~\floor. rocks. De\~clopment of the intertidal bench on At Ainchitka these turbid currents wcrc probably initiated by eartllquake-induced sliding. The resulting dcposit is moderately coarse-grained tiul.bidites* of the Ua~ijo characterized by graded bedding; i.e., beds tvbich are Point Fo~nlatiouiis shown in Figs. 6, 7, and 8. The coarse-grained at the base, grading up\v*'ardto fine-grained at bench is inferred to be the result of frost action on the top. 20 Gard Fig. 6-Discontinuous sandstone blocks in turbidite bed of the Banjo Point Formation. Note the upward decrease in grain size of tnrbidite. Kelp-covered intertidal bench is shown in foreground. valley cut in breccia of Eocene or Oligocene Banjo Point Formation. Note fanned colunlns lying normal to the steep vallev wall at center of island. The intertidal bench at lo\\, tide shows in foreground. Geologic History 21 Fig. 8-Intertidal bench cut on breccia of Banjo Point Formation exposed at low tide. Firmly cemented erosion-resistant boulders project above bench surface. The berm at outer edge of bench is probably presewed from freeze and thaw mechatical weathering by constant \vave splash. Note the water-filled eroded joints. Fig. 9-Sea-cliff retreat caused by freeze and thaw. Bedrock cone is protected from freezing by a stream cascading over its surface. Bedrock is breccia of Banjo Point Formation. Rote lo15.-angle joint that estends froin cliff at right across bedrock cone. Near high tide covers the intertidal bench. Fig. 10-Pilloxv lava in tipper part of Amchitkn Formation 1100 m rvest of Constantine Hxbor. Devitrification of glassy rinds allorr-s diffcrcntial \veathering. The hanmer in the center of the photograph is 30 cm long. GEO1,OGIC HISTORY Eocene and Oligoce~~eEpochs The geologic history of the Aleutian Ridge is, iii a gcncral \\.ay, reflected by the history of Amchitka Island. The rocks of Anlchitlca (Fig. 2) recoi-d a long history of Tertiary volca~~icacti\,ity. The oldest rocks on i\~nchitk;l Island, the Amchitka For~nationof early Tertiary age (Carr, Quinlivan, iuld Gard, 1970), provide a local record of deposition and altelxtion of Inore than 2300 m of submarine pillo\\r lavas, Breccias, and tttrbidites of basic to intermediate composition. Toward thc end of deposition of tlic Amchitka Fol.matio11, a thick pile of monolithologic glassy l~recciasand at least two pillo~vlevas (Fig 10) \\,ere estrltded ou the sea bottom. Follo\ving tliis episode of local volcanic acti\r- ity, after only a short interval of time, the Banjo Point Forlnatio~l\\,;ts laid down during laic Eocene or early Oligocetle time. Althot~ghloc:tlly there are submarine lavas at tlie base of tlie for~iiation,the Banjo Point For~~latio~lconsists aln~ostentirely of fossilifcrous mari~ic volcenicl;~stic rocks. These rocks are mei11ly turbidites that range fro111 coarse breccias to mudstones composed of basaltic rock fragments (Fig. 11). Pig. 11-Coarse-gl-ainrd turbiditc bed in Banjo Point There is considentble evidcnce to indicate that Formation. 'l'hc bed grades from angular bonldcter~ thc turbidites nrere derived from muteri;~L shed sized clasts at base to angular pebbles and sand at top. Geologic History 23 Fig. 12-Rounded boulders of previously deposited breccia in a coarse-grained turbidite bed of the Banjo Point Fo~mation. from tlre sides of nearby ancient subaerial volca- crops out over much of the eastern half of the noes. 'll~isdebris slid do\vn the submarine flanks of island, the sectio~ihas been repeated by nortnal the volcanoes as turbiclity currents and was depos- faulting. The ~nasimumknown thickness of Banjo ited on the sea floor. Because of their density these Point Formation is 967 1x1, penetrated in an esplor- sub~narineslides lrad the ability to transport huge atory drill hole at site B (~i~.2). blocks of eroded material; yet, curiously, they occasionally \\,ere renlarkably gentle in their han- dlirlp~ --~~ of these laree blocks. Rounded blocks of '\liocene 0 -- ~ .3 displaced breccia (Fig. 12) as large as 5 111 in In mid-Tertiary time, uplift, erosion, and initia- dian~eterand sheets of sandstone 1 111 thick and tion of nornlal faulting along mainly northeast- 100 n~ long have been obser\,ed. i\3any of the trending faults occurred 011 Anichitka. Uplift ap- discontinuous sandstone sheets confornl to the pears to have occurred throughout the central and attitude of the enclosing turbidite beds but pinch \\restern Aleutian Ridge (Gates, Eraser, and Snyder, out abruptly or degenerate laterally into discontin- 1954) and \\,as acconlpanied by intrusion of uous blocks of sandstone (Fig. 6). \\ridespread epizonal plutons. A concordancc of Foraminifera1 and molluscan assenlblages potassium-argon isotopic dates of plutonic rocks found in the finer grained turbidites indicate a late from Shemya to Unalaska (Scholl, 1972) tends to Eocene or Oligocene age for the Banjo Point confirn a \videspread period of plutonism and, For~iiatio~i(Carr, Q~~inlivan,and Gard, 1970). Ilence, uplift clnring middle and late hliocene time. Because of a maior erosional unconforn~ityat the top of the formation, no complete section of the On Amchitka this intrusive episode is rcpre- Banjo Point Formation is csposcd. Although it sentetl by t~voseparate s~nallplutons, one cropping 24 Gard Fig. 13-Typical outcrop of East Cape pluton showing wide-spaced blocky jointing. out at East Cape and the other at White House series of cones, is believed to have stretched from Cove. The East Cape pluton (Fig. 13), which varies the middle of Amchitka northwestward to the in composition from gabhro to biotite granite, eastern part of Rat Island (Carr et al., 1971) and invaded rocks of the Amchitka Formation. Contact may have heen 17 km wide. metamorphish has locally changed \\,all rock and Cauldro~lsubsidence associated with the Chitka roof pendants to hornfels and amphibolite gneiss. Point volca~licrocks is suggested by a prominent A potassiiun-argon date of 15.8 i 0.7 million arcitate negative ~lnagnetic anon~aly, the IVhite years was obtained from a biotite-bearing phase of House Cove anomaly (Bath et al., 1972, p. 22). the pluto11 (Carr, Quinlivan, and Gard, 1970). Subsidence is fiurther suggested by drill-hole infor- ;:The F\'hite House Cove pluton (Fig. l4), a mation at sites E and F (Fig. 2), by anomalous diorite porphyry that is exposed in only a few seismic \relocities within the postulated cauldron places, probably underlies much of the western (Engdabl, 1973), and by an arcuate fractnre third of the island and is believed to be genetically pattern (not shown in Fig. 2) north of I\'indy related to the Chitka Point Formation. The un- Island. Three potassium-argon dates of 12.4 + 1.1, dated White House Cove pluton has the same 13.2 -t 1.2, and 14.1 i 1.1 million years on Chitka negative rei~lna~ltmaglletization as the Chitka Point lava flo~vsconfirm a middle to late Miocene Point For~nationbut is different from the weakly age for the Chitka Point Formation, as indicated positive magnetization of the East Cape plutou. by a fossil-pollen assemblage (Carr, Quinlivan, and The mountainous \vesten1 third of Amchitka Gard, 1970), and thus the same age for the White Island consists of remnants of a subaerial volcano. House Cove pluton. These rocks, the Chitka Point Formation, clearly At that time sea level )\.as relatively lower-or indicate that in late Miocene time, after the earlier the ridge \\.as relatively higher-than it is today. Tertiary marine rocks of the Banjo Point Forma- Stu~npsof trees in growth position found in a tion had been uplifted, faulted, and eroded, strato- laharic breccia 23 m below sea level (Bar, Ellis, volcanoes \\,ere built on the remnants of the ridge. and Helle, 1973; Humphrey and Branstetter, 1973) The \rolcanic complex, ~vhichprobably was a linear provide evidence that relative sea level may have Geologic History 25 Fig. 14-Intrusive contact of light-colored White House Cove pluton with darker pyroclastics of the Chitka Point Formation. A geologist is standing on the fine-grained chill border. Note porphyritic texture of diorite in pluton on left. been at least 23 m lower. The site of these stumps, Pollen from a thin coaly seam identified by IS. B. however, lies near the offshore extension of the Leopold (Carr, Qi~inliva~~,and Gard, 1970) indi- Teal Creek fault, which has dropped the Chitka cates that such trees as cypress, yew, pine, alder, Point Formation down on the north side. It is not spruce, birch, poplar, hickory, and walnut were certain on which side of the fault this site lies, but present. She reported that "the sample count submarine topogvaphy suggests it is 011 the up- contains 6.5 percent of deciduous broadleaved thrown side. Irregular knobby sea-bottom topogra- trees now exotic to Alaska, and about a fifth of the phy and small offshore islands composed of sub- generic list includes forms foreign to the aerial Chitka Point columnar lavas aroulld the region . . . ." western half of the island snggcst that sea level may Allmy of the Chitka Point lava outcrops are have been relatively as much as 80 or 90 m lower long, narrow, thick remnants of valley-filling flo~vs than at present. which now form ridges such as Column Ridge The climate during Chitka Point time must (Fig. 15) and Andesite Point, suggesting that con- have been considerably less severe than the present siderable erosion was concomitant with construc- climate. Laltaric breccias, conglomerates, and tion of the volcallic complex. Resistant narrow bedded tuffs colltaitl remnants of large trees and submarine ridges extellding offshore as deep as wood fragmellts in various stages of silicification. 90 m below sea level are probably extellsions of 26 Gard Fig. 15-Column Ridge fonned by lava flow tbat filled a valley cut in pyroclastics of the Chitka Point volcanic comples. Differential erosion has removed pyroclastics and left lava flow as a ridge. Cliff is about 100 ~n high. these flo\\c remnants, Ales Island (Fig. 7) is a int~i~desrocks near the castern end of the island. Chitka Point andesite flow remnant that fills a Crosscutti~lgrelationships between a ilornblende slilall valley cut in basaltic breccia of the Banjo andesite dike and a basalt sill at St. htlakarius Point Point Formation, and a subnlarine ridge extelids indicate that the hornblende andesite intrusives arc offshore from it for a distance of 11 km. younger than the basalt. Pliocene Epoch Pleistoce~ieEpoch During and after Chitka Point volcanism, the Graben falilting Ins presel-\zed a rather remark- area of Amchitka Island appears to have remained able series of Pleistocene deposits on Amchitka above sea level. Two periods of nlillor intrusive Island. Sediments are prcscri~edat and near South activity occ~irredduring the Pliocene Epocl~.Slliall Bight, at Constantine Harbor, and along the Rifle basaltic dikes and sills crop out on the eastern half Range fault. Although the Pleistocene llistory is far of the islalid and were also found mainly intruding from completely known, these strata reveal lilore the Banjo Point Formation in the exploratory drill than has been reported so far from ally other holes. Potassium-argon dates on t\vo basalt dikes tileutian Island. Evidence is found for at least two are 8.9 f 0.G and 10.2 f 1.1 million years (Can., glaciations separated by a fossilifcrous marine Quinlivan, and Gard, 1970). If these intrusives had deposit. broken througll to the surface, nearly all c\zidence h,larine tlansgressions during interglacial tinles is now gone; ho\\~ever,one slilall outcrop of \\.hat I~mecar\aed at least four well-developed liiarilie appears to be a basalt no\\, remnant ;uld an terraces (G to 9, 15 lo 18, 37 to 49, and 67 to accolnpanying floxv breccia is located at Cyril 76 111) on Anlcllitka Island (Fig. 3), the lattcr t~vo Cove. ti swan11 of hornblende andesite dikes of which 2u.e x\r;~~-pcd,tilted, and faultccl and are Geologic History 27 correlated \\,it11 fossiliferous marine and beach stratum of logs. Four such beds have been noted in deposits. Marine regressions during glacial advances the esposed section of lake beds. The grabcn block 11;n.e cut one broacl estcnsi\~ctcrl-acc (-91 111) and mi~sthax,c been subsiding evenly during this time at least two snlaller discontinuous terraces (-46 since 110 angular iciiconformities have been ob- antl -77 111) \vhich are now belo\\, sea le\,el. Only served in the fine-grained lake beds. Evenness of the -46- and -91-111 terraces arc cstcnsi\ze enough the bedding suggests that the lake was deep enough to bc sho\\~non Fig. 3. that the bottom setliments \\.ere not tlisturbcd by Marine terraces are 1n~1c1ibetter de\~elopeclon wave action. In the upper half of the section, the eastern half of Amchitka Islantl Oeciunse pre- holvever, a slight erosion;il ~unconfor~nityoverlain Chitka Point erosion and bc\,eling hiltl already by 35 111 of cross-stratified sand and strongly reduced the rocks of earlier Tertiary age to a low iron-stained \veathcred pebble gravel records surface. On the steep flauks of most of the \vestern changed conditions. \\'Iiether this depositioll rc- part of the island, Pleistocene terraces are re- cords uplift nearby or whether it is outwash from stricted to narro\\r discontinuous ledges. an early glaciation is not known. Subsequently, the Fossil sea cliffs ant1 terraces generally arc better graben again reverted to qtuiet lacustrine conditions developed on the south sidc of Amchitka that1 on for a short time, and 3 111 Illore of carbonaceous the north. This same phenomenon holds true for siultl and silt were deposited. the Holocene tide-level bench (Fig. 8), \vhicli is The lacustrille strata in the lower part of the better de\seloped illotlg the Pacific coast than it is sea cliff co~ltainfreshxvater diatoms, pollen, car- along the Bering coast of the isl;u~cl.1 believe this is bonizecl logs, and wood fmgmeiits. Thin as11 layers mainly because the southcrn esposurc to the lox\, tell of iliter~llitte~lte111pti011s of nearby \wlcanoes. \\,inter sn11 allox\rs inore thaxving ;und freezing to Spluce, pine, and, perhaps, alder grew on f\mchitka take place on the s<)uth-facingPacific cliffs during at this time, iu~dferns dominated the local vegeta- tlie winter months \vhen the sea cliffs arc com- tion. The pollen assemblage "suggests that the monly ice coated than on the north-facing Bering average July teliiperatures were 2 or more degees cliffs, ~vhicllare continually in the shade. Possibly, C wanner than no\v in Jnly. These samples coulcl too, Pacific storms are stronger ancl generate larger, be either Pliocene or early Qtratemary age, and of Inore po~verfulwaves \vhicli ~vouldaid in Inore these all early Qt~aternuyage seems lnore likely" rapid remo\,al of frost-riven rnbble. Fraser and (I,eopold, 1970). The presence of trees suggests B;trnctt (1959, p. 240) also noted the better that this lake rnay have existed during the early clevelop~ne~ltof the modern tide-level bench on the Pleistocene before the advent of glaciation. south sidc of I EXPLANATION ( Dune sand South Bight I1 interglacial beach deposit QUATERNARY A Hydrodamalis location I Soutn Bight I interglacial marine deposit NNW Lacustrine and fluvial (1) deposit SSE FEET METERS 2001 a Amchitka formation 0 Ba) FEET 0 100 Fig. 17-Diagrammatic geologic cross section of South Bight graben, Amchitka Island. Faults are generalized in part. Vertical exaggeration, 2 X. Geologic History 29 A thin dcposit of very coarse gravel lies at w are similav to present-day temperatures of 3.9"C altitude of 65 ~n on a prominent marine terrace for February and 10.O°C for August. about 1.5 knl northwest of South Bight (Fig. 4); Remains of four marine mammal taxa have sand in this deposit contains worn and broken been found in this deposit. These are Steller's sea sponge spicules. The base of the abaudoned sea lion (Etrmetopias cf. E. jtrbata Schreber), walrus cliff of the terrace lies between 67 and 76 m above (Odobetzzcs ~ostnarus Linnaeus), whale, and present sea level (Figs. 3 and 4). I consider this Steller's sea cow (Hyd~odarnalisgigcis Zimmer- terrace to be related to the South Bight I marine mann) identified by G. E. Lewis and F. C. \Vhit- transgressioll and the gravel to be the remnant of a more, Jr. (Gard and Szabo, 1971). bouldery beach deposit of the same age. A thin The Amchitka fossil is the first recorded fossiliferous marine deposit cropping out in the discovery of Steller's sea cow remains in place in cliff at the northwest corner of Constantine Harbor Pleistocene deposits (Gard, Lewis, atld Whitmore, is tentatively correlated with the South Bight I 1972). The giant beast, now extinct, was discov- marine tra~lsgressionand is discussed in more detail ered in 1741by the German naturalist G. \V. Steller later in this chapter. along the coast of Bering Island, one of the The South Bight I marine transgression is IComandorskie Islands, the most westerly isla~ds clearly older than the overlying South Bight I1 in the Aleutian chain. The sea cow, whose closest transgression, but at present I am unable to living relatives are the manatee and dugong, prob- correlate it with marine transgressions recorded on ably had survived the onslaughts of hunting man the coast of mainland Alaska. only because the islands were itnkuown to man The South Bight I marine deposit and the until the Bering expedition was shipwrecked there underlying lacustriue deposit in the South Bight in 1741. After that it was only a matter of graben have been tilted southward and broken by a time-27 years-until these huge sluggish beasts myriad of small faults generally of less than 3-m were slaughtered to extinction. They were easy to displacement. catch and good to eat, and Russia] ships regularly stopped there for victualing. South Bight I1 Marine Transgression. After Parts of the skeleton of a young but apparently tilting and faulting, both earlier deposits were nearly full-gown individual, as well as parts of two beveled by a new eustatic higll-level stand of the others, were found in the deposits of the South sea-the South Bight I1 marine transgression. This Bight I1 marine transgression (F. C. Whitmore, Jr., high-level stand of the sea must have persisted for a and L, h'l. Gard, Jr., 1974). The Pleistocene animal considerable length of time, for it cut a widespread was toothless, as was his eighteenth-century coun- terrace rvhose sea-cliff base now lies between 37 terpart, and undoubtedly also subsisted on kelp. and 49 m above present sea level (Figs. 3 and 4). The .South Bight specimen is indistinguishable from Continued uneven subsidence of the South Bight the modern species although it was probably graben lowered and tilted the unconformity south- somewhat larger. A female specimen found at ward and preserved a 6- to 30-m-thick wedge of Bering Island was measured by Steller and found to highly fossiliferous coarse sand to boulder beach be 7.5 m (25 ft) long and was estimated to have gravel that was deposited on the unconformity weighed as much as 10 metric tons (22,000 lb) (Fig. 17). (Scheffer, 1972). Activity of nearby volcanoes is recorded by Powers, Coats, and Nelson (1960) were first to beds of pumice pebbles. Pumice fragments appar- recognize the South Bight I1 deposit as Pleistocene. ently floated to this area, were rounded by wave They speculated that it was "probably interglacial, action and were stranded on the interglacial beach. and perhaps the youngest interglacial age, inas- This beach deposit contains both vertebrate much as the deposit has not been removed by and invertebrate fossils as well as fossil coralline erosion." algae. Uranium-series dating of bone and shell Although others have suggested various pre- material yielded a preliminary age of about Illinoian ages for this deposit (Hopkins, 1967; 135,000 + 12,000 years for this deposit (Gard and Allison, 1973), mainly on the basis of faunal Szabo, 1971). Further refinement of this date assemblages, my conclusion is that these beds and indicates an age of 127,000 f 8,000 years (Szabo the 37- to 49-m terrace were constructed during and Gard, 1975). A recent study of the rich the last major interglaciation, the Sangamon. This invertebrate fauna provides evidence for paleotem- conclusion is based on the extensiveness of the peratures and paleoecology. Allison (1973) reports terrace, the freshness of the deposit, and the that the meall February water temperature prob- radiometric age date of 127,000 + 8,000 yews, ably was about 3.9"C and the mean Augitst sea which is roughly equivalent to dates accepted temperature was between 10.0 and 11.7OC; these elsewhere for the age of the Sangamon Interglacia- 30 Gnrd tion (Richmond and Obradovich, 1972; h~lesolella Sea level is believed to have been relatively et al., 1969; Szabo and Rosholt, 1969; Birkeland, stable for the past 2000 to 4000 ycals, on the basis Crandell, and Richmond, 1971). of the presence of a broad intertidal bench (Fig. 8) The ~viclcspreadnature of the terrace suggests fouud at many places around Amchitka :uld other that the high-level stand of the sea persisted longer Aleutian Islailds (Morris and Bucknam, 1972). than ~vouldhave bee11 likely during substages of Such gco~norphicevidence indicates that, despite the \\'isconsinal Stage. It is eve11 less likely that the the high seismicity of the Alcutian arc, most of the deposit is as old as pre-Illinoian age. A recent islands have renlained isostatically stable during fission-track date of 600,000 years for ash beds in this time, although abandoned sea cliffs 011 some of late I findings on Amchitka that an ice sheet of \Viscon- Finks, R. M., 1972, Queens College, Flushing, N. Y., per- sina11 and probably late IVisconsinan age covered sonal communication. Amchitka Island. Fraser, G. D., and H. F. Bamett, 1959, Geology of the Delarof and IYesternrnost Andreanof Islands, Alefctinrr Islands, Alaskn, U. S. Geological Survey, Bulletin No. REFERENCES 1028-1, pp. 211-248. Funk, J. A., 1973, The Late Quaternary Histoly of Cold Bay, Alaska, and Its Implications to the Configuration Addicott. I\'. 0.. 1972. U. S. Geoloeical- Su~vev.,. nersonal. of the Bering Land Bridge, Geol. Soc. Anrer., Abstr. communication. with Programs, 5(2): 162. Alaska Glacial M~DCommittee of the U. S. Geolaeical Gard, L. M., Jr., G. E. Lewis, and F. C. Whitmore, Jr., Sutvey (11. IV. ~oulter,D. M. Hopkins, T. N. ~arlstyom, 1972, Steller's Sea Cow in Pleistocene Interglacial T. L. PC\\,&, Clyde Wahrhaftig, and J. R. XVilliams), Beach Deposits on Amchitka, Aleutian Islands, Geol. 1965, ilIap Showing Extent of Glaciations in Alnskn, Soc. Anrer., Ball., 83(3): 867-870. U. S. Geological Survey, Miscellaneous Geologic Investi- -, and B. J. Szabo, 1971, Age of the Pleistocene Deposits gations hlap 1-415. at South Bight, Amchitka Island, Alaska, Geol. Soc. Allison, R. C., 1973, h,larine Paleoclimatology and Paleo- Amer., Abstr. with Programs, 3(7): 577. ecology of a Pleistocene Invertebrate Fauna from Gates, Olcott, G. D. Fraser, and G. L. Snyder, 1954, Amchitka Island, Aleutian Islands, Alaska, Paleogeog- Preliminary Report on the Geology of the Aleutian rophy, Paleoclimntology, Pnleoecology, 13(1): 15-48. Islands, Science, 119(3092): 446-447. Anderson, R. E., 1971, Tectonic Setthf~of Amchitka Hamilton, E. L., 1967, Marine Geology of Abyssal Plains in Isbnd, Alnska, USAEC Report USGS-474-75 (Rev. I), the Gulf of Alaska, J. Geoplzys. Res., 72(16): U. S. Geological Survey. 4189-4213. Bar. Louis., R. "I. Ellis. and "1. I-I. Welle. 1973. Fossil Tree Hopkins, D. M., 1967, Quaternary Marine Transpessions in Stumps Found In Situ on Submerged Ridge at Alaska, in The Bering Land Bridge, D. M. Hopkins Amchitka Island, Alaska, U. S., i\'at. Alar. Fish. Seru., (Ed.), pp. 47-90, Stanford University Press, Stanford, Fishery Bf~ll.,71(4): 1099-1103. Calif. Bath, G. D. \\I. J. Car, L. M. Gard, Jr., and \V. D. Humphrey, T. M., Jr., and D. L. Branstetter, 1973, Quinlivan, 1972, Interpretation of an Aeromagnetic Geological Investigation of the Bering Sea Floor Sflruey of the Anfchitkn Islnnd Area, Alnskn, U. S. Adjacent to the Cannikin Detonation Site on Amchitka Geological Survey, Professional Paper 707. Island, Alaska, USAEC Report NVO-124, Nevada Oper- Birkeland, P. I\'., D. R. Crandell, and G. hl, Richmond, ations Office. 1971, Status of Correlation of Quaternary Stratigraphic Jordan, J. N., J. F. Lander, and R. A. Black, 1965, Units in the Western Conterminous United States, Qua- Aftershocks of the 4 Feb~uary1965 Rat Island Earth- ternnry Res., l(2): 208-227. quake, Scieace, 148(3675): 1323-1325. Black, R. F., 1972, University of Connecticut, Storrs, Leopold, E. B., 1970, U. S. Geological Survey, personal Conn., personal communication. communication. Byers, F. M., Jr., 1959, Geology of Umnak nad Bogoslof Mesolella, K. J., R. K. h'Iatther\rs, I!'. S. Broecker, and D. L. Islands, Aleutian Isln~frls,fllaska, U. S. Geological Sur- Thurber, 1969, The Ash.onomical Theory of Climatic vey, Bulletin No. 1028-L, pp. 267-369 (1960). Change-Barbados Data, J. Geol., 77(3): 250-274. Car, \\I. J., L. M. Gard, G. D. Bath, and D. L. Healey, 1971, A,Ior~is, R. H., and R. C. Bucknam, 1972, Geomorphic Earth-Science Studies of a Nuclear Test Area in the Evidence of Late Holocene Vertical Stability in the Western Aleutian Islands, Alaska-An Interim Sum- Aleutian Islands, Alaska, Seismol. Soc. Anfer. Bell., mary of Results, Geol. Soc. Amer., Btcll., 82(3): 62(6): 1365.1375, 699-705. - V D. Quinlivan, and I,. h.1. Gard, 1970, Age and Powers, 1.1. A,, 1961, The Emerged Shoreline at 2-3 Meters Stratigraphic Relations of Amchitka, Banjo Point, and in the Aleutian Islands, in Pacific Island Termces- Chitka Point Formations, Amchitka Island, Alaska, in Eustatic?-(A Symposium), Zeitschr. Geonforplrol., Chn,?ges in Stratigraphic ~Vomenclntureby the U. S. Snppl.. 3: 36-38. Geologicnl Suruey, 1969, G. V. Cohee, R. G. Bates, and -, R. R. Coats, and \\I. H. Nelson, 1960, Geology and \\I. B. Wright (Eds.), pp. A16-A22, U. S. Geological Subtnarine Physiography of Atifchitkn Island, Alaska, Survey, Bulletin No. 1324-A. U. S. Geological Survey, Bulletin No. 1028-P, pp. Coats, R. R., 1962, Magma Type and Crustal Structure in 521-554. the Aleutian Arc, in The Curst of the Pacific Basin, G. Richmond, G. M.,and J. D. Obradovich, 1972, Radiometric A. hlacDonald and H. Kilno (Eds.), pp. 92-109, Amer- Correlation of Some Continental Quaternary ican Geophysical Union, Geophysical Alonograph No. 6 Deposits-A Review, American Quaternary Associa- (National Academy of Sciences-National Research tion, Abstracts, Second National Conference, University Council, Publication 1035), National Research Cooncil, of hliami, Coral Gables, Florida, Dec. 2-5, 1972, pp. l\'ashington, D. C. 47-49. - 1950, Volcnnic Activity in the Aleatiatr Arc, U. S. Schafer, J. P., 1971, Surficial Geology, in Geology of the Geological Survey, Bulletin No. 974-B, pp. 35-49. Nenr Islands, Alaska, Olcott Gates, H. A. Powers, and Engdahl, E. R., 1973, Cooperative Institi~teof Rescarclt in R. E. IVilcox, U. S. Geological Sulvey, Bulletin No. Environtnental Sciences, personal communication. 1028-U, pp. 779.812. -, 1971, Explosion Effects and Earthquakes in the Am- cllitka Island Ilegion, Science, 173(4003): 1232-1235. Scheffer, V.B., 1972, The Weight of the Steller Sea Cow, J. Everett, K. R., 1971, Composition and Genesis of the ~llammnlogy,53(4): 91 2-914. Organic Soils of Amchitka Island, Aleutian Islands, Scholl, D. I\'., 1972, U. S. Geological Survey, personal Alaska, Arctic Alphre Res. (Borrlder, Cola.), 3(1): 1-16, communication. Szaho, ll. J., and L. &I. Gard, 1975, Age of the South \.on Huene, R. E., and G. C. Shor, Jr., 1969, The Structure Bight 11 Marine Transmission at Amchitka Island, ,\leu- and Tectonic History of the Eastern Aleutian Trench, tians, Geolo~),,3(8): 457-459. Ceol. Soc. A~iier.,Btcll., 80(10): 1889-1902. -, and J. N. Rosholt, 1969, Uranium-Series Dating of \\'illis, D. E., et al., 1972, Final Report-Seismological and Pleistocene Molluscan Shells from Southem Cali- Related Effects of the Cannikin Underground Nilclear fornia-An Open System hlodel, J. Ceoplr),~.Res., Esplosion, USAEC Report COO-2138-9, University of 74(12): 3253-3260. \Visconsin. H~~PO~OUWilliam W. Dudley, h: Tilbur C. Ballance V. Mo Clanzrnan U. S. Geological Survey, Lakewood, Colorado il,,rchitkn Isln,rd is cor~~poserlof stratified uolcnrric rocks depthr of nrore tlrmt 1000 nr where it then ?>roueslnternlly thnt unry zvidely hl hyrl,'nztlic properties. Becnrlse of the low n,td rrpwnrd dong on interfnce with snltlunter to discltarge interstitin1 per,?reabilitg i~rmost of the rocks, gronndrunter at the ocenrr floor. ,,roues ,nost nctiuely in tlre upper feu Itandred meters Tile collnpse of the cavity produced by tlte Ca~znikin iohwe fmctrcres irr tlre rocks nre nttr,re~orrsrrtzd ,,tore open. nuclenr esplosios created nn icrrsnt~lrntedrabble chii~rire), figether iuitlt /Ire tltick srirficinl ,,rnr~tle of uegelntio,> ntrd rind n,r nren ofsnrfafnce suhsirleiice >LOWfilled with ruater n~rl pent, the shnllo~onqaifers conrprire n gro~r~rrlzunterreservoir krrolun ns Contrikin Lake. Both surface ra~loff aiirl flolu that resf>onr/s strofigly to infiltratioa of precipitntion. from the shnllozu groandcunter system ruere cnptrrred for llorus of azost stren~nsrind leuels of i,tn,t), of the lakes are ,>lore then n yeor to fill tlre chisr~rey voids and the lnke sustnitzed during rlry periods by disclrnrgefio~ntlre slzallo~u basi>r. Tlrc lnke nrrd rubble clrimrrey erny /roue alelost groa,rdmnter spstetfr. Uii.ect surface rrnroff ofprecipitntio,r dozcbled the rlepth of fresh ~/ori~~dlunterflozu, nllozuing occrirs jkeqrrerrtly, hozoeue,; ntrd the qunlity of tlre srrrfnce cii.c~ilntio,t iuto the rubhlelfilled explosion cnuity. If the [onlei. dlcri,rg tlrese periods is i,flrre,rced by snlt spray front hydrnrrlic properties rleter~,riiied by testing in the esplor- the oceans. "tor), hole nt this site are sii~zilai. to tlrosc alorlg n H)~drorrlictests nnd tevrpem twe eteosr 1 7S000' 1 7S030' Bird Cape BERING SEA - 51-30' - PACIFIC OCEAN C hole and hole UA-1 LONG SHOT rite MILROW GZ EXPLANATION ALASKA A Gaging station 0 Site and hale designation Hale UA-I-HTH-1 Kilometers I 17S000' 17CJ030' Fig. I-Hydrologic monitoring points and site locations, Amchitka Isla~d. the drainage basin. Average basin altitudes and streamflow sustained by groundwater discharge drainage areas \vere.determined from topographic (Figs. 3 and 4). Figure 3 sholc~sthe base flow of maps. Results of the coinparison, using discharge Falls Creek gradually declining during May and records for calendar years 1969 and 1970 (U.S. June 1968, a period of little precipitation, as water Geological Survey, 1970, 1971, and 1972) at in the surficial materials and underlying bedrock Bridge Creek, Falls Creek, and Lintpet Creek was depleted. The increases in discharge in July (Fig. I), are as follo\vs: and August are small in relation to the correspond- ing increases in total volu~lle of precipitation, Drainage Average Dist~.iboted which is estimated by multiplying the daily precipi- basin altitnde, nl annual runoff, mm 'tation at the airport (the only point at ~vhich Bridge Creek 38.4 598 Falls Creek 74.4 949 Limpet Crcek 88.7 1198 Figure 2 shows these results. The average aitnual precipitation at tlte airport of 953 nlm (37.5 in.) for alntost 5 years of record (Armstrong, Chap. 4, this volume) is also plotted against the altitude of the gage. These data indicate that distributed runoff, and thus precipitation, does 20 I I I I I I I increase with altitude. Furthermore, the compari- 500 600 700 800 900 1000 1100 1200 1300 AVERAGE ANNUAL DISTRIBUTED RUNOFF son of precipitation at the airport ~vith the AND PRECIPITATION. nlm runoff-altitude relationship in Fig. 2 indicates that most of the precipitation results in surface-\\rater Fig. 2-Distributed annual stream runoff (January 1969-December 1970) and average annual precipi- runoff. tation at the airport (October 1967-June 1972) Stream-discharge records collected during dry versus altitude. Airport precipitation from Arm- periods and between storm periods sho~va base strong, Chap. 2, this volume. Fig. 3-Discharge and estimated precipitation in Falls Creek drainage basin, April through September 1968. October I November j December / January February March Fig. 4-Discharge and estin~atedprecipitation in Falls Creek drainage basin, October 1968 through h~larch1969. 38 Dl~dley,Ballattce, and Glat~rma?~ long-term data \\'ere collected) by the area of the lost all their water through evaporatio~iand infil- Falls Creek drainage basin. The difference indicates tration. Through this period some deeper lakes, that the suidicial materials and underlying bedrock such as Pumphouse Lake, Long Lake, and Joiies had to be resatitrated by recharge from precipita- Lake (Fig. I), which occupy bedrock depressions tion before discharge increased in September. and are apparently fed by perennial springs, contin- Prccipitntion and discharge (shown in Figs. 3 and ued to discharge water to streams. 4) may not altvays coi~lcidebecause of the 22-km distance between the precipitation and strea~nflo\c~ Canriikin Lake. Cannikin Lake (Fig. 5) is gages. unique among the many lakes on the island E 646 6W m E 646 2W m E 646 800 m E 647 400 m rn GAGING STATION m ' m L I I I I Univerral Transverse Mereator Proiection IUTMI 600-m grid ficr, zone 60 Fig. 5-Cannikin Lake and IVhite Alice Creek drainage basin, hlodified from Go~~zalezet at. (1974). Lakes because it fills the surface depression that resulted from the collapse of the cavity formed by the Hundreds of small shallo\v lakes dot the land- Cannikin nuclear explosion. Figure 6 is a photo- scape on Amchitka, especially on the nlarine graph of the lake taken in May 1974. This lake is terraces on the eastern two-thirds of the island. one of the largest and the deepest on Amchitka. It These lakes range in width from less than 30 111 to has a surface area of 12 ha (30 acres), a depth of lnore than 100 m and in depth from less than 1 m about 10 nl (33 ft) at its deepest point, and a to more than 3 m. Lakes in the mountainous areas volume of 4 x lo5 m3 (1.4 x lo7 ft3) (Gonzalez, occupy bedrock depressions; lakes in the southeast \Vollitz, and Brethaucr, 1974). part of the island are more numerous, however, The Cannikin nuclear device was detonated at a and most are confined entirely by the thick depth of 1791 nl belo~vland surface on Nov. 6, vegetation and peat. These lakes occupy depres- 1971. The explosio~lcavity, wliiclt collapsed 38 hr sions in the peat and have bottoms composed of after detonation, fonned a near-vertical rubble organic material, silt, and clay of low permeability. chimney extending to the land surface. Tlte col- During rainy seasons water flolvs from the lakes lapse at the surface formed a closed depression and intermittent springs, but during dry seasons ~vithassociated fractures and faults centered about many lakes and springs become d~y.he121y and June 360 ni southeast of ground zero. (For a discussion 1968 was a dry period during ~vhichmany lakes of the asymmetry of the collapse, see Xlorris, Gard, Pig. 6-Aerial view of Cannikin Lake. and Snyder, 1972.) This closed depression cap- containing less than 200 mg of dissolved solids per tured surface-water discharge from 84% of the liter. Sodium (Nat) and chloride (C1-) arc the surrounding (\\kite Alice Creek) drainage area dominant cation and anion, respectively. In terms (Gonzalez ct al., 1974). of reacting-value percent (100 times milliequiva- The rubble chimney created a hydraulic sink, lents of ion per liter divided by total miliiequi\~a- alld groundrvater flolved toward and infiltrated the lents of cations or anions per liter), Na' consti- chimney. The decline in groundwater level around tutes about 70% of the catiolis and CI- about 52% the site also induced infiltration of surface water. of the anions. The relative collcentratiolis of both Except for snvall ponds, no visible filling of the of these ions are greater than those generally foiu~~d surface depression occurred for more than in cotitillental surfacc water or precipitation, which 10 months after the chimney collapsed. indicates that thc quality of precipitation on The closed subsidence depression began to fill Amchitka Island is strongly influenced by salt in September 1972, which indicates that the spray from the nearby Pacific Ocean and Bering infiltration rate changed significalltly that month. Sea. Gonzalez et al. (1974) deduced that the rate Chemical chxactcristics of water from sources change \\,as caused by the saturation of the rubble on Amchitka Island and various other sources of chimney. By December 1972 the surface depres- interest are given in Table 1 for comparison, sion had filled and had begun to overflow into the Examples of the chemical character of precipita- lo\ver reach of \ilhite Alice Creek and thence to the tion in continental areas are given for rainfall in Bering Sea. After Cannikin Lake had filled, dis- North Carolina and Virginia (A in Table 1) and for charge records (Fig. 7) show that \\'bite Alice Ion-conductance water from Europcan Alpine Creek again responded llormally to precipitation, lake (B in Table l)..TIle data for A and B contrast which also indicates that the rubble chimney \\,as lnarkedly with those for oceanic \later (C in saturated. Table 1) and, in terms of the ratio Naf/C1-, with those for precipitation in the Pacific Ocean (D in Chemical Quality of Surface I\'ater Table 1). The Pacific Ocean samples were collected The water in streams, fakes, and springs on during September 1961 on the flying bridge of the Amchitka Island is of excellent quality, generally U.S. Coast and Geodetic Sui7rey ship Pioneer 40 Dudley, Bnlla~~ce,alld Gln~~za~a~z October 1971 November 1971 Fig. 7-Discharge at the \irI,ite i\li< 1973. between latitudes 25'25% to 51°45' N and longi- GROUNDWATER tudes 177"02'\\' to 177°31'\\'. Several samples of precipitation on Amchitka Groundwater investigations on Amchitka Is- Island were taken by collecting drainage from the land were conducted to provide informatioll for roof of the airport hanger. Values of the ratio assessing the feasibility of milling emplacement Naf/C1- for these samples ranged from about 0.9 to chambers for nuclear devices and for evaluating the 1.3, \vhicll possibly indicates variability in salt- possibility of migration of radioactive contami- spray influeilce under differing wind conditions. nants away from potential sites. The illvestigation Sample suites E, F, al~dG ill Table 1 have Na* included reconrtaissatlce surface study of hydro- and C1- contents (in terms of percent reacting logic features (springs, seeps, and perennial lakes values) that are intermediate between colltents of and streatns), long-term monitoring of Ivater levels these ions in continental precipitation and in in test wells and other available holes, and detailed oceanic water. 'l'he ratios Na'/Cl- for averages of testing of deep (1000 to 2134 111) exploratory these samples of Amchitka surface water do not holes at actual and potential explosioil sites. shoxv a clear similarity to oceanic water. In Fig. 8, ho\\~c\~er,Nat/C1- ratios are plotted against specific Terminology and i\leasurement of Parameters conductances for individual samples. Although the The configuration of the water table is basic to data are scattered, it is evident that sainples having hydrologic in\restigations. For most practical pur- lolver conductance, i.c., those having dissolved less poses, the \vater table is the surface below which mineral matter from the island, have Naf/Cl-ratios all openings in the soil or rock are saturated with that closely approach the ratio for oceanic \vatel: water. Because capillary rise may cause fine-pined This strongly indicates that salt spray does signifi- materials above the water table to be saturated cantly influence the chemical character of precipi- with water at less tlralr atmospheric pressure, the tation on Amchitka Island. \\rater table is best detected in the subs~~rfaceby hydro log)^ 41 Table 1-Chen~ical Characteristics of \irater Samples from Alnchitka Island and Selected Other Sources Percent of eationst Percent of anionst Sample Po- hlag- Sodium/ sourcc So- tas- ne- Cal- Chlo- Bicar- Sul- Specific chlo- Number of and dioy sium sium cium ride bonate fate cortduc- ride samples reference* (Na) (K*) (81g2+) (cazf) (CI) (HCOj) (SO:) tance$ (N~+/cI) avcraged Inland precipitation A 34.2 3.9 16.2 45.7 22.8 12.9 64.3 Alpine lake B e28.2s e2.0 7.4 62.4 18.4 65.4 16.2 e45 1.5 Unknown Oceans C 77.2 1.6 17.8 3.4 91.1 0.4 8.5 Pacific Ocean precipitation D Amchitka Island E (streams) 64.2 2.3 14.7 18.8 55.1 36.5 8.4 F (lakes) 81.0 2.7 9.7 6.6 49.7 41.0 9.3 G (springs) 71.5 3.9 14.3 10.3 48.5 41.5 10.0 *A, average values for 1-year period in five-basin area of North Carolina and Virginia; data from Gambell and Fisher (1966, p. K33). B, European Alpine lake; data from Chehotarev (1955, p. 43). C, data from Ballance and Beetem (1972, p. 10). D, Rain samples collected in the Pacific Ocean on board the USCGS ship Pioneer; data collected by F. H. Rainwater and reported by Gambell and Pisher (1966, p. K25). E, F, and G, data from Ballance and Beetem (1972, pp. 10-12). tlOO times ratio of reacting value of individual ion to total reacting values of cations or anions. iMicromhos per centimeter at 25O~. Se, estimated. 400 I I I I I I I Static head, or simply head as usually applied o ID - in groundwater hydraulics, is a measure of poten- & " 350 -Oceanic water tial energy at any point within the groundwater Z 1 N~+/cI-= 0.85 0 0 body. It is the sum of the hydrostatic pressure (the 2°z3~- DA A - 3 .. height of a column of water that the pressure at 0 - - - the point will support) and the elevation of the 0"E 250 OO@ 00 0 'j given point above a datum, usually chosen as sea 0 2 0 00 E $200- 00 0 - level. Rapidly flowing water, a rarity in ground- " 0 0, Streams W water, lnay have sufficient kinetic energy to o A 0, Laker a 150- A, A, Springs - require the addition of the velocity head to obtain total head. Head measurements at Atnchitka were 1W 6 made during the testing of drill holes by isolating SODIUMICHLORIDE, AS RATIO OF REACTING VALUES intervals at de~thwit21 airs of ltvdraulicallv inflated packing elements (straddle packers) and Fig. 8-Relation of the ratio N~'/cI- to specific conductance of stream, lake, and spring \vater on measuring water levels in tubing open to the zone Amchitka Island. bettveen packers. The hydrostatic conlponent of the head is affected by the density of the water column in the tubing. To provide comparability the standing water level in drill holes rather than among measurements at great11 depth, one must by laboratory measurements of saturation in soil or normalize the apparent hydrostatic heads by cor- rock samples. Springs, streams, and lakes that recting for the influence of the temperature and persist during prolonged clry periods usually indi- salinity on the density of the water column. In this cate intersections of the water table with the land chapter head is expressed as altitude (in meters) surface and ground\vater discharge to the surface- above sea level. water feature. The configuration of the water table Groundwater flo\vs in response to differences defines the gravity-driven potential field that in head within the three-dimensional potential causes groundwater flow in the saturated zone. field. The hydraulic gradient is a ditnensionless qirantity that expresses change of head per unit rapidly through the effective porosity (n,), the length of flow path. It is a ~neasureof the rate of particle velocity (vp) of water for use in predicting consumption of energ! ~.equiredto move water contaminant velocity (ignoring retardation by ad- through soil and rock. sorption) is given by The voltnne of water (at existing viscosity) per unit time passing through a unit cross-sectional area of rock under a unit gradient is the hydraulic conductivity. Permeability is the more co~nnlon term expressing essentially the same quantity, but, Shallo\\~Groundwater System in the terminology of ground\vater hydraulics, permeability is a property of the rock as detcr- Tlie materials in the upper few meters to mined wit11 (or corrected to) a fluid ha\,ing a perhaps a few hundred meters beneath the surface viscosity of 1 cetitipoise. \\Then expressed quantita- of Amchitka Island are quite permeable and, where tively, the distinction may be significant; in quali- unsaturated, are capable of accepting recharge tative expressions, ho\vever, permeability and its readily. These materials consist of tundra, soil, derivatives are the more easily used terms. peat, and fractured and meatliered volcanic rocks. Tlie tl~icknessof a rock unit multiplied by its The observed strong response of water levels in average hydraulic conductivity is termed the trans- shallow observation wells to infiltration of precipi- missivity. Under field conditions average hydraulic tation provides evidence of the ease of infiltration. conductivity is deterlni~iedfroin measured trans- Fluctuations detertnined by periodic measurements inissivity and thickness. The same episodes of in four closely spaced holes at the Long Shot site straddle-packer testing as those used in measuring (see Fig. 1) are shown in Fig. 9. static heads were used on Amchitka to determine Figure 9 also slio~vsthat, at the Long Shot site, transmissivity. Tlie usual technique was to swab heads decrease as the depth of penetration in- (remove wit11 rubber lifting elements on wire line) creases, \vhich indicates do\vn\vard flow. The great- water from the tubing after closing ports that est down~vard hydraulic gradient (about 0.05) connect the tubing to tlie interval of hole betxveen occurs between depths of 30 and 90 in, which the packers. Openi~lgthe ports caused an immedi- indicates either an increase it1 permeability or a ate drop of head in the tested interval, and the reduced vertical (increased horizontal) component subsequent recovery of head (water-level rise in the of flolv at greater depths. For the ground\vater to tubing) was analyzed by a curve-matching tech- reach points of discharge (streams, lakes, and nique (Cooper, Bredehoeft, and Papadopulos, springs for the shallow aquifers aid the coastline or 1967) to deternline transmissivity. Where a suffi- beyond for the cleepec aquifers), the initial down- cient difference between the land-surface altitude ~vard flow must, of course, change to predomi- and static head existed, similar slug-recovery tests nantly horizontal or upward flow. \\.ere perfornied after the tubing had been filled The water table is at or very near land surface wit11 water. over most of the island, as indicated by the many Transmissivity is the most iinportant parameter perennial lakes and streams aicl by water le\cels in for estimating the sustained infloxv that tvould shnllo\v holes. Surficial tuaterials at inany places occur ~vhiiemining an underground emplacement. are saturated except for about 2 months of each For the prediction of the rate of radionuclide year \\.lien there is usually less precipitation. migration, lio\vever, gradient, hydraulic conductiv- Consequently the configuration of the water table ity, and effective porosity are needed. Porosity is is essentially that of the land surface, althougl~the the proportion of the voliune of openings to the water table ltas less relief than the land. In general, total volume of soil or rock. Effective porosity thc hydraulic gradient is from the axis of the island includes only that part of interconnected total toward the coasts. Many seepagc points occur porosity s\~liich, because of its relatively higher along the coasts at the contact of turf and hydraulic conductivity, most readily trailsnlits wa- underlying rock or along rock fractures. Ho\vever, ter. Fractures in rock of loxv to lnoderatc priiuary few major springs issue fro111 the ~uiderlyingconsol- permeability most cominonly determine effective idated rocks above sea level near the shoreline. This porosity, and the tern1 fracture porosity is essen- suggests that, although the hydraulic potential tially synonymous with effective porosity for such cxists for ocean\vard movement of groiund~vater, rocks. the translnissivity of tlie rocks in the upper fetv The units of hydraulic conductivity (k), vol- hundred meters is not high enough for large rates ume per unit area per unit time per unit gradient of flow. Most groundwater beneath Amchitka (i), rednce to tlie units of velocity. Because flolv apparently moves in very local systems to discharge occurs only through the pore spaces, and most in the lakes and streams, a conclusion supported by 42 - E 2 ~41- - > U J 'c P Hole depth, 152.4 m DEC. JAN. FEO. MAR. APR. MAY JUNEJULY AUG.SEPT. OCT. NOV. DEC. JAN. FEO. MAR. APR. MAY. 1969 1970 1971 Pig. 9-U'ater levels in shallo~vholes at the Long Shot site. the earlier observation that most precipitation on varied geologic conditions provide a complex the island runs off in stream channels. framework for ground\vater movement. Because In the northtvestern part of the island, where the stratified rock units have greatly differing the slopes are steep aud there are many cliffs, soil hydraulic properties, flow in the vertical direction and peat are very thin or nonexistent. \\rater is generally retarded throughout the island. collects in small surface depressions perched above The exploratory holes at Sites B through F the water table. Some steep stream valleys have were tested both by pumping and by slug-recovery benches of accitmulated erosional debris (col- tests after s~vabbingwater from or injecting it into luvium) made up largely of angular rock fragments. zones isolated with straddle packers. Sites B and C, At lower altitudes the surface of this colluvium is the eventual locations of the Milrow and Cannikin covered by soil and a thick mat of vegetation. explosions, were tested more tl~oroughlythan the Recharge from precipitation percolates through the others and thus provide the best data on the mantle of vegetation and colluviu~n and flo~vs variation of the head and transtnissivity with do~vnslopealong the surface of the underlying less depth. The zones tested were generally about 60 m permeable volcanic rocks to issue as springs. One thick, although packer spacings were varied some- such spring supplied a military camp at Top Side what so that packers could be seatcd in suitably (Fig. 1) during World 11. smooth parts of the drill holes. Transmissivities deteril~inedfroin the packcr Hydraulic Characteristics of Deep Rocks tests vary over a range of five orders of magnitude, Deep exploratory holes ranging in depth from from about 30 m2/day in a 60-111 zone centered 1000 to 2134 m were drilled at Sites B through F about 700 m below sea level at Site C to about (Fig. 1). The holes principally penetrated volcanic 0.0003 m2/day. Such a range is expected in strati- breccias containing dikes, sills, and beds of poorly fied materials of various origins, mineralogy and sorted volcanic sandstone, siltstone, and tuff. The illechanical properties. youngest (uppermost) breccias are mostly subma- 12igure 10, tvhich is based on measurements rine turbidites consisting of coarse frag~llentsof reported by Ballance (1972, 1973), sho\vs that at volcanic materials often in a sandy to clayey Sites B and C there is a gcilcral decrease of matrix. The older (deeper) breccias generally con- transmissivity with depth, although the decrease is tain more zeolite and chlorite in the fractures and far from uniform. Transmissivities greater than cavities, which probably reduces the effective 0.1 m2/day in zones deeper than about 1000 111 porosity and permeability of these rocks. These belo\\r sea level were obseivcd commonly only at Y -2200 I l lll I I 111 I l Ill I l lll I 1111 I 0.0001 0.0004 0.001 0.004 0.01 0.04 0.1 0.4 1 4 10 40 TRANSMlSSlVlTY OF ZONE ABOUT 60 rn THICK, m2/day Fig. 10-Transmissivities determined from hydraulic tests of isolated zones in drill holes at Sites B and C. ? indicates possible bypass around packers during test, causing falsely high values for transmissivity. [Adapted from Ballance (1972; 1973).] Sites C and E. Site F is the least permeable, no Temperature profiles reported by Sass and transmissivities greater than about 0.05 mz /day Munroe (1970) were made in October 1968 at having been measured in 15 intervals tested be- Sites B, C, D, and F. The profiles for Sites B, D, tween altitudes of -21 1 and -1883 In. I-Iowever, and F (sho~vnin Fig. 11) sholv a uniform increase the interval from -486 to -1468m \\.as not of temperature with depth. The rate of increase is tested, and the test in the deepest interval (-1886 about 28OC/km, which is similar to the geother- to -1952 tn) sho~ved a transmissivity of about mal gradient in many places in the world. Recovery 0.15 m2/day. Data obtained at Site D indicated from disruption of the normal thermal environ- low transmissivities similar to those at Site F. ment by drilling at these sites was probably Results of the hydra~ilictesting point strongly to fractures (joints and faults) as the primary avenues for fluid movement. The fractures tend to close under greater lithostatic loads at depth. In addition, the de\relopment of zeolites and chlorites in fractures in the deeper rocks further impedes the mo\rement of \\rater. Thus the decrease of transmis- sivity with depth is to be expected. Low transmis- sivities measured in boreholes may be misleading in fractured rocks, ho\vever, because dense but brittle rocks penetrated by the hole may be fractured a short distance away. Temperatures in Drill Holes During drilling of the deep exploratory holes, water was used as a circulating fluid through most of the inter\~alsthat \Irere later tested l~ydra~~lically. The introduction of cold surface \\rater to the deep fornvations undoubtedly lo\vered rock tempera- tures temporarily. In addition, the drill holes Fig. 11-Temperature-depth profiles in drill holes provided open pathways for water to move from at Sites B, C, D, and F. [Modified from Sass and aquifers of higher head to those of lo\ver head. h'lunroe (1970).] Hydrology 45 completed it1 less than 1 year; only 4 months commonly referred to as the Ghyben-Ilerzberg elapsed between the completion of testing at relationship, after the tn~oindependent origilvators Site D and the temperature sluivey. of thc concept, \\I. Badon Ghyben in 1888 and However, the profile for Site C, measured B. Hemberg in 1901. In the Ghyben-Herzberg about 12 n~otltlisafter the completio~lof hydraulic relationship, a balance of head is assumed to exist testing, indicates cooler te~nperatures below at the interface of freshwater and saltxvater, the -400-m altitude and less unifornlity than do the height of saltwater from the interface to sea level profiles fro111 the other sites. This indicates that exerting the same pressure as a somewhat higher cool water from shallot\, aquifers \\,as flo~vingdown columll of less dense freshwater. If specific gravi- the hole and entering deeper rocks in which heads ties of 1.000 for fresh~\,aterand 1.025 for seawater were lower. These data demonstrate that the are assumed, this head balance u~lderstatic condi- hydraulic gradient at Site C is dotvn\vard. tions would occur at a depth beloxv sea level (Ai on Fig. 12-Schematic comparison of positions of the interface between fresltrvater and saltwater under static and dynamic conditions. (a) Simplified approximation, assuming static conditions. (b)Approximation accounting for flow and head changes with depth. Deep Circulatiol~of Freshwater Fig. 12) equal to 40 times the height of the water table (h) above sea level. The resulting idealized Beneath oceanic islands atld coastal areas, fresh lcns takes the form shol\m in Fig. 12a. ground~vaterflowing oceanward overlies saltwater that saturates openings in rocks at depth. In the In modern applications of the Ghyben- ideal case of an island undcrlaitl by homogeneous Herzberg relationship (Todd, 1959), it is recog- and isotropic material, the freshrvater body as- nized that such a lens could not be sustained under sumes a lenticular shape, bounded at its base by a static conditions. The actual lcns must extend out concave-up~vard surface (Fig. 12). As noted by under the sea floor to provide a discharge area for Todd (1959), the relationship between the altitude the fresh ground\vater (see Fig. 12b). Furtherll~ore, of the water table and the depth of the interface is as the fresh ground\vater seeps through the rocks, the liead loss (Ah) due to friction reduces tlic higher than the present water table, if the lalid pressure component of head at depth below that surface xvas at the same altitude as it is today. If calculated for hydrostatic conditions. Conse- cstimate sea level to have averaged through several quently the interface bet~veen fresh~vater and thousand years 10 m lolver than it is at present, the saltwater must occur at a shallo~verdeptli than that freshxvater head n,ould liave been about 20 m predicted by tlie simplified Gliyben-Ilerzberg rela- greater than at tlie present. Freshwater circulation tionship, Ai = -4011. Finally, tlie interface is actu- beneath central i\mcliitka might liave been as deep ally a zone of mixed fresli~sater and salt\\'ater as 2500 m. Ho\vever, Gard (Chap. 2, tliis volume) resulting from dispersion, ocean tides, seasonal indicates that sea level has been essentially at its fluctuations of the water table, and diffusion in present level for tlic past 2000 to 4000 years. response to salinity and temperature gxdients. \\'hcther the rocks beneath Anichitka are suffi- ciently impermedble or isolated from the oceans to Effects of Geology and Paleohydrology, Al- have resisted adjustment of head to a rising or though tlie distortio~iof the simplified static lens stable sea level in tlie past several thousand years is by dyna~iiicconditions is considered above, tlie unknown, but intuitively it seems unlikely. A conceptual model of fresh gonnd~vaterflolv bc- broad zone of mixing, relict saltwater in relatively neath t\mchitka Island niust be modified further to sliallo~ubut impermeable rocks, and relict fresli- take into acconnt tlie geologic fraiiie\vork as nrell water in deep impermeable rocks are among the as the hydrologic history of tlie island. coniplicatioi~s that have been caused by Do\vn\vard and up~vardflobv through stratified sea-level changes. rock sequences requires largc gradients as com- pared wit11 gradients for flo\v parallel to stratifica- Salinity of Deep Gromitl\vater. Samples of the tion in tlie Inore transnlissive units. Because tlie water swabbed from intervals isolated by straddle strata at Amchitka dip moderately (about 10' to packers in tlic exploratory holes at Sites B and C 40' south\\~est\vardnear Site C), even units that are generally increased in salinity with depth of sam- several hundred meters or more beneath the pling. Dissolved-solids concentrations for these surface at the center of the islwd probably samples, reported by Beetem et al. (1971), are intersect the sea floor. This geotiietric setting shown and compared \\it11 tlie concentratioil in i~iiplies that most of the head loss in the deep sealvater (about 35,000 mg/liter) in Fig. 13. groundwater s)~steili occurs during flo\v in the At an altitude bet\eeen about -1050m and more-resistant do\vnward direction hereas as rela- about -1150 m, the dissolved-solids concentration tively little gradient is then required to expel tlie at Site B reached tlie niaxi~iiumobsctved, about water to tlie sea through the aquifers. The result is 30,000 mglliter. Considering that quantities tliat head declines nit11 depth beneath the island at swrabbcd mere probably not sufficient to recover all a greater rate thrui \vould be calculated for a the writer that was injected during drilling, this homogeneous isotropic system. The actual depth of the fresh~vaterlens, therefore, is less than it \vould be in tlie idealized setting. Sea-level changes cause corresponding changes E level d in the position of tlie Gliyben-Herzberg interface. W r' However, tlie hydrology of Anichitka Island is not 9 known in sufficient detail to predict the length of I - ',site B I time needed to reestablish equilibrium flo~vcondi- Z , tions after significant chaiges in sea level. Gard ;; 8W (Chap. 2, this volume) indicates tliat sea level varied by probably several tens of meters as varying ainounts of water formed ice during advances and Z E -1200 Site c recessions of Pleistocene lacie~s.The last glacial 0 advance, on the basis of f C dating: of peat (Gard, "J ' 0 Chap. 2, tliis volume), apparently ended about 3 10,000 years ago \vIien sea level !\,as about 30 m II Q belo\v that at present. For an unkno~vnperiod byparr before the contiiientd glaciers had begun to melt, -2OW sea level was possibly as low as -90 m. As melting 10W 2WO 4WO IO.000 20,000 40.WO began on Anicliitka, tlie potential recharge \\,as DISSOLVED SOLIOS, rnglliter probably so great that the water table was at the Fig. 13-Dissolved-solids coticentrations of water subglacial land surface, i.e., on tlie order of 10 111 satnples from drill holes at Sites U and C. concentration indicates that the interface bet~veen ciiused by dilution of the sample froln \jrat.er salt\vater and fresh\cwter is above -1 150 m. The ~uni- bypassing the upper packer during swabbing. form concentration of about 21,000 mglliter in the Head Relationships at Sites B and C. The interval -800 to -1050 111 may indicate \crater that correction of heads nleasured in straddle-packed was either diluted by drilling fluids or \\.as drawn intervals in the exploratory holes at Sites B and C from a natural zone of mixing. These four salnples \\.as based on the salinity data in Fig. 13 and on \\,ere collected during tests conducted in a 24-hr t\vo sets of temperature data. The lllaxil~lulll period before the liole \eras deepened; the three tcmperaturcs tliat might have predominated in the samples immediately belo\\' (betr\reetl -1150 and zollcs that are tested are those reported by Sass -1525 111) \\,ere obtained during a similar period a and i\,lunroe (1970) and sho\vn in Fig. 11. The month later, whereas the deepest sample \\,as iiiiniliium teiiiperatiu~es,caused by introduction of swabbed after 3 \\reeks of additiollal drilling. There- cold drilling fluids, arc those determined ill geo- fore tlie differellcc in dissol\red-solids conccntra- physical logs run before episodes of hydraulic tion bet\veen tlie zone -800 to -1050 111 md the testing. zones below ca~inotbe considered sigiiificallt un- The uncorrected and corrected head data for less it is strongly corroborated by other data. the t\vo sites are sholvn in Fig. 14, beginning in the The maximum dissolved-solids concentration in upper right corner of each graph. The uncorrected xvater samples taken as deep as -1800 111 at Site C data sIio\\~a distinct decrease of head \\it11 depth \\.as only about 12% of the concentratiotl in tniiforiiily distributed at Site B but occurring seawater (Fig. 13). Fresh surface \\rater had been mainly above -450 111 at Site C. used in drilling the hole, but volumes of \t7ater Correcting the height of the \Irater columns siniilar to those lost in drilling \\,ere pul~lpedfrom above the zones tested at Site B to an equi\ralent each interval of liole before hydraulic testing. height of \\rater at a specific gravity of 1.000 results Additional data \\'ere obtained in another hole only in a significant increase of head because of the high 35 ni away. salinity of tlie \\rater, At an altitude of --I450 111, The Caunikin emplacement hole at Site C was the correction is about i25 111; corrections for the drilled over a 2-year period beginning near tlie end possible temperature extremes, lio\ve\~er, differ of hydraulic testing in the nearby exploratory hole. oiily by 4 in at that depth and are negative because The emplacement hole, 2.3 111 (7.5 ft) in diameter, \\rater is less dense at temperatures above 4OC. The \\,as drilled with only minor losses of the mud general do\\~nrverd gradient apparently persists circulant belo\\' an altitude of -1000 m. At the after correctio~isat Site B, although iunder both of beginning of mining a chamber at a11 altitude of the extreme temperature assinnptions the gradient about -1725 m for the Cannikin device, a large appears to be slightly up~c'ard bet\veen -700 and inflow of mud, unhardened cement slurry, and -1050 m. water forced remedial measures and hydrologic Because the \crater swabbed from intervals in monitoring. As reported by Ballance and Beetelii the Site C esplol.atory liole \\,as much less saline (1972), the inflo~vsoon cleared. The voltune of than tliat from Site B, the salinity component of water pumped from the cl~a~lbertuid attendant thc density correction is small. The temperature tunnels during the follo\ving 5 months exceeded corrections doniiilate and result in corrected heads 7000 in3. Tlie temperature of the discharge soon that are some\vhat lo\ver than uncorrected heads. stabilized at about 41°C. Tlie specific conductance The head \'alues based on the ininimum tempera- of the \\rater, about 6100 + 600 pmhos/cm (at tures arc believed to be the more realistic for zones 25'C), remained about the same throughout the belo\\( --I600 m because testing immediately fol- period. On the basis of chemical cl~aracteristicsand lo\\led the drilling of this intcr\'al, d~uringwhich temperature, Balla~lce and Beetem (1972) con- there was a large loss of fluids to the rock. The cluded that the inflow \\,as derived from a zone at gradient is small in the don~m\~arddirection, \\~llich about -1650 In altitude. The specific conductalice is consistent with the temperature profile of of the water \\'as essentially tlie same as that of Fig. 1I. sanples from the interval -1590 to -1750 m in the esplorato~~hole. Relatioil of Data to Ghybe~l-IIerzberg The- That the correlatioll of cliemical data from the ory. Earlier in this chapter the altitude at \\4lich a two holes is coincidental seeliis highly unlikely. \\'e balance of head \\rould exist bet\veen fresh~vatcr conclude that the data sho\\'n in Fig. 13 for Site C above and salt\vater below \\'as expressed as are substantially correct and that the freshx\~atcr Ai = -4011, ~vhere 11 is the corrected frcsh\\~ater lens extends to an altitude of at least -1700 111. head at the depth of the interface. Because the The anollxalously lo\\, dissol\red-solids concentra- temperature of the salt\crater iucreases \\,ith geo- tion in tlie salllple from -1775 111 is tliought to be thermal gradient (28°C/km as dctcrmincd from 48 Dudley, Ballnnce, rind Gla~zzi~ta~a +2W I I I I I of the curve expressing the corrected freshn~ater Estimated position of Sea head with the corrected Ghyben-Herzberg curve. ~SYBI Head measurements (expressed as freshwater-) be- lo\\, the depth of intersection, ho\vever, should coincide \\,it21 the corrected Ghyben-Herzberg 400 curve. As shown in Fig. 14, the head data from Sites B and C quite obviously do fall belo\\' the Ghyben-Herzberg curve. This is an enigma that -'0°- apparently can be explained only in terms of Site B inadequacy of data or test methods. Residual head dra\crdo\\~nsof 8 to 12 m in the lo\\rermost zones of - both holes, resulting from inco~nplete recovery -1200 from pumping before the sxsabbing tests, is a possible but too convenient explanation. If the inconsistency of the data from the -16W deepest zones is ignored, the head data suggest 0 +lo +20 +30 ALTITUDE OF STATIC HEAD, m Ghyben--Herzbcrg interfaces at an altitude of -1000 to -1200 m at Site B and at an altitude of 1100to -1300 m at Site C. \\'hen compared \\'it11 the salinity data (Fig. 13), this range for Site B is reasonable if we attach significance to the salinity differences between the zone -800 to -1050 m and the zones below -1100 m. Earlier (see section on Salinity of Deep Ground~vater)this difference \\'as presented as probably not significant {vithout strong corroboration. The head data do not pro- vide sufficiently strong corroboration because of their inconsistency with theoretical head relntion- ships below the freshwater lens. There is one physically plausible explanation for the salinity difference at Site B, ho>\rever. The sample at -720 m (Fig. 13) \\,as obtained from a zone that has a transmissivity of about 3 m2 /day, a relatively high value for the deep groundwater system (Fig. 10). Tests in intervals at -800 m and -980 In, in contrast, indicated transmissivities that are among the lo\\~estmeasured. Presumably fresh- water can circulate rvith relative ease only to an Fig. 14-Uncorrected and corrected static heads at altitude bet~veen -720 and -800 m. The zone SitesB and C and altitudes of Ghyben-Herzberg from -800 to -1050 m, sho~vnin Fig. 13 to interface as a function of static head. Curve 1, uncorrected static heads. Cu~ve2, static heads cor- contain water of apparently intermediate salinity, rected for salinity and maximum temperature. could reasonably be interpreted as a zone of Curve 3, static heads corrected for salinity and mixing or diffusion above the Ghyben-Herzberg minimum temperature. interface as selected by head criteria. For Site B we are left, then, with the conclu- sions that a Ghyben-Herzberg interface does exist Fig. ll), a density correction must be applied to at an altitude not Ioxver than about -1200 m, that the saltwater head also. As shown in Fig. 14, the zone from about -800 to -10501n lnay beginning at sea level the uncorrected Ghyben- represent a zone of mixing or diffusion, and that Herzberg curve expresses the relationship the data are not sufficiently precise to eliminate Ai = -4011. The corrected curve \vas conlputed uncertainties. nulnerically in 50-111 increments of depth and The data from Site C are even more perplexing. sho\ss an increasing slope with depth. At -1400-m Figure 13 sho\\rs that the pattern of salinity chauge altitude, the slope is -48, \\fillereasat -1800 m it is with depth at Site C, although the water Isas more -52. dilute, is strikingly similar to that at Site B. In theory the depth of the Ghyben-Herzberg Further~norc, in Fig. 14 intersections xvith the interface should occur at the point of intersection Ghyben-Herzberg line occur at similar depths at both sites. Finally, an apparent deficiency of head However, if exit paths from the deeper aquifers to exists below the iutersection depths at both sites. the oceans are available, .r\41ich is probable based Referring again to the discussion on salinity of on currently known geologic conditions, disruption deep ground\vater, ho\vever, the salinity of the of the vertical resistance to downward flo\v by an inflo~vto the Cannikiu chamber so strongly corrob- explosion-produced rubble chimney \vould illcrease orates the salinity data from the exploratory hole the depth of active or moderately active flonr. that we nlust conclude that the freshwater leus The introduction of a permeable vertical path extends at least as deep as an altitude of -1750 m would also decrease the head loss associated \\'it11 at Site C. It is possible to speculate that adjustment downward flow. Because of the greater residual of the salinity interface has not occurred after deep head at depth, the Ghyben-Herzberg interface flushing at a time of lolver sea level. Supportiug must also move do~vnxvard. Beneath Ca~mikin evidcuce is lacking, however. Quite possibly the Lake, \vhich now n~aintaiusa head at the surface of head data are misleading, and the salinity data +35 m (115 ft) (Gonzalez et al., 1974), the perme- indeed sho~vcirculation of groundxvater to at least able rubble chimney may allow most of this head -1750 m altitude. to be applied to an altitude of at least -1800 m. From consideration of head relationships alone, the Ghyben-IIerzberg interface could be depressed EFFECT OF EXPLOSIONS ON DEEP to below -1700 in, or well xvithin the cavity region GROUND'\\'ATER FLOIV of the Canuikin explosion. The salinity data suggest that the interface might have been that Nuclear explosions as large as Milro\v (about deep even before the Cannikin explosion. In the 1 h*Ityield) and Cannikin (approximately 5 hflt) have altitude range of -1500 to -1800 m, which significant effects on hydrologic features. A com- encompasses the cavity and lo~verrubble chimney, plete discussion of these effects is beyond the several zones were tested aud found (Fig. 10) to scope of this chapter. have trausnlissivities bet~veen0.15 and 0.7 mZ/day Disturbances of streamflolv and oscillatio~lsof for 60-m depth intervals. fluid pressure in the ground\crater system in re- Data from Site B indicate that the rubble spouse to the i\'Iilro\c~explosiou have bee11 de- cltimney from the h3ilrow explosion, which \\!as scribed by Ballance aud Dudley (1970). Gonzalez detouated at an altitude of -1178 m in October and \\'ollitz (1972) have presented similar data for 1969, opened hydraulic comlllunication to rela- the Cannikin event and have related hydrologic tively permeable zones at -700 and -1056 In response to seismic stress and motion. The thermal, altitude. The explosion cavity is xvithin rock units hydraulic, and radiochemical processes acting in l~avingtransmissivities of 0.04 m2 /day or less. the Cannikin rubble chimney are described by H. C. Claassen (nrritten communication, 1976). Groundwater Velocity The reports cited above discuss the relatively The possibility of fresh\vater circulation deep transient effects of Milro\v and Cannikin. Holvever, into nuclear-explosion sites makes it necessary to the penetratiou of the rubble chimneys deep consider the velocity with \vhicl~ groundxvater beueath the surface of Amchitka Island has dis- could flo~vand transport radionuclides from the rupted the geologic fran~eworksufficiently that explosion sites to points of discharge at the surface long-term effects on the deep £lo\\, system are a or on the sea floors. The parameters that nlust be certainty. evaluated to perform the calculations are effective The movement of water in the deep ground- porosity, hydraulic conductivity, and hydraulic water system under natural couditions is extremely gradient. Average hydraulic conductivity can be slo\e, as indicated by the following four items: calculated as the transmissivity divided by the 1. Stream ruiloff is sufficiently great to remove thickness of the iuterval tested. The data in Fig. 10 most precipitation from Amchitka Island. sho\v the average hydraulic conductivit)~of the 2. Transinissivities measured belo\\' altitudes of zones between -1500 aud -1800 111 to be about about -750 m are small as are do\\'n\vard hydraulic 3 x m/day. gradients. The hydraulic gradient from the Cauuikiu site 3. The geologic frame\vork suggests that verti- (Site C) to the Bering Sea is the head (+35 m) of cal hydraulic conductivity is extremely small. Cannikin Lake, less the head lost in do\vnward 4. The dissolved-solids concentfatio~~sin Sam- seepage through the rubble chimtiey, divided by plcs from belo\\, altitudes of -400 m at Sites B aud the leugth of the flow path from the lower rubble C are an order of maguitt~deor more gseater than chimuey to the discharge point. The probable the conce~ltrationsin surfacc \\rater and shallow flolv-path leugth is 2500 m, a value chosen from a ground\vater. model presented by Penske (1972) and consistent 50 Dlcrlley, Bnllance, n11d Gla~lz?narl \\,it11 the geologic setting. If the hydraulic conduc- The velocity range of 0.07 to 0.3 m/day, which tivity of the rubble cliin~neyis on thc order of correspond, respecti\'cly, to brecciated basalt arid 1 m/cI;~y,a value comparable to those of a unifornl moderately fractured basalt, results in estimated medium-grained sand or nioderately fractured mas- travel times of 23 to 103 years over a 2500-m flo\cr sive rocks, the loss of head in do\\~n\\.ardflo\\' path. These estimates are bascd on the assumption through 1800 111 of the rubble \\,ould be less than that the hydraulic conductivities measured at 1% of the total head available. Consequently the depth at Site C are representative of the entire flow average hydraulic gradient over the flolv path from path. Since the southwest\\rard dip of the geologic the lower rubble chimney to the Bering Sea is units provides probable intersection ~viththe sea about 35 m/2500 m, or about 0.014. floor, thereby eliminating the need for cross-bed The fracture porosity can be estimated from flolv, this ass~un~ptionis reasonable. It seems likely measured hydraulic condr~ctivityand fracture spac- that \\rater carrying the niore mobile radionuclides, ing as obser\fed in drill holes and mined chambers. such as tritium, could begin to discharge on the The most intact rocks beneath Amchitka have floor of the Bering Sea within a century and fracture spacings of several meters; 10 m call be possibly \\,ithin a fe\v decades. takc~ias an upper limit. The nleasured hydra~~lic The dol\wward flo\v in the Cannikin rubble cond~rctivityis consistent ~i~itlithat of a mode^ chimney will keep radioactivity concentrated near ately but not highly fractured massive rock; a the base of the chimney, or below about -1500 m fracture spacing of 0.5 m is a reasonable estimate altitude. On the basis of thc total transmissivity for moderately fractured rock. The minimuln measured below this depth, the 0.014 gradient average spacing-occurring in the basalt breccia ~vouldallow this region to be flushed by a flo\v of penetrated by the Site C exploratory hole (Bal- about 20 m3/day, a very lo\\. rate. lance, 1972)-is on the order of 0.05 m. Among the factors that !$rill lo\ver the concen- Sno\u (1968) related these pamieters in the tration of radioactivity in tlie water before dis- follo\\~ingequation, con\~crtcdto units in use here: charge to the sea are: 1. Radioactive decay. 2. Adsorption on rock surfaces (except tri- tium). 3. Dispersive mixing with uncontaminated \va- \\,here 11f is the fracture porosity (dimensionless), k ter in the rubblc chimney. is the hydraulic conductivity (meters per day), and 4. Spreading of the contaminant plume, both d is the fracture spacing in the drill hole (meters). by dispersive lnixing and by divergence of flow The solution of Sno\\"s equation for fracturc lines from the chimney to a prohtbly broad zone spacings of 0.05, 0.5, and 10 111 provides fracture of discharge. porosities of 6.3 x 1.4 x and Although there are i~lsufficientdata to estimate the 1.9 x lo-', respectively. Corresponding average effects of these processes, the dilution will bc particle \'elocities are 0.07 m/day for the 0.05-111 large-from a few to probably se\rcral orders of spacing, 0.3 m/day for the 0.5-m spacing, and magnitude. Further dilution by mixing of tlie 2.2 m/day for the 10-nl spacing. The n'idth of slon~lydischarging ground\vater with seaitrater ~vill openings in a cubic set of uniform fractures, be sufficient that detection of radioactivity in calculated from effective porosity and assumed above-background concentrations is considered spacings, \\rould be about 5 pm for the closest very unlikely. spacing, 23 pm for the moderatc spacing, and A sinlilar analysis of the h.Iilro\v site provides 63 p~nfor the 10-m spacing. comparable results. The moderately transmissive Snow's (1968) investigations in drill holes to zones below an altitude of 1000m, \\41ich Before depths of about 120 111 at dam sites in igneous and the explosion wcre belo\\' thc freshtvater lens, metamorphic rocks indicate that fracture openings probably are now being flushed by chimney xvatcr greater than 40 pm \vould bc unusual belo\\, 100 111 tolvard thc Pacific Ocean. Thc overall tr;msniis- but that there is only a very gradual rcdi~ctionof sivity of the zone interscctccl by the lower part of average fracturc x\~idthbelo\\, the 50-m depth. At the rubble chimney is about the same as that at depths as great as 1500 In, hotvevcr, openings Site C, ancl the available lxydraulic gradicnt is only geatcr than 20 pm must be considered as highly slightly less. improbable. It is more improbable, ho\\~ever,that the measured average hydraulic conductivity is conccntratcd in the interstices (between grains or crystals) of the basalts and basalt breccias at that flmchitka Island is drained principally by depth at Site C. streams that carry direct runoff and sustained base flow. The base flo\v is discharged to the streams the Alilrow Event, Amchitka Island, Aleutian Islands, and malty 1;ikcs from a illoclerately permeable ~\laska,USAEC Report USGS-474-71, pp. 51-65, U.S. shallo\\~groundxvater reservoir. Because the water Geological Survey. Beetem, \\I. A,, K. A. Young, C. L. \\'ashington, and I,. J. table is at or very near the laild surface oser l~lost Schroder, 1971, Che~nicalAnalyses of \\later Samples of the island, thc shallo\c. groundwater flow occurs Collected on t\n~cbitkaIsland, Alaska, USAEC Report ~vithinthe tundr;~,soil, and peat as \c~clIas \c.ithin USGS-474-135, U.S. Geological Survey. the upper few hundred meters of underlying Chebotarev, I. I., 1955, hletamorphism of Natural \\'aters in volcatlic rocks. The quality of the surface and the Crust of \\'eathering-1, Geochim. Cosmochi~n.Acta, 8.- . -77.48 - - - . near-surface \\rater is goocl. Salt spray drivell by Coats, 11. R., 1956, Recoinlnissnltce Geology of Some \\rind over the narrotv island from the adjacent Il'estern Aleutian Islands, Alaska. U.S. Geological Sur- ocean illflueilces the composition of rainfall, as vey, Bulletin 1028-E. illdicatcd by ratios of sodil~mto chloride in water Cooper, H. H., Jr., J. D. Bredehoeft,and I. S. Papadopulos, of the surface and shallow subsurface systems. 1967, Response of a IFinite-Diameter \Vell to an Instantaneous Charge of \\later, Il'ater Resorrr. Res., Geologic setting, hydraulic characteristics of 3(1): 263-269. the rocks, and the salinity of water within the deep I'enske, P. R., 1972, Event-Related IIydrology and Radio- ground\vater system provide e\ridence that ground- nuclide Transport at the Cannikin Site, Anlcbitka \eater movement under natural conditiolls is very Island, Alaska, USAEC Report NVO-1253-1, Nevada slo\v or nonexistent at the dcpths at which thc Operations Office. h4ilro\\. and Cannikin iluclear explosions were Gambell, A. \\'., and D. \V. Pisher, 1966, Clre,~ticnlCos~po- sitiolt of Rainfall hr Enstern ~Vorth Cnroli~rn nlid detonated. The L,Iilro\\r device was sited \\,ithi11 south en st err^ I7irgi,ria, U.S. Geological Survey, \\later- ilcarly static saltwater beloxv the Icns of slowly Supply Paper 1535-K. circulatiilg freshrvater $\hereas Canllikill \\,as proba- Gonzalez, D. D., and L. E. \\'ollitz, 1972, Hydrological bly within the fresh~vaterlens. Effects of the Cannikin Event, Seisnlol. Soc. Amer. Disruption of the rocks by rubble chimneys at Bull., 62(6): 1527-1542. -, L. E. \Yolti[z, and G. E;~Bretlrauer, 1974, Batbyo~etr)? both explosion sites is believcd to have increased the depth and velocity of freshwater circulation. of Cannikin Lake, Atnchitka Island, Alaska, with an Evaluation of Computer Alapping Techniqoes, USAEC Althougl~the data are sparsc, it is estimated that, Report USGS-474'.203, U.S. Geological Sulvey. within a fen, decades to perhaps a century, \\rater Morris, R. I-I., L. hl. Gard, and R. P. Snyder, 1972, Visible containing radioactivity from deep in the rubble Geological Effects of the Cannikin Event, Seis~~rol.Soc. chin~ncyswill begin seeping slolvly onto the ocean Anler. Bull., GZ(6): 1519-1526. floor, ~vhere the radioactivity will be greatly Sass, J. H., and R. J. h'llunroe, 1970, Heat Flow from Deep diluted by mixing with ocean water. Boreholes on T\vo Island Arcs, J. Geopltys. Res., 75(23): 4387-4395. Snow, D. T., 1968, Rock Fracture Spacings, Openings, and REFERENCES Porosities, N-oc. Anrer. Soc. Ciuil Errg., 94(1): 73-91. Todd. D. K.. 1959. Ground ll'ater tl~~drolov.". -lohn \\'iles P; Ballance, \ C., 1973, Hydraulic Tests in Elole UAe-2, Sons, Inc., New York. Amchitka Island, Alaska, USAEC Report USGS- U.S. Geolonical Sulvev. 1972. \\later Resources Data for 474-103, U.S. Geological Survey. t\laska.-~art 1. ~ui.face\~ater Ilecords. Part 2. !\later ,1972, Hydraulic Tests in Hole UAe-I, Amchitka Quality Records, U.S. Geological Survey Opcn-File Island, Alaska, USAEC Report USGS-474.102, U.S. Report. Geological Sunrey. -, 197l,\\'ater ResourcesData for Alaska. Part 1. Surface -,and \V. A. Beetem, 1972, \\later Inflow into Hole \\later Records. Part 2. \\later Quality Records, U.S. UA-I, Amchitka Island, Alaska, USAEC Report USGS- Geological Survey Open-File Report. 4i4-156, U.S. Geological Survey. -, 1970,\\'ater ResourcesData for ~\laska.Part 1. Surface -,and I\'. 11'. Dudley, 1970, Hydrologic Effects of the \\later Records. Part 2. \\'ater Quality Records, U.S. Milrow Event, in Geologic and Hydrologic Effects of Geological Survey Open-File Report. This page intentionally left blank Weeather and Robert H. Armstrong Air Resources I,;iborato~y-Las Vegas, National Climate Ocea~iicand At~iiosp\iericAdniinistration, u. s. Department of Commerce, Las \legas, Nevada Amchitka Islai~dhas a pronounced maritime cli- TEMPERATURE mate. The day-to-day \\.eather is marked by cliangeability because of the great frequency 114th The moderating effect of the surrounding seas \vhich migratory pressure systems pass along the is evidenced by the limited ranges of tempcrature North Pacific storm track. In the absence of at i-\mchitka (Fig. 1). The niean daily range is significant local effects, such as surface heating and 3.g°C (7'F) or less in all n~onths,and the annual nocturnal cooling, \c'hicli esert a large influence on range of the mean daily temperature is only 9.4OC weather conditions at a continental location, the (17OF). By contrast, a continental station (Saska- Aleutian weather results alnlost entirely fro111 toon, Saskatche\van) at approximately the same large-scale pressure systenis and thcir associated latitude has an anliual range of 36.1°C (65OF). eather her fronts. During the sulnlner season fog predominates as LO\\' CLOUDINESS a result of the ad\rection of relatively warm, moist air over the colder ocean surface. The air in the Clear skies arc extreniely rare at Anichitka air-sea interface layer is cooled to the saturation (Fig. 2). Fog or clouds below 2972 111 (9751 ft) point, and extensive fog results. The summcr fog above the airport elevation (67 m, 220 ft) cover often persists for days at a time. nore than half the sky most of the time. I,o\v The Air Resources Laboratory-Las Vegas ceilin%s (more tl~anhalf the sky covered) and fog (t\RL-LV) experience during relatively short pe- occur with greatest frequency during the summer riods of residence on Amchitka in connection with months. In the other three seasons, brief periods of Atomic Energy Coniniissioli (AEC) activities there clearing skies follol\r the passage of migratory bore out my hypothesis that climatological data storni systems. available from a 5-)gear period in the 1940s adequately defined the local climate and weather. During most of the period that Anlchitka was in FOG, RAIN, AND SNO\\' use by the AEC, airlvays \\,cather observations were made by airport control-tower operators during Figure 3 points up the prevalence of fog during periods of tower operations. These data were not the summer mondis. The high frequency occurs continuous, \\,ere not under ARL-LV control, and because thc seawater is colder tlian the air that were not summarized. The control-tower data did, flox\rs over it. The air is cooled, condensing its ho\\rc\rer, substantiate the validity of the earlier moisture. cliniatological clata. Except for the solar-radiation Although mean a~rnualprecipitation at Am- data, data given here are from the 5-year nearly chitka is not escessi\re, 828 mm (32.6 in.) corn- continuous weather observations of the 1940s. parcd with 991 nim (39 in.) for Seattle, precipita- Table 1 contains general clin~atologicaldata. tion in sonie form is experienced one-fifth or niore Solar radiation \\'as not illeasured during mili- of the time in all months, with the most frequent tary occupation. A pypanorneter installed at the hourly occurrence duri~lgthe swnmer months (Fig. airport control tonrer measured the solar radiation, 3). The average nuniber of days per month on and the liniited data obtained are included. \vltich prccipitation in the indicated amounts is 54 Ari>~stro~zg Table l-Cli~natoloeical Data for Amchitka Island. Alaska. from Februarv 1943 Throueh lune 1948* Jan. Feb. Mar. Apr. May June Jnly Aog. Scpt. Oct. Nov. Dec. Annual Temperature, 'C Record high Mean daily max. Daily mean Mean daily min. Record lorv Av. No. of days with a miniml~mof c0.ooc Av. No. of days with a maximum of <0.o0c Precipitation, mm Mean monthly Av. No. of days with a0.25 mm >6.35 mm >12.70 mm Percent of hours with rain Percent of hours with snow Snow depth at 0800 local standard time, cm Av. No. of days with depths of a2.54 cm >5.08 cm >15.24 cm >30.48 cm Surface wind Mean speed, m/sec Percent of hours with speed a 14 mlsec *From U. S. Department of the Air Force, no date (a) and no date (b). tLess than 0.5 day or percent. LOW overcast ...Z - Lev, broken MONTH MONTH Fig. 1-h'lonthly temperature data Fig. 2-Sky condition pressure area that persists in the Gulf of Alaska during the sttmlner months. Figure 7 shoxes the inondlly mean wind speed and percentage frequency of hours with wind speeds equal to or greater than 14 and 21 mlsec 132 and 47 miles per hoar (mph), respecti\~ely]. \\'ind speeds may exceed 45 mlsec (100 mph) in the Aleutians. 111 January 1969 the wind speed exceeded the 100-knot (51 m/sec, 115 mph) instrument scale at the Amchitka base camp. CLIMATE OF B AND C SITES In a clilnatological sense, the data for the Rain and drizzle airport [elevation, 67 m (220 ft)] are considered entirely applicable to the B site [40 m (131 ft)] and the C site [63 m (207 ft)] (see Fig. 11, Chap. 1). \\re ha\,e obsenred that over the relati\rely low JFMAMJ JASOND southeast part of the island there is no sigmificant MONTH spatial \,xiability in the large-scale wind direction Fig. ?--Fog and precipitation frcqucnc).. and speed. True, eddies may be induced by small hills, gullies, ctc., quite near the g~o~ul~din the microscale circulations, hut no instrutnentation sufficient to define these microscale circul a t'~on patterns \\,as ever used at Amchitka. observed and the mean monthly precipitation (in lnillimetcrs and inches) are sho~vtlin Fig. 4. Data for the period October 1967 through June 1972 are given for comparison tvith mean monthly precipitation data for the 1940s. The mean annual total precipitation for Febrnary 1943 through Jnne 1948 was 828 mm (32.6 in.); that for October 1967 through June 1972 was 953 mln (37.5 in.). The 15% excess in mean annual precipitation of the later data over the earlier is not considered significant \vhen one notes (from the later data) that the highest values in each month exceed the monthly means by values ranging from 20 to 107%. 111111111111 Average number of SURFACE IVINDS days per year 25 with >0.25 mrn The extreme variability of surface-\\rind direc- 20 >", tions at Amchitka is a result of the high incidence 4 of migratory pressure systems. Figure 5 sho\\'s the 0 seasonal paths of surface low-pressure centers. 10 These storm tracks are baseti on data available from 1899 to 1954. There is no pre\'ailing \vind direction as such at 0 f\mchitka, as sho\vn by Fig. 6. There is little, if JFMAMJ J ASOND any, topographic channeling of the \\rinds at the MONTH 10x1, entl of the island \\'here the data were taken. Fig. 4-Precipitation amounts and ficquenc)r of Summer winds tend to blo\v from the south~\~st occurrence. 0---0,October 1967 through June 1972. quatlrant because of the semipermanent high- @a,February 1943 througll June 1948. I 1 WINTER SPRING 180" f (Jan.) 180" IApr.1 / / 6 d 0, 'S 3 0' / a' a' / +40°N +40°N 180' 180" I I SUMMER (Julyl / / /' //- +40°N +‘lo"~ 180' 180° Fig. 5-Seasonal storm tracks, North Pacific Ocean (from Klein, 1957). Solid lines denote clearly defined (primary) tracks; broken lines denote less frequent or less well defined (secondary) tracks. In a marine environment such as An~cl~itka's, CLIMATE OF THE IIIGHLANDS one cannot conceive of significant temperature changes over such relati\~ely short distances [B is Clilnatological statistics presented here are 8 km (5 miles) and C is 16 km (10 miles) from the applicable to the lo\\, southeast end of Amchitka airport] or through a height cbailge of only 4 to for which long-term data are available. They only 27 m (13 to 89 ft), and, in the absence of approximate collditioils for the relatively nloun- pronounced terrain-feature differences, precipita- tainous areas of the island since most meteorolog- tion amounts or intensities would not be expected to ical elements are affected by elevation, slope, and vary beyond the limits of ineasurement accuracy. terrain shielding. Thus the long-term airport climatological data I11 general, one would expect the weather to be represent the best available data for the B and C more severe in the highlands. Temperatures, in the sites. absence of surface in\rersions, are usually lower WINTER SPRING (December - February) (March - Mayl N 7/14 W 81 11/21 E W 71 S S SUMMER FALL (June-Augurtl (September - November) N N 10% 12 E W 71 S S Fig. 6-Seasonal surface wind roses. The length of each direction bar represents the percent frequency of wind from that direction. Numbers at the end of each bar represent mean wind speed (meters per secondjknots). The number in the center of the rose is the percent of calms. \\,it11 increasing altitude. I,o\v clouds over the SOLAR RADIATION southeast end of the island would often be fog in the mountains. h,Ioderate rain over flat terrain is A pyranometer was i~lstalled at the airport oftcn accon~pa~~icdby heavier rain in nearby co~ltroltower in February 1969 to obtain solar mountains, and, certainly, more snowr accumulates radiation data for the bioenviro~llnentalprogsams. and remains longer at high than at 101s elevations. Fairly con~pletedata were obtained from August The effect of elevation on temperature and the 1969 through November 1971. increase in \\rind speed caused by channeling over The pyranometer measures global solar radi- rugged terrain result in more scvere storms in the ation received on a horizontal surface. This in- mountains than in the lo~vlands. cludes bod1 radiation received directly from the 30 5 20 w 0 c WL Obrewationr 10 svilh wind 0 JFMAMJ JASOND hlONTH MONTH Fig. 8-Solar-radiation data. Fig. 7-Surface-wind speeds. solid angle of the sun's disk and radiation that has accurate solar-radiation measurements. Frozen been scattered or diffusely reflected in traversing precipitation on the exposed envelope affects the atmosphere (sky radiation). The ratio of the instrument accuracy, a factor that should be sky radiation to the global solar radiation fluctu- considered during die winter months. For these ates markedly, from near unity when the sky is and other reasons, the data are not to be con- densely overcast to less than 0.1 under extremely sidered comparable 114th data from an obscrva- clear conditions. tory \\'here particular emphasis is placed on ob- Figure 8 gives maxiniuni, minimum, and aver- taining solar-radiation data. age daily totals of solar radiation at Amchitka and, for purposes of comparison, the average daily totals for Seattle (-47.5ON) and Annette Island, REFERENCES Alaska (-55%). The large differences bct\veen Klein, \Y. I]., 1957, Principal Tracks and Mean Frequencies average daily totals for Amchitka and Seattle of Cyclones and t\nticyclones in the Nortl~ernIlemi- during the summer months may be accounted for sobere. llesearcll Pa~erNo. 40. U. S. Weather Bureau. by the fact that Seattle has a niean sky cover \i1ashi&ton, D. C. ' (sunrise to sunset) of about 50% (U. S. Department U. S. Department of the Air Force. Air \\'eather Se~~rice.no of Commerce, 1968) in summcr, wvliereas fog date '(a), Swface Winds, 45702, Amchitka, Alaska, February 1943 ThroughJune 1948 (Less hlay 1947). predominates in the summcr months at Amchitka. U. S. Department of the Air Force, Air Weather Seneice, no The Amchitka data must he used o~ilyin a date (h), Uniform Summary of Surface \\'eather Obser- qualitative sensc because of several uncertainties. vations, Parts A & R, 45702, i\mcl>itka Island, Aleutian Py~a~ionictcrsin continuous operation require fre- Islands, February 1943 Through June 1948 (Less hlay quent inspection and cleaning (daily or oftener). 1947). U. S. llepartment of Commerce, Environmental Science Thcre is no assurance that this was done since care Senrices Administration, 1968, Cliniotir Atlas of the of the pyraliotneters was by control-to\ver opera- United States, Superintendent of Documents, U. S. tors whose primary concern was certainly iiot Govenlment Printing Office, \\'ashington, D. C. Prehistoric Human Allen P. Mct2artne.y- Department of Anthropology, University of Occupation of the Arkansas, fiyetteville, Arkansas Rat Islands Prehistoric Rnt Isloaders had an almost exclusiuely marine littoral zones, or inland areas. These eigltt systems provided adaptation to their islandgroup of 10 large to smnll islonds. a stable subsistence pattern that, in turn, supported n large A large part of the Aleut poprrlntion lived on Amchitka but maritinte populntion. Biotic variations between the Rnt interacted widely within the group for subsistence mtd Islnnds suggest that Amcltitka, Kiska, Rat, and Bt~ldirwere sociol rensons. the most rewarding islands to exploit. dlost previous archaeologicnl investigations in the Rat group have bee11 on Amchitko. Stress is placed in this Because of their pritnitive technology and manner of chapter ?rot on resrrlting artifnct collectiot~sbut o~rdeterrni- building, prehistoric Aleuts had only limited long-range nnnts of site locatiota and how prehistoric Aleuts subsisted. impact on the islands they occupied or on tlte fauna and Island size and sl~ape, isolation, ecologic richness, and flora they trtilized. Their intpact on the Rat Islands was uttlcanism are environment01 unriobles titought to be impor- indeed smnll when compared with that of modern mmt, tant in differentiating one island front nttother. For any one especiofly il'orld llrorIlpersonnel. islnnd, such considerations as accessibility, exDosrrre, re- Rat Island culture was si~nilarto that deuelooed in sources, and defensibility make sonte coast01 nrens more adjacent island groups. Conformity of artifact styles srcitable for uillage locotions thmr others. Most Rat Island througlzotrt the western Aleutians brdicates the extensive sites nre relatively low sites with easy access to tire sea. sociol interaction that occurred ainong conrntu,rities of this Subsistence is organized into eight ntojor procurement snnte area. systems of hunting, fishing, nnd gothering on open seas, -- Modern man's impact on the western Aleutian HUMAN ECOLOGY Islands is all too obvious. Artifacts of our civiliza- tion are unavoidable to today's visitor. World Owing to the extensive environmental studies a I roads, quonset huts, gun emplacen~ents, of Amchitka Island under AEC sponsorship, that recent Atomic Energy Commission (AEC) con- island stands out as the best studied of any in the struction, and plastic trash accumulating on Aleutians. Scientists have accumulated much infor- beaches from offshore fishing fleets all give evi- mation about the island's physical environment dence of man's intermittent presence. and the biota inhabiting it and the surro~nnding Given the present scars on this fragile, subarctic seas. Because the majority of the biologic studies biome, one might wonder about the counterparts stress an ecological perspective, it is appropriate deep in the area's prehistory. \\'hat debris and that man's participation also bc suminarized from constructions did ancient Aleuts leave on the the perspective of human interaction in various shores, and what were their collective effects on ecosystems. local ecosystems? \\re can provide only partial A,lan adapts, as other organisms, by genetic and answers to these queries, and they are outlined pltysiologic means, but man principally adapts here. This chapter summarizes what is known of through cultural mechanisms and therefore is human spatial factors, exploitation patterns, and unique ainong all life native to the islands. Such ciiltural interaction. Our attention is focused on adaptive mechanisms are treated here to demon- the reconstruction, incomplete as it is, of prehis- strate how prehistoric Aleuts subsisted in the toric Aleut participation in the prevalent marine various ecosystems discussed in the other chapters ecosystem. Specifically, this participation is viewed of this book. It is subsistence that impinges most within the Rat Island group, one of the major on the collateral Amchitka studies. island clusters comprising the Aleutian archipelago. However, this summary describes not the total- Thus this approach is sirnultaneoitsly ecological ity of studied ecosystems but rather those in which and regional. man principally interacted. In delimiting man's niche, we selecti\~elyintegrate tlie relevant ecologic Travel depended on large and small skin boats, factors cutting across disparate ecosystems but thc nlajority of food \\,as securcd from the ocean, characterizing man's existence. In this manner and man)' ra\v lilaterials \\.ere derived from sea- habitats and ecosystems treated by otlier investi- matnli~al exploitation. To the degree that Aleut gators \\rill be rcfcr~edto only as they are csploited teclniology via \\ratercraft made essentially all the or affectcd by man. coastline accessil~le, the varied ~narinc habitats This summary does not attempt to recreate all along these shores \\.ere exploitable. Thus mail, as aspects of Aleut culture through archaeological or tlie niost flexible or generalized omnivore, was able ethnohistorical approaches. For instance, we have to expand his dependence in more-varied habitats no living Aleut population subsisting at a hunting than any otlier Aleutian animal. At the same time, and gathering level similar to that of the past. Aleuts essentially ignored many natural commu- Thercforc we cannot study ongoing Aleut energy nities studied recently by ecologists because these flo\\r as has been done among otlier contemporary could contribute few food resources. Exploitation hunters or farmers (lcemp, 1971; Rappaport, of natural terrestrial communities was generally 1971) to discern past subsistelice or ecologic inversely proportional to their elevation above sea interaction. At best, we can combine our arcliaeo- level and their distance from tlie shore. logical and ethnohistorical data to suggest the Our ultimate measure of s~~ccessfi~ladaptation general marine exploitation patterns follo\ved by is found in tlie evidence of man's spread and the ancient Alcuts. Midden remains display direct cluration of his occupancy throughout not only evidence of subsistence, and therefore it is this this island group but also the entire archipelago. broad area that we shall emphasize. Aleuts lived on all large and most small islands and did so almost continuoi~slyfor at lcast 2500 years MARINE ADAPTATION in the \vest and 4000 years in the east. Today's visitor \\,ill note that the total native occupation of tlie western Aleutian Islands for A\ICHITIBuldir Island to the west. This island group forms a preliistoric Aleuts interacted with a basically dif- natural spatial unit in \vhich to operate, although ferent aspect of tlie physical environment, the Buldir is comparatively isolated and is included in nlarine part, and less wit11 tlie terrestrial part. The the Rat group somewhat arbitrarily. Because dan- Aleuts had an almost total co~nmitlne~itto the sea. ger from boating accidents is largely a fiunction of Land was a stable base of operations and retreat water distance and turbulence, we may safely from foul weather, but, in subsistence terms, land presume that early Aleuts also moved less fre- was only important as it formed beaches that could quently across these major passes than among the be combed, lookouts for aninla1 activity at sea, or Rat Islands proper. Tides and currents between the natural concentration areas for marine mammals, Pacific Ocean and the Bering Sea in these passes birds, and fis1i.t today present tlie same boating problems faced by past kayakers. \\'ithi11 tlie Rat group, however, *Midden sites are pockmarked wit11 house depressions; have lush dark-green vegetation stands dominated by travel was relatively easy. On clear or selniclear Iferncleanl lnitatum, Angelicn lucidn, Ligusticsm hulteaii, days, otlier islands loomed large on tlie horizon Elymus nrennrius ~nollis,Clnyto~~in sibericn, and Achillen (see island profiles on pp. 6 to 8 in hlurie, 1959). borealis; and have soils of high porosity, loxs temperature, In summary, water barriers around the Rat group and high organic matter, calcium, nitrate, and phosphorus created hnma~ibehavioral boundaries that limited (Bank, 1953). Figure 5 in Chap. 1 illustrates the manner in nrhich midden sites stand out on the coastal landscape. intergroup interaction. fFor a discussion of island ecosystems and human Travel within the Rat group was not only adaptations to them, see Fosberg (1963). possible but was required. hiterisla~id travel, for Preltistoric Hzu~inttOcc~lpatio~t of tlte Rat Islaltds 61 example, was a requisite part of snbsistence exploi- s~onsoredresearch are presented in publicatiolls by tation. The Rat Islands are dissimilar in their food Turner (1970), Desautels et al. (1971), Cook, animal population sizes aitd in the presence or Dixon, alld Holmes (1972), and Merritt (1972). abundance of various rock aid mineral resources Thcsc results are s~uumarizedin the sectiolts that serving as essential raw materials. A normal sea- follow. sonal round ~vouldtypically take an Aleut hunter, Other recent archaeological research includes if not his family, to several islailds in pursuit of site surveys in the Rat group. Bank (n.tl.) inven- foodstuffs and materials. Feasts or festivals often toried four sites on Rat Island and four on Iciska occurrecl bet\veen villages from different islands. Island during the 1950s. A,lcCartney (1972, Volcanic eruptions in any of the group's several pp. 25-26) surveyed 22% of the Rat group coast in volcanoes likely caused a mass exodus to nearby 1972 by boat and added five possible sites on islands for safety from raining ash. Attack from Little Sitkin, ICl~vostof,and Segula islands. outside Aleuts also could have forced villagers to The only archaeological excavations and testing flee to another island for refuge. \vithin the Rat group have occurred on Amchitka, Thus, to understand the relationship between Iciska, aitd Little Iciska. The coastlines of all the Amchitka islanders or any other island population islands except Semisopochnoi have been surveyed within the entire Rat Island sphere, we must for sites, but islands other than Amchitka have not approach the whole island group as an integrated been suiveyed with equal rigor. Additional sites on series of ecosystems. these other islands will undoubtedly be found in Finally, this chapter admittedly contains some the future. necessary speculation owing to the illcomplete archaeological record and our difficulty in inter- 1968 Reconnaissance preting that record. Precise delineation of settle- ment and ecologic factors is accordingly limited. During 1968 Sense and Turner began with the Relatively few sites have been excavated in the Rat Guggenheiln (1945) site inventory of 40 known group, and a larger sample ~vouldpermit interpreta- sites and expanded the total to 76. I\'. S. Laughlin, tion and explanation at a higher level of confi- R. F. Black, and L. A,I. Gard also assisted Sense and dence. Turner in their testing of 14 of the sites. Approxi- mately 1400 stone and bone artifacts came from these tests, but their descriptions have not been HISTORY OF ARCHAEOLOGICAL RESEARCH published (Desautels et al., 1971, p. 21). Turner (1970) briefly described test excava- Archaeological investigations of the Rat Islands tions at Sites 1, 2, 3, 13, 14, and 15 and human span a period of a century. Dall (1877, pp. 47-48 skeletons and fragments from Sites 3, 14, 15, 16, and 51; 1878, p. 8) first excavated at Constantine and 23. He estimated, on the basis of midden Harbor, Amchitka (Site 3), in 1873. HrdliEka (1945, pp. 220-237 and 347-364) dug on Iciska accumulation rate, that Amchitka sites dated to as long as 2500 years ago. This figure is substantiated and Little Iciska in 1936 and at Amchitka Sites 2 and 3 in 1938. l\'orld War11 brought a large by radiocarbon dates from lower levels of Sites 31 f ir inilitary populatioll to the islands, which caused and 36 (range, 890 90 to 2550 95 B.P.; the most extensive damage to prehistoric sites since Desautels et al., 1971, p. 384). No sites as large as their initial occupation (Guggenheim, 1945). Sum- those known in the eastern Aleutians (e.g., Chaluka maries of archaeological in\restigations during and on Umnak) were noted. prior to 1945 in the western Aleutians are found in The most important summation provided by A,lcCartney (1967, pp. 10-20), Turner (1970, Turner is that regarding determinants of site pp. 118-119), Desautels et al. (1971, pp. 20-21), location. He notes that (1)site distribution is not and Fuller (Chap. 7, this volume). random, most sites occurring in the southeastern The interest of AEC in t\mchitka as a testing half of the island; (2) sites typically were protected ground renewed archaeological reconnaissance in embayinents, were close to freshwater and uncler the aegis of that agency. Fieldwork during food-rich intertidal habitats, had water approaches 1968, 1969, and 1971 centered primarily on protected by kelp beds, and had easy access to the (I) inventorying all sites seen on the surface beach; and (3) no correlation was apparent be- today,* (2) making test cxca\~ationsin some of tween sites and drift\vood sources or gross geologic these sites, and (3) evaluating the effects of nuclear formations. These site characteristics are explored detonations on such sites. Results of the AEC- at greater length in this chapter. The 1968 collections have not, to date, been fully analyzed and published; they are now in the *Site locations on Amchitka are indicated in Fig. 1. possession of the University of Alaska R,luseum with the other AEC-sponsored collections. I have artifact analysis is a formal, hierarchic classification not personally inspected any of these collections. of descriptive categories including four levels. No major stylistic or cultural changes were noted by the excavators in any mitlden sequences. ti late The report of Desautels et al. (1971) is the artifact complex was suggested for the period AD primary descriptive document for Amcbitka exca- 1000 historic contact. This complex includes niue vations aud artifacts. Six sites (10, 14, 31, 35, 36, classes or general techniques from specific kiuds of and 60) were tested during 1969, and 6800 bone ueedles to "new styles of boue ornamenta- artifacts were recovered from these tests. These six tion." sites were chosen because they seemed most likely A faunal analysis of Site 31 dcbris indicates to be disturbed by nuclear detonations at the that the bone sample breaks clown into approxi- ceutral and southeastcnl portions of Amchitka. mately 85% fish, 10% sea mammal, and 5% bird Two aclclitional sites (77 and 78) were added to the bones by x\reigllt. Of the sea mammals, 59% are sea site inventory.* Six radiocarbon dates \\,ere deter- otter, 33% are harbor seal, 6% are sea lion, and 2% mined, three each for Sites 31 and 36. Site 31 arc fur seal by bone count (Desautels et al., 1971, dates range between 600 BC and AD 1060 and pp. 314-340 and 359-367). \ilhale bones are also come from a depth of 70 to 190 cm; Site 36 dates found in low unspecified frequency (Desautels cluster between 295 aud 105 BC, all being from et al., 1971, pp. 38-39). Sea urchins comprise over 210 to 270 cm deep (Desautels et al., 1971, pp. 38, 90% of the ~narine invertebrate remains. This 48, and 384). The report fully presents the analysis is consistent with the midden matrixes of stratigraphy and features for the midden tests. other eastern and western island groups. Almost all the features are fire hearths (Sites 31, 35, aud 36) or burials under wooden covers or 1971 Excavations within xvootlen sarcophagi (Sites 10 and 14) [simi- lar burial features are described by Dall (1877, Additional site excavations were made in 1971 at Site 32 in the middle of the Pacific Ocean side pp.62-63; 1878, p. 8), Jochelson (1925, pp. 45-46), \\leyer (1929), and HrdliEka (1945, of Amchitka (Cook, Dixon, and Holmes, 1972). pp. 179-180)]. Three skeletons, one adult male, This site was considered to be endangered by the one adult female, and one chilcl, were found in projected 1971 Cannikin detonation. The nlajor Sites 10 and 14. No houses were uncovered. contribution of this work !\.as the excavatiou of a The large artifact collcctious incl~tde11 iron complete house dating to about AD 1500. Four fragments, primarily of knife blades, and some of radiocarbon estimates date the Site 32 midden these, if not all, are of prehistoric age. These iron between AD 85 and 1750. A total of 546 artifacts fragments are from Sites 14, 35, and 36. The were recovered in the 1971 tests. Of a total of seven possible houses uoted at the site, only one was completely excavated. This *Of the 78 sites thus far designated, 2 (Sites 20 and 21) llouse is the only fiilly uncovered house yet are of Ilistoric age. Sites 24, 77, and 78 are designations for described in detail in the chain. Another house or blo~\~oots(areas of \$finddeflation) atop a high bluff at the two l~avebeen fully excavated at Chaluka midden, center of the Bering Sca coast. Besides the fact that no midden debris or recognizable features were noted by Sense Umnak, but no details Ilave been published. Prcvi- or Desautels at tlrese blowoots, the "artifacts" alone are ous tests at other Amcllitkan sites have uncovered questionable. Because there are essentially no recognizable only portions of house features. The Site 32 house Aleut artifact forms among the illustrated examples (De- measures 7.5 by 6.0 nl in outsicle dimensions, \\'as sautels et al., 1971, pp. 376-383) rvllich relate to the other built directly 011 bedrock, and was built al~nost midden sites, we must assume that, if these are, in fact, entirely of driftwood r;thcr that1 whale bones. human occupation sites, they are much earlier than the middens that are as ntuch as 2500 years old. If rve assume Because of thc rock base, which prevented drain- that these pieces represent a prior and heretofore unrecog- age, the structural inembers of the house have been nized Aleutian culture, tBcy should shor\- some relationship preserved by groundxvarer saturation since the to tlle only early human assen>hlage in the chain, the house collapsed. The Aleutian pattern of covering Anangala core and blade industry off Umnak (BlcCartney and Turner, 196G). However, no prismatic blades, polyhe- house roofs ~eithsod made it possible for the wet dral cores, transverse burins, and related tools were fotind sod to seal in and protect the ~voodenrvall and at Sites 24, 77, and 78. Some cobble hammerstones are roof pieces from \\rind dispersal or weathering. individually similar to recent specimens. But, taken as a The site reports by Desautels et al. (1971) and tool complex, these pieces are not recognizable as a human Cook, Dixon, and Holmes (1972) are inlportant tool assemblage. They are quite likely naturally fractured rock fragments (eoliths) and must be lleld in abeyance until contributions to \vestern Aleutian archaeolog)~. fiirther proof of their human origin is offered. Excluding Only Jochelson (1925, pp. 24-26) and Spaulding these 5 sites, the ,\mcllitka site total described here is 73. (1962) report on systematic site excavations in the Prehistoric ~II-~~IRILOcclipatio~t of tlre Rnt Isla~rds 63 Near Islands to the west. HrdliFka's (1945) excava- collected from a few of these sitcs will serve as an tions on Amcliitka, IGska, a~dLittle Iciska were important ilucleits for future cultural analyses. never properly documented. No archaeological reports are available from the adjacent A~ldrcanof group to the cast except Jocl~elson's (1925, SITE CHAKACTERISTICS AND DISTRIBUTION: pp. 26-29) short section on Atka excavations. THE HUMAN SPATIAL DIMENSIONS Although the two recent Amchitka reports shy away from cultural compariso~isand syntheses of As a natttral geographic unit, the Rat Islands* the Amchitka collectiol~s\\rithin the extant litera- collstitute an archipelago of 10 large to small titre on Aleutian and south~vestemAlaskan archae- islartds totaling approximately 1000 km2. These ology, they do provide us \\,it11 accurate and islands are distributed irregularly as exposures of specific arcl~aeological information for these the tlleutian riclge between the Andreanof iuld broader cultural comparisons. Near Island groups. The islands vary horizontally bet\veen rounded, linear, or irregular shapes and vertically from lo\\' to high. The highest point on Detoliatioll Effects Little IGska is less than 61 m (200 ft) it1 elevation, Merritt (1972) provides a short summary of the whereas peaks on Iciska, Segula, Little Sitkin, alld effects that the three nuclear detoltatioils [Long Semisopochnoi reach about 1220 111 (4000 ft). Shot (1965), kIilrore (1969), altd Camliki11 (1971)] hacl 0x1 archaeological sites. Such sites are particu- N~u~iibersa~id I Prehistoric I-lzrmu~tOcc~rputio~~ of tlte Rut Islurlds 65 Altltough we have close to 100% of all represents a different settlen>ent at different pe- Amchitka sites to analyze, we know little about riods of time. how each futlctioned or differed from others. We The following suininary excludes those site- must presume that large midden areas represent location deteril~iila~~tsrelated to the Aleut spiritual stable villages that were occupied for most, if not world or those which have no clear ecological all, of the year. Small sites are presumed to have significance. t\hiIe stressing a kind of mechanistic been seasonal camps, such as fishing stations, but approach to man's use of the physical and biologi- sites do not aggregate into clearly separable types cal environment, we recognize that other cultural on the basis of size, depth, elevation above the pressures exist for establishing communities, beach, sampled artifact content, or other ob\''lous moving about, and exploiting certain ecosystems. characteristics. Therefore this section is devoted to A village may have been consistelltly occupied it1 a review of spatial variables or determinants it1 an part because strong ties to ancestors or spirits attempt to better ullderstalld why Aleuts estab- residing there dictate such a behavior. One food lished sites where they did. resource may have been overlooked in lieu of Three related factors restrict our current another bcca~tseof taste preferences. Motivation kno\vledge of site function. Most sites have llcver for moving an encampment may have sprui~gfrom beell systematically excavated for archaeological desires to "keep moving" rather than depletion of data that would bear on the question of function, the local food supply. seasonality, and duration of occupation. Second, In short, mall has illany con~plexiilducemellts the World War I1 military occupation is estimated guiding behavior, and we should not ininimize to have damaged 60% of Amchitka's sites i~ollecologicalcultural pressures when vie\ving ma11 (Desautels et al., 1971, p. 2), leaving, in some interacting in the Aleutian environment. But, cases, no ~utdisturbed strata to be tested today. because we have little knowledge of ~toi~ecological Third, the 1968 artifact samples from 14 different factors, the determinants described here portray sites have gone unpublished. Specific information mat1 iilteractiilg with spatial and ecological limita- about intersite variation that might have come tions. from such an analysis is therefore not available. Because we cannot ascertain specifically how Interisland Determinants past Aleuts moved from base settlelne~ltsto sea- Of first interest in iilspectiilg the Rat group sonal camps during the yearly subsisteilce round, sites is \thy 80% of them fall on Amchitka. The ~vemust consider each site a cultural entity by fact that Amchitka is a relatively large island is itself, representing a distiilctive social/political insufficient to explain the site concentration. A group over an uilspecified period. Yet, in at least partial explanation is that site surveying has not seven illstailces on Amchitka, two sites are located tdkcn place as itltensively on the other islailds as it very close together (<0.5 km or 0.3 mile). If these has recently on Amchitka. But a more critical were occupied contemporaneously, these double factor appeals to be operating; i.e., the other sites may have been different occupation areas for isla~ldsare not as suitable for human occupation. the same people. Six of these double-area sites On probing into differential site suitability and (2-3, 6-43, 22-23, 31-56, 32 high and low, and irregular site distributioil of past Aleuts, 'ive find 35-36) comprise a high area and a low area several overridillg patterns. First, different islands essentially side by side. It is possible that the were not occupied with equal intensity. Second, on different elevation areas represent a seasonal shift the same islaud site distribution was never uniform up and dorvll a common slope (Desaictels et al., along the coast, and sites were never located in the 1971, p. 347). The seventh site pair (72-73) is an interior. Third, on the coast oilly lower bencl~es example of two high sites. and terraces were commollly occupied. Finally, Because the distance separating any site pair is even in these lower coastal zones, still other so slight, it appears unlikely that the sites represent considerations apparently influenced the location siinultaneous but discrete political/social groups. of a habitation site. As an alternative, ho\vever, these site pairs may In this section some of the key determii~allts have been occupied sequentially, with a later that help explain the utilizatioil of the land inass village or camp being established in proximity to, for seasonal or ycar-routld settlements are exam- but 1101 directly on, a former settlement. ined. Variation between islands is treated first, Since we lack precise cross dating between ally folloxved by variation around the same island. of these site pairs or analytical tools for discerlliilg Future field observatio~ls and excavatiolls if the same group occupied two slightly separated should be directed at testing these determillants as areas, we assume here that each iildividual area hypotheses. Enough evidence is available to suggest strong positive associations between site localities islands, but Semisopochnoi is distinguished from and these variables, but grcater analytical precision the former two by having a lo\ver index figt~rc and model testing may permit us to derive causal (4.69). Semisopochnoi's circular shape, relatively explanations for the phenomena obse~-\red. regular coast, and comparatively small perinleter confirnl this low index. Besides Amchitka and Island Size and Shape. Depending 011 its size Itiska, Rat Island also has a larger index and shape, one island rnay provide greater or lesser (7.02) than most, indicating that it falls between coastal living space to prehistoric inhabitants than Amchitka and ICiska on the one hand and tlle another islancl. Perimeter is a fl~nctionof size, but rounder islands (3.94 to 5.07) on the othel.. it also increases \\,it11 the degree of deviation alvay Although Rat has less arca (30.3 km2) than Segula fro111 circular plans and/or smooth coasts. A (36.5 km2), it has a larger index because of the narrow island or one wit11 convoluted and indented longer coast. margins \vill have a greater perimeter than a round island or one ~vitliunindcnted coasts. As a measure I11 terms of human occupation, island area of the relationship between area and coastal length, done is not so important as perimeter or index of a ratio or index of irregt~lnrity is calculated for irregularity. Only tlie coast or periphery is utilized each of the Rat Islands to sho\v their degree of for subsistence, and therefore island perinleter is a deviation from circular, regular coastlines more meaningful figure than area. The greater the (Table 1). Because area and linear measures are not coast leugth per unit of area, the larger the proportional, this index is necessary to compare potential population that could be supported by islands of differing shape and shoreline indentation resources along it. regardless of island size. This table clearly sho~vs Given the limiteel size range of Aleutian that Amchitka (298.4 km2), Iiiska (296.6 km2), Islands, the index figures make it possible to and Seniisopochnoi (238.6 km2) are the largest conipare the likelihood of travel across any island if elevation and terrain are disregarded. Distances across islands \\rill be less on islands \\.it11 higher indexes. Altliough sites are not found inland, away Table 1-Rat Island Area and Perimeter hleasares from the coast, the ability to traverse an island's Area,* Pcrin~eter,t Index of interior rnay have been crucial when considering Island km2 km irrcgolarityi escape fronl attackers arriving by boat or exploita- tion of aclditional coastal zones. Howevcr, cleva- Amchitka 298.4 194.7 11.27 tion and topography are important in determining Kiska 296.6 160.9 9.35 thc case \\,it11 which Aleuts might cross from one Rat 30.3 38.6 7.02 side to anotl~er.Vertical shape, therefore, may be Little Kiska 8.8 17.7 5.96 Davidof 4.7 12.9 5.94 rather important under certain cultural conditions. Segula 36.5 30.6 5.07 Most of the Rat Islands are volcanic cones Khvostof 2.6 8.0 5.00 Little Sitkin 68.8 39.4 4.75 (Huldir and Segula), caltleras (Little Sitkin alcl Selnisopochnoi 238.6 72.4 4.69 Semisopochnoi), or caldera remnants (Davidof and BuI Type 1: Low depositional shores 7 // C~eachdeposits, alluvium, colluvium with gas cover Kirka Harbor -EAST Type 2: Spit shores Active rand, pebble beach facing reaward -NORTH Bechevin Bay Unimak lrland J Type 3: Wave-cut exarpment shores Glacial deposits --ilteau with vegetation cover Rock plateau Anderitic breccia covered with colluvium and vegetation Stabilized rcarp face; vegetation cover Bluff adjacent to Site 32 +SOUTH Amchitka lrland Active scarp face Talus over exposed barat rocks on active beach beel Rock and volcanic plateau EOIIUV~U~,and vegetation Stabilized rarp face: vegetation cover Western shore 6 Kagamil lrland Meters tlYEST Fig. 2-Types of Aleutian shore profiles on which archaeological sites occur. Types 1 to 3 are lo\\. coasts with landi~~gedges and flat living spaces above. A sunrey of the 10 Rat Islands sho\\rs that only \\'hen low-coast percentages are compared with part of any island's shore is relatively lo>v in the actual kno\vn site distribution (Fig 1) the sites manner of Types 1, 2, or 3. Table 4 summarizes almost always fall along low coasts. The majority these lo\v coastal zones. The table illustratcs that, (82%) of Amchitka sites are foiund on the south- although suitable low coasts vary bct\veen 0% for castern 60% of that island, ~vhichis lo\vcr than the Davidof and I Buldir lrland Interspersed vegetation Interspersed vegetation and Adak lrland -SOUTH -NORTH Active cobble beach &Active cobble beach covered b'lilh talus covered with taiur Active cobble beach covered with talus Type 5: Cliff shores with higher constructional slope and talus beach r Vegetation over volcanic mckr and arh Kagamil lrland South shore Davidof lrland cliff with talus Exposed rock cliff with talus Offshore rock islet and stack Type 6: Steep cliff shores with no beach Wert shore / Kagamil lrland over volcanic racks and ash South share Stabilized rarp of Little Sitkin lrland volcanic rock covered Pwith vegetation +WEST Exposed rock cliff Active erosional cliff +SOUTH of expored rock Fig. 3--Types of Aleutian shore profiles on r\.hich archaeological sites do not occur. Types 4 to 6 are too high and/or too steep for hu~nanoccupation. terraces. Again must remember that intensive Using this ratio and disregarding other island site surveys comparable to those on Amchitka have determillatits outlined above, we may projcct a not been collducted on the other Rat Islands, but total of 110 sites for the Rat Islands. We alrcady recent surveying and records of previous investi- have 73 for Amchitka plus 19 fou~ldon seven gators (AlcCartncy, 1972) all corroborate the lo\$, other islatids. \\'e might expect to find about 18 coast-prehistoric sitc association. additio~lalsites on islands other than Amchitka in In speculating about the total number of the future. If accessibility is, in fact, shown to be prehistoric sites in the Rat Islands, we use Am- the most critical factor in site location, then this chitka as our referent because we kno\\. the postulated total capacity for sites \\,auld be ap- locations of essentially all the sites (N = 73). If we proxinlatcly accurate. If we assume a 110-site total colilpare ~lu~ilbcrsof sites to low coast, nrc find a for all the Rat Islands, Amchitka will have about ratio of about 0.63 sitc per low-coast kilometer. 66% of these sites. Amchitka, Iciska, and Rat, the Preliistoric HILIIIOI~Occupatiotl 0.f the Rat Zsla~tds 71 Table 4-Low Coast Distribution on Rat Islands Lo\\. coast Percentage Perim- suitable for of total Island cter.* km sites,t km ~erin~eter hlaior lorn coast areas - Atnchitka SE 60% of the island. Rat SE end near Ayugadak Pt.; N\\' 30% of the island. S coast; N\\'coast between Zapad 'lead and Gula Pi. Little Kiska \Vend of the island. Kiska Kiska Harbor on the E; \\'coast be- tween \\?tchcraft Pt. and \'illcan Pt. Semisopochnoi Alluvial valleys on E and W bays. Little Sitkin \Villiwaw Cove on the N; Sealy or Prokhoda Pt. on the SE. Buldir N\\'alluvial sallev. Davidof Khvostof *See Table 1 for mar, source. Vhese approxinlate figares are estimated from the same map sources as in the preceding footaote;lo~vcoasts are here taken to be coastal zones of <30 nl elevation, comprising flat marine terraces, alluvial terraces,or sloping sand and cobble beaches. Only general coastal areas are measured; a small, isolated spot ,nay be suitable for a village but may not be detected on the maps used for checking these coasts. three islands with the maximuin lo\\' coasts, \\rill have about 87% of all Rat Island sites. Recent t\EC study of Amchitka has made that island our best referencc base for site distribution. Additional site surveys on other Rat Islands tray test this coast-site relationship for the enti1.e group. By checking and adjilsting the regularity of this relationship tsitliin the Rat group, one may find such a ratio to be of great value in testing site-population capacities in other parts of the chain. \\re have referred (Table 4) to lo\\, coasts being liinitecl to approximately 30 in above sea levcl. There is no clear division at this elevation, but nlost Amchitka sites appear to ]lave this elevation as an upper limit. Desautels et al. (1971, p. 22) note that sites fall into two general classes on Amchitka, high bluff sitcs ant1 low beach terrace sites. Site elevations are shown in Fig. 4. There are ;ELEVATION. o, v - " discrepancies in height and size cstimates betweell LOWsites High rites Desautels et al. (1971) and preliminary suivey 3815481 3314611 records of 1968 (Scnse and Turncr, 1970). Fig- Fig. 4-Histogram of Amchitka siie elevation fre- ure 4 treats the preliminary survey ele\ration esti- quencies based primarily on 1968 site sutvey forms. mates; precise site measureinents will probably hleasured from approximate high-tide line. alter these figurcs slightly. The ele\ration distribution sho\vs a break at 13 111, those belo\\, referred to as lolv sites and those above as Iligh sites [low coasts (<30 m) are The high bluff sites fall predominantly at three not to be corifused witli low sites ((13 111) as elevatioil levels, 15.3, 22.9, and 30.5 n~,reflecting discussed in thc text]. Slightly over half of these the nataral clevation of different parts of tlie total sites are lo\\' sitcs. Almost all the sites at island's periphery, or possibly reflectiilg thc general 8.9 111 or less are found on or behind bcacli storin use of roundetl ele\ration estinlatcs in uliits of 50, terraces; half tlie sites bet\\reen 9 and 12.9 m are on 75, and 100 ft, resl~ecti\rcly. Tllcre appears to be beach terraces and half are on low bluffs. lcss of a Iieight contin~uu~ifor high sitcs, but precise measurements of elevation may sho\v more oftell have old sites, isthmuses, necks, aud spits are variation than is reflectetl in these figures. relatively rare 011 ally Aleutian island. The Am- The highest site (Site 75; 68 111) is found 011 chitka site distribution substantiates Veniaminof's l\'incly Island off the Pacific coast of Amchitka. observation that protected emba)~mentsare where The site is atop a hill on this islet which falls sites were usually located. directly to the low beach belo\e. The elevation Another variable of site exposure is aspect, or alternatives for a calnp are therefore severely direction faced in relation to the sea. Aspect and limited. its relati011 to \\rind aud storm directious duriug the seasons the site was occupied would be important Exposure. Site exposure tnay be measured illsofar as it influenced the ability to survey the sea roughly in terms of protection offered from wind surface and easily enco~cntersea mammals. Fig- aud other \\reather elements as well as hy aspect, ure 5 is a directional rose iudicating percentage i.e., directiou faced in relation to the trend of the directious to\vard which Amchitka sites face. Site seacoast. Settlemeuts theoretically could be established (1) 011 straight regular coast, (2) on headlailds and points, or (3) 011 embayments. Although we find all three situations in the Rat Islands, the large majority of Alnchitka sites is found in concave etnbayments (see Fig. 1). By settliug in bays, bights, or coves, Aleuts fouud protection from the ever-present \\rind along beaches isolated between headlands. Bay shores are allnost universally lo\ser than headland shores on either sidc because head- lands are formed by steep, resistaut rock exposures while bayhead shores are formed by alluvium aud beach deposition. Thus bayhead shores are typi- cally Type 1 profiles, and lieadla~ld shores are typically Type 5 or 6 profiles. Beaching a skin boat would be simpler in the quiet waters of such embaytuents, especially if at1 exposed reef or kelp beds preveuted breakers from enteriug the beach area. Reefs 30 to 90 m wvide are almost always found before Amchitka's sites, and kelp beds altnost totally surroiuild the island. Operating from the head of a bay, a villager \vould Fig. 5-Directional rose showing frequencies of the survey a curving shore that could be exploited way Amchitka archaeological sites face the open sea. easier that1 a straight or convex coast since he could travel by skin boat across the water at chords facing is defined as the directiou perpeudicular to to the cut-ved shore. the shore innnediately adjacent to the site looking The index of irregularity ctescribed earlier mill outward from the site. suggest the degree to which oue islaud is likely to The two major directions faced, sonth\vest aud have greater or lesser local protectiou for shore use. northeast, are perpe~~dicularto the long axis of the The higher the index, the greater the probability island. Hoivever, more sites face south than north, that there will be numerous coves and embayments making the orientation toward S-S\\' slightly offeriug protectiou from the wind and ~veather. predominant over that to\vard N-NE. If we tally The ethnobistoric literature is confnsing about sites on the txvo long sides of Amchitka, we find 31 the relatiou between site locatio~laud coast con- or 44% facing the Bering Sea aud 40 or 50% faciug figuration. Veuiaminof (1840, Vol. 11, p. 200) the Pacific Ocean. Two other sites face the states that villages (age unspecified) are "laid north~vest at that eud of the island. There is no out. . .for the most part, in bays and gulfs in order significant associatio~lbetn~een ocean frontage and to liave as convenient a lauding station as pos- site elevation; of 39 Pacific Ocean sites with sible." Jochelsou (1925, p. 23), hoxvever, states elevation data, 20 or 51% are loxv and 19 or 49% that "ancient Aleut villages" \\'ere normally located are high. Bering Sea and Pacific Ocean site fre- "on narrow isthmuses, on uecks of land between quencies correspond closely to present-day sea two ridges, on promontories, or narrow sand- otter distributiou (Bering Sea, 40%; Pacific Ocean, banks." Although it is true that such localities 60%) (Estes aud Smith, 1973, p. 65). \\'hether Prelristoric Hzrntan Occz~patiotrof the Rat Islands 73 distributions of sites and sea otter are meaningfully Further, any shifts in sublnarine and terrestrial associated, however, is impossible to determine on habitats may ha\.e caused corresponding shifts in current evidence. allillla1 occupation of such habitats over several The relative proportion of sites on the Pacific mille~l~liaof human occupation. We cannot assume Ocean and the Bering Sea is of interest because that present habitats have gone totally unchanged Veniamillof (1840, Vol. 11, p. 200) states that over time. during the early ~titleteetlthcentury "the greater Freshwater close to a site, preferably in the number of villages" faced the Bering Sea. Sarichev form of a stream, was desired (Veniaminof, 1840, (1807, p. 72) earlier reported that Fox Island Vol. 11, p. 201). Certainly a number of Amchitka settlements \\.ere not located on the southern or sites typically ha\^ streams adjacent or close to Pacific Ocean coasts. Amchitka sites, at least, do them. Living by streams may also be important not corroborate these conclusions for the pre- because of salmon spawning in some of them historic period. during the summer, thus concentrating a valuable food resource. However, drinking water from Resources. Resources (food, water, and raw streams is not considered to be critical owing to materials) are grouped here as a major determinant the wet climate and abundant springs and seeps of site establishment. Yet resources and defense, that are almost ubiquitous to the Type 1 and 3 described next, appear to be more problematic as shores. determinants than those already discussed. Be- Nonfood material resources usitally come from havior patterns surrounding resource use and de- animals eaten, especially sea ~nammalsand birds, or fense are understood only indirectly,. \\re can only from driftwood abundantly scattered on beaches. roughly assess past resource abundance at ally Stone and mineral resources were collected at particular coastal area on the basis of modern localized areas, often away from the shoreline, and abundance. Likelvise, the efficiency with \vhich therefore would hardly be determillallts for more Aleuts may have procured such resources can only permanently occupied sites. be assumed. Cotlsideri~lg only resource exploitation, we Resource availability and original population may expect coastal hunters and gatherers to size should be directly related. Large villages arrange their villages at unequal intervals around require abundant resources for support. \\re have these islands. hc1ore than an upper threshold num- some direct evidence of resource utilizatio~lfrom ber of villages per coastal length or people per the archaeological site record, but exactly how a village ~vould strain the resource network too particular resource was used is not clear in all cases. much, and population declines would result. One measure of resource abu~tdancemight be Beginning with food, we may presume that the ratio of sites per total island perimeter. This some coastal areas were favored variously for their may be expressed as the distance between sites. sea lion rookeries, seal hauling grounds, sea otter- Given Amchitka's 73 sites and the estimated island kelp beds, urchin and shellfish habitats, fishing perimeter of 194.7 km, we find an average intersite spots, bird rookeries, plant collecti~lgzones, etc. distance along the coast of 2.67 km. If we measure Yet how important these specific resource areas straight-line distances between adjacent sites, how- were to people choosing a site locality is difficult ever, we find a mean intersite distance of 2.09 km to evaluate because \cre have almost no historic- ancl a range of 0.2 to 10.5 km. These figures, of period settlement data for comparison. course, ignore a time dimension in that not all sites Aleuts probably avoided living immediately were simultaneously occupied. adjacent to sea mammal or bird rookeries to avoid disturbing these animals. To choose a site just for Defense. Venialninof [1840, Vol. 11, pp. its access to marine invertebrates or summer plants 184-186 (cited in HrdliEka, 1945, pp. 144-148)j would seem unlikely, although these food orga- reports late prehistoric internecine contacts be- nisms were usually available (Veniaminof, 1840, tween islanders and describes defensive refuges as Vol. 11, p. 200). Access to a major resource like steep offshore islets providing temporary fortresses \vhales could not be facilitated except in a gross that \\.ere easily defensible. Coxe (1787, manner as in choosiltg a site overlookillg an pp. 110-121), among others, describes early interisland pass where cetaceans commonly travel. Aleut-Russian conflicts that amply demonstrate Since the Aleuts were quite mobile in ski11 boats. the Aleuts' \varring abilities. Jochelson (1925, for almost all the year, we must not coltsider them p. 23) reports that sites were chosen especially for bound to specific areas for resources. Hunting and their defe~tsivepotential in casc of attack. But, as fishing from boats required extetlsivc travel, often remarked above, Jochelson's i~lterpretatiotls of aroitnd or out from lllorc than one island. wvllat coastal terrain constitutes appropriate habita- tion sites is questionable. Still another question is figures. Sites are separated between high and lo\\, \\,hether a defense variable was important during sites as previously described to discrilninate rela- the elltire Aleutian occof:tion. Prehistoric \varfare tionships between site size and elevation. Sites vary has not yet been identified in the archaeological \videly in size from 25 to 2500 m2. Eighty-three record. percent (59) of the sites are less than 800 in2 in hsIost Amchitka sites are not located on narrow area. The most frequent size class is 200 to 300 m2 spits or points assuring easy escape by some "back for both high and lo\\, sites. There appears to be no ~vay," nor are they usually found above high positive correlation bet\veen size and elevation precipitous cliffs serving as natural defenses. It is since the distributions of both high and lo\\, sites clear from Vcniaminof's record that the normal follo\v one another closcly. ltabitation site is not, at the same time, the Another nleasure of population size or site defensive refuge, or "fort." Rather, such refuges duration is depth of cultural debris. Large or stable were steep coastal spots or islets close to the village settletnents should evidence greater depth than \ehich offered protection only under attack. There- small or nlore transient ones. The majority of fore defensive elements, such as sheer cliffs and Amchitka sites have uot been tested, and only high elevation, were not normal deternli~latltsof depth approxiinations call be made. Of 72 sites for midden site placement. which depth has been estimated, we find a mean \\re note, however, that in contrast to Am- depth of 1.2 n~ and a range of 0.2 to 3.8 111. Only chitka and other narro\jr islands, being able to 15%, or 11 sites, have estimated depths of 2 m or escape attack by walking across the island to a greater. Unfortiui~ately we have no coinparable frie~tdlyvillage is not possible on maily islands. figures from other major islands with ~vllich to High volca~licconc islands \\~ouldpresent a poor contrast these depth estimates. opportunity for cross-island travel. Hiding in high- On the basis of present estimates, there appears elevation caws or other rccloubts among interior to be no correlation between depth and size of rugged peaks \voutd be a11 alternative in seeking sites. 'This may be thc result of inaccurate depth safety. estimates, or it may mean that large village populatioils li\wl at settlements for a relati\rely DEMOGRAPHY short periotl. As \\,ill be deinonstmted later, site size, if not depth, is helpful in deriving a site Aboriginal population callnot be ascertaiited population. directly but call be estimated on the basis of site A'Iany educated guesses enter into the prc- size and ethnographic analogy to historic sites. historic population estimate for Amchitka and the Amchitka site size distribution is given in other Rat Islands. If we look to the extant Fig. 6, \\,hich is based on 1968 prelin~i~~arysur\.ey ethnohistoric literature describing carly historic SITE AREA. rn2 Fig. 6-Histogram of Alnchitka site size frequencies based primarily on 1968 site sursey forms. ,historic Human Occ~~pntio~fof tire Rat Isln~cds 75 (AD 1750 to 1800) house and village sizes, we are radiocarbon dated; so we cannot verif~, their cotifounded by two problems. First, most of the simultaneous occupation. literature describing Aleut dwellings deals with the Hoxvever, let us speculate on some possible eastern islands. And, in late prehistoric times, the population sizes. If all seven houses held six to single-family bnrabnrn was replaced or snpple- eight persons each, xve \voulcl estimate the upper niented by a much larger multiple+family bnrabarn midden population at 42 to 56. The lower iniclden housing as many as 100 or Inore natives (Coxe, area next to the stream at Site 32 may or may not 1787, pp. 119 and 214; Lantis, 1970, have heen occupied at the same time. Cook, Dixon, pp. 172-174). The only fitlly excavated house in and I-Iolmes (1972, pp. 5-6) give no specific site the Rat Islands is the 400-year-old wooden struc- sizes for either the upper or lolver ~lliddenareas, ture described earlier (Cook, Dixon, and Holmes, nor do they indicate whether possible house 1972, pp. 91-101), \vhich \\.as of single-family size. depressions \\.ere found in the lotver midden. We have no direct evidence of tlle Larger multiple- Since the most prevalent site size 011 Amchitka family houses occurring in the western Aleutians, varies between 200 and 300 m2, we might expect although they niay eventually be found. Therefore these settlenients to have between 24 (four houses large multiple-family dwellings of the eastern times six occupants) and 56 (seven llouses times islands cannot he meaningfully compared to small, eight occupants) total population. If we assume single-family dwelli~lgs of the \vestem islands that 45 m2 is typical for Amcliitka house space during the late prehistoric and early contact and that as Inany houses as possible \\ill cover any period. site, then Amchitka's 73 sites (33,044 m2) should A second pitfall lies in taking village size during have about 7 34 total Iiouses. Further, assun~ingsix the late eighteenth or early nineteenth century to eight persons per house and that all houses were Russian period as i~ldicative of the precontact si~~lultaneouslyoccupied, we arrive at an upper village size. As Jochelson (1925, p. 119) points limit of het\veen 4400 and 5900 persons for out, the Russian colonists congregated the Aleut Amchitka. populatio~lat particular village locatio~isfor easier Going beyond Amchitka, we noted above that exploitation and administration. Population de- about 37 (33%) sites shoulcl be foinid in the Rat cline was great during the first 40 years of contact, group besides Amchitka's 73 (67%) sites. \\re havc and the population sizes of established Russian- no accurate size estimates for non-Amchitkau sites, period villages may not represent those of previous but, if we assume an equal population density, villages. Veniaminof (1840, Vol. 11, pp. 202-203) then \\re would estimate that an additional 2200 to does sho~vthat during 1825 to 1830 the average 2900 persons occnpied the other islands. This village size was 55, with a range bet~veen13 and \vould bring the Rat Island total population to 196. Suggestions of Unalaska village size during the het~vecn 6600 and 8800. Ho\vc\rer, we have no last quarter of the eighteenth century range be- current way of kno\ving holv many sites \Irere tween 32 and 45 (Lads, 1970, pp. 173-174). occupied simultaneoi~sly. Sample calculations for The nuinher of persons per I~oi~sel~old,or 100%, 50%, and 25% are as follows: single-family d~velling,times the number of houses 1. Rat population (100% of sites occupied provides one estimate of site population. But three concurrently): 6600 to 8800. factors cannot be controlled. We do not kno\v 2. Rat population (50% of sites occupied con- exactly llo~vmany houses existed at a particular currently): 3300 to 4400. site, how many houses were sitnultaneously oc- 3. Rat population (25% of sites occupied con- cupied, or how many persons lived in a house. currently): 1650 to 2200. At this point the single excavated house from Site 32 is relevant. This ~voodenstructure measures \\'hich of these estimates is most realistic? If we 6.0 by 7.5 111 in outside dime~lsions(Cook, Dixon, accept the judicious opinions of Veniaminof and Holmes, 1972, pp. 91.101). This house oc- (1840, Vol. 11, p. 177), Lantis (1970, p. 174), and cupies 45 n12. If the upper midden area of the sitc Polonskii (cited in Fedorova, 1973, p. 160) that in fact measures 300 n12 as reported in the 1968 the total precontact populatioii for the archipelago survey, then only seven houses of roughly equal fell somewhere between 12,000 ant1 15,000, then size could fit into the site area. In fact, the clearly the first estimate is too Iiigh. Because the excavators report seven depressio~isat this upper Rat group constitutes only about 11% of the bluff area. We may only surmise the number of chain's total coastline, the lowest figure (1650) fits occupants living in the house, but, taking the best if we assume a uniform relationship between bench sleeping area available, it would appear that number of sites and kilometers of coastline. If six to eight adults ~voaldbe a niaximum number. continued scholarship substantiates the approxi- None of the other six house depressions were inate 15,000 chain m;~simum,our largest error in calculsting population is probably in assuming races or tablelands, ~notultains, lakes, streams; simult;rneous site occupation. other zones discri~ili~iatedon the basis of elevation, Given at least 2500 years of Aleut occupation, slope, drainage, etc.). All these habitats were the 25% rate seems reasonable for simultaneous exploited with differential intensity through the site occupancy. On Amchitka this 25% rate could various procurenlent systems described later. be accounted for in the following \tray. Eighteen By using ski11 boats Aleuts maximized their sites of 61 perso~ls each \\rould comprise 1100 exploitation of marine resources. Their mobility people (67% of the Rat Island populatio~iof 1650 enabled them to sample more diverse marine and 25% of the Amchitka sites of 73). These 18 vil- habitats and prey species tliail ally other competing lages ivould be separated from one another by 10.8 Aleutiail predator. h,Iatl could hunt sea mammals coastal kilometers (based on the 194.7-k~ncircum- on the sea surface, catch fish deep beneath the ference). Every person \\~ouldbe dependent on an surface, collect invertebrates from the intertidal average coastliile stretch of 0.18 km. Given the zone, snare birds from a sea cliff, or dig roots in an kinds of marille resources actually depended on island's interior. On any island or island gsoup, and the sea space used, this area is realistic. then, man's exploitatirle capacity made him the The preceding ~lumericalspeculations are pre- most successful omnivore at the highest trophie sented as just that. But we need constantly to level in the food web. examine population variables to learn more about Aleitts were specialized in subsisteilce to the the ecological capacity to support human groups. degree that procuremeilt systems dealt with rela- tively few kitlds of animals and plants. Ho\vever, MARINE ADAPTATION AND EXPLOITATION they were not so specialized that population PATTERNS decline was inevitable if one or two animal species failed to appear over a brief period. Within Abosiginal man living in the Aleutian chain had maritime limitations, Aleuts subsisted on a broad no choice but to depend 011 marine resources for range of organisms in diverse habitats. survival. Food and ra\v materials came predomi- Water Travel and Resource Accessibility. In- nailtly from the sea; a native could not sur\ri\re on teraction among itihabitants of the Rat Island terrestrial organisn~salone. Besides comprisiilg the group was stressed in the introductioil. The obvi- principal larder, the sea provided for transportation ous mechanism for suc11 interaction was interisland bet\veen island masses and \cras an important travel by skin boats. \\'l~etherby small kayaks determinant of \\reather affecting those islands. (Daidarkas) or large umiaks (Daidax), male hunters This section treats the manner in which man and fishermen or entire families could exploit used the marine environment for food, raw mate- esselltially all the Rat Island coasts and intervening rials, and travel. This utilizatioil is organized and surrounding waters for food and raw material. through subsistence procurement systems, the di- \\re have demoilstrated that not all the Rat Islands rect interactions mail had \\,it11 marine ecosystems. were inhabited as densely as Amchitka. I-Io\crever, Finally, an assessment of prehistoric man's impact all the islands were probably exploited for food on his physical and biological eilvironme~itis also and other materials to some degree. Although steep provided. coastal profiles prevent persistent land occupa- tioils, these coasts may be abundant sea bird, sea Resource Utilization mammal, and fish habitats. Further, skin boats not Habitats. Cultural adaptations to habitats and only open up more different kinds of liabitats than their associated biota dictate that hunters and those exploitable by land-bound hunters and gath- gatherers shall live near the seashore in this atid erers but also make more similar habitats available. similar archipelagoes. The littoral zone and the If one sea lion rookery or bird colony \\,as open sea surface are by far the most important temporarily vacant, then similar localities \\,ere sources of food and raw material; inland terrestrial available to the boatme11 on the same or a nearby liabitats contribute more raw materials than food. island. Three major macrohabitats can be isolated with Finally, in their skin boats Aleuts could travel examples of particular habitats within these cate- \vith almost equal ease in any directioil on the sea gories: (1) o/~et~seas (inshore and offshore zones; surface, weather permitting. Therefore, with good many distincti\~eareas discriminated on the basis of visibility betftreen the Rat Islands, nve most pre- currents, tides, and waves), (2) littoral zone (sea sume that boating bet\\reen isla~lds\vas as likely as cliffs, beaches, reefs, and islets or rocks; other areas travel around islands. Aleuts li\ring on the north- discriminated on the basis of ele\ration, angle, rock \crestern end of Amchitka could reach Rat Island structure, etc.), and (3) terrestrial areas (low ter- with less effort than paddling to the southeasteril Prehistoric Hu?nan Occupatiot~of the Iiat Islat~ds 77 half of Amchitka. Or Aleuts camped on the lnary interacting elements into integrated \\,holes, northrvestern end of Iciska could cross to Segula (2) process or interaction betlvcen elements, and for hunting quicker than they could travel to (3) ecologic relationships between cultural and southwestern ICiska. A predilection to traveling biological spheres. These systems, though abstrac- around an island rnass is consonant with a land- tions of hnmali behavior, reflect as realistically as based world view but not to one based on full possible the manner in xvhich Aleuts interacted maritime adaptation. with their islands and surrounding seas. They suggest the major subsistence themes and at the Procuremellt Systems same time show the patterned, nonrandom be- Analysis of Aleut adaptation shows that there havior in relation to the environment. Procurement existed several major ways of procuring food and systems are the cultural adaptations or survival raw nlaterials. Rather than applying a specialized schemes that most directly and abundantly extract technology and behavior pattern to procure each required nutrients and other resources from the animal and plant species used, Aleuts @ansferred environment. Systems are complex, and therefore these patterns to several habitat-related species. we cannot analyze them reduced to their separable These ways of subsisting or acquiring requisite parts. In other words, the organization of the parts materials from the environment are called procure- is the target for study. Finally, by integrating ment systems (McCartney, 1975, pp. 295-300). archaeological and ethnographic data, we can Flannery (1968) applied this same general ecosys- synthesize the bare archaeological finds into a tem concept to h,Iesoamerican cultural devel- dynamic cultural process. opment. A system is conceived to be "an intercommuni- Aleutian Procurement Systems. Aleutian pro- cating network of attributes. . . forming a complex curement systems, as typical of most held by whole" (Clarke, 1968, p. 42) or "a functioning set hunters and gatherers, are largely determined by of elements that are interrelated so that change in habitat, seasonal occurre~lceof important species, one affects the others" (Redman, 1973, p. 16). and specific animal behavior. As a result, Aleuts And, of course, systems pertain in this instance had relatively little latitude for effectively manipu- specifically to networks or functional sets of lating their environment. They raised no crops, procurement elements. As part of every procure- kept no domesticated animals with the possible ment system, we find the follo\ving elements: (1)a exception of dogs, and had no sure method of food habitat where sought-after organislns or materials storage. Their adaptive flexibility lay not in jug- occur; (2) the organisms or material proper; (3) the gling a particular system but in switching systems if implements or tools used to extract the organisms one failed to produce necessary food or materials. or materials from the habitat (technology); (4) the Eight procurement systems are tentatively sug- dynamic extraction process; (5) the prerequisite gested for the Aleutian chain. These systems fall in kno~vledgeand skill guiding the extraction process inclusiveness between higher level divisions (hunt- and application of technology (techniques); (6) the ing, fishing, and gathering) and lower level divisions human procurers xvho singly or in groups represent (e.g., winter sea otter hunting, summer halibut larger kin or residential units as families or villages; fishing, etc.) and are a compromise in degree of (7) the food or rawmaterial products that emerge specificity. These eight systems reflect the majority as usable input into subsistence, or the component of procurement patterns for the entire chain, parts of an extracted organism after additional although some island population will be expected processing (e.g., butchering); and (8) the time to deviate slightly from these as local conditions during which the preparation, extraction, and demand. processing occur, or duration and seasonality. Although these systems have been distilled From the archaeological record, rve only observe from historic-period literature, they are thought to directly remains of organisms used, inorganic mate- have cotlsiderable antiquity. Because the archipel- rials, and tools; but organislns and materials can ago falls beyond the southerly limit of west usitally be associated tvith particular kinds of Alaska1 drift ice and is dominated by maritime habitats if not exact localities. Ethnographic and climate rather than continental climate, tempera- ethnohistoric documents shed light not only on ture and other weather fluctuations over the past tools but also on process (4), know-how (5), several millennia probably caused only minor organization of hunting and gathering groups (6), deviation from today's pattern. I\,Iodern biota are resulting products (7), and amount of time and sholvn in the midden debris to have had very little seasonality invol\led (8). change in character for thousands of years. There- Thc usefulness of procurement systcnls as fore procurement methods are presuned to have heuristic devices lics in their inclusion of (1) pri- been relatively stable as well. The Aleutian procurenlent systeins and major I11 Fig. 8 food primarily and raw materials habitats are as follows: secondarily affect the population of a region like Hzt~tti~tg: the Rat Islands. Feedback acljljustments are shown 1. Sea mammal hunting offshore (inshore and wit21 food supply controlling extraction techniques offshorc \\raters). and raw materials controlling technology. In other 2. Sea mammal hunting onshore (active words, ainouilts of food directly limit (-) or beach, low terrace, islets, or rocks). expand (+) the intensity of food exploitation 3. Bird hunting on water (inshore, offshore, through the eight systems. Wit11 the exception of lagoons, and inland lakes). system 8, ranr materials are usually derived from 4. Bird hunting on nesting sites (sea cliffs and the by-products of food ani~nalsextracted through other island margins). the remaining se\ren systems. Therefore feedback Fislz i~zg: (0) to technology shows the flo\v of animal 5. Fishing offshore (offshore \\raters from by-products to finished tools rather than suggesting boat). that a shortage of mur materials will intensify 6. Fishing onshore (stream mouths and in- extraction systems to provide thosc supplies. Prob- shore waters wit11 hook and line or nets). ably sinall anlounts of food and/or raw materials Gntlterbzg: derive from exchange or trade of a formal or 7. Intertidal and beach collecting (active infor~nalnature. This input and output is dia- bcacl~csand intertidal zones or reefs). gramed to reflect such flo\vs of goods (see Ci~ltural 8. Onshore collecting (sea cliffs and island Patterns and Interaction section). interiors). The double arrows between habitat and tech- Figure 7 illustrates a compositc habitat profile niques reflect the interaction between these t\vo indicating where the systenls interact. Figure 8 elements; the effects of nlan's exploitation of the illustrates these systeins as part of a total ecosys- various habitats are outlined belo\\,. tem equilibri~unmodel; likely causes of disequilib- The population of a village or community, riu~nare also indicated. l~o~vevertransitory, is largely a function of food 7. INTERTIDAL AND BEACH I. SEA MAMMAL I HUNTING 8. ONSHORE bOLLECTIN.. r. I COLLECTING .--.~ ~ ..- UkkSHUHE Ber~le~,ro015,51olk5;bo1kelgrosr; Shellf~bh.~r~h~n~.~~lopu~; marme o1goe:dead sea mammols Whales.rea Iho?r,reols, blrd eggr;rtones for orlifocfs; i ~eootters mineral$ I washed anhore;dnflwood 3. BIRD HUNTING I I ON WATER L. 1.5tA MAMMAL HUNTING ON SHORE I Seolr.reol~ons. I roielv ~ennt~crs 6. FISHING ON SHORE ! OL Fig. 7-Composite profile of an island edge showing habitat locations of the major procurement systems described in the text. and ra\v materials to meet the needs of sustenance, variatiolls will occur from island group to island shelter, and mobility. An important consideratioll group, but differences are considered here much in our model is that isolatioll from external less sig~lificantthan gcileral uniformity. Because cultitral influence and insulation from vagaries in the Rat group constitutes a major segment of this food supply (based on relatively stable ~narine archipelago, the presumption here is that pre- habitats) translate into a relatively stable human historic Aleuts in that group participated in these population. The human carrying capacity is high as systems as much as did any other island group are all Pacific coastal groups because of abulldatlt population. marine resources. Population size will also reflect The archaeological record at least shows the the minimum food supply rather than the maxi- use of the procurement implements included in mum amount available during the most productive Table 5 rather than some radically different tool season, or potential population adjusts to the assemblage. We cannot, as yet, derive total systems minimal food supply (latc wi~ttcrand early spring) from the archaeological records. However, since to sustain the population through the annual cycle. the Rat Island artifacts, features, settlement pat- The effects of positive feedback are therefore terns, habitats, and species correspond closely to restrained by habitat and organism limits. those of other parts of the chain, we can safely Disequilibrium or a decrease in population can assume that past exploitation techniques also result from three areas of influence. Such inter- conformed to the Aleutian subsistence pattern, related physical phenomena as ashfalls, earth- which is the total of these eight individual systems. quakes, and tsunamis, characterizing the Pacific Ocean rim, can quickly interrupt normal marine Rat Island Food and Material Resources and terrestrial ecosystems by altering habitats or causing massive die-offs. Bad weather, certainly a Food. Consider the food limitations of aborig- relative term in the Aleutians, can restrict exploita- inal Aleuts. [See Zimmerman (1963, p. 57) for tion by keeping Aleuts off the ocean surface; or similar native food restrictions in the Hawaiian slo~vto rapid and transient to permanent shifts in Islands as typical of other Pacific island groups.] animal ranges can affect available food and raw- There are no terrestrial mammals found west of material resources. Finally, diseases, group acci- Unimak Island and the tip of the Alaska Peninsula dents, and warfare can affect population directly besides foxes, lemmings, mice, and shrews (Clark, by killing natives or by indirectly making it 1945, pp. 48-49; hslurie, 1959), and even these impossible for dependents to survive without pri- were not to be found in the Rat Islands. No mary providers in the family. Unfortunately evi- domesticated crop plants were known, and the dence of such disequilibrium is difficult to detect damp climate precluded storing dry plant products in the archaeological record. The most massive and for long periods. Unlike the Alaskan interior, best-known state of disequilibrium resulted from which has relatively high summer temperatures, the Aleut-Russian contact of the eighteenth century small island masses of the Aleutian chain are when disease and genocide were particularly acute. dominated by the cold north Pacific Ocean waters Table 5 summarizes the major techniques and with summer temperatures of about 4OoF. Am- implements used in procurement. This table places chitka is about the southernmost point in Alaska, many archaeologically derived artifacts into ac- but the sea environment negates any appreciable tivity spheres or contexts which stress their use summer warmth. Lolv insolation may be equally as rather than form. Figure 8 shoxvs the kinds of important as temperature in limiting plant growth. products that result. The few native plants that do cover the islands Rat Island Syste~ns. I-Iow well do these eight provide little caloric input, and these are only available during the summer months. systems describe prehistoric subsistence and eco- logic exploitation in the Rat Islands? They appear Thus sea resources, directly or indirectly, be- to be the best models we have at present. come human food staples. Sea mammals, fish, and Note that the citation examples in Table 5 span marine invertebrates variously provide meat, blood, 200 years since Russian contact aud specifically and fat. Marine algae are a source of vegetable describe procurement techniques from throughout matter. The most important birds by size and the chain. There is a preponderance of examples abundance are those dependent on marine orga- from the eastern islands for \vhich we have more nisms. Terrestrial plants, as mentioned above, are sources, but \\re find subsistence uniformity controlled by a maritime climate. Survival in the throughout the chain and over time largely because chain, then, depends on some mix of hunting, the physical environment and biota are fairly fishing, or gathering these marine-related organisms uniform throughout. Of course, minor abundance over the annual cycle. Prehistoric Hutna~tOcct~patio~l of the Rat Islnttds 81 Table 5-Procurement System Techniques and Implements 1. System: Sea-mammal hunting offshore Techniques Sight and surround animal with boats; tire diving animal (except large whales); harpoon and spear animal or club it in water; large whales wash ashore when dead, but smaller sea mammals must he dragged or carried to shore. Implements Kayaklumiak (paddles and oars, sucking-tube bailers, sponges, gut kamleika suit,* wooden visor hat, stone ballast, and other boat accoutrements). Harpoon (toggle head with end blade or point, loose shaft, socket piece, wooden shaft, long thong line, and bladder float or nontoggling but detachable harpoon head with barbs, sometimes with end blades or points, socket piece, wooden shaft, long thong line, and bladder float). Spearjlance (nondetachable barbed head, often with end blade or point, socket piece, and wooden shaft). Harpoon dart (nontoggling harpoon head, socket piece, wooden shaft, some with feathers, and thong line between head and shaft; propelled rvitlt throrving board). Club. References Cook,Clerke, and Gore, 1818, p. 37; Coxe, 1787, pp. 62,87, and 120; Elliott, 1886,pp. 140-141, 151-152; Heizer, 1943, 1960, p. 134; Jacheison, 1933, pp. 11, 24-25, and 55-58; Langsdorff, 1817, pp. 340-343; Ransom, 1946, p. 614; Sarichev, 1807, p. 74; Turner, 1886, pp. 200-201; VanStone (Choris), 1960, pp. 155-156;Veniaminof, 1840, Vol. 11, pp. 105-106, 132-133, 217-225, 231, 245-248, 342-344, 388, and 395-396. 2. System: Sea-mammal hunting onshore Techniques Surprise seals, sea otters, or sea lions on island shore or on islets or rocks; hunters approach by foot or boat, depending on locale; kill animals by harpooning, spearing, or cltthbing; possibly nets for entangling used prehistorically; decoy seals. Im~lements Kayaklumiak (as in sea-mammal hunting offshore). Ilarpoon, lance and/or spear (as in sea-mammal hunting offshore) Club. Net. (Decoy). Refcrenccs Elliott, 1886, pp. 142-144 and 366-368; Veniaminof, 1840, Vol. 11, pp. 344-345, 362-363, 383-384, and 388, 3. System: Bird hunting on watel Techniques Stalk birds resting on sea surface or lakes or surfacing; capture birds by darting with leistert or shooting with arrow; kill birds by breaking their necks; secondarily, net birds on lakes from blind. llnplements Kayaklumiak (as in sea-mammal hunting offshore). Leister darts (multiple-pronged wooden darts propelled with throwing board). Bow and arrow (barbed head or point, usually with no end blade, wooden shaft with feathers; short wooden bar\.). (Net with trip line). References Cook, Clerke, and Gore, 1818, p. 37; Coxe, 1787, p. 215; Jochelson, 1933, p. 11; Sarichev, 1807, p. 74; Turner, 1886, p 140. 4. System: Bird hunting at nesting sites Techniques Sheer bird cliffs approached by boat from below or from above by rope; snares set at burrow entrances or on ledges where birds perch; strangled or snared birds collected later; or birds clubbed on nest or caught with hands and killed by breaking their necks; bolas, handnets, and leisters used to down flying birds around rookery; birds also captured away from the nesting areas. Implements Kayak and/or umiak (as in sea-mammal hunting offshore). Ropes of thong or seaxsced. Collecting container of woven grass or stretched gut. Snare. Club. Bolas (stones and thong). lland net/loose net. References Coxe, 1787, p. 86; Elliott, 1886, p. 209; Jochelson, 1933, pp. 8, 10, and 53; Sarichev, 1807, p. 70; Staehlin, 1774, p. 28; Tmner, 1886, pp. 117.1 18, 130, 137, and 141; \'anstone (Choris), 1960, p. 155. 5. Systen~:Fishing offshore Techniques Fishing from boats with hook and line or sometimes spearing with leister spears; club used to kill large fish (halibut). (Table continues on page 82.) Table 5-(Co~ltixmed) Implements Kayak/umiak (as in sea-mammal hunting offshore). Pish line (bone one- or two-piece fishhook, long seaweed and/or kelp or sinew line, stone weight). Leister spear (multiple prongs on wooden shaft). Club. References Cook, Clerke, and Gore, 1818, p. 37; Coxe, 1787, pp. 82, 87, and 215; Elliott, 1886, pp. 168 and 212-213;Jochelson, 1933, pp. 5, 8, 11, and 5 1; Langsdorff, 1817, pp. 333-334; Turner, 1886; Veniarninof, 1840, Vol. 11, pp. 248-249. 6. System: Fishing onshore 'Techniques Fishing with hand lines from shore; catching salmon, Dolly Varden trout, and other massed fish at stream mouths with nets, leisters, and with hands; dams and weirs sometimes used. Implements Net of sinelr, or vegetable material (large nets used with stone weights and possibly wooden floats). 1:isbing line (as in fishing offshore); sometimes used wit11 bladder floats for sets. Leister spears (as in fishing offshore). (Kayaklumiak possibly used prehistorically in setting nets out from a beach). References Cook, Clerke, and Gore, 1818, p. 37; Coxe, 1787, pp. 62 and 82;Jochelson, 1933, pp. 7, 11, and 51-52; Ransom, 1946, p. 608; Sarichev, 1807, p. 7l;Turner, 1886; Veniaminof, 1840, Vol. 11, pp. 248-249. 7. System: Intertidal and beach collecting 'Techniques Combing exposed reefs and intertidal shores far urchins and other marine invertebrates and algae; combing active beaches for washed op sea mammals, birds, fish, and driftwood. Implements Prying implements far removing shellfish and invertebrates from rocks. Grass baskets or stretched gut containers (adze, maul, and wedge for sectioning logs on the spot). References Elliott, 1886, pp. 214-215; Jochelson, 1933, pp. 8 and 11; Ransom, 1946, pp. 609.611 and 616-617;Staehlin, 1774, pp. 29 and 36-37, 8. System: Onshore collecting Technique Collecting terrestrial plant materials, bird eggs,: stone tool materials, minerals, etc., from specific localities landward from the active beach. Implements Baskets, nets, and other containers. Cord:!ge (sinew or grass for bundling gathered materials). hlattock (bone head hafted to vpaoden handle). Unhsfted probe or digger. hlaul (stone hammerhead hafted to wooden handle; for breaking up stone materials). References Cook, Clerke, and Gore, 1818, pp. 39-40; Coxe, 1787, pp. 62, 69, 86, 196, and 213; Elliott, 1886, pp. 148, 168.169, 181, and 222; Jochelson, 1933, pp. 7-8, 10-11, and 59-63; Langsdorff, 1817, pp. 334-335 anrl 343; Ransom, 1946, pp. 613 and 619;Sarichev, 1807, p. 70;'T\lrner, 1886; Veniaminof, 1840, Vol. 11, pp. 94-95, 109, 229, and 232-233. *\\'aterproof parka of stretched gut. tSpcar with multiple side or end prongs. $It may be argued that bird egg collecting from bird colonics on shore should be placed under bird hunting at oestingsites rather than here, espr-cially if technirlues of habitat esploit+~tionare stressed. I have sixlllinarized the Aleut food sphere become especially important for survival into the elsewhere (McCartney, 1975,p. 315) and restate the next productive period. Part of this low ebb of conclusions here. Besides the fact that food staples food is due to inadequate storage techniques for arc of marine origin, with sea mammals, fish, and preserving seasonal foods into later months. Fat or birds providing the greatest input, most of these oil from sea lnalnnlal blubber is a necessary part of foods are also transient and are only found at the cliet; a constant diet of lean meat or vegetable pa-ticolar tinles during the year. Because of the products does not provide the requisite caloric f~tel cyclical and soi~letimesi~ligrating nature of the for survival. Finally, the annual dependency on fauna attd the sharp seasonality of the flora, there nlarine resources froill slllall sltellfish to the largest is an uneven food supply from one month to the \vltales plus such incidentals as land plants demon- next. At the poorest month (about March), "crisis" strates holv Aleuts had to exploit their total foods, such as shellfisl~, sea urchins, and algae, arcllipelago en\~lrolllllellt. ' hlany of the above characteristics may also p,loitation throughout the chain and, by iniplica- apply to Northwest Coast Iudian groups. Accord- tlon, the uniformity of techniques required in ing to Piddocke (1965, p. 247): procuring these species. . . .the abundance of the resources of the K\r,akiotl has Materials. Aleuts used a wide range of raw been some~vhat overestimated and its significance materials for their tools, clothing, decoration, misinterpreted. It was great enough to support a population larger than the usual size reported for boats, houses, and other equipment for adapting. hunting and gathering societies; but this population lived Three major categories \\it11 types of material and sufficiently close to the margins of subsistence so that some finislted products are given in Table 7. variations in productivity which fell below normal Archaeological relnaius are usually stone, bone, could threaten parts of the population with famine and and tooth, and only occasionally water-soaked death from starvation. \\rood (for wood, see Cook, Dixon and I-Iolmcs, Just as "starvation \\,as no stranger to the 1972;Turner, 1973, p. 9). I<\vakiutl" because of "prolonged periods of bad Besides driftwood and stone (systems 7 ancl 8), ~veatherwliicli prevented llunting and fishing, and most of these ra\v materials for manufacturing are the failure of fish runs," so Aleutian starvation \\'as primarily from sea niammals and secondarily from knolvn for much the same reasons (I'eniaminof, birds. An esample of tlie intensive utilization of 1840, Vol. 11, p. 186, cited in Pctroff, 1884, sea lions ancl resulting materials is described by p. 151; Jocl~elson, 1925, p. 43; 1933, pp. 8 and Elliott (1886, pp. 369-370). 11). Fauna from only one Aleutian site has been Island Resource Locations. Thc Rat Islands studied for quantitative estiniates of food con- vary today in richness and abundance of species, sumed. Demiiston (1972, pp. 180-220) calculated and this disproportionate distribution undoubtedly tlie amounts of protein, carbohydrates, fats, aud occurred in the past. This section outlines tlie kilocalories available from the Ashishik Point inajor resources available on the \rarious islands as a (Umnak) faunal collection. She also computed the suggestion of probable prehistoric exploit at' 1011. relative amoinits of food that could have been However, several consiclerations should be kcpt in derived from tlie major types of anilnals repre- lnind in rcviening resource localities. sented in her excavation units. \\'eights of edible 1 e cannot presume that nlodern faunal products derivable from invertebrates, birds, fish, distribution esactly matches that of several thou- and sea mammals were in the ratio of about sands of years ago. hlodenl distribution data, 1 : 2 : 36 : 52, respectively. This ratio does not holvever, provide tlie best iildication of past fauna necessarily hold for tlie relative amounts of pro- available. Since aniliials pursued are habitat spe- tein, carbohydrates, and total calories supplied by cific, we kno\v that there is no random distribution the different animal groups, but still it is a more of sea mammals, fish, birds, invertebrates, algae, precise estimate of the Aleut diet than that 'and terrestrial plants today or in the past. R'Iodern suggestecl by the ethnol~istoricliterature (McCart- habitat locations are assunled to correspond closely ney, 1975, pp. 293-295). This ratio of to past locations on the same island. 1 : 2 : 36 : 52 for thc various food sources is our Population sizes for most of the species best estimate of relative source importance for the hin~ted,fished, or gathered have not becn drasti- entire chain, including, of course, the Rat Islands. cally reduced from the prehistoric period with tlie Table 6 lists the identified faunal organisms exception of \vhalcs and pelagic fish, ~vhicharc from Rat Island sites and from Near and Andrea- taken by the high-seas fisheries. Sea otter popula- nof sites for comparison. Most of tlie faunal debris tions have rebounded in the Rat Islands since they has not been studiecl q~~antitatively,nor have all were placed under protection in 191 1, and l~arbor lo\\,-frequency species been identified. Sea 111a111- seals and sea lions have not been exploited for pelts mals and birds particularly were used not only for in a way conlparable to the controlled fur seal food but also for ranr materials (skins, feathers, harvesting in the Pribilofs. Except for islands on gut, bones, teeth, ctc.), and some species may have .rvhich foxes are fotuid, bird colonies are relatively been gathered Inore for one requirement than for free of disturbances. h,Iarine invertebrates, algae, the other. A comparison between these Rat Island and terrestrial plants are thought to be essentially species and a more coniplete listing of Aleutian tuncliangetl since the prehistoric period. Bec;~use species known to have been utilized aboriginally Aleuts have not depended on these tilajor food (R'IcCartney, 1975, Table 1) shows a high correspon- animals for over half a century and because the deuce. All species considered to be important Aleut population over that period was but a slilall because of their size and/or number are foulid in fraction of thc prehistoric population size, Aleuts the few Rat Island sites sampled. \\re find substanti- have probably uot seriously affected fau~laltotals ation tlielefore for tlie uniformity of faiuial ex- in recent years. Table 6-Identified Faunal Remains from Rat Island Midden Sites Rat Island sites Comparative sites Amchitka Site 31 Little Kiska and Amchitka Attu, Amchitka, Attu and Atka (Desautels et al.. 1971, (Hrdli'cka, 1945, Little Kiska Adak, and Kiska (Jochelson. 1925, Species pp. 314340) pp. 613-616)* (Friedman, 1937)f (DaU, 1877, p. 50)$ pp. 25.27-29)s Sea mammals Sea otter (Enhydro lutrir) 59% of sea mammals (744)q X (common) L. Kiska** X Attu, Atka Hair or harbor seal iPhocn vitulinai 33% of sea mammals (418) X (most abundant) L. Kiska X X Atka Northern sea lion (Eumotopias jubata) 6% of sea mammals (79) ' X (common) L. Kiska X X Attu. Atka Northern fur seal (Callorhinus urrinus) 2% of sea mammals (30) X (common) L. Kiska Walrus (Odubenus rosmarus) X (1 bone) L. Kiska Pacific harbor porpoise (Phocoena vomerina) X (occasional) L. Kiska Dall porpoise (Phocoenoides dalli) X (occasional) L. Kiska Whales X X X Attu. Atka Land mammals Dog (Canis familiaris) X (5 innominates) L. Kiska Fish Rock greenling (Ilexopmmos lagocephalus) X (abundant) X Attu Red Irish lord ilfemileoidotus hemilepidotus) X (abundant) Pacific cod (Codus mncrocephnlus) X (abundant) X Attu, Atka Pacific halibut (Nippoglorsus stenolepis) X X Attu, Atka Rockfish (Sebastes sp.) X Salmon (Oncorhy~rcusSD.). . Unidentified herring, sculpins, and flounder X Atka Marine Invertebrates (in descending order of Sitc 31 abundance) Green sea urchin (Strongylocentrotur sp.) X (very abundant) Kiska X Attu. Atka and L. Kiska Limpet (Acmaeo scutum) X Amchitka X Attu, Atka Limpet (Acmoen pelto) X Amchitka X Attu, Atka Blue mussel (Mytilus edulis) X Amchitka X Attu, Atka Giant horse mussel (Modiolus modiolus) X Attu Mussel (~Modiolusuul,?aris) Butter clam (Saxidomus giganteus) X Atka Bent-nosed clam (Macoma nnsutn) Clam (Mactm ponderoso) X Attu, Atka Pearly monia (Pododesmus macroschisma) Black katy (Knthnrina tunicotn) X Attu tiairy triton (Fusitriton orgeonense) X Sitka periwinkle (Littorinn sitkana) X X Amchitka X X Attu Channeled dogwinkle (Thais canaliculata) X ? (Thais lima) X Amchitka X Island scallop (Colomys islandicils) X Northern opal-shell (Epitonium wroblewskii) X Amchitka Rock cockie (~roto&cn stamineo) Whelk X ? iVolutoosius rebvlaris) X Amchitka ? (~olliseilapelta) Barnacle (Bolanus sp.) X Birds (by family, in descending order of Site 31 abundance) Phalacrocoracidae (cormoranu) 37% of first 6 families X (occasional) Amchitka 24% of bird bones X Attu. Atka (250) 11191,---, Anatidae (swans, geese, ducks) 35% of first 6 families X (abundant) Amchitka 34% of bird bones X X Attu, Atka 17151 \---, (185) Alcidae~ ~ (murres. murrelets. auklets. 11% of first 6 families X (rare occasional) 22% of bird hones X X Attu. Atka puffins) (77) Amchitka (118) Procellariidae (shearwaters, petrels) 7% of first 6 families X (occasional) Amchitka X (2 bones) X Attu Diomedeidae (albatrosses) X (common) Amchitka 12% of bird bones X X Attu, Atka (68) Laridae (gulls, terns) X (rare) Amchitka 3% of bird bones X X Attu ,--, (16) Gaviidae iloonsi X (less than 4% each) 1% of bird bones (6) Ln Podicipedidae (grebes) X Falconidae (falcons) X Accipitridae (eagles) X 1% of bird bones (3) X Corvidae (ravens) X X (rare) Amchitka 3% of bird bones (15) Strigidae (owls) X (I bone) 0 Fringillidae (sparrows, longspur, X Attu C:5: buntings) - -J w2 *Little Kiska Site at west end of the island, sampled in 1936: Amchitka Sites 2 and/or 3. sampled in 1938: birds = 10 species. +Friedman's Little Kiska sample is that excavated by Hrdlirkain 1936: 34 species: R = 1.132: the following 7 species total 408 bones: (75%) in descending order: Pelagic 0 Cormorant, Pacific Eider, Short-tailed Albatross, Emperor Goose, Horned Puffin, Crested Auklet, and Tufted Puffin. 2 :Dall does not specifically identify the source of his material: at least some comes from Amchitka and/or Kiska: with the exception of urchins. the remaining * invertebrates "do not form altogether more than one.tenth of onc per centum of the stratum." 5. SJochelson's specimens from Attu and Atka: frequencies not specified. - YIndividual bones, not individual animals; whale fragments not included in counts. **No details about Hrdlirka's Amchitka faunal samples, but sea otter and harbor seal are the most abundant sea mammals. e, -+> +?Turner (1970. p. 121) also notes the abundance of albatross bones in Amchitka middcns. (D Thc modern distribution and ccnsus figures for seeking the best hunting and fishing waters (sys- these iulimals and plants, then, ought to be a clue tems 1, 3, 5, and 6), not the land masses with the to their location and relative abundance in the most organisms on shore. \\'hales and porpoises are past. alxvays water bound, and migrating fur seals rarely haul up ~vhilcmoving through the chain's passes. Sea otters and harbor seals are so wary that they are less frequently surprised on shore than en- Table 7--hlaior Raw hlate~.ialsand Inlportant Prodtlcts countered in the water. Because these species Vegetable caninot be fixed to an island but rather are fixed to Driftwood specific water habitats, we realize that shore counts Boat frames, house frames, burial sarcophagi, harpoon can only suggest abundance for a surrounding shafts, clubs, bowls, trays, spoons, handles, throwing water space or for species procured from shore boards, masks, and some firervood. Grass (systems 2,4, and 7). Baskets, mats, cordage, nets, cloaks, house co~terings, Table 8 summarizes the present-day distribu- linine.". and insulation material. Twigs tion and abiundance of the major Rat Island Fuel and coshionine. resources that were important items of s~tbsistcnce Kelpjalgae for prehistoric Aleuts. Becausc not all the islands Fishing and other lines. have been surveyed wit11 equal intensity for compa- Other rable specics, only a few geileralizations can be Dyes and medicinal herbs and plants. Animal derived from the present quantitative data. Occur- Skins rencc only is considered in Table 8, and seasonal Clothing and boat coverings. variation is ignoreci; maximum animal population Gutjinternal organs estimates are given rather than minimums or Kamleikas, windows, containers, drumheads, and har- means. For the inost part, all these species are poon floats. recognized in prehistoric midden clebris (Table 6). Sinerv Nets and cordage for lines and lasbing. A few generalizations are as follows: (1) The Bones best sea lion islands are, in order, Buldir, Semi- House frames, boat pieces, tools, rveapons, spoons, and containcrs. sopocl~noi, ICiska, Amchitka, and Rat. (2) The Teeth inost sea otters are found around Amchitka, Kiska, Tools, decoration, and gaming pieces. and Rat Islands. (3) The best harbor seal island is Feathers probably Amchitka. (4) Fur seals inay have been Decoration. more abundant on and around Buldir if that Whiskers Decoration. suggested rookery is accurately identified Blubber (Tablc 8). (5) Rlost colony birds, except puffins, Lamp oil and waterproofing for skins. xvhich center on IChvostof and Daviclof, and cormo- Blood rants, ~vhichccnter on Amchitka, are concentrated Glue. Inorganic on Buldir. (6)Duck ancl geese estimates are avail- Ocher and/or other minerals able only for An~cl~itka,but, because these birds Paint. primarily occupy littoral waters or shores, they Sulfur should be most abundant on islands wit11 the Fire starting. greatest perimeter [Amchitka (195 km) and IGska Stone Tools, bowls, lamps, weights, ounaments, fish weirs, (16 1 km) stand out in shorcliile lengtll] . house fountlations, boat ballast, and storage-pit liners. (7) Stream runs of saln~onand Dolly Varden are probably most abundant on Amchitka, Kiska, Rat, ancl Semisopochnoi, roughly in that order; no comparative measures of major inshore and off- shore fish species are available. (8) R,Iajor differ- 2. Island size and species abundance are not ences in wvhale, fish, invertebrate, algae, terrestrial ahvays proportiollal to one another. Sufficient plant, and drift\soocl distribution are unknown; the evidence of this is seen in the size of Buldir's sea last four could be grossly comparecl by island size lion colony or Kh\~ostof's and Davidof's puffin and circumference. (9) No land mammals, with the colonies (see Tablc 8). possible exception of dogs, are kno\vn on my of 3. Because fish and sea lllamnnals are the t\vo the islands preceding Russian lats and later intro- most important food sources, we can only partially ductions. (10) Although sulfitr was limited to estimate their abunclance by obselving animals Little Sitkin, IUlr,ostof, and Davidof, \~olcanic from shore. In other \vords, Aleuts were often ejecta is more abundantly distributed; basalt alcl Preltistoric Htrvinn Occtcpation qf the Rat Isln~~rls 87 Table 8-hlaior Rat Island Resources Organic Resources Sea mammals Gulls: Total Rat Island estimate. 6500t: Buldir. 5000: Sea lions: Estimates given in descending order. Buldir, Amchitka, 3000 (\Villiamson and Emison, 1969; 3450; Semisopochnoi, 1590; Kiska, 1470; Amchitka, Sekora, 1973). 900; Rat, 570; Little Sitkin, 335; Segola, 250; Rat Cormorants: Total Rat lsland estimate, 2400; Am- Island total, 8565; major rookeries fairly evenly chitka, 1600; Semisopochnoi. 700 (\\'illiamson and spaced around any one island; many offshore rocks Emison, 1969; Sekora, 1973). and islets have sea lions; usually stay in waters of Fulmars: Total Rat Island estimate, 250 (Sekora, <50 fathoms (Kenyon and Rice, 1961; Kenyon and 1973). Kine.". 1965: Sekora. 1973). Ilarlcquin Duck: Amchitka total, 3500 (l\'illian~sonand Emison, 1969). Sea otters: Estimates for 1965 given in descending Common Eider: Amchitka total, 1500 (Williamson and order. Kiska, 1652; i\mchitka, 1525; Rat, 435; Emison, 1969). Semisopochnoi, 271; Little Sitkin, 180; Segala, 75; Emperor Goose: Amchitka total, 5000 (l\'illiarnson, Klivostof and/or Davidof, 52; Buldir, 30; other Emison, and White, 1971). Ainchitka estimates: about 1961, 4000 to 7000; Common Goldeneye: Amchitka total, 4000 (\\'illiamson 1970, 4000; 1970-1972, 7601; Rat Island total, and Emison, 1969). 4196t; Rat and Delarof Islands total, 23,000 to Bufflehead: Amchitka total, 2000 (\\'illiamson and 32,000; almost all islands have sea otters offshore, Einison, 1969). especially in waters of <30 fathoms and \vhere kelp Rock Ptarmigan: Amchitka total, 1800 (\Villiamson and beds are extensive. (Kenyon, 1969; Spencer, 1969; Emison, 1969). Estes and Smith, 1973; Sekora, 1973). ..>..n:",. Harbor seals: Amchitka has an estimated 1200; other Salmonids: Approximate number of streams potentially Rat Islands combined have 1600; usually found in suitable for salmon, in descending order: Amchitka, waters of <30 fathoms (Sekora, 1973). 30; Kiska, 27 plus 2 lagoons and 3 landlocked lagoon Fur seals: Rarely haul up in the Aleutians when lakes; Rat, 7; Semisopochnoi, 7; Segula, 3; Little migrating through to and from Pribilofs; only knotvn Sitkin, 3; Little Kiska, 3; Butdir, 1; IChvostof, 1; possible Aleutian rookery is Buldir; only hunted Davidof, 0; Amchitka has 21 streams supporting pink offshore in other areas; most fur seals use eastern and a few coho salmon plus Dolly Varden; red salmon rather than western island passes (blurie, 1959; are known from Semisopochnoi; Kiska supports coho Sekora, 1973). and probably pink salmon plus Dolly Varden in at Birds least three areas (USGS Bulletin 1028 maps; Neuhold, Anklets: Total Rat Island estimate, 631,000; individual Helm, and \'aldez, 1971; A.C. Hartt and C. A. island totals given in descending order: Kiska, Simenstad, personal communication, 1974). 125,000 to 400,000; Segula, 70,000; Buldir, 40,000 (Sekora, 1973). Inorganic Resources BIurres: Total Rat Island estimate. 30.600: individual Sulfur: Large deposit on Little Sitkin; sulfnrous fun~aroles island estimates: Buldir, 30,000; ~emiso~ochnoi,700 on Little Sitkin, Davidof, and Khvastof; for fire making (Sekora, 1973). (personal obse~vation;Sekora, 1973). Pumice/scoria ejecta: Semisopochnoi, Little Sitkin, Puffins: Total Rat Island estimate, 28,000; individual Dadof, Kiska, Segula, and Buldir probably all have island totals given in descending order: Davidof, surface pumice and/or scoria for abraders and other 7720; IU~ovostof,6540; Semisopochnoi, 2200; Pyra- implements (Sekora, 1973; L. hl. Gard, written con mid Island (near Davidof), 1500; Amchitka, 1300; n~unication,1974). Kiska, 1000t (Sekora, 1973). Lignite: Amchitka; for labrets and other carved ornaments Kittiwakes: Total Rat Island estimate, 40,000; Buldir, (Sekora, 1973; L. hl. Gard, written communication, 35,000 (Sekora, 1973). 1974). argillite have an even wider distribution; no obsid- Because Buldir was so isolated and afforded only iall source is known.* limited coastal living space, ho~\~cvcr,it attracted From these distribtttio~laldata, sketchy as they felv, if any, permanent human residents. The arc, we co~lcludethat sea mammal, fish, and bird interrelated dimensions of island size, circt1111- foods and materials are more abundant on Buldir, ference, and irregularity appear to determine abun- Amchitka, Kiska, Semisopochnoi, and Rat Islands. dance of primary species except for Buldir. Buldir's sea mammal and bird populatio~ls may be a function of isolatio~l~\~hcre scarcity of land con- *L. hI. Gard (written communication, 1974) states that centrates populations. Inorzanic materials as a obsidian is common in the upper part of the Amchitka group are more localized than are organic mate- Formation (see Chap. 2) but is unsuitable for tool manufac- ture because it tends to shatter into small pieces. The rarity rials. t\ resource such as sulfur lends itself to trade of obsidian in Alnchitka artifact collections confirms the because its acquisition can be more closely con- unsuitability of local obsidian. trolled. I~Iuma~iImpact on tlie E~~viro~~~nent (Veniaminof, 1840, Vol. 11, pp. 46-47, 196-197, and 234-235; Jochelson, 1933, p. 11).This is the All the evidence one can muster suggests that Aleutian version of Liebig's law of the minimum. prehistoric Aleut occupation throughout the chain Aleuts never devisecl an effective leveling system to had little lasting impact on isla~ldterrain. Remain- spread the abundance of late spring, summer, and ing testimony of past physical alteration is li~i~ited fall througllout tlie annual cycle. TIlus population to such slowly decaying features as house ruins, leveled off, not becatise marine foods were ex- middcn debris, burial sarcophagi, or mummy bur- hausted but because storage techniques \\?ere insuf- ials in caves and crevasses. If unbuffered by thick ficient and/or weather precluded the normal calciferous dcposits (urchins, shellfish, n~ammal, dependence on hunting and fishing in marine and bird bones) or not Jvaterlogged, organic mate- habitats. In other words, confinement to land was rials quickly decay, leaving little or no residue. very restrictive to a people used to extracting food, Thus the most apparent relics of past occupation mainly sea mammals and fish, from the sea. one notes today are the coastal midden sites, which Population size probably never rose to the point are covered 514th luxuriant vegetation because of that major staples were being overly taxed. their added soil nutrients. As pointed out in the introduction, the ancient village remains are rather 2. Given the luitnting, fislting, and gathering insignificant compared \crith modern excavatiolls subsistence tecliniques and their prinlitive if effi- and land-surface alterations. cient technology, it would be extremely nnlikely that prehistoric Aleuts could have caused drastic Unlike Eskimos of the Canadian Arctic, xvho reductions in animal or plant resources. Sea mam- built so many features of stone (e.g., burial cairns, mals from sea otters to whales \Irere abundantly boat supports, markers, meat caches, etc.), Aleuts available at Russian contact. There were no pan- rarely used beach boulders except for house-\\,all Aleutian endangered species until the promysltlell- foundations. Drifttvood \\'as the abundant raw nik (Russian fur trapper and trader) arrived. Fish building material, and, unless co\rered with sod and and algae were essentially limitless given the uraterlogged, pieces xvould soon blo\v away or techtlology iused to extract them. Perhaps over- decay. i~ldulgellce in egg collecting or bird snaring at a Aleut impact on past biota, ho~ve\,er,is more colony site or o~'ercollecting sea urchins and difficult to evaluate and can only be postulated. shellfislt might cause short-term declines in those The first colonizatio~~by ~nanof the chain, as yet shore localities, but such fauna would probably poorly docume~ltedarchaeologically, undoubtedly rebound rapidly if left alone for several years. had a significant impact on various species in- Terrestrial plants never figured too importantly in corporated into the contemporary procurement the diet, and only grasses constitutcd a major systems. Nonetheless, nre have 110 pre-man baseline ran-material resource. The majority of plants used against which to judge man's initial effects. If had leaves, stalks, twigs, or berries removed with- species easily preyed on by man were present in out disturbing the root structure. Therefore thcse the western Aleutians, suc11 as the sluggish nortll- plants were not removed from their respective ern sea co~vand the flightless spectacled cormo- plant communities. Only a few plants provided rant, both fonncl in the Commander Islands at roots (e.g., the ICamchatka lily, F~itillaria Russian contact (Stejneger, 1936, p. 351), then ccrn~scllatce~~sis)for food, and only the collectillg Aleuts probably affected their rapid demise. of these plants \\~oulclcause their removal. halore ililportant than initial impact was the 3. Aleuts depended heavily on migrating ma- continuing interaction between Aleuts and avail- rine species for food and materials. \\'lvales, fur able species for several millennia since. In general, seals, many birds, and salmo~lidswere available in a archaeologists consistently see stability and equilib- cyclical fashion and were not year-rotuxld residents. rium in the archaeological record; rapid populatioll Therefore, although migratory sea mammal, fish, decline or expansion of Aleuts (and hence their and bird populations might be considered finite on supportive species) Itas not been documetlted. an island's shore for ally one day, week, or season, Several conditions relating to populational stability a seemillgly infinite number of these animals and ecologic adjustmellt are germane and are passed through this archipelago filter over the outlined belo\\,. course of a year. Like mainland Eskimos exploiting 1. Resources available during the worst part of migratory caribou, Aleuts sampled from a large the year ultimately regulated populatioll size. The faunal itniverse not bound by sharp limits. The bleakest period was during late \\$inter and early Aleut influence on the north Pacific marine fa~u~la spring when stormy seas prevented Aleuts from was insignificailt ~vlten compared to European- liu~~tingand stored food had been depleted controlled \vhaling, other sea mammal hunting, and Prehistoric H~onanOccupatio~t of the Rat Isla~trls 89 fishing, \vIiich reached this area during the eigh- In sum, prehistoric Aleuts possessed neither teenth century.* domesticated crops nor animals and could only 4. Fitlally, Aleut population stability was both collect food, not produce it. Like other hunters cilhanced and undermined by food sharing as one and gatherers, little control could be exerted over kind of reciprocity (see, for example, Sahlins, potential dictary resources. Over thousands of 1972, pp. 215-219 and 235-236). Prior to the late years of occupation, humatl population cqui- ivitlter food shortage was a period of relative librium was reached on the basis of hunting aild abundance. Winter was the time for large feasts or gathering. Disequilibrium must Iiave been tlie festivals in xvhich whole villages participated. Other exccption rather than the norm; as illustrated in villages from the same or different islands were Fig. 8, such disequilibrium derived from upsetting invited as guests (see, for example, Coxe, 1787, either habitat, organisms, extractive techniques, pp. 199-201; Veniamiiiof, 1840, Vol. 11, pp. 85-91 food products, or population or some combi~latioil and 208). This pattern follows the common of thesc. As participants in a generally stable Eskimo messenger feasts of maiulalld Alaska (see, environment, Aleuts were causing few detectable, for example, Nelson, 1899, pp. 357-363; Spencer, long-term effects that \ve can document today. 1959, pp. 210-228), at which time food consump- Obviously, by their very presence, Aleuts affected tion and sharing is an important function. It is their various habitats and resource organisms on obvious from the following passage from \vIiich they depended. However, these effects were Veilia~ninof (1840, Vol. 11, pp. 85-86) that such illdeed miniinal when compared to the historic festivals paralleled the Northwvest Coast potlatches period and the impact that the rifle, motorized in their spirit of rivalry: boat, and other industrialized technology and The formal festivals consisted of scenic stage exploitation had on that same environment. representations. These were always held in the winter and alternated between one settlement and another. They were organized by the entire population and for CULTURAL PATTERNS AND INTERACTION them each and every inhabitant gave almost all that he had-particularly in the zvny of food supplies. Thus after each festival everyone of them inevitably \vent Considering the several thousand artifacts col- hungry in the full sense of the word. Such a famine was lected on Amchitka, the nonarchaeologist iuight not considered a shameful thing but one of glory. Each \\.ell suppose that we have a firm understanding of settlement, in its turn, sought to outdo each other in ancient Aleut culture. Uufortitnately this is not the the invention of scenes, in the artistry of staging, and in the profusion of their hospitality. (Emphasis added.) case. This section begins wit11 a rdsitmd of ~vliat little cultural informatio~lwe actually have. Al- Therefore sharing available food resources in early though this sitnmary applies to the Rat group, our winter enabled many Aleuts to gain from the archaeological knowledge of the Near and Andrea- accumulatio~~sof a few. Hunger and privation, nof groups is equally scanty if not more so. Only however, were the rewards for feast givers prior to the eastern archipelago has received more intensive the beginning of the spring food cycle. investigations (see, for example, Aigner, 1970; Denniston, 1972; A'IcCartney, 1974; Turner and Tunler, 1974). Even so, taken as a cultural whole, *Hett and O'Neill (1974, pp. 37-38) conoborate the limited effects of man on the Aleutian ecosystem through Aleut prehistory remains poorly known today. their study of a carbon cycle model: ". . . the Aleut This is ironic in light of Dall's haviug initiated population played a relatively minor role in maintaining or Alaska11 archaeology a century ago by excavating affecting the stability of the ecosystem as defined by the aloiig the Aleutian chain. rate of recovery from disturbance. It should be noted that The state of Rat Islalid archaeology can be the Aleut is considered in the context of his natural state where his major effect was as a consumer. These conclu- summarized as follo\vs: sions are quite invalid when applied to civilization's present 1. Only Ainchitka has bee11 adequately sur- influence in the subarctic." veyed for sites; tlie nine other islands, comprising Rappaport (1963) notes a similar insignificant effect of 67% of the coastline, have not been inteilsively native peoples in middle and South Pacific Islands: "On the searched for additional sites. reef, in the lagoon, and in the sea there continued to be relatively unmodified communities, composed of generally 2. Of kno\vn sites, only about 20% have been the same elements in generally the same relationsl~ipsas tested in ally fashion; no one site has been those that obtained prior to man's arrival. The ecological exteiisively excavated with systematic procedures; niche or niches exploited by hnman populations belonging the largest excavations \sere those of HrdliEka on to such communities depended mainly upon the biotic Amchitka (Sites 2 and 3) and Little Kiska, but elements present before the advent of man; the presence, location, and quantity of these elements were subject to these provided no cultural infor~lvatioilbesides the little or no control by human beings" (Rappaport, 1963, artifacts, a few human burials, and a small faunal pp 166-167). sample. 3. Only tltrce sites have been radiocarbon This section briefly revielvs the nature of Rat dated, and all thcse arc on Amchitka (Sites 31, 32, Island collections and the kinds of interaction that and 36). occurred ~vitliinthis isiaiid group and between it 4. Only one house feature ltas bccn u~tcovered and neighboring groups. (Site 32); other featnres include sexreral hearths and burials; all features arc found in Amchitka sites, Rat Island Artifacts ancl none are known from other islands. Because tlte artifacts recovered under AEC- 5. Altltough several human skeletons ha\)e been sponsored projects are now stored in the University found, thesc have not been studied as a serics nor of Alaska museum and could not be studied easily co~nparativelyreported on in detail. in preparation for this chapter, no synthesis of tlte 6. h3ajor Ainchitka collections ivave cither gone 1968, 1969, and 1971 artifacts is presented here. unanalyzed (HrdliEka and Scnse) or ltave been Itisteacl, the Smithsonian collections of HrdliFka st~tcliedanalytically with no coinparative syntheses and othcrs were inspected to evaluate artifact along temporal, spatial, or fuitctio~laldimensions variations within thc Rat Islands. These artifacts (Desautels and Cook). from Amcltitka, ICiska, and Little ICiska constitute 7. Finally, the Rat Island native population \\,as the largest collections besides those of the AEC.* eliminated or removed early cluriiig the Russian The Smithsonian collection has never been ana- colonization, and there is very little etltnohistoric lyzed and pnblished; only a fc~speci~neils appear literature from whiclt to tie together historic and in HrdliEka's (1945) Aleutian report. Fifty-five of prehistoric cultures. the Ainchitka pieces are illustrated in Figs. 9 to 15 of this chapter; all but one of these are front Becausc prehistoric rc~naiiis(sites, burials, arti- Sites 2 and 3. These show some of the variety of facts, fa~uialdebris, etc.) from this island group are inlplements and weapons usccl in procurement and generally similar to those found clse\vhere in the other activities plus the toolmaking skills of an- chain, we vie\\, Rat Islanders as taking part in the cient Aleuts. larger, archipelago-\vicle cultl~ralmilieu. \\'e have Table 9 sunl~narizcsmajor artifact classes as adequate evidence that these past peoples adapted grouped by activit)' spheres or use contests. These similarly to the Aleutian en\~ironment. For tltc classes are takcn from reports and collection tiotcs above reasons, Ito~\~cver,\\re must often speculate coveri~tgall island groups, but tthosc specifically about specific details of the prehistoric past. knox\,n from thc Rat Islands arc marked. Tltis Aleutian archacology may be accurately char- listing attempts to organize artifacts of known acterized as becoming marc oriented to settlement function; artifacts of unkno\\rn function are ex- pattern and to ccologic and systemic approaches cluded. In some cases we cannot be certain that an over tlie past decade since these approaches have implement was used in only one context. Knives developed in contemporary archacology generally aud scrapers, for instance, were probably used in (see, for example, Flannery, 1967; Leone, 1972; multiple contexts on multiple materials. Redman, 1973). Yet tlie organization of Aleutian It is reatlily apparent that almost all these prehistory \\.itltin these kinds of moclels is hardly typical Aleutian ilnptements are known from tlte an accomplislied fact. Archaeologists working in Rat group, indicating strong cultural continuity many otlter geographic regions have made great strides with these approaches because they operate *The U. S. bluseum of Natural Ilistory, Smithsonian \vitltin a \\,ell-blocked-out cultural-historical Institution, is the repository for a total of 2059 catalogued frame. 'l'he major problems of time, space, and artifacts from the Rat Islands collected over the past function have bcci~acidrcssed, but this cultural- century. A breakdown of this total by island is as follo\\.s: liistorical foundation is largely lacking in tlte Amcl~itka,1526 (74%); Little Kiska, 481; Kiska, 34; and Rat, 18. ,\ breakdown by collector is as follows: HrdliTka, ?\lcutians, and \ve are hindered because of it. Unit 1916 (93%); Goggenheim, 34; Dall, 25; and others, 84. concepts (e.g., [\'llley and Phillips, 1958) have Because of Hrdlirka's 1930s field work, the Smithsoniao gone undefinetl ancl have not been used. For has the largest Rat Island collection with the exception of \>;trioas reasons, namely, poor stratigraphic control the recent AEC-sponsored one. The small Amcllitka collection that Guggenheim gave to in many excavations, few radiocarbon dates, few the Smithsonian has been rather picked over, and feu, fine published reports, and nmny mistudied tnaseum pieces are included. It is typical of many collections made collections, Aleutian arcltaeologists have operated by \irorld \Var 11 servicemen in that, by the time they largely in a ~toncomparati~,emanner except at tlte bcca~nemuseum acquisitions, mostly broken bits and pieces artifact level. As a result of all these factors, we remained. Hundreds if not thousands of artifacts left Amchitka with \\'orld War 11 military perso~mel,but very can offer only a sketchy and speculative synthesis few of thesc were later donated to museums for safe- of prehistory for one island group or the entire keeping. One Amchitka exception is a collection no\%.in the archipelagc~. possession of Rockmont College, Denver. .ehisto~icfIzr?nn?l Occtrpntio?t of the Rat Islnnds 91 over space. Examples of classes not yet found in rable sampling techniques, lack of provcnieiice the Rat Islands \\rill probably be found in the specifications, and lack of typological analyses. fnturc. As stressed throughout this chapter, Aleut These are briefly outlined before offering iilterinl lifeways are rather similar \vherevcr noted in the cultural generalizations. chain, and preserved technology is accorclingly Sanlple size is critical because, although Rat similar. Differences do occur, however, at the Island artifacts total about 10,810 (AEC, 8750; typological or stylistic level beneath that of major Smithsonian, 2060), collections from the Near and classes. Andreanof groups flanking the Rat group total much less. The largest museum collcction, to my Cultoral Interaction I\'itl~inthe Rat Group knowledge, is that of the Smithsonian: Near Islands, 2010 (Attu, 410; Agattu, 1600); Andrea- b1uch of the preceding discussion suggests that nof Islands, 456 (Ilak, 23; Icanaga, 84; Adak, 132; Rat Islanders interacted in direct proportion to Atka, 105; Amlia, 112). In addition, 936 other their proxin~ity to one another. Aleuts living Near Island artifacts from Attu, Agattu, and closest together \\~ouldhave shared in a greater Shemya are treated by h3cCartney (1971), in- number of common cultural elenieltts than those cluding those published by Spaulding (1962), living on the most isolated islands. \\'e would Jochelson (1925), and Hurt (1950). Artifacts from expect similarity to be displayed, particularly in any one island, much less any one site, are artifact assemblages, in that style reflects shared relatively few in comparison \\,it11 Amchitka and manufacturing norms. On a single island, such as Little ICiska. HrdliEka's large Agattu collection is Amchitka, \re \vould expect that two synchro- the single exception. nously occupied villagcs close together \vould h(1ore important than merc sa~nplcsize is the display more common cnltilral expressions be- fact that ncithcr HrdliEka nor military personnel tween them than either \crould to a distant village. responsible for other collections retained all esca- Of course, the conlparison of assemblages from vatecl materials. In other words, the collections do nearby sites occupied at different periods !\vould not represent the artifact assemblages originally indicate nothing about cultural sharing. deposited at the sites. Better, complete, and unique It is not possible to test this hypothesis of artifacts xvere selected over poor, broken, and proximity-similarity on Amchitka by merely re- redundant ones. \\'om and broken flake scrapers viexving the t~voAEC artifact reports (Desautels and knives, notched stone weights, and fraglnentcd et al., 1971; Cook, Dixon, and IIolmes, 1972). A bone and stone implements, for cxainple, were complete study of all the 1968, 1969, and 1971 often left in the field. The AEC collections were collections would be required to detect the degree systematically assembled, with attention paid to all to which proximity and cultural similarity are ci~lt~traldebris. These reccnt Ainchitkan collectio~~s related. Even tlien the existing sample may be are difficult to cotupare with the older collections insufficient, and additional artifact samples may from the Near Islands and the Rat Islands (except be required to substantiate this postulate. Some of for Spaulding's Agattu collection) because the the materials are not comparable teinporally. &','lost analyst never kno\vs what \\.as actually excavated of the 1971 artifacts date to about 400 years ago, in these acljacent island groups. \\,hereas the 1969 artifacts span a long period, Besidcs making inconiplete collcctions, beginning about 2500 years ago. I-IrdliEka and nlilitary collectors failed to assernble Furthermore it is not possible to adequately provenience measurements, detailed notes about test this proximity-similarity hypothesis between site context, or relevant maps, drawings, and islands because the only sizable non-Amchitkan photographs. \\re have essentially no inforination collcction is that from Little Kiska. A detuonstra- about miclden stratigraphy and features 01. \\'here tion supporting this hypothesis would requirc the collected artifacts originate vertically or hori- additional collcctions from sc\reral different zontally in the illidden matrix. ICnowing relative islands. artifact positions \\,oultl make these western island collections much marc irseful than they arc cur- Cultural Interaction Betx\xeen Adjacent rently. Island Groops Finally, no one has classified central or \vestern Aleutian artifacts in either a more traditional Coniparative Limitations. If comparisons typological manner or through cluster analysis of within the Rat Islands are limited, con~parisons attributes. This condiiion prevails for the majority between major island groups are little better. of all Aleutian sites. e are thus left \\'it11 Limitations in making typological and st)~listic individual artiracts to conlpare rather than artifact comparisons include small salnple sizes, noncompa- clusters of demonstrated cultural significance. ('rest continues on page 107.) Fig. 9-Typical Amchitka artifacts (U. S. hluseum of Natural History, Sites 2 and 3). Ifunting implements: a-g Chipped-stone projectile points or end blades for use in arrows, darts, harpoons, or spearheads such as h-k (basalt and similar stone). (Catalog numbers 396219, 396215, 396289,396215,396215,396289, and 396289.) 11 Bone arrow or dart head wit11 conical tang and slotted tip for a stone projectile point. (Catalog No. 396130.) i Bone spearhead with missing tip; the plug tang was inserted into a bone socket piece and probably secured with thong lashing through the basal hole; the original tip was either slotted or had an open basin for a stone projectile point. (Catalog No. 386125.) j Bone spearhead with basal lashing lip, slotted tip for a stone projectile point, and an owner's mark incised into one face for identifying sea mammal kills; this head was probably lashed directly to a wooden shaft. (Catalog No. 396138.) k Bone spearhead with slotted tip for a stone projectile point and finely incised line decoration; this late style head was embedded directly into a wooden shaft end and lashed around the tang notches; possible use as a fish spear instead of a sea mammal spear. (Catalog No. 387673; this specimen only from Site 20.) Prehistoric Huntan Occz6pation of tlte Rat Islands 93 Fig. 10-Typical Amchitka artifacts (U. S. h'luseum of Natural History, Sites 2 and 3). FItt~rling hnplenten ts: a-e Bone dart or harpoon heads; all these beads have rvedge-shaped tangs that were held secure in socket pieces such as i and j below. (Catalog numbers 396272, 396127, 396272,396127, and 396127.) f Bone unbarged arrow or dart point with conical tang. (Catalog No. 396142.) g-11 Bone barbed points with asymmetric tangs; possibly for use on a bird or fish leister. (Catalog numbers 396129 and 396139.) i Bone socket piece rvith peg tang for insertion into a \vooden shaft end or into another intermediate bone socket; trvo wooden inserts in the oval socket end made it possible to wedge a point tang securely hetween them. (Catalog No. 396287.) j Bone socket piece with bifurcated tang; the tang fits directly into a wooden shaft end; wooden inserts were also used in the socket end. (Catalog No. 396151.) Fishing inrple~~teats: k Hone leister side prong; for use on a bird or fish leister shaft. (Catalog No. 396092.) I Cobble sinker or weight with full groove for line or nrt attachment. (Catalog No. 396191.) Prehistoric Hu~nanOcct~pation of the Rat Isla~c~ls 95 Pig. 11-Typical Amchitka artifacts (U. S. Museum of Natural History, Sites 2 and 3). Skin-preparation implements: a Bone handle for a metal knife blade with step-scarf tang for attachment to a wooden-handle extension; a rusted iron-blade fragment is embedded in the end slot; the side-groove function is unknown. (Catalog No. 396145.) b Aivl of split proximal end of an albatross humerus. (Catalog No. 396088.) c Awl of sharpened proximal end of a bird radius. (Catalog No. 396085.) d Chipped-stone scraper blade; hafted for use in wooden handle (dark gray andesitic stone. (Catalog No. 396208.) e-g Chipped and partially ground stone ulu blades: probably hafted for use in wooden, bone, or woven-grass handles (argillitic stone). (Catalog numbers 396218, 396334, and 396218.) h Small ground stone ulu or knife blade: probably hafted in a wooden handle for use (argillitic stone). (Catalog No. 396218.) Prehistoric Human Occt~pationof the Rat Isla~tds 97 Fig. 12-Typical A~nchitkaartifacts (U. S. hIuseu~n of Natural History, Sites 2 and 3). a Bone adze head; an adze blade such as b was lashed into the basin at the upper end, and a \$rooden handle was lashed onto the longer end by means of the lashing grooves. (Catalog No. 396288.) h Chipped and partially ground stone adze blade (dark gray hard stone). (Catalog No. 396210.) c Bone wedge tuith boles resulting from using the wedge as an upper bearing or holder for a bow drill. (Catalog No. 396220.) d Small peg-like bone wedge. (Catalog No. 396219.) e Scoria abrader. (Catalog No. 396306.) f Chipped-stone scraper; the angled steep edge is unifacially flaked. The scraper is hafted to a wooden handle for use (basalt). (Catalog No. 396292.) g Chipped-stone drill bit; the bit is hafted to a drill shaft for use (dark brown hard stone). (Catalog No. 396293.) 11 Flake drill (gray andesitic stone). (Catalog No. 396340.) Prehistoric IIutna18 Occzcpatioll of tlze Rat Isla~tds 99 Fig. 13-Typical Amchitka artifacts (U. S. Museum of Natural History, Sites 2 and 3). Household implements: a-b Bifacially chipped knife blades; used hafted or unhafted (basalt and gray andesitic stone). (Catalog numbers 396213 and 396214.) c Bifacially chipped knife blade with tang notches for lashing to a wooden handle (basalt). (Catalog No. 396290.) d Bifacially chipped asymmetric knife blade with one tang notch for lashing to a wooden handle (basalt). (Catalog No. 396291.) e Bifacially chipped knife blade; probably hafted in a wooden handle before use (basalt). (Catalog No. 396290.) f Flake knife or scraper with unifacial chipping along both cutting edges (obsidian). (Catalog No. 396296.) Prehistoric IIu~nntzOccupatio~a of the Rnt Isla~~ds 101 0 5 I 1 I I crn Fig. 14-Typical Amcliitka artifacts (U. S. hluseum of Natural History, Sites 2 and 3). Ho#kseholditnplen~ents: a Carved stone lamp with charred oil residue. (Catalog No. 396186.) b Combination hammerstone and grinder; both ends have been used for pounding and two side facets have been worked flat by grinding. (Catalog No. 369307.) c Combination hammerstone and grinder; the edges and particularly the ends have served as pounding surfaces: a flat grinding facet is worn on one face. (Catalog No. 369197.) Prel~istoricHz~~na~t Occtrpatioi~ of the Rat Islatlds 103 ~ 0 5 IIIIII crn Fig. 15-Typical Amcbitka artifacts (U. S. i\Iuseum of Natural History, Sites 2 and 3). Collectif~gimplements: a Ivory digger with carved handle and incised line decoration. (Catalog No. 396242.) b Bone probe or digger with decorative end caving. (Catalog No. 396163.) Articles of adornment/miscella~~eous: c lvory/tooth labret. (Catalog No. 369100.) d Ivory/tooth labret. (Catalog No. 396097.) e Bone labret. (Catalog No. 396107.) f Bone bead. (Catalog No. 396115.) g Bone "gauge"; use unknown; surface polish is often found on these pieces at the ends below the pips. (Catalog No. 396112.) Prelristoric I Contextual category Artifact class Hunting implements Bone harpoon heads (detachable)* Bone spear/lance beads (nondctacI~able)* Bone dart heads* Bone arrowheads* Bone bird blunts* Bone loose shafts* \\'ooden covers fbr stone end blades \irooden shafts for harpoonlspearslda~ts* Bone bladder nozzles Bone and wooden clubs* Bone bird gorges Bola stones (?)* Fishine-. im~lements Bone fishhooks* Bone fish leister pointslside prongs* Stone weiahtslsinkers* Collecting implements Bone diggerslpmbes* Bone mattock and pick heads* Net and basket containers \Vood\vorking and Stone maul heads* bonervorking Hand hammer stones* implements Bone wedges* Bone adze heads* Stone adze blades* Bone drill bearings* Stone drill bits* Stone scraDer blades* Stone flake scrapers* Bone knife handles* Stone knife blades* Iron knife blades* Scorialpumice abraders* Stone\r~orkingimplements \\'hetstones* Bone flaking tools Skin preparation Bone ulu handles implements Stone ulu blades* Stone scraper blades* Stone flake scrapers* Iron knife blades* Bone scrapers Bone cutting/trimming tablets* Bone arvls/ponches* Bone needles* Bone needle cases Household implements Bone ulu handles Stone itlu blades* Stone knife blades* Bone fire dl-ill bearings* \\'ooden fire drill base Sulfur (for fire starting) Hand hammer stones* Ocher grinding slabslstones* Stone lamps* Stone griddles Stone bowls Bone bowls/containers Bone container sides* \\'ooden bowlsltrays Bone and \\rooden s~)oons/scoops/shovels Articles of adornn~ent Bonc/tooth labrets* and decoration Boneltooth spools* Bone/tooth pins Bone/tooth pendants Preliistoric Hzrma~~Occupatio~i of tlie Knt Isln~ids 107 Table 9-(Continued) Contextual category Artifact class Fieure Pendants made of drilled or grooved mammal teeth* Bone beads* 15(f) Shell headslpendants* Bone chains Boneltooth animal and l~umanfigurines* Bone combs Hone and wooden masks *Known from Rat Island colfections. As with tlie intragroup comparisons above, evidence of contact and cultural flow. Our mediuni artifact or cultural similarity between island groups is artifacts and other physical manifestations of directly expresses contact between Aleuts of tliese cultural patterning. The physical a~lthropologist groups. Altl~ougli not all contact need imply would test such contact by studying gene flow cultural transmission, stylistic sinlilarity may be expressed in liuma~i skeletal series. A linguist explai~iedrealistically by sucli transmission. \vould try to establish coliti~luityand discontinuity on the basis of present Aleutspeakers and historic hltergoup Comparisons. I suggested several language documents. Distributional artifact studies, years ago (IvIcCartney, 1971) that distinctive char- in aid of themsel\~es,may seem a bit liumrlrum; acteristics of Near Island material culture ulti- yet results of tliese comparisons directly indicate mately derived from the relative isolatio~iof tliat adaptations, populatio~imovement, in situ develop- group from the rest of the archipelago. After I ment, trade, or other i~iteractio~land adjustment. completed that article, the Desautels et al. (1971) The critical questions revolve around why these monograph appeared, and I commented that tliis distributio~lsare what they are. \\'hat cultural atid report suggested that Amcliitka was Inore closely social regularities, in other words, arc being demon- related to central atid eastern assemblages tlian to strated by particular artifact clusterings in space? those of the far xvesteni isla~lds(klcCartney, 1971, Although tlie explanatory phase will benefit p. 106, postscript). Since I have studied additional greatly from accurate distributional descriptions Rat and Andreanof artifacts, a general updating of based on large-scale comparative studies, an at- the 1971 co~iclusionsis in order. But tlie compari- tempt is made in a later section to find etlnlo- sons are now in txvo directions rather tlian one, as historic reasons for stylistic spread. \\,as tlie case with the Near Isla~idsat the chain's Long artifact attribute and type lists will not end. be i~lcludedin this chapter. kly previous listing of \Vlrat can we hope to learn from artifact shared pan-Aleutian traits (McCartney, 1971, distribution studies? In regard to the ecologic, pp. 138-139, Table IV) sta~~dsas presented. These systemic study of marine subsistence emphasized 32 iteills are indicative of artifacts fou~ldthrougli- above, we should ascertain if, in fact, all islanders out the cliain, and more will certainly be added in adapted similarly. Even casual study shows us that the future. The listing of cc~itral and eastern indeed all these various island collections display Aleutian artifacts (A,lcCartney, 1971, p. 141, the same artifact core as is listed, for instance, in Table V) also remains generally accurate, with Table 9. At the stylistic level, however, we are lnariy Rat Isla~ldexamples :low added that demoti- primarily concerned with ascertai~li~lgcultural strate the western distribution of most of these 25 spread over time and space. As mentioned previ- itenis to that isla~idgroup. Cotiversely, Near Island oilsly, most areas of North America have a cul- characteristics (A,IcCartney, 1971, I). 136, tural-historical frame~vork established through Table 111) have been significantly altered. At least past decades of archaeological study. The Aleutia~ls nine of tlie original 29 items arc now known from lack any such framework. Therefore, seeking evi- tlie Rat Islands as well, but these may be restricted dence of coltural flow-population contact along to these two groups only. We could add several tliis atte~iuatedisland chain requires distributio~ial artifacts whose distributio~iis currently limited to illput about the spread of cultural characteristics. the Rat group. These ~\rouldi~icli~de iron-bladed \\re have postulated that dista~icebet\seen islands k~iives (Desautels et al., 1971, p. 246, Fig. 124) and island groups is inversely proportio~lal to ant1 conical-tanged arrow points (I>esautels et al., cultural contact and spread. To establish such a 1971, p. 197, Fig. 92 a and 1); Cook, Dixon, and relatio~lship requires tliat we demolistrate some Nol~~ies,1972, plate 16). As a result of further research since 1971, 1 rials to choose from may be. Exa~nplesof mate- no\\, offer a subjecti\~ebut considered opinion that rial-tool associations are as follo\vs: slate or there is far more continuity throughout the chain argillite (soft, fine grain), ulu blades; basalt (hard, than ciiscontinuity and that the Ncar Island group brittle), knife and scraper blades and projectile is probably not as different or distinctive from the points; ~veldedtuffs (soft, \resicular), carved bowls remaining groups as previously suggested. As more and lamps; and scoria and pumice (soft to hard, is learned about artifact variation and spread from vesicular), abraders. Only rarely is one material newly excavated as \\,ell as existing museum collec- substituted for another, lvhich causes a given tions, it appears that every major island group will material's propelties to be disregarded. i\luch less have distinctive implement styles not found else- frequently, form rather than material dictates \\,here and at the same time will share artifacts \\'it11 persistent use. Drilled sea-mammal teeth with no one or more other adjacent groups. L\'ithout a other alteration are commonly used for pendants complete analysis of all available artifacts, one may or beads, or bird-wing bones are usually sharpened intentionally or unintentionally stress continuity at one end to make awls and punches. by listing shared traits or discontinuity by listing 6. Artifactual distribution carries no assumed distinctive traits. A further complicatio~~is lack of implications about spread of physical type (Paleo- time control to judge contetnporaneity; the large Aleuts or Neo-Aleuts) or language dialects (eastern, majority of extant Aleutian artifacts are not central, or western). The entire chain, plus the accurately dated. adjoining Alaska Peninsula, is a geographic contin- The follo\ving nine hypotheses su~nmarizeour uum along which race, language, and culture current interpretation of artifact distribution; all assorted tllenlselves independently since their demand rigorous examination in the future: modes of transmission are distinct (klccartney, 1. All island groups have been in more or less 1974). conti~luouscontact for at least the past 2500 to 7. On the basis of proximity, the Rat Islanders 4000+ years; no group has remained isolated for a had more contact with adjacent Near and Andrea- period sufficient to produce high-level artifact nof (including Delarof) people than with any other dissimilarity. group. Because the Delarof Islands and Buldir 2. Standardizatio~lof artifact form and style is lsland are intermediate between the Andreanof and due in part to common environmental adaptation Near groups, respectively, and because these islands throughout the chain. Continuity over time and probably either supported no resident villages space is due to a relatively stable environment for (Buldir) or only a few (Delarofs), these islands approximately 4000+ years; readaptations to sub- would likely be peripheral hunting, fishing, and sistence alternatives were not required whic11 collectillg areas for the remaining Rat Islanders. would be reflected in subsistence technology and Recent Delarof estimates show a combined colony artifact inventory. bird total of 250,000 and a sea lion total of 6,600 3. Cultural interactio~land population contact (Sekora, 1973); these compare favorably with was Iteightened by geographic restrictions on abundance figures for Buldir and the remaining Rat travel; homogeneity resulted from east-west travel Islands. only along the chain with no interior to occupy via 8. Because the Aleutian chain extends from the river systems as in mainland Alaska. mainland far into the North Pacific, the western 4. Artifacts appear to have been spread inde- islands, including the Rat group, mill tend to show pendently of one another; no large clustering less cultural interaction with mai~llandAlaska than appears to occupy the same space simultaneously. the closcr eastern islands. A firm denlonstration of 'This conclusion may, ho\vever, be based on in- this attraits additional study of central ancl western co~npleteknowledge of actual cluster spread. A Aleutian artifact collections. small complex of late prehistoric artifacts previ- 9. Finally, trade or exchange was not con- ously described (McCartney, 1971, p. 96) may ducted at the pan-Aleutian level. Rather, \\,hat exemplify implement styles spreading to varying little evidence we have suggests trade predomi- degrees through the chain. !\'hether artifacts spread nantly \\'itllin island groups (compare Lantis, 1970, singly or in clusters, the critical questions are \\.hat p. 274, for supporting ethnographic data). Part of functional, adaptive, or other cultural processes the problem lies in establishing sources of land- were responsible. bo~ind~naterials that, because of their scarcity, 5. hflaterial specificity is an important factor in \vould lend themselves to trade. Obsidian is an artifact commonality. For technologic reasons, one cxcclletlt choice. Because it is harcl, brittle, and has material may be better suited for a particular conclloidal fracture, it is an excellent material for artifact than some other. This is not unique to the projectile points or knife and scraper blades, but Aleutians, but the particular combination of mate- obsidian was not xvidely tfaded throughout the cliain. Of all the Near Island specimens examined, early iron and copper suggest fairly \vide distribu- none are of obsidian (ArIcCartney, 1971, p. 136, tion at tlie early contact period. Steller reported Table 111). The Rat group collections (AEC and iron knives in the Shumagins in 1741, sl~ip\vrecks Smithsonian) reveal only two pieces, one of which lvere noted on Itodiak, Atka, and in the western is illustrated in Fig. 13f of this chapter (Turner, islands, and iron exchange betwecn Aleuts and 1970, p. 121; Desautels et al., 1971, pp. 78 and Asiatics or Europeans is referred to on Avatatiak 109; Cook, Dixon, and Holmes, 1972, p. 23). I (east of Unalaska) and Attu (Coxe, 1787, p. 50; have located only one obsidian artifact in the Jochelson, 1933, pp. 21-23; HrdliEka, 1945, Smithsonia~lAndreanof collections. In tlie eastern pp. 100-101; deLaguna, 1956, pp. 62-63). Iron Fox Islands, however, where at least three or four from Icodiak or the Alaska Pellinsula shores was obsidiati sources are reported (Umnak, Unimak, traded to the eastern Fox Islands (HrdliEka, 1945, Unalaska, and Akutan; Elliott, 1886, p. 148; p. 142). HrdliEka, 1945, pp. 98-99), obsidian is not at all Arcl~aeologically we do not find prehistoric rare. It makes up, for instance, about 15% of the evidence of wide metal use in the chain. Of all the 8000-yea~oldAnangula core and blade complex prehistoric collections I have studied or have (h,lcCartney and Laughlin, 1964), 5 to 48% of kno\vledge of, I have never found metal fragments Clialuka (Umnak) artifacts (Denniston, 1966, or bone knife handles with iron or copper oxida- p. 91, Fig. 3), 4.5% of Akun waste flakes (Turner tion stain except in the Rat group. The Smith- and Turner, 1974, p. 47, Table 3), and 5.4% of sonian has one bone knife handle from Amaknak Izembek flakes (tip of tlie Alaska Peninsula; (Fox Islands) that is similar to, but not exactly A,lcCartney, 1974, p. 72). Tlie highest use of like, those found in the Rat Islands, but the piece obsidian reported is at the Ashishik Point site has no metal stain in the blade slot. Presiunably (northeastern Umnak), ~crhichis very close to the relatively rare shipwrecks during the late pre- obsidian source on that same island end (Dennis- historic era supplied the iron for the Rat Islanders ton, 1972, pp. 94-95 and 296, Fig. 17). Obsidian since no terrestrial or meteoric sources are known here ranged bet\veen 40 and 90% in different in the Aleutian chain. Shipwrecks would be ex- levels. These indications of distribution suggest pected to occur wit11 greater frequency the more that obsidian was quarried only in the Fox Islands recent the period. arid that there was hardly large-scale trafficking of Obsidian and iron, then, are examples of two this material throughout the chain. Tlie two Rat scarce but useful commodities that might have pieces must have passed from east to \vest by been exchanged widely. Of course, iron may have undirected exchange. been even more limited atid insufficiently abun- The Rat Islanders exploited one regional re- dant for trade. The fact tliat long-distance trade source apparently found in few other localities. net\vorks did not develop may have been due to This is pre-eighteenth-century foreign iron, and it, the fact that material such as obsidian and iron was too, failed to be traded ~videly to other areas. not required for successful toolmaking and sur- Desautels et al. (1971, pp. 243-248) found 12 vival. They were luxuries to be used if available. iron-bladed knives at Sites 10, 14, 35, and 36, and They seem to substantiate the hypothesis that Cook, Dixon, and Holrnes (1972, p. 21, plate 24) ideas of style spread widely but materials spread found two pieces at Site 32. HrdliEka's Smith- only regionally. sonian collections include seven iron-bladed knives Other precious materials traded include denta- or iron-stained bone handles (Amchitka, four, lium, amber, mineral pigments, and porphyry including tlie one sho~v~iin Fig. I la; Little Kiska, (Veniaminov, 1840, Vol. 11, pp. 94-95; Lantis, t~so;and Itiska, one). Because none of these were 1970, p. 275). associated with Russian- or American-period trade goods and clearly are uot just surface finds, Explanation of Artifact Similarity. It is ap- may safely take them to be precontact in age. The propriate to suggest possible mechanisms respon- fact that all the blades are hafted in typical Aleut sible for the spread of artifact styles. Althougli we style bone handles rather than having originally have stressed Aleut seasonal mo\rement as an manufactured handles suggests that the iron alone adaptive mechanism, little has been said thus far was acquired from an Asiatic source and made into about social interaction and movement. Given that knives once in the hallds of Aleuts. The ultimate interaction is expressed archaeologically in shared source of iron prehistorically \\.as China or Japan; artifact styles, then how did short-term social it Isas obtained by the Aleuts from boats wrecked interaction facilitate the sharing of normative on Aleutian shorcs. There is little support for patterns? Tx\ro major ways of interacting are Jochclson's (1933, p. 22) hypothesis of Chukclli- imliiediately apparent. One is tliat resulting in Eskimo-Aleut distribution. Several references to finished artifacts being exchat~ged between dif- ferent villages, islands, or island groups. Second, arcliaeological collections primarily express spread inovement of people rather than items \\rill spread of male cultural behavior. styles even more effectively than exchange. Trade and festivals are two previously men- tioned modes of social interaction between Aleuts CONCLUSIONS of different villages. In fact, festivals were likely the most important occasions for social as well as This chapter relates to Aleutian archaeology of material exchange. Gifts, or more realistically tlie past, present, and future. Archaeology over tlie investments of presents demaiiding reciprocity, past century has largely been eastern-oriented, were given during the winter festivals or feasts artifact-centered, nonclassificatory, and lion- (Veniaminof, 1840, Vol. 11, p. 91). Such artifact comparative. It quite logically has treated origins scattering contributed to\vard homogenization of and development as central issues. Integration of styles within an island group at least. Short-range ethnohistoric and archaeological data has profit- trade of weapons, masks, clothing, skins, etc., also ably stressed temporal continuity out of the occurred between villages in other trading situa- distant past. But, as one part of Alaska most tions (Coxe, 1787, p. 197; Veniaminof, 1840, intensively occupied and dominated by Russian Vol. 11, pp. 110-111). As pointed out above, long- and later American trading companies, tlie range trade throughout the chain is not borne out Aleutians have never been studied commensurate by archaeological evidence or by ethnoliistoric with their historic importance. accounts. This chapter centers on the Near and Rat Perhaps more important than the movement of Islands, especially on Amchitka. Archaeological items \\,as tlie movement of people, either tempo- data from that island provide a much needed iarily or permanently, from one residential alliance western counterpoint to eastern-island information to another. There is abiundant ctl~nol~istoricin- to reveal what is culturally representative of the formation about both xiiales and females changing entire chain. Artifacts have been given only second- their residence willingly or uiir\~illingl)~.For in- ary coilsideration here because most of them have stance, men moved for a year or two to their wife's been adequately treated in previous publications. village at marriage to perform bride service. Matri- Another reason is because archaeologists today are lineal clans appear to have established kin, and clan looking beyond artifacts at less tangible, but no exogamy would likely have a man marrying his less real, cultural phenomena. By stressing the mother's brother's daughter, a member of another spatial dimension, one immediately becomes en- iiiatrician. Or, if he did not marry his cousin, a man grossed with such interrelated cultnral aspects as ~souldlikely marry another nonclan woman from settlenieiit pattenis, subsistence, demography, another village to expand political bonds to that social organization, trade, atid religion as xvell as village. Besides, in small \villages of related families, technology. \ire are pressing for cnltural dynamics it might be difficult to find the proper spouse that operated within tlie Rat Islands to understand witliiil the village at marrying time. Because resi- the relative importance of determinants as de- dence after marriage was at tlie original village of scribed above. tlie husband (patrilocality), the wife \\rould be Future Aleutian archaeology will progress the moving into a new conimunity. Boys were prefera- furthest by addressing hypotheses such as those set bly instructed during adolescence by their maternal forth in this chapter. Amassing more artifacts \\rill uncles (avunculate), and this would i~suallyoccur not, in and of itself, answer the sorts of human at the mother's original village prior to marriage. behavior questions raised here. Salvaging prc- [\'hen we add \varfare captives or slaves to tlie historic remains from endangerment by man or above, \vc obviously find significant social mo\fe- from natural erosion is important, but it will be ment (for details of kinship, marriage, education, more fruitful to an understanding of Aleutian aid warfare, see Veniaminof, 1840, Vol. 11, prehistory if our excavations also aim at higher pp. 73-108; Lantis, 1970, pp. 205-271). level questions, as, for instance, those dealing with Because people xvill express learned styles when adaptation, interaction, and development. nianufacturing in~plcincnts and other lnaterial Future arcliaeological research ~vill coiltiilue goods, reshuffling Aleuts from village to village and the i~iterdisciplinaryapproach exemplified by the even from island group to island group resulted in AEC ecologic studies on Atnchitka. The kinds of spreading stylistic behavior. Ho~vcver,since Aleut questions now raised cannot be adequately an- men are believed to be the makers of more durable swered without input from other scientists of the artifacts (bone, tooth, stone, and wood) than those physical and biological environment and integra- made by woinen (grass, skin, gut, hair, etc.), our tion of their information. .ehistoric Hz~?nanOcc~cpntio?~ of the Rnt Islnt~ds 111 U. S. Geographical and Geological Survey of the Rocky hlountain Region, Washington. de Laguna, F., 1956, Chtrgnch Prehistory: Tlfe/irchaeology Several persons contributed to this Rat Island of Prince IVillianl Sound, Alosko, University of \\'ashing- summary. John Cook initially suggested to the ton Press, Seattle. editors that I write this chapter, \Villiarn Fitzhugh Denniston, G., 1972, Ashishik Point: An Economic Analy- assisted my i~lspectionof Rat Island collections in sis of a Prehistoric Aleutian Community, unpublished Ph. D. dissertation, Department of Anthropology, Uni- the Smithsonian Institution, Janles Nybakken pro- versity of \\'isconsin. vided contemporary nomenclature for some marine - 1966, Cultural Change at Chaluka, Umnak Island: invertebrates, and Michael Schiffer contributed Stone Artifacts and Featores, Arctic Antlfropol., 3: valuable archaeological comments. Schiffer, Sandra 84-1 24. Scholtz, and Paul Long assisted me in developing Desautels, R. J., A. J. I\lcCurdy, J. D. Flynn, and R. R. the irregular island measure described herein. R3. L. Ellis, 1971, Archaeological Report, Amchitka Island, Alaska, 1969-1970, USAEC Report TID-25481. Merritt provided numerous data about AEC Am- Elliott, H. \V., 1886, Our Arctic Prouince: Alaska and the chitka research since its inccption. The editors and SenlIslnnds, Charles Scribner's Sons, New York. Leonard Gard, Christy Turner, and Nancy R,lcCart- Estes, J. A,, and N. S. Smith, 1973, Ainchitka Bioenviron- ney offered divers but pertinent criticisms that mental Program, Research on the Sea Otter, t\mchitka have benefited this chapter. I wish to thank all Island, Alaska, Final Report, USAEC Report NVO- 520.1, Arizona Cooperative \\'ildlife Research Unit, these persons for their various considerations; none University of Arizona, Tucson. should be held responsible for the ideas or data Fedorova, S. G., 1973, Tl~eRussian Populntiotz in Alnskn expressed here. and Californin, Late 18th Century-1867, The Lime- Thc Department of A~~tl~ropolog)~,Smith- stone Press, Kingston, Ontario. sonian Institution, kindly loaned me the Rat Island Flannery, K. V., 1968, Archaeological Systems Theory and Early Mesoamerica, in A~rthropologicalArchneology in artifacts illustrated in Figs. 9 to 15. The illustra- tlfe Anfericns, pp. 67-87, The Anthropological Society tions \sere prepared by Louis Gregoire. of \\'ashington. -, 1967, Culture History vs. Cultural Process: A Debate in American Archaeology (Review), Sci. Arner., 217(2): REFERENCES 119-122. Fosberg, F. R., 1963, Thc Island Ecosystem, in illan's Place in the Islmfd Ecosystem, F. R. Fosberg (Ed.), pp. 1-6, Aigner, J., 1970, The Unifacial Core and Blade Site on Bishop hiuscum Press, Honolulu. Anangula Island, r\leutians, Arctic A~ttl~ropol.,7: Friedman, H., 1937, Bird Bones from Arcl~aeotogicatSites 59-88. in Alaska, J. Il'nsh. Acad. Sci., 27: 431-438. Bank, T. P., 11, 1953, Ecology of Prehistoric Aleutian Guggenheim, P., 1945, An Anthropological Campaign on Village Sites, Ecology, 34: 246-264. Amchitka, Sci. dfonthly, GI: 21-32. -, n.d., Loose Map Sheets \\'ith h,Iarked Sites: Uliaga, Heizer, R. F., 1960, The Aleut Sea Otter Hunt in the Late Kagamil, Near Islands, Kiska, Rat, Delarof Islands, and Nineteenth Century, Anthropol. Papers Uniu. Alaska, 8: Atka. 131-135. -, 1943, A Pacific Eskimo Invention in \\'hale 1.Iunting in Clark, A. H., 1945, Animal Life of the Aleutian Islands, in Historic Times,Amer. Anthropol., 45: 120-122. The Aleutian Islands: Their People and Natfcml History, Hett, J. M., and R. V. O'Neill, 1974, Systetns Analysis of No. 21, Smithsonian Institution \\'a Background the r\leut Ecosystem, Arctic Anthropol., 11: 31-40. Studies, pp. 31-62. Hrdlizka, I\., 1945, The lileutinn and Com~nanderIslands Clarke, D. L., 1968, A~mlyticalArclraeology, hlethuen & n~:d Their Inhabitntfts, \\'istar Institute of i\natomy and Co., Ltd., London. Biology, Philadelphia. Coats, R. R., 1950, I'olcanic Activity in the Aleutian Arc, Hurt, I\'. R., 1950, Artifacts from Shemya, Aleutian U. S. Geological Survey, Bulletin No. 974-B. Islands, Anfer. Antiquity, 16: 68-69. Cook, J., C. Clerke, and J. Gore, 1818, A Voyage lo the Jochelson, \\'., 1933, IIistory, Ethtfology and A~ftl~ropology Pacific Ocean in the Years 1776-1780, Robert De- of the Alefrt, Publication No. 432, Carnegie llistitution silver, Philadelphia. of \I1ashington. Cook, J. P., E. J. Dixon, and C. E. IIolmes, 1972, -, 1925, Archaeologicnl Investigations in the Aleutian i\lchaeologicat Report, Site 49 Rat 32, Amchitka Islands, Publication No. 367, Carnegie Institution of Island, Alaska, USAEC Report 'IN-20-1045, Holmes \\'ashington. and Narver, Inc. Jones, R. D., Jr., 1963, BuMir Island, Site of n Remnant Coxe, \ 1787, Account of the Russian Discoveries Breeding Popfclatio~f of Aleutian Canada Geese, 14th Between Asia and America, Augustus BI. Kelley, Pubs., Annual Report, pp. 80-84, The \\'ildfowl Trliost. Clifton, N. J. Kemp, \\I. B., 1971, The Flow of Energy in a IIunting Dall, \Y. H., 1878, On the Remains of Later Pre-Historic Society, Sci. Amer., 225: 104-115. illon Obtained from Caues in tlfe Cntlrerinn Archipelago, Kenyon, K. \\I., 1969, The Seo Otter ir~the Enstern Pncific Alaska Territory, No. 318, p. 8, Smithsonian Contribu- Ocean, U. S. Department of Commerce, Bureau of tions to Knowledge. Sport Fisheries and \\?Idlife, North American Fauna, - 1877, On Succession in the Shell-Heaps of the No. 68. Aleutian Islands, in Tribes of the Extreme Northwest, -, and J. G. King, 1965, Aerial Strrvey of Sea Otters, I\'. H. Dall, J. Furuhelm, aud George Gihbs, pp. 41-106, Other AInrine ilmnmals nnd Birds, Alnskn Peninsuln nttd Aletttia,t Islnnds, Report to the U. S. Bureau of Sport Rappaport, R. A,, 1971, The Flow of Energy in an Fisheries and I\'ildlife. Agricultural Society, Sci. Amer., 225: 116-132. -, and D. I\'. Rice, 1961, Abundance and Distribution of -, 1963, Aspects of Man's Influence Upon Island the Steller Sea Lion, J. dla~nmology,42: 223-234. Ecosystems: Alteration and Control, in dlatt's Place in Langsdorff, G. H., 1817, Voyages and Trnuels in Vnrious the Island Ecosystem, F. R. Fosberg (Ed.), pp. 155-174, Parts of the il'orld During the Years 1803-1807, Bishop Museum Press, Honolulu. 11. Colburn, London. Redman, C. I,. (Ed.), 1973, Research and Tlteory in Lantis, hl., 1970, The Aleut Social System, 1750 to 1810, Ctrrrettt Archaeology, John \\'iley 8r Sons, Inc., New from Early llistorical Sources, in Ethnohistory in York. Southwestern Alaska nnd the Southern Yukon, Sahlins, hl., 1972, Stone Age Economics, Aldine Publishing &I. Lantis (Ed.), The University of Kentucky Press, Co., Chicago. Lexington. Leone, hl. P. (Ed.), 1972, Contemporary Archneology, Sarichev, G., 1807, Account of a I'oyage of Discovery to Southern Illinois University Press, Carbondale. the North-East of Siberia, the Frozen Ocean and the Malde, H. E., 1964, Ecologic Significance of Some Un- North-East Sea, Richard Phillips, London. familiar Geologic Processes, in The Reconstructio~~of Sekora, P., 1973, Aleutian Islands National IYildlife Refuge Past Enviromnents, J. J. Hester and J. Schoen~vetter \\'ilderness Study Report, preliminary draft, U. S. (Eds.), pp. 7-15, Fort Burgwin Research Center, New Bureau of Sport Fisheries and Wildlife. hlexico. Sense, R., and C. G. Turner 11, 1970, Catalog of Archaeo- McCartney, A. P., 1975, Maritime Adaptations in Cold logical Site Survey Records, Amchitka Island, Alaska, Archipelagoes: An Analysis of Environment and Cul- Holmes and Narver, Inc., internal report. ture in the Aleutian and Other Island Chains, in Spaulding, A. C., 1962, Archaeological Investigations on Prehistoric ~VlaritimeAdaptations of the Circumpolar Agattu, Aleutian Islands, Anthropology Papers No. 18, Zone, \\I. Fitzhugh (Ed.), pp. 281-358, Aldine Publish- University of h,lichigan hlusenm of Anthropology. ing Co., Chicago. Spencer, D. L., 19G9, Amchitka Bioenvironmental Progcam. -, 1974, Prehistoric Cultural Integration 'Along the Interim Report, Sea Otter Surveys, Amchitka Island, Alaska Peninsula, Atttltropol. Papers Uniu. Alaska, 16: Alaska, August 27-October 6, 1968, USAEC Report 59-84. -, 1972, An Archaeological Site Stcrvey and I~~ventory BhlI-171-120, Battelle blemorial Institute. for the Aleutiatt Islands National Ilrildl$e Refuge, Spencer, R. F., 1959, The North Alaskaf~Eskbno: A Study Alaska, 1972, U. S. Bureau of Sport Fisheries and in Ecology aad Society, Bulletin 171, Bureau of l!'ildlife. American Ethnology, Smithsonian Institution. -, 1971, h Proposed Western Aleutian Phase in the Near Staehlin, V. S., 1774, An Account of the New Northern Islands, Alaska, Arctic Anthropol., 8: 92-142. Archipelago, C. IIeydinger, London. -, 1967, An Analysis of the Bone Industry from Stejneger, L., 1936, Georg Il'illteln~ Steller: The Pioneer of Amaknak Island, Alaska, unpublished h1.A. Thesis, Alnskan Natural History, Harvard University Press. Department of Anthropology, University of Wisconsin. -, and I\'. S. Laughlin, 1964, Earliest Aleutian Pre- Turner, C. G., 11, 1973, Report oft 1973 Antlrropological history: The Anangula Island Core and Blade Site, Fieldwork on Aktrn Island, Alaska, Department of unpublished paper read at the 29th Annual Meeting of Anthropology, Arizona State University, Tempe. the Society for American Archaeology, Chapel Hill. -, 1970, Arcllaeological Reconnaissance of Amchitka -, and C. G. Turner, 1966, Stratigraphy of the Anangula Island, Alaska, Arctic Anthropol., 7: 118-128. Unifacial Core and Blade Site, Arctic Attthropol., 3: -,and J. A. Turner, 1974, Akun, Anthropol. Papers 28-40. Univ. Alnska, 16: 27-57. Merritt, M. L., 1972, Cannikin: Effects on Archaeological Turner, L. M., 1886, Contributions to the Natural History Sites, USAEC Report SC-RR-720359, Sandia Lahora- of Alaska, No. 155, Senate Miscellaneotrs Documents, tories, Albuquerque. 1st Session, 49th Congress, Vol. 8, pp. 3-226. Murie, 0. J., 1959, Fauna of the Aleutian Isla~~dsand U. S. Coast Pilot 9, 1955, Alaska, Cape Spencer to Arctic Alaska Peninsula, U. S. Department of the Interior, Bureau of S~ortFisheries and \\'ildlife. North American Ocean, U. S. Coast and Geodetic Survey. Fauna, No. 61. Vanstone, J. IV., 1960, An Early Nineteenth-Century Artist Nelson, E. U'., 1899, The Eskimo Aborct Beringstrait, 18th in Alaska,Pac. Northwest Quart., 51: 145-158. Annual Report, Bureau of American Ethnology. Veniaminof, l., 1840, Notes on the Unalaska Islnnds Neuhold, J. M., W. T. Helm, and R. A. Valdez, 1971, District, Russian American Company, St. Petershurg Freshwater Vertebrate and Invertebrate Ecology of (Human Relations Area File translation of Vol. I1 and Amchitka Island, Annual Progress Report, Joky 1, part of Vol. Ill). 1970-June 30, 1971, USAEC Report BbII-171-142, IYeyer, E. M., Jr., 1929, An Aleutinn Burial, Anthropologi- Battelle RIemorial Institute. cal Papers, Vol. 31, No. 3, American bluseum of Natu- Orth, D. J., 1967, Dictio~tary of Alaska Place Names, ral 'Xistory. U. S. Geological Survey, Professional Paper No. 567. Wilcox, R. E., 1959, Some Effects of Recent I'olcanic Ash Petroff, I., 1884, Report on the Population, Industries mtd Falls with Especial Reference to Alaska, U. S. Geologi- Resources of Alaska, Tenth Census (1880), U. S. cal Survey, Bulletin No. 1028-N. Department of the Interior. Piddocke, S., 1965, The Potlatch System of the Southern I\'illey, G. R., and P. Phillips, 1958, A'fethod and Theory in Kwakiutl: A New Perspective, Southzvest. J. An- American Archneology, University of Chicago Press, thropol., 21: 244-264. Chicago. Ransom, J. E., 1946, Aleut Natural Food Economy,Amer. N'illiamson, F. S. L., and W. B. Emison, 1969, Studies of Anthropol., 48: 607-623. the Avifauna on Amchitka Island, Alaska, Annual Prel~istoricHuma~l Occt~pation of the Rat Isla~lds 113 Progress Repott, June 1968-July 1969, USAEC Report Report, JuIy I, 1969-June 30, 1970, USAEC Report BhII-171.125, Battelle hlemorial Institute. Bhll-171-131, Battelle Memorial institute. -,IV. B. Etnison, and C. hl. White, 1971, Studies of the Zimmerman, E. C., 1963, Nature of the Land Biota, in Avifauna on Amchitka Island, Alaska, Annual Progress dlo,t'r Place it, the Islntzd Ecosystem, P. R. Fosberg (Ed.), pp. 57-64, Bishop hfuseum Press, Honolulu. This page intentionally left blank History, M. L. Merritt 1741-1967 Sandia Laboratories, Albuquerque, New Mexico Amchithn hns long been subject to man's inflttetrce. II%en Cn~tadaGoose wns extirpated from the islnnd. The foxes discovered by the R~rssians in 1741, the Aleutians had tltemselues werefinally eradicated from the islond by 1960 10,000 to 20,000 inhabitants and Antchitkn itself perltaps by an extensive poisoningprogra~it. 100. IIord treotmetlt cnused their nunrber to fall dmsticnlly, During IVorld li'ar II Atnchitka becnnte the forward and by 1849 Antclzitka was abandoned. No one has lived fighter-bomber base in the cnmpnign ngainst Japnnese-held there otr a permnnettt basis since, bat old livitrg sites ore still Attrr and Kiska. About 5000 troops were present oft the visible around the edges of the islatrd. islnnd then, and facilities were later added to house 011 Pressure by Rrcssintt and American hunters led to the authorized troop stre~zgtlr of about 16,000, Matry struc- near estermiltation of sea otters, collseruation nteasures not tures, roads, three airstrips, and much miscellmteous debrir hauitzg been kept zip after the 1867purchnse; the 1911 Fur remain front the war yenrs. Seal Treaty, howeuer, served also to protect sea otters, and In 1965, 1969, and 1971, nuclear shots were fired they hnve recovered well. 11: 1913 the Aleutians were set underground at Amchitka by the U. S. Department of aside as what is ~tozucalled the Aleutian Islands National Defense and the U. S. Atomic Energy Commission. During Wildlife Refuge. the first of these, there was a stnall bioenuironntental In 1921 blue foses were introduced onto Amchitka for program, and, during the last two, a large one, the resrclts of st~pport of ~leukfrotn Atkn. As a result the Aleutkn which are reported in other chapters of this book. No one ktio\vs xvllen man first set foot on 1741 (Old Style*), and named by them St. Amchitka. It is believed that Paleo-Aleuts came to Markiana. Golder disagrced; he believed it was Amchitka from the east at least 2500 years ago as Kiska that was sighted that day. But it does not part of their slo\cr clockwise advance around the matter; if Amchitka \\.as not seen in 1741, it was shores of the Bering Sea. Thereafter for centuries within the next two decades. these people lived on these islands and called them Bering's ship came to an end on Bering Island, home, drawing their living from the rich seas of the I Year* Total Russiansl Creoles6 1) Year* Total Aleuts? Russians$ Creoles6 Islands Cellsus IVhite I I Year Total District Aleuts men Creoles I I *Sources: 1741-1880: Fedorova (1973, pp. 161, 203-204, 275-279). Obvious errors in arithmetic have been corrected. Others appear to exist and are marked 7. Totals sometimes disagree with Tikhmenev (1863), as quoted by Petroff (1884); these are marked -*. 1880: Petroff (1884, p. 33). 1880-1970: Decennial reports of the U. S. Bureau of the Census. ?Russian counts of Aleuts included Koniags as well. U. S. coilnts of Aleuts were not always made, and, when they were made, varying criteria of who rvas an Aleut were used. $The column labeled Russians included others before 1870 and includes all whites after 1870. §In Russian usage a Creole was a person of mixed Russian and native blood. 7At these points Fedorova or her source made an arithmetic error that cannot be corrected with the information available to me. **At these points Fedorova's totals disagree with Tikhmenev's totals. History, 1741-2967 119 Table 2-Population of Native t\vo years and prohibition of firearms during the Villaees in the Western Aleutians protected season (Fedorova, 1973, p. 189). In Year* Atta Atka 1840 Governor Etholen extended the fallow peri- ods to ten years (Bancroft, 1886, p. 582). By the end of the Russian period, the Atka Aleuts were allox\~ed to take only 300 sea otters per year (Bergsland, 1959, p. 75). The proble~nsof the Russian government in holding a distant profitless territory in the face of vigorous expansion across the continent by the British and Americans and its general international posture, including its need to consolidate its inflite~lcein Asia along the Chinese bordcr, led to the decision to give up Russian America. These problems are illustrated by the facts that the Russian-American Company actually played do~vn the discovery of gold in 1848 lest it lose its monopoly, and tltcn, xvhen gold \\.as discovered in Canada along the Stikine River in 1862, the Company asked for an armed cruiser to protect their possessions front being overrun by American *Sources: Fedorova (1973); and Canaclian prospectors, but they never got that Petroff (1884); decennial reports of help (Gsovski, 1950, p. 29; Okun, 1951, pp. the U. S. Bureau of the Census. tOnlg heads of families 245-249). counted. $Returned to Atka, not Attu. U. S. TERRITORY "I Tile importance of birds in the blue fox diet is evident . . . they furnish 57.8 percent of the food, though the percentage varies on different islands, depending on availability. Land birds are relatively unimportant. They are hard to capture and do not gather in large groups. But the concentrated colonies of petrel, auklets, and related species furnish rich hunting grounds. Today the Aleutian Canada Goose is on the Secretary of the Interior's list of threatened species (Bureau of Sport Fisheries and Wildlife, 1973, p. 288) and apart from game farms apparently survives only on Buldir Island. \ire may wonder, how did the Biological Survey come to commit the ecological blunder of introducing foxes onto these islands? Three reasons present themselves. Exec- utive Order 1733, tvhich set up the Aleutian Islands Reservation, specified ". . . the propagation of reindeer and fur bearing animals" as one function of the Resel~ration. H. Douglas Gray (private communication), retired Alaskan Wildlife Agent, told me that in 1921 the Biological Siuvey *Source: Annual reports of the Alaska Game was staffed in this part of the world with old Comn~ission. jl cannot account for the fact that these trappers and other fur-oriented people. They were avuragr prices sorneti~t~esdiffer markedly from interested in what the land could produce and those to be inferred from Table 3. naturally turned to this "highest and best use." Probably too they wanted to help the econom- Amchitka the report said (klurie et al., 1937, p. ically depressed Aleut people, with \\'horn they 196): could not help but sympathize. . . . the ideal use of this island would be for Before \\'orld IVar I1 the Aleutian Islands wildlife, such as waterfowl and sea otters. However this Rese~vation was administered from ICodiak. In island plays an important part in the life of the Atka October 1948 the refuge received its own manager, natives an4 under present conditions we could hardly Robert D. Jones, Jr., to be stationed at Cold Bay withdraw the island from fox raising. However, the following solution is suggested. First, and furnished \\'it11 a dory to get about in. Jones it is barely possible that the importance of fox raising had come to the Aleutians during the war, spend- may be lessened in the future due to events at present ing a little over a year on Amchitka itself. unforeseen. In that case Amchitka could be put to its most fitting use as outlined above. Or, the natives may The goal of eliminating foxes from Amchitka be induced and assisted to practice more controlled first appears on the record in March 1939 in Gray's and intensive fox farming on Atka Island itself so that "Proposecl Plans for the Administration of the Amchitka Island will not be nceded. It should be Aleutian Islands Wildlife Refuge," which said, recalled that at present the Atka natives have Atka, "Amchitka Island has been taken out of fox Amlia, and Amchitka, all large islands. \Ye feel that it is well, in some manncr, to be alert for an opportunity to farming production." Then the war put thcsc plans release Arnchitka \vithout working undue hardship on aside. In 1948 Jones took them up again. How to the native Atka community. do it? Control was not enough; complete elimi- nation \\rould be the only satisfactory answer. In wife, Charles Foster Jones and Etta Jones. [Gol- 1950 Jones experimented with strychnine-loaded odoff (1966), an Attu Aleut, remembered them as pellets of seal blubber left along established fox Gene and Jane Foster.] Jones died in the after- trails. That siimnler it did not xvork, but, in the math of the occupation; it is not clear whether he harsher conditions of tlle follo~vingwinter, the was shot trying to escape or committed suicide pellets were eaten and were effective. Rat poison rather than surrender (Garfield, 1969, pp. 81, 310; (1080) was also set out, and, although the rat New I'orlz Til~les,6/21/43, p. 3; Bergsland, 1959, population of the island has never been more than p. 126). Mrs. Jones was separated from the others, reduced, their 1080-laden carcasses killed off the sent to a prisoner-of-war camp at Zentsuji, Japan, feral dogs and cats left by departing soldiers. The and returned to the United States in October 1945 poison-pellet campaign continued and was en- (11'ezu York Tilnes, 6/21/43, p. 3; 10/19/45, p. larged, supplemented by the shooting of such foxes 3). In September 1942 the Japanese evacuated as were seen. By 1960 tracks of a single fox were the Attu Aleuts to Otaru, Hokkaido. (The ~\'ezu found around an old mess hall not far from York Times erroneously reported thein as having Constantine Harbor. Signs of four to six others been taken to Sakhalin, but in 1943 their sources were found elselvhere on the island but nowhere of information from Japan were poor.) The Attu any signs of litters or dens. These, it turned out, people were put to work digging pottery clay. were the last foxes on Amchitka. After 40 years After the war the survivors were flown to they were gone (Jones, 1960; Berns, 1960). Okina~va and then to Manila, from \vhich they went by ship to San Francisco, by train to Seattle, and by ship to Adak. Golodoff says that they wanted to return to Attu, but that "we were not On June 3 and 4, 1942, planes from a Japanese enough people so the Government wouldn't let us carrier force struck Dutch Harbor on Unalaska. On go to Attu. The Government told us to live \\.it11 June 7 Japanese troops landed on Iciska and Attu. the Atka people." The reason that the Bureau of Indian Affairs gave was that Attu was too distant It was part of a plan to draw support away froin the impending Battle of h,Iidway. In that respect it to supply, and the villagers were sure to need a failed; Admiral Yaiilamoto lost all four of the subsidy and other assistance. The Attuans went to carriers he had committed to the h4idway engage- Atka, arriving there on Dee. 11, 1945, 18 of them: ment, and the tide of battle was turned in the 4 men, 3 women, and 10 children, the oldest 15, in Pacific. On the northern front, ho~vever,Japan was five families plus a baby boy born to R,Irs. Alfred ahead. War had come to Alaska, and henceforth Prokopioff about the time they left Japan (~Vezu service in Alaska warranted the Asiatic-Pacific York Times, 10/19/45). The rest of them had starved on the scanty and unfamiliar diet tlle Campaign Medal and ribbon. On ICiska there had been a 10-man Navy Japanese had given them. weather observation team. All were captured, In the meantime, in June 1942 the Americans although one man held out for 50 days, eating evacuated all remaining Aleuts west of Unimak, grass, wrorms, and shellfish. These men lived out including those on the Pribilofs. Atka itself was tlle war as prisoners of war in Japan; they were put evacuated on June 13-14, just ahead of a Japanese to work in forced-labor prisoner gangs in ship- air attack from Iciska. The Aleuts were taken to yards and steelworks. All survived to return home various places in southeastern Alaska: the Pribilof after the war (Garfield, 1969, pp. 80-81, 310). Islanders to Funter Bay on the north end of Attu village had been there since aboriginal Admiralty Island near Juneau, the Atka people to times and \\.as thus one of the oldest villages in the Killisnoo near Angoon on Admiralty Island, the Nikolski people to \\lards Cove near Icetchikan, and chain as \\,ell as the most distant from the others to Bernet Bay near Sitka. Here they mainland. Accounts differ as to ho\v many Aleuts were on Attu, various accounts putting it at remained for three years, until June 1945, suf- fering some\vhat from the different environment anything from 39 to 45.* Fifteen of then1 were and resi~lting respiratory diseases and from the children. There was also the schoolteacher and his cultural shock. Duncan (1945) ends his article on their prospcctive return, "But 11o\\v are you going *Mew I'ork Times (6/21/43): 42 Aleuts and 2 whites. to keep 'em out on the Chain after they've seen Duncan (1945): 45 Aleuts and 2 whites. Juneau? Indian Se~-\ricemen have discovered a little A'ew I'ork Tiines (10/19/45): 40 Aleuts and 2 whites. Berreman (1955, p. 57): 42 (not clear if whites wavering, especially among the younger Aleuts. included). '\\'e'll go back to the island no\\~,'someof them are Golodoff (1966): 40 Aleuts and 2 white. saying, 'and see how we like it. If we don't like it Garfield (1969, p. 79): 39 Aleuts and 2 \shites. there, maybe someIio\\~ \vc can get back here History, 1741-1967 125 again.' " More seriously, Berre~nan (1955, pp. child, 1964, p. 270; Morison, 1968, pp. 12-13; 52-53) says of Nikolski that the wartime removal, Craven and Cate, 1950, p. 374). Scouts were put the postwar return to find their village a shambles, ashore on August 28 to check for the presence of slow and inadequate resolution of their losses, and the enemy; there was no sign of the enemy, and general frustration at not being able to satisfy the the main force of 4556 men ment asllore on the \\,ants acquired during the removal \\,ere important 30th. \\'itllin 10 days an airstrip had been built by new factors in the continning disintegration of the putting a dike around a shallow lagoon and closing old Aleut ways and social integration. off the tide from it. The strip was hard-packed These evacuations left the Aleutians an empty sand and was later covered with steel matting. and open field for the prosecution of the military Today's airfield on Adak is concrete, at the same campaign. It \\.as to prove a messy, wet, inefficient place, and it still has to bc kept dry by pumping operation. The Americans concentrated on harass- the ditches alongside. ing the Japanese-held positions with as many On the 10th of the next month, a stnall party bombing raids as weather and supplies would reoccupied Atka. Atka had been nsed at various permit and developed a series of ever more forward times before for seaplane operations, and a land bases to facilitate that effort and to seive as bases strip was now built there. for finally dislodging thetn. The Japanese concen- In between these islands and Iciska lay two trated on holding, supplying, and reinforcing the other large islands, Tanaga and Amchitka, with positions they held; they started a few desultory land flat and lo\\' enough to be considered seriously movements fonvard, but these were never calried as possible sites for airstrips. The official army through. For both sides the Aleutian front was a history indicates that the initial plan was to move war poorly supplied and equipped. onto Tanaga (Conn et a]., 1964, pp. 270-276). Conditions made the \\Tar dangerous in more Reconnaissance parties had been sent to both \trays than just being shot at. Today Reeve Aleutian islands at the end of September, the one to Airways flies the chain on a regular basis; you can Amchitka being made by 34 scouts on September count on delivery safely to your destination ~vith 23. The History of the Alaskan Air Command says seldo~nmore than a fe\v hours' or a day's delay due of this reconnaissance: "In the report made by Col. to fog or winds. During \\'orldlVar 11 there were few \'elbeck, it would seem impossible to construct a of the aids used today: only beacons and radio suitable runway of any type on the island. With the direction finders, and those were scarce. There was exception of the recommendation that a landing "a complete lack of radio ranges, lighting facilities, field could not be built on Amchitka, Colonel and other navigational aids \vest of Umnak" Verbeck's report was of great value in the prep- (Craven and Cate, 1950, p. 397). Aircraft flew aration and planning of construction"! (History of visually where often there was little visibility, and the Alaskan Air Command, 1952). A plan was ships went where there were no charts more recent drawn up to move onto Tanaga on November 1, than those from Russian times. R~Ianyan airplane but reviews carried all the way up to General crew \\.as lost for no reason other than that they h,Iarshall eventually brought back the decision to could not find their way back or down. use Amchitka, the more forward island (Conn et In June 1942, just after the Americans dis- al., 1964, p. 275). covered the presence of the Japanese .on Kiska, the Preceded by an aerial strike to find out if any submarine S-27 was sent out to find out if the Japanese troops were on the island aud to destroy Japanese had yet occupied Amchitka. The S-27 the village on Constantine Harbor, including the lvent onto a reef, and the crew had to abandon church, so that there would he no hiding place, a ship, setting up camp in the church in the new sulvey was made on December 17-19. It \\'as unoccupied village at the head of the Constantine led by Colonel Talley, who had made his repu- Ilarbor. Six days later a patrolling PBY plane tation in the speedy job of airstrip building on found them there, and the next day, June 26, they Adak. His reports were instrun~entalin confirming were evacuated back to Dutch Harbor. There had the change of the proposed base from Tanaga to been no Japauese on Amchitka. Amchitka. He predicted two to three weeks would Bombing missions could be run from Dutch be needed for a fighter strip and three to four Harbor 01 Umnak, but it \\!as a Ions run and costly months for a main airfield (Conn et al., 1964, p. in men and machines. A more forward base \\.as 275; C~avenand Cate, 1950, p. 374). His party needed. The Army Air Corps nomiuated Tanaga; found signs that Japanese had been there-a boat the Navy nominated Adak. The Joint Chiefs of with no water in it from s~lo\vthat had fallen two Staff had to adjudicate and chose Adak because the nights before and nn\veathered rifle cleaning fine harbor there made it a natural for staging later patches along a trail. A story told during the AEC amphibions operations (Conn, Engelman, and Pair- occupation of Amchitka was that the airstrip was built according to Japanese survey stakes, but Talley (private communication, 1976) himself says that xvhat they found \\'as several test pits shorviug soil cross sections. The Japanesc had, in fact, made two abortive sorties tolvard occupying Amchitka. One in Oc- tober 1942 tumcd back \\,hen their intelligence service warned Admiral Hosagaya, falsely as we kno\v and he did not, that a strong U. S. Navy task force \vas in the vicinity. The other in late Novctnbcr 'ivas deterred by heavy air raids (Gar- field, 1969, pp. 148-149). On Jan. 12, 1943, 1945 men, under the command of Brigadier General Lloyd E. Jones,* \vent ashore at the head of Constalltine Harbor Fig. 2-Troops landing at the head of Constantine (Figs. 2 and 3). It was in the tail end of a storm, Ilarhor, Jan. 13, 1943. (U. S. Army photograpll SC and the uporst of the \\inter xvas still ahead, 17 1235-B.) probably as miserable ~veathcr for a landing as U. S. troops have ever had to make. Their first priority task \\,as to make a landing strip, and this they did in 5 weeks' time; so on February 16 a P-40 squadron \\,as able to come in and land. Aleutian xveather is such that often even in the worst of clouds and fog there is a hundretl feet or so of some\vliat clearer air down near the \\rater. A graben crosses Amchitka, dipping gently under the \\later at both ends to form Coustantine Harbor and hlakarius Bay. The nclv Amchitka fighter strip (Fox Rumva)') took advantage of these climatic and gco~norphologicalpeculiarities by going right from the water's edge gently uphill to the south- west (Fig. 4). This water-level strip saved many a pilot's life. The Navy also Itad a PBY patrol squadroll on the island that used a ramp into the sea at the foot of Fox Runxvay. After Fox Rumvay \\.as complete, \\rork was started on a bomber run\vay (Baker Runrvay). In Fig. 3-Scene of 1943 landing in May 1974. mid-June 1943 a third cross nulrvay (Charlie Runxvay) was started. In early 1944, after the Aleutian campaign itself was over, constructiou Fox Runrvay is no longer useful. It was began to enlarge Baker Runrvay to support B-29's concreted in the 1944 construction program, but, in bombing attacks on the Japiu~esehomeland. The sometime before the eventual pullout in 1950, run\vay construction \\,as completed, but the trenches were cut across it since it nras no longer B-29's never arrived: Rapid advances in the Central usable. \\'inter storms and thc 1957 tsunami have and South Pacific ancl poor \\reather in the Aleu- built a beach ridge acrus its foot; and the Rat tians made an Amchitka base lcss desirable than Island earthquake of Feb. 3, 1965, caused some of bases in the klariana Islands and in China (Craven the fill land in the foreground of Fig. 4 to settle and Cate, 1950, pp. 396, 400-401; History of the and block its drainage to the harbor (Fig. 5). Baker Alaskan Air Command, 1952). [B-29's finall) Runrvay \vas reuovated and improved cluring AEC made it to Amchitka 4 years later when t~voof occupancy of the island and is a good runway. A thetn landed therc after making observations of the beacon and landing lights remain, but these arc eclipse of May 8-9, 1948. This was the first time only operational ~vl~cntherc is a party of pcople on that aircraft were ever used for this purpose the island. Charlie Runxvay has not been main- (I Fig. 4-Foot of Fox Runway, Mar. 4, 1943. Motor pool is in foreground, P-40's on taxiway in rear, and derelict 8-24 in water beyond strip. (U. S. Army Fig. 5-Foot of Fox Runway, May 1974. Note photograph SC 571335.) subsidence of fill land from the 1965 Rat Islands earthquake. The American presence on Amchitka \\,as discovered by the Japanese on Jan. 23, 1943. 29, \vhen only 28 Japanese remained alive to Bombers immediately started hitting the runrvay become prisoners. The American forces suffered under construction, but these attacks were kept up 3829 casualties, including 549 dead, 1148 only a month. The very geography is against a1 xvounded, 1200 \\.it11 severe injuries from the cold, attacker acting from Iciska. Bird Cape, at the 614 with disease (inclttding exposure), and 318 northwest cnd of Amchitka, is only 50 miles from other casualties. h,lany mistakes were made, espc- IZiska; a small camp and radar site \\,as set up there cially in not taking good advice about the con- \vhich, ~veatherpermitting, was able to \tram those ditions to be expected in the Aleutians. \\re are told at the airstrip 30 miles to the east of approachi~tg that the services profited greatly from this expe- airplanes. rience (Garfield, 1969, pp. 208-258). Occupation of Amchitka, although protlucing On h.lay 11, of 168 Air Force planes in the no im~nediatespcctac~llar results, did help hake forward area, 86, or 51%, were on Amchitka the Japanese hold on the Aleutians hopeless. (Craven and Cate, 1950, p. 381). Systematic supply of their garrisons became haz- Kiska remained in Japanese hands. A force of ardous, requiring a pol\rerful task force to drive a 18,006 men, including 4983 Canatlians, landed convoy of ships through the air and ~lavalblockade. there on Aug. 18, 1943. As it titrned out, the Just such iul attempt was foiled 11y the Navy on Japanese had been able to evacuate their forces on bkarch 26, and thc Japanese doom was sealed. July 28, and the U. S. and Canadian troops IZiska \\,as the chief target of planes from encountered only mines and booby traps left by Amcllitka; any intermittent letup in the cloud the departing Japanese. There were 208 casualtip- cover permitted a bombing run on a target only 80 28 clead, 50 xvounded, and 130 with trench fuot iniles anray. \\'cather also played its part in making (Garfield, 1969, p. 288). IZiska the main target in another way. Bombers The expulsion of the enemy left the North sent to Attu frequently fo~undthat island closed in p.dc~fic " forces without an immediate mission. One and unloaded on Iciska on the return trip; so in possible one was using the Aleutians to attack April only 30 sorties \\,ere actitally made against northenl Japan or the IZuriles. The IZuriles had Attu. These did other important work, however; been Russian since the seventeenth century but the aerial pictures they took constituted almost the had been taken by the Japanese as part of the sole source of intelligence about Japanese troop spoils of the Russo-Japanese \\'ar of 1904 and strength (Cravcn and Cate, 1950, pp. 376, 379). 1905. They had come to be considcrcd part of the The overriding purpose of the i\merican mili- Japanese homeland, and Shimushu and Para- tary forces \vas to tlislotlge the Japanese from the mushiru, the first and second of the I TIIF. POSTWAR YEARS Fig. 8-Pacific hut protected by peat berms, winter 1944-1945. (Photograph courtes). of R. F. Shq, Some troops stayed in the western Aleutians Anchorage.) for a while after the war. Atnchitka itself was evacuated by August 1950 (Jones, 1960, p. 9). they required a site having "a minimum thickness Attu was empty by R,Iarch 1949, although at some of 150 feet of incoherent detrital material in which time since the Coast Guard has put 35 men there the seismic velocities \\~ould be essentially uni- to run a loran station. Atka was placed in caretaker form" (Piper, 1952). For some reasou it was status in January 1944. ICiska was not held for long thought this could be found on Amchitka. T\vo after it was found that the Japanese had aban- drilling parties were sent to the island, and in early doned it, perhaps because there is much unex- 1951 they diilled 34 test holes at various positions ploded ordnance on it. Shemya was evacuated in to the northwest of where Cannikin was fired in 1951 but was later used as a refueliug stop by 1971 (Chap. 1, Fig. 11). No such conditions were Northmest Airlines on the Anchorage-Tokyo found, and the exploratory drilliug was stopped 011 route and now again has an Air Force station on it. May 10, 1951. The scope of the project was Adak has been occupied continuously and indeed reduced to the 1.2-kt level and became the Jangle has gained importance when the Navy headquarters Surface and Jangle Underground shots fired in was removed from ICodiak in 1971 and the refuge November 1951 at the Nevada Test Site. headquarters to Adak from Cold Bay in 1972. In the early 1950s much of ~vltatis no\\, old, Adak now has the distinction of being the station rotten, and weatherbeaten was still in very good farthest west in the United States ~vheremilitary shape. This, of course, was also true in other dependents are allo~ved. former fields of war, and there too a great deal of In 1951 a proposed new project involved material remained, slurplus to the military but Amchitka, foreshadowing its recent use by the potentially valuable to someone. Some time after Department of Defense (DOD) aud the U. S. the war, probably about 1951, salvage rights were Atomic Energy Commission (AEC) as a site for put out to bid. The original bid invitations brought nuclear explosions. Project \\'indstorm \\'as a re- no response; so the Navy, \\.hich was acting as the sponse to a nlilitary need for information about executive agent for the military in this matter, the cratering potential of nuclear explosions. The entered into negotiation with Bankers Life and DOD x\~ished to fire tmo 20-kt explosions, one Casualty Conlpany of Illinois. After prolonged surface burst and one sl~allowvlyb~tried. For this negotiations, a figure of 8318,000 \\'as agreed on 130 Merritt for all the salvage rights on 13 islancls of the Refuge, these other activitics being tolerated icnder Aleutian chain (Anonymous, 1961). This firm the original provisio~lof Esecutive Order 1733 that intencled to do the job with Japanese labor, but the "the establisl~ment of this resenration shall not unions ol~jected;so they were forced to abandon interfere with the use of the islands for lighthouse, this idea and sublet the rights (Robert At\\,ood, military, or naval purposes, . . ." The Department private communication, 1973). The first salvage of the Interior resu~nedmanagement ovcr most of operation \\'as by Aleutco, a Seattle firm \vhose the refuge territory in 1948 \\'it11 the appointment name suggests that it was formed ad hoc for this of Jones as Refuge hjlanagcr. Reading his reports particular job. Aleutco took only items of large from 1950 on, \i7c see a series of activities that value that could be easily transportecl, things on continuecl year after year. Some of them were pure \\!heels so to speak: trncks, generators, compres- l~ousekceping matters, such as permits issned, sors, and installed real property, such as motors. A poachers caught, and trips made. There are also second salvage was co~lductedin 1953 by General extcnsi\~c lists of birds seen and identified \\,it11 Arletals of Tacoma. They salvaged metals, partic- their ti~tlesof arrival and nesting. There are details ularly copper and lead (R. D. Jones, private on rats trapped on Amchitka (Brecltbill, Chap. 12, commutlication, 1973). These were "high-grading" this volume). Thcrc arc lists of plants. But the operations; for instance, the copper of telephone do~ni~la~ltthemes, cspecially for Amchitka, are the wires \\.as often taken by felling the poles, but this sea otter and the Alcutian Canada Goose. was only done in the southeaster~lpart of Am- In 1958 the refuge \\,as described as follows: chitka where there was much of it. The salvage Tlle Aleutian Islands National \\'ildlife Refuge was rights ha\>e since either been bought back by the estahlisl~edin 1913 to insure protection of the native government, as in 1955 \vhen Shemya \vas to be buds and fur bearers, to provide for the propagation of made a refueling stop for Northxvest Airlines, or reindeer, and to encourage and develop tlte fisheries. simply expired, as seems to have happened to 11s we now know, the propagation of reindeer is often Amchitka. inin~ical to populations of nesting birds, and to tlle range itself if the animals are not carefully managed. During the years 1954 to 1957, a Distant Early Tbat and the Aleut people's lack of interest in reindeer \\'arning radar netx\~ork(the so-callecl DE\\' line) has dictated a "go slo\v" policy in reindeer propaga- was built across northern Alaska and Ca~lada.A tion, Because of remoteness and the existence of severe DE\\' line extension uras I-1111into the Aleutians xvcathcr conditions, development of tl~efishery has starting in 1957. The communications portion of proceeded slo~vly.Even today the inshore fishery of all but the eastern end of the Refuge is rarely utilized. The the DEL\' line was called the \\'hitc Alice system. Japanese 1,igh seas fishing fleet is currently engaged in (Since then the \\Ihitc Alice system first became a the development of the off-shore fishery. It is mainly separate entity selling also the isolated bush to\vns in the field of native bird protection and the manage- of Alaska, and II~IVlarge portions of it are ~nelded ment of fur bearers that Refuge administration has moved. The redevelopn~cntof a sizeable population of into the state\vide telephone ser\rice nun by RCA sea otters bas been the most significant step in the Communications.) In con~lectiotlwith H1hite Alice history of tile Refuge. When in 1913 thc Refuge came operations and the DE\\I line estcnsion, \\'cstcrn into being a sea otter was a rare specimen, hut there are Electric creurs \\.ere on Amchitka from 1959 to now more than 20,000 of tlle curious little animals witbin tile Refuge boundaries. 1961. Their ~nissio~l\\,as to run a relay st a t'1011 ~~ . ~~ ...~..~. ~ using a 30-ft antenna dish to bridge the 400-mile (In spite of thc con~~nentabout reindeer, caribou, a gap bet\\reetl Adak and Shemya during the time closely related species, were-introduced to Adak in those stations had only 60-ft clishes. T11ese and 1958 and 1959. Thcy are kept under close control other main DE\\' line stations ~lolvuse 120-ft dishes by hunting; indeed, G. Vernon Byrd, the present and have ample power and sensitivity to briclge Assistant Rcfuge A,Ianager, \vhile in the Navy on such gaps. The actual \\Thite Alice site is in an area Aclak had as part of his cluties as Base Consewation not damagecl in IVorlcl \\'ar 11; it is a little over a Officer assisting the Fish and \\'ildlife Service in the kilometer north\\~cstof \\here Cannikin \\,as later lnanagelllellt of this game herd.) The sea otters of to be fired (Chap. 1, Fig. ll), coi~lcidentallyfairly the \vestern Aleutians, and especially Amchitka, close to the 1951 proposed sitc for Project were estensi\rely studied, resulting in a Ph. D. \\'indstorm. From this short time activity co~llcs thesis by Lensink (1962) and a monograph by the llanle of the \\'bite Alice Creek that drains the ICenyon (1969). Renyon describes at length the Cannikin sitc and figures clse\vhere in this book. esperi~nc~ttsthat have led to rcccnt apparently successfitl transpla~ltsof sea otters from Amcllitka POST\\'AR REFUGE I\IANf\GEMENT and elsewhere where the population is now stable into parts of their fornter rangc where they llad Anlchitka and thc westcrn Aleutians \\.ere still been completely estirpated by hunting (Abegglen, part of the i\leutian Islands National \\'iltllife Chal~. 20, this volume). I have already indicated the plight of the beginning in 1913 it was reindeer, fits bearers, and Aleutian Ca~radaGoose and how the introduction fisheries. In 1958 it \\.as not reindeer but sea otters, of foxes to most of the Aleutians all but wiped birds, and, to a much smaller extent, fisheries. The them out. Jones' reports give a picture of a most recent statement of purposes is in a recent dedicated long-term effort to reverse this sad Wilderness Study Report (Sekora, 1973): history. Poxes were extirpated from Amchitka by 1. To maintain at minimum recovery levels all 1960 and from Agattu by 1968,* the latter island native species normally associated with the environ- being the one that IvIurie (1959) had believed \\'as ments found on the Refuges-more specifically, to once the very best island for the geese. Jones protect and preserve populations of colonial nesting suspected from A,Iuric's reports that these geese had birds, marine mammals, and other \vildlifc indigenous survived in small numbers on Buldir and verified to the Aleutian Islands and their adjacent waters. 2. To assure the survival in a natural state of each this in 1961. In 1963, 18 goslings were removed of this Nation's plant and animal species-nlore from Buldir and sent to a game farm in Colorado. specifically to restore the Aleutian Canada goose and Sixteen of these sulvived and formed the nucleus sea otter populations to former ranges and levels of of a captive population now at the Pati~xentgame abundance. farm in hflaryland. Thus Joncs (not single handedly, 3. To contain all lands or networks of lands of of course, but as the driving force) had done away national significance whose benefits to the public can best be achieved by the distinctive competence of the with the primary enemy of the geese on two National Wildlife Refuge Systen-more specifically to islands, had found a suviving nucleus of them, and preserve watersheds rvhich contribute materially to the had arranged for them to be cared for and to production of salmon stocks. multiply in numbers. It only remained to reintro- 4. To seek out, identify, designate, preserve, and duce the geese to these islands, tvhen the DOD and appropriately use sites and objects on refuges that are recognized to have esthetic, historic, geologic, mcheo- later the AEC took over Amchitka for other logic or scientific values. purposes and frustrated the long-term plan. 5. To raise to optimum levels the kinds, range, Actually it has not becn quite as bad as all that. amount, and quality of wildlife and wildlands- In the spring of 1971, the Fish and 'IVildlife Service oriented recreation-more specifically, to manage did bring back Aleutian Canada geese to Amchitka game populations [caribou] on Adak Island as a halvestable resource. using AEC support to transport them. They did 6.To establish and preserve in a natural state not bring them back in an effort to transplant, selected areas for reference observation, scientific horvever, so much as in an experiment on trans- study, and/or specialized public use, and in ~vhicl~the planting methods. The proven systcm for trans- major ecological communities in the system are repre- sented. planting migratory birds consists of bringing a group of adults into a place and holding them in Statehood came to Alaska in 1959 after a long cages. The adults rear young, and these young, tvho and difficult struggle. The Territory of Alaska had are inlprinted \\it11 this place, are released. This never had much control over its natural resources, standard method requires a crew to provide con- and indeed this had becn one of the sore points ti1111ous care for several years, usually about 3. that led to tlte demand for statehood. In the What was tried in 1971 was to see \\'hat could be Aleutian Isla~ldsNational \\'ildlife Refuge, state- done on a much shorter time scale. Seventy-five hood has meant some sharing of responsibility by geese were brought in and released shortly before the Department of the Interior xvith the nexvly the nesting season. Very few live birds were seen established Alaska Department of Fish and Game. after the release, and the few dead birds seen had (The old Alaska Game Com~nission ceased to either been killed by eagles or eaten by then1 after exist.) Today one must have a state fishing license death from other causes. There \\,as no evidence of to fish for salmon or Dolly Vxden on Amchitka. production of young birds or the returli of tlte More important was the control and management geese in the follo~vingyears. Thus this experiment of the sea otter. Technically the State acquired provecl unsuccessful. The Patuxent game farm has jurisdiction over sea otters at sea, between the yet other Aleittian Canada geese, and in 1974 the lotver tide line and thc 3-mile (5-km) limit of U. S. Fish and \\lildlifc Service reintroduced them to sovereignty, and the Department of the Interior Agattu and in 1976 started to return them to kept jurisdiction over sea otters on land. This was a Amchitka. fine and artificial line that did not meet the Therc has been a slow change in management practicalities of the situation. In practice, the State purposes for the refuge over the years. At the has managed the sea otter resource, issuing permits for their taking for such purposes as research, transplanting them into other parts of their former *So reported Jones, but he \%.asmistaken. A crerv bringing geese to Agattu in tbe spring of 1974 found over range, and hanresting thcir furs for sale where thcir a hundred foxes still there. density of numbers permitted. All this, of course, has been rendered moot by passage of the A'larine to haul the dory higher on the Island. Tbis we did and Aclammal Protection Act of 1972, \\~hichhas re- continued to watch with apprehension the silent ebb turned responsibility for all sea mammals to the and flow. Clearly the sea was in a state of unrest doe to some seismic disturbance. Later we learned of a Department of the Interior and has resulted in a damaging earthquake at that time on the coast of Chile moratorium on taking them for any pnrpose in South America (Jones, 1960, p. 20). (Abegglen, Chap. 20, this volume). The Great Alaskan or Good Friday earthquake of Mar. 27, 1964, produced a tsunami that was TSUNAMIS very destructive in the epicentral region and along the Pacific coast as far as California, but, because Before finally turning to the events that led to of the directionality of the source, the resulting nuclcar testing on Amchitka, mention should be tsunami was little felt in tlte Aleutians. made of another external influence on the island, No one was on Amcltitka on Feb. 4, 1965, one provided by nature, not man. It is well known xvhen tlie magnitude 7% Rat Island eartliquake that the Aleutians are a region of the \vorld that is took place .rvitlt its epicenter just 20 miles south of very active seismically (Gard, Chap. 2, this Amcltitka. This \\,as during the period leading up to volume). Earthquakes sometimes give rise to tsu- Long Shot (belo\\'), and so a party was sent to see namis or seismic sea xvaves (often miscalled "tidal \\,hat damage had been caused on the island. Gard \vaves"). Cox and Pararas-Carayannis (1969) list (1965, p. 4) reports that a water wave ltad xvaslted tsunamis that have been felt in Alaska so far as the up on the beaches, "no higher than about 6 feet historical record reveals them. Presumably tsu- above high tide," and that the fill land along tlie namis have been felt in the western Alclitians for foot of Fox Runway had settled (Fig. 4), but no as long as those islands have existed, but they do major damage was done. not appear in the record until 1944, after which there are 15 entries (Table 5). These were all recent enough that measurements at standard tidal sta- ACTIVITIES LEADING UP TO THE AEC tions exist on them; four of them had amplitudes USE OF ARICHITKA of a meter or greater at Attu or Adak. For three of them observations at Amchitka itself are available. The westcm Aleutians must ltave cropped up The 1957 tsunami, which arose from a source numerous times in the postlvar years as possible near Unimak in the eastern Aleutians, is mentioned sites for projects requiring isolation from masses of by Jones in Itis Narrative Report of 1957: people. I have in my files, for instance, a short memo written in 1960, during the days of great . . . . A tidal wave reaching 13 ft above high tide followed the most severe of the ewthquakes, washing optimism in the Plowshare program,* that dis- debris onto the lower part of the islands and creating cusses the feasibility of using nuclear explosives to additional damage at Adak. At Adak agitation of the dig a harbor at Sltelnya. It was just a paper water sources for trout hatchery apparently released exercise, one among many such that were not the accumi~latedproducts of decomposition and tem. taken seriously and not followed up on. porarily polluted water flowing in the troughs. Ten to eleven thousand fingerlings were lost before the situa- tion was restored. \\'ltile the Refuge Manager was on Early Site-Selection Efforts the Island in mid May the settling basin and piping system were completely cleaned of sediment as an The period of the early 1960s \\,as one of many effort to prevent or ameliorate a recurrence. Most of searches for test sites. During these years limi- the cabins and other facilities remaining from the fox tations began to become evident in the use of the farming days were demolished by this tidal wave. existing Nevada Test Site (NTS). After the unof- Jones' report concerns Adak; Everett (Chap. 8, this ficial mo~atoriumon nuclear testing was broken by volume) gives some evidence of tlte effects of this the U.S.S.R. on Sept. 1, 1961, the NTS ltad been tsunami on Amchitka. used almost exclusively for underground tests. The At the time of the 1960 Chilean earthquake city of Las Vegas, 60 miles front Yucca Flat, was and resultant tsunami, Jones was on Amcltitka. He beginning to acquire high-rise structures along tlte reports: "strip." High-rise buildings tend to be sensitive, one might almost say near-resonant, to tlte ground A curio;,s event, one that made us uneasy took place on the 23rd and 24th of May [later than the time shocks from the larger nuclear explosions. The indicated in Table 5 because of the travel time from initial step in adjusting to this gro\ving problcnl South America]. At noon of the former we noted a \\'as to enlarge tlie test site to the northwest to visible tidal current flowing into the Harbor. Thoughts of a damaging tidal wave leapt to mind and we watched with anxious eyes, expecting a bore to develop. The *Plowshare rvas the AEC program of study of non- current oscillated in and out and we thought it prudent military uses of nuclear explosions. Table 5-Tsunamis in the \\'estern Aleutians* Rin~opheight (111) at Date Earthquake (GhIT) source hlagnitude Attu Adak Unalaska Dec. 7,1944 Japan h,Iar. 4, 1952 Japan Nov. 5,1952 Kamchatka Nov. 26,1953 Japan Mar. 30, 1956 Karnchatka hlar. 9, 1957 Unimak Nov. 7,1958 S. Kurile Nov. 13,1958 S. Kurile 0.2 <0.1 An aftcrshock? May 4,1959 N. Kurile 0.2 h'iay 22, 1960 Chile >1.8 1.5 0.8 Dec. 20,1962 Fox Island 0.1? Uncertain if real. Mar. 27,1964 Prince li'illiam Sound 0.4 0.3 0.4 Feb. 3, 1965t Rat Island 3.2 NA 0.2 Mar. 29,1965 Rat Island 0.2 An aftershock. Oct. 17,1966 Peru 0.1 0.1 *Source: CoxandPararas-Carayannis (1969). A.~Tlme rvas 0521, February 4, GhIT, which was 1821, February 3, local time (BST). include Pahute hslesa, \vhich raised the distance to of the AEC's Nevada Operations Office, James Las Vegas to over 100 miles. bl subsequent years Reevcs, insisted that it is just not possible in the use of Pallute R~Iesa has permitted tests of American society to carry out such a test without soniewhat larger than a megatoll at the NTS. ktlo~vledgeof its preparation leakiug out. So the These limitatio~lsled in 1964 to an exteusive secrecy \\,as dropped and with it the feature of an search for an alteruative to the NTS. This search unannounced seismic event, but the test went \\.as confined to areas in the 48 contiguous states ahead on the issue of determining travel times from \vest of the Cordilleran front aud turned out to be sources near oceanic trenches. Thus was born Long a mere paper study; no visits to possible sites were Shot. Joues (1965, p. 35) co~lfirrnsthe misgivillgs permitted, and the 1964 alternate test site efforts about being able to keep such a test secret, saying, died away without ever being formally canceled. "In Marclt 1964 we learned of a major, highly classified project soon to begill on Amchitka, aud Long Shot assumed \ere were again in the presence of the In late 1963 planuing started in secret for what Bomb. . . . In the spring a series of heavily loaded is norv known as Lolig Shot. At that time there military aircraft flights proceeded to Amchitka, were unans\\rered questions about the ability of the altd tlle nature of the cargo (\vhich became U. S. detection system and the \vorldwide seismic common knowledge along the route) collfirmed community to detect and distinguish nuclear ex- our assumption." The cargo inclurled a Canadian- plosions, especially if one happened to be fired in made drilling rig xvltose presence became kno\vn the Icuriles or in Kamchatka, naturally seismic \vhen the U. S. customs held up its passage through areas llext to the sea. Could an explosion in such a Anchorage some hours for lack of proper papers. place be distinguished from natural seismic shocks? During the sumnler of 1964, six test holes were To \\,bat extent ~souldtlte Aleutian trench, a major drilled about the eventual Long Sllot site to exception to the usual seismic assumption of a confirm the Geological Survey's a~~alysisthat there horizo~ltallyunifonn eartli, speed up or delay a was no local faulting. Preparation for even a simple signal and cause its epicenter to be mislocated? It test is not short, and on Oct. 29, 1965, Lo11g Shot was proposed to shoot an underground nuclear test was finally detonated. It was an 80-kt shot 720 n~ in that region unannounced and see how tlte (2350 ft) deep 2 kni southwest of Cyril Cove seismic commu~lity \vould do (Test Command, (Chap. 1, Fig. 11). Defense Atomic Support Agency, 1967). From tlle refuge vie\vpoint, Long SIlot was an Ho\vever, the AEC's statutory responsibility unexplained interference with the long-term plans for test safety included even those tests carried out for Amchitka by nien illsensitive to refuge valoes. by the Depatment of Defense, aud the AEC had Icenyon (1966), in his internal report on the shot, to be brought into the planning. The then manager speaks several times about the "progressive destroc- tion of refuge values that are added by each new Amchitka invading activity." He mentions meeting nlen on Between NOV. 30 and Dec. 16, 1966, a U. S. the project \\rho Itad no idea before coming that Geological Sulvey team made a geological recon- the island was part of a National Wildlife Refuge. naissance and aeromagnetic stu-\'ey of most of the Normally a happy man, he was dismayed that he southeastern half of Amchitka. Their immediate and Spencer, the Associate Supelvisor of Alaska purpose was to select five sites for deep cxplor- Refuges, had to "join, as temporary employees, the atory drill holes. They picked five, plus one staff of the Laboratory of Radiation Biology of the University of \\'ashington in order to make post- alternate location, and reported their findings to blast examination of refuge wildlife resources at the Site Selection Committee on Mar. 2, 1967. (Unimaginatively, these six sites were called Sites A, Amchitka." B, C, D, E, and F.) A generalization of their What Long Shot and the two later nuclear geological findings during this reconnaissance to- shots did to these wildlife resources is discussed in gether with data gathered earlier is given in Carr, Chap. 26 of this volume. Suffice it to say here that Garcl, and Qninlivan (1967). the blast-caused effects of Long Shot were almost The following criteria mere used by the Site nil, but there resulted a small (and still continuing) Selection Committee in its 1966-1967 site- tritium leak into drilling mud pits next to the Long selection activities: The rock needed to be com- Shot pad (see Seymour and Nelson, Chap. 24, this petent ellough for mining at shot depth and volume). The island was evacuated in December impermeable enough to avoid the risk of ground- 1965, with little effort having been made to clean water contamination, all at a depth sufficient for up the island. containment of radioactivity. (Amchitka's isolation ~vouldhelp if containment were to fail.) There Search for a Supplemental Test Site should be adequate separation of individual sites In mid-1966 the search for another test site on the island; sites were to be 4 miles apart, 3000 was revived, this time impelled by the need for a ft from either coast, and at least 3 miles from place to test a series of devices leading up to the Baker Runway. Logistical factors favoring Am- warhead for the U. S. Anti-Intercontinelltal Ballis- cllitka were year-round access by sea or air, an tic Missile (AICBM) system, \vhich \vould include existing road network, a harbor, airstrips, and a yields considerably greater than what experience land mass large enough that people at a Control since 1963 had indicated could be tolerated at the Point at the north~vestend of the island ~vouldbe NTS's Pahute Aslesa. (The NTS limitation continued far enough from any prospective firing site to to be high-rise structures in Las Vegas, whose protect them from the effects of ground shock. As numbers and heights were steadily increasing.) The to environmental concerns, it was recognized that investigatioll proceeded informally from June until Amchitka is in the Aleutian Islands National Wildlife Refuge and that operations ~vouldhave to Nov, 18, 1966, at \vhich time a new Site Selection Colnmittee was formally set up. The search had by be carried out with minimnm disturbance to the then tlarrowed do\vn to an area in central Nevada natural environment and wildlife. north-northeast of the NTS and to Amchitka It would seem that there ought to be a formal Island in the Aleutian chain, with the \vestern end Memorandum of Understanding on record between of the north slope of the Brooks Range in Alaska the AEC and the Department of the Interior on the considered an "insurance" alternative in the event use of the island, but a search for one has not the other sites did not meet programmatic require- turned it up. Apparently the AEC's original entry ments. onto the island was based on an informal agee- ment made in late 1966 between AEC Chairman Much of the early thinking of this Site Selec- tion Committee was devoted to central Nevada. Seaborg and Secretary of the Interior Udall. Eventually a general location in Hot Creek Valley was chosen; particular sites within it were sug- ORGANIZATION OF BIOLOGICAL STUDIES gested by the Geological Survey; cxploratory holes ON AbICHITKA were drilled; and, after the long process of prep- aration, on Jan. 19, 1968, a calibration shot called In 1966 Battelle Columbus Laboratories (BCL) Faultless was fired. It was an intermediate yield was nlanaging a program of ecological studies in shot (200 to 1000 kt) at a depth of 975 m (3200 Panama for the AEC and the Atlantic-Pacific ft). No fiurther work has been done at the Central Interoceanic Canal Coinmission associated with the Nevada Supplemental Test Site. The site is not possibility of digging a new canal. They were asked suitable for shots as large as were rcquircd to test to organize studies at Amchitka as well. Their ideas the AICBM warhead. and plans were discussed nit11 the Fish and \\'ildlife History, 1741-1967 135 Service, first in Marc11 1967 and periodically responsibilities. In general, the island was to be thereafter. Actual on-island biological studies returned insofm as possible to its condition in started in the fall of 1967. These investigations 1967 before coilstruction activities started. The were designed to predict, evaluate, and document debris left from Long Sliot was to be cleaned up, the effects on the biota and the environment from but not World War I1 debris or buildings, except in the proposed nuclear tests, to recommend mea- specific instances along roads: it was believed that sures for minimizing adverse effects, and to predict many of these wartime remains were so covered and evaluate the potential hazards to man that \\,it11 subsequellt gco\\~tl~that their removal \\~ould might result from the accidental release of radio- entail more damage than benefit. iluclides to the environ~neiltand their s~~bsequent The sto~yof activities on the island during its trausport to humans via marine food chains. use by the AEC is left for later chapters of this These broad objectives had to he brought down book. All people evacuated the island on Sept. 8, to the level of specific concerns and particular 1973, after demobilization and cleanup, and AEC projects. The first list of specific concerns, agreed rights of use and control over Amchitka were to \vith the Fish and Wildlife Selvice in 1967, is as relinquished on Sept. 18, 1973. T\vo continuing follows: stipulations remain from the AEC occupancy of 1. Effects on the sea otter population. the island. The Department of the Interior has 2. Transport of radionuclides to man via ma- been asked to prohibit excavation, drilling, and rine processes. removal of materials near the surface zeroes of the 3. Trmsport of radionuclides to man via migra- threc nuclear tests that have been executed there, tory birds. Long Shot, h,lilro~\~,and Cannikin. Also, the AEC 4. Disturbance of terrestrial habitats. has asked for visitation rights and use of residual 5. Disturbance of freshwater habitats. facilities and equipment remaining on Amchitka The broad objectives and the specific concerns for scientific parties examining possible long-tern~ within them changed somewhat as time went on; consequences of those tests. So far there have beell for instance, the concern about sea otters changed three wc11 return parties, in IvIay and August 1974 from an early study of their ability to ~vithstand and in August 1975 (Kirkwood, 1974; 1975; shock \craves in the water to a study of their Nelson, 1975). population distribution and behavior. Also, the early emphasis on radionuclide transport chauged REFERENCES to an emphasis on possible physical effects on the environment. Anonymous, 1947, Building the Navy's Bnses in IVorld IVarII. History of the Bureau of Yards and Docks and Competent investigators had to be recruited. the Ciuil Engir~eerittgCorps, 1940-1946, Vol. 2, Super- Some of these came from ~vithinthe Battelle staff; intendent of Documents, U. S. Government Printing others were enlisted from the academic community Office, Washington, D. C. aud government laboratories. Later a few inves- -, 1961, Opinion Estimate of Improvements at Shemva,. . tigators entered the program under direct contract Alaska. From Corps of ~n~ineersfiles, Ancborage (May 16, 1961). with the AEC. Most of the contributors to this Atwood, Robert (Ed.),1973, At~chora~eDaily Titnes. book are drawn from this group of investigators. Bancroft, 11. H., 1886, History of Alaska, 1730-1885, The Fish and Wildlife Sewice of the Depart- A. L. Bancroft & Co., San Francisco, Calif. ment of the Interior assigned four men to tlle Bank, T. P., 1958, The Aleuts, Sci. Anter., 199: 113. Bensin, 8. hI., 1967, Russian Orthodox Church in Alaska, Ainchitka activities: txvo refuge-management biolo- 1794-1967, Russian Orthodox Greek Catllolic Church gists who were rotated to keep one man on the of North America, Diocese of Alaska, Sitka. island at all times to moilitor construction ac- Bergsland, K., 1959, Aleut Dialects of Atka and Attu, tivities and othcnvise to look out for Interior's Tram Anler. Pltil. Soc., 49(3): 1-128. interests and two research-oriented men to monitor Berkh, V. N., 1823, A Chronological History of the Discove~yof the Aleutian Islands. Translated in 1963 and participate in the research being carried on. by M.B. Ricks as Berkh's IIisto~y of the Aleutian (These latter men, Merrell and Abegglen, are also Islands, 1823, in Tlte Earliest History of Alaska, Cook anlong the authors of chapters in this volume.) Inlet Historical Society, Ancborage. Also translated by R. A. Pierce, Limestone Press, Kingston, Ontario, 1974. In April 1967 the Department of the Interior lhlaterials for the Study of Alaska History, No. 5.) aid the AEC agreed that demobilization at the Berns, V. D., 1960, Fox Report of Amchitka and Surround- conclusion of the program would include cleanup, ing Islands. Report to the Predator and Rodent Control removal of all movable material brought onto the Supelvisor, Dec. 13, 1960. From U. S. Fish and \\'ildlife Selvice files, Anchorage. island, and demolition of buildings built for the Berreman, G. S., 1955, Inquiry into Community Integra- project except for xvhat the Department of the tion in an Aleutian Village, Anter. Attthropol., 57(1): Interior \\rould need for its continui~lg refuge 49-59. Bureau of Sport Fisheries and IVildlife, 1974, Proposed Golodoff, I., 1966, The Last Days of Attu Village, Alaska Aleutian Islands Il'ilderness. Aleutiott Islands Natiottal Sportsman, December 1966, pp. 8-9, as told to Karl W. IVildlife Reficge, Alaska, Department of the Interior, Kenyon. Draft Environmental Statement DES 74-21. Gray, H. Douglas, 1973, former Alaska Game Warden, -, Threatened IVildlife of the United States, 1973 Alaska Game Commission, Juneau, private communica- edition, Resource Publication 114 (revision of Resource tion. Publication 34). Gray, H. D. (?), 1937, Memo Notes on Amchitka Status. From U. S. Fish and Wildlife Service files, Juneau. Carr, \V. J., L. h,l. Gard, Jr., and \\I. D. Quinlivan, 1967, Geological Reconnaissance of Amchitka island, Decem- Gruening, E., 1968, The State of Alaska (rev. ed.), ber 1966, USAEC Report USGS-474-42, U. S. Geolng- Random House, Inc., New York. Gsovski, V., 1950, Rtcssian Ad~ninistrationof Alaska and ical Survey. the Status of the Alaskan Natives. Foreign Law Section, Chevigny, H., 1942, Lord of Alaska: Baranov and the Library of Congress. Senate Document 152, 81st Russian Aduentsre, Viking Press, Inc., New York (re- Congress, 2nd Session. printed by Binsford & Mort Publishers, Portland, Ore., Guggenheim, P., 1945, An Anthropological Campaign on 1965.) Amchitka,Sci. ~llonthly,61: 21-31. Clark, A. H., 1910, The Birds Collected and Observed Hinckley, T. C., 1972, Tlte Antericanization of Alaska, During the Cruise of the United States Fisheries 1867-1897, Pacific Books, Palo Alto, Calif. Steamer Albatross in the North Pacific Ocean, and in History of the Alaskan Air Command, 1952, Information the Bering, Okhotsk, Japan and Eastern Seas, from on the Amchitka Air Force Base, included in a letter April to December, 1906, Proc. U. S. Nnt. dlus., 38: from G. E. Hasselrvander, Historical Research Branch, 25-74. hlaxwell AFB, Alabama, to C. M. Hardy, Bureau of Conn, S., R. C. Engelman, and B. Fairchild, 1964, Guarding Sport Fisheries and Wildlife, Anchorage, dated Apr. 15, the United States and Its Outposts, in series U. S. Army 1971. Source appears to be a History of the Alaska Air in \\'orld War 11. The \Yestern Hemisphere, Office, Chief Command, 1 July-31 December, 1951, Vol. 111. of blilitary History, Department of the Army, \Vashing- Hulley, C. C., 1970, Alnska: Past and Present, third edition, ton. Binsford & Mort, Publishers, Portland, Ore. Corps of Engineers, 1945, Records of the Seattle District, Jones, R. D., 1973, U. S. Fish and Wildlife Service, Corps of Engineers, for the Period Ending 30 June, Cold Bay, Alaska, private communication. 1945. Manuscript sheet in the author's possession. -, 1965, Narrative Report, Jan. 1, 1965-Dec. 1, 1965. Cox, D. C., and G. Pararas-Carayannis, 1969, Catalog of Aleutian Islands National FViidlife Refuge. From U. S. Tsunamis in Alaska, U. S. Coast and Geodetic Survey, Fish and \Vildlife Se~vicefiles, Anchorage. \Vorld Data Center Report WCDA-T-69-1. - 1960, Refuge Narrative Report, April Through August Craven, \Y. F., and J. L. Cate (Eds.), 1950, The Army Air 1960. Aleutian Islands National Wildlife Refuge. From Forces in IVorld lVar II, Vol. 4, The Pacific: Guadalca- U. S. Fish and Wildlife Service files, Anchorage. nal to Saipan, Auptst 1942-July 1944, The University -, 1957, Quarterly Refuge Narrative Report. Izemheck of Chicago Press, Chicago. and Cold Bay Area and the Aleutian islands National Dall, IV. H., 1877, On Succession in the Shell-Heaps of the \Vildlife Refuge, 1June, 1956, to 1June, 1957. From Aleutian Islands, in Tribes of the Extrente Northwest, U. S. Fish and JVildlife Service files, Anchorage. Jordan, D. S., and G. A. Clark, 1898, The History, \\I. H. Dall, J. Furuhelm, and George Gihhs, Contribu- tions to North American Ethnology, Vol. 1, pp. 41-91, Condition, and Needs of the Herd of Fur Seals U. S. Geographical and Geologic Survey. Resorting to the Prihilof Islands, in The Fur Seals and Fur-Sen1 Islands of the North Pacific Ocean, Vol. 1, Duncan, Sgt. Ray, 1945, The Aleuts Go Home, in Yank, D. S. Jordan and various associates, Superintendent of Alaska edition, May 18, 1945. Documents, U. S. Government Printing Office, \Vashing- Fedorova, S. G., 1973, Tlze Russian Population itt Alaska ton, D. C. ottd Califonria, Late 18th Century-1867 (translated Kenyon, K. \V., 1969, The Sea Otter in the EnsternPacific and edited from the 1971 Russian original by R. A. Oceas, U. S. Department of the Interior, Bureau of Pierce and A. S. Donnelly), Limestone Press, Kingston, Sport Fisheries and \\'ildlife, North American Fauna, Ontario (Materials for the Study of Alaska History, No. 68. No. 4). , 1966, hlilitary Use of Island Refuges, Including Findlay, A. G., 1886, Directory for the Nauigatio~iof the Operation Long Shot, Sea Otter Surveys and Sea Otter North Pacific Ocean aith Descriptions of Its Coasts, Transportation Experiment, October-December 1965. Islands, etc., from Pnnama to Behring Strait, and Japan; From U. S. Fish and IYildlife Sentice files. Anchorage. Its Il'inds, Currents, and Passages, third edition, Lon- Kinney, \V. A,, 1949, Operation Eclipse: 1948,;~ot. ~eogr. don. Ma.e.. 95: 325-372. Gard, L. M., Jr., 1974, U. S. Geological Survey, Denver. irkw wood, J. B. (Comp.), 1975, Bioenvironmental and -, 1965, Effects of theFebruary 3, 1965, Earthquake on Hydrologic Studies, Amchitka Island, Alaska. Fall. Amchitka Islo~id, Alaska, U. S. Geological Sulvey, 1974 Task Force Report, ERDA Report BMI-171-158, Technical Letter, Long Shot-2, Mar. 12, 1965. Battelle Memorial Institute. - 1974, Amchitka Bioenvironmental Program. Bioen- Garfield, B., 1969, The Tltousand-Mile ll'ar. IVorld Il'ar II in vironmental Safety Studies, Amchitka Island, Alaska. Alaska and the Alerrtiatis, Douhleday & Co. Inc., Spring, 1974 Task Force Report, USAEC Report New I'ork. BMI-171-157, Battelle hlemorial Institute. Golder, F. A., 1922, Beriizg's Voyages. An Account of the Lantis, M., 1970, The Aleut Social System, 1750 to 1810, Efforts of the Russians to Deterntine the Relation of from Early Historical Sources, in Ethnohistory in Asia and America. 2 Vols., American Geographical Soutltwestern Alaska and the Yukon: ,\lethod and Society, New York (reprinted by Octagon Books, Content, hl. Lantis (Ed.), University Press of Kentucky, New York, 1969). Lexington, Ky. Laughlin, IS. S., 1967, Human Migration and Permanent Petroff, I., 1884, Alaska: Its Population, Industries, and Occupation in the Bering Sea Area, in The Beri~~gSea Resources, in Tenth Census of tire United States, 1880, Land Brirtge, D. &I. Hopkins (Ed.), Stanford University Vol. VIII. Press, Stanford, Calif. Pierce, R. A,, 1968, New Light on Ivan Petroff, Historian of Lensink, C. J., 1962, The History and Status of Sea Otters Alaska.Pac. Northwest Quart.. 59(1): 1-10. in Alaska, Ph. D. Thesis, Purdue University, Xerox Piper, A. M., 1952, operations WIGDSTORM and JAN- University Microfilms, Ann Arbor, Mich. GLE. Proiect Ii8)a. Geologic.-. Hvdroloaic. , andThermal Makarova, R. V., 1975, Russians on the Pacific, Features bf the'sites, U. S. Geological &key, Portland. 1743-1799 (translated and edited from the 1968 Quimby, G. I., 1944, Aleutian Islanders. Eskimos of the Russian original by R. A. Pierce and A. S. Donnelly), North Pacific, Anthropology Leaflet Number 35, Chi- Limestone Press, Kingston, Ontario (Materials for the cago Natural History Museum. Study of Alaska History, No. 6). Sekora, P., 1973, Aleutian Islands National Wildlife Refuge, Martin, F., 1946, The Hunting of the Silver Fleece: Epic of IVilderness Study Report. From U. S. Fish and Wildlife the Fur Seal, Greenberg, New York, N. Y. Se~vicefiles, Anchorage. hlorison, S. E., 1968, History of United States A'aval Shiels, A. W., 1967, The Purchnse of Alaska, University of Operations in Tl'orld IVarII. Vol. VII. Aleutians, Gil- Alaska Press, Fairbanks, Alaska. berts and i\larshalls, June 1942-April 1944, Little, Sokolov, A,, 1851, Syevernaya ekspeditsiya, 1733-43 goda Brown and Company, Boston. (The Northern Expedition, 1733-43), Znp. Gydro- hlurie, 0. J., 1959, Fauna of the Aleutian Isla~~dsa12d ,p-aficheskago Departamento, 9: 190-469 (not seen by Alaska Pesinsuln, U. S. Department of the Interior, Fish author). and ISildlife Service, North American Fauna, No. 61. Test Command, Defense Atomic Support Agency, 1967, -, V. B. Scheffer, J. H. Steens, and H. D. Gray, 1937, Report of Operations, Vela Uniform: Project Long Report on Biological Investigations in Aleutian Islands, Shot, Report VUF-2702. Alaska, hlay 4-September 21, 1937. From U. S. Fish U. S. Government Alnnual, 1974-1975, Superintendent of and Wildlife Service files, Anchorage. Documents, U. S. Government Printing Office, \\'ash- Nelson, V. A. (Comp.), 1975, Bioenvironmentd Studies, ington, D. C. Amchitka Island, Alaska. 1975 Task Force Report, Veniaminov, I., 1840, Notes on the Unalnska District, ERDA ReportNVO-269-26, University of Seattle. Russian American Company, St. Petersburg (translation New York Times, Various News Reports. from the original Russian, Human Relations Area File, Okun, S., 1951, The Russian-Americon Conlpany, Haward New Haven). University Press, Cambridge, hlass. (translated from the Wickersham, J., 1938, Old Yukon-Tales-Trails-Trids, 1939 Russian original). West Publishing Company, St. Paul, Minn. APPENDIX A: THE EXECUTIVE ORDER THAT ESTABLISHED THE ALEUTIANS AS A RESERVATION AND NO\\' AS A NATIONAL \\'ILDLIFE REFUGE* It is hereby ordered that all islands of the Aleutian chain, Alaska, including Unimak and Sannak Islands on the east, and extending to and including Attu Island on the west, be and the same are hereby reserved and set apart as a preserve and breeding ground for native birds, for the propagation of reindeer and fur bearing animals, and for the encourage- ment and development of the fisheries. Jurisdiction over the wild birds and game and the propagation of reindeer and fur bearing animals is hereby placed with the Department of Agriculture, and jurisdiction over the fisheries, seals, sea otter, cetaceans and other aquatic species, is placed with the Department of Commerce and Labor. It is unlawful for any person to kill any otter, mink, marten, sable or fur seal, or other fur bearing animal within the limits of Alaska Territory, except under soch regulations as may be prescribed by the Secretary of Commerce and Labor; and it is unlawful for any person to kill any game animals or birds in Alaska or ship suc11 animals or birds out of Alaska except under the provisions of law and under st~chregulations as may be prescribed by the Secretary of Agriculture. Within the limits of this reservation it is unlawful for any person to hunt, trap, capture, wilfully disturb, or kill any bird of any kind whatever, or take the eggs of any such bird, except under such rules and regulations as may be prescribed by the Secretary of Agriculture. Warning is expressly given to all persons not to commit any of the acts herein enumerated and which are prohibited by law. The establishment of this reservation shall not interfere with the use of the islands for lighthouse, military, or naval purposes, or with the extension of the work of the Bureau of Education on Unalaska and Atka Islands. This reservation to be known as the Aleutian Islands Reservation. WM H TAFT THEWHITE HOUSE. March 3rd, 1913. "This Executive order has been retyped verbatim from the original copy. ALEUTIAN ISLANDS RESERVATION For Protection ofNafiveBirdsthe Fropagation of Reindeer and FmBearing Animals and theDevelopment of Fisheries ALASKA fmbracing the ls/and!offheA/eut/bn&ah assegregafed&broken he anddesgnated 2/eut/hn/s/ands Reservat;on." DEPARTMENT OF THE INTERIOR GENERAL LAND OFFICE Fred Dennett Commissioner APPENDIX B: THE WARTIME MEMORANDUM TO TROOPS ON AMCHITICA THAT CALLED ATTENTION TO ARCHAEOLOGICAL SITES ON THE ISLAND* RESTRlCTED HEADQUARTERS U. S. TROOPS APO 986, U. S. Army MEMORANDUM) 16 June 1943 No...... 35) ARCHEOLOGICAL OPPORTUNITIES, THIS STATION I. It is the desire of this Headquarters, both from the scientific and recreatiotial standpoints, to encourage the collection and preserva- tion of archeological specimens. 2. The Aleutian Islands contain the buried remains of ancient races, important in the peopling of the New \Vorld. The soldiers of this garrison have an unparalleled opportunity to collect and preserve these specimens and tlu~sto contribute materially to Science. The United States Govern- ment's Smithsonian institote is interested in tile remains found on this island. For a iluinber of years The S~nithsooiaoinstitote sent an expedi- tion to the Aleutian Islands. An officer of this Post, FLT Paul Guggeoheim, 444th Signal Battalion, was a member of an expedition in 1937. 3. In order to develop our o~porti~nitiesto the fullest, it is re- quested that individuals and organizations, wherever they find specinlens (bonesand tools. etc.) that are suspected to be of archeolorical interest. ~ICSL~IVC~t11c s11cci111~11sa11t1 110tiry 1.t. Gtt~e,c~~l~ei~ntltrouglt l)ispatcl~16-Rl. 'll~cqpec,i~net~ sI~c)t~ld II~ tt~rtletl o\'cr tu 1.t. GIIK<~IIIIC'~III311tla11 tile infurtna- tion tltnt C:~II bc givc~~co~~ccrl~i~i): tllc p1.1ce fuund, etc., so that inore \~>rcin~c~~~tnnv be lucatccl. Str:ci~~~c~~swill he sent to 'll~cS~nithqonikt~ 111- stittltc ;ttllt c~~.ditSvill bc gi~c11to I~I:lli~~uv~rcr. Ally persoll wllo now II;,~ ktto\\~lctl~!col tlte t>lncc \\fltcrc sixilolcns II~:IY 11c I'uu~l(l$llo~tld rivc 111c infoniiation to Lt. Guggenlieim. 4. It is desired that Orgaliizatio~iConimanders and all other officers explain the possibilities of this endeavor to their meti. With interest and entliosiasm on tlie part of tlie officers, much interest will be develop- ed. The recreational advantages are great. 5. Excavations for specimens tilust be made in a scientific manner so as not to disturb or break up specimens in the ground. 111 the near future, this Headquarters will publish inforoiation as to the scientific manner in which to work a site containing specimens, so that men under the sopervi- sioa of an officer may themselves engage in excavating work. Until that time, sites that are known to contain specimens sliould not be disturbed unless required for military reasons; but information as to location of tlie site should be eiven to Lt. Guaeenheim-- (Disvatcli 16-R1). This Head- quarters will auth&ize sites for excavation iG11icii will not be disturbed bv, uersonnel. other than those suecificallv authorized. Siens will be res- pected by all personnel. By command of Brigadier General JONES: &ARTHUR d!.fd~ A. LUNDIN. Major, F. A,, /jgl Adjutant. DISTRlBUTlON "A" RESTRICTED *This mcmarandum has been retyped \.erbatim from the original copy. Previous Scientqic R. G. Fuller* Battelle Colulnbus Laboratories, Investigations, ~~s vegas, ~evada 1867-1967 Durirrg the first century after the United Stoles ncqnired much was learned nbout the naturnl history of the western Alaska, progress ill the scietftific erplorntion of the western ~l~~~~i~~~a,,d about the yrin,itiue peoples lvho jrlhabited Aleutians zons erratic. The islai~ds were distant from the the irlnnds from prehistoric times. This body of knowledge Alnsknn mninlntfd, nnd the Alelctian c1i111atewas a deterrent to exploration. Despite these obstacles, marry irruestigators 1"s a useful base forplanniffgthe 6ioenuir0111nentalsttrdies were dmwn to the remote islands. Between 1867 and 1967 reported iffthis uoh~n~e. During the first 126 years following the discovery which began after Alaska was purchased by the of tbe Aleutians by Vitus Bering in 1741, some United States in 1867. I have chosen to treat the useful infor~nationabout the geography, human advances in chronological order rather than by inhabitants, flora and fauna, and general character scientific discipline. Many of the investigators who of the western Aleutianst was collected and have added to our kno~vledgeof the Aleutians were recorded by various explorers, by the Russian fur generalists whose interests ranged widely. It there- hunters, and by Russian administrators and mis- fore seems more logical to discuss their contribu- sionaries \\rho acco~npaniedthe hunters and traders tions in the order in which they were made,* during their exploitation of the natives and the instead of separating them among different areas of natural resources of the islands. But, however interest. The history of geographic exploration is valuable tltese observations may have been, most outside the scope of this chapter; for a good infor~nationwas collected only incidentally to the general survey that covers such exploration in the prin~arymissions of conquest, colnlnercial exploi- North Pacific Oceall-Bering Sea region up to the tation, and religious conversion that engaged the early 1960s, see Friis (1967). observers. The progress of scientific discovery in the Sherwvood (1967) reviews the more important western Aleutians after 1867 was erratic and for investigations of this period, for which he says he good reasons. Two themes are common to the ". . . used the standard sources in English and in accounts of most investigators: (1) the obstacles German, and in English and German tralwlations imposed by adverse weather and (2) the difficulties from the Russian." I am not surprised that the of getting logistic support and transportation Aleutians are scarcely mentioned in this review. among the remote islands. Most exploring and collecting expeditions bypassed Armstrong (Chap. 4, this volume) discusses the the western Aleutians, concentrating on the main- Aleutian climate, and the subject needs only brief land, the Alaskan peninsula, and the eastern mention here. Problelns that weather posed for islands. The Russian center of power for trade and field investigators on the foggy windssvept islands administration, after the first few years of free-for- are easily imagined: it was difficult, and sometimes all exploitation, was located first on Icodiak Island impossible, to preserve supplies and specimen and later at Sitka on Baranof Island. The scattered collections against the dampness; favorable condi- islands to the west must have been even nlore tions for observing wildlife were rare, and the forbidding and difficult of access to the early severe storms that often swept the islands some- explorers than they were to later investigators. times stopped field work entirely. The discussio~lthat follows xvill focus on those investigations of a more strictly scientific mature *This plan is adhered to except for the treatment of Russian research in oceanography and marine biology)..That *Present address: Oracle, Arizona. subject is discussed briefly in a section at the end of this tThe Andreanof, Delarof, Rat, and Near Island groups. chapter. 141 TI-ansport \\,as the other big obstacle. Attu enabled him to make extensive natural-history Island, at the western end of the chain, is over 900 collections and to pursue ethnological and archaeo- km (about 550 miles) from Atka. This section of logical investigations in the Aleutians. the Aleutians comprises some 40 islands and islets, Sherwood (1965) says of Dall: "The collec- nlauy of which lack safe harborage for support tions he made in his leisure time \\?ereextensive. He vessels. The usual base from which to explore the sent seven boxes of dry shells, echinoderms, and chain was Unalaska, some 500 km (350 miles) east fossils to Professor [Louis] Agassiz* at the Mu- of Atka, and until recently no regularly scheduled seum of Comparative Zoology [Harvard Univer- transport Ivas a~ciilablealnong the islands.* Early sity]. Plant specimens were mailed direct to the investigators were forced to provide their own enlinent American botanist, Asa Gray. In one year vessels or to depend on the support of trading alone, Dall forlvarded t~venty-sevenboxes and kegs ships, go\~enlmentvessels, or any other ships of of general zoological and ethnological nlaterials to opportunity that happened to sail through the the Smithsonian." Hultin (1960) credits Dall with area. making ". . . fairly good collections. . ." of plants, Comme~.cialand government vessels had their Ivhich were deposited in the National Herbarium assigned mirsions, which understandably took pre- and in Gray Herbarium at Harvard. cedence over rendering assistance to visiting scien- Da11 (1877a) reported the results of his archae- tists. Few exploring parties were fortunate enough ological and ethnological studies and the conclu- to work from their own support ships; so a sions he drew from them in the first volume of catch-as-catch-can ulode of travel was normal. Cont~ib~tioi~sto ~Vortlr Awrericntt Etlt?rolog)~One Inevitably, investigators experienced frustratiug de- of his three papcrs in the volume is "On Succession lays, were often unable to visit important target in the Shell-I-leaps of the Aleutian Islands." In the islands, or had their visits cut short by the early introduction to this paper, he explains the circum- departure of a support ship. So it is not surprising stances under which he was able to investigate the that scientific studies in the western t\leutians archaeological remains of the islands: "The notes seem to have proceeded unevenly with long periods of which this paper is the result were made rvhile when advances \sere slight. The in\restigators engaged in a hydrographic and geographical recon- rvorked against formidable odds, and \\'hat is naissance of the Aleutian Islands, under the aus- remarkable is that they succeeded as well as they pices of the United States Coast Survey. They were did against those odds. made at enforced intel-vals of leisure, occasioned by iveather which would not permit the ordinary surveyiug operations of the party to be carried DALL on, . . . ." Considering the limited opportunities Dall had to pursue this effort, it is note~vorthythat The American naturalist William Healey Dall be was able to collect as much iufor~natiouas he had a distinguished scientific career in Alaska, a did. re'sume' of which is presented by Sher\vood (1965). Dall's report on the shell heaps (mounds of His work in the far nortb\sest began in 1865 when, debris that constitute the principal repositories of at the age of 19, he joined the Scieutific Corps of material remains of the prehistoric inhabitants of the Russian i\merican Telegraph Expedition. I have the western Aleutians) is prefaced by a brief found no evidence that this assignment took him description of the topography of the islands and into the western Aleutians,t but it did establish his the flora and fauna of the islands and adjacent seas. reputation as a reliable and productive field natu- He lists the islands, from Unalaska westward, on ralist and awakened in him a lifelong iuterest in which shell mouuds \Irere observed or reported to Alaska. When he joined the staff of the Coast exist (a total of eight). Of these, Amchitka Island is Survey in 1870, he embarked on a course that discussed at greatest length, and he stresses the abuudance of shell heaps on that island, ivhicll he took to be cvidence of a large resident human "Even today commercial air transport is available on a population in earlier times. scheduled basis to only a few of the western islands (Adak, Dall describes in some detail the methods he Shemya, and Attu) whereas ship transport is available from Adak to Atka. used in excavating the heaps, methods that were tIn 1865 DaII crossed the Bering Sea on the expedition remarkably sound as judged by modern standards. vessel r\'ighfhrgnle rvitll stops at St. Lawrence Island and at Turner (1970) nlentioned the fact that Dall in his tr\.o ports on the Siberian coast. The captain of the ". . . excavations at Coustantine Harbor employed ~\'ighfhtg.nle was Charles hl. Scammon, who a ferv years later the principle of .tratigmphy-preceding by 50 achieved some scientific recognition \\,it11 publication of The dlnrine rllon~ntnlsof il3e h'orth-IVesfer~tConst of~\'ortlr Atnericn (Scammon, 1874). *Dall had studied earlier with Professor Agassiz. Preuio~isScientific I?iuestigntions, 1867-1 96 7 143 years its use by the archaeologist N. C. Nelson."* In discussing articles of persolral ornamentation Dall reported that excavations were carried out on found in the shell heaps, Dall notes that the early Attu, Amchitka, Adak, and Atka. As an example peoples made beads of amber, among other nlate- of the ground plan of a typical shell heap, he rials. He adds, "This substance occurs sparingly in inclitdcs a sketch of one examined at the head of the lignitic deposits of Tanaga, Unalashka (sic), Constantine Harbor, Amchitka Island (probably Atka, and Amchitka, and was reckoned of the RAT3 as reported by Desautels et al., 1970). greatest value by the Aleuts." His reference to The conclusions Dall reached as a consequence "lignitic deposits" calls to mind the later discovery of his archaeological investigations in the Aleutians of partially carbonized tree remains on Amchitka will be discussed only briefly here. \Vhat is notable (Powers el al., 1960; Humphrey and Branstetter, is that some of his judgments, though by no means 1973; Barr, Ellis, and Helle, 1973). all of them, have been supported by the much Da11 (1877b) attempted to determine the rela- more extensive subsequent studies. Among his tionship between the Aleuts and the aborigines of generalizations, probably the most important are: mainland Alaska, with whom he had been in I. That the islands were populated at a very contact during his early Alaskan field work. He distant period. recognized the affinities between Aleuts and 11. That the population entered the chain fronl the Eskimos, on the bases of language, geographic eastward. 111. That they were, when they first scttled on the distribution, and a common dependence on mari- islands, in a very different condition froin that in time resources for subsistence. He listed 20 divi- which they were found by the first civilized trav- sions of Eskimos or Innuit, drawing 011 his personal elers . . . . observations of those groups occupying the coast VI. That a gradual progression from the lorv Innuit of western Alaska and on reports by others on stage to the present Aleut condition,t xvithout serious intei~uption,is plainly indicated by the materials of, more-distant divisions. The Aleuts were treated in and utensils in, the shell-heaps of the islands . . . . less detail since he had less information to draw on. VIII. That the stratification of the shell-heaps sllorv He recognized two principal geographic divisions: a tolerably uniform division into three stages, charac- the Eastern or Unalashkans (sic) and the Western terized by the food which formed their staple of or Atkans. He estimated the total Aleut population subsistence and by the weapons for obtaining, and utensils for preparing this food, as found in the in 1871 to be about 2450 persons, of whom less separate strata; these stages being than 500 lived in the western islands. I.The Littoral Period, represented by the A review in detail of Dall's contributions to Ecllinus Layer. knowledge of the geography and natural history of II. Tile Fishing Period, represented by the Fish- the western Aleutians is beyond the scope of this bone Layer. 111. The Hunting Period, represented by the chapter; they appear to have been substantial by Alammalian Layer. any standard. Bartsch, Rehder, and Shields (1947) give an account of Dall's entire career and a Finally, Dall concludes, "On the wllole, I am bibliography of his volu~ninouspublications. These inclined to think that three thousand years is a authors, staff members of the U. S. National moderate estimate for the time required to form iVlaseum, write: these mounds of ~efuse." At the close of the 1874 season Dall was recalled Later workers have challenged Dall's neat to the Il'ashington office [of the Coast Survey, later division of Aleutian prehistory into the "Littoral," the Coast and Geodetic Survey] to assemble all the "Fishing," and "Hunting" periods as well as some known data of use to navigation in Alaskan waters. The information thus assembled was published in the of his interpretations of cultural evolution among 'Pacific Coast Pilot of Alaska,' the foundation upon the prehistoric peoples of the chain (Jochelson, r\,hich all subsequent Alaskan studies by the Coast and 1925; HrdliEka, 1945; Laughlin, 1951; Spaulding, Geodetic Survey were based. . . . .During his period of 1953). On balance Dall appears to have reached office at the United States Coast and Geodetic Survey some so~undconclusions regarding the prehistory of his spare hours were devoted to the arranging of the the Aleuts, a notable achievement \\,hen we con- immense collections that he had made in Alaskan waters as well as the material that had accl~mi~latedat sider-the l~ardshipsand limitations under u~liichhe the Smithsonian Institution. His journals reveal that no worked. opportunity \\*as missed to collect on land or with dredge in the sea during his work in the north. Our knowledge of tllc fauna of that region is still largely *Ceram (1971) reports that as early as 1781 Thomas based on Dall's collecting.* Jefferson described the use of this valuable technique in his excavation of prehistoric earthern bilrial inoonds in Vir- *Although this last assertion (made less than 30 years ginia. ago!) may be something of an overstatement today, it tDall here no doubt refers to the conditions of the suggests the esteem in rvhich Dall's work was held by his Aleuts at the time of first contact with so-called civilized peers three-quarters of a century after he began his Alaskan matt. field studies. Dall's primary fields of inquiry were mala- he, nevertheless, managed to make obsen~ations cology and paleontology, but he was a true and substa~ltialcollections of Aleutian birds (his naturalist, and his observatio~lsand collections special area of interest), fishes, and mammals and ranged far beyond his own areas of special interest to record some interesting eth~lologicalnotes on into every aspect of natural history. As a result, the Aleuts of the western islands. Turner never specialists in many diverse fields benefited from maintained a station on Amchitka, but he appar- the collections he made. For example, much of our ently managed to spend the latter part of May and early kno\vledge of the marine fauna of the the early part of June 1881 on that island while be western Aleutians rested on the collections made was en route from Attu to Unalaska via schooner.* there by Dall during his Coast Survey cruises. While there he added considerably to his observa- tions of the nati~ralhistory of the islands. After Turner returned to the States, he worked up his BEAN Alaskan material for publication. Turner (1886) devoted most of his report to Tarleton H. Bean of the U. S. National Museum the birds and fishes of Alaska, often supplementing participated in at least a portion of the 1880 the results of his own investigations with those of summer cruise of the U. S. Coast Survey vessel others in such a way that it is not always possible Yt~koncomma~lded by [\I. H. Dall. The following to identify the source of the information. Among year he published what appears to be the first the fishes he reports as occurring in the western attempt to enumerate the fishes of Alaska and Aleutian region are 11 species not listed from that adjacent marine waters (Bean, 1881). Bean's list area by Bean (1881). contained 116 species, of which only 16 were Turner appears to have been the first investiga- collected from the western Aleutians; it repre- tor to publish an annotated list of the birds of sented all the Alaska collections then held by the Alaska. His listing includes over 50 species from U. S. National Museum plus a few species that had the western Aleutian Islands. Over 20 of these been reported from Alaska waters by other investi- species are specifically noted as having been ob- gators.* In an Appendix are listed an additional 99 served on Amchitka. Turner also commented on species of fish that ". . . are known to occur in the adverse effects of native foxes on the breeding waters bordering upon the limits of Alaska, and populations of Aleutian Canada geese on those will doubtless be found by future investigations." islands \vhere foxes occur naturally. His observa- A review of the Alaskan species enumerated by tions thus appear to have anticipated the disappear- Bean shows that nearly all those reported from the ance of breeding populations of geese from islands western islands were collected by Dall. Amchitka to which foxes \irere introduced in the 1920s. was among the collection locations cited for 10 of Turner tried to make plant collections during the 16 species from the western Aleutians. his stay in the Aleutians, but his attempts were largely thwarted by the weather. He reports that, TURNER ". . . of all great difficulties the most troublesome was to preserve the plants after I had collected Lucien McShan Turner, a naturalist, served in them. The constant moisture of the climate has Alaska as a meteorological observer with the U. S. frequently ruined my entire collection of a sum- Army Signal Corps from 1874 to 1881. In May mer's work." Hultdn (1960) notes that Turner left 1874 he arrived at St. Michael, on Norton Sound, "medium sized collections" in the National Herbar- where he was stationed until July 1877. After his ium and Gray Herbarium. However, Hult6n is return to the States and a brief separation from the critical of Turner's contribution to the kno\vledge Signal Service, he returned to Alaska in A4ay 1878 of Aleutian flora. He points out tltat, since Turner and began his tour of duty in the Aleutians, an was not a botanist, his observations regarding tlle assignment that took him to the western islands of distribution of plants in the Aleutians are not the chain. Turner established and operated a wholly reliable. meteorological station on Atka from May through August 1879 and one on Attu from July 1880 EVERMANN AND GOLDSBOROUGH through mid-May 1881. His official duties as a weather observer occupied much of his time, but In 1903 a special comnlission appointed by the U. S. Comnlissio~ler of Fisheries carried out an *Bean says of his listing: "The species named are almost wholly shore species, or such as arc found in comparatively shatlorv rvater; the deep-water fishes of Alaska are still *This can be inferred from references to Amchitka undiscovered.. . ." obsetvations mentioned in his published report. Previozls Scientific I~zuestigntio?zs,1867-1 967 145 in\restigation of the salmon fisheries of Alaska and, GILBERT AND BURKE in the course of the study, added considerably to kno~vledgeof the fish fauna of the North Pacific The U. S. Fisheries vessel Albatross carried out Ocean and the Bering Sea. Evermann and Golds- exploratory fisheries investigations in the north- borough (1907) incorporated this information into western Pacific in the summer of 1906. Gilbert and a neur catalog of "The Fishes of Alaska."* The Burke (1912) reported on the results of this cruise. listing includes more than twice as many Alaskan They note: species as Bean (1881) reported, but the authors On the outward voyage the vessel passed the rely to a considerable extent on the reports of Aleutian chain, touching at Unalaska, Atka, Agattu, Bean (1881) and Turner (1886) for data on the and Attu Islands. . . . Shore collecting was carried on western Aleutian fauna. The catalog includes at in these localities, and some 37 l~aulsof intermediate least seven species from that area that are not in net or dredge were made along the route, several of the earlier listings. These are attributed mainly to these hauls being highly successful. Rich ground, rvhich collectio~tsinade during the cruises of the U. S. would repay thorough investigation, was found on Petrel Bank (north and east of Semisopochnoi Island), Fisheries vessel Albatross in the 1890s and during [and] in the vicinity of Attu and Agattu Islands . . . . the salmon investigations of 1903. These authors concluded that the Albatross reconnaissance produced no evidence of a sharply defined marine faunal break anywhere along the Aleutian chain or, in fact, between the Aleutian A. H. Clark of the U. S. National Museum and Commander Islands. They did note differences participated in a long northern cruise of the U. S. between the marine fauna off ICamchatka vs. that Fisheries steamer Albatross in 1906. The cruise around the Commander Islands, which they at- started from San Francisco, Calif., in April, pro- tribute to the barrier effect of the deep channel ceeded north along the coast to the Alaskan between the two areas. peninsula, west through the Aleutian Islands, to Of the 123 species listed from the 1906 the Okhotsk Sea, to Japan, to the Hawaiian Albatross collections, 34 are described as new Islands, and back to Sau Francisco in December. species (8 new genera are also recognized). Sixty of The Albatross passed close to Amchitka without the species listed are noted as occurring in the stopping, but brief stops were made at Atka, western Aleutians if the Petrel Bank collections are Agattu, and Attu [see Gilbert and Burke (later in included. Among these are 17 of the species this chapter) for a reference to the fisheries results described as new. of the cruise]. Clark (1907) published an account of 18 new species and 1 new genus of birds from eastern Asia JOCHELSON and the Aleutian Islands which incl~tdeda new subspecies of Lagopzrs rz~prestris (Rock Ptarmi- Russian scientific interest in the Aleutians did gan)? from Adak Island, presumably collected not terminate with the sale of Alaska to the United during the 1906 cruise of the Albatross, although States in 1867. Russian scientists continued to this is not stated. Clark (1911) gives a complete inquire into possible geological, floristic, and fau- listing of the birds collected or seen during the nistic relationships among the western Aleutians, cruise. Of the 172 species and subspecies listed, 43 the Commander Islands, the adjacent Siberian are noted as occurring in the western Aleutians mainland (especially the ICamchatka peninsula) and (Atka, Agattu, Attu, Semichi, and ICiska are the Kurile Islands. Thc origin and the prehistoric specifically mentioned). R4ost of them are birds antecedents of the Aleuts also constituted a prob- associated with the pelagic and seashore habitats, lem of great interest to the Russians, under- and most are to be found among the species now standably so because they were the first Europeans known from Amchitka Island (see White, William- to meet and observe the Aleutian aborigines. son, and Emison, Chap. 11, this volu~ue). The Kamchatka Expedition (sometimes referred to as the Aleut-Kamchatka Expedition) was initiated in 1907 by a Russian banker, F. P. Riaboushinsky, and was later carried out under the auspices of the Imperial Russian Geographical *The title is not entirely accorate; the 288 species listed Society. The expedition inclitded 20 scientific staff include 22 species not found north of Dixon's Entrance. Technically these would not be considered Alaskan species. members and was organized under several divisions tlagopus mutrcs of current nomenclature (h,furie, representing the disciplines of anthropology, 1959). botany, zoology, geology, and meteorology. The anthropology* division, headcd by \Valdemar brought his manuscript to America, and, after Jochelson, contributed most to our k~lo\vIedgeof making further comparative studies of Eskimo- the western Aleutians. Aleut collections in this country, he rewrote it in Jochelsolt and his !\rife, a medical doctor who English. The history, ethnology, and physical assisted him in the field work, reached the anthropology of the Aleuts are discussed in Jochel- Aleutians by \tray of New York, Washington, D. C., son (1933). San Francisco, and Seattle. En route, official In 1us 1925 report, Jochelsoil discusses at some permits \\,ere obtained, supplies were purchased length the collclusions reached earlier by Dall, who and packed for sl~ipment, and Aleutian Island had excavated on some of the same wester11 islands collections in the United States were studied. as Jochelso~l (Attu aud Atka). In particular, he Transportation for Jocl~elson's party from Seattle takes exception to Dall's division of Aleut culture to Unalaska, to various other Alelttiall Islands, to history into three periods (Littoral, Fishing, and the Pribilofs, and, finally to ICamchatka was Hunting) and to Dall's contention that the first provided by the Alaska Conl~nercial Company, peoples reaching the islands possessed only a various U. S. Go\?ernment agencies, and a Russian rudimentary culture, lacking fire, stone imple- ~nilitarytransport. Jochelson and his party ap- ments, and many other traits kno\vn to the Aleuts parently had the wholehearted cooperation of at first contact with Europeans. almost everyone they encountered, but the un- On the other hand, Jochelson agrees fully with certainties of travel arrangements and the exi- Dall on the important point that the Aleutians gencies of Aleutian weather hanlpered their field were peopled from the east, i.e., from the conti- studies and extended the time they spent in the nent of North America. And, although he is islallds \\,ell beyond \\.hat they 1~ad originally cautious in attempting to fix a date for the first scheduled. habitation of the islands, his judgment is not Jochelson (1925) writes, ". . . according to the greatly different from that of Dall. Jochelson estimate of our original plans, only one year was to concludes: "All the foregoing considerations fail to be devoted to the Aleut problem. Actually, the give us a definite measure of time which could be field-work in the Aleutians stretched over a period expressed in figures; nevertheless they show clearly of 19 months,t while the journey from Petrograd that the Aleut came to the islands many centuries to Unalaska and the preparatory work occupied 3 ago . . . ." His". . . many centuries. . ." can easily months more." He also reports: "In all, we accommodate the 3000 years cited by Dall as a excavated at 13 ancient village sites, investigated reasonable estimate for the time required to form 57 pits of various size, 3 burial caves, and 3 other the shell heaps he excavated. caves . . . . Of the 47 days devoted to excavation, only 8 were dry and calm; the remainder of the time \xre xvorked in rainy and stor~nyx\.eather . . . ." BENT ET AL. Jocl~elsonapparently made no excavatiolls on Amchitka, but he did work on Atka and on Attu, A. C. Bent, R. H. Beck, F. B. A,lcI HUTCHISON A British botanist, Isobel CVylie Hutchison, *The hi. V. Brolutt Bear was later used to conduct the made a brief plant collecting tour through the oceanographic surveys of the Chukchi Sea in 1959 and Aleutians in the summer of 1936 while traveling 1960 in connection with the AEC-supported study of the Cape Thompson Region (\\'ilimovsky and \\'olfe, 1966). aboard the U. S. Coast Guard cutter C/ze/nn. In the tone marine vascular plant, eel grass (Zostern sp.], is western islands she collected at Atka, Amchitka, mentioned as occurring near Umnak Island ". . . but \*,as Iciska, and Attu. Her collections, which included not seen farther west." (Scheffer in BIurie, 1959). Previous Scientgic I~~vestigations,1867-1 967 149 National \\'ildlifc Refuge).* As Miirie writes in the 1959 report: "In 1920, the Uuited States Bureau Victor B. Scheffer, scientific leader of the 1938 of Biological Survey was given the responsibility cruise of the Brozo~z Bear, contributed a section for enforcetnent of the Alaska fur laws and entitled "Invertebrates and Fishes Collected in the ad~ninistrationof the bltte-fox indzlstry (emphasis Aleutians, 1936-1938" to the report by kIurie. added) in the Aleutians. As time went on, it Some inferences concerning the western islands became apparent that proper supervision of this visited during 1938 can be drawn from collection important wildlife refuge would necessitate an sites aud dates given for fish specimens mentioned extensive inventory of the resources of these in his contribution. Scheffer (1959) cites examples islands." It Isas the purpose of Murie and his of marine and anadro~nousfish collected at Attu colleagues to provide such an inventory. (August lG), ICiska (August l9), and Amchitka I have not found a narrative account of the (August 22). Apparently the expedition was mak- Brozon Bear expeditions which would indicate just ing its way eastward from the end of the chain in how much of the effort \\'as devoted to the western mid-August, and many other western islands were Aleutians, but there is evidence that the western- probably visited duriug this final cruise. most islands were visited during each of the 3 years The Brown Bear survey of the Aleutians involved and that Amchitka Island received a accomplished a great deal, although it is unfortu- considerable amount of attention. nate that so long a delay occurred before publica- tion of the official report of the expeditions 1936 (klurie, 1959). 111 keeping with the stated purposes An unpublished "Report on Amchitka Island," of the survey, field studies concentrated on birds provided from the files of the U. S. Fish and and mammals. Voluminous, although unpublished, Wildlife Service, Anchorage, Alaska, dated Aug. 11, internal U. S. Fish and Wildlife Service (FWS) 1936, and written by H. \V. Jewel1 and H. D. Gray, reports were made pro~nptlyto guide decisions for notes: "The dates of our visits to Amchitka Island wildlife management of the many diverse islands of are July 24 and August 1, 1936." A revie\v of the Refuge; these appear to have served their Tvfurie (1959) reveals, from his comments on purpose xvell. various species of avifauna, that the 1936 cruise Other aspects of the island biota were not included, in addition to Amchitka, at least the \vholly neglected. Extensive, although by no means following other western islands: Adak, Kanaga, exhaustive, collections were made of marine and Iciska, aud Attu. It seems reasonable to assume terrestrial plants; marine, freshwater, and terrestrial that some additional islands among the many from invertebrates; and marine and freshwater fishcs. The Atka westward were visited in 1936. collections were made available to appropriate specialists in the scientific comm~unityand formed the basis for a number of separate publications in technical journals (see Bartsch and Rehder, 1939; Another unpublished report provided from the Clark, 1939; Murie, 1940a, 1940b, 1945; Schultz, files of the U. S. Fish and \\'ildlife Service, Anchor- 1939). age, Alaska, gives the itinerary of the Brotou Bear The following section will show that the Brozon party for the 1937 season. In brief, the ship Bear party of 1936 also made a contribution to provisioned at Unalaska in .late May and then the archaeological exploration of the Aleutians, proceeded to Attu with stops at Umnak, Atka, and although this endeavor was somewhat tangential to ICiska. Attu was reached on June 7, and from there the main purposes of the survey. the party worked eastward, stopping at some islands not visited in 1936. From July 4 to July 18, klurie was left on Amchitka and Steenis on ICanaga \vhile the Brozu~~Bear returned to Unalaska for supplies with Scheffer and Gray remaining aboard While the U. S. Bureau of Biological Survey to continue their scientific duties. The cruise was investigating the flora and fauna of the continued after hrlurie and Steenis were picked up, Aleutians, a small party from the Smithsonian the Aleutian work ending in early September at Institution, led by the physical anthropologist Ales Unimak. From there the party departed for HrdliEka, was digging into the question of pre- Seattle. historic human occupation of the islands. HrdliFka's work in the Aleutians, which reached *See Chaps. 6 and 20 of this volume for a history of the westward to include the Comn~anderIsland group, administrative permutations experienced by the Reserva- was a natural extension of his earlier studies on the tion or Refuge and changing concepts of management. Alaskan nlaiuland and on Kodiak Island. Those investigations were, in turn, stimulated by Itis The Smithsonian Aleutian expeditions, like belief (HrdliEka, 1930) that "Alaska and the those of most earlier investigators, depended for opposite parts of Asia hold, in all probability, the logistic support on the help of the U. S. Govcrn- key to the problenl of the peopling of America." lnent agencies (Coast Guard, Bureau of Biological I-IrdliEka's most extensive field study of the Survey, and Coast and Geodetic Survey) and on arcl~aeologyof mainland Alaska and a concurrent the Alaska Con~mercialCompany. Doubtless this ethnological investigation of Alaskan Indians and kind of dependence determined to some extent Eskimos were carried out in the summer of 1926. xvhere the expeditions were able to work. But a Ilis account of this rvork and the conclusions he review of the daily logs for the three seasons reached are given in Hrdli'cka (1930). In that report indicates that the western islands, including Am- he reached no definitive judgment regarding the chitka, \\?ere the scene of a considerable amount of role of the Aleutian Islands in the "peopling of the work. America" or on the relationship of the Aleuts to other northwestern aborigines. \\'hat is somc~vhat surprising is that he suggests (in Fig. 29 of the 1930 report) the I Table I-Parts of USGS Bulletin 1028 Devoted to \Vestern Aleutians Part Subject (Bolletin 1028) Author and date Great Sitkin Island Simons and h~lathewson (1955) Northern Adak Coats (1956a) Island Kanaga Island Coats (1956b) Some western islands Coats (1956~) Ocean floor, northeast Snyder (1957) Rat Islands Little Sitkin Island Snyder (1959) Delarof and West Fraser and Barnett Andreanof Islands (1959) Gareloi Island Coats (1959a) Segula, Davidof, and Khvostof Islands Adak and Kagalaska Fraser and Snyder Islands (1959) Semisopochnoi Island cdats (i959b) Amchitka Island Po\vers, Coats, n and Nelson (1960) Rat Island Lewis, Nelson, and Powers (1960) Kiska Island Coats, Nelson, Lewis, and Powers (1961) Near Islands Gates, Powers, and \Vilcox (1971) emphasis on skeletal remains recovered by HrdliEka field work and was the author of most of the and Jochelson. New data are presented on anthro- publications resulting from the project. pometry and distribution of blood groups among In a popular account of the first summer present-day Aleut populations. These, together season in the islands, Bank (1956) lists the objec- with linguistic and cultural features, are used as the tives of the expedition: (1) to "make a complete basis for elucidating relationships between the collection of plants"; (2) ". . . to explore and map Aleuts inhabiting the eastern and western parts of the least known of the islands"; (3) ". . . to visit the chain and between Aleuts and other Eskimo the few scattered villages where Aleuts were still peoples. living and to record their fast-disappearing native In this paper Laughlin retains the Pre-Aleut and customs, especially those pertaining to old uses of Aleut terminology recognized by HrdliEka. In later plants for foods, medicines, and poisons"; and (4) papers he and other anthropologists have substi- ". . . to compile a manual of useful plants and tuted the terms Paleo- and Neo-Aleut to distinguish animals in the Aleutians." The field party during the two somewhat different aboriginal stocks that the first season was a modest one: Bank, then a peopled the Aleutians (Spaulding, 1953; Laughlin, graduate student, and one assistant. They got as far 1963). west as Atka, traveling by military ships and aircraft, made plant collections, explored some UNIVERSITY OF MICHIGAN archaeological sites, and established working rela- EXPEDITIONS tions with Aleut informants. I have found no comprehensive description of Ethnobotru~ical and archaeological studies in the studies carried out by the University of the Aleutian Islands, under the auspices of the Michigan expeditions and no evidence that any of University of h'lichigan, were started in 1948. the four stated objectives was fully realized. These field investigations were continued for sever- However, in a paper subinitted for publication in al years thereafter with support from the Office of 1952, Bank (1953a) reported: "In all, four trips Naval Research and the Michigan A,lemorial Phoe- have been made to the Aleutians, during which nix Project. T. P. Bank 11, of the University of more than 20 islands were visited." In a footnote hlichigan, Department of Botany, initiated the to this paper, Bank is identified as "Field Director, University of Michigan Expeditions to the Aleutian islands, under the protection initiated in 1911 Islands, 1948-52," fronl which it can be inferred (Abegglen, Chap. 20, this volume). There was an that the investigations tended at least over the acute need for inore information on sea otter 5-year period indicated. distribution and biology to guide nlanagement of Bank (1952) gives a generalized view of the the resource. It was also recognized that the plant communities associatcd with the topographic Aleutians are an important breeding ground for features of the ttvo main island types represented waterfowl, although fox farming, established ear- in the Aleutians: "n~ountainous" and "plateau" lier on many of the islands, was hardly compatible islands. He also (Bank 1952, 1953a) notes the wit11 successfi~luse of the habitat by nesting birds. importance of drainage as a determinant of plant- Amchitka Island was a logical choice as a field comlnunity type and calls attention to the inherent laboratory for sea otter and avifaunal studies. It instability of plant communities in the islands: had a large sea otter population and was poten- "The ecological concept of succession is hard to tially a good breeding habitat for waterfowl. The apply to communities ~vhichcontinue to fluctuate island had additional advantages as a site for field in an unstable way so that a so-called climax work: an airstrip, a harbor, and a system of grouping today may be sho\vn later to have been roadways developed by the military. only temporary." Ethnobotanical findings of the Betmeen 1949 and 1952, R. D. Jones, Refuge expeditions are given in Bank (1953b, 1953c), h,Ianager, collected carcasses of a number of dead which discuss the uses made of plants by the or lnoribund sea otters on Amchitka beaches and Aleuts and the vegetational cover associatcd with submitted them to patl~ologistsat the Arctic Public ancient Aleut village sites. Health Center (then in Anchorage, Alaska) for Spaulding (1962) reported the results of the examination. Rausch and Locker (1951) and principal arcl~aeological investigation carried out Rausch (1953) reported the results of their parasi- by the Michigan group in 1949, i.e., excavations in tological studies of these animals. Several internal two prehistoric dwelling sites on Agattu Island* parasites, at least two of which are pathogenic, near the western end of the chain. An important were isolated; some parasites were found in most outcome of this work was the recovery of xvood carcasses examined. Ho!vever, the investigators and charcoal samples suitable for radiocarbon could not ilnplicate the parasites as a major cause dating from the refuse mounds. Charcoal from an of sea otter mortality at Amcllitka. Rausch (1953) intermediate depth in one of the excavations was hinted tl~ata too high sea otter population density assigned an age of 1300 f 150 years. The average may have been a contributing factor and recom- age of t\vo \\rood sanlples from near the base of the mended efforts to redistribute (transplant) sea same mound was estimated as 2565 + 300 years. otters from Amchitka to otller suitable habitats. These findings suggest that the prehistoric peoples Early transplant attempts were unsuccessful; reached thc western islands at an early time, captive ani~nalsdied before they could be removed supporting Dall's estimate made three-quarters of a fro111 the island from causes that were not readily century earlier. apparent. In 1954 a study of the physiological reactions of captive otters was carried out at Amchitka by investigators from Pttrdue University UNITED STATES FISH AND assisted by FIVS personnel and supported by FIVS \I7ILDLIFE SERVICE (FWS) funding. The results of this study were reported by ICirkpatrick, Stullken, and Jones (1955) and Stull- As IVorld \\far I1 ended and military control of ken and ICirkpatrick (1955). The conclusion was the Aleutians \\,as relaxed, the FIVS undertook that captive sea otters generally died from the field investigations in some of the islands com- combined effects of inadequate (too infrequent) prising the Aleutian Islands National \Vildlife Ref- feeding and environmental stress. Some progress uge. T\vo subjects, sea otters and birds, received was made in maintaining animals in captivity for special attention, with principal stress being placed long periods, but the problenls involved in trans- on the former. This emphasis was a natural one. planting \Irere not entirely solved. On the basis of a nuntber of observations, it was kno~vnthat the once-decimated Aleutian sea One F\VS staff lnenlber who collaborated with otter population had grown to sizeable propor- the Purdue investigators was Calvin J. Lensink, tions, especially around some of the western ~vhowent on to conduct an intensive study of the status and distribution of the sea otter in Alaska. The findi~~gsof this investigation, reported by *Ale: HrdliZka excaeated sites on Agattu in 1937 but at Lensink (1960, 1962), established the fact that a location several miles from the Krugloi Point sites where northerll sea otter stocks, nearly wiped out before the hlichigan party worked. 191 1, had increased substantially after protection Previous Scientific Irtvestigcctio?ts, 1867-1 9G7 155 was imposed. Lensink also found that a large research in tlle western Aleutian region by Amcri- proportion of the Alaska sea otter population was can investigators. Some infor~nationwas accumu- concentrated around some of the western Aleu- lating on the oceanograpllic features and com- tians; Amchitka was estimated to have the largest mercial fisheries poteutial of the area, but there sea otter population of any island in the chain. was no attempt at organized exploration until the In 1955, Karl \\I. Itenyon, F\\'S \\'ildlife Biolo- second half of the twentieth century. The state of gist, began a study of sea otter biology that knowledge of oceanography of the North Pacific continued, in the field and in the laboratory, for and Bering Sea and the sources of inform at' 1011 over a decade. A~Iuchof his rcseafch was done in were reviewed by Fleming (1955) and Dodimead, the western Aleutians and especially at Amchitka. Favorite, and I-Iirano (1963). These reviews were The work culminated in the publication of a published as Bulletins 2 and 13, respectively, of the non no graph covering essentially every aspect of the International North Pacific Fisheries Commission biology and natural h~story of the sea otter (INPFC). (Kenyon, 1969). It was largely through Itenyon's The INPFC was established in 1953 by an observations and experiments that the problem of International Convention entered into by Canada, captive mortality was solved, making it feasible to Japan, and the United States. Since the inception successfully transport animals to suitable habitats of the Commission, it has been the chief coordi- hundreds of miles from the point of capture. Sea nating agency and the principal source of support otters have since been reintrotluced into areas from for oceanographic and fisheries research in the seas which they had earlier been eliminated (Abegglen, around the Aleutian chain. The main focus of Chap. 20, this volume). INPFC research is on conservation of comlnercial Other facets of natural history were not ne- fisheries resources (principally Pacific salmon) and glected during the F\VS sea otter investigations. on allocation of rights to these resources among Itrog (1953) reported his observations on birds of the participating nations. However, much basic Amchitka made during his visit to the island in information on the marine biota and the oceanog- 1952 in connection with sea otter parasitology raphy of the region derives from INPFC activities. studies. He i~lcludesa section on food habits of the The scope of INPFC research in the oceans Bald Eagle, noting evidence of eagle predation on adjacent to the \vestern Aleutians between 1955 young otters. Kenyon (1961) published an anno- and 1960 is sho~sn in Alanzer et al. (1965, tated list of the birds of Amchitka Island, a list Figs. A-1 to A-6). These figures plot the location of totaling 69 species. He gives a brief history of the stations occupied by INPFC research vessels during FWS progain to exterminate foxes and rats and each year of the 6-year period; Japanese or speculates on the probable adverse effects of these American vessels, or both, occupied stations in the predators on avian populations. western Aleutian Islands during at least part of The report by Kenyon and Rice (1961) on the every year of the period covered. Many stations abundance and distribution of the Steller sea lion is were well offshore, reflecting the objectives of the based mainly on surveys made in 1959 and 1960. cruises, but a number were located near the chain, A largr proportion of the total estimated popula- particularly in the vicinity of major passes between tion was in the western Aleutians; a total of 1750 the North Pacific Ocean and the Bering Sea. animals n8as estimated for Amchitka. Kenyon Resolts of INPFC research are presented in annual (1965) also reported on the food habits of harbor reports, a bulletin series, and in \~arioustechnical seals at Amchitka. journals and government publications (Canadian, Other investigations of valious aspects of the Japanese, and American). Amchitka ecosysteln have been made in recent years by F\VS pelsonnel. For the most part, the results of these studies are unpublished and are available only in F\\'S files. The first systematic effort to investigate the inshore fish fauna of the Aleutians xvas made by INTERNATIONAL NORTH PACIFIC Wilimovsky and his associates as part of a regional FISHERIES COhll\lISSION ichthyological inventory. Wilimovsky (1964) re- ports: "The purpose of this phase of the study was After the Alaskan salnlon fisheries study of to conduct an ichthyological reconnaissance of the 1903 (see the catalog of Alaskan fishes by Ever- heretofore little knoxvn inshore fauna of the mann and Goldsborough, 1907) and the Aleutian Aleutian Islands. Field in\restigations extended over cruise of the Albatross in 1906 (Gilbert and Burke, a three-year period and were supplemented by 1912), there seems to have been a long period of limited material presently extant in museum collec- relative inactivity in oceanographic end fisheries tions." It is significant that, in his brief review of previous investigations, \Vilimovsky refers exclu- dealing with particular classes of invertebrates are sively to early collections, e.g., those reported by cited in other sections of the Atlas. Few of these Bean, Turner, and Evermann and Goldsborough. numerous references are available in English trans- He also notes that "Recent Russian general lation. works. . . make no major increase in the knowl- Finally, mention should be made of the Bering edge of the area." Sea Comprehensive Scientific-Commercial Expedi- \Vilimovsky lists 135 inshore fish species* from tion of the Soviet institutes TINRO* and VNIROt, the Aleutian Chain. Of these, about two-thirds of more commonly referred to simply as the Bering the species were found to occur in the \vestem Sea Expedition, which initiated a series of oceano- islands and about one-third were found at An* graphic cruises in the Bering Sea and North Pacific chitka. He emphasizes the reconnaissance nature of Ocean in 1958. The purposes of the expedition are the study, the iinplication being that a more best set forth by Moiseev (1964): intensive investigation 'ivould probably add to the Thus, the vast potential of the northeastern inventory. hslore important, he points out that Pacific Ocean remained untapped, especially with "The really diverse and rich fauna is to be found in respect to benthic and demersal items. . . . At the same time, catches of most stocks of flatfish, herring, crabs, the immediate sub-tidal zone. We hardly know the and other items had reached their limit due to natural composition of this stratum at a single location and population fluctuations (sic), and fishing even had to future efforts should be concentrated on this belt." he reduced because of the depressed condition of these Siiuenstad et al. (Chap. 19, this volum?) show that staple stocks of Far Eastern fishery. the recent Ainchitka investigations did, in fact, The Bering Sea Expedition therefore had the problem of discovering nerv fertile fishing grounds for provide an opportunity for more intensive sam- the Far Eastern trawling fleets, highly experienced and pling of the fish fauna in deeper waters around the xvell equipped for this type of fishing. island. Wilimovsky concludes, on the basis of his This statement, by one of the architects of the extensive collections, that "After eliminating spe- expedition, suggests that the principal objective cies occurring throughout the North Pacific basin, was the discovery of commercially exploitable the remaining Aleutian Island fish fauna shows a fishery stocks, especially those susceptible to har- closer relation to the American continent than to vesting by Soviet trawlers. Studies of oceanography boreal Asia." I-Ie also notes the high proportion of and marine biology were focused mainly on fea- Cottidae in the fauna and the existence of several tures that had a direct bearing on the fishery. Most endemic cottid forms that ". . . range from the cruises worked primarily in the eastern Bering Sea mid-Aleutians westward to the I Stullken, Donald E., and Charles hl. Kirkpatrick, 1955, sance of Amchitka Island, Alaska, Arctic A,ithropology, Physiological Investigations of Captive hlortality in the VII-2: 118-128. Sea Otter (E,rhyd~ohctris), in Tm,tsactio~~sof the ZOtlt Turner, Lucien McShan, 1886, Co~itributiottsto the Natrrral r\'ortlr American il'ildlife nnd Naturnl Resorcrces Confer- History of Alaskn, in Arctic Series Publication No. 2, ence, pp. 476-494, \Vildlife hlanagement Institute, Signal Service, U. S. Army, \$'ashington. \\'ashington, D. C. \Vilimovsky, Norman J., 1964, Inshore Fish Fauna of the Sotton, G. hl., and R. S. \\'ilson, 1946, Notes on the \\'inter Aleutian Archipelago, in Proceerlings of the 14th Birds of Attu, Condor, 48: 83-91. Alaska Science Co~zferetrce,pp. 172.190. Tatewaki, hlisao, and Yoshio Kohayashi, 1934, A Contribu- -, and Joh~lN. i\'olfe (Eds.), 1966, Environment qf the tion to the Flora of the Aleutian Islands, J. Fat. Agr., Cape Thompson Region, Alaska, USAEC Report PNE- Hokkaido Imp. Univ., 3G(Pt. 1): 1-119. 481. Turner, Christy G., 11, 1970, Archaeological Reconnais- \Vilson, R. S., 1948, The Summer Bird Life of Attu, Co~tdor.50: 124-129. This page intentionally left blank