ATOLL RESFARCH BULLETIN No. 31 Expedition to Raroia, 'Hamotus Part

Home , Bengt Danielsson, Hikueru, Niau, Raroia, Takume, Tuamotus

ATOLL RESFARCH BULLETIN

No. 31

Expedition to Raroia, 'hamotus

Part 1. Ekpedition to Raroia, Tuamotus

ty Normanl). Nesell

Part 2. Physicdi Characteristics of Raraia

. . by Norman D..: Mev&ll'

Part 3. General Mas of Rarcia Atoll

by Norman D. Ncwell

Issued by

THE PACIFIC SCIENCE BOARD

National Academy of Sciences--Na-tional Research Council

Washi~gton, D. C.

November 30, 1954 Part 1 EXPEDITION TO RhROIA. TUAMOTUS by Norman D . Nevie1.I.

Introduction ...... 1 Preface ...... 1 Cooperation of the French Administration in Tahiti . 2 Field Operations ...... 3 TheMap ...... 3 Collaboration in Fieid Nark ...... 4 General Results ...... 4 Some Fund.=aienta.l Problems of Atolls ...... 5 Regional Setting of the Tuamotu Archipelago ...... 7 General 1ntmducti.on ...... , . 7 Cliniate ...... 10 Currents ...... 10 SouthernSweLl ...... 10 Hurricmes and Tsunami ...... 10 bibliography ...... 12

Illustrations 1 . Map of Pacific Ocean shoxing relationship of euuatorial currents to the Tuamotus and other island groups 2 . Uap of Pacific Ocean shov:~~gandesite line and Eastern Pacif'ic barrier to animal and plant migratioi~s 3 . Stmcture Lines in ?art of Vne %u%h Pacific li . Re:l.ationship between ship passes and prevailing winds in the Tuaino tus EXPEDITION TO RAROIA, TUMOTUS by Norman D. Newell

IN'rHODUCTPON Preface

An intensive ecological reconnaissance of Raroia, a Polynesian atoll about 450 miles northeast of Tahiti, was successfully completed during July and August, 1952, by a seven man research team representing the biological and geological sciences. This study, the third project in a program of atoll. studies organized by the Pacific Science Board of the National Research Council and supported principally by contract N7-om-291(04), NR 388-001 with the Office of Naval Research, enjoyed the active aid of the French colonial government at Tahiti, without which the study of Raroia could not have been undertaken. Funds equal to about 20 per cent of the total bvdget were supplied to some of the team members by The American Museum of Natural Histo~j,under a grant from the Humble Oil and Refining Company, and by the U. S. National Museum, The Univer- sity of Hanaii, and Mr. George Vanderbilt,.

Although splendid results were obtained, the team was handicapped by failure to receive approxjmately one-half of the expedition equipment, strike- bound on the eve of departure in a California port. Part of this highly specialized equipment, assembled at considerable cost over many months, was generally irreplaceable and the success of the expedition was, therefore, jeopardized. Essential items withheld from the team in this way included a boat, several rafts, outboard motors, a compound microscope, books for field determinations of corals and plants, an underwater carnera, fishing gear, warfarin for studies in rat extermination, fish poison and gasoline. Nithout replacement of some of the most crucial of these items a comprehensive study of Raroia could not have been undertaken.

The problem of supply was difficult for the expedition because freight sailings from the United States to Tahitiare infrequent and there are none from Hawaii, where the team members were convening early in June. The expedition was reprieved by the circumstance that the Tahiti government had just purchased a small vessel which was being prepared in Honolulu for an early voyage to Papeete. Through the courtesy of H.: RenB Yetitbon, Governor of French Oceania, certain replacement items were carried on this vessel toTa- hiti. These included fishing gear and poison, warfarin, special fish drums, cans and canning equipment for sea weeds, and one outboard motor. The equipment and supplies were delivered to the team on Raroia in time for use of most items but too late for use of the warfarin in the projected studies on

Pronounded Rah-ro-ee '-a Members of expedition, in addition to author, were John V. Byme, Bengt Danielsson, Maxwell S. Doty, Robert R. Harry, J. P. E. Mor- rison, and V? Jan Newhouse. rats.. The yat'poison wk, returned to Papeete and turned over to the Institut de Recherches M6dicales.de 110c.6anieFrwcaise ,.. . for studies in Tahiti...... The selection of ~ar$iaGdng the numerous atolls pf French Oceania was most fortunate. The selection was madebecause the team anthropol.ogist, Bengt Danielsson, had recently. spent 1.8 montbs'dn the atoll studying acculturation of thepeople. He was thoroughly familiar with the, language.,-the inhabitants, and working conditions. More important, he was assured of tHe complete support of the people.. He was Gnthusiastically celcomed'a~an..adopted son of .$: Raroian fmily and a member 'of the community. Having come .them, first.:as a,., tQ; . . member of the intrepid Kon Tiki crew, later as scholar an8 'st$en.t :of their: . , ways, the people had wquired respect for Danielsson that.amognted'ah&t to veneration...... Because of these.special'circumstnnces the people made available to..;thq: ., expedition team their considerable resource6 including boats, outboard motors,,. cleLan cistern water, several houses, a shoiver bath house, a toile-t, refriger- ators, and all-vave 'radio receivers. The transportation facilities 'itere I neither adequate rlor reliable, but tKey permitted a successful comprehensiye: reconnaissance of the a.toll.. . . ., . ... . !.' i : !.. , .., .., . .' .. ... , '. , . .. Coooeration of the French Administration in ~ahiti' ::. . , -...... , . . , , ...: , :V ' I ... . .Besides supplying transportation for part of.&r'equipment fronii%lonolulu to Tahiti; Governor Eetitbon provided 'ro+nd trip .

,'that.a medical officer ..?obl.d visit ,'~ard%&~.Hikueru and knaa.' f6r ' routi~eexam- inatibns,b,f the populatidn. He also'persvaded Colonel Cnavat, chief of a. ..~renckmapping missio;, to acco,mpany tfie';fl.igIfr, to 3a.roia in order to:ht$~ist . . in obtiining-suitable 6hotographs of; the ato& ...... , ,.. . . The staff of the Institut de Recherches M~dicalesde 1''0~6anie~~~'cai~e, Dr; 'Jol~n:K.essel,Dr. Georges Torres, and Mr. an& IWS.'Glen Pkrrish, were helpful and co&eous topembers of the expedition. Helpful aid ind counsel were

freely giGen 6y-m&ny.of .the cititfens of Papeete., ... : . . , . .. 2 Valerie Zirkle Neive2.l and G. l?obert.Adlingt& f The: ~se&cmMuseum of Natural History devoted many weeks in the selection, .pu&chase,. md'packaging of expedition supplies in New York, ard Mrs. Nei:.ell,"~:ho.preceded the research team to Tahiti, gave valuable aid there in completion of arrangements with the Government. Without Mrs. Newell's intervention it is doubtful that t,he difficult but urgent radio communication between Papeete ana Raroia v~ouldhave been finally established. Untiring efforts of the staff of the Pacific Sci- ence Board, Mr. Harold J. Coolidge, Mrs. Lenore Smith, and especially Miss Ernestine Akers, on behalf of the expedition mere in large measure responsible for the successful conclusion of the field work. Through the courtesy of Dr. Preston E. Cloud, Jr., a pH meter and a plane table, alids.de, and surveybg rod of the U. S. Geological Survey mere made available on Loan for ow stuciies. Dr. John W. Wells identified the corals and Mr. J. Sperrazza, the Foraminifera.

Fortunately for the needs of the ex?edition, the people of Raroia are moderately prosperous. Eley live in a single vi-llage, ~arumaod(~~arurnaova), situated on the lagoon cettr the single ship pass. Tha nume;.ous facilities of the village of 127 persons mere xore or less continuoilsly tivail8.ble to the research group. Trvo radio receivers monitored t,he Papeete program almost continuously, so that official messages :n the TaXtian language sent to us from Tehiti were promptly received and delivered. This was a valuable service during three difficul-t weeks before our contact with Papeete, via Hikueru, was established by means of our small hand-ge:lerated txo-,ivay radio.

The clim'te of Raroia is ideal. md altogether the working conditions were very comfortable.

--Field -Oaerations All. direct negoti&tions with -the peopl" mere turned over to Danielsson, who was aided by Miss Aurora Matua, a Tahitian scnolar. Besides his om sci- enttfic investigations of the people, Danielsson was given charge of the medical supplies, the kitchen help, md th9 paying of wages. Geriesal camp policies and project plans were developed by the combined team.

Scientific activities were distributed as follows: P'nysical ecology of the a-toll (geology) and corai distribution, Nor~anD. Neiwell, assisted by John V. Syrne; biological factors of ecology, and plant distribution, Idaxwell S. Doty, assf sted by Jan Newhouse; animal distribution, excepting co,rals and fishes, J. P. E. Morrison; fish distribution and ecology, Robert R. Harry; ethnology and human ecology, Bengt Danielsson.

The [email protected] for aerial photographs was taken during the early morning of June 20, 1952, ~hichluckily happened to be the only clear day of a long series of overcast days. Tvo 35 nun Argus C-3 cameras with K2 filter were used interchangeaSly for vertical views. One camera at a time ws mounted in a blister replacing a side window. Bengt Danielsson and E3Pamrell Doty operated these two cameras, making exposures at 15 second intervals. The flight was made at an altitude of 3000 neters in order to obtain negatives of convenient scale (1:60,000) and the scale was later confirmed on the groucd by pla?e- table traverse. Surprisingly good resu1.t~ were obtained in t~ocircuits of

Tehave adopted stadard Polynesian rzther than French forms of the native names. The Polynesia1 -g is pronounced adoften spelled ng.- the atoll with nearly complete coverage (95 'per cent) bf the rim. In addition, oblique photographs of the lagoon shovr hundreds of small pa.tch reefs many of which eventual.1~may be approximately plotted. on the map after calculating cor- rections for oblique perspective. The vertical photsgraphs were printed at 1:20,000 and .this scale was ampl.oyod for compilat5on of a base map. The negatives have been filed with tile Colonial. Government in Papeete. The final map is Part ? of this Bulletin.

-Colla.borm in Field Integration of heterogeneous fleid data from diverse di.scipli.nes is essential in an ecological reconnai.ssance sixh as this study of Raroia. Personal bias in the collecting and evaluation of data is not easily weighed, however, and it is usually difficult for the specialist to accommodate iris attention to problems outside the fringe of his personal experience. Efforts were made at Raroia to overcome these natural difficulties. Time was taken by,each partic- ipant from his special tasks to accompany the other members in the field in orderto establish and maintain contacts in overlapping oreas of investigation end to learn about more divergent matters. There ?:ere several rmnd-tab1.e discussions and innumerable con;rersations involving various combinations of per- sonnel. In generai, the smaller groups vrere more satisfactory for discussions than the assembled team.

Pjithln the framework of -the project a ncmber of special studies were undertdcen involving two or more collak.orsting embers of the team. Epmples are: 1) sever& transect surveys, 2) land tenure and copra production, 3) classification of native narnes of, and utilization of plants and animals, f+) native term for geologic features, and many others.

General Results Raroia. is similar to other Tgamotuan atolls, but cons~icuouslyunlike recently stud.ied atolls of Micronesia (Bikini, Axno, and Onotoa). Since Poly- nesian atolls generally have been inadequately knovm, Raroia may well serve as a standard of reference for Polynesia. Great quantities of field d.ata and specimens were obtained. These include photograinmetric data probably super- ior to those available for other South Pacific atolls. The collections indicate an unexpectedly rich biota comparable to those of Samoa and Bikini. Our results suggest that previous conclusions have been based on inadequate col- 1.ecting. For example, 286 species of fishes have been known from all of the Tuamotus, but our collections from Raroia include approximately 140 species. Fifteen species of corals have been known previously from the 'i'uamotus, but our collections from Raroia contain 53 species. This work shows that very many species range muck farther to the east in the Pa.cific than had been believed, and knowledge of the biota of the Central South Pacific has been greatly expanded. Data on tine people and 'their hj.story are especially complete. The first results on the expedition to Raroia have been issued as Bulletin 18, Ichthyological Field. Data of Raroia Atoll, Atoll Research Bulletin, 190 pp., 1953. Other reports appear in the present and subsequent numbers of the Atoll Research Bulletin. Some Fundamental Problems of Atolls

Cora' atolls, justiy celebrated in a voluminous scientific literature running into tens cf thousands of pages, have long ~ttractedthe fascinated attention of voyagers in tropic seas. The center of interest uctil recently has lain chief1.y in the genesis of reef forms and the nature and' origin of the central lagoon vihich gives atolls their most cnaracteristic expression.

Atolls support more or less balanced marine and usually also terrestrial communities surrounded by relatively sterile waters such as charac.l;erize the plankton-poor waters of tie tropics aid subtropics (Sargent & Austin, 19491, Tne origjn and economy of these relatively simple communities provide many un- solved ecological problems of the greatest fundamental importance.

Consider ths difficult and complex initid colonization of a newly formed shoal area or volcanic island surromde2 br deep sea. In the great erup4;ion of ICralcatau in 1883 the local is?.an.-l life and shaol-water organisms were largely exterminated an2 the area fgr a time was rendered uninhabitable for most species (Cammenuan, 1948). Ei,$hin a fex ~eeksreinoculation by air'brne and seaborne spores, seeds, larvae, and adu1.t~of hardy pioneers was begur~. Colonization of Krakatau, carefuily studied by Dutch biologists, seems to have followed predictable chance fr~quencedS.rect3.y related to larval hardiness, length of migratory stage, and various 0th~fa.ctors that paice up the repertory of di~persalfaci1itj.e~. Some sgecies arrivad early, others later, the relative time of introduction being directly related to the frequency with which colonizers of each species could successfully bridge the barrier of deep water. Only the barest outline in this ccmnplex series of events is really known. Much must still be inferred.

The colonization of a volcanic mountain like Kraltatau on first considera- tion may appear much more comp1.e~than the co!.oni.zation of a low, relatbely sSerile atoll. However, avoiding for the moment the involved question of the various origins of atolls, it is generally agreed that many atolls undoubtedly did. originate through the gradual. sjnking of reef-encircled volcanic mountains. The biota of such a.tolls, in a sense, are in some quarters considered relicts derived from the axcestral high islands, restricted and. modified by relatively homogeneous and almost monotonous ecclogic conditions. The ecologic simplicity of atolls rnakes them especially attractive to those who mould try to understand the interrelationships and processes among the orgwisms of a primitive sea- land community. This s?.mplicity, however, is only apparent and misleading. Most of the ecologic problems of atolls are indeed poorly understood.

The ecologic interrelationships betnreen coralline an$ other kinds of algae on %he one hand and t'ne coral animals so characteristic of tropical. reefs are not understood. Tne intimate and almost universal association in reefs of these specialized plmts and animals strongly suggests that they may be interdependent. Reef-buildhg (hematypic) corals invariably contain in their tissues zooxanthellae, unicellular algal. symbionts, wiiich a;id the corels by utilization of animal waste and by secretion of oxygen. However, many cords may feed exclusively on animal plankton (Yonge, i940), which in turn must feed on phytoplankton. Sargent & Austin (1949) have determined that the quantity of plankton swept across the reefs of the northern Marshalls by wind-driven currents is gross1.y inadequate to support the reef animals. They conclude, therefore, that the reefs are self-supporting; that is, the reef algae pro- duce at least as much organic matter. as consumed by the reef animals, and that the rate of production of organic matter by an at'bll as a whole is several times as high as that of the surrounding ocenn. Rather than filte~ingplank- ton, the marine corn~nunityof the atoll, a~cordingto these investigators, ab- sorbs inorganic nutrients from the passing equatorial curreat (Sargent & Austin, 1949, pp. 245-2491. ... Surface tropical waters in midocean are classicaLly considered deficient in nutrient sa3.t~. The sit,uation over Cieep ~xdersnear coral reefs has not been sufficiently investigated, bu't it sesms proba9le that turbulence extends to considerable depth where coral reefs are bathe6 in the equatorial currents and this turbulence would certainly bring nutrient rich deeper waters to the surface.. Orr (~933~p. 62) found evidence of uprelling along the Great Ba,rrier reef. This might well be most prmounced &wing rough weather ?:hen adequate observa-Lions have not been made.

Corsl animals, as well as coralline algae, generally ere most pro8uckive at the outer margins of seaward reefs. This growth vigor on the seaward side, accord'ig to Yonge (19~0,p. 392) is most probably related to the greater trans- parency of the water there as compared to the lagoon which all.6~~photosynthesis of zooxantheilae and coralline dlgae to sxtend to maximum depths. He considers this factor more important than greater :;up-,~l.iesof oxygen and animal. plankton at the outer reef edge. !ki.s matter requires furtier study.

It is a matter of consi3erable interest that a 1.om bank of coral gravel throw up above high tide level by a storm very quickly supports a characteristic varied assemblage of higher planks nithou% the development of humus and a soil profile. Coconuts germinate and become healthy trees in apparently sterile gravel. It has been demonstrated that several kinds of algae fix nitrogen from the air. Some of t'nese (Myxopbyta) grow bettor on an alkaline sub- stratum than elsovthere. Algae aided by 'oa.cteria fix nitrogen better than do the bacteria alone (Chapman, 19W, p. 304). It may be considered as probable that the fertility of newly exposed accumulations of coral gravel and sand is to a large extent a result of the activities of nitrogen-fixing algae and bacteria having a range of osmotic tolerance such as characterizes the flora of the littoral zone. Some of these very likely are normal inhabitants of the sea. Sea birds, which rest on rocks and gravel bars, precede the terrestrial. vegetation and enrich the ground with fertilizer gathered over a wide radius of open sea. Investigations into the colonization and proiiuctivit,y of a,tolls, al.ong these and other lines, promise to yield significant results. fiathemay (Atoll Research Bulletin 16) has emphasized the relationship betvieen atoll vegetation and variations in the bird populations and phosphate concentrations in the soil.

Coral atolls are enormous accumulat~onsof calcium carbonate. Interest in the conditions of origin of limestones )as greatly stimulated the inves-ti- gatien by geologists of calciurr, carbonate in modern' sass. In most areas of the sea floor at the present time and dur-hg the geologic past calcareous material has been predominzntly !n the form cf slceletal structures of animals and plants, hence, the sedimentationi,st, tie paleontologist ond the marine biolo- gist are all directly concerned rith the problem of fornation of such sediments. Calcium carbonate is also precipitated directly from sea~yaterto form inorganic hu&s and sands (Necell., &t al 1951). ~lthoueh:the proE?Lem has been - --, , . ,,, '. uhder s.hid~~formay yegys; fi.nal conclusions cannot b'6 rezched regardifig .the . r'elative.'imr,ortance Sf inorganid precipitation because the S&ihiii.ii of 'Cal- . . . . ci& carbnate in sea water is not well und&stood. , . . ., :. . . .,,:;: .... .:. Surveys by many investiga.tors (Trask, 1937), shorn that the calcium car- "' hnate content' of ixodern marin'e sedinients is greatest- at low latitudes. This is related certa.inly to the hi.ghsr -production of calcargous $inimals'~d&ants . - in m&?m waters, and to the fact that warm rnarjpe waters'often are 'S'iipers&$ur- ktedVith. calcium carbonate. It'is well known tba.b, ahiinals with cnlcareo* skeletons, such as corals, mollusks ana foraminifera, c!an extract'lime'frorn un- saWrated waters, but the exact mmner in vhich this is' done is not fuj:ly'.: - understood. In any case, the secretion of lime try these orgabss is enbrmously . . gr~ate~.inwarm., shallow seas.where the waters (are supersaturated with respect to calcium.cartonate. in the case of lime-secretiug algae, extraction of car- -ban dioxide during photosynthesis causes concen.tratfon of 'carbonate ions +rid precipitation of lime. Hsating of the water and concent,rution due to evapora- tion will. also cause precipi'tstion of c6lcium carbonate from sea water, par- ticularly when tile water is agitated. ,. . , .. The coral reefs of the South Pacific we among the chief areas of cjl'cium carbona3e deposition of the world. It is 20 c?oubt significant, that thes6, like other important areas of lime' accumul.ation in the Atlantic and Indian Ocea~s,;are bathed by tvarm equatorial currents (Fig. 1) probably su~erskturated with respect to calcium carbonate. It is highly probable that depositionof liae sediments in all of these areas nould be greaJ;l.y diminished or even' neg- ligible ~Jithouta "conveyor beltn soxrce of supply. . . . ', .<

Regional Setting of the Tuamotu Archipelago --General ---introduction.-Lying near -the center of tie Pacific Ocean the Tua- mot0 Archipelago is exceptionally isolated from both eastern and mestern continents. The atolls of this group are outposts at the southeastem ffi.inge:of the great Indo-Pacific biological realm, separa3ed from the hericas by the most effective water barrier on earth to migrations of shallow marine aid terrestrial organisms, the broad and uninterrupted deep waters ofthe eastern'' Pacific (Fig. 2). The biota of the Tuamotu is an attenuated indo-pacific 24 assemblage having little in common with thst of the Americas (Ekman, '1953, p. 72). 1t was derived mainly from tne west,'.partly br innumerable island jumps and perhaps via the equ~torialcounter current. (Fig. 1) md otiner occasional drifts opposed to the prevslent westerly circulation at these low latitudes.

Most of the Polynesi,an islhds are oceanic in the sense that they lie well inside the andesite line (Fig 2) vlhich is rather generally held by geologists to form ths structural margin of tne Pacific Ocean.

Locally, as at Clipperton Islands, less thm 700 miles from :!lexica, Indo-Pacific elements hsve successfully crossed most of the eastern Pacific without really obtainiag a foothold along the Americas (Hertlein and Emerson, 195:). Many geologists now favor the view that the continents are gradually ex- panding by successive orogenies at the expense of the ocean basins. Implicit in this view, the island'arcs of the western Pacific rather than represenl;ing old and collapsed areas are late increments welded to the Austral-Asiatic continent.

Next to the Maldives the Tuamotu Archipelago conttiins the greatest number of atoils, spread in a broad belt two hundred miles wide and more than a %how sand miles long from northwest to southeast. Pie 78 islands of the Archipelago of which 76 are atolls are arranged in several linear series reminiscent of island arcs (Fig. 3). The distribu.tion and proximity of these atolls to %he volcanic Society Islands, which ideally illustrate the successj.ve stages in ~a&inrs theory of subsidence, encourage the view t,hat the Tuamotu atolls rest on chains of volcanic mountains which have sunk beneath the sea so slowly that the reef growth has been successfully maintained near the surface. This is not the .same as Stenrns (1946) suggestion that the reefs o low islands were formed in shallow water on the crests of suharine folds.2 There are rather striking features that characterize most of the Tuamotu atolls and Raroia is in most respects typical of the group. The regional trend of islands in the south central Pacific is from southeast to northwest, but a few of the atolls along the northeast side of the archipelago, including Raroia, are elongate toward the northeast, at right angles to the regional trend. The prevailing currents in this area are also from the northeast. However, since there is no correspondence of form v,ith current directions in most of the atolls, it seems safe to conclude that the northeast trend of Raroia and a few other atoll6 reflects structural conditions of the basement divergent from the regional trend.

The configuration cf the sea floor around the Tuamotu group is very poorly known. A few soundings taken by the Albatross suggest that the northwestern atolls of the archipelago rise from a platform abut 800 fathoms beneath the surface. Soundings around the central and eastern atolls indicate smaller pla- teaus or spurs uniting adjoining islands and show that the eastern atolls are sepwated by channels of great depths (Agassia, 1903, p. xxviii).

Kuenen (191;) points out that the postulated folds would all have to rise to about the same critical level of reef growth near the surface, an extremely unlikely coincidence. Follovdng Daly's conclusion (1910) that a11 pre- Pleistocene reefs were killed and destroyed during low rater levels of the Pleistocene Kuenen argues that solution, rather than mechanical destruction, may have reduced many limestone banks and islands to the temporary low water levels. I am inclined to believe, that only the smallest limestone islands were truncated during Pleistocene lows. For example, there are innumerable small limestone cays and islands in the Bahamas, B.W.I., which were terraced during Pleistocene low levels, but the general forms of the islands were not much affected. Terraces cut at, this time are rarely more ttan a half mile wide fe.g: Newell, al:, 1951). Most likely the Pleistocene low levels resulted m some denudtXion of exposed surfaces but there is no compelling evidence for extensive marine planation during he short duration of the low levels. Evidently the reef organisms were not ~liminatedbut they were maintained near the level of the fluctuating sea. Of course, this argument does not affect the possibility that some atolls may lie over submerged guyots (see Hamilton, 1953). 1. Map of Pacific Ocean showing relationship of equatorial currents to the Tuamotus and other island groups. 2. Map of Pacific Ocean showing andesite line and Eastern Pacific barrier to animal and plant migrations. 3. Structure lines in part of the South Pacific. The bottom deposits in the deep pwts of the oce* (1700 to 2500 fathoms) separat5xg the groups of Tuamotu atolls are predominantly red c1.a~. In shallower waters pteropod and coral sandy ooze is the principc.1 sediment (&. G., p. i.Solxidings near the etolls show steep slopes characteristi.c of atolls (5.*., p. mix). Several observers have recognized eleva5ed reef limestone in some of the atolls at the northwest end of the archipel.ago. From this evidence Agassiz h8.s concluded that Miiakntea, whish is somewhax dissected, was uplifted almost 230 feet (Agassiz, a. sit., p. Z), >fiatdiva., l.0 to 12 feet (@. &., p. 55), Tikahau, several feet (@. &., p. 521, Raqiroa, 15 to 16 feet !&. G., p. 20), Kauhra, 17 feet (a.cit,., p. 17), and Niau, 20 feet (0~.a., p. 20). In addition might be cited elevated coral recf limestone examined by us on Anaa, which rises at leasf 15 feet above low water, new the village, along the northern shore of the atoll.

Evidence of a.pprecj.ab1.e and uneven. recen:t uplift throughout the northwestern part of the ~umotusis v:ell established. Agassizf s obsersa;tio:ls bear- ing on uplift farther to the sou-Lheast must be te!ser? witin reserva-tions because he frequently mistook large reef bl-ocks throw. up by stom;s end welded to the reef flet a.s erosion remnants of ome higher platforms.g Dana had earlier cor- rectly interpreted

As pointed oyt long ago by bo%h Daa (1890) and hgassiz (1903) the Tuamotu atolls are characterizes by a narrow rim and by reiatlveiy few end narrow ship passes. Of the 76 stolls of the group 47 are without a ship pass, 21 have a single pass, and 10 are each provi.decl with tmc ship passes (carte de la hlarine, No. 1716). The passes mark gaps in tie atol: rim where organic accretions evidently ha.ve not kept piice with the general. upwa~dgrowth during subsidence. The majority of the passes occur approximately on the leeward side (R.g. L). This suggests that the distribution of passes may in some way be controlled by the prevailing winds. Production and deposition of calcium carbonate sediments apparently is greatest on the windward side of Raroia atoll si-ce the dsepest part of the legoon is dcwnminc? from tine center. Early gaps along the windward rim aoubtless rere filled before those of the leeward rini. Possibly the flow of waters thrcugh the leeward passes, especi.ally concentrated at ebb tido, -- mn general, Agassi.zts work on coral reefs is poor1.y documen-bed and strongly colored by his crusade for the view expressed by 3emper (1863) that subsidence is not involved in the origh of atolls. idmy of Agassiz's assertions, such as the prevalence of outcrops of basaltlc rocks on barrier reefs the Society Islands, lack documentation and subeecl;jen"uonfFrmation. His general explanation of the origin of the T-~amotusseems fa~tastic. "In the Tuarnotus there is a great development of Tertiary coralliferous limestone, the last remnants of the forner elevated laid once covering a large area of the etolls as found also 2.n +he Gilberts" (a.G., p. xvii). "It is to the cutting down of the elevations of the old ledge to the general level, and tne subsequent building up of the ctolls by the material supplied from the reef flel,s and from :he coral slopes Vflat is 5ue the great uniformi.ty of all these atcl.Ls" !&. G.,p. 13). This casiua7. dating of elevated liqestones of lkmot,u tolls was based entirely or: supposed lithologic re- semblances to certain Tertiaz:y iimestones of Fiji (Agzssia, a. &., p. 18). jnhibits growth of corals and coralline algae, but the particul.ar factors involved are not clearlj understood. --C1inate.-The %amotu Archipelago lies viithin the belt of southeast trade winds and enjoys a mild d equable climate. Danielsson (1951) reports dom- inantly easterly winds 2at Rarcia riith a maximum velocity in 1950 of 7 (Beaufort). There were 37 calm days in 1950. The mean diurnal high tempera-. . : ture in 1950 was 30.4' C and the'mean low 23.3OC, giving an average of 26.8OC. :. The rainfall for that year was 1181 millimeters. Out of the entire year 2x5 days were clear (Danielsson, 1951). kccordb-g to the inhabitmts of Raroia, 1950 was not an exceptional year. However, Danieleso~does not record westerly winds such as are supposed to occur 3uring the rainy season (Sailing directions, Pacific. Islands, vol. 2, Eastern ,groups, 191!.0, hbl. 166, U. S. Hydrographic Off ice). .. ,

. Currents.-In accordance with the Coriolis effect the circulation of surface waters in the Tuamotus moved along by the southeast trzdes is usually toward the southwest at rates ranging from five to 25 miles a day. "In the rainy season, from October to %arch, vhen westerly viindc, squalls, and rains are frequent, the currents vary most and occasionalljj set to the eastward at rakes from one-half to two knots. I' (Saling directi.ons, Paci-fic Islands, vol. 2, ?astern groups, 19L0, Publ. 166, U. S. Hydrographic Office).

~i@ersal;f larvae of shallow-mate* anim.al.s tpward the east may be largely . , accomplished at these times.': . . ----Southern -swell.-Atolls of the Tuimo-bus are 'pounded on the south and south- rest'by a strong swell which produces continuous surf usually heavier than that of the windward side.. . Yne soutn margin of the rim in many of the atolls is relatively broad, low, and free firomislets, characteris-tics v~hichresult apparently from the exceptionally vigorousand continuous sheet flow of water across this sector.

"The effect of the prevalent southwesterly gales in high couL,hern latitudes in transmi'cting :%he'heavy sea, which is felt many hundreds of miles from the latitudes of its origin, occasions a serious obstacle to landing on these low isl.ands by rolling in on the shore in an opposite direction to the trade wind, making it often more dangerous to land on the lee thm on the weather side of the islands." (Sailing directions, Pacific Islands, vol. 2, Eastern groups, 19110, Publ. 166, U. S. Hydrographic Office).

Hurricanes and tsunami.-According to native tradition violent storms have played an inportant part in modification of Tuamo-Luan atolls and t,his is abundantly confirmed by the physical characteristics of the islands. Accord- ing to the Raroians w: ,n ds of hurricane ve1ocitj.e~that occasionally strike Raroia and neighboring Takume generally blow from the northwest, ordinarily the lee side of the atolls. These and many of &he Tuamotu,ato:Lls have the highest and most extensive land, the coarsest coral rubble, and the largest displaced reef blocks on the lee side.

JaVindirections given by Danielsson (personal comniunication) were uncor- rected for magnetic declination. Tne declination at Rar0i.a for 1950, when his observations were made, ras approximately 120 02' east. 4. Relationship between ship passes end prevailing winds in the Tuamotus. The actual frequency of tropical cyclones in these waters is un7knom, but there are published records of seven severe hurricmes in the Tumotus bek;~een 1878 and 1906. There are also less well substantiated accounts of a hurricane in 1823 and another in tne 1850's. The latest hurricane, of ininor conseqJence, touched Raroi~.in i926. lke wind of -this storm is said to have blom from the southeast quarter &tBaroia.

During the storm of 1903, still, clearly recalled bymany Fahabi-1;ants at Raroia, sea waves were 40 feet high at the oui;er rtef and water was higher than the house tops at the village. Opposing waves ars sz3.d to have swept over the highesC land from both set and lagoon, meeting wi$hin tie islands. Topographic forms of the islands, including cons?icuous flow-lines, in th.e coarse coral rubble, generally at right angles to tine atoll rim strongly suggest that. the land has been inun2.atcd many timesj indeed built up in places, by transia.tion waves that sweep across the atoll rim to the 1e.goon. . . . . Native tradi%io:?s refer to tsunami, but appmently these had but little geological sigaiflcace in the 'luamotus. The latest occurrence was on April 1, 1946, v:'nen the water of the lagoon at Bsroia rose some tmo feet above the level of splng highs . ficodi~g. thevill&ge, Garumaoa, to a dept3 of about ono foot. Bibliography

Agassiz, Alexander (190?) The coral reefs of the tropical Pacific, IuIus. Comp. Zool., Ivlem., vol. 28, a0pp. , . . Chapman, V. J. (19U) & introduction to the s.LLdy. of the algoo, Cambridge UnYv. Press, England, 387 pp.

Dammerman, IC. W. (1948) , The fauna .o.f ,Krakatm, 1883-1933; Verh. Kon. Neder- landsche Akad. Neterisch., Natuur., 594 pp. :.

Danielsson, Bengt (1951) Quelques observations meterdo~iquesfaites .& Raroia (amotu), Soc. Etudes Oceaniennes, Bull., p. 192-199, p. 236-243.

Dana, J. D. (1890) Corals and coral islands, Dodd, Meade and Company, New Yo&, 3d ed., 440 pp.

Dal-y, i?. A. (1910) Seistocene ~!I.aciat5-0nand the coral ref'problem, Am. . . Jour. Sci., vol. 30, p. 297-?08...... ,.. . :.. . , Ekman, Svcii (1953) Zooi7eo.iraghy of the sea, Si.dp!ick adJackson, Ltd., :: London, W7 pp.

Hamilton, Eciwin L. (1.953) Upper Cretaceous, Tertiary, and Recent plankton& -Ycreminifera from mid-.Pacific fiat touped seamounts, Jour. Paleont., vol. 27, p. 204-237.

Hertlein, Leo George; and Emerson, William K. (1953) Mollus!cs Lri _Cl.ipperton Island (Eastern Pacific), San Diego Soc. Nat. Hist. Tr., vol. xi, p. 315-364.

Kuenen, Ph. H. (1947) J& problems f marine eeolo~~i:Atolls canqrons, Verb. Konjnklijde Med. Kcad. vkten~Chappell,vol. 43, p. 1-69.

Nevell, N. D. 2. (1951) Sh.oal-water ,~eoloeyand. environments, eastern -Andros --Islaad, - --Bahamas, An:. ?&IS. Nat. Hist., Bull., vol. 97, art. 1.

Sci. Repts., wl. 2, no. 3, p. 37-86.

Sargent, Narston C. ; end Austin, Thomas S. (1949) Orqanic productivitr of gg -at011, 11. Geophysical Union, Tr. , vol. 30, p. 245-249, Semper, C. (1863) --Reisebericht (Palau-Inseln),-- Zeitschr. fEir Wiss. Zool., VOS. 3 p. 558-570.

Stearns, H. T. (1946) & integration of coral-reef kwotheses, Am. Journal Sci., -:ol. 2,p. 2L5-262. Trask, Parker (1937) Helztion of salinity to % calcium carbonate content of --marine -sediments, u .s. ~eolT~urve~Prof. Paper 186, p. 273-299. Yonge, C. i8. (1940) j& bioloq of reef-buildin& corals, Gre8.t Barrier Reef &peckition 1928-1929, Sci. Repts., vol. I, p. 353-391. Part 2

by- Nor!wn O . Ne~iell

General Tntro~duction ...... 13 Characte.iistics of Reroia ...... 13

Introducti.on ...... Charnels arid Islets .... Islad conglomwate Platform Zoint. s ...... Ro.np~rts beaches ... lslond gravel ...... Submerged te~races .... Zrosion of the land .... Ground water ...... Soils ...... Stratigraphy ...... Bibliography ...... 21

I?.lustra-Lioss 5 . Hydrographic map of Ra:oia atoll., and loca-Lion of prof'ile~ 6 . Rindward isiwts, prcfilos; Loca'i;lon on fig. 5 7 . Leemv2 isle.ts, profiles; iocatio:~on fig. 5 8 . Correlat.ion of strata 2n .'G est pits. Gamrnaoa 1sl.et PWYSICAI, CIIARACTERISTICS OF RAROIA

by Norman D. Newell

. . Accordj.ng to a survey 'hi!' i'4roia md Tahme, made in.1.950 by the Service 3ydrographique (L. 'J. NZ~),.,:Rhroia lies between 15O 55' and 160 14' sou-th latitude and 1.420 18' and l&?O32' west longiitude. It is elliptical with the long dimension toward the nort,hea,s-l. In size it is intermediate among the atolls uf the Tuamotus with dl.mensions of 14.4 km by 1,L km. The most obvious features ol' Karoia are n relatively lai-ge proportion of land surface to water &oil@ the'rim; .md a large number of shallow, narrow channels,. many of which are storm spillviays, incomplete a.t the seaward end. The lmd surface is low, rare1.y exceeding four meters above normal high waterg and for tne most part the surface is less t.hm one meter high. Tfie normal tidal fluctnation is sniall, about one-half meter, attainhLllf: perhaps one meter at spring tides; kence ,?;ind direction and velocity influence the water level far more than do the tXdal fluctus.tj.ons.

Raroir., like many other Tuamo-ti~a;ia.tol1.s /,Fig. 5) bas a single ship pass on the le&ard side of the atoll.. The lagoon waters are clear, and relatively .. deep (55 ni) in the center. Neigh3orir.g Takume on tine other hand, arid several . . other at,olls of t,he Tuanotus lack passes. In every case, these lagoons are '. :. shaLlow, being iwgely filled wiYn muctdy sediments. Evidently sedimnts ac- cumulate slovdy in open lagoons. . :

Charncteristics 92 Raroia

,: .: ., ,.. : ~ocation15O. 55'. s to J6O.:ii+' S iit,, and. i@ 181 B! to 14z0 32' W Long. . . : Knteq temperature.,oi' surfaca: sea .:a.ter : , . 260 c Prevailjag winds Easterly Cjclones (hurricanes.) . . Westerly Rairifall (195G) ., . . 110 cm Tides 0.6 m Area of. atoll. (including lagoon) . , LkOO km2 .:rer; of.:lagoon .: : 340 kin2 . .. Area of atoiJ rim.. . . 60 km2 .L+ngth .;f atoll . .. 44 km ,. Breadth of atoll: . . . 14 kin ximu mum, depth ,of lagoon 55 m :.:~qimulp heigbt of. iand 6 m Circumference at outer reef edge 90 kii . . Averege brnad';h of atoil. rim 0.6 km ...... -...... , ... .

A/ The outer.gi.abe1 ra!npart a-t tle elbow .of QnerOa, the mesternmost poin-t of -' Raroin, reaches the extreme e:i.evation of 19 b20 feet above normal high me. A few other poiil%s on the atoll are 12 to 15 feet above normal high ~nater.

.{ j.:! 7". Land area (35% of rim) 21 km2 Vegetated area 6 km2 Lagoon patch reefs (number) -ca 2000 Ship passes (number) 1 Visibility of 8" secchi disk Ocean 3.& m Lagoon 28 m DistrtbutionY~J~I, of underwater terraces, reef, and land forms recognizable on the photographs were mapped for the entire atoll and salient geologic features were recorded by means of aerial and gromd photographs and by maps and cross-sections. A number of topogr~~hkp-ofiles mere constructed across the atoll and in certain~criticiilplases levels wre surveyed by means of alidade qd plane table. Underw~~eribservztions and photographs were made with tiie aid of two Browne mesks and a notor-compressor built at The American Museum of Katural History. i~ bnthp~bricmap of the lagoon was prepared by Bj:rne from handline soundings. All of these observations reveal a number of facts or" interest.

The Atoll Rim --1ntrouction.-h rim of Raroia atoil is narrowest, averaging 530 to 700 meters wide, along the nortttwest side 3where these are relatively few channels and the greatest extent, of vegetated islnnd upland. Much of the surface here rises r'rom one and a half mefers to thee meters above normal high water level. On the other hand, the broadest sector of tiie 800 to 12.50 meters wicie, is at the south end of the atoll (see general map, Section 3 of th.is Bulletin), 71he2e the surface is more or less continuously awash. Thi.s i.s the lowest part of the rim. In other sect,ors intermediate conLitions between these extremes are found.

Although it cannot, of course, be supposed hat the rates of reef growth are uniform all around the atoll, it is probable that wave truncafion of the existing land areas and deposition of the resulting debris in the 1.agoon would result in a. more nearly uniform rim width than is nw the case. We may sur- mise with confidence that the relative narrowness and height of the northwestern rim are a result of full eq;osclre to the most violent storms.

Channels and islets.-There is only one ship pass through the leeward rim and on the neighboring twin a-Loll, Takume, there is none. The Raroia pass js narrow (700 m) and shallow (five and one-half to eight meters) and choked by vigorous coral growth v:hich may eventually block the pass. At the sout'l end of the atoll facins the strong soui;hwn swell characteristic of the region, there is a broad inundated sector of t'ne rim approximately five kilometers long and one and a half kilometers v:ide vihich is barely awash at low tide. Else- where, there are approximately 260 shsLlom charlnels across -the rim ~hichare drained or are awash during ].ow water ot. the seaw3rd ends and covered by two to two and a haif meters of water zear the lagoon ends. There are roughly 280 islets around the atoll, of which only 60, inclilding most of the larger and higher land areas, are on the leeward (northwest) side of the atoll. Althnugh the larger islets also lie on the leeward side of neighboring Takume, this is I 5. Hydrographic map of Raroia ahll, wd location of profiles. unusual among atolls, even in the Tuamotus, because most frequently the land areas are concentrated to the windward vhere gravel and sand are most vigor- ously heaped up by waves. The storms of greatest violence regularly strike the west (lee) side of Raroia and Takume and there is much evidence of island building by Yiumicanes there.

In addition to the shailow channels or spillways between islets, there are some 160 deep, angular clefts or notches (incomplete channels) in the lagoon shore similar to the channels except that they do not ex*end across the island to the seaward side. These clearly are being lengthened headward as storm vaters cross the islets toward the lagoon, arid they represent various steps in forrflation of shallov~channels. The metnod of formation of the channels is it-- self an interesting probiem considered more fully on subsequent pages. It is concluded that t,he shallow channeLs are all of very recent origin and were formed chiefly by mechanical (hydraulic) erosion of the uplifted rim, but the ship pass pr0babI.y has been inherlteii f-con? a Pleistocene lor water level.

Island conglomera~jjlat,form.-The smal!.er islets are formed of a flat platform of c~arse'limestor~econglomerate (p&okota in Tuamotuan) from about one- half to a little less than one meter above normal high vater on vffiich, here and there, are small patches of sand and coral debris generally lagoonward of the islet center (Figs. 6, 7). A fen of these patches of rubble are vegetated; the roclc plztform is bare. The larger islands are bordered by the rock platform, but wells dug in the interior (Garumaoa, Oneroa, Tetou) reveal that the central areas are underlain by uncernented sand and 'gravel to 'the water table, becoming' loosely cemented below the vrater table.

Joints.-Irregular joints are visible here a:?d there on the conglomerate . . pavement but 'they do not show lineation. Joints somewhat opened by solution are occasionally visible on bare areas of th-e outer reer' flat near the shore. 'These arc slightly sinuous generally 8;re approximately parallel vii-th the reef irnnt. None can be followed more tnan about 100 meters. Presumably they have 'been produced by adjustments or slump in the marginal talus of the atoll. There is not any visible displacement along tilese joints, but the effects of aifferential movement vould probably be quickly effaced by erosion.

Ramaarts beaches.-One or more rampart ridges of gravel on the rock platform defend the seaward sides of nearly all the islands. These are coarsest and highest along exposed promonJ;ories. The highest surface on the atoll (approximately six meters) is the crest of a rampart on the leeward side of , . the atoll (at Oneroa, Fig. 7,R). In more sheltered places the ramparts tend. to flatten, becoming sandy beaches. These are all storm features which lie inland upon the roclc platform at sonie distance from -the normal high water line. Eoiian sediments are completely lxlring on Raroia.

Island gravel.4 striking feature of -the surface away from the lagoon shore on the lee side of the atoll is a veneer of cosrse, zlgzl-blackened coral rubble with very litK1e sand. Reef 'blocks up to a meter or so in diameter, in- clu6.ing corals which live only in fij.irly deep water outside the atoll. (&- coni~era),are scattered over the. land on the leeward side of the atoll as far as the lcgoon shore. These are storm gravels deposited by sheet flood during kurricanes. The gravel is coarsest on the northwest rim of the atoll, appreciab1.y finer on the ~outheastside, which generally is the sheltered side during hurricanes. This surface gravel does not contain fine sediment, but the interstices are sandEilled a few centimeters be lo^: the surface. Presum- ably, the water v~hichpasses over the islands cluri.ng storms has capacity to carry sand and humus lagoonward, 'J.eavi.ng well-trashed coarser rubble on the higher ground near the outer shore and somewhat finer sediments near the lagoon. Some sand is trapped between pebbles, however, and doubtless it is tveshed dolv- nard by the rain. Pits'dug on the larger islands shorn that the upper one to tometers of en.rth. is composed of coarse rubble and entrapped sand rlhich could . .. have been depositedonly by violent storms. ,

Subnerged terraces.-There is evidence of an outer terrace believed to lie at approximately 20 meters similor to t~osedescribed for i3i.kini (Tracey, Ladd, and Hor'fmcister, 1948) md Andros Islofid (Nev:ell, Eigby, $':hiteman ad Bradley, 1951) and there is a shallower terrace rrhich slopes'outward- at the reef front from about low water level to eight meters. A subaikrged rock' terrace, probably also at about eigh'tmetere, lies along the lagoon shore. This is largdy

Fsosion of t3eland.-The rocky outer shore all around the atoll is bstng cut w~zyand the land area is decreasing'. Outlying erosion.'remnants on tile reef flat show clearly that the shore has retrea.ted along a broadly curved front as much as 20 or 30 meterssince the conglom&cate platform was formed and elevated. Most of the products of this erosion' together with the coral gravel and form- iniferd. sand swept through 'the channels during storms: accumulates in the lagoon, which in spite .of heavy sedixentation is moderately Seep (55 m) and is floored bjr firie sand and gravel rather than mud: . . . The v~aterof the channels on the windwardside and at the broad reef flat at the south e2d of the atoll. generally flows lagoonwzrd, even during low vjater, because of tine strong swel.1. There is reversal of flon in the larger channels on the leemrd side of the atoll, where the wind generates a feeble sezward cur- relit during high wzter-,bu't this current carries hardly any 'sediment. ..., ...... Lagoon shore currents receive and rework the debris .frdm the, sejward reef T,together with Tagoonal sediments, fornts a nearly ~onhuousband of beaches, beach ridges, spi.ts and bars; these are mainly sadalong tht;.leeward (southeastern) shore and sand and gravel on the mh&:ard (northwestern) shore of +he lqoon. Each che~ncli.smarke6 at the 1.agoon by a small c'rescentlc c?.elta ol more or less loose sediments, and there is little or no reef"'growth at these places. . . Accretion of sedixents along the lagoon shore tends to msk sips of ero-

:sion'and retreat of 'cne ihore, bu-6 almost all' the islan6s have lorn heaSands an& 6uil.i.w~of beach rock rising a few centimeters abo-ie: i&gh ?at;er. - The edge af tine beschrock has heen cut bacb at mahy @aces sume 15: or 20 ineters since i.t was formed. Hoviever, it appears that the lagoon siore.may ke.advancing la- -'

~~onwaraat theno'rttieast,ancl southwest ends of the atoll. ,: . ' ...... WINDWARD ISLETS LAGOON OCEAN

.". .." .... m W of Tikqera ...... lriet north of Piu~iu

LEGEND

Reef Limestone

......

Island Conglomerare and Beachrock

Islet north of Porou Younger Sand and Gravel I G

200m 150 I4

I ... 6. Windward islets, profiles. OCEAN

Chcrnnel between Kumekume and Korere -15m

-10

n Korere and Huohine SE N - 5

150m 100 50 0 Incomplete channel between Huohine and Tornogogie .1 .2 .3 0 -4 - 5

7. Leeward islets, profiles. --Ground ---.nrater.-~na&es of nine well vwtnrs were made at the Garumaoa vil- 1::ge near the close of a dry period (June 29) when tine wells were being used 4..U.I capiici-by. Tnerei'o-e, the salini.ty was probabkf higher than the average for a year. Salinity was deteriniried by hycironteter measurement. Values for pH were obtained witn a Ganma electric meter loaned by the U. S. Geological Survey for the puypose. Tots hardness was determined by titfation. Because of in- aiivertent omission of an essential reagent, chlorinity of the vater could not be deQ.ennined. The walls ranged from about 14 to 17 parts per thousand salinity;n:it,h pII of ?.5 to 7.8. Tots1 hardness ranged betmeen 2&0% 13 to 360% . . I? par'ts per miilioa. .. ---Garumaoa --and --other -well waters--. V!ell i'f Tenqera%ure C0 Salinity o/oo Hardness o/ooo PI1

+"Ezil at 'Ietou, little used, Augusf 23; the others are at Garumaoa, June 29.

Five test ?its ware dug to the prnter table in a series across the north end of Garumaoa Islat (~'ig.8). The pits at each end ef the series wsre wXthin 75 meters of the sea. Analyses of water samples from these pits &re show below. four and five vere not dug until after the onset of sho~~erswhich came shortly after the middle of July, breaking a' drought of many weeks.. . --V

Pond # 1 2i..7 2 23.5

All of the weL1.s and moat ponds rise azd fall with the tides, but the 6.e- tails of effects of! tiries on Vnne lens of ground wai;er.were cot studied. In ge:neral. it appears that the Cresh watzr lens is adequate for plants and animals on the large? islancis even during periods of drought. Rain-water cisterns supply nearly all of tine drinking water, however, at the village. In all of the we1:i.s end pi-ts the sand ai~dgravel were unconsol.idated down to the highest level of the fluctuating water table. Below .irk1is level. the sediments are loosely cemented to form friable calcai.enite.

Soils.-The land areas of Raroia are almosli riestitute of soil over all but the iwvest surfaces. ProixUy leaf ruold :mc! humus are periodically stripped, @.wayor burled bysheetflood duiling hurrrmes. SuXphurous muck is found in a fe::i st:aq?y areas just inside the :.aqoon beach ridges. These' are natural . . moah ar;d .the aaters are quite sallie (see preceding table, pond 1). Farther ialaxlc; are small taro sits all dug before the present century. Th8 muck of thse pits is acid.. As might be expected, the surface rubble of the islands gellorally is alkaline. The conglomerate, coral gravel ald for&niniferal sand :!hhich compose the islanc?s are neariy pure cdcium carbonate. Clay minerals and sl.lica occur on:Ly as trace3 zxd do nct form significant accumulaiAons. High alkalFn?.'ty is assured by the local custom of periodically burning over under- b~ushand coconut husks, wnich resul.ts in the disintegation and calcining of the surface gravel. -Strsti.!ga&.-Five test pits txo to three meters c'eep 'were 6ug across. iIcmohomo (n,or-th end of Gammaoa Jslet) i.n crcier to study the succession of deposits and to secure samples of ground kter (Fig. P). Tke outer pits were nea.7 the tivo shores and the ocemmard hole was d.ug vrithin. 20 meters of the conglomerate platform. ~ontrnryto expectgtion tino ccr,gl.omer&:e, composed of firmly cemented coarse debris oi' corals, was not encounkered in any of the pits, d.though dl urere'dee? enough. Thus apparently it does not form the foun- dation of this large islaid as 'with the small islan6s. The successioi?&f strata punekated in these pits is uniform across the island. There is a lower unit of well sorted medium l$esaild whi,ch extends upmard from the high water level some 50 to 240 ceritime.~ers. This is. overla-h by coarse, sancly gra-. vel 150 to 350 centimetersthick containing one or two buried soil zones end . . a poorly developed humus zone at the, top. %ells tit Gsrunaoa, at Cneroa and.. Tetou also show tha: the sediments are generally finer below c*nd coarser above. Smples of the sedimen.ts eqioseti on the five test pits were co'iiec.ted for soil anal.ysis and carbon 14 determizzation. it seems probable that the onset of deposition of the coarse mfitcrials may have originated in a amarlted climatic change, silch as developmenl of, or shift h,a hur~icanetrack. The slight drop in sea level discussed earlier probably resulted in increased rates of erosion of reef and island conglomertite. . . Logs of five test i"Ats-----.---- at the north end.of Gaiumaoa Islet hg. g)

Test pit A (5). I\lortheast corner of Garamaoa Islet, Ijepth of hole, 79 inches.

Top Inches Bed I. Loose gravel 2. Algei crilst on pebbles 3. Gray-buff sane ad gravel. 4.. Gray sand with some humus 5. Gravel, subangular pebbles up to 4" an Gravel Sandy Gravel

Approximate Low

8. Correlation of strata in test pkts, Garurnaoa Islet, Bed 6. Sandy gravel. with roots 2 7. Sand, light-buff with scattered pebbles 5 c. Coarse gravel, washed, angular, a few boulders 9 9. Sand, medinm, well sorted, Light-buff 29 10. Sand, gravelly, grayish 16 11. Sandstone, like above 3 *

Salinity 4.1; orlginal tenperature 2V°C. N.B. Rc;ots occur scattered throughout, but predominate at levels indi- cated above. me section looks as Z.hough beds above #/6 are result of burial of older profile, perhaps by 1903 hurricane.

Test pit B (2). North end Garumaoa Islet. Dept'n of hole, 121 inches.

Top Inches

Bed 1. Mantle of coarse angular coral fragments, thb: whole 2 being coated black at surfa.ce by blue-green algae . . 2. Dark-gray sandy gravel'kith organic material 2 3. Gray sandy' gravel with' many roots 12. . . : . 4. Gray sandy gravel 20.. ~. 5. Open coarse gravel 10 . . 6. Small. cobbles vtith open interstices several inches across; lowest root zone 10 7. Poorly sorted sand and gravel 10. 8. Poorly sorted subangular gravel, no sand, unconsoli- dated 9 9. Semi-consoli$ated, coarse sand, light-bnff, rather xell sorted (beach sand) 46

Water table

Salinity 9.0; original temperature 27.5 C; pH 7.92; hardness A20 513 parts per million.

Test pit C (4). North end of Garumaoa Islet. Depth of hole, 113 inches

TOP Inches ...... 1 Bed ,1. Unsorted coral grave1;washed :3~.: : 2. Algal film on pebbles, black - 3. Gravelly sand; dark--gray, ~iithsmall roots -.62 . . 4. Sand and grivel, darker,:-major root zone 4 5. Sand and gravel, light-gray, buff, some cobbles 14 :; 6. Angularjsmall cobbles and gravel,.. . grading dom into sand, light-gray . . : ... Brormish dark-gray, ~an~~,"rvitl?'a few: roots, . .., . , humic soil with a few pebbles ..6 . . Brom, pel1 sorted medium tofine sand. 4 .: Brow gra.vel, sandy, fragments up to 8 inches ', 16 Sand, coarse, lignt-tan, well sorted with a feu

small pebbles; lightly cemented .. . . . A9 -9 Water table Salinity 3.5 (diluted by recent rain) ; original temperature 28OC, pH 8.0. iest pit D (3). North end of Garumaoa Islet. Depth of hole, 90 inches.

Top Inches

. . Bed 1.. Surface veneer gravel 1 . . 2. Black humus and leaf mold 1 3. Gravel, gray, sandy, with roots 12 4. Gravel and cobbles, sandy, light-gra:~, 'with a few small roots 1'2 5. Gravel, washed, subangular, up to 3 inches with open interstices, fine roots 7 6. Sand and coarse gravel, light-hff, fine roots 5 7. Brom sandy humic soil, gravelly, darker at top ., 18 8. Cravelly sand, yellomish tan 10 9. Sand, coarse, light-buff, well sorted with a fen small pebbles 2L

Rater table

Fluctuation of water level at least 10 inches.

N.B. The sediments below #7 are tan, those above are light-buff to bone colored.

Test pit E (1). Northeast corner of Garumaoa Islet (see Figwe for.1.ocation). Depth of hole, 68 inches.

Profile

TOP Inches Bed I. Surface coral rubble, partly calcined by burning; no

sand visible; f~agmentssmall pebbles to boulders.. '

2. Upper half inch black, humic, lower part dark brovmish- gray sandy gravel full of.,srnall roots, mainly horizontal (2 samples) 3. Sand with pebbles, light-gray (sample taken) I,. Pebbles arid co5bles, subangular, light-gray, no sand 5. Sandy gravel, light-gray 6. Hunic soil, dark-gray, sandy gravel, full of living roots, both horizontal or vertical (sample) 7. Gravel, hardly my sand (sample) 8. Sand, brosm, pebbly (sample) 9. Buff sand and pebbles (sample) 10. Gravel, washed, no sand 11. Sand, buff, with a fev pebbles (sample) Water table

Original temperature 27.5OC. Bibliography

Ladd, H. S.; and Tracey, J. I., Jr. (1949) Tie problem of coral reefs, Sci. Nonthly, vol. lxix, p. 297-305. Newell, N . D. -et -al. (1951) -Shoal-water-- -mow environrn~~t~~,e~.stem An- -dros Island, Ba.hamas, Am. Mus. Nzt. Hist., Bull., vol. 97, art. 1 Tracey, J. I., Jr.; Ladd, H. S.; and EIof:fmeister, J. E. (1948) Reefs_ of Bikini, Marshall islunds, Geol. Soc. Am., Bull., vol. 59, p. 863.478. P&Pt 3

GENEJUU MAP OF MOIA

by Norman D. Newel1

Abundaat patch reefs across the central. part of the lagoon have not been mapped. Stippled areas, vegetation cover; black areas, bare land surface (limestone, gravel and sand). Original wind date corrected for nmgnetic declination.

0 GARUE PASS