Thgi Natural Histoiy o Enewetak Atoll DOeEV/06703-T1-Vbl. 1 (DE87006110) Volume I The Ecosystem: Environments, Biotas, and Processes United states Department of Energy Office of Energy Research Office of Health and Environmental Research Ecological Research Division >»:k' ■h Volume I The Ecosystem: Environments, Biotas, and Processes o« ? J) ■ o Top: Aerial vit-w of Enewetak Atoll from an altitude of 10.000 ft looking north. The wide south passage to the lagoon is at the bottom of the pic- ^ tiirc. The three islands to the right of the passage are Enewetak, Medren, and Japtan. The deep east pass is seen between Medren and Japtan. Thei five southwest islands are seen to the left of the wide south passage. Ikuren is the first one. North of these islands is the shallow southwest pass. The Aloll is elliptical in shape measuring about 41 km from north to south and 33 km from east to west. [Photography by P. L. Colin.] H<>lti>ni: Aerial view of the northern end of Enewetak Island showing the cluster of buildings of the Mid-Pacific Research Laboratory. The quarry is visible on the reef flat. The small island immediately to the north is Bokandretak. [Photography by E. S. Reese.] DOE/EV/00703-TI-Vol. 1 (DE87006110) The Natural History of Enewetak Atoll /9 3 V. I Volume I The Ecosystem: Environments, Biotas, and Processes Edited by: United States Department of Energy Office of Energy Research Office of Health and Environmental Research Ecological Research Division Prepared by Office of Scientific and Technical Information U.S. Department of Energy ^^tKWOG,^ *° Dennis M. Devaney Bernice P. Bishop Museum Honolulu, Hawaii Ernst S. Reese University of Hawaii Honolulu, Hawaii Beatrice L. Burch Bernice P. Bishop Museum Honolulu, Hawaii Philip Helfrich University of Hawaii Honolulu, Hawaii MARINE BIOLCGlCAl LABORATORY LtSR/ HY V,t)ODS HCLl ma:: W. H. I. _J ><-CW-i|J [Photograph by Annabelle Lyman.] Bok in, kon menninmour ko im menin eddok ko ion Enewetak, ej kein kememej im kautiej ri Enewetak. This volume on the natural history of Enewetak Atoll is dedicated to the people of Enewetak. International Copyright,® U. S. Department of Energy, 1987, under the provisions of the Universal Copyright Convention. United States copyright is not asserted under the United States Copyright Law, Title 17, United States Code. Library of Congress Cataloging-in-Pablication Data The natural history of Enewetak Atoll. DOE/EV00703T1-Vol. 1 (DE87006110). DOE/EV00703-Tl-Vol II (DE87006111). Includes indexes. 1. Natural history — Marshall Islands — Enewetak Atoll 1 Devaney. Dennis M II. United States. Dept. of Energy Office of Scientific and Technical Information QH198,E53N38 1987 508 96'83 87-24863 ISBN 0-87079-5791 (set) ISBN 0-87079-580-5 (pbk : set) ISBN 0-87079-581-3 (microfiche : set) Work performed under contract No. DEAC08-76EV00703 The set of two volumes is available as DE87006110 (DOE/EV/0073-Tl-Vol 1) and DE87006111 (DOE/EV/0073-Tl-Vol.2) from NTIS Energy Distribution Center P. O. Box 1300 Oak Ridge, Tennessee 37831 Price Code: Paper Copy A99 Microfiche AOl Printed in the United States of America 1987 Foreword As activity and funding at the Mid-Pacific Research Laboratory began to diminish in the early 1980s, it seemed fitting that a synthesis be prepared of the three decades of research that had been conducted at this Laboratory on Enewetak Atoll. For 30 years the Atoll served as a convenient, accessible location for studies of Mid-Pacific island ecosystems, and several hundred scientists utilized the facility. Primary fund- ing was provided by the Office of Health and Environmental Research, Ecological Research Division, U. S. Department of Energy (formerly the Atomic Energy Commission and the Energy Research and Development Administration). This is an attempt to synthesize in two volumes the results of the Mid-Pacific Research Laboratory studies that have been published in hundreds of widely dispersed publications. It is hoped that present and future scientists involved in studies of Mid-Pacific islands will find this synthesis a convenient resource for their research. Considerable time and effort were expended by many contributors to make this synthesis [X)ssible. Thanks are extended to all these authors for their manuscripts. Special appreciation is expressed for Dr. Dennis Devancy's dedication in filling gaps in the taxonomic descriptions of several invertebrate groups. This publication would not have been possible, however, without the determination and persis- tence of Dr. Ernst Reese in organizing and collecting the material. Deepest gratitude is acknowledged for his conscientious efforts. Helen M. McCammon, Director Ecological Research Division Office of Health and Environmental Research United States Department of Energy Acknowledgments Many people have contributed in many ways to the production of these two volunncs. Regardless of the nature of the contribution, everyone listed below has given thought and time, that most precious com- modity of thinking individuals, to bring The Natural Histor\^ of Enewetak Atoll into publication. Authors of the chapters are not listed separately, even though, in most cases, they critically read other chapters. No doubt we have overlooked many who have contributed in important ways, and for these oversights we apologize. To all of you we wish to extend our deepest thanks. Donald P. Abbott Isabella A. Abbott Hazel K. Asher George H. Balazs Jerry L. Barnard Frederick M Bayer Henry R. Bennett Richard A. Boolootian Thomas E. Bowman Harry U. Brown Fenner A. Chace, Jr. G. Arthur Cooper Edward B. Cutler Mae G. De Rego Maxwell S. Doty Iraneus Eibl-Eibesfeldt Robert E. Elbel William O Forster Vicki S. Frey John T. Harrison Janet F. Heavenridge Derral Herbst Robert W. Hiatt Lipke B. Holthuis Richard Houbrick Arthur Humes Edwin Janss Robert E. Johannes Victor R. Johnson, Jr. Irene D. Keller Phillip B. Lamberson Jere Lipps Frangois Mautin Helen M. McCammon Ellen Moore William Newman Jeri-Lyn Palacio David Pawson William F. Perrin Marian Pettibone Colin S. Ramage Anita Savacool Hajo Schmidt Stephen V. Smith Sy Sohmer William J. Stanley Lyn Sweetapple Lori N. Yamamura IX Preface The two volumes of The Natural History/ of Enewetak Atoll summarize research done at the Mid-Pacific Research Laboratory from 1954 to 1984 under the auspices of the Department of Energy. The history of the laboratory and the reasons for its support by the United States Depart- ment of Energy are described in Chapter 1 of Volume 1 . Over a thousand persons — established scientists, their assistants, and graduate students — conducted research at the laboratory during the 30-year period. Their efforts resulted in 223 publications. These have been collected in four volumes of reprints entitled Mid-Pacific Marine Labora- tory Contributions. 1955-1979, U. S. Department of Energy, Publication NVO 628-1. The laboratory has con- tinued operation on a limited scale to the present. A col- lection of papers recently appeared in the Bulletin of Marine Science, Volume 38, 1986. Much of the research conducted at the laboratory was on the marine environment. The reason was that the majority of scientists applying to work at Enewetak were marine biologists. For many, this was the first opportunity to study the biota of a coral atoll. Fewer studies were con- ducted in the terrestrial environment and its biota. Nevertheless, as these volumes attest, the coverage is amazingly complete and thorough, and there are few, if any, studies of an equivalent ecosystem that equal the total research effort reported in these volumes. Volume I provides a synthesis of the research carried out under the subject headings of the respective chapters. Certain of the chapters, e.g., those on geology, subtidal and intertidal environments and ecology, and those on reef processes and trophic relationships, summarize a great diversity of research carried out by many scientists for many years. In contrast, the chapters on meteorology and oceanography summarize research carried out under one integrated program involving fewer scientists working over a shorter period. Volume II of The Natural Historic of Enewetak Atoll provides information on the taxonomy of animals and plants known to occur at Enewetak Atoll. This taxonomy represents a fulfillment of one of the first assignments to the laboratory — to determine the scientific names of the biota of the atoll. The collections on which the checklists in each chapter are based are housed at the Bcrnice P. Bishop Museum in Honolulu and the U. S. National Museum of Natural History, Smithsonian Institution, Wash- ington, D. C. In addition to the sp>ecies checklists, each chapter in Volume II provides a succinct summary of the biota with respect to endemism, range extensions, and other features that set the Enewetak biota apart from those one might expect to find on equivalent Indo-Pacific islands. This com- pendium of taxonomic information for an atoll should prove of immense value to scientists interested in biogeog- raphy and evolutionary biology of island ecosystems for years to come. One of the problems of editing these volumes has been the correct use of place names. In some cases authors used the military code names for islands while others used the native names. Even the native names have changed from early phonetic spellings to the sp>ellings currently in use and preferred by the Enewetak people. For example, the name of the atoll has changed from Eniwetok to Enewetak, and, although the correct current sf>elling is used throughout, the old spelling occurs in older references and maps which appear in these volumes. Maps giving the military code names and the native names preferred by the Enewetak people are located in Chapter 1 of Volume I. Surprisingly, it is difficult to determine the exact number of islands. Due to the effects of storms, small islands are ephemeral, and two islands and part of a third were ob- literated by nuclear explosions. Currently there arc 39 rec- ognizable islands, and these are shown on the map used throughout the book. These volumes do not report on the extensive radiolog- ical surveys and studies which have been conducted by the Lawrence Livermore Laboratory, University of California, and the Radiation Laboratory, University of Washington, also under the auspices of the U. S. Department of Energy. Dennis M. Devaney, senior editor of this volume, disaf)- peared while collecting specimens off the Island of Hawaii on August 13, 1983. Dennis was doing what he loved best, collecting marine invertebrates, at the time of his death. He collected extensively at Enewetak, and he under- took the task of organizing the systematic chapters of Volume II. Beatrice L. Burch, Dcvaney's assistant at the XI Bishop Museum, completed the task, and she has written of human history. In a small way, this book stands as the introduction to Volume II. something good that has resulted from those years. It is fitting that the two volumes of this book are dedi cated to the people of Enewetak Atoll. They, like so many Ernst S. Reese other human beings, were caught up by forces beyond Professor of Zoology their control and understanding in an immense cataclysm University of Hawaii, Honolulu Contributors Marlin J. Atkinson University of Western Australia, Nedlands, Australia Robert K. Bastian U. S. Environmental Protection Agency, Washington, D.C. Andrew J. Berger University of Hawaii, Honolulu, Hawaii Patrick L. Colin University of Papua New Guinea, Port Moresby, Papua New Guinea Robert A. Duce University of Rhode Island, Kingston, Rhode Island Ray P. Gerber St. Joseph's College, North Windham, Maine Philip Helfrich University of Hawaii, Honolulu, Hawaii William B. Jackson Bowling Green State University, Bowling Green, Ohio Robert C. Kiste University of Hawaii, Honolulu, Hawaii Alan J Kohn University of Washington, Seattle, Washington James A. Marsh, Jr. University of Guam, Mangilao, Guam Nelson Marshall University of Rhode Island, Kingston, Rhode Island John T. Merrill University of Rhode Island, Kingston, Rhode Island Roger Ray U.S. Department of Energy, Las Vegas, Nevada Ernst S. Reese University of Hawaii at Manoa, Honolulu, Hawaii Byron L. Ristvet S-CUBED, Division of Maxwell Laboratories, Albuquerque, New Mexico Stephen H. Vessey Bowling Green State University, Bowling Green, Ohio Contents Chapter Pagg Introduction ^^■^^ Ernst S. Reese 1 Research at Enewetak Atoll: A Historical Perspective 1 Philip Helfhch and Roger Ra\^ 2 History of the People of Enewetak Atoll 17 Robert C Kiste 3 Physiography of Enewetak Atoll 27 Patrick L. Colin 4 Geology and Geohydrology of Enewetak Atoll 37 Bi^ron L. Ristuet 5 Oceanography of Enewetak Atoll 57 Marlin J. Atkinson 6 Meteorology and Atmospheric Chemistry of Enewetak Atoll 71 John T. Merrill and Robert A. Duce 1 Subtidal Environments and Ecology of Enewetak Atoll 91 Patrick L. Colin 8 Intertidal Ecology of Enewetak Atoll 139 Alan J. Kohn 9 Reef Processes: Energy and Materials Flux 159 James A. Marsh. Jr 10 Trophic Relationships at Enewetak Atoll 181 Nelson Marshall and Rav P. Gerber 11 Terrestrial Environments and Ecology of Enewetak Atoll 187 Ernst S. Reese 12 Biology of the Rodents of Enewetak Atoll 203 Williann B. Jackson. Stephen H. Uessey, and Robert K. Bastion 13 Avifauna of Enewetak Atoll 215 Andrew J. Berger Author Index 221 Subject Index 223 XV Introduction Ernst S. Reese University/ of Hawaii at Manoa Honolulu. Hawaii 96822 The first volume of The Natural Histori/ of Enewetak Atoll provides a summary of the research carried out over the 30-year period from 1954 to 1984. The frontispiece illustrates the dramatic contrasts between the immensity of the lagoon and the seemingly fragile necklace of small islands which surrounds it, and also between the sea condi- tion on the windward, seaward side of the reef and the relatively sheltered waters of the lagoon. The first chapter discusses the history of research at Enewetak Atoll. The reasons behind the establishment of the Enewetak Marine Biological Laboratory are described. The authors, Philip Helfrich and Roger Ray, have been associated with activities at Enewetak from the very early days. They conferred with Robert W. Hiatt, the first direc- tor of the laboratory. In Chapter 2, Robert C. Kiste, a foremost authority on the f)cople of Micronesia, provides a history of the Enewetak people to whom these volumes are dedicated. The next four chapters deal with the physical environ- ments of Enewetak Atoll. In Chapter 3, Patrick L. Colin describes the physiography of Enewetak. Colin served as resident scientist in charge of the laboratory from 1979 to the end of 1983 when all resident scientific staff left the atoll. Following the description of the atoll, Byron L. Rist- vet, a frequent scientific visitor to Enewetak, provides a summary of the geology and geohydrology in Chapter 4. Next, in Chapter 5, Marlin J. Atkinson describes the oceanography. Under the direction of Stephen V. Smith, Atkinson participated in an important study of the lagoon circulation. Chapter 6 on the meteorology and atmos- pheric chemistry is the final chapter in the group of chapters dealing with the physical environment of Enewetak Atoll. Written by John T. Merrill and Robert A. Duce, the chapter is based on the results of the SEAREX Project. Duce served as the director and principal investi- gator of the project. The next four chapters are devoted to the marine ecosystem and its biota. They summarize the large amount of research carried out at the Mid-Pacific Research Labora- tory in the marine environment. All of the authors were frequent visitors to the laboratory, and they have done a splendid job of reviewing the research carried out in their area of interest. In Chapter 7, Patrick L. Colin describes the subtidal environments of Enewetak and reports on the research done on the subtidal biota. This is followed in Chapter 8 by Alan J. Kohn's masterful summary of research in the intertidal environment. Kohn has been a student of tropical intertidal ecology for 30 years. He tack- led a particularly difficult task because of the extensive study of the intertidal environment and its biota by many scientists over the years. Chapters 9 and 10 deal with processes and relation- ships in the marine environment. In Chapter 9, James A. Marsh, another frequent visitor to the laboratory and a recognized authority on coral reef processes, reviews the extensive work which was carried out at Enewetak on the community metabolism of coral reefs and related topics such as calcification processes, nitrogen and phosphorus cycles, and the role of detritus in the ecosystem. Nelson Marshall and Ray P. Gerber extend the ecosystem approach in Chapter 10 to include the entire atoll. They discuss the trophic relationship between the shallow reefs and the lagoon. Both Gerber and Marshall conducted research at Enewetak. The final three chapters are devoted to the terrestriEd environment. Because fewer scientists applied to conduct research in the terrestrial environment, less work was accomplished, and an integrated overview is not possible. In Chapter 11,1 rep>ort on the life history, behavior, and ecology of land crabs, review what is known about atoU soils, and conjecture on the carrying capacity of an atoll such as Enewetak. For a description of the vegetation, the reader is referred to Chapter 3 in Volume II by Janet O. Lamberson. William B. Jackson, a frequent visitor to Enewetak over the years, and his co-workers Stephen H. Vessey and Robert K. Bastian report on their long-term study of the rodents in Chapter 12, and Andrew J. Berger summarizes our knowledge of the bird life of the atoll in Chapter 13. Berger, a noted ornithologist and the foremost authority on Hawaiian birds, made a number of trips to Enewetak. 1 suspect that few readers will read this volume from cover to cover, but those who do will gain an appreciation for the complexity of the atoll ecosystem and a better XVII understanding of the intimate relationships between the seemingly fragile components of the ecosystem: the lagoon, the reefs, the islands and their biotas, all perched on a volcanic and coral pinnacle in the vastness of the Pacific Ocean. In the final analysis, however, the book will serve its purpose best if the reader comes away with more questions than answers and a desire to find the answers to these questions in future research on the natural history of coral reefs and islands. Chapter 1 Research at Enewetak Atoll: A Historical Perspective PHILIP HELFRICH' and ROGER RAYf 'Hawaii Institute of Marine Biologi^. Uniuersiti/ of Hawaii, Kaneoke. Hawaii 96744: fNevada Operations Office, U. S^ Department of Energi^, Las Vegas, Nevada 89114: current address is 10252 Hatherleigh Dr., Bethesda. Maryland 20814 INTRODUCTION The Pacific theater of operations in World War II brought millions of military personnel to the tropical Pacific, and their activities on the Pacific Islands afforded close contact and awareness of the physiography and natural history of these small dots of land scattered in the vast expanse of ocean. This enhanced awareness, coupled with a recognized need by the military establishment for increased knowledge of Pacific Island areas, led to government-sponsored investigations, complemented by efforts of many individual scientists whose interest had been stimulated by wartime visits to these islands. In the postwar period, two activities of the U. S. government focused further interest on the coral atoll of the tropical Pacific and influenced the future of research at Enewetak Atoll (Figs. 1 and 2). The origin of the spelling "Eniwetok" is lost but would appear to be a phonetic rendering of what the people called their atoll. In 1973 it gave way to the current spelling, consistent with written Marshallese, and meaning "island which points to the east." World War II demonstrated the importance of these small, scattered land masses to any military confrontation in the Pacific basin. After the war, the U. S. Navy moved to develop a series of permanent bases from among the many temporary wartime bases and outposts which had been established across the Pacific. With the prominent role of the Navy in developing and maintaining these bases, it is not surprising that the Navy's research arm, the Office of Naval Research (ONR), inaugurated a scien- tific program in the late 1940s aimed at a better under- standing of atoll morphology and of all aspects of island life from microorganisms to human inhabitants. The ONR funded a series of expeditions in conjunction with the Pacific Science Association, many of which were to atolls in the central and western tropical Pacific. Arno Atoll in the southern Marshall Islands and Onotoa Atoll in the Gil- bert Islands (now Kiribati) were subjects of intensive inves- tigation in 1950 and 1953, respectively. Scientists involved in these atoll studies contributed to the establish- ment of the Eniwetok Marine Biological Laboratory (EMBL) on Medren Island, Enewetak Atoll, in 1954. The second postwar activity which served to focus attention on the mid-Pacific area was the atomic wcafwns testing program in the northern Marshall Islands. Two atomic weapons had inflicted mortal damage upon Japan and had brought a precipitous end to the war in the Pacific. Military planners and strategists knew very little about this new and awesome strategic resource. Thus, an area was sought which might accommodate full-scale test- ing of atomic weapons. Neil Mines (1962) in his book Proving Ground describes the process of choosing the northern Marshall Islands as the testing site. First Bikini Atoll and then Enewetak Atoll became test sites, to be known together as the Pacific Proving Ground. National security considerations soon led to research and develop- ment testing and, with the impetus of the cold war, to the testing of thermonuclear weapons in these islands. In all, between 1946 and 1958, 43 nuclear devices were tested at Enewetak and 23 on Bikini — events which were to have profound and lasting environmental, social, and cultural effects upon these two atolls as well as others nearby. The nuclear testing program provided a setting, a focus of interest, and an opportunity for research in the northern Marshall Islands which eventually led to the establishment of the EMBL. THE WEAPONS TESTING PROGRAM Soon after the 1946 tests at Bikini (Operation Crossroads), which had been designed to assess the mili- tary significance of atomic weapons, the United States Congress created the Atomic Energy Commission (AEC), a civilian agency charged with responsibility for the research, development, testing, and production of nuclear weapons. This new agency was to become host and manager of the Pacific Proving Ground and, later, sponsor of EMBL. Operation Crossroads was largely a seaborne opera- tion, with logistic support from the naval base at HELFRICH AND RAY -^LJ \ J ^y^JAPAN yiOVAr i 1 -A.AHAN ^^ ISLANDS W MAft)ANA5 •WAKE ISLANDS JOHNSTON GUAM. ENFWFTAK ATOLL ^ UJELANG ATOU _^ Af BIKINI ATOLL * -'■ MAffSHALL - '. I5LAM0S KWAJALEIH^^ ■ > . .. - ■■6 ; , • . CAftOLINE ISLANDS ■ 51LBC«T ' ISLANDS ^CAHTON ISLAND ' ,^^,5,^,, » 'SlanD N€* CUINCA^"'"^, SOLOMON .^,,s..s REGIONAL MAP CMaPhiC SCAI.E IN •«Aotm:ai. wilCS ^jMIOWAY / ,0^ y _ioii.^ — ^~"^"«f -'-"^HONOLULU In ,-''"/■*«( J^ y. ^^ 1 ^^ ,.8^- / ,»>v^ ^^ ---^ItXNSTON OUAIl\ \ — — i°-'0___^^^ENEWETAIt . / AIRLINE DISTANCES MAP NO SCALE Fig. 1 Regional and airline distances maps of the Pacific and tlie Marshall Islands showing location of Enewetak Atoll. Kwajalein. It consisted of two tests, one an airdrop and the other an underwater detonation. The radiation and other effects of both of these tests — code-named Abie and Baker — were largely confined to Bikini Atoll, with such fallout as left the Bikini area being deposited in areas of open ocean. The same could be said of the early develop- ment tests, which began at Enewetak in 1947. The selec- tion of these atolls had been strongly influenced by their remoteness and by the predictability of wind conditions. The 1954 operation, code-named Castle, was planned contemplating use of both atolls. Detonation of Bravo, the first test of Castle, drastically altered that plan. The explo- sive power (yield) of Bravo was more than twice that which had been predicted, and local winds carried the debris, or local fallout, directly across Bikini Atoll, contami- nating much of the land area and rendering the control area and many of the experimental sites unusable for the remainder of the Castle operation (Hines, 1962). Some testing continued at Bikini, but Enewetak, after Bravo, took on even greater importance in the atmospheric nuclear testing program. During the period which ended on October 31, 1958, Enewetak was the site of 43 nuclear A HISTORICAL PERSPECTIVE 10 MILES Fig. 2 Islands of Enewetak Atoll with Marshallese names shown on the lagoon side and English code names on the ocean side. weapon tests. Enewetak, Medren, and Japtan Islands housed the command, administrative, logistic, and techni- cal support facilities, and the islets in the northern and eastern portions of the atoll served as test areas. Table 1 lists the detonations at Enewetak, and Fig. 3 illustrates the test locations on the atoll. The nuclear testing program required the mobilization of a vast assemblage of scientists, technicians, and support personnel and the establishment of laboratories, shops, and living quarters, in addition to port facilities and an air ter- minal to connect with a supply system extending through Hawaii to mainland bases as far as 8000 miles away. Test operations over more than a decade were conducted by a series of Joint Task Forces (JTFs), consisting of Army, Navy, Air Force, and AEC elements, in a coordinated operational command. The commander was a senior mili- tary officer of flag rank and had as his deputy a senior AEC scientist. The test detonations were grouped in series which, typically, lasted several months. During the times between series — usually a year or more — the support apparatus continued to function. This availability of logistic and administrative support made it feasible to consider the establishment of a laboratory facility. The AEC interest in HELFRICH AND RAY TABLE 1 Nuclear Tests at Enewetak Atoll Operation Type and event name Date height, ft Yield Location Sandstone X-ray 4/14/48 Tower 200 37 KT Janet, west tip Yoke 4/30/48 Tower 200 49 KT Sally Zebra 5/14/48 Tower 200 18 KT Yvonne, north end Greenhouse Dog 4/7/51 Tower 300 Yvonne, north end Easy 4/20/51 Tower 300 47 KT Janet, west tip George 5/8/51 Tower 200 Ruby Item 5/24/51 Tower 200 Janet, north tip Ivy Mike 10/31/52 Surface 10.4 MT Flora King 11/15/52 Airdrop 1500 500 KT Yvonne, 2000' N Castle Nectar 5/13/54 Barge 1.69 MT Mike Crater Redwing Lacrosse 5/4/56 Surface 40 KT Yvonne, north end Yuma 5/27/56 Tower 200 Sally, west tip Erie 5/30/56 Tower 300 Yvonne, by airstrip Seminole 6/6/56 Surface 13.7 KT Irene Blackfoot 6/11/56 Tower 200 Yvonne, middle Kickapoo 6/13/56 Tower 300 Sally, north tip Osage 6/16/56 Airdrop 670 Yvonne, middle Inca 6/21/56 Tower 200 Pearl Mohawk 7/2/56 Tower 300 Ruby Apache 7/8/56 Barge Mike Crater Huron 7/21/56 Barge Mike Crater Hardtack, Phase I Cactus 5/5/58 Surface 18 KT Yvonne, north end Butternut 5/11/58 Barge Yvonne, 4000' SW Koa 5/12/58 Surface 1.37 MT Gene Wahoo 5/16/58 Underwater 500 James, 7400' S Holly 5/20/58 Barge Yvonne, 2075' SW Yellowwood 5/26/58 Barge Janet, 6000' SW Magnolia 5/26/58 Barge Yvonne, 3000' SW Tobacco 5/30/58 Barge Janet, 4000' SW Rose 6/2/58 Barge Yvonne, 4000' SW Umbrella 6/8/58 Underwater 150 Glenn, 7400' N Walnut 6/14/58 Barge Janet, 6000' SW Linden 6/18/58 Barge Yvonne, 2000' SW Elder 6/27/58 Barge Janet, 4000' SW Oak 6/28/58 Barge 8.9 MT Alice reef, 3 mi SW Sequoia 7/1/58 Barge Yvonne, 2000' SW Dogwood 7/5/58 Barge Janet, 4000' SW Scaevola 7/14/58 Barge Yvonne, 561' SW Pisonia 7/17/58 Barge Yvonne, 12000' W Olive 7/22/58 Barge Janet, 4000' SW Pine 7/26/58 Barge Janet, 8500' SW Quince 8/6/58 Surface Yvonne, middle Fig 8/18/58 Surface Yvonne, middle A HISTORICAL PERSPECTIVE 162-10'E 11° 40 N FLORA EDNA'S DiUGHTEl^ EDNA DAISY -CLARA 162-20E JANET KATE LUCY PERCY , , , MARY /_/_/_/_MARyS DAUGHTER. NANCY OLIVE ■ PEARL I62°I0 E I62''20 £ H 10 MILES Fig. 3 Enewetak Atoll nuclear tests with name, year of detonation, and approximate locations. expanding knowledge of the environmental setting in which the tests were being conducted provided the basis for discussions which led to the establishment of the EMBL. ESTABLISHMENT OF EMBL Of necessity, the nuclear testing program of the 1940s and 1950s was conducted in a climate of national urgency and classification security. Important scientific and strategic information had been lost to foreign powers in the immedi- ate postwar period, and the pace of atomic weapons research and development had become a vital indicator of political power. In this environment, the establishment of a university-associated research laboratory, with its traditions of academic freedom and open publication of research results, was nothing less than remarkable. It reflected the enlightened scientific climate of the AEC and the AEC's concern regarding the long-term consequences of applica- HELFRICH AND RAY tions of nuclear technology. There was a need for more complete knowledge of the dynamic biogeochcmical processes which might lead to the transp)ort of radioactive contaminants in the atoll system to man. More fundamental was the acknowledged inadequacy of our understanding of the systematics and ecology of the highly diverse atoll biota. Early records of environmental monitoring during the test series included entries such as "red fish" and "green filamentous algae," reflecting the lack of any pertinent tax- onomic descriptions of the local biota. The College of Fisheries of the University of Washington, under contract to the AEC, had conducted studies at Bikini and Enewetak of the interaction of environmental radioactivity with vari- ous species and had made substantial contributions to the literature regarding these nuclear-affected atolls (Mines, 1962). There remained, however, a need for a broader base of information about the systematics, ecology, and life history of the atoll flora and fauna. Details of the discussions leading to the establishment of EMBL are unavailable. In the early 1950s, however, the eminent biologist, H. Burr Steinbach, then of the Univer- sity of Chicago and later of Woods Hole Oceanographic Institution, was asked by Sidney Caller of the Office of Naval Research to travel to Enewetak Atoll to explore the feasibility of establishing a marine biological laboratory. Steinbach's trip and his subsequent report recommending the establishment of a laboratory on Enewetak Atoll were instrumental in AEC's action to contract with the Univer- sity of Hawaii to establish and operate the EMBL. The contract, signed on June 3, 1954, required the university to manage the laboratory and to direct and coor- dinate its scientific programs. Policy direction and sponsor- ship were provided "by the Division of Biology and Medi- cine of the AEC Headquarters in Washington, D. C. Robert W. Hiatt, Director of the Hawaii Marine Labora- tory, became the first director of EMBL. The first orders of business were to provide supplies, equipment, and work areas for visiting investigators and to establish a reference collection of animals and plants with an ecological index for their use. To facilitate scientific investigations of terrestrial and intertidal biota, two islets on Enewetak Atoll — Ikuren and Mut — were set aside as reserves for the exclusive use of EMBL scientists. This was done to ensure that a continu- ously available source of typical fauna and flora would be protected, to the extent possible, from proving ground activities. During these early years, EMBL scientists were permitted to use the laboratory only in the intervals between test series. However, marine scientists from the University of Washington Applied Fisheries Laboratory, under separate contract to the AEC, were in residence during the actual test events. Their work at Enewetak and elsewhere in the Pacific is recounted by Hines (1962) and is reported in numerous published papers. The laboratory was first quartered in a rectangular metal building, with an aquarium lanai, located on the southwest shore of Medren Island. The building was equipped with a simple seawater system, a single air- conditioned instrument room containing microscopes, a small library, and an assortment of nets, diving gear, and other field equipment. Being a sponsored tenant in the proving ground — which in peak periods accommodated hundreds of scientists, technicians, and supp)ort per- sonnel — the laboratory enjoyed superb facilities for dining, housing, recreation, and medical care. During the 1950s, 1960s, and early 1970s, the labora- tory was operated on a part-time basis, with the active periods generally dictated by university class schedules. Thus, most investigators visited during the summer months and the periods of winter or spring academic holidays. Also during this period, visit authorizations were restricted to male U. S. citizens who had passed a security screen- ing. Travel to Enewetak from Honolulu was by military or military charter aircraft. The flight time from Honolulu to Enewetak was about 10 hours, usually with stops at John- ston Island and at Kwajalein and/or Wake Island. It is noteworthy that, despite considerable resistance to the invasion by women of what had been traditionally an exclusively male territory, arrangements were made to accommodate the eminent zoologist E. Alison Kay at the Enewetak Laboratory in December 1970. Her arrival sig- naled a new era in which the merits of the scientific research proposed were the only criteria for acceptance of a researcher at EMBL. Initially, the research emphasis at EMBL was toward the establishment of a reference collection of the local marine flora and fauna. This was accomplished by special- ists, who made extensive collections of particular groups of animals and plants, identified the individual specimens (including those new to science), labeled, cataloged and preserved them, and placed them in the laboratory collec- tion room. To complement the reference collection, a small library was established on site, providing convenient access not only to published references and texts but also to the works, both published and unpublished, of visiting investi- gators. Notices placed annually in the journal Science served to call this facility and its superb atoll environment to the attention of the community of marine scientists. This early research and subsequent publicity regarding the EMBL facility, combined with the availability of modest research grants, brought an enthusiastic response. From 1954 until this writing, 1028 scientists have worked at Enewetak, many returning for several periods of field col- lection and investigation. Notable was the response of tem- perate zone biologists who had not previously worked in the tropics. Entering the strikingly clear lagoon waters for the first time, with no more complex equipment than a face mask, was an exciting experience. Examination of a coral pinnacle, with its enormous diversity of organisms, brought a whole series of new dimensions to the work of these scientists. The limitations of the physical facilities and the remoteness of the EMBL field station were offset by an abundance of exciting research opportunities and vir- tual freedom from the pressures and distractions of cam- pus life. These features resulted in a level of scientific pro- ductivity unequaled in the experience of most researchers. A HISTORICAL PERSPECTIVE The original EMBL building eventually proved inade- quate to the needs of the scientists and in 1956 was expanded to include an extension for storage and a 4' X 20' concrete tank to hold experimental animals. Further expansion of the laboratory occurred in 1959 when Albert L. Tester of the University of Hawaii initiated a major program in shark physiology and behavior. For this program, two interconnected parallel tanks were con- structed, which allowed sharks to swim in an oval pattern. This facility permitted Tester and his colleagues to hold and condition sharks, to test their reactions to various chemical stimulae, and to elucidate some of the anatomical and neurological bases for their aggressive behavior. Nuclear testing activities at Enewetak ended in late 1958 with the declaration by President Eisenhower of a moratorium (accompanied by a similar Soviet moratorium) on all nuclear testing. The 1958 moratorium, originally a 1-year commitment, was actually continued until Sep*- tember 1961. At that time, the Soviets suddenly resumed testing at a high rate. Even then, however, the United States, in its response, did not return to testing in the Marshall Islands. Although the AEC continued to adminis- ter the Pacific Proving Ground until it was transferred to the Navy in 1960, AEC gradually withdrew activities and support on Medren until EMBL was the only active facility on that island. This made support such as power, water, housekeeping and messing, and logistics difficult. In 1961 EMBL moved from Medren to Enewetak Island where an active support infrastructure still existed. The laboratory's new home became a building on the lagoon side of Enewetak Island, previously used as a recreation center (Figs. 4 and 5). This building was modified to provide two small air-conditioned rooms for the protection of instru- ments and chemicals. A rectangular aquarium was con- structed in the center of the large main room which was enclosed on three sides and open to the lagoon. A sea- water system was installed, and living quarters were pro- vided for EMBL personnel and visiting scientists in a build- ing across the lagoon road from the laboratory complex. Although adequate, this facility had one imp)ortant draw- back. Boat operations required the use of the utility pier at the northeast end of the island, making loading and unloading difficult, and necessitating the carrying of equip*- ment and specimens between the pier and the laboratory. By 1969, another move was in order. In this same year, the directorship of EMBL passed first from Robert W. Hiatt to Vernon E. Brock, and then, a few months later, to Philip Helfrich. Helfrich continued as director until January 1, 1975. In 1969, military activities at Enewetak dictated another move for EMBL, this time to the vicinity of a large, three-story dormitory building which had been con- structed on the ocean side, toward the middle of Enewetak Island. The new location was a complex of aluminum build- ings, previously used as library, recreation center, and darkroom. This location was more desirable because of its proximity to sleeping quarters, food service facilities, and the boat launching ramp. In addition, it included a large, covered lanai — which was supplied with running seawater for aquaria — and two portable swimming pools used as holding tanks. With about twice the space that had previ- ously been allocated, the new facility included a large gen- eral laboratory, a shop, photo darkroom, library, equip- ment room, communications room, a dive locker, and a separate building for the storage of hazardous chemicals (Fig. 6). In the early 1970s, EMBL acquired its own com- munication system, providing a voice and teletype link to the University of Hawaii. MOVES TOWARD RESETTLEMENT The year 1972 brought significant fxjlitical develof)- ments which were to have a lasting effect upon the future of the people of Enewetak and upon the fortunes of EMBL. Political status talks had been going on for several years between the government of the United States and representatives of the people of the Trust Territory of the Pacific Islands (TTPI). These talks were aimed at ultimate termination of the United Nations trusteeship over the Micronesian Islands (with the United States as trustee) and the establishment of one or more new and independent self-administering political entities. During the 1972 talks, responding to the pleas of the people of Enewetak for the return of their home islands, the United States took the first steps toward that return. In April, Ambassador Hay- den Williams, the President's personal representative to the talks, was joined by High Commissioner Edward John- ston of the TTPI in a public statement of U. S. intentions. It provided that military use of Enewetak would shortly be completed, thus permitting the atoll to be returned to the administration of the Trust Territory, and that steps neces- sary to rehabilitate the islands for resettlement could then begin. Later in 1972, the AEC's Nevada Operations Office, using the resources of its national laboratories and contrac- tors, mounted a massive radiological survey of Enewetak Atoll as a preliminary step toward cleanup and rehabilita- tion. These activities are described in official reports (U. S. AEC, 1973; U. S. DOE, 1982; Holmes and Narver, 1973; and U. S. DNA, 1975). Although EMBL did not participate directly in either the 1972 survey or the cleanup, the director and other scientists consulted and assisted in many ways. While applied science and engineer- ing were at work to restore the atoll, the basic studies of EMBL continued apace. Although this tiny, remote research station might have been overwhelmed by the enormity of the cleanup effort (thousands of men, over 3 years, at a cost of more than $100 million), those respon- sible in the AEC (now the U. S. Department of Energy) and the U. S. Defense Nuclear Agency (DNA), recognized the lasting worth of the science program and saw to it that the laboratory's interests were protected. In 1978, the U. S. Coast Guard LORAN Station, which had occupied a complex of buildings at the eastern end of Enewetak Island, was closed. By agreement with DNA and with the p)eople of Enewetak, DOE obtained the HELFRICH AND RAY o z < -I CO LU LJJ z LU z o 0) < o o > GC O QC O m < -I vj\ Ui v^" \ > Yl \ O W \ m \ \ < \;*»?\ I v^M □. w^H < B^ \ d \ \ o w\ o 1- pl o I vP^l Q. ^^ z W\ Q fti'.'^ LU If ■^- « 1- ma? o Kr mI Q. IH'/n UJ f /'111 Q /M'/il Q // /'if 1 z (Plli 1 < jjL oi 1 _I //a/ //i/l M (fl /i£ /M 1/ //'In u. //^ i o / n/'//-' < // / //i/ri m /i / //'// / (T /// ll'iv < /// / / //'/^^ CO //^ n/B 1 UJ ^/^fi/ ///'k J o ^ / h ///'Jj /7 Q / Q( sM u/,'j|-J, // z (^ffl/ *l /if"^ (A/ \a ' /]/'/) 1 O \ V (J '/ 1 // z \Vn \'l~jl Q \\? /'/ // < ] V lI// I ^ Cfl i§ e u c o o a o ■a r e a Q. a bu '**§»ii A HISTORICAL PERSPECTIVE ■mm. ■ "^*^ Fig. 5 The second laboratory facility was located on Enewetak Island from 1961 to 1969. [Photo by E. S. Reese.] long-term use of these facilities and allocated them to the laboratory (Fig. 7). Over the next 2 years, in anticipation of the demobilization of the cleanup force and the sharp reduction in available logistics and life support facilities, steps were taken to make the laboratory ready to "stand alone." The complex was augmented with several portable housing and laboratory units, and plans were made for local power, fresh and salt water systems, and other needed support. The new location was a considerable improvement, consolidating all operational and support activities in one location. The new facilities included a main air-conditioned laboratory building with work benches and equipment space, a library, communications room, dark room, reference collection roorr and several storage rooms. Attached to the main building were a generator room and a storage shed. Four additional buildings pro- vided sleeping quarters accommodating as many as 18 per- sons. Other buildings provided a kitchen, food storage, a chemistry laboratory, a scientific shop, a dive locker, a general maintenance shop, and a covered seawater lanai. A 50-foot tower on which two 600-gal tanks were located provided gravity feed for a seawater system. Good quality unfiltered seawater for this system was pumped from a former quarry in the reef. Access to the lagoon for boats and personnel was pro- vided by a conveniently located concrete ramp and a wooden pier. Laboratory boats were moored offshore or launched and retrieved from trailers at the ramp. Fresh water was provided by catchment of rain from the roofs of several buildings and stored in four 10,000-gal cisterns. Diesel and gasoline fuels were stored in tanks on the lagoon side of the laboratory complex. These fuels, along with other supplies, were delivered to the laboratory approximately every 2 months by the DOE research vessel L\kianur, which was based at Kwajalein and supported DOE's environmental research, radiation protection, and medical programs in the northern Marshall Islands. Person- nel, mail, and light cargo were usually transported via the Airline of the Marshall Islands (AMI) on approximately a biweekly schedule and occasionally on a chartered flight. 10 HELFRICH AND RAY Fig. 6 The third laboratorv facility was larger and in a more convenient location on Enewetak Island from 1969 to 1978. The name was changed to the Mid-Pacific Marine Laboratory (MPML) to emphasize the broader research purview of the laboratory. [Photos by E. S. Reese.] A HISTORICAL PERSPECTIVE 11 B Fig. 7 The fourth and final location of the laboratory was in the former U. S. Coast Guard LORAN Station on Enewetalt Island from 1978 to the present; a. The dormitory is to the left and the mess hall to the right; b. View of the laboratory complex from the 50-ft-hlgh water tower with one of the cisterns in the foreground. The name was again changed to the Mid- Pacific Research Laboratory (MPRL) to note the inclusion of terrestrial as well as marine research. [Photos by P. Helfrich.] 12 HELFRICH AND RAY RESEARCH EMPHASIS There were two major periods of research at Enewetak conducted by the University of Hawaii under contract with DOE and its predecessors. During the first 20 years (1954 to 1974), the AEC supported independent research that was broadly aimed at increasing our knowledge of this rich and diverse coral atoll ecosystem. The rationale for suf)- porting this broadly based research was that it was impos- sible to predict what aspects of the system might be most perturbed by the test activities or what the lasting effects of these perturbations might be. Thus, a broad spectrum of investigations was considered appropriate. In retrosp>ect this was a wise choice because later events and decisions depended upon information resulting from this early research. Scientists from EMBL, with their acquired data base, were frequently called upon for advice and assis- tance, especially during the period of preparation of the atoll for the return of the Enewetak people. The modest cost of maintaining and op)erating the laboratory over these years provided the AEC with a bargain in science because the support systems were in place for AEC and defense department programs. The incremental cost of supporting the laboratory was, therefore, relatively small. The scien- tific research was accomplished at low cost because most of the participating scientists were salaried by their home institutions. Much outstanding research was accomplished at EMBL (Fig. 8). The record of accomplishment is set forth in the volumes of collected reprints of scientific publications which were issued in 1976 and 1979 (U. S. ERDA, 1976; U. S. DOE, 1979). As knowledge of coral reef ecosystems advanced, it was deemed advisable to mount a major effort to understand the metabolism of an entire atoll (Fig. 8). Discussions and planning conferences culminated in the initiation of a major program in the summer of 1971 under the name SYMBIOS. This program lasted for 12 weeks and involved the research vessel Alpha Helix, 25 participating scientists, and numerous support (>ersonnel under the leadership of Robert Johannes. SYMBIOS was jointly sponsored by the National Science Foundation, the AEC, and the Janss Foundation. Its initial objective — to study the metabolism of an entire atoll — proved to be too ambitious, but a thorough study of the windward reef was accomplished and some major advancements were redized in our knowledge of reef metabolism. As with other research, this effort posed many new questions and chal- lenges, and resulted in repeat visits to Enewetak by SYM- BIOS scientists to further pursue work initiated in this landmark study. The results of SYMBIOS are summarized in Chapters 9 and 10 of this volume. In 1972, the DNA began a series of studies to better understand cratering effects of nuclear explosions. Craters formed by the nuclear explosions of earlier years were analyzed by direct observation, seismic response measure- ments, and dynamic experiments utilizing chemical explo- sives. Scientists from EMBL were called UF>on to advise the defense department, especially upon the expected impact of their experiments on the marine environment. Later, fol- lowing a strong protest and legal action by lawyers for the people of Enewetak, the dynamic experiments were can- celed and only shallow coring of the atoll rim and seismic studies of the reef structure were pursued to complete this project. The second period of research began with the reorgani- zation of the laboratory in 1974. Following discussions with the Chairman of the Atomic Energy Commission, Dixie Lee Ray, a visit was made to the laboratory by an ad hoc advisory group, including officials and scientists from the University of Hawaii, the AEC, and several indep>endent consultants. Chairman Ray had expressed an interest in reorganizing and upgrading the laboratory to a full-time oF>eration, with research objectives more directly relevant to AEC interests. The advisory group met at Enewetak in February 1974 and later made brief visits to Bikini and to Majuro, the capital of the Marshall Islands. Participants were William O. Forster, Nathaniel Barr, and Charles Osterberg of AEC Headquarters; Roger Ray of the Nevada Operations Office of the AEC; Philip Helfrich of the University of Hawaii prector of EMBL); William R. Coops of the Research Corporation of the University of Hawaii; Robert Hiatt of the University of Alaska (first Director of EMBL); and Glen Fredholm, an independent consultant. The advisory group: (1) articulated in some detail its recommended objectives for a laboratory agenda which would be responsive to AEC direction, (2) suggested that the field station at Enewetak be up>graded to full-time activity with a resident staff, and (3) recommended that the name of the laboratory be changed to the Mid-Pacific Marine Laboratory (MPML) to reflect its interest in a wider geographical area, including such areas as Bikini, where the AEC continued to have an active interest. In March 1974, following the advisory group meetings, Roger Ray and Philip Helfrich returned to Majuro to meet with officials of the government of the Marshall Islands and with members of the Enewetak Municipal Council. The latter meetings were hosted by Micronesian Legal Services Corporation, counselors for the people of Enewetak. The Enewetak Council expressed its desire that the laboratory continue to function in the Enewetak community after the return and resettlement of the atoll residents. It approved the site of the Coast Guard LORAN Station as the ulti- mate home of MPML. With the approval of reorganization and redirection of goals, the laboratory entered a new and productive phase. Support and encouragement of basic studies continued under AEC sfxjnsorship, while mission-oriented research was being planned and implemented. The major AEC- oriented projects of the 1975 to 1980 period were (1) a study of the circulation of the Enewetak Lagoon, (2) research on the d^ amies of groundwater resources of Enewetak Atoll, and (3) studies of ciguatera fish poisoning at Enewetak. On Jan. 1, 1975, Philip Helfrich left the University of Hawaii and was replaced as director of MPML by Stephen V. Smith, who served in that capacity until 1977. During A HISTORICAL PERSPECTIVE 13 flHHMl ^s^luilhii^ ai^Bn^^K^Blti ■L ■':,''J4--v-. ^ mmy^'-^fmA ^^ . *^ # C 2m^J •" f^wurV^- 3rl^^ 0^ii^^^ lik.ll^H '.J ■ •■ ' ■ ^hC 7 •".* •k'h %^' ,^^u y^^L-^ f . -a^ < u a b i o I •s c3| II o 3 11 5 § •o » *■ E SI 1 t3 ?=§ I E J3 OB is W CD 00 "S a 14 HELFRICH AND RAY Smith's tenure the three research projects mentioned above dominated the activities of the laboratory. A study of the oceanography of Enewetak Lagoon was prompted because — despite intensive studies of various facets of Enewetak's geology, physiography, biota, ecosystem dy- namics, radiation contamination, etc. — only cursory infor- mation existed on the circulation patterns of the lagoon (Chapter 5 of this volume). This comprehensive study directed by Smith resulted in information on the physical and chemical dynamics of the entire lagoon. The topic of the second investigation was the dynamics of groundwater resources of Enewetak, a study that developed information vital to the returning Enewetak people who required uncontaminated water for drinking and agriculture. This investigation was directed by Robert W. Buddemeier (Chapter 4 of this volume). Ciguatera fish poisoning, the topic of the third study, had plagued the people of the Marshall Islands for many years, waxing and waning in an inexplicable manner. The return of the people and their dependency on fish for sustenance placed a special urgency on the results of this study that was directed by John E. Randall (Chapter 7 of this volume). During 1975, the AEC was reorganized, and the func- tions pertinent to MPML were assigned to the newly formed Energy Research and Development Administration (ERDA). In turn, ERDA gave way to the U. S. DOE in 1977. Resident managers were established at MPML on a year-round basis in 1975, and these individuals became integrated into the Enewetak community. This was an important aspect of MPML's operations because these scientists represented a benign, if not benevolent, element among the numerous government-sp>onsored activities related to the radiological survey, cleanup operations, and various medical and agricultural programs. The individuals who served as the resident laboratory managers were all exemplary in their dedication, and there were numerous examples of extraordinary service. From 1975 to 1977 the resident laboratory managers were Philip and Janet Lamberson. In June 1977, Ernst S. Reese assumed directorship of MPML, replacing Smith. During Reese's tenure (1977 to 1979), the research on lagoon oceanography, groundwater dynamics, ciguatera, and other aspects of atoll research continued. Planning and implementation of the move to the former Coast Guard LORAN Station took place. In addi- tion to continuing to fully support the research mission of MPML, the laboratory personnel cooperated in many ways with the DNA. A highlight of this coojseration was the pro- duction of an audio-slide presentation to acquaint the mili- tary personnel of the DNA with the natural history of a coral atoll and to describe the recreational opportunities offered by the atoll environment. There was also a caution- ary note about the dangers of the atoll environment rang- ing from severe sunburn to the presence of sharks. The audio-slide presentation contained an important message about conservation of the atoll environment as well: observe and enjoy but do not destroy. Following the cleanup, support services were with- drawn, and the laboratory was placed on a "stand alone" status, having to provide for all of its own life support and laboratory operations needs, with resupply from infrequent supply ships and light aircraft. During this challenging period, Reese was ably assisted by Victor R. Johnson and Maridell Foster and by several capable resident laboratory managers: Paul M. Allen, Michael V. DeGruy, and Gary Long (1977 to 1979). In 1979, Patrick L. Colin and John T. Harrison (1979 to 1983) took over the operation of the laboratory. Throughout this period the laboratory contin- ued to accommodate a few visiting scientists as transporta- tion and logistics could be arranged. In 1979, with the cleanup of Enewetak nearing comple- tion and the return of the atoll's residents imminent, a workshop was held at the Asilomar Conference Center, Monterey, Calif., to consider the future role of the labora- tory and its relationship to the other DOE scientific pro- grams in the Marshall Islands. The DOE headquarters sponsor at that time was the Division of Biomedical and Environmental Research under the direction of Helen M. McCammon. The fXDE policy enunciated at this time sig- naled the ultimate phase down of the laboratory over the following 2 to 3 years and the determination that signifi- cant effort should he devoted to synthesizing the research product of the laboratory's entire history into a publishable work. The present volumes are the result. It was decided also that, to the extent that the laboratory continued active research programs during the phase down years, these should not be confined to the marine environment. This latter decision was reflected in yet another name change: MPML became MPRL, the Mid-Pacific Research Labora- tory. In 1980, soon after the Asilomar meeting, Helfrich again assumed the directorship of MPRL. For most of the time between 1977 and 1980, a large, joint military force was at Enewetak — with a peak popula- tion of about 1000 drawn from the Army, the Navy, the Air Force, civilian government agencies, predominately DOE and civilian contractors. Research at MPRL continued through this period and in some ways the laboratory thrived upon the ready availability of logistic support, espe- cially frequent and dependable airlifts, and a generally har- monious relationship with the joint cleanup command. In fact, through the cleanup years, the resident manager of the MPRL facility met daily with the Joint Task Group Commander and his staff to discuss mutual interferences and mutual supp)ort. Many interesting aspects of the cleanup effort required an intimate knowledge of the atoll system, and the laboratory was often called up)on for con- sultation and advice. Selection of a suitable site for lagoon disposal of debris, protection and exploitation of food resources, and the preservation of scientifically valuable artifacts were but a few examples. On one occasion a major earth-moving eftort was planned for an island which had unexpectedly become a nesting ground for a very large flock of migratory birds. The laboratory's data base facilitated an immediate assessment of the length of time these birds would require protection, and it was possible to A HISTORICAL PERSPECTIVE 16 reschedule the cleanup activities so as to have only a minimal effect upon them. The atoll rehabilitation program consisted of the re- moval and disposal or isolation of debris and contaminated materials, the construction of homes and community build- ings and facilities, and the planting of more than 30,000 coconut, pandanus, and breadfruit trees. The cost was over $100 million. In April 1980, a ceremony was held at Enewctak, commemorating completion of the cleanup and the return of 543 Enewetak people to their ancestral home. A short time later, the last elements of the Joint Task Group departed Enewetak, leaving the laboratory as the only American presence in the community. Over the next 3 years, major emphasis was placed upon studies of a portion of the atoll ecosystem which had until then been largely unexplored — the soft lagoon sub- stratum. This research was directed by Patrick L. Colin. Much of the fallout material which remained from the nuclear tests had settled in the lagoon floor, and the dynamics of this biotope were little understood. As a result of this research, a fresh perspective was acquired. What had formerly been considered to be a largely passive sys- tem into which materials were sedimented from the water column was revealed to be an area in which burrowing organisms were continually reintroducing material into the water column — a process which led to some revision of the understanding of important biogeochemical processes. Interest in these processes helped to stimulate interest, in 1981, in one more interdisciplinary initiative at Enewetak. A significant improvement in understanding of the deeper sediments of the lagoon required direct observation and sampling, and these techniques required the use of a research submersible. With the cooperation of the Hawaii Undersea Research Laboratory, the research submersible Makali'i was made available for a period in the summer of 1981 (Fig. 9). Other sponsors of the expedition were the National Oceanographic and Atmospheric Agency (NOAA) and the DOE. The DOE support included use of the research vessel Liktanur. Fifteen scientists and seven sup- port personnel participated in a program which included Fig. 9 The research submersible Makali'i operated by the University of Hawaii shown on one of its 53 research dives in the Enewetak Lagoon in the summer of 1981. [Photo courtesy of HURL Program, University of Hawaii.] 16 HELFRICH AND RAY 52 successful research dives between July 7 and Sept. 29, 1981. The results were presented in a special symposium of the Western Society of Naturalists in Los Angeles in December 1982 and were published in Bulletin of Marine Science (Harrison, 1985). AN ERA ENDS Although the plans for an autonomous laboratory after the 1980 departure of the cleanup forces were thought- fully and thoroughly prepared and enthusiastically carried out, and despite the welcome that MPRL had received from the returning Enewetak community, its anticipated position as a permanent fixture in that community was not to be. At a time of constrained research dollars in the DOE, and with support grants from all sources limited, the cost of maintaining a resident staff and operating the MPRL facility as a self-sustaining field station became prohibitive. Support from the Division of Biomedical and Environmental Research was terminated in 1982, whereupon [X)E's Nevada Operations Office sought and obtained funding for one more year through the DOE Ofiice of Defense Programs. This additional year of fund- ing permitted an orderly phase down of the laboratory activities and the preservation of some of MPRL's unique assets. The reference collection which had been started during Hiatt's early tenure had grown and had been well preserved and cataloged. For several years this was accomplished through a contract with the Bernice P. Bishop Museum, under the able sufjervision of the late Dennis M. Devaney. The collections were carefully pack- aged and shipped to Hawaii to be placed in the temporary custody of the Bishop Museum. Early in 1985, negotia- tions were completed by the DOE with the Smithsonian's National Museum of Natural History and with the Bishop Museum for the permanent transfer of the reference collec- tion to the latter institution. The MPRL's library and much of the laboratory equipment were transferred to Hawaii Institute of Marine Biology. The remaining U. S. government activity at Enewetak is now conducted on a campaign basis, usually supported by the research vessel Liktanur. At this writing, however, two [X)E contractor employees remain at the atoll, and the field station remains intact and capable of limited sup)- port. Philip Helfrich retains the title of Director of MPRL and, with modest funding from DOE, entertains inquiries from scientists who desire to explore the feasibility of con- tinuing studies at the atoll. There is every indication that the people of Enewetak would welcome such visits. ACKNOWLEDGMENTS The wisdom and foresight of H. Burr Steinbach and Robert W. Hiatt and of those in the Office of Naval Research and the AEC who spawned and nurtured the idea of a research facility at Enewetak deserve special note. Time has proven that the decisions to establish, maintain, and support EMBL and its successors were wise and fruitful commitments which resulted in important con- tributions to our knowledge of atoll ecosystems and more broadly to marine science. Assuredly, there are still many unanswered questions, but just as surely new knowledge will continue to be built up>on the foundation of about 250 published scientific papers which have resulted from research conducted at Enewetak Atoll over the past 30 years. The writers of this chapter, who have been partners in the administration and support of the laboratory for almost half of that period, record their hop)e that new ways will be found by interested scientists and their spon- sors to continue, even on a limited scale, the exciting and rewarding experience of research at this remote and iso- lated atoll. REFERENCES Harrison, J T III, 1986, Recent Marine Studies at Enewetak Atoll, Marshall Islands, Bull. Mar. Sci.. 38: 1-3. Mines, N. O., 1962, Prouing Ground: An Account of the Radiobiological Studies in the Pacific, 1946-1961, University of Washington Press, Seattle. Trust Territory of the Pacific Islands, Enewetak Atoll Master Plan, 1975, 3 volumes. Holmes and Narver, Inc., Anaheim, Califor- nia. U. S. Atomic Energy Commission, 1973, Enewetak Radiological Surve^i. 3 volumes, Nevada Operations Office, Las Vegas, Nvauo. U. S. Defense Nuclear Agency, 1975, Environntental Impact Statement: Cleanup, Rehabilitation, Resettlement of Ene- wetak-Marshall Islands, 4 volumes, Washington, D.C. U. S. Department of Energy, 1979, Mid-Pacific Marine Labora- tory Contributions, 1 volume, Nevada Operations Office, Las Vegas, NVa628-l. — , 1982, Enewetak Radiological Support Project, Nevada Opera- tions Office, Us Vegas, NVO-213. U. S. Energy Research eind Development Administration, 1976, Eniwetok Marine Biological Laboratori/ Contributions, 3 volumes, Nevada Operations Office, Las Vegas, NVO-628-1. Chapter 2 History of the People of Enewetak Atoll ROBERT C. KISTE Director, Pacific Islands Studi/ Program Uniuersify of Hawaii, Honolulu, Hawaii 96822 INTRODUCTION The names of Enewetak and Bikini Atolls are linked in history, and they are well-known around the world because of their use as nuclear test sites by the United States. Indeed, once the atolls became available as research sites, a vast amount of research resulted; this volume is just one of the results. Most of the research has been in the biologi- cal and physical sciences, and the sheer volume of it has tended to obscure a very important fact — Enewetak and Bikini could be used for nuclear and other research pur- poses only after their indigenous human populations had been moved elsewhere. Much less is known about the people than about the flora, fauna, and physical properties of their atoll homelands. This chapter focuses upon the people of Enewetak. It examines their history, the struc- ture of their culture and society, the ways they have coped with the colonial powers that governed the islands, and their response to their resettlement on Ujilang Atoll. Some mention is necessarily made of the Bikini community because the histories of the two peoples are intertwined. Data about the Enewetakese are mainly derived from the research of four anthropxjlogists, all of whom worked with the p>cople after their relocation. Jack A. Tobin was the first. He served as Marshall Islands District Anthropolo- gist between 1950 and 1957. He resided with the Enewetakese on several occasions, and portions of this work resulted in his doctoral dissertation (Tobin, 1968). In 1964, Leonard Mason and I spent several months on Ujilang, and during the academic year 1972-73, I was involved in a legal suit (to be discussed later) which involved the Enewetakese and the U. S. Dejjartment of Defense (Kiste, 1976). More recently, a younger anthrof)ol- ogist, Laurence Carucci, spent 1977 and 1978 with the Enewetakese, and he too produced a doctoral dissertation (Carucci, 1980). THE ANCIENT PAST The research findings of prehistorians and linguists indi- cate that the Marshalls and other islands of Micronesia were settled by peoples who migrated from the general area of island southeast Asia into the insular Pacific many centuries ago (Bellwood, 1979). Indeed this particular migration probably began about 5000 years ago. Reflect- ing the ancient migration patterns out of island southeast Asia, the Marshallese language belongs to the large Aus- tronesian (also known as the Malayo-Polynesian) language family which is spread from Madagascar, through southeast Asia and across Micronesia, Polynesia, and many regions of Melanesia. Exactly when the early migrants arrived in the Marshalls is not known. The earliest archaeo- logical date currently available for the Marshalls is from a site at Majuro Atoll which was occupied at the time of Christ. In all probability, future archaeological research will push the date for the settlement of the Marshalls further back in time. No archaeological research has ever been conducted at Enewetak Atoll, however, and it seems safe to assume that remains of the past once deposited in its soil were obliterated with the preparations for and by the nuclear test program. The Enewetakese, however, have their own version of the distant past. According to their oreil litera- ture, they had always lived on Enewetak. In their own words: "We were there from the beginning." At the same time, their legends also recount how at least some of their ancestors purportedly came from Bikini, Ujac, Wotto, and other atolls also located in the northern Marsheills (Tobin, 1968). Regardless of the time of the settlement of Enewetak, two things are certain. Enewetak Atoll is isolated, and once the ancestors of the current population were in place, they had relatively little contact with other communities. As a consequence, the language and culture of the Enewetak people became differentiated from those of other Marshallese, and the people did not identify them- selves with the others. Indeed, they thought of themselves as a people who were separate and unique, "the people of Enewetak Atoll" as opposed to the islanders in the rest of the Marshallese archipelago. 17 18 KISTE The contact that the Enewetakese had with others, lit- tle as it was, was not limited to the Marshalls. The oral accounts associated with genealogies relate that some Enewetak people, mainly males, occasionally sailed to the south and west, contacting the ancient population of Ujilang (included in the Marshalls) and on to the high vol- canic and culturally and linguistically different island of Ponape. Contact with Ponape was to continue well into historic times and up until World War II. Long before the advent of Europeans, the people of Enewetak had developed a culture which represented a good adaptation to the limited atoll environment which is quite restrictive when compared to the high volcanic islands of the Pacific. The people were skilled navigators (an art which has been lost with the availability of travel on the vessels of foreigners), and they were expert builders of outrigger sailing canoes which were among the largest in the entire Marshalls. (Well into the 1960s, the Enewetak people were still constructing canoes that measured over 55 feet in length with masts that soared 30 feet above the vessels' decks.) In the relatively dry northern Marshalls and with the poor soil of the northern atolls, terrestrial resources were quite limited. Subsistence resources from the land were limited to coconuts, pandanus, papaya, bananas, and arrowroot. One or two breadfruit trees produced poorly. None of these crops required much care, and the people were very casual in their attitude about their maintenance. A similar attitude was evidenced regarding domestic animals. A few pigs and chickens were allowed to more or less fend for themselves, and their flesh was mainly reserved for holiday occasions. Thus, in part, ecological necessity had caused the Enewetak people to develop an economy which was heavily reliant upon marine resources. They knew the behavior and the monthly and annual movements of the large inventory of marine fauna. The fish of the lagoon and sea were caught, and expeditions were organized to collect shellfish, capture lobsters and turtles, and gather turtle eggs. In addition, several species of birds were also cap- tured as food resources. Shortly after the beginning of the German colonial era, old patterns were altered and the people became involved in the copra trade. Coconuts were converted to copra for cash and/or trade goods. Rice, flour, sugar, coffee, tea, canned meats and fish were eventually added to the diet. Several other features of the people's lifestyle deserve mention. Like most atoll dwellers, the people located their residences on the largest islands of their atoll. In the case of Enewetak Atoll, only the two largest islands were inhab- ited: Enewetak Island in the southeastern quadrant of the atoll and Enjebi Island on the atoll's northern rim. Although permanent residences were located on Enewetak and Enjebi Islands, the people were quite mobile within the atoll. Fishing and collecting activities penetrated every niche of the environment. Regular expeditions were made to all islands in the atoll to make copra and to col- lect food resources. Clearing brush and planting were done during these visits. Except for holiday seasons, it was not unusual for half of the population to be away from the two main islands as the p>eople dispersed in pursuit of a liveli- hood and for pleasure. Such expeditions broke the monot- ony of life on a small island and provided relief from one's fellows. SOCIAL ORGANIZATION Although the people had a collective identity as Enewetakese when juxtaposed to other Marshallese, they were divided internally into two separate communities that resided on Enewetak and Enjebi Islands. Community is defined as "the maximum group of persons who normally reside together in face-to-face association" (Murdock, 1949). Members of the two communities intermarried and cooperated in a variety of activities. Each functioned, how- ever, as a separate social and pwlitical unit, and its members had separate identities. The people of the Enewetak community called and thought of themselves as riEnewetak (the people of Enewetak Island) and those of the Enjebi community were riEnjebi, (the people of Enjebi Island). The traditional settlement pattern of both communities was dispersed. Residences were located on separate land parcels known as wato and were scattered along both sides of a sand and coral roadway which ran parallel to the length of the lagoon beach. In most cases, a uxjto was a strip of land which cut across the width of an island from lagoon beach to oceanside reef. They varied in size from about 1 to 5 acres. Each wato had a name, and the people who lived on Kabnene wato on Enewetak Island were sometimes referred to as riKabnene. The two communities had the same political structure. Each was headed by a hereditary chief known as iroij (Fig. 1). The chiefs directed the affairs of their respective com- munities, arbitrated disputes, and consulted one another with regard to concerns of the entire atoll and the total population's relations with outsiders (Fig. 2). In contrast to other Marshallese communities, which are organized around matrilineal principles, succession to the chieftain- ship was patrilineal, i.e., a man was succeeded by his eld- est son; the eldest son was succeeded by his younger brothers in the order of their birth; and when the last of them died, the eldest son of the eldest son succeeded. Like other Marshallese, the people of Enewetak Atoll were divided among several matriclans. The clans were named, and every individual automatically became a member of his or her mother's clan at birth. Clan member- ship could not be altered. The clans were vehicles for the provision of hospitality. One was obligated to protect fel- low clansmen and to provide them with food and shelter (Fig. 3). The clans were exogamous, i.e., members were required to marry outside of their clan. Members treated their clansmen as if they were parents or siblings, and sex within the clan was tantamount to incest. The preferred marriage partner was a real or classificatory cross-cousin HISTORY OF THE PEOPLE "19 ,.*«W J^ ■■■*?' '^^ Fig. 1 Iroij (Chief) Joannes Peter and his wife Bela. Ujilang Atoll, Nov. 17, 1976. [Photo by Janet Lamberson.] I Fig. 2 Luther, an Enewetak elder and a repository of tradi- tional cultural wisdom. Ujilang Atoll, Nov. 17, 1976. [Photo by Jcinet Lamberson.] (father's sister's daughter or mother's brother's daughter), and a very high percentage of marital unions were of the preferred type. Ideally, postmarital residence was patrilocal. A male took his bride to live on his father's land. Sometimes newlyweds lived with the man's parents, but the couple usually built a separate dwelling nearby. Quite commonly, a man and his married sons occupied adjacent dwellings but shared a common cooking house which was a separate structure. Thus, a patrilocal extended family was the most common family group located on a given wato. Another facet of Enewetak Atoll culture that differed from that of the rest of the Marshalls was the system of land tenure and inheritance. In contrast to the rest of the Marshalls where matrilineages (subunits within the matri- clans) constitute landholding corporations, the land tenure system at Enev,(etak Atoll was bilateral. In most cases, a married couple divided the land they had each inherited among their children, and a child usually received some land from both his or her father and mother. As the paren- tal generation died and as members of the next generation married and produced children, the process was repeated with parents allocating land among their offspring (Fig. 4). The people had an almost mystical attachment to their land, and their ties to it were deep. They could trace the history of their holdings back about a half-dozen genera- tions. As indicated previously, an individual's identity was, at least in part, defined by one's urate and one's island of residence. A final important social institution was an import. The people of Enewetak Atoll were the very last in the Marshalls to experience missionization because of their iso- lation and distance from the wetter, more richly endowed southern atolls where colonial powers always had their 20 KISTE Fig. 3 Aruo, a canoe builder and sailor, was lost at sea at Enewetalc Atoll in 1983. [Photo was taken at Ujiland Atoll in 1977 by Janet Lamberson.] headquarters. Not until 1927 did a Protestant missionary arrive to bring fundamental change to the people's world view. The first missionary was an islander from Mokil Atoll in the eastern Carolines, and he was followed by another missionary from Kosrae. The outsiders did not remain long, however, because within a few years a member of the ri£neaieta/c community was trained to lead the spiritual life of the people. The church took firm root. As in most places throughout the Pacific, the pjcople fully embraced Chris- tianity. Its teachings were mixed with traditional beliefs about ancestral and nature spirits and other notions about the supernatural, and the result was a hybrid that had become an integral part of the local culture and society. Work and play were tabu on Sundays. Other church ser- vices were held during the week. Christmas and Easter were the major holidays of the calendar year. COLONIAL HISTORY The Spanish explorer Alvaro de Saavedra is given credit for the European discovery of Enewetak Atoll in 1529. After his initial contact, like many other islands and atolls in the Marshalls and Carolines, Enewetak was not visited again by Europeans for many decades. The next known sighting of the atoll occurred in 1792, and 2 years later another European vessel called. In 1798, Enewetak Atoll was mapped by a Captain Fearn in command of the Hunter (Tobin, 1968). Although contact with the outside world surely has made some impression on the people, it seems somewhat odd that no accounts of early Europ>ean visitors were found in the oral history of the people. In 1898, shortly after the Germans had declared the Marshalls to be a Protectorate, a German trading company contracted the Enewetakese to extend their plantings of coconut palms for the copra trade. Some of the people traveled to Ujilang Atoll to work on the copra plantation there under a German supervisor. German rule was brief, however, and no German or other outsider actually took up residence on Enewetak during German times. In fact, the people were still adjusting to the European interlopers when Japanese colonial rule replaced that of the Germans in 1914 (Kiste, 1977). Because they are much closer to Ponap)e Island in the eastern Carolines than the old colonial headquarters at Jaluit Atoll in the southern Marshalls, Enewetak and Ujilang Atolls were administered and visited by Japanese vessels from Ponape during Japanese rule. Consequently, the Enewetakese were separated even more from other Marshallese. It was also during Japanese times that the people lost some of their autonomy and lessened their con- trol over their land. Japan began its rule with a show of force by sending naval vessels to confirm Japan's author- ity. In the early 1920s, a Japanese trader established him- self on the atoll. He falsely claimed that the colonial government had granted him p)ermission to acquire land and develop coconut groves. He also claimed that the peo- ple were required to assist him with the venture. Initially the Enewetakese did not resist and worked for modest rewards in trade goods, but as they became more familiar with the Japanese, they realized they had been duped, and the two chiefs filed a complaint with officials. The issue was not resolved before the Japanese military began to for- tify the atoll in the late 1930s as part of the preparations that led to World War II. The war years brought tragedy. First, the Japanese constructed an airstrip on Enjebi Island and evicted the riEnjebi to a small corner at the eastern end of their island. The American invasion in 1944 devastated and practically denuded both the Enjebi and Enewetak Islands. Ten per- cent of the local population was killed. At the end, both communities were moved to two small islands in the east side of the atoll. The Americans constructed a large mili- tary base on Enewetak Island, and the people acquired their third colonial master. When the Americans asked HISTORY OF THE PEOPLE 21 Fig. 4 The Enewetak children represent the promise for the future. UJilang Atoll, Nov. 17, 1976. [Photo by Janet Lamberson.] them to abandon their homeland, the Enewetakese correctly concluded that they had no real alternative, so they offered no resistance (Kiste, 1977). THE UJILANG RESETTLEMENT Ujilang is 124 miles southwest of Enewetak. It had been inhabited by a Marshallese population, but in the late 1800s a typhoon decimated the atoll and killed all but a handful of its people, most of whom were moved to the southern Marshalls. Ujilang was then developed as a com- mercial copra plantation during the German eind Japanese eras, and as noted, some of the p>eople of Enewetak Atoll had experiences there as laborers during German times. Ujilang was abandoned during World War II, and thus it was available to receive a population. American authorities initially thought of Ujilang as a site for the relocation of the Bikinians. They were the first to be moved to make way for the nuclear tests. Their first relocation occurred in March 1946 when they were moved to nearby Rongerik Atoll. It had never had a permanent population of any size, and the reason soon beceime apparent. Rongerik's resources, greatly overestimated by American planners, were inadequate to support the com- munity. After considerable delay and many complications, the Americans decided to move the Bikinians to Ujileing, and in November 1947, an advance party of Bikini men and navy Sea bees arrived to construct a village. In less than 2 weeks, however, officieils in Washington, D. C. announced plans to use Eneweteik as a second test site, necessitating a relocation of its inhabitants. They were moved to Ujilang on Dec. 21. The Bikinians were eventu- ally resettled on small Kili Islemd in the southern Meu'sheills where they have never made a satisfactory adjustment (Kiste, 1974). Ujilang has only one sizeable island, 2tnd both the riEnewetak and riEnjebi communities were resettled there. The islemd was evenly divided by an Americiin naval offi- cer who ckllotted one half to each community. A rather compact village was constructed in the middle of the island, with the Enewetak and Enjebi people residing on their respective sides of the dividing line. No longer separated by Enewetak 's large lagoon emd with the more compact settlement pattern, the two groups became a sin- gle community while retaining their dual political structure. 22 KISTE The years on Ujilang were quite difficult. The atoll is much smaller than Enewetak and has correspondingly fewer resources. Enewetak has 39 islands with a total land area of 2.75 square miles; its large lagoon covers 387.99 square miles. In contrast, Ujilang has 21 islands which col- lectively constitute only 0.67 square miles (Holmes and Narver, 1975; Tobin, 1968). Compounding the problem of living on a smaller atoll with a greatly reduced resource base, the people, like other Micronesians, have rapidly increased in numbers. The total population at the time of relocation was only 141. By the early 1950s, the number had increased to about 170. By 1977, the population had reached 400 (Kiste, 1977). A census taken in 1978 reported 540 (Carucci, 1980), and today the number is probably in the vicinity of 600, a four-fold increase since relocation. Population pressures on Ujilang's resources obviously increased during the people's years on the atoll, and on numerous occasions, food supplies from the land were depleted. Coconuts that might have been converted into copra were needed for sustenance, and as a consequence, the people had little cash to purchase imports. The situa- tion was exacerbated because Ujilang is distant from the government center at Majuro, and ships carrying food and other supplies frequently failed to service the atoll. As a result, the people suffered considerable physical depriva- tion. For those who knew it well, memories of life at Enewetak brought despair, and younger people became convinced that they had been deprived of their true home where want was unknown. The desire to return to Enewetak increased with each passing year (Kiste, 1977). In spite of the adversities suffered and the periods of discouragement, the people always maintained a great sense of pride in themselves and a determination to control as much of their destiny as possible. EXiring the initial years of U. S. rule, the jseople sized up the Americans and attempted to determine the best ways of dealing with them. Until the mid-1960s, they tried to get help by mak- ing complaints and fjetitions to the administration. Welfare measures were occasionally implemented, but more often than not, the people's pleas went unheeded. During this period, the traditional political structure remained intact. The chiefs functioned in their usual roles, and as many traditional leaders elsewhere, they resisted American efforts to introduce Western political forms — in this instance, a council form of government headed by an elected magistrate. By the early 1960s, however, some change was observable. The two chiefs were by then older men. Some contemporary issues required that the decision-making processes be opened to include younger men who had attended American schools and/or had been employed by the administration. Meetings of adult males were occasionally held, and some decisions about commu- nity affairs were decided by a majority vote. In 1967, exceptionaDy poor conditions on Ujilang and a realization that previous pleas to the administration had largely been ineffective prompted the people to take a much more aggressive stance. After an absence of 6 months, a field trip vessel called. Much to the surprise of the official in charge, the people boarded the ship and announced their intention to abandon the atoll. A poten- tially dangerous voyage on an overloaded ship was avoided when the officials volunteered to remain on the atoll and "suffer from starvation" until the administration responded to the situation. The display of assertiveness produced results. Substantial amounts of food and other supplies were soon delivered, and the District Administra- tor of the Marshalls came to hear the people's grievances. The sit-in aboard ship and another threat to abandon Ujilang a year later had the greatest support from younger adults. The sit-in also seems to have been linked to a major transformation in the community's political structure. Sometime during late 1967, the two chiefs had yielded to younger men. A magistrate and a council of 12 were elected. Reflecting the traditional division of the popula- tion, the riEnjebi and the riEnewetak each elected six coun- cilmen. The magistrate became the head of the entire com- munity; the council became its legislative body. The chiefs, however, continued to function importantly as advisers and men of substantial influence (Kiste, 1977). In 1968, the people evidenced considerable sophistica- tion about the larger world when they petitioned the United Nations for assistance in returning to Enewetak. In August, it was learned that Bikini was judged to be safe from radiation and that it could be returned to its people.' The news caused great resentment among the riEnjebi and riEnewetak, and they strongly protested their continued alienation from home. The protest produced results. In 1970, in an effort to satisfy the people, the United States Congress authorized a payment of $1,020,000 to the peo- ple of Enewetak. Other payments were to follow in later years. The initial attempt to placate the people was not suc- cessful. In late 1971, they announced their intention to return home before the end of the following year. Depart- ment of Defense (DOD) officials contended, however, that it was necessary for Enewetak to remain under DOD's con- trol. This was rejected, and by early 1972, the people obtained legal counsel from the recently created Microne- sian Legal Services Corporation (MLSC). The people then informed officials that they would institute legal action if Enewetak was not returned to them. On April 18, 1972, the long-awaited day arrived; it was announced that the U. S. would surrender Enewetak by the end of 1973 after certain "unspecified activities" had been completed there. The p)eople had won a major victory, but it soon became apparent that the "unsp>ecified activities" were a threat to their future well-being. The activities were part of *ln 1968-69 a cleanup was conducted at Bikini Atoll, and a residential complex was established About 140 Bikini people returned to Bikini in the early 1970s, but by 1978 it became apparent that the radiation content of foods grown at Bikini made permanent residence there inadvisable. The Bikinians were again removed from their atoll and, at this writing, have not yet returned. HISTORY OF THE PEOPLE 23 the Pacific Cratering Experiments (PACE) project and were sponsored by the DOD and related agencies. PACE had commenced with small explosions and was projected to culminate in several multiple ton detonations of high explo- sives and one final 500-ton blast. It was hoped that this series of experiments would help to provide a better understanding of many of the effects of the tests of the 1950s. The f)eople of Enewetak, represented by their MLSC lawyers, invoked the provisions of the National Environmental Policy Act, and they filed suit in the Federal District Court in Honolulu in September. At Ujilang, PACE scientists explained their project, claiming that it would cause no long-term damages. The people listened politely and responded with a brief but very firm statement. In essence, they stated: "PACE is evil, and we will do what- ever we can to prevent it." The magistrate gave an elo- quent speech which reflected the people's values and feelings. I do not know if you have made an attempt to compare your sense of values, you who live in America or else- where, with ours. You live with gold and money and we have to depend on land and whatever life we can find on land and in the water. Without these, we are nothing. We do not have to explain further that Enewetak, with what- ever land resources and whatever marine resources it has, is our homeland, and seeing that you understand this, we do not know why you continue to insist to do these things on Enewetak, when for us there is really nothing else to look forward to. For this reason we must continue to ask that you refrain from proceeding with this program. PACE is no good . . . Enewetak has undergone severe damage. There are islands that are missing. There is a considerable amount of land that has been destroyed. The question then comes: Has not Enewetak done enough for your testing? We do not know who you will take this message to — perhaps you will take it to Washington or the Depart- ment of Defense — but, the point still remains that we feel Fig. 5 Official ceremony returning Enewetak Atoll to its former inhabitants. Enewetak Atoll, Sept. 16, 1976. From left to right, seated at the table, are Oscar DeBurum, then District Administrator of the Marshall Islands; Binton Abraham, Iroij (Chief) of the liEnewetak, now deceased; Thomas Lacy, Brigadier General, U. S. Air Force, then Field Commander, Defense Nuclear Agency; Peter Tali Coleman, then Deputy High Commissioner of the Trust Territory of the Pacific and later Governor of American Samoa; Joannes Peter, Iroij (Chief) of the riEnewetak; Hcrtes John, magistrate of Ujilang Atoll. [Photo by Janet Lamberson.] 24 KISTE that Enewetak has done enough. We have sacrificed enough and PACE should not be continued because it only means further destruction of our homeland. [Office of the Judge Advocate Pacific Air Forces, 1973.] The legal suit was never brought to trial as the DOD cancelled the PACE project soon after the public hearings (Kiste, 1976). That the magistrate and not the chiefs spoke for the people reflected the changes that had occurred in their po- litical organization. By the time of the PACE affair, further change had occurred because the process of electing coun- cilmen had been altered. In elections subsequent to 1967, the 12 councilmen were elected from the population at large and not half from the Enewetak and half from the Enjebi sides of the community. It appeared that the old division between the two sides had lost some of its mean- ing. RETURN TO ENEWETAK After the PACE affair, the people exjjerienced some reversals. Radiological surveys revealed that some islands of Enewetak Atoll are more heavily contaminated by radioactive debris than previously thought, and they can- not be inhabited for decades to come. In 1976, after extensive radiological surveys, it was determined that Enewetak Island and several others on the atoll's eastern rim could be partially restored with reasonable safety. The U. S. Congress provided funds for their cleanup and reha- bilitation. The full-scale cleanup effort began in late 1977. The Enewetakese were consulted in the planning and some were employed to help with the work. The cleanup of Enewetak Atoll, the construction of dwellings and commu- nity buildings, and extensive replanting was completed in 1979, and the atoll was officially returned to the people in April 1980 (Figs. 5 and 6). The event was celebrated by virtually the entire papulation with 542 people attending. Fig. 6 Iroij Joannes Peter signing documents returning Enewetak Atoll to the liEnetoetak and riEngebt. Enewetak Atoll, September 16, 1976. [Photo by Janet Lamberson.] HISTORY OF THE PEOPLE 25 OTHER ISSUES Although the Enewetak case is unique, the people share some historical trends with other Micronesians. Like other islanders, the people of Enewetak have had to become familiar with the representatives of the successive colonial administrations. The Enewetakese had to learn the customs of the new foreigners and had to develop ways to cope with them. The initial years of American rule followed on the footsteps of World War II, and it was a time when memories were still fresh of the destructive powers that the U. S. had unleashed during its crushing defeat of Japan in the Pacific. Understandably, Micronesians were cautious and even timid in their dealings with Americans. With the passing of time, Micronesians everywhere grew bolder and became more skilled as they managed their relations with Americans. Encouraged by this relation- ship, Micronesians have modified their traditional institu- tions and have adopted more democratic p>olitical structures. In recent years, and very much like the people of Enewetak, they have become more assertive as they have negotiated for what they believe are their own best interests. Inspired by the general wave of decolonization in the Pacific, and as the end of the U. S. trusteeship draws near, Micronesians have been struggling to take control of their own lands and destinies. Self-government is coming to the U. S. territory, and it seems unlikely that situations such as those which occurred at Enewetak or Bikini will ever occur again. REFERENCES Bellwood, P., 1979, Man's Conquest of the Pacific, Oxford University Press, New York Carucci, L., 1980, The Renewal of Life: A Ritual Encounter in the Marshall Islands, unpublished Ph.D. dissertation, University of Chicago Kiste, R. C, 1974, The Bikinians A Studv in Forced Migration, Benjamin/Cummings Publishing Company, Menio Park, California — , 1976, The Peoples of Enewetak vs. the U. S Department of Defense, Ethics and Anthropology, M. A. Rynkiewich and J. P. Spradley (Eds.), John Wiley and Sons, New York. — , 1977, The People of Enewetak: Past and Present, Micronesian Perspectiue. 1(2): 18-23. Murdock, G. P., 1949, Scxial Structure, Macmillan Company, New York. Office of the Judge Advocate Pacific Air Forces, 1973, Transcript of Testimoniyi Enuironmental Hearings "Project PACE," Hono- lulu, p. 79. Tobin, J. A., 1968, The Resettlement of the Enewetak People, unpublished Ph.D. dissertation. University of California, Berkeley, pp 18, 22, 57. Trust Territory of the Pacific Islands, Enewetak Atoll Master Plan, 1975, 3 volumes. Holmes and Narver, Inc., Anaheim, California Chapter 3 Ph\;siograph\; of Eneivetak Atoll PATRICK L. COLIN Motupore Island Research Department (Jniuersity of Papua New Guinea Port Moresby/. Papua New Guinea LOCATION AND SIZE Coral atolls have been variously defined and, without considering unusual cases, can be described as more or less continuous reef (largely corals and other calcium car- bonate producing organisms), which surrounds a deeper lagoon and drops steeply to oceanic depths on the seaward margin. All islands are typically low, derived from reef rub- ble and sand. Enewetak Atoll conforms to all aspects of this description and in many respects is a "textbook" atoll. It has a large elliptical lagoon, approximately 41 islands on its rim, a few passages between the lagoon and ocean, and narrow shelves dropping steeply into deep water on all sides. The subsurface geology of Enewetak and Bikini have been extensively examined, and these results are reported in the U. S. Geological Survey Professional Papers 260 series. Enewetak Atoll is located in the northwestern Marshall Islands with its center at approximately 11°30'N; 162n5'E (Fig. 1). It is 220 km from the nearest land, Ujelang Atoll to the southwest; 310 km from Bikini Atoll to the east; and about 410 to 460 km from other atolls (Ujae, Wotho, Ailinginae, Rongelap) to the southeast to east. To the north occur Wake Island, about 1000 km northeast, and Marcus (Tora Shima) Island, about 1600 km northwest. To the west are the Marianas, the nearest being about 1700 km. All islands of the Marshall Islands are low, most being coral atolls. The high islands nearest to Enewetak are Ponape, to the southwest, and Kusaie, to the south, both about 580 km distant. Enewetak is a relatively large atoll, somewhat elliptical in shape, about 33 by 41 km in size, with the islands, reef flat, and lagoon covering about 1000 km . It is the third largest atoll in the Marshall Islands, exceeded by Kwajalein (the largest atoll in the world) and Rongelap. By world- wide standards, it is not exceptionally large. The majority of the area of Enewetak is the lagoon, with the reef flat and the islands covering progressively less area. Table 1 provides information on the area covered by various environments at Enewetak. WEATHER AND CLIMATE Weather at Enewetak is dominated by the surrounding marine conditions. Since all islands are low and of small area, they do not alter weather conditions by their pres- ence. The atoll is semiarid, with rainfall averaging only about 1700 mm per year, and has a distinct wet-dry annual cycle. Air temperatures are relatively high and very stable, with a mean annual temperature of about 28°C. Solar radiation is intense, and humidity is consistently high. At almost 12°N, Enewetak is within the trade wind belt with nearly consistent easterly winds. The atoll is sub- ject to tropical storms and typhoons at irregular intervals which greatly affect the marine and terrestrial environ- ments. The meteorology of Enewetak is discussed in Chapter 6 of this volume. ENVIRONMENTS OF ENEWETAK The Lagoon The lagoon is the largest component of the atoll. It is relatively deep by atoll standards, averaging about 54 m, with a reported maximum of 71 m. The lagoon bottom generally slopes from the lagoon rim toward the center. At a distance of 2 to 4 km from the rim, the lagoon bottom is essentially flat at a depth of about 45 m. Even the outer- most 2 to 4 km of the lagoon has generally low slope gra- dients on its bottom because of the horizontal distance required to reach 45 m depth. The only areas with signifi- cant slopes, except along the flanks of patch reefs and coral pinnacles, on the soft bottom of the lagoon occur shallower than 25 m. Below that depth, except for small- scale undulations, there is little variation in the soft bottom from the flat and horizontal. The area above 25 m depth is also affected by wave action and currents which can affect sediment distribution. Most of the lagoon bottom is relatively inaccessible to human observers. The depths are below those practical for sustained diving operations and, generally, must be observed or sampled remotely. The area of the lagoon bo' 27 28 COLIN s < iS u o291 3 PHYSIOGRAPHY 29 TABLE 1 Areas of Environments at Enewetak Atoll (in kilometers) Total atoll (land and shallow water 1022 less than 100 m deep) Total land 7.125 Total marine environment 1015 Total lagoon depth in meters 938 Oto 10 47 10 to 20 56 20 to 30 75 30 to 40 103 40 to 50 253 50 to 60 310 Over 60 94 Outer reef slope est. 13 Reef flat (less than 1 m at low tide) 64 torn visible in aerial photographs is limited to depths of 15 to 20 m and is usually located only on the rim of the lagoon. The only structures which are visible from the sur- face in the central lagoon are coral pinnacles which reach within less than 15 to 20 m of the surface. There arc two major channels between the lagoon and ocean (Figs. 2 and 3). The first is the "deep" channel, between Medren and Japtan, which is nearly 60 m deep in places but is relatively narrow. It averages only about 1.4 km in width between Japtan and Medren, but the deepest portion (below 40 m depth) is only about 600 m wide. During tidal changes, swift currents flow in and out of this channel It is exposed to the easterly swell from the ocean and allows such swell to enter the lagoon in its vicinity. The swell, combined with wind-produced chop due to the open fetch of the channel and currents flowing out of the lagoon (counter to the wind direction), often produces extremely rough conditions in the channel. The deep channel splits into two branches just west of Jedrol Island leaving an area of shallow reef in between with minimum water depths of about 6 m (Fig. 2). This wedge-shaped reef gradually deepens both to the west and north until it essentially merges with the lagoon bottom. Near its easternmost extremity, a ferro-cement barge — the "Concrete No. 9," locally called the "cement ship" — ran aground, resulting in a distinctive marker of this site. The bottom slopes away at about a 45° angle into the branches of the deep channel which begin to flatten out at about 40 m depth. The second major passage, the "wide" channel, is located at the south end of the atoll between Enewetak and Ikuren. It is no more than 15 to 18 m deep but stretches 10 km between the islands (Fig. 3). Since it is considerably shallower than the deep lagoon bottom, it resembles a sill. The currents in its vicinity are essentially unidirectional, out of the lagoon (Atkinson et al., 1981), but their speed is determined by the tide. Although the wide passage does not directly face the ocean swells, the swells are refracted somewhat around the southern end of Enewetak Island and enter the lagoon through this open- ing. This, combined with waves from the lagoon and the Fig. 2 Aerial view of the deep, narrow channel entrtuice to the lagoon between Medren and Japtan Islands on the eastern, windward side of the atoll. [Photo by P. L. CoUn.] 30 COLIN Fig. 3 Aerial view of the wide, south channel passage to the lagoon looking from Enewetak Island (lower right) to Ikuren Island (upper left). The shallow bottom of the sill at the passage is visible. [Photo by P. L. CoHn.] shoaling nature of the bottom at the wide passage, produces rough conditions with standing waves and steep waves in the western half of the wide passage. A series of shallow open passes with fingers of emer- gent to near emergent reef intersjjersed between them is called the "southwest passage," an additional passage between the lagoon and ocean. These openings cover about 6.7 km of the atoll margin from the island of Biken to the beginning of unbroken shallow reef to the southeast. The sand-bottomed passes appear deeper to the south — as much as 8 m deep in places. While significant, the southwest passage is p>erhaps an order of magnitude less important in lagoon-ocean water transport than the deep and wide channels (Chapter 5 of this volume). The reef flat is also a major source of water movement into or out of the lagoon. The amount of such transport is dependent on the height of the tide and the wind and waves which influence the wave pumping of water from ocean to lagoon. Where islands disrupt the free flow of water across the reef flat into the lagoon, water flow is channeled into narrow, deeper areas where current speed can be relatively high. These channels are variously termed "rips" or "gutters" and can also occur on intraisland reef flats where there are areas of higher current flow. The biological communities and environments of the lagoon are discussed in Chapters 7 and 8 of this volume. They are quite variable from place to place, varying from sediment-bottomed areas devoid of hard substratum to well-developed coral reefs. The diversity of plants and animals is as high in the lagoon as it is in other areas of the marine environment. Emery et al. (1954) reported over 2000 "coral knolls" in the lagoon with some suggestion that they "belong to 2 distinct size categories; nearly all the large coral knolls have a diameter in excess of 1 mi whereas nearly all the rest are smaller than Vi mi, and intermediate sizes are not common." Most of these do not reach sufficiently close to the surface to be visible and can be detected only by echo sounding. Emery et al. (1954) distinguished between the term "coral pinnacle" and "coral knoll," preferring the latter term, but did not clarify how the reef structures of the lagoon margin were considered. In essence an inter- grading series of reef structures exists within the lagoon. Although distinct types — such as coral knolls (broad, rela- tively low structures), coral pinnacles (high relief relative to diameter), and patch reefs (small structures, often in sheil- low water) — can be identified, intermediates are common. Those reef structures that are present on the bottom and visible from the air are generally, in this treatment, con- sidered to be "patch reefs." The Reef Flat The shallow reef flat, much of which is emergent at low tides, around the rim of the atoll has been the most intensively examined marine environment. It consists of areas of rock pavement with seaward algal ridge structures and lagoonward rubbly bottom. The reef flat varies consid- erably in different areas of the atoll, particularly between the windward and leeward sides but also over relatively short distances on the windward shore. Very little of the algal ridge, normally produced by coralline algae, is "live" at Enewetak. Instead of the healthy pink corraline areas PHYSKDGRAPHY 31 which have the characteristics of typical algal ridge struc- tures, they are covered with fleshy algae. Indications are that these areas were live algal ridges sometime within the relatively recent past, but whether man has played a role in their demise is uncertain. There is one small area of live algal ridge still present at Enewetak, near the island of Ananij, which occurs at the easternmost extension of the reef flat. This and the ecology of the reef flat are discussed in subsequent chapters. The Seaward Slope The seaward slope from the reef flat to the dropoffs to depths over hundreds of meters is narrow all around Enewetak. The edge of the seaward slop>e is marked by "spur and groove," alternating reef and rubble fingers projecting seaward where the waves break. On leeward reefs, there are no distinct spur and groove formations but a deeper series of promentories and reentrants in the upper 15 m. On windward reefs, a rock bottom then slopes away gradually to a break point at the 18 to 30 m depth where the bottom begins to slope much more steeply. Oceanic depths are quickly reached. The width of the seaward shelf varies around the atoll. It is widest off Enewetak Island, being about 400 m wide. Other areas of the windward reefs are narrower so that it is only 100 to 200 m wide on the northeastern reefs between Lojwa and Enjebi. On leeward shores the shelf is very narrow, only a few tens of meters wide. It is literally possible to stand on the reef flat and throw a stone into depths of 100 fathoms. The Islands There arc approximately 40 islands at Enewetak, excluding a few small sand islands remaining above water at high tide. Two islands (Elugelab and Lidilbut, not shown in Fig. 1) were vaporized by nuclear testing, and three oth- ers were so severely altered that only small remnants remain (located in the northwest part of the atoll). The vegetation of most islands at Enewetak has been progres- sively and increasingly altered compared to the nondis- turbed state. The alterations occurred initially by the estab- lishment of coconut groves, later by wartime construction and damage, and finally by nuclear testing activities and subsequent military activity. Extreme alteration occurred again during the Enewetak cleanup with the aim of re- establishing the coconut groves. The only islands which have essentially undisturbed vegetation are the five south- western islands and Biken (see Chapter 11 of this volume and Chapter 3 of Volume II for further details). The islands can be grouped into several reasonably natural units defined by significant gaps between units and identified by direction location. Often these units are identi- fied by compass location and are defined here. The "southwest islands" of Enewetak are the five islands from the southerly Kidrenen through Ikuren. They are separated by both distance and intervening passes from isolate Biken, which will be termed the "western island," and from Enewetak Island. The islands of Enewetak, Bokandretok, and Medren are called "south- eastern islands"; they share a common reef flat and are separated from all others by passes. The ten "central islands" are those of the windward side from Japtan through Runit, including Jedrol. The "northern islands" are those 15 to 16 islands from Bijire or Billae through the northern Boken. They are separated from the central islands by several kilometers of op)en reef flat. The last group, the "northwest islands," from Bokoluo to Luoj, are separated from the northern islands by the large MIKE and KOA craters and consist of four islands and one sand bar. The islands consist largely of coral sand, rubble, and boulders with areas of exposed beach rock and reef flat pavement. In certain areas, large quantities of cement debris are incorporated among coral boulders and rubble. All the islands are low, the highest elevation being approxi- mately 4 m on Enewetak Island. Beaches occur on many lagoon shores, the most extensive and continuous today found on Medren. Enewetak Island in pretesting days pos- sessed an apparently continuous beach, but the lagoon shore has been so altered by the construction of seawalls or by the dumping of riprap that sand beaches occur only in short stretches today. The ocean shore of islands on the windward side of the atoll facing the reef flat often have alternating beach-beach-rock shores. Sand beaches here, however, do not extend below the intertidal, merging with the reef flat or rock which extends offshore. Vegetation occurs above the high tide line on all shores. There are no mangroves or mangrove-like terres- trial plants extending into salt water at Enewetak. With the exception of Biken and the five southwestern islands, the vegetation has been extremely altered. Enewetak, Medren, and Japtan Islands are residence islands. Houses were constructed and other buildings were converted during the Enewetak cleanup. These three residence islands, plus Ananij and the islands from Billae through Aej, were planted with coconut palms between 1978 and 1979. Coconut palms have not been planted on Enjebi, the second largest island of the atoll, except for an experimental garden plot that was established in 1975 by Lawrence Livermore National Laboratory and that contains coconuts, pandanus, and breadfruit. The soil of Enewetak Atoll islands consists of little more than calcium carbonate sand and rubble (Chapter 11 of this volume). This material has virtually all its origin from the sea and is derived from corals, calcareous algae, foraminifera, and a wide variety of organisms producing smaller amounts of carbonate materials. Occasionally, pieces of pumice which have drifted to Enewetak are found near beaches. More rarely, noncarbonate rocks, car- ried by rafting debris such as fallen trees, are found. Enewetak soils have very little organic matter or nutrients. This is particularly true for the highly disturbed islands where human activity has eliminated the normal ground cover of vegetation and nesting birds. On normally vegetated islands, a limited amount of organic materia! is 32 COLIN tied up in leaf-litter on the soil surface, but relatively little is actually found in the soil. The larger islands of the atoll have good freshwater lenses beneath them. All the islands are quite low, so the water table lies very close to the ground's surface. It is not necessary to drill more than about 3 m deep to hit water. The groundwaters of Encwetak have been studied in some detail (Chapter 4 of this volume). MAN-MADE FEATURES Quarries Areas of reef flat adjacent to several islands at Encwetak were quarried or excavated for building or road construction purposes. A single quarried area is at the rKjrth end of Enewetak Island adjacent to MPRL (Fig. 4). This area was quarried during the Japanese occupation. Because a wide area of reef flat was left seaward to reduce wave swell entering the quarry, the Enewetak quarry is calm during low tides and is an ideal location for snorkling and diving. Numerous investigators at Enewetak have taken advantage of this. The Enewetak quarry covers about 2.75 hectares and averages about 1.5 m in depth, with the deepest sfX)t being 3 m. The biological communi- ties present in it are discussed in Chapters 7 and 8 of this volume. The reef flat at the south end of Medren Island was also quarried. Although slightly larger than the Enewetak quarry, little protective reef flat was left seaward of it; therefore, it is more open to wave action from the open ocean. A small quarry occurs at the north end of Medren on the reef flat. Seven relatively small areas were quarried on the reef flat near the middle portion of Runit Island. All are well inside the seaward margin of the reef flat and are well pro- tected from waves at low tides. At Enjebi, there are a few areas toward the north end where the reef flat was quarried. There is one elongate rectangular quarry and two small round ones. Also, on the western side of the island are three irregular areas next to shore, deeper than the adjacent bottom, which were prob- ably quarried for construction of the Japanese airstrip there during World War II. Craters Six craters remain from nuclear weafwns testing at Enewetak. Three craters are the result of atomic bomb tests. The other three are from thermonuclear weapons tests and are roughly three orders of magnitude larger in area and volume. Two atomic bomb craters are at the north end of Runit Island (Fig. 5). The histories, morphol- ogy, and subsurface geology of the Enewetak craters are extensively discussed by Ristvct (1978), resulting from work done by the Air Force Weapons Laboratory, Albu- querque. Both Runit craters are about 120 m in diameter. The most lagoonward. Cactus crater, was used for con- struction of the Cactus crater crypt during the Enewetak cleanup from 1977 through 1979 in which the crater was filled with cement, contaminated debris, and soil. A Fig. 4 Aarlai view ot the north end ot tnewetaK isiana snowing the buildings of the Mid- Pacific Rasearch Laboratory and to the right of them the quarry on the reef fiat. [Photo by E. S. RecM.] PHYSIOGRAPHY 33 Fig. 5 The north end of Runit Island with La Crosse crater (lower right) wnd the Cactus Crater Crypt (upper left). La Crosse crater is about 120 m across and about 10 m deep. The Cactus Crater Crypt was built in the crater to contain contaminated soil and debris. [Photo by P. L. Colin.] Fig. 6 The island of Boken (north) with the Seminole crater, a small atomic bomb crater. The island and adjacent islands In the foreground have been drastically altered by the forma- tion of the crater. [Photo by P. L. Colin.] 34 COLIN 25-foot-high cap of poured cement plates covers the crypt. The seaward crater, Lacrosse crater, was not altered dur- ing construction of the Cactus crypt. Nolan et al. (1975) described the distribution of substrate types in these craters and fish assemblages occurring in them (Chapter 7 of this volume). The third atomic crater is on the west side of Boken Island in the north of Encwetak (Fig. 6). It is similar in size to the Runit craters, roughly 200 m in diameter, 10 m deep, and is connected to the sea via the reef flat. Two of the thermonuclear craters are located between Boken and Bocinwotme Islands (Fig. 7). The MIKE crater is the western one, roughly 1.8 km in diameter and 56 m The last crater is some 7 km southwest of Bokoluo, the westernmost of the northern islands. The device was exploded from a boat anchored over the shallow lagoon margin. This produced a crater which excavated northwestward in the shallow reef and reef flat but is very open to the lagoon to the southwest. It is roughly 1.7 km in diameter. Other Physiographic Effects from Nuclear Tests A large area of reef flat and seaward reef face cleaved away in the area north of the MIKE crater sometime Fig. 7 Mike (right) and KOA thermonuclear craters on the northern reef at Enewetak Atoll, photographed from 10,000 feet. The larger Mike crater Is about 1.6 km across. The island of Boken with the small Seminole crater is seen on the left side of the photograph. The section of outer reef face which cleaved off after the KOA test can he seen seaward of the Mike crater at the bottom of the photograph. [Photo by P. L. Colin.] deep. The KOA crater to the east is slightly smaller, about 1.5 km in diameter. Both blasts were detonated on islands which disappeared with formation of the craters. A third island, Bogairikk (not shown in Fig. 1), was largely elim- inated with formation of the craters and is now represented solely by remnants on the sandbar west of Boken. The MIKE crater breaches the shallow reef into the lagoon at the 10 to 15 m depth contour. The KOA crater is still separated from the actual lagoon by a shallow txjt- tom of less than 6 to 10 m depth but is confluent with the MIKE crater on its west side. A minimum of about 400 m of reef flat separated the MIKE crater from the op>en ocean; a slightly greater margin exists between the ocean and KOA crater. between 1952 and 1958 (Fig. 7). The section of reef did not break away as a result of the MIKE test but was split off sometime later. About 300 m of the reef face, running as much as 60 m inward on the reef flat, fell away, and there is no bottom visible in aerial photos over what was once reef flat. This represents an exposure of underlying reef structure which is of unprecedented magnitude (see Chapter 4 of this volume for details). Direct examination of this scarp reveals that it is vertical to slightly overhanging with relatively sparse benthic organisms on its upper sur- face. Other nuclear-produced phenomena still visible at Enewetak include ejecta trails on the reef flat produced by thermonuclear tests, particularly in the area of the craters, PHYSIOGRAPHY 35 plus small depressions on the reef flat probably produced by single ejecta blocks. REFERENCES Atkinson, M J,, S V Smith, and E. D, Stroup, 1981, Circulation in Enewetak Atoll Lagoon, Limnol. Oceanogr . 26: 1074-1083 Emery, K. O., J. T Tracy, Jr , and H. S. Ladd, 1954, Geology of Bikini and Nearby Atolls, U. S. Geo/ Suru. Prof Pap.. 260 A. pp 1 265. Nolan, R. S , R. R. McConnaughy, and C. R. Stearns, 1975, Fishes Inhabiting Two Small Nuclear Test Craters at Enewetak Atoll, Marshall Islands, Micronesica, 11: 205-217. Ristvet, B. L , 1978, Geologic and Geophysical Inuestigations of the Enewetak Nuclear Craters, Air Force Weapons Lab., Air Force Systems Command, Kirtland Air Force Base, New Mex- ico, AFWL-TR-77-242. Chapter 4 Geologi; and Geohydrologi; of Enewetak Atoll BYRON L. RISTVET S-CUBED. A Division of Maxwell Laboratories, Inc. Albuquerque, New Mexico 87198 INTRODUCTION Enewetak Atoll is located at 162° east longitude and 11° north latitude in the Pacific Ocean. It is the north- westernmost member of the western Ralik (Sunset) Chain of the Marshall Islands. Enewetak Atoll is one of the larger atolls; it is roughly elliptical in shape, with a north-south length of 40 km and an east-west width of 32 km (Fig. 1). The reef is cut by three passes. The Deep Channel on the southeast side is only 1.5 km wide, but it has a depth of 55 m between Japtan and Medren Islands. The South Channel is approximately 9.5 km wide but is only 10 to 20 m deep. The Southwest Passage is even shallower, only 2 to 4 m in depth. Maximum tidal currents of nearly 1 m s~' in the Deep Channel and 0.5 m s~' in the South Channel have been observed (Emery ct al., 1954). The reef may be divided into four parts with distinct morpholo- gies related to their positions relative to the prevailing northeast trade winds. The parts are the windward reef on the northeast, the leeward reef on the southwest, and the two transitional reefs on the northwest and southeast (Fig. 1). The reef encloses a lagoon of 920 km^ with a maximum depth of 65 m. The lagoon has a relatively smooth carbonate sediment bottom studded with hundreds of coral pinnacle and patch reefs (Emery ct al., 1954). Forty-two low-relief islands and islets composed of car- bonate sands and gravels exist on the atoll with a total dry land area of 6.7 km^ with the largest islands being about 1 km^ in area. Enewetak Atoll receives an average annual rainfall of 1470 mm, mostly during August to December. Rainfall is highly variable with annual totals ranging from 605 to 2422 mm (Buddemeier, 1981). Tides are of the mixed semidiurnal type with a maximum range of about 1.8 m. The purpose of this chapter is to summarize the vast wealth of data on the geological aspects of Enewetak gath- ered over the last 40 years. A SUMMARY OF GEOLOGIC INVESTIGATIONS The history of investigations of atoll geology in general and Enewetak Atoll in fiarticular may be divided into three periods: pre-1946, 1946 to 1964, and post-1964. Th« first period was one of discovery and initial exploration. These early observations became the framework for many hypotheses on the origin and evolution of atolls. Most of the early studies focused on the surficial geologic features and lacked the direct sampling of subsurface data to evalu- ate the many hypotheses of the day. Beginning In 1946, there was a significant increase in knowledge of atolls resulting from a series of comprehensive scientific studies of the northern atolls of the Marshall Islands, particularly Bikini and Enewetak. These geologic investigations were conducted by U. S. Geological Survey (USGS) scientists for the U. S. Atomic Energy Commission (AEC) to estab- lish baselines to assess effects from nuclear weapons test- ing conducted at Enewetak and Bikini between 1946 and 1958. A vast amount of surface and subsurface geologic data was gathered and analyzed, and the results were pub- lished through 1964 (cf. Emery et al., 1954; Schlanger, 1963). From 1964 to the present, scientific studies have been of two types: those which have continued to addrcs* the problems conceptualized by earlier studies and those which have addressed the effects of the nuclear weapons testing at the two atolls. Enewetak Atoll continues to this day as one of the sites of significant studies of atoll geol- ogy, carbonate sedimentology, and organism/sedinrKnt interrelationships. Prc-1946 Period The first geologic studies of the Marshall Islands were conducted in 1816 and 1817 by Albert Chamisso. Cha- misso, a naturalist with the Russian Von Kotzebue expedi- tion to the northern and western Pacific Ocean, described the reefs, islands, and lagoons of the eastern chain of the Marshalls. For the rest of the 19**^ century the MesrshAls were visited only by general surveying expcdftkxN (Em«T? et al., 1954). 37 38 RISTVET 2 ^ SOUTHWEST \ PaSSAGE SOUTH CHANNEL 10 20 STATUTE MILES CONTOUR INTERVAL 100 FATHOMS DATUM IS MEAN LOW TIDE Fig. 1 Location map of Enewetak Atoll, Marshall Islands. Meanwhile, expeditions in other areas of the world were contributing to an understanding of atoll geology. Darwin (1842), during the voyage of the HMS Beagle from 1831 to 1836, studied reef building organisms and reef morphologies. He established a three-fold classification of reefs that is still used today: fringing reefs, barrier reefs, and atolls. Darwin (1842) integrated his findings into a theory of atoll formation on subsiding island foundations with antecedent fringing and barrier reef stages. However, other workers (cf. Agassiz, 1903; Daly, 1915; and Gar- diner, 1931) later proposed alternate theories postulating that atoll reefs grew upwards from still standing submarine summits of various origins. By the turn of the century, direct subsurface sampling became a paramount issue to understanding the origin of atolls. Funafuti Atoll, in the Ellice Islands 2400 km south GEOLOGY AND GEOHYDROLOGY 39 of Enewetak Atoll, was the site of the first sampling well drilled on an atoll (David ct a!., 1904). This well, drilled from a ship in the lagoon, penetrated 337 m of carbonate sediments, demonstrating the great thickness of atoll reef sediments. A major review of the theories of atoll formation was written by Davis (1928). Davis carefully evaluated the data and hypotheses; his evaluation supported Darwin's (1842) subsidence theory as correct and rebutted alternate theories. Although Davis (1928) rejected Daly's (1915) the- ory that glacial period sea level histories resulted in atoll formation, he did enter them as an important new element to consider in evaluating the geologic history of atolls. Before 1946, atolls of the Marshall Islands provided lit- tle evidence for the aforementioned theories. During the period of 1918 to 1944, the Marshalls were under the control of Japan; although Japanese scientists conducted studies on the atolls, much of the resulting data are not readily available (Emery et al., 1954). Stearns (1945) made some general comments on possible battle damage to the reef of Enewetak Atoll following the American occupation in 1944. In summary, at the end of 1945, only a small body of data existed on atoll geology. The general locations and morphologies were described, and the types of reef- building organisms and their environmental requirements were known in a general sense. Conclusive evidence on atoll formation had not been found, and sparse data existed on subjects such as atoll foundations, ages, lagoon and outer slojDe sediments, reef zonations and productivi- ties, and ecology. 1946 to 1964 Period A period of intense scientific study on the northern Marshall Islands began in 1946 to establish baselines from which damage could be assessed from the U. S. Nuclear Weapons Testing Program. Bikini Atoll was chosen to be the site of the first nuclear weapons effects tests conducted by the United States. Operation Crossroads, consisting of the detonation of two atomic bombs over and under naval ships in Bikini Lagoon, was conducted in 1946. Two expeditions to Bikini were made in 1946 and 1947 to study the atoll environment. The 1946 effort included gen- eral surficial geologic studies of the rcjf, lagoon floor and outer slopes, and a seismic refraction study of the subsur- face structure of the atoll (Emery et al., 1954). The 1947 studies yielded much geologic information on the subsur- face through the drilling of three holes on Bikini Island and reef, with one hole penetrating 775 m of carbonate sedi- ments (Ladd et al., 1948). In 1950 additional seismic refraction studies were completed at Bikini and the adja- cent Sylvania Guyot and the southern part of Kwajalein Lagoon (Dobrin and Perkins, 1954; Raitt, 1954). An aeromagnetic survey of Bikini Atoll (Keller, 1954) was also completed. Nuclear testing began at Enewetak Atoll in 1948 with three events of Operation Sandstone. Shortly thereafter. the USGS began a series of geological and scientific inves- tigations again to establish baselines to measure the effects of the nuclear detonations. In 1950 four shallow holes were drilled by the AEC in the reef on the seaward side of Engebi (Enjebi) Island to locate a suitable rock quarry (Ladd and Schlanger, 1960). In 1951 the AEC drilled 17 shallow holes on six different islands for soils engineering studies related to the construction of structures for the nuclear testing. The AEC also drilled three deep holes on the atoll in 1951 and 1952 under the technical guidance of the USGS: K-IB was drilled to 390 m on Engebi (Enjebi), F-1 was drilled to 1411 m on Elugelab, and E-1 was drilled to 1287 m on Medren Island. Both F-1 and E-1 reached volcanic basement with 5 m of olivine basalt being recovered from E-1. The confirmation of a basaltic foundation beneath Enewetak Atoll substantiated Darwin's subsidence theory of atoll formation (Ladd et al., 1953). Drill holes K-IB and F-1 were subsequently destroyed during nuclear tests, but the E-1 hole is still open to at least 609 m (Daniels et al., 1984). After 1952 field study of the geology of the northern Marshall Islands was reduced significantly, although the nuclear testing continued through August 1958. The con- tinued availability of Enewetak for future field studies was ensured by AEC's establishment in 1954 of the Enewetak Marine Biology Laboratory, now known as the Mid-Pacific Research Laboratory. The AEC completed some additional shallow drilling in 1953 and 1956 for soils engineering (Pratt and Cooper, 1968), but no more drilling for geologic study was completed until 1971. However, the vast amount of field data and samples yielded in the 1948 to 1952 efforts were studied and evaluated through 1964. Formal presentation of the completed studies was com- piled in the USGS Professional Papers 260 Series com- pleted in 1964. This 28-paper series comprises the most comprehensive single body of geologic, geophysical, and oceanographic data ever assembled on a group of atolls. Much of the rest of this chapter will draw heavily on the data presented in these papers. A major pap)er by Emery et al. (1954) is a comprehen- sive study of the surface geology of Bikini, Enewetak, and nearby atolls. It also presents data on the sediments of the lagoons, reefs and islands, reef morphologies and lithologies, and coral zonations of different reefs as well as many other topics. Also presented in the paper are the lithologic sections for the deep holes drilled on Bikini. Sub- surface zones of calcitic limestones are described which are overlain and underlain by aragonitic sediments. These lime- stones are postulated to represent times of subaerial exp)0- sure of the atoll (Ladd et al., 1948; Emery et al., 1954). Another paper in the 260 series by Munk and Sargent (1954) describes the variation in the spur and groove structure of the Bikini reefs and relates them to distribu- tion and direction of wave energy. This relationship demonstrates that these are not relict Pleistocene erosional forms. Wells (1954) defined ecological zones of windward reefs in the northern Marshalls on the basis of dominant coral faunas and compared these zonations with reefs else- 40 RISTVET where. The great organic productivity of atoll reefs versus the surrounding oceans is demonstrated by Sargent and Austin (1954). The subsurface geology and geophysics of Enewetak Atoll somewhat dominates the 260 series. The penetration of a basaltic basement by drilling and the aeromagnetic and seismic refraction surveys indicated the presence of volcanoes beneath Enewetak, Bikini, and Kwajalein Atolls. Ladd and Schlangcr (1960) present the locations and drill- ing data for the Enewetak drill holes. They conclude that most of the near surface material to 60 m depth is uncon- solidated, whereas deeper zones of recrystallized and leached carbonate are postulated to represent fjeriods of subaerial emergence of the atoll. Foraminifera were used to establish a Tertiary biostratigraphy of the Enewetak sub- surface and to document continuous shallow water condi- tions in which the entire carbonate section had been de- posited (Cole, 1957; Todd and Low, 1960). The oldest carbonates were identified as Upper Eocene in age. The general subsurface geology of Enewetak was defined by Schlanger (1963), who presents detailed litho- logic logs of the Enewetak deep holes and provides an interpretation of the geologic history. Schlanger (1963) noted the presence of numerous "solution unconformities" within the Enewetak geologic column. The term solution was used because Schlanger felt these unconformities represented karstic surfaces. The scientific programs in the northern Marshalls had a stimulating effect on the academic interests in atolls. The interest in the geology and biology of carbonate reefs is still a dominant field of study. The geologic studies of this period answered many of the basic questions about atoll formation. Atolls rested on subsided volcanic foundations. The compositions and dep>ositional environments of the subsurface sediments were characterized and interpreted. Specific zones of altered carbonates were identified and interpreted to represent periods of atoll emergence and given paleohydrologic meaning. Reef zonation and mor- phology, as products of interacting biological aggradation and mechanical and biological erosion, became better understood. 1964 to Present Geological studies conducted during this time in the Marshall Islands have been centered on Enewetak Atoll. Numerous studies primarily concerned with sediment/ organism interrelationships, the distribution of radionu- clides within the atoll sediments, and geohydrology have been conducted under the auspices of the Mid-Pacific Research Laboratory, which is sponsored by the Depart- ment of Energy (DOE). The Defense Nuclear Agency has sponsored four major field programs to understand the craters resulting from the near-surface detonation of nuclear weapons: the Pacific Cratering Experiment (PACE), 1971 to 1972; the Exploratory Program on Enewetak (EXPOE), 1973 to 1974; the Enewetak Atoll Seismic Investigation (EASI), 1980; and the Pacific-Enewetak Atoll Craters Exploration (PEACE), 1984 to present. Until the early 1970s, studies of Enewetak geology consisted of reviews or extensions of previous work. Gross and Tracey (1966) used stable carbon and oxygen isotope data to substantiate the hypothesis that the calcific lime- stones in the subsurface were formed in fresh water environments (Ladd et al., 1948; Ladd and Schlangcr, 1960; Schlanger, 1963). Thurber et al. (1965) performed U/Th radiometric dating of corals of the Enewetak subsur- face from the Quaternary [>eriod and revealed an absence of corals dating between 6000 and 100,000 years before present (ybp), indicating a significant hiatus in def)osition. At Enewetak, PACE was conducted to evaluate the influence of the shallow subsurface geology on the dimen- sions of nuclear explosion craters. It consisted of two phases: (1) geological and geophysical investigations of the shallow (<70 m) subsurface of the atoll and (2) a series of high explosive cratering experiments. A federal court order cancelled PACE before most of the high explosive crater- ing experiments were conducted. However, much of the first phase was completed, and nearly 250 shallow boreholes were completed on seven islands, with 235 being drilled on Aomon or Runit Island (Henny et al., 1974). Most of the holes were soils engineering borings which returned little or no sample. Sample recovery in the cored boreholes was variable but was generally quite fxxjr. A generalized four-layer engineering geology model for the shallow subsurface at Aomon Island was developed by Henny et al. (1974) using these limited samples and seis- mic refraction survey data. The follow-on geologic and geophysical program to PACE was EXPOE. The objective of the EXPOE program was to develop a model of the near surface geology of the atoll for the nuclear crater regions in the northern islands. Forty-six cored boreholes and 13 water sampling wells were completed on 11 islands on the windward, leeward, and transitional sides of the atoll, and 250,000 lineal feet of shallow seismic refraction surveys were completed from 1973 to 1974. The EXPOE program was notable for the excellent sample recovery: recovery of 4-in. cores of both consolidated and unconsolidated materials averaged over 80% (Couch et al., 1975). This recovery was far greater than any previous drilling, especially in the pxjorly and unconsolidated near surface sediments, and allowed for a more detailed picture of the stratigraphy and petrology of the upper 100 m of the Enewetak subsurface than gained in previous studies (Ristvet et al., 1974, 1977). As will be discussed in greater detail later in this chapter, the EXPOE findings indicate that the atoll pe- riphery to at least 80 m depth consists of subordinate reef and dominant back reef and marginal lagoon deposits of the Holocene and Pleistocene ages. Five subaerial surfaces were recognized in the Pleistocene section associated with sea level drops during glacial periods (Ristvet et al., 1974, 1977). The EASI field program consisted of overwater high resolution multichannel seismic reflection surveys of the KOA and OAK nuclear craters and the undistu''bed lagoon off of Engebi (Enjebi) Island (Ristvet et al., 19ti0; Tremba GEOLOGY AND GEOHYDROLOGY 41 et al., 1982; Trcmba, 1985) and participation in the MPRL sponsored R/V Makali'i submersible dives in 1981. The EASI seismic reflection profiles showed that shallow unconformities recognized in the EXPOE drilling continued across the lagoon paralleling the present day bathymetry. Additional deeper reflectors at 150 and 245 m and a series of reflectors between 320 and 365 m were noted and compared to the unconformities described by Schlanger (1963). It was hypothesized that the Middle Miocene reflectors between 320 and 365 m may be a representation of a series of closely spaced unconformities much like the Pleistocene section described for Enewetak and Bikini (Ristvet et al., 1974, 1977; Tracey and Ladd, 1974). Unfortunately, only deep drilling with high core/sample recovery would resolve this issue. The PEACE program was a two-phase program with the objective of understanding the surface and subsurface morphologies of OAK and KOA nuclear craters. The first phase of the PEACE field program was performed during the summer of 1984 and included high resolution mul- tichannel seismic reflection, bathymetric, side-scan sonar, and submersible studies primarily of the two cratered areas but included some studies of atoll-wide nature (Folger, 1986). The second phase of PEACE was conducted during the summer of 1985 and consisted of overwater drilling into and adjacent to the two nuclear craters. High core recovery was obtained in drill holes as deep as 490 m beneath the lagoon floor. The PEACE drilling data are in the analysis phase. During this post- 1964 period, Enewetak was the site of several geologic investigations sponsored by MPRL. Exam- ples of these investigations include rates of calcification of the windward reef (Smith and Harrison, 1977), studies of Holocene sea level histories which suggest a higher than present sea level 4000 to 2200 ybp (Tracey and Ladd, 1974; Buddemeier et al., 1975), and investigations of the Quaternary history of the reef flat (Szabo et al., 1985). Submersible studies of the outer slope have been con- ducted by Colin et al. (1986) and Halley and Slater (1985) to define the morphology of the outer reef slope. SURFACE GEOLOGY General Enewetak surficial geology is best divided on the basis of depositional environments: the outer slope, the reef, the islands, and the lagoon. Outer Slope The topography around Enewetak Atoll was first deter- mined by 85 radial and five partially complete concentric lines of soundings made by the USS Bowditch in 1944 (Emery et al., 1954). The profiles show a steep slope of 18° to 49° from the reef edge to 450 m depth changing to a more gentle slope of 10° between 450 and 2000 m. Sediments collected from a profile seaward of the South Channel showed a predominance of fine grain and Halimeda debris to 1500 m depth (Emery et al., 1954). In 1981, 22 submersible dives were made on the outer slope of the southern half of Enewetak to depths as great as 360 m (Colin et al., 1986). The outer slope was found to be quite steep, averaging about 60° between 90 and 360 m on the windward and transitional side and slightly greater on the leeward side. No terraces or grooves were noted below 30 m. Vertical grooves were noted on the lee- ward side below 150 m depth. Talus accumulations were noted below 150 m, with significant sediment slopes being found seaward of the South Channel below 200 m depth. Below 90 to 100 m depth, it appeared that no significant reef framework was being constructed. Significant quanti- ties of sediment are being transported down the face of the outer slope on the windward side with little or none being transjaorted on the leeward side. In 1984 and 1985, Halley and Slater (1985) investi- gated the outer slope of the reef north of the MIKE nuclear crater utilizing the research submersible R/V Delta. Halley and Slater (1985) noted that the slope is character- ized by three zones: (1) the reef plate, algal ridge and near fore reef, from sea level to 16 m depth with less than a 10° slope; (2) the by-pass slop>e, from 16 to 275 m, with slopes of 55° decreasing to 35° near the base; and (3) a debris slop)e less than 35° below 272 m depth. Halley and Slater (1985) also examined an exp)osed cross section through the reef and fore reef deposits within a rockfall scarp created by the KOA nuclear detonation. The slump scarp exposes three stratigraphic units that are differentiated by the surficial apf)earance: (1) a near-vertical wall from the reef crest to 76 m that appears rubbly and is composed mainly of coral heads; (2) a vertical to overhang- ing wall from 76 to 220 m that is massive and fractured, producing smooth, blocky surfaces; and (3) inclined bed- ding below 220 m along which the slump block has frac- tured, exposing a dip slope of hard, dense white carbonate rock that extends to below 400 m. Caves occur in all three units. Fore reef boulder beds dipping seaward at 30° are truncated by the current outer slope surface, thus revealing the erosional nature of the bypass slop)e. Atoll Reefs The Enewetak reefs, like those described elsewhere in the Marshalls and other localities in the world, show a strong zonation in bands parallel to the front (Emery et al., 1954). These bands are defined by both coral and coral-algal communities (Odum and Odum, 1955) and by sediment deposition patterns (Emery et al., 1954). Differ- ences in the zonation types are recognizable for the three reef types: windward, leeward, and transitional. Most previ- ous studies have concentrated on the zonation of the wind- ward reef (Emery et al., 1954; Odum and Odum, 1955; Wells, 1954); however, description of the leeward and transitional reefs are presented by Emery et al. (1954). Figure 2 presents the zonation of the windward reef. The zones consist of fore reef, algal ridge, coral-algal, reef 42 RISTVET SU3J.3N Nl Hidaa GE0LCX3Y AND GEOHYDROLOGY 43 flat, and back reef flat. Figure 2 also displays the relation- ship of the reef to the islands and lagoon. Each of the five zones has unique biologic and geologic characteristics. Each of these zonations provides a model for what is seen in the subsurface. However, as will be seen in subsequent sections of this chapter, the Pleistocene subsurface appears to consist dominantly of subtidal deposits, whereas the modern reef flat consists of predominantly intertidal environments. Tracey and Ladd (1974) and Buddemeier et al. (1975) present evidence that the broad intertidal, rocky platform of the modem windward reef flat consists of lithified sub- tidal sediments implying a previous higher-than-present Holocene sea level. The modern windward reef is an ero- sional platform develo[)ed after a growth of the Holocene reef to a higher sea level. Hence, there is the possibility that the modern Enewetak windward reef flat is not a good model to use to interpret former aggradational reef environments seen in the subsurface. The windward fore reef consists of an area 30 to 50 m wide, sloping gently seaward at 10° to 15° and covered with coral and Halimeda sp. These gentle slopes do not exist on the leeward reef where the fore reef has 40° to 60° slopes. The same biological communities exist on the leeward fore reef as on the windward side (Colin et di., 1986). The fore reef extends to a depth of 30 m where the slope rapidly steepens, and the presence of stony corals and Halimeda declines drastically. At fairly regular intervals along the slop>e, there are nearly straight grooves perpendicular to the reef face. These grooves are from 2 to 3 m wide and 8 to 15 m long and are separated by spurs 5 to 10 m or more wide. The spurs are composed of living encrusting coralline algae (Emery et al., 1954). The origin of the grooves and spurs has been suggested by Munk and Sargent (1954) to dissipate the wave energy against the reef front. These grooves often extend into the algal ridge, especially on the transitional and leeward edges. The fore reef appears to be a site of active reef building with the sediments being cemented by biologic binding and penecontemporaneous marine cementation. The algal ridge is primarily composed of encrusting red algae, primarily Porolithon. The algal ridge may actually grow above the reef flat elevation to as much as 0.3 to 0.6 m above the lowest low water due to wave action keeping the living algae wet during low water. The algal ridge with its biological and marine cementation provides the framework for the preservation of the back reef and lagoonal sediments from the erosion of ocean waves (Emery et al. 1954). Algal ridges occur on both the lagoon and ocean sides on the leeward reef. Both of these leeward algal ridges are poorly developed and do not rise much above the lowest low water. On the inner side of the algal ridge, there is a belt of rich coral growth from 50 to 150 m wide. Stony corals cover more than 50% of the reef surface. Shallow pools contain most of the coral. The remainder of the zone is a pavement of encrusting red algae. The growth forms of the coral colonies are low or encrusting to withstand the wave action and low tides. Corals are predominantly Acropora. Pocilhpora, and Montipora. Again the coral-algal zone through biological and marine cementation provides well-cemented sediments for incorporation into the subsurface. The windward reef flat at Enewetak is a fairly level rock surface that may be divided into two rather distinct parts: (1) a barren rock surface that appears to be the ero- sional surface of an older reef and (2) a rock substrate with a thin veneer of organisms, primarily the articulate red alga, Jania, giving the surface an appearance of being covered by a mat which Smith and Kinsey (1976) dubbed the "algal-turf." Tracey and Udd (1974) and Buddemeier et al. (1975) present evidence to suppxsrt a higher-than-present sea level between 4000 to 2200 ybp. This higher sea level may have been 1 m or more greater than the present. The ero- sional nature of the present reef flat is postulated to be due to the lowering of sea level to near its present datum around 2000 ybp. Tracey and Ladd (1974) support their hypothesis with age dates of planed coral heads in the present windward reef flat seaward of Runit and Aomon Islands. Additional evidence is provided by Buddemeier et al. (1975), who through age dating show that much of the windward reef flat seaward of Aomon Island is composed of cemented subtidal deposits now present in an intertidal zone, the result of a recently lowered sea level. Additional evidence for a higher-than-present Holocene sea level around 4000 ybp for other Pacific islands may be found in Curray et al. (1970) and Chappell and Veeh (1978). Despite its apparent erosional character, the present windward Enewetak reef flat is a highly productive reef environment in terms of the mass of carbonate sediments produced (Smith and Harrison, 1977). The algal-ridge, coral-algal zone, and the reef flat compose what is termed the "reef plate" (Henny et al., 1974). The reef plate con- sists of well-cemented rock resulting from penecontem- poraneous biologic and marine cementation. During PACE and EXPOE, several holes were drilled on the reef plate seaward of Aomon and Runit Islands. These holes, in addition to the outcrops exposed in quar- ries on the Enewetak, Medren, Runit, and Engebi (Enjcbi) reef flats and the outcrops exposed in the LaCrosse nuclear crater on the Runit reef flat, show that the Holo- cene reef plate is a lagoonward prograding wedge of well- cemented sediments overlying unconsolidated subtidal car- bonate sands and gravels. The seaward edge of the wedge begins approximately at the reef plate /coral-algal zone boundary. Within the shallow Quaternary subsurface, sedi- ments beneath the coral-algal zone appear to be continu- ously well cemented with depth. Beneath the reef flat, the thickness of the wedge tapwrs from 3 to 4 m at the center of the reef flat to <1 m at the back reef /reef flat bound- ary (Ristvet et al., 1977). The back reef is characterized by small to large solitetfy coral heads of Pontes and Heliopora in a rocky to sandy substrate. Little or no marine cementation app>ears to be occurring, and the sands and silts have their origin from 44 RISTVET sediment production and bioerosion of the reef flat. This environment extends from the reef flat to the islands. Atoll Islands The present islands of Enewetak represent wave and eoiian deposits of excess sediment production from the reef stabilized in part by the formation of beachrock. Islands are present on the reef except on the northwest transitional reef. The islands are all approximately 3 to 4 m in elevation above the lowest low water. Two basic island shapes exist for Enewetak Atoll: (1) long linear islands that parallel the reef h-ont, such as Runit, Enewetak, and Bokoluo and (2) the triangle-shaped islands with the base on the lagoon side parallel to the reef front and the point facing the seaward reef, such as Engebi (Enjebi), Aomon, and Louj. The origin of these two island shapes is not understood. The islands are covered with vegetation and have fairly well-developed soil profiles. The origin of beachrock has been the subject of several investigations at Enewetak and other carbonate beaches in the world. Beachrock at Enewetak is present on 30 to 40% of all beaches. The formation of beachrock appears to be a fairly recent phenomenon with significant formation continuing today. The author has collected samples of beachrock at Enewetak encapsulating World War 11 shell casings and cables from the nuclear testing pjeriod. The origin of beachrock was first investigated in the Marshalls by Emery et al. (1954), who looked at interstitial water chemistry and concluded that evaporation and heating of interstitial seawater resulted in carbonate precipitation. Schmalz (1971) studied the interstitial water of beach sediments on the lagoon side of Bijire Island in 1967. He concluded that precipitation of the dominant acicular aragonite and minor micritic magnesian calcite cements in the interstices of the carbonate sand was caused by the mixing of seawater with the brackish meteoric water in the thin Gyben-Herzberg lens. A succession of studies on the origin of beachrock cements followed Schmalz (1971). Commonly invoked processes for the precipitation of beachrock cements include evaporation of seawater, mixed fresh-saline waters, and vague types of biological involvement (Manor, 1978). Current models show that degassing carbon dioxide from beach groundwaters appears to be the primary phenomenon that forms beachrock (Manor, 1978). Atoll Lagoon The bathymetry of the lagoon was mapped in detail by the U. S. Navy in 1944. Nearly 180,000 soundings were made, and the results were contoured (Emery et al., 1954, chart 5). The lagoon bathymetry is somewhat irregular due to the presence of numerous coral knolls (patch and pinna- cle reefs). The lagoon consists of four major bathymetric features: (1) lagoon terrace; (2) lagoon basin; (3) coral knolls; and (4) the reef openings. The lagoon bathymetry shows a terrace between 15 and 22 m depth (Emery et al., 1954). The terrace borders all edges of the lagoon except the northwest and southern margins, where it is absent. The width is variable with 3 km being the greatest attained. The lagoon terrace is dot- ted with numerous patch reefs. The slopes from the islands to the terrace are gentle, averaging <2.5°. An even gentler slope, averaging 1.25°, separates the terrace from deep basin (Emery et al., 1954). The main lagoon basin is a relatively flat area with slopes of 0.10°. The greatest depths are nearly 65 m in the northwestern half of the lagoon. The mean depth of the basin is approximately 55 m (Emery et al., 1954). Within the lagoon are a large number of individueJ coral knolls or patch and pinnacle reefs. Emery et al. (1954) reported the presence of 2293 individual coral knolls. About 10% of the knolls rise to within 8 m of the surface. Most have tops between 30 and 36 m depth. The distribution of the coral knolls within the lagoon apf)ears to be random. Seismic reflection profiles from EASl and PEACE through knolls suggest that they are predominately Molocene features. Nearly half of the knolls are formed over what is interpreted to be preexisting eroded Pleisto- cene patch or pinnacle reefs, whereas the other half of the lagoonal coral knolls do not appear to have an antecedent structure beneath them (Tremba, 1985; Grow et al., 1986). The bottom sediments of the Enewetak Lagoon were first characterized by Emery et al. (1954) and most recently by T. W. Menry and B. R. Wardlaw (personal communication). Emery et al. (1954) found that the sedi- ments consist of the following chief components: Halimeda sand, coral sand and gravel, foraminifera sand, mollusc shell sand and gravel, and fine debris. Fine debris was defined as all grains <0.25 mm in diameter. Emery et al. (1954) show the terrace to be dominated by fine debris and the basin by Halimeda and foraminifera sand. Menry and Wardlaw (1985) show a similar distribution but reiport much more mud-sized (<0.062 mm) carbonate sediment on the terraces and in the basin than Emery et al. (1954). McMurtry et al. (1985) have investigated the magni- tude of bioturbation of the lagoonal sediments off Runit Island and found that the burrowing shrimp of the family Callianassidae nearly completely mix and redistribute the surface sediments to a subbottom depth of at least 1.5 m. SUBSURFACE GEOLOGY AND GEOPHYSICS The subsurface geology of Enewetak Atoll consists of an approximately 1370 m thick carbonate sediment caprock overlying the summit of a basaltic volcano that rises 5000 m above the floor of the ocean (Ladd et al., 1953). Most o* the drill hole data for the interpretation of the subsurface geology of Enewetak are derived from drilling on islands or the reef flat. The PEACE program has added data to 490 m subbottom depth on the north- ern lagoon terrace and northwestern shallow lagoon. The subsurface geology, as deduced from the analysis of the borehole samples and seismic profiles, is very similar to ' n GEOLOGY AND GEOHYDROLOGY 45 the subsurface geology of Bikini Atoll (Emery et al., 1954) and Midway Atoll (Ladd ct al., 1970). Basement Rocks In 1952, two deep holes (Fig. 3) reached the volcanic basement below the carbonate sediment caprock. Hole F-1 on Elugelab Island encountered hard basement rock at 1405 m depth. In hole E-1 on Medren Island, unweath- ered basalt cuttings were recovered from 1267 m, and solid basalt cores were taken from 1282.5 m to 1287 m. The basalt was an alkali olivine basalt containing analcime (Schlanger, 1963). The rock is similar to the late lavas of INSERTS APPROXIMATE PAKRY ISLAND 'Vl>,,^ ' \ '^ • ENIWETOK ■■i HAWAIIAN IS \^ -^^K>c f'T^^ ^ 1 180' 140- INDEX MAP J_ _L 162*00' 162MO' 162'20 162-30- Fig. 3 Location map of three deep AEC holes drilled in 1951 and 1952. [From Ladd and Schlanger, I960.] 46 RISTVET the Hawaiian volcanoes and other islands of the central Pacific Ocean. Kulp (1963) found the basalt to be Eocene in age with a whole rock K/Ar radiometric date of 59 ± 2 million ybp and a pyroxene K/Ar date of 51 ±5 mil- lion ybp. The shape of the volcano is characterized by the two drill holes at the north and southeast edges of the atoll, the seismic refraction profiles of Raitt (1957) and the recent seismic reflection profiles of the PEACE program (Grow et a!., 1986). Figure 4 displays the subsurface ve- locity structure of the atoll from the surface to the upper mantle as interpreted by Raitt. Although not depicted in Fig. 4, the uppermost velocity layer is a thin (105 m thick), low velocity (1920 m s~^) unit detected beneath the reef northwest of Elugelab Island. The second and third layers have apparent harmonic mean velocities of 2440 m s~' and 3050 m s~\ respectively. Raitt (1957) suggests that both velocities are characteristic of partly con- solidated calcareous sediments. The fourth through sixth layers comprise the volcanic basement underlying the car- bonate caprock. Finally, the seventh layer has a seismic velocity, 8.1 km s~\ characteristic of the upper mantle. Grow et al. (1986) show, in seismic reflection profiles, the top of the volcanics to be a relatively flat surface with only minor topograpy. Carbonate Rocks Figure 5 displays Schlanger's (1963) generalized interpretation of the subsurface of Enewetak Atoll based on the three deep holes drilled in 1951 and 1952. Also used for comparison is the interpretation of the subsurface of Bikini (Emery et al., 1954) which shows a strong simi- larity in the vertical extent of these zones for both atolls. Schlanger (1963) recognized that beneath both Enewetak and Bikini, there are zones characterized by the presence of fossil molds and solution features (leached and altered sediments) alternating vertically with zones containing pri- mary skeletal aragonite (unaltered sediments) separated by relatively sharp boundaries. Schlanger (1963) termed the upper surfaces of the leached zones "solution unconformi- ties," because they resemble karstic surfaces. Ladd et eil. (1948) suggested that the leached zones at Bikini represented periods of atoll emergence and exposure of the marine sediments to subaerial conditions. The unal- tered zones were believed to represent sediments that were never emergent. Schlanger (1963) identified three major solution uncon- formities in the subsurface of Enewetak at depths of 850 m (Early Miocene), 335 m (Middle Miocene), and 90 m (Pleistocene) below the surface (Fig. 5). Schlanger (1963) also described an interval of partially leached and altered sediments within the upper 90 m of the Enewetak subsurface. However, due to limited sample recovery he was unable to identify a solution unconformity within this interval. He did conclude that at least one additional p>eriod of atoll emergence had occurred during the Pleisto- cene following the emergence related to the major solution unconformity at 90 m depth. The high percentage of recognizable fossil material in the three deep drill holes allowed Schlanger (1963) to interpret the depositional environments of the sediments. Figure 6 presents the interpreted paleoecology of holes E-1 and F-1 at Enewetak and 2A-B at Bikini. The sec- tion of Eocene fore reef deposits in hole F-1 between 1280 and 1385 m represents outer slop)e deposits contem- fwraneous with near-reef, shallow-water deposits in hole E-1 from 845 to 950 m depth. The section of fore reef deposits in F-1 from 822 to 1280 m has no counterpart in E-1. Schlanger (1963) interpreted this as evidence that the earliest reef building at Enewetak began on the southeast side of the atoll near Enewetak and Medren Islands of today. Reef production and p>ossible erosion of the southeast reef during the lower Miocene emergence resulted in the wedge of fore reef sediments seen in F-1. The seismic reflection studies of Grow et al. (1986) app>ear to confirm a prograding reef front from southeast to northwest starting in the presumed Eocene sediments and continuing to the Middle Miocene unconformity. Post-Eocene carbonate sediments are all of shallow- water origin as sampled in the three deep holes, the EXPOE holes, and the PEACE holes. By the Middle Miocene unconformity, the location of the reef tract apF>ears to have been very close to the present position of the modern day islands (B. R. Wardlaw, personal com- munication). In the substantial time represented by the upper 335 m of carbonate sediments, the reef tract has only migrated seaward 200 to 300 m. Deposition of shallow-water sediments under conditions of slow atoll subsidence continued through the Middle Miocene (Cole, 1957). However, the presence of several disconformities noted in the PEACE drilling from 335 to 490 m (T. W. Henry and B. R. Wardlaw, personal communica- tion) and Schlanger's (1963) reported presence of recrystallized limestone from 603 to 650 m suggest some periods of atoll emergence. Minor amounts of dolomitized limestone were noted within the Eocene stratigraphic section in both F-1 and E-1 and below the assumed Lower Miocene solution unconformity in F-1 (Schlanger, 1963). The dolomite appears as protodolomite replacing calcite. Schlanger (1963) felt that its origin may have resulted from the alteration of high magnesian coralline algae. Other hypotheses have been proposed including dolomite forma- tion in the mixing zone of meteoric groundwaters with sea- water during atoll emergence and/or formation from hyf)ersaline conditions in a restricted shallow-water environ- ment penecontemp)oraneous with dep>osition. Sailer (1984) presents new evidence using stable Sr isotope data that the Enewetak dolomite precipitated from normal seawater significantly following deposition at burial depths greater than 900 m. The 335 m unconformity described by Schlanger (1963) indicates a major emergence occurred after the depKJsition of Middle Miocene sediments. Ristvet et eJ. (1980) postulated, on the basis of the EASI seismic reflec- tion profiles, that the 335 m solution unconformity con- GEOLOGY AND GEOHYDROLOGY 47 UJ (- UJ o liJ Q 15 18 NORTH DATUM MSL OCEAN BOTTOM m < < UJ SOUTH T^T^^^r-^^g^f^^^^^ 4,145 ^^^ VAVVWVVVVW" 5,578 6,888 t^v^t:vvv;-- 8,077 \ \ ~^7 ? WVV\\V\V\\\Vi? ? AVVV\\\\V\VVVV ? 0- q: UJ H _ UJ 6- O a. UJ Q 12 15- WEST 2 EAST _ UJ _i or. i I DATUM MSL _ |^ t ^^ w OCEAN BOTTOM -"^ "^''"^L^^^VX- ^^^ __2,438 ^y ^^^vTTTVrVX^^'^ \ --T\W--' ,- 5,578 \ V^2738 V ^"~ \\\\\\\ 6,888 VCVr^ SHOT FIRED ABOVE LAYER EXTRAPOLATIONS AND INTERPOLATIONS ▼ LOCATION OF RECEIVING STATIONS -p^^TTvrTTTTTV^ 2,438 SEISMIC VELOCITY IN METERS/SEC. ^Sy^'7 VERTICAL EXAGGERATION = 5X Fig. 4 Deep seismic refraction profiles of Enewetalt Atoll. [Adapted from Raitt, 1957.} 48 RISTVET O o in o o o — 1 o o in GEOLOGY AND GEOHYDROLOGY 49 2 at c JS B o o < CQ ■a e H 5> 3 1^ c U (0 s < o 1 j: t r ■o I seaje J3je«-M0|ieys ado|S-(ueid 05 UJ 2 en X H lu a o o 1^ o o o o o in so RISTVET sisted of several closely spaced unconformities similar to the Pleistocene section described by Ristvet et al. (1977). Preliminary results from the PEACE program show the presence of at least four subaerial surfaces between 310 and 350 m subbottom depths. This suggests that the Mid- dle Miocene may have had episodic continental glaciation conditions similar to those well documented for the Pleistocene/Late Pliocene epKjchs. At least two of these unconformities show karstic features suggesting relatively long periods of subaerial exposure (B. R. Wardlaw, per- sonal communication). Resubmergence of the atoll occurred in Tertiary / time with the deposition of shallow-water sediments. From 210 to 252 m the sediments represent very organic-rich, nor- mal lagoonal, or shallow-water deposits. Preliminary PEACE data suggest that the sea level did not fall during this depositional interval. Lignitic material is scattered throughout the interval. Leopold (1969) reported a polli- niferous interval from the early deep holes from 205 to 270 m. This interval is interpreted as being a time when the atoll had rather large islands and large mangrove swamps developed (B. R. Wardlaw, personal communica- tion). At a depth of more than 210 m, the sediments indi- cate normal shallow-water deposition and a return to the small island configuration. Schlanger (1963) describes the presence of a major solution unconformity at a 90 m depth. Preliminary PEACE data show this to be the top of the Pliocene (T. M. Cronin, personal communication). A second subaerial exposure surface is recognized approximately 15 m below the 90 m solution unconformity. The PEACE drilling program has provided nearly con- tinuous sampling of the upper 490 m of the Enewetak sub- surface near its northern and northwestern lagoonal edges. Unfortunately, results of this recent drilling program are still forthcoming. Preliminary results of the PEACE drilling confirm the general interpretations made by Schlanger (1963); however, they provide a significant increase in the detailed understanding of the post-Lower Miocene strati- graphic section unavailable to Schlanger (1963) due to the poor sample recovery of the 1951 and 1952 drilling. It is anticipated that the PEACE results will lead to redefinition of the biostratigraphy, based both on foraminifera and ostracods of the post-Eocene of Enewetak and the Pacific in general. A detailed understanding of the Enewetak sea level history will also be forthcoming as well as additioneil insight into the processes of the diagenesis of carbonate sediments. The Quaternary subsurface of Enewetak is well-defined by the data obtained during the EXPOE drilling and is now further supplemented by the PEACE drilling. Five major unconformities were recognized by Ristvet et al. (1974, 1977); Goter (1979); and Szabo et al. (1985). Figure 7 presents an ocean reef to lagoon cross section across Engebi (Enjebi) Island constructed from the logs of EXPOE drill holes (Couch et al., 1975) and supplemented by data from the geologic rcdescriptions of several of these holes for the PEACE program (B. R. Wardlaw, personal com- munication). Each of the five unconformities represents paleosubaerial exposure surface and is marked by the pres- ence of paleosols (terra rosa type), soil base features (lam- inated crusts, rhizoconcrctions, etc.), and/or prominent changes in the mineralogical and chemical compwsition and cementation of the sediments. These five unconformities arc within the upper zone of leached and altered sediments described by Schlanger (1963). Because of the excellent core recovery during the EXPOE drilling, the identification of these Quaternary unconformities was easily made. Szabo et al. (1985) have dated three of the first four litho- somes using a variety of radiochemical techniques. These ages are shown in Fig. 7, The first unconformity at approximately 10 m depth is the Holoceiie/Pleistocene contact. Radiocarbon dates indi- cate that the Holocene sea transgressed the emergent plat- form reef by about 8000 ybp. The reef grew rapidly upward (about 5 to 10 mm yr~') until approximately 6500 ybp. Following 6500 ybp, vertical growth slowed to about 0.5 mm yr~' prompting lateral development of the reef (Szabo et al., 1985; Tracey and Ladd, 1974). As pre- viously discussed, sea level may have been nearly 1 m higher than present between 4000 and 2200 ybp. Current relative sea level rise at Enewetak may be near that of the long-term subsidence rate of 0.02 to 0.04 mm yr~' (Bud- demeier et al., 1975). Smith and Kinsey (1976) estimate that the present Enewetak reef has potential for upward growth of approximately 1 mm yr~^ The difference in growth potential versus modem relative sea level rise may explain why the reef plate is prograding lagoonward as noted by Ristvet et al. (1977) for the windward reef off Runit and Aomon. Pleistocene rocks in the lithosome directly below the first unconformity are dated at 131,000 ± 3000 ybp by Szabo et al. (1985) and 100,000 to 120,000 ybp by Thurber et al. (1965). There is also a significant change in the mineralogical and chemical composition of the sedi- ments below this first unconformity versus the Holocene sediments above. Ristvet et al. (1974) document the near total loss of high-magnesian calcite below this layer and significant decreases in the whole rock trace element con- centrations of Mg, Fe, and Mn. Calcitic vadose and phreatic carbonate cements are first encountered in this lithosome. The development of the unconformities and the associ- ated diagenesis of the underlying carbonate sediments are the result of relative sea level changes during the past. Periods of worldwide continental glaciations cause a lower- ing of sea level. At Enewetak during the Quaternary, this may have been in excess of 100 m (Walcott, 1972) during each major glacial advance. During these periods of sea level lowstands, the Enewetak Lagoon is above sea level and the atoll becomes a large, high carbonate island, resulting in severe changes to both the hydrologic regime and sediment production of the atoll. Because the reefs are subaerially exposed, only minor reef growth occurs as a fringing reef on the outer slopes of the atoll-island. The atoll-island undergoes subaerieJ erosion and soil develop)- GEOLOGY AND GEOHYDROLOGY 51 CO < UJ o o UJ LU Q UJ 2 O Z 3 Q UJ UJ 2 Q O 2 UJ o z o H o 2 o ffi tr. o o 3 o UJ UJ o X - Tl p- "c. -* I / ■a 1! ■8 c 10 •a to 1 S ■§ •a e u O) c u ID IS s O w 01 -isw Monaa Sd3i3kN ni Hidaa 52 RISTVET merit from colian sources. Such subaerial exposure to meteoric waters results in the development of an extensive Gyben-Herzberg lens within the island which is conducive to the alteration and cementation of the sediments. During atoll emergence, several processes acting alone or in various combinations can produce significant modifi- cations in those carbonate sediments exposed to meteoric waters. Most of these processes arc dependent on the initial dissolution of carbonate minerals into an aqueous phase. Subsequent precipitation of calcium carbonate may be caused by changes in carbon dioxide pressure, tempera- ture, evaporation, mixing of waters of differing ionic strength, and other mechanisms (Bathhurst, 1971). Precipi- tation appears to be highly variable in both space and time. It may be contempKsraneous with dissolution or may involve transport over large distances. The model prop>osed for the diagenesis of Enewetak sediments is similar to that profxssed for other carbonate sequences (Thorstenson et al., 1972). Meteoric waters passing through a soil approach equilibrium with the ambient CO2 pressure which is normally significantly higher than atmospheric CO2. These high CO2 waters promote dissolution of the metastable aragonite and magnesian calcite mineralogy of recent carbonate sedi- ments and approach equilibrium solubility. The saturated waters at a later stage encounter an environment of lower CO2, causing degassing of CO2 and the subsequent precipi- tation of calcite. The process of CO2 control on solution-precipitation of carbonates occurs within both the vadose and phraetic zones. At standard pressures and tem- peratures, the loss of high-magnesian calcite to calcite generally precedes the solution of aragonite and the con- current development of moldic porosity before the precipi- tation of calcite. As may be seen in Fig. 7, several [>eriods of atoll emergence have been followed by submergence during the Quaternary. For the Quaternary, it appears that following each sea level rise, the new depositional environment parallels that below the unconformity and buries it with new sediments as the platform subsides. The processes involved in subaerial diagenesis of the sediments during each f)eriod of emergence are multiple upon the older lithosome below each unconformity. In other words, for any depth within the meteoric vadose and phraetic regime, there is a potential for the solution reprecipitation process to occur as many times as there are subaerial exposures above that depth. This multiple diagenesis results in pro- gressive increases in cementation and mineral stability with increasing depth for at least the Quaternary section of the Enewetak subsurface. The Quaternary subsurface of Engebi (Enjebi) (Fig. 7) consists of a complex mosaic of depositional lithofacies, which have subsequently been affected by diagenetic processes. In general, cementation increases with depth and towards the reef within each stratigraphic unit. This lateral change in cementation and, as shown by Ristvet et al. (1974), corresponding changes in the rates of mineral stabilization and trace element petrochemistry may be in part due to (1) the occurrence of marine cements in those sediments near the reef flat versus those deposited lagoon- ward and (2) to diagenetic processes affecting the sedi- ments as a function of the paleohydrologic regime and the paleochemistry of the meteoric lens (Ristvet et al., 1977). Shallow seismic refraction surveys were conducted on windward, leeward, and transitional islands during EXPOE and yielded consistent profiles for the Quaternary Enewetak subsurface (Ristvet et al., 1977). As shown on Fig. 7, four distinct velocity intervals exist. The velocity in the unsaturated island sediments, Vq, is 330 to 600 m s~'; \Ji is the velocity in saturated, unconsolidated Holo- cene sediments and is typically about 1600 m s~^ The velocity in poorly to moderately cemented Pleistocene sedi- ments, V2, is typically 2500 m s ^ The V1/V2 interface corresponds to the first unconformity. The higher velocities of well-cemented sediments which occur on the reefward side of the island and at depths below 60 m as inferred from lithologic descriptions of drill holes are represented by V3 (Ristvet et al., 1977). The unconformities recognized by the drilling on the atoll edges may also be followed into the lagoon on seismic reflection profiles obtained during EASI and PEACE (Ristvet et al., 1980; Tremba et al., 1982; Tremba, 1985; Grow et al., 1986). Figure 8 is the interpretation of a seismic reflection record which is a lagoonward extension of the Engebi (Enjebi) reef to lagoon geologic cross section shown in Fig. 7. The seismic profile is perpendicular to the reef front and crosses the lagoonal terrace into the lagoon basin. In Fig. 8, the first reflector/refractor corresponds to the Holocene/Pleistocene unconformity at 15 m subbottom depth. The reflector at 66 m subbottom depth seems to correspond to a Pleistocene unconformity seen in the Engebi (Enjebi) drill holes. From the PEACE drilling, it is apparent that the deeper reflectors between 150 and 330 m correspond to lithologic changes and do not neces- sarily represent unconformities. The 330 m reflector does represent the top of a series of closely spaced reflectors corresponding to the Middle Miocene unconformities recog- nized in the PEACE boreholes. Of interest is that parallel- ism of the reflectors to the present bathymetry. This feature of the seismic records was noted atoll-wide for reflectors above the Middle Miocene unconformities helping to confirm the hypothesis that the present-day reef environments have shown little lateral migration since the Middle Miocene. GEOHYDROLOGY Studies of the hydrology of Enewetak Atoll were ini- tiated in 1972 to evaluate possible environmental effects of the proposed PACE high explosive craters on the ground- water resources of the islands (Koopman, 1973). Addi- tional studies sponsored by the DOE have been conducted as part of a program to determine the physical, chemical, and biological mechanisms controlling the distribution and transport of radionuclides in the atoll environment (cf. Bud- demeier and Holladay, 1977; Wheatcraft and Buddemeier, GEOLOGY AND GEOHYDROLOGY 53 UJ Q Z _l < CO CO CQ ^ / E — T «5 E E e CD E GO ID o CD rt- 00 o I - I (M O o I _I -■ to < o »- q: UJ > o o cvj o o rO O o N g •3 w E o e e « o i o a e o u 1 OB i ■g & w Sd3i3W ' Hid3a 54 RISTVET 1981; Buddemcier, 1981). An additional study was spon- sored by the DNA (Buddemeier and Jansen, 1976) to investigate the groundwater potential for use in the Enewetak Radiological Cleanup. Atkinson et al. (1981) investigated the water budget and circulation of water in the Enewetak Lagoon and found that essentially all the water input to the lagoonal system comes from wind-driven transport across the wind- ward reef. Since the windward reef crest is typically near mean sea level, waves drive water from the ocean into the lagoon at nearly all times. The windward reef blocks any return flow. Atkinson et al. (1981) determined that nearly all of the outflow occurs through the South Channel. The Deep Channel had a balanced inflow and outflow. Other input/output pathways, i.e., transport over the leeward reef was insignificant in comparison to input over the wind- ward reef and output through the South Channel. Atkinson et al. (1981) calculated a mean residence time for lagoon waters of 1 month with a maximum of 4 months for water in the northeast section of the lagoon. Although water lev- els were not directly observed, the circulation pattern requires the existence of a net lagoon to ocean gradient (Buddemeier, 1981). Koopman (1973) first noted that, for the islands of Enewetak, a significant discrepancy existed between the calculated thicknesses of a fresh water Gyben-Herzberg lens and that observed in trenches and borings in the field. Koopman (1973) observed that the islands of Engebi (Enjebi) and Aomon had only thin brackish water lenses approximately one-tenth as thick as would be predicted using conservative calculations. Buddemeier and Holladay (1977) measured tidal lags in wells on Engebi (Enjebi) Island and noted that there was a sharp discontinuity in the plot of tidal lag time versus depth between 10 and 20 m subsurface depth. They hypothesized that the effect might be due to a more p)ermeable aquifer below the first unconformity of Ristvet et al. (1977). Wheatcraft and Bud- demeier (1981) demonstrated, using tidal data from Engebi (Enjebi) Island, that the classical Gyben-Herzberg lens model does not describe the hydrologic system observed, which is controlled by vertical transmission of tidal signals from deeper and more permeable Pleistocene aquifer(s). Buddemeier (1981) noted that total fresh water content of island groundwater was essentially independent of island area and radius and that the southern islands have approx- imately 50% more fresh water volume than the northern islands. In addition to this difference in gross fresh water inventory, Buddemeier (1981) noted the northern islands have thinner layers of p>otable water and more extensive brackish water transition zones than do the southern islands. Buddemeier (1981) made additional tidal measure- ments on Japtan, Biken (Rigili), Enewetak, Aomon, and Engebi (Enjebi) Islands and concluded that significant differ- ences were present between the amplitudes of reef and lagoon tide stations on the falling tide resulting in a net lagoon to ocean head. Buddemeier (1981) concluded that this net head of water will tend to set up a lagoon to ocean flow of water through the permeable Pleistocene aquifer and that the amount and quality of fresh island groundwater is controlled by the rate of lagoon to ocean flow through the Pleistocene aquifer. The estimated lagoon to ocean transit times are on the order of 3 to 6 years, which corresponds well to the fresh water residence time estimates of the islands based on inventory and recharge. The rate of flow from lagoon to ocean dep>endency explains why islands in close proximity to reef channels, such as the southern islands, have greater volumes of fresh water than others. ACKNOWLEDGMENTS I wish to express much appreciation to Edward Tremba for critically reviewing this manuscript and many hours of stimulating discussion. I wish to acknowledge the fine technical support provided by J. MacCornack and L. D. Evans in preparing this manuscript. I express my apprecia- tion to the many personnel of the Defense Nuclear Agency, Department of Energy, Mid-Pacific Research Laboratory, University of Hawaii, Air Force Weapons Laboratory, U. S. Geological Survey, and Holmes and Narver, Inc., who have participated and supported the geo- logic investigations of Enewetak Atoll for the last 40 years. Funding for this effort was provided by the Defense Nuclear Agency. REFERENCES Agassiz, A., 1903, Reports on the Scientific Results of the Expedition to the Tropical Pacific, the Marshall Islands, Mem. Mus Comp. Zool . Haruard College. 28: 271-329. Atkinson, M. J., S. V. Smith, and E. D. Stroup, 1981, Circula- tion in Enewetak Atoll Lagoon, Limnol. Oceanogr., 26: 10741083. Bathurst, R. G. C, 1971, Carbonate Sediments and Their Diagenesis, Elsevier, Amsterdam. Buddemeier, R. W., 1981, The Geohydrology of Enewetak Atoll Islands and Reefs, in Proceedings of the Fourth International Coral Reef Siimposium, Manila, 1: 339-345. , S. V Smith, and R. A. Kinzie, 1975, Holocene Windward Reef Flat History, Enewetak Atoll, Geo/ Soc Am. Bull., 86: 1581 1584. , and R. C. Jansen, 1976, Groundwater Resources Eixiluation, Enewetak Atoll. Defense Nuclear Agency Project Report, Hawaii Inst, of Geophysics, University of Hawaii, Honolulu. , and G. Holladay, 1977, Atoll Hydrology: Island Groundwater Characteristics and Their Relationship to Diagenesis, in Proceedings of the Third International Coral Reef Svmposium. Miami, 2: 167-173. Chappell, J., and H. H. Veech, 1978, Late Quaternary Tectonic Movements and Sea Level Changes at Timor and Atauro Islands. Geol Soc of Am. Bull.. 89: 356-368. Cole, W. S., 1957, Large Foraminifera from Eniwetok Drill Holes, U. S. Geol Suru Prof Pap. 260 V, pp. 743-784. Colin, P. L., D. M. Devaney, L. Hlllis-Colinvaux, T. H. Suschanek, and J. T. Harrison III, 1986, Geology and Biolog- ical Zonation of the Reef Slope, 50-360 m Depth at Ene- wetak Atoll, Marshall Islands, Bull. Mar. Sci.. 38: 111-128. GEOLOGY AND GEOHYDROLOGY 55 Couch, R. F., Jr., J D Fetzer, E. R. Goter, B L Ristvet, E. L Tremba, D. R. Walter, and V. P. Wendland, 1975, Drilling Operations on Eniwetok Atoll During Project EXPOE. Air Force Weapons Laboratory Air Force Systems Command, Kirtland Air Force Base, New Mexico, AFWL-TR-75-216. Curray, J R, F P Shepard, and H. H. Veeh, 1970, Ute Quaternary Sea Level Studies in Micronesia: Camarsel Expedition, Geol Soc Am. Bull. 81: 1865-1880 Daly, R A , 1915, Pleistocene Glaciation and the Coral Reef Problem, Am. J. Sci.. 30: 297-308. Daniels, J. J., F. Clutsom, D. Glover, and B. L. Ristvet, 1984, Borehole Grauiti^ Suruet; of Drill Hole El. Enewetak Atoll. Marshall Islands, unpublished manuscript. Darwin. C , 1842, The Structure and Distribution of Coral Reefs. D Appleton, New York. David, T W E., G. H. Halligan, A. E. Finckh, and G. Sweet, 1904, The Atoll of Funafuti. Narrative of the Second and Third Expeditions. Royal Society, London. Davis, W M , 1928, The Coral Reef Problem, Am. Geog Soc. Spec Pub. 9. pp. 1-596. Dobrin, M B , and B Perkins, Jr., 1954, Seismic Studies of Bikini Atoll, U S Geol Surv. Prof Pap 260-J. pp. 487-506. Emery, K O , J 1 Tracey, Jr., and H. S. Ladd, 1954, Geology of Bikini and Nearby Atolls, U. S Geol Surv. Prof Pap. 260 A. pp 1-265 Folger. D. W., 1986, Introduction, Sea-Floor Observations and Subbottom Seismic Characteristics of OAK and KOA Craters. Enewetak Atoll. Marshall Islands. D. W. Folger (Ed.), U, S. Geol Surv. Bull , 1678: 1-7 Gardiner, J. S., 1931, Coral Reefs and Atolls. Macmillan, London. Goter, E. R., 1979, Depositional and Diagenetic History of the Windward Reef of Enewetak Atoll During the Mid to Late Pleistocene and Holocene, Ph.D. dissertation, Rensselaer Polytechnic Institute, Troy, N. Y. Gross, M G., J D Milliman, J. I Tracey, Jr., and H. S. Ladd, 1969, Marine Geology of Kure and Midway Atolls, Hawaii: A Preliminary Report, Pac. Sci., 23: 17-25 — , and J I. Tracey, Jr., 1966, Oxygen and Carbon Isotopic Composition of Limestones and Dolomites, Bikini and Enewetak Atolls, Science. 151: 1082-1084. Grow, J. A., M. W Lee, J. J. Miller, W Agena, J. C. Hampson, D. S. Foster, and R. A. Woellner, 1986, Multichannel Seismic-Survey of OAK and KOA Craters, Sea-Floor Observa- tions and Subbottom Seismic Characteristics of OAK and KOA Craters. Enewetak Atoll. Marshall Islands. D. W. Folger (Ed.). U. S. Geol. Surv Bull., 1678: D1-D46 Halley, R. B , and R. A. Slater, 1985, Geologic Reconnaissance of the Natural Forereef Slope and a Large Rockfall/Slump Exposure. Enewetak Atoll. U. S. Geol. Surv., unpublished manuscript. Manor, J. S , 1978, Precipitation of Beachrock Cements: Mixing of Marine and Meteoric Waters vs. C02-Degassing, J Sied Petro. 48:489-501. Henny, R. W., J. W. Mercer, and R. T. Zbur, 1974, Near Sur- face Geologic Investigations at Eniwetok Atoll, in Proceedings of the Second International Coral Reef Symposium. Great Bar- rier Reef, Australia, 1: 615-626. Keller, F., Jr., 1954, Aeromagnetic Surveys in the Aleutian, Marshall, and Bermuda Islands, Trans. Am. Geophys Union., 35: 558-572. Koopman, F C, 1973, Ground Water Conditions on Eniwetok Atoll, with Comments on Effects of Use of Conventional Explosives. U S. Geol. Surv., Water Resources Div., unpub- lished rept. to Air Force Weapon Laboratory, Kirtland Air Force Base, New Mexico Kulp, J L , 1963, Potassium-Argon Dating of Volcanic Rocks, Bull. Volcaque. 26: 247-258. Ladd, H S., J. I. Tracey, Jr., and G. G. Lill, 1948, DriUing on Bikini Atoll, M. 1., Science, 107: 51-55. E Ingerson. R. C. Townsend, M. Russell, and H K Stephenson, 1953, Drilling on Eniwetok Atoll, M. I., Am Assoc. Petroleum Geologists. 37: 2257-2280. and S. O Schlanger, 1960, Drilling Operations on Eniwetok Atoll, U. S Geol. Surv. Prof. Pap. 260 Y. pp. 863-903. , J I Tracey, Jr , and M. G. Gross, 1970, Deep [drilling on Midway Atoll, U. S. Geol Surv Prof Pap. 680 A. pp. 1-22. Leopold, E B , 1969, Miocene Pollen and Spore Rora of Eniwetok Atoll, Marshall Islands, U. S Geol Surv. Prof. Pap. 260 II. pp 1133-1184 McMurtry, G M , R C Schneider, P L. Colin, R. W. Bud- demeier, and T H Suchanek, 1985, Redistribution of Fall- out Radionuclides in Enewetak Atoll Lagoon Sediments by Callianassid Bioturbation, Nature. 313: 674-677. Munk, W H , and M. C. Sargent, 1954, Adjustment of Bikini Atoll to Ocean Waves, U S Geol Surv Prof Pap 260 C. pp. 275-280 Odum, H. T., and E P Odum, 1955, Trophic Structure and Pro- ductivity of a Windward Coral Reef Community on Eniwetok Atoll, Ecol. Monogr.. 25: 291-320. Pratt, H. R., and H. R. Cooper, Jr , 1968, The Near Surface Geology at Eniwetok and Bikini Atolls. Air Force Weapons Laboratory. Air Force Systems Command, Kirtland Air Force Base, New Mexico. AFWL-TR-68-68. Raitt, R. W., 1954, Seismic-Refraction Studies of Bikini and Kwajalein Atolls, U. S Geol Surv Prof Pap 260 K. pp. 507-528. , 1957, Seismic-Refraction Studies of Eniwetok Atoll, U. S. Geol. Surv. Prof Pap 260 S. pp. 685-698. Ristvet, B L., R. F. Couch. J. A Fetzer, E. R. Goter, E L. Tremba, D R. Walter, and V. P Wendland, 1974, A Quaternary Diagenetic History of Eniwetok Atoll, Geol. Soc Am.. Abstracts, 6: 928-929 , E. L. Tremba, R F Couch. Jr.. J. A. Fetzer, E. R. Goter. D. R Walter, and V. P Wendland, 1977, Geologic and Geo- physical Investigations of the Enewetak Nuclear Craters. Final Report. Air Force Weapons L-aboratory, Air Force Systems Command, Kirtland Air Force Base, New Mexico, AFWL- 77-242. , R F. Couch, and E L Tremba, 1980, Late Cenozoic Solu- tion Unconformities at Enewetak Atoll, Geol. Soc Am.. Abstracts, 12: 510 Sailer, A. H., 1984, Petrologic and Chemical Constraints on the Origin of Subsurface Dolomites, Enewetak Atoll: An Example of Dolomitization by Normal Seawater, Geology, 12: 217-220. Sargent, M C , and T S. Austin, 1954, Biologic Economy of Coral Reefs, U S. Geol Surv. Prof. Pap 260 E, pp 293-300. Schlanger, S. O., 1963, Subsurface Geology of Eniwetok Atoll, U. S. Geol. Surv. Prof. Pap. 260-BB, pp. 991-1066. Schmalz, R. F., 1971, Formation of Beachrock at Eniwetok Atoll, Carbonate Cements, O. P Bricker (Ed), Johns Hopkins Press, Baltimore, pp. 17-24. Smith, S. v.. and J T. Harrison. 1977. Calcium Carbonate Pro- duction of the Mare Incognitum, the Upper Windward Reef Slope, at Enewetak Atoll, Science, 197: 556-558. 56 RISTVET , and D W. Kinsey, 1976, Calcium Carbonate PrcxJuction. Coral Reef Growth and Sea Level Change, Science, 194: 937939. Stearns, H T , 1945, Decadent Coral Reef on Eniwetok Island, Marshall Group, Geo/. See Am Bull . 56: 783-788. Szabo, B. J., J I. Tracey, Jr., and E. R. Goter, 1985, Ages of Subsurface Stratigraphic Intervals In the Quaternary of Enewetak Atoll, Marshall Islands, Quaternary Res . 23: 54-61. Thorstenson, D C, F T Mackenzie, and B. L Ristvet, 1972, Experimental Vadose and Phraetic Cementation of Skeletal Carbonate Sand, J Sed Petro . 42: 162 167 Thurber, D L , W S Broecker, R. L Blanchard, and H A Potratz, 1965, Uranium Series Ages of Pacific Atoll Coral, Science. 149: 55-58 Todd, R., and D Low, 1960, Smaller Foraminifera from Eniwetok Drill Holes, U S Geo/. Suru. Proj^ Pap 260 X. pp. 162. Tracey, J 1., Jr , and H S. Ladd, 1974, Quaternary History of Eniwetok and Bikini Atolls, Marshall Islands, Proceedings of the Second International Coral Reef Syrnposium. Great Barrier Reef, Australia, 2: 537 550 Tremba, E L., 1985, Project EASI: Phase III. Air Force Weapons Laboratory, Air Force Systems Command, Kirtland Air Force Base, New Mexico, AFWLTR 84-105 R. F. Couch, Jr , and B. L Ristvet, 1982, Enewetak Atoll Seismic Investigation (EASI) Phases I and 11, Air Force Weapons Laboratory, Air Force Systems Command, Kirtland Air Force Base, New Mexico, AFWL-TR-82-20. Walcott, R I , 1972, Past Sea Levels: Eustasy and Deformation of the Earth, Quaternary Res , 2: 1-15 Wells, J W , 1954, Recent Corals of the Marshall Islands, U S Geo! Surv Prof Pap 2601. pp. 385-486 Wheatcraft, S. W., and R. W. Buddemeier, 1981, Atoll Island Hydrology, Grounduiater. 19: 311-319. Chapter 5 Oceanographx; of Enewetak Atoll MARLIN J. ATKINSON Zoology Department. Uniuersiti; of Western Australia Nedlands, Western Australia INTRODUCTION Enewetak and Bikini lagoons are large, deep atoll lagoons. The circulation systems of both lagoons are dom- inated by wind-driven currents (von Arx, 1948, for Bikini, Atkinson et al., 1981, for Enewetak). However, the full dynamics of the Enewetak circulation system is explained by a combination of wind-driven currents, slope currents from water input by waves, and tidal currents. The findings at Enewetak indicate that the internal circulation and flush- ing of deep atoll lagoons is affected by atoll morphology and local wave and tidal conditions, features which in gen- eral control circulation in shallow atoll lagoons (Mllliman, 1967; Gallagher et al., 1971; Henderson et al., 1978; Ludington, 1979). This chapter begins with the general oceanography of the northern Marshall Islands and then concentrates on the oceanography of Enewetak Lagoon. The oceanography of Bikini Lagoon and vicinity have been well studied com- pared to Enewetak. In this chapter frequent comparisons are made between Enewetak and Bikini. NORTH EQUATORIAL CURRENT Enewetak resides in the North Equatorial Current. In the region of the Marshall Islands, the current Is between 6° to 8° and 15° to 17° north latitude. The southern boundary of the current moves northward with the sun during northern hemisphere summer and shifts back toward the south In winter. The current has a general westward drift between 20 to 50 cm s"^ Surface water is isothermal to 75 m and varies seasonally between 26° and 29°C. The main thermocline is between 150 and 300 m with a temperature of 10°C at 300 m. By 1500 m the temperature drops to 3°C. Between the region of 3° to 11°N the salinity Is rela- tively low (34.1 to 34.5 °/oo) reflecting the annual net rainfall in the region and the eastern flow of the Equatorial Countercurrent. Higher salinities occur to the north of 11° (the latitude of Enewetak) due to increased evaporation. Isohalines show development of Intermediate Water about 11°N. Figure 1 shows the temperature-salinity relation- ships of the Western North Pacific Central Water and Pacific Equatorial Water in the region of the northern Marshall Islands (Barnes et al., 1948). The solid lines In the diagram indicate the temperature-salinity correlations at different latitudes; at 20°N the water is all North Pacific Water and at 4°N it is all Pacific Equatorial Water. The Insert In the diagram shows the depth of transition zones between the two water masses. Enewetak Atoll resides In the region where the transition zone is above 200 m and is only 50 m thick (Barnes et al., 1948). Dynamic topography near Enewetak has never been measured. However, data collected during the Operation Crossroads project and by the Japanese indicate dynamic topography to be complex near Bikini, with the presence of eddies northwest and northeast of the atoll (Barnes et al., 1948). Rather permanent eddies probably exist near Enewetak because they do for other islands (Hamner and Hauri, 1981). The complexity of dynamic heights suggests that currents near the atolls may vary in both sp)eed and direction. WAVES AND TIDES Waves formed by the northeast trade winds break on the northern and eastern reef perimeters of the atoll. This constant pounding of the fore-reef shapes the spurs and grooves on the windward side. For Bikini Atoll, Munk and Sargent (1949) used wind data to calculate wave direction, wave height, and wave energy. The spur and groove sys- tems on the windward side of Bikini dissipate 95% of the calculated wave energy as frictlonal heat and channel 5% of the energy upward to maintain a head of water on the reef flat. The head of water establishes water flow from the ocean to lagoon across the windward reef flats. Waves within the lagoon are generated by local wind patterns and have little Influence In shaping the reef structures, but they do Influence sand transport. 57 58 ATKINSON Salinity % 20' 34.0 15' O o O 3 a E .* 10< 35.0 — I — crt26 4°N /' O^^' / ff/ // 10°N /Vo / Ot 27.0 600 M -800 I PE Fig. 1 Temperature-salinity relationships of Western North Pacific Central Water (WNPC) and Pacific Equatorial Water (PE) in the region of the northern Marshall Islands. The solid lines show the temperature-salinity relationship at different latitudes. The insert shows depth of transition zone between the water masses. [Drawn from Barnes et al.. 1949.] Breaking waves on the fore-reef and the back-reef determine sand transport in the following ways Cross-reef currents carry sand from the fore-reef and the reef flat to the lagoonward rim of the reef, building and eroding islands. Ephemeral sand spits develop on the margins of the islands; this sand is sorted and distributed by long- shore transport from lagoon waves and back-reef currents. Two general patterns of sand grain size have been deter- mined for Bikini (Emery et al., 1954): (1) grain size increases across the reef flat from ocean to lagoon, then inside the lagoon, grain size decreases until a depth of approximately 15 m is reached; and (2) grain size decreases from the middle of the seaward beaches toward the ends of the islands and decreases from each end of the island to the middle of the lagoon beach. Two processes are apparently responsible for the distri- bution of sand: (1) high energy cross-reef currents carry a large suspended load, depositing sand as they slow down; and (2) the continual breaking of seaward and lagoonward waves on the islands transports sand along the shore. The high energy currents are formed from oceanic swells breaking on the fore-reef, and the long-shore currents are formed by lagoon wind waves breaking on the beach. In the future, sand transport by currents at Enewetak could be studied as a function of wind speed and direction, surf height, and swell direction. The tides at Enewetak Island are usually in good agree- ment with the U. S. Navy Tide Tables. However, lagoon and ocean tide records show differences in amplitude, tim- ing, and tide curve shape. When the reef is awash at Enjebi, wave setup produces ocean tides with a mean water level 0.3 to 0.5 m above the mean lagoon level; at Enewetak Island, the differences are small but significant (Buddemeier, 1981) (Fig. 2). Buddemeier also analyzed long-term differences between the Japtan gauge and a lagoon gauge at Biken (see Fig. 2 for location). His analy- sis showed that while the tide records were similar in amplitude and frequency composition, the Biken highs are broadened and the lows narrowed by about 1 hour. Based on an average difference in tide elevation betwfeen Japtan OCEANOGRAPHY 59 Enewetak Isl. Fig. 2 Location of islands and currents. See text for discus- sion. [From Atkinson et al. with permission.] reef flats are shallow (0 to 2 m deep), and the deepest part of the Deep Entrance is about 57 m. Because the sur- face North Equatorial Water is well mixed to a depth of 75 m, water flowing into the lagoon, either over the reefs or through the channels, is well-mixed ocean surface water. Salinity, temperature, dissolved inorganic plant nutrients, and dissolved carbon dioxide suggest little stratification within the lagoon water (Table 1). For the data at hand, surface water (0 to 10 m) appears to be slightly cooler (0.2°C) and less saline (0.06 °/oo) than deeper water. During the data collection period (July 26 to August 16, 1974), the weather was unusually rainy and cool (S. V. Smith, personal communication). August is a period of low wind; therefore, stratification should occur most dramatically during this month, yet no major stratification is evident in these data. There is only a slight indication of the rain in the surface water. Nutrient data collected by S. V. Smith and M. J. Atkinson during June 1979 in the lagoon and passages also showed no vertical structure. Several detailed nutrient profiles taken between to 2 m above the bottom, at 10 cm intervals, revealed extremely low and unchanging concentrations. Vertically averaged phosphate and nitrate-nitrite concentrations are contoured and suggest a weak minimum toward the center of the lagoon (Fig. 3). and Biken, Buddemeier estimated that the water level is about 6.5 cm higher at Biken than at Japtan. CURRENTS LAGOON WATER The lagoon at Enewetak is well isolated from the gen- eral westward flow of the North Equatorial Current. The Cross-Reef Currents Cross-reef currents involve shallow flow over the wind- ward and leeward reef margins of the atoll. The area of TABLE 1 Chemical Data for Ocean and Lagoon Water* Temp (°C) Sal ("/oo) Total alk (eq m^ PO« NO, NH^ (mmoles m ) Si pH Total CO2 Pco, (moles m ^) (uatm) Ocean Surface 29.5 34 30 2.29 0.12 0.21 0.50 44 8.31 1.88 297 S. Dev. 0.2 0.04 0.03 0.05 0.12 0.20 2.3 0.02 0.04 22 N. 7 7 7 6 6 5 7 7 7 7 Lagoon Surface (0.10 m) 29.6 34.22 2.29 0.14 0.15 0.37 3.4 829 1.89 313 S. Dev. 0.5 0.19 0.05 0.05 0.08 0.21 1.9 0.04 0.04 31 N. 125 117 118 116 111 101 118 118 116 116 Mid-depth (10 to 30 m) 29.8 34.28 2.31 0.13 0.13 0.28 3.4 8.25 1.90 311 S. Dev. 0.5 0.08 0.05 0.04 0.07 0.17 1.7 0.04 0.03 38 N. 56 56 56 56 55 53 56 56 56 56 Deep (30 to 50 m) 29.7 34.28 230 0.15 0.18 0.27 34 8.27 1.91 336 S. Dev. 0.3 0.08 0,04 0.05 0.12 0.19 1.8 0.03 0.03 26 N. 28 27 28 27 27 26 28 28 28 28 'Data collected by S. V. Smith, July and August 1974. 60 ATKINSON (a) (b) Fig. 3 Phosphate (a) and nitrate (b) contours for lagoon water. Concentrations in mmole m^'. windward cross-reef currents is shown in Fig. 2 by horizon- tal lines along the eastern boundary of the atoll. These currents are a result of breaking waves on the windward reef; they vary in response to surf height (and therefore regional wind patterns) and tide height. From Enjebi to Enewetak Island (Fig. 2), water crosses the reef from the ocean to the lagoon in a direction approximately perpen- dicular to the reef front. The windward cross-reef currents do not reverse direction, flowing from lagoon to ocean. The current speed ranges from 10 to 150 cm s ^'. These currents range in volume transport from 0.05 m s ' per meter of reef front during low tide and low surf to about 1.5 m'^ s ' during high tide and high surf. A mean volume transport value of 0.56 m s~ m~ was calculated; this is equivalent to 6.6 X 10 m per tidal cycle across the windward reef (a tidal cycle is used to facilitate comparison with other volume transports). The volume transport of the entire windward cross-reef current probably varies by a fac- tor of 2 to 3. Winter tropical storms drive water over the reef in massive amounts, building and eroding atoll islands. The area of leeward cross-reef currents is represented by vertical lines in Fig. 2. These currents do not flow in any well-developed pattern. Transport along the leeward reef, rather than across it, is common. During a period of high surf from a north swell, S. V. Smith and E. D. Stroup (January 1976, unpublished data) measured inward flowing currents along the northwest leeward reef. Current speeds and volume transports at 10 different loca- tions ranged from about 15 to 50 cm s~ and from 0.15 to 0.57 m'^ s^'m"\ respectively. Significant inflowing and outflowing currents were measured in the region north of Kildrenen Island and south of the Southwest Passage (Fig. 2). Noshkin et al. (1974), using surface concentra- tions of 2^*Pu, 239pu, ^'*°Pu, and ^^^Cs, have also shown that, during winter high tide periods, significant amounts of oceanic water enter the lagoon across the northwest and southwest leeward reef. Dye releases on the leeward reef flat demonstrate a slow drift either oceanward or lagoonward over the lee- ward reef margin A maximum value for oceanward flow might be the speed of the net oceanic drift of the lagoon surface current. A characteristic current speed over the entire leeward reef might be about 50% of the lagoon sur- face current speed. Much of the reef margin bares at low tide, so an estimate of the average depth of the reef is near 50% of the mean tidal range, or 0.42 m. The net transport of water out of the atoll over the leeward reef margin is estimated to be 0.4 X 10^ m'' per tidal cycle (i.e., only 6% of the windward reef input). Channel Currents The Deep Entrance current (Fig. 2) reverses approxi- mately every 6.2 hours, with the tide. The current speed ranges from to 80 cm s~\ increasing from zero to a maximum in about 3 hours and decreasing to slack water in another 3 hours. The period of slack water in the chan- nel is no more than a few minutes. The direction and speed of the current are nearly constant throughout the water column. The volume transport of the Deep Entrance current varies between neap and spring tides. On May 31, 1979, near maximum spring tide, the current transported 3.0 X 10* m'^ per 6.2 hours over an entire tidal cycle. OCEANOGRAPHY 61 Surface and deep drogues placed in the channel on a flooding tide reversed on the ebbing tide and returned to near their original position. It was estimated that the net volume transport of the Deep Entrance is approximately zero over a tidal cycle. The South Channel (Fig. 2) has a nearly continuous outflow. During flood tide, surface water drifts westward across the channel; on ebb tide, the surface current turns from westward, to southwest, to south. The surface water in the South Channel tends to move westward as wind drift, while water below a depth of 5 m moves southwest to south, depending on the tide condition. The current speeds range from 8 to 30 cm s^'. Based on 20 drogue measurements and dye releases over a complete tidal cycle, the average outflow was estimated to be 6.9 X 10 m'^ per tidal cycle; this represents 75% of the total lagoon surface current volume transport and 105% of the estimated water flowing inward over the windward reef flats (Table 2, and material presented later). The Southwest Passage is a shallow break in the lee- ward reef, yet it has a reversing current similar to the Deep Entrance (Fig. 2). The calculated volume exchange between ocean and lagoon is approximately 0.8 X 10* m^ per tidal cycle. Because the currents are reversing (see pre- vious discussion), the net outflow through these channels is small in terms of the water budget. The calculation of volume transport of water over the entire leeward reef includes this net outflow through this passage. Lagoon Currents Currents of the central lagoon may be characterized by a surface current, a mid depth current, and a deep current. The currents are distinguishable by their characteristic speed and direction The water column in Enewetak Lagoon is nearly isohaline and isothermal; salinity ranges 0.20 °/oo at most (average near 34.4 °/oo), and tempera- ture varies by no more than 0.5°C (annual range, 27°C to 29°C). The surface current is wind-driven. The general surface drift is southwesterly, or downwind (Fig. 4). The spatial and temporal variations in the current directions are considerable. In the central lagoon, drogues move south, west, and north, appearing to respond to the wind direc- TABLE 2 Water Budget; Estimates of Mean and Range Current Transport (range) 10* m^ per 12.4 hour (+ is to lagoon; — is from lagoon) Bases for calculation Comments on current Windward cross-reef Leeward cross-reef Deep Entrance South Channel Southwest Passage Surface Mid-depth Deep -6.6 ( + 2 2 to +198) -0.4 (0 to -0.8) Net = (-1.0 to +1.0) (3.0 X 10* m^ transport each way) -6.9 (-4 5 to -8.5) Net = (-0.2 to -0.2) (0.8 X 10* m^ transport each way) 9.2 (3 to 30) 8.6 (unknown) 2.2 (unknown) 0.56 m s ' m ' reef front 27,000 m open reef front 0.05 ms"' 0.4 m (half tide range) 47,000 m open reef front 40 ms^' 34,000 m^ (cross-sectional area) 0.15 (0.07 to 0.23) m s"' 145,000 m^ (cross-sectional area) (2/2)"* (conversion to normal direction) 0.40 ms"' 9,000 m^ (cross-sectional area) 0.06 m s~' 10 m X 34,600 m (maximum cross- sectional area at 5 m depth) 0.03 m s"' 20 m X 32,000 m (maximum cross- sectional area at 20 m depth) 0.01 ms"' 18 m X 28,000 m (maximum cross- sectional area at 39 m) June 21-29, 1971 continuous inflow. Variable flow. None to fast channel currents. Reversing. Typical tidal currents to 0.80 m s" Pulsing, continuous outflow. Reversing. Typical tidal currents. Variable. Functional to wind speed Variable. Functional to wind speed. Variable. Functional to windward cross-reef inout 62 ATKINSON NE TRADES //^/// Fig. 4 Lagoon surface currents from drogue data. Arrows represent smoothed drogue tra- jectories over varying lengths of time; they arc not vectors. Some drogue runs were made during calm or variable wind. [From Atkinson et al. with permission.] tion of the previous 6 to 12 hours. Drogue paths over 6 hours showed no rapid changes in directions; however, they often traced slight curves, suggesting they were slowly changing direction with the wind. The speed of the surface current is approximately 2% of the wind speed (Fig. 5). Data for Fig. 5 were taken from days when both the average wind direction and the average wind speed had small standard deviations. Correla- tions between wind and current speed and wind and current direction for all data are poor, probably because cross-reef currents and tidal channel currents influence the surface current at least 5 km into the lagoon. The surface current moves in a layer which varies from 5 to 15 m thick. The average thickness of the surface layer is approximately 10 m. Downwind volume transport of the surface layer is approximately 9.2 X 10 m per tidal cycle (Table 2). Von Arx (1948) reported that the sur- face current at Bikini is 5 to 20 m thick and changed depending on the wind conditions. The mid-depth current lies between 10 and 30 m in depth This current generally flows northeastward, oppo- 20 Current = 0.0246 -^ 2.06 (Wind) r2 = 0.96 _L- -o _1_ 1 2 3 4 5 6 7 Wind speed (m s-i) Fig. 5 Surface current speed as a function of wind speed for the center of the lagoon. OCEANOGRAPHY 63 site the surface flow, at speeds of 2 to 4 cm s The volume transport of this current is approximately 8.6 X 10* m^ per tidal cycle (Table 2). The deep current flows southward between 30 and 50 m. This current is slow, ranging from 0.5 to 1.5 cm s Drogues in this current were followed for up to 10 days; while the cumulative direction and speed were consistent and predictable, a 6- to 12-hour east-west variability ("slosh") was noticed in their movement. This motion was attributed to flow around the lagoon pinnacles and/or tidal pulsing. The volume transport of the deep current is approximately 2.2 X 10* m^ per tidal cycle through a cross section near the middle of the lagoon (Table 2). Vertical Current Profiles in Lagoon Figures 6a and 6b are photographs of vertically suspended fluorescine dye dispensers. These profiles reveal the spiral current structure of the lagoon water. The pro- files show that the deeper currents are offset to the right (clockwise) from the shallower currents. Figure 7 is a graphic summary of the vertical dye profiles in the lagoon The number at the end of each arrow is the depth in meters of the observation. The arrows have no magnitude because current speeds were not determined. In all stations across the lagoon, from Runit to West Spit (Fig. 7), the current spiraled to the right, forming a substantial east- ward flow which is referred to as the mid-depth current. At two stations, deep scuba dives were made to verify the southern flow of the deep current previously documented with the deep current drogues. Figure 8 is constructed from all the deep-drogue mea- surements and selected surface-drogue measurements made during the summer and winter periods. These data points represent end points of current vectors emanating from the origin. The shaded spiral indicates the resulting current structure. There arc not sufficient data to resolve the spiral more accurately. The spiral reveals the basic three-current system: the surface current (0 to 10 m) is southwesterly, the mid-depth current (10 to 30 m) is northeasterly, and the deep current (30 to 50 m) flows southward. The vertical current structure, as summarized by Fig. 8, can be altered by the cross-reef currents and tidal currents. The diagonal lines in Fig. 9 delineate the area of the lagoon directly affected by the windward cross-reef currents. At the northern end of the lagoon (region 1 in Fig. 9) these currents follow the contour of the atoll. Along the central part of the windward back-reef (region 2) the current may be going north, west, or south, depending on the tide and surf conditions. Near Enewetak Island (region 3) the current also follows the contour of the atoll. The surface current and deep currents in regions 1 and 2 move in the same direction when large volumes of water cascade over the reef. During spring low tide, however, when little water enters the lagoon over the reef, a surface current, mid-depth current, and deep current characteristic of the open lagoon can be observed (Fig. 7). The currents directly behind the windward reef are variable in speed, being fastest when large surf drives water into these regions. Figure 10 is a plot on two suc- cessive days, showing the current increase with rising tide. Notice that the second date had higher surf and a slightly higher wind speed than the first date. The two linear regression coefficients are significantly different at the 95% significance level. These data were taken at the site denoted "A" in Fig. 9. The cross-hatched areas in Fig. 9 delineate water that experiences reversing current through the Deep Entrance and the Southwest Passage. The area near the northwestern leeward reef, marked by circles in Fig. 9, is an area of convergence. The lagoon surface water cannot escape over the leeward reef, particularly when large surf drives oceanic water over these reefs into the lagoon. Large aggregations of jellyfish have been observed in this region, as well as strong southwesterly flow along the lagoonward margin of the reef. WATER BUDGET Table 2 is a summary of the volume transports for the important components of the water budget. Input The water can flow into the lagoon from the windward reef, the Deep Entrance, and the Southwest Passage. The windward cross-reef current transports about twice as much water as the Deep Entrance current. Because the windward cross-reef current never reverses, the volume transport over the windward reef represents net input of water into the lagoon. The Deep Entrance and the Southwest Passage show net transports of approximately zero over each tidal cycle. Output The water can flow out of the lagoon from the leeward reef, the Deep Entrance, and the Southwest Passage. Because the Deep Entrance and the Southwest Passage have net transports near zero over each tidal cycle, the net inflow from the windward reef must exit as outflow over the leeward reef and out the South Channel. Because the flow over the leeward reef is relatively small (Table 2), most of the water flows south, exiting out of the South Channel. The numbers in Table 2 do not sum to zero over a tidal cycle; however, these data were collected during dif- ferent tide stages. Ranges were included in the table to indicate the natural variability of the system. CIRCULATION MODEL Lagoon circulation can be explained as a response to three sources of energy: (1) the surf on the windward ocean reef, (2) the wind, and (3) the tides. 64 ATKINSON Fig. 6 Photographs a and b show right-handed vertical current profile. Photograph c shows the left-handed vertical current pro- file 2 km north of Enewetak Island. Direction of current and depth of dye dispenser are shown in the line drawing below the pho- tographs. See Fig. 7 for location. [From Atkinson et al. with permission.] Surf The breaking waves on the windward reef drive water over the windward reef flat and into the lagoon primarily on the eastern (prevailing windward) side of the atoll. The cross-reef currents and the currents behind the reef are, therefore, dependent on the surf height and the depth of water on the reef. This oceanic water spreads into the lagoon, moving downwind and mixing vertically and hor- izontally. Since the South Channel is the only significant region of outflow, the water column has a net transport to the south. This southerly net volume transport must OCEANOGRAPHY 65 2.4 - ^ J T ■ 16 2 4 6 Fig. 7 Vertical current profiles. Number at end of arrow gives depth in meters. Arrow gives direction of the current. Circled number gives the depth of the bottom in meters. [From Atkinson et al. with permission.] TRADES Surface current (0-10m) Deep-water current (30-50m) _L _L _L cm s-1 A 0-10m A 10- 20m O 20- 30m • 30- 40m ■ 40- 50m Fig. 8 Summary of drogue results. The shaded spiral represents the approximate endpoints of current vectors from the origin. [From Atkinson et al. with permission.] Windward Reef jy peep ;^;; Entrance South Channel Fig. 9 General current patterns in the lagoon. See text for key. increase toward the south end of the atoll to acconnmodate the inflowing water across the windward reef. Figure 11 shows the relative increase in southerly volume transport versus distance from the north end of the atoll. These relative volume transports are based on cumulative transport across the windward reef. These are reported as relative values, since variation in swell and wind direction alter this relationship an unknown amount. Wind The wind creates the downwind drift of the surface water and the upwind drift of the mid-depth water. These can be qualitatively described as a special case of Ekman wind-driven circulation. This pattern is superimposed on the net drift of the entire water column toward the South Channel. This southerly drift can be observed in the deep water, below surface layers affected by the wind. The northern end of the atoll has a relatively small southerly drift based on cumulative net input (Fig. 11); therefore, the effect of the wind can be observed at a deeper depth in the north end than in the south end of the atoll. Drogues suspended at 38 m in the north end moved northeast to east, whereas drogues at a similar depth in the southern end moved south. The increasing volume transport from north to south, due to the increasing net input over the windward reef, creates a southerly deep current that thickens toward the southern end of the lagoon. Figure 12 is a plot of the 38 m drogue directions versus distance from the north end of the atoll. By the middle of the lagoon, the layer at 38 m is well within the southerly deep current (Fig. 12). 66 ATKINSON w 10 - E o a c 0) o 270 1200 0.15 1300 1400 1500 Time of day (h] 1600 1700 1.5 Tide height (m) Fig. 10 Change in current speed as a function of tide helgfit near Runit on successive days. Average current speed plotted at midpoint of time interval. HIGH > < 3 O > < L^ 1 I I . 10 20 30 DISTANCE (km) 40 Fig. 11 Increase in cumulative net input over the windward reef flat as a function of distance from the north end of the lagoon. 360 - -o 090 O UJ 180 - 270 10 20 DISTANCE (km; Fig. 12 Change in direction of drogues suspended at 38 m as a function of distance from the north end of the lagoon. The observed pattern of wind-driven currents resembles in many ways the pattern predicted by Ekman for an enclosed sea in which the following conditions apply: (1) impermeable, closed boundary; (2) constant, unidirectional windstress over the entire surface; (3) homogeneous water; (4) uniform depth; and (5) constant eddy viscosity. At Enewetak these conditions are only partially met. The lagoon rim is closed neither to leeward nor windward; in particular, large quantities of water are introduced along the windward edge (Table 2). In a fully enclosed sea, the Ekman flow integrated over depth is zero at every point. In a lagoon such as Enewetak this will not be the case, but the detailed effects of the "leaky" boundary and the irregular bathymetry have not been estimated from present data. The remarkably shallow spiral pattern of currents is a new finding which should be further investigated and modeled. Surface current speeds are 5 to 20 cm s~', approxi- mately 2% of the wind speed. The surface drift is generally downwind and seems responsive to the wind direction of the previous 6 to 12 hours. The mid-depth upwind current speeds are about one-half of the surface current speeds. These wind-driven currents would cause the surface water to overturn in 5 to 10 days if there was no vertical mixing. Von Arx (1948) estimated approximately the same time for turnover at Bikini. Von Arx (1949), Munk et al. (1949), and Ford (1949) suggested that the surface water at Bikini sinks in the western portion of the lagoon and upwells in a small band in the eastern portion of the lagoon. No direct evidence of upwelling has been found at Enewetak. Upwelling, if it exists as such, will be largely intermittent, because of the intermittent (tidal) pulsing of OCEANOGRAPHY 67 the windward cross-reef inflow of surface water. At a max- imum (high tide, active surf), this inflow is approximately equal to the downwind transport of the lagoon surface layer; during these intervals, upwelling is not required by continuity to supply the wind-driven surface transport. The essentially vertical homogenous water in Enewetak Lagoon suggests that surface water mixes with bottom water before reaching the leeward side. It also does not allow any conclusions regarding the presence or absence of upwellings from distributions of water properties. At Bikini, Ford (1949) was able to follow the motion of discrete water distinguished by salinity variations. Surface, mid-depth, and deep water salinities at Enewetak are shown in Figs. 13a, b, c. These salinity con- tours show some of the general features of lagoon circulation. Surface water was slightly less saline than deep (b) (c) Fig. 13a, b, c Salinity for surface (0 to 10 m). mid-depth (10 to 30 m). and deep (30 to 50 m) water. (Collected by S. V. Smith, July 26 to Aug. 16, 1974.) 68 ATKINSON water because the weather was rainy during the collection period. Relatively high salinity ocean water cascades over the windward reefs and flows in through the Deep Entrance. Northeast trade winds blow less saline surface water downwind with a buildup in the northwest region of the lagoon. Because water is trapped in the leeward side of the lagoon, return flow develops in the deeper water. The low salinity return flow is mixed with surface water, creat- ing a relatively vertically well-mixed water column with low salinity downwind and high salinity upwind. During long dry periods, opposite salinity gradients might be expected, with high salinities downwind and relatively low salinities upwind. Only a small portion of downwind surface water escapes out of the Southwest Passage. The excess water must move south toward the South Channel; consequently isohalines bend toward the south (Fig. 13). There is no evi- dence of a discrete water mass sinking on the downwind side of the lagoon, flowing upwind as deep water, and upwelling on the leeward side of the lagoon (as reported by Ford, 1949 at Bikini). The water column appears vertically well mixed (Table 1 and Fig. 13). There is also no suggestion that water can maintain vertical structure for 5 to 10 days at Encwetak. As ocean water pours over the windward reefs and into the lagoon, it mixes ver- tically and horizontally as it moves downwind. Conse- quently the salinity gradient is low to high, west to east, regardless of depth. The water on the windward side of the lagoon is predominately ocean water, but water on the western side of the lagoon reflects net processes in the lagoon. Scuba divers can observe strong mixing on the upper vertical wall of the West Spit. Lower salinity lagoon water mixes with high salinity ocean water in this region. Phosphate and nitrate are lower in the western lagoon water than in eastern water. Because water on the eastern side in general reflects net lagoon processes, low nutrients in that water suggest net uptake of these nutrients into the ecosystem. Net organic production of benthic ecosystems has been estimated by net uptake of nutrients (Smith and Jokiel, 1976; Atkinson, 1981; Smith and Atkinson, 1983). The observed decrease in these nutrients indicates a rea- sonably low, net organic production for the atoll. Ford suggested oceanic eddies might move through the broad open channel at Bikini, the Enyu Channel. Perhaps this process might occur in open lagoons; however, it docs not appear to occur at Enewetak. Large eddies would be destroyed when flowing into the lagoon by strong tidal currents in the channels. Although a large eddy could not be maintained, large oceanic eddies moving by the atoll could influence the chemical and biological composition of inflowing water. Tide Tidal currents directly influence the flow of water within several kilometers of the passes, especially in the southern part of the lagoon. These tidal currents can overwhelm the wind-driven circulation, leading to such local effects as the "left-hand" spiral observed two kilome- ters north of Enewetak Island (Fig. 6c). RESIDENCE TIMES In the most elementary analysis, the average residence time of water in the lagoon can be estimated by dividing the lagoon volume by the net rate of water input. The cal- culation yields a residence time of 33 days. Clearly there is a variation of actual residence time from one part of the lagoon to another because: (1) the water is introduced all along the windward reef, but exists primarily through the South Channel; and (2) there is no major north-south recirculation mixing northern waters with southern water. Thus, the residence time for water entering the north end of the lagoon will be relatively long; water entering across the southern reef will have a short residence time. Because the water entering the northern lagoon must transit the entire lagoon before exiting and because it undergoes mixing by the superimposed wind-driven circula- tion during that transit, a very simple estimate of the residence time for that part of the inflow will have at least qualitative validity. If it is estimated that the northern part of the lagoon receives one-quarter of the total inflow, then the residence time for this water (under the same very sim- ple assumptions) will be four times longer than that for the lagoon water as a whole, or 132 days. Water entering the system in the north is of particular interest because it flows across the areas with high bottom-sediment concentrations of transuranic radionu- clides (Nelson and Noshkin, 1972). Figure 14 is a general- ized plot of sediment radionuclide activity; it indicates that if release into the water column is proportional to the con- centration in the sediment then most of the radionuclides E >- > < o 20 DISTANCE (km) 40 Fig. 14 Decrease of sediment radionuclide activity as a function of distance from the north end of the lagoon. Radionuclides include "Sr, ^^u, '^Cs. "Co. OCEANOGRAPHY arc released into northern lagoon water, which has residence times well above the average for the whole lagoon. The concentrations of radionuclides in the water column decrease from the northern end of the atoll to the southern end, by a factor of 2 to 5 (Noshkin et al., 1974). This horizontal gradient reflects the general increase in flushing rate in the south end of the lagoon, as well as hor- izontal diffusion from the north end. The water column is vertically well mixed in terms of temperature and salinity. However, in the central lagoon the horizontal diffusion rates for certain materials may be greater in the surface water than in deep water. Near the windward reef, where both surface and deep currents respond to the cross-reef currents, vertical transport may be greater than in the central lagoon, and there may be no difference in horizontal diffusion rates between surface water and deep water. VON ARX MODEL FOR BIKINI Von Arx's (1948) model conceptualizes lagoon circula- tion by linking two basic patterns: a "primary circulation" and a "secondary circulation." The primary circulation consists of wind-driven surface water moving downwind, sinking, and then returning upwind to the windward (eastern) side of the atoll lagoon as deep water. The secondary circulation consists of horizontal recircu- lation of deep water. Von Arx reported that the volume transport of the eastern flowing deep current is greater than the volume transport of the surface current. He con- cluded that some of the deep water is shoaled upward or "upwelled" in the eastern part of the lagoon, becoming the surface current. The remaining portion of the deep water diverges at the leeward edge of the windward reef. Some water moves northward following the bathymetric contour of the basin. The deep water circulation forms two counter-rotating bodies of water, the northern one moving counterclockwise and the southern one moving clockwise. Von Arx estimated that the exchange of lagoon water through all channels and passes during winter is approxi- mately 3.8% of the total lagoon volume per tidal cycle. At a 30% exchange efficiency, von Arx estimated the winter Bikini lagoon flushing to be 39 days. The summer flushing time was estimated to be twice as long as that in the winter. The conspicuous feature of von Arx's model for deep atoll lagoon circulation is the deep return flow toward the windward side of the atoll. This return flow connects the primary circulation with the secondary circulation. The model for the circulation system of Enewetak has some similarities to the model proposed by von Arx for Bikini. The primary circulation system consisting of an overturning wind-driven surface current is the same in terms of speed and volume transjxjrt. The secondary system, or deep circulation, is not the same as that proposed by von Arx. The deep current at Enewetak flows southward, toward the channel having net outflow. Von Arx described a horizontally recirculating deep current with a volume transport greater than the sur- face current, hence upwelling on the windward side. At Bikini the large open channel (Enyu Channel) is at the southeastern end of the lagoon. A net transport toward this channel would create an eastward flowing deep current. The eastward mid-depth current and the "pass- ward" deep current would then appear to be a single deep current with a mass transport greater than the surface current. The excess volume transport of von Arx's deep current might largely be balanced by net outflow through Bikini's southeastern channel. Von Arx did not report a large net outflow; however, recalculation of his data sug- gests net outflow through the Enyu Channel. Outflow was also shown in the distribution of indigenous zooplankton (Johnson, 1949) and was observed in surface radionuclide patterns (Noshkin et al., 1974). To reach the Enyu Chan- nel, the deep water in Bikini Lagoon must move east- wards. In Enewetak Lagoon the only effective exit is at the southernmost part of the atoll; therefore, the deep water must move southward. Note that in the model derived from Enewetak, the deep motion is primarily controlled by the location of the major exit points from the lagoon. Water flow through other atoll lagoons seems to be regulated by atoll morphol- ogy and local wave and tidal conditions (Milliman, 1967; Gallagher et al., 1971; Henderson et al., 1978; Ludington, 1979). Studies of deep currents in other deep lagoons could be valuable in testing this interpretation. CONCLUSION Windward and leeward cross-reef currents, channel currents, and tidal flow are the major factors influencing the exchange of water between atoll lagoons and the sur- rounding ocean. Because these factors are specific to local wave climate, tidal conditions, and atoll morphology, atoll lagoons have widely varying flush characteristics. Wind- driven circulation, a pervasive feature of lagoons, con- tributes primarily to internal circulation rather than flush- ing. Upwelling on the windward side of lagoons may occur as a summation of the above phenomena but does not seem to be a generalizable feature of deep lagoon circula- tion. Deep water flow appears to orient itself toward the channels of net water output. ACKNOWLEDGMENTS This chapter is based on the final report of EXDE con- tract number EY-77-5-08-1529, Water CiTculation of Enewetak Atoll Lagoon and Circulation of Enewetak Atoll Lagoon, by M. J. Atkinson, S. V. Smith, and E. D. Stroup. Parts of the research were done under the auspices of the Mid Pacific Research Laboratory. Thanks to S. V. Smith for chemical data and review of the manuscript. 70 ATKINSON REFERENCES Atkinson, M J., 1982, Phosphate Flux as a Measure of the Net Coral Reef Productivity, in Proceedings of the Fourth Interna- tional Coral Reef Symposium. Manila, 1: 412-418. S. V Smith, and E D Stroup, 1979, Water Circulation of Enewetak Atoll Lagoon, Final Report, DOE Contract EY- 77-5-08-1529. S. V. Smith, and E. D. Stroup, 1981, Circulation in Enewetak Atoll Lagoon, Limnol. Oceanogr , 26: 1074-1083. Barnes, C A., D. F Burmpus, and J Lyman, 1948, Ocean Cir culation in the Marshall Islands Area, Trans Am Geophys. Union. 29: 871-876. Buddemeier, R. W., 1982, The Geohydrology of Enewetak Atoll Islands and Reefs, in Proceedings of the Fourth International Coral Reef Symposium, Manila, 1: 339-345. Ekman, V W , 1905, On the Influence of the Earth's Rotation on Ocean Currents, Ark Mat, Astron. Fysik, 2: 1-53. Emery, K. O., J. 1 Tracey, Jr , and H. S. Ladd, 1954, Geology of Bikini and Nearby Atolls, U S Geol Sum. Prof. Pap , 260-A. pp. 1-265. Ford, W L , 1949, Radiological and Salinity Relationships in the Water at Bikini Atoll, Trans, Am, Geophys Union. 30: 46-54 Gallagher, B S., K. M. Shimada, F. I. Gonzalez, Jr., and E. D. Stroup, 1971, Tides and Currents in Fanning Atoll Lagoon, Pac. Sci., 25: 191-205 Hamner, W. M , and 1 R. Hauri, 1981, Effects of Island Mass: Water Flow and Plankton Pattern Around a Reef in the Great Barrier Reef Lagoon, Australia, Limnol. Oceanogr . 26: 1084-1102. Henderson, R. S., P. L. Jokiel, S. V. Smith, and J. G. Grovhoug, 1978, Canton Atoll Lagoon Physiography and General Oceanographic Observations, Atoll Res. Bull., 221: 514. Johnson, M W., 1949, Zooplankton as an Index of Water Exchange Between Bikini Lagoon and the Open Sea, Trans, Am. Geophvs Union, 30: 238-244. Ludington, C. A.. 1979, Tidal Modifications and Associated Circulation in a Platform Reef Lagoon, Aust J Mar, Freshwater Res., 30: 425-430. Milliman, J P., 1967, Carbonate Sedimentation in Hogsty Reef, a Bahamian Atoll, J Sediment. Petrol., 37: 658-676. Montgomery, R B , and E. D Stroup, 1962, Equatorial Waters and Currents at 150°W in July-August, 1952, Johns Hopkins Oceanogr Stud , 1: 1-205. Munk, W H , G C. Ewing, and R. R. Revelle, 1949, Diffusion in Bikini Lagoon, Trans. Am. Geophvs. Union, 29: 59-66. and M. C. Sargent, 1949, Adjustment of Bikini Atoll to Ocean Waves, Trans. Am, Geop/iys. Union, 29: 855-860 (also U. S. Geol. Surv Prof Pap 260 C). Nelson, V., and V. E. Noshkin, 1973, Enewetak Radiological Sur uey. U. S Atomic Energy Commission, NVO-140V, pp 131 225 Noshkin, V E, K M Wong, R. J. Eagle, and C Gatrovsis, 1974, Transuranjcs at Pacific Atolls, 1, Concentrations in the Wafers at Enewetak and Bikini, University of California, Liver- more Rep , 51612: 1 30 Smith, S. v., and P L Jokiel, 1978, Water Composition and Biogeochemical Gradients in Canton Atoll Lagoon, Atoll Res, Bull , 221: 15-53. , and M. J. Atkinson, Mass Balance of Carbon and Phos- phorus in Shark Bay, Western Australia, Limnol. Oceanogr , 28(4): 625-639. Von Arx, W W., 1948, The Circulation Systems of Bikini and Rongelap Lagoons, Trans. Am. Geoph^/s. Union, 29: 861-870 , 1954, The Circulation Systems of Bikini and Rongelap Lagoons, U S. Geol Suru. Prof Pap. 260-B, pp. 265-273. Chapter 6 Meteorologi; and Atmospheric Chemistry; of Enewetak Atoll JOHN T. MERRILL and ROBERT A. DUCE Center for Atmospheric Chemistr\/ Studies Graduate School of Oceanography^ University^ of Rhode Island, Kingston, Rhode Island 02881 INTRODUCTION The Marshall Islands area has a marine climate that varies from tropical to subtropical; near Enewetak Atoll the weather is characterized by brisk steady winds, moderate rainfall, and unvarying high temperatures with typical partial cloudiness. The atoll lies near the northern edge of the tropical zone dominated by the migrating equa- torial trough of low pressure, with its heavy rains. It lies well within the northeast trade wind area of the North Pacific; that is, the surface winds are from the east and northeast on average. There have been more than 20 years of careful meteorological observations at the airstrip on Enewetak Island, and we make use of some of the archived data to discuss, in turn, the various aspects of the weather. In the section on climate and weather, we cover briefly the mean and variation for each observed quantity of interest and note our state of knowledge of these fac- tors. Also in that section we set out an annotated bibliog- raphy of sources of additional data and of specialized dis- cussions. In the section on the atmospheric chemistry of the atoll, we make use of the extensive data collected dur- ing experiments there in 1979. We discuss both the mean value and exfjected range of variation because neither alone covers all of the weather. The variability of the weather is the combined effect of dis- turbances of various scales which may have well-defined structures in space and time and of phenomena that can be taken as random. We begin by discussing some of the more common structured disturbances. Over the years diurnal variations at island sites have been discussed and analyzed. While there is no doubt that there are diurnal cycles in cloudiness and precipitation, no attempt is made here to provide explanations for them in terms of first causes because the interaction can be both subtle and com- plex. Also, at short periods there is the atmospheric tide, primarily a thermally driven effect that produces global pressure fluctuations and wind patterns that are rather complex. The influence of the atmospheric tide at the sur- face, though greatest in the tropics, is relatively small, and we mention it only briefly. The semidiurnal fluctuation is the dominant mode of the tide and has an amplitude of about 1 mbar, or 85% of the diurnal variance about the annual mean of 1010 mbar pressure at Enewetak. Chap- man and Lindzen (1970) developed the presently accepted theory of the tide. The discussion by Lavoie (1963) of cal- culations available at that time is superseded, despite the absence of seasonal effects in Chapman and Lindzen's basic model. Nevertheless, the data presented by Lavoie for the monthly variation of tide parameters are correct and illustrative, despite the relatively short record. Disturbances lasting a day or more are common in the tropics, and we discuss them in the sections entitled "Trop- ical Storms and Disturbances" and "Winds Aloft." There are two seasons at Enewetak, the dry season from December through March and the wet season from April through November. Annual variation is crucial to under- standing the weather in the tropical marine environment, and this influence is included in each section, particularly in the section on precipitation. Although there has been much work recently on the variability of climate over periods of a year to a decade, we cannot say much yet about how such changes affect the tropical islands. It is known that there are quasi- periodic fluctuations in the strength of the Pacific trade winds correlated with equatorial sea surface temperature variations at very large scales and that there follows a chain of consequences that includes changes in both tropi- cal and mid-latitude circulations. A clear exposition on this subject, the Southern Oscillation, is to be found in Tren- berth (1976). While much of the present interest stems from the possibility that disturbances in mid-latitude weather and coastal upwelling could be forecast months ahead, we will certainly learn much about the tropical cli- mate itself from the numerous studies now under way. 71 72 MERRILL AND DUCE CLIMATE AND WEATHER OF ENEWETAK ATOLL Temperature and Humidity It is obvious that high surface temperature and hu- midity are to be expected on tropical islands. It is less obvious, but well documented, that it is difficult to obtain accurate temperature measurements in an op)erational pro- gram in such environments because of such factors as radi- ational heating of the shelter in daylight. Thus it is likely that the air temperature range rep>orted below is exag- gerated by about 0.5°C (see Lavoie, 1963, for a discus- sion). This is a small enough error for temperature, but it significantly degrades the accuracy of the relative humidity. Nevertheless, we can see that there is relatively little change in these quantities through the year and that a reg- ular diurnal cycle is evident. The temperature and hu- midity both respond noticeably and regularly to rain showers, but in the data presented here the nearly random occurrence of rain with time has smoothed out this effect; in fact, even hourly data do not show the full effect of short-lived rain events. The temperature and humidity data shown in Fig. la-c are from the U. S. Air Force measurements now archived by the National Climatic Center. Made at hourly intervals between 1945 and 1969 (with irregular breaks), the obser- vations correspond to 14.1 years of uninterrupted mea- surement. This record is sufficiently long that the overall pattern and its variability can be perceived. The data aver- aged over 3-hour periods are displayed as a function of the hour and of the month; the draft plot is extended beyond the borders shown so that edge effects are minimized. That the temperature depends very little upon the time of year can be seen in the mostly horizontal contours shown in Fig. la. Also note that the highest daily tempera- ture is recorded between 12 and 15 hours local standard time and lies between 28.5 and 30°C; the lower temperature is observed in the dry season and the higher during the wet season. In the morning and in the evening, the temperature depends even less upon the time of year, with values increasing and decreasing daily through the upper 20s. In the hours after midnight, the decrease of temperature slows in the dry season and ceases in the wet season, with the lowest average value reaching 26 to 27°C. Now these are monthly and 3-hour averages over years of data, and even though these patterns are generally valid, there are fluctuations. The representativeness of this pattern can be seen in the small variances: just over 0.5°C at night to a maximum of under 1.5°C in the afternoon in October. (These are variances of hourly data averaged over 3-hour periods for each month, i.e., variances about the mean shown in Fig. la.) This method of averaging does not accurately record the average maximum and average mmimum hourly temperatures for each day. These are given in Table 1 for each month of the year. The range of temperature is greater here, as expected. The average minimum temperature for each month is nearly independent of month at about 23°C, whereas the average maximum exceeds 32°C in August and September and is 30°C during the dry season. (c) M J J A MONTH Fig. 1 Temperature and humidity data for Enewetak. (a) Dry bulb temperature, °C; (b) Relative humidity in percent: (c) Dew point tempicrature, °C. These are three hour averages for each month. Contour interval is 0.5°C for temperatures. 2.5% for humidity. METEOROLOGY AND ATMOSPHERIC CHEMISTRY 73 TABLE 1 Average Minimum and Maximum Temperatures, °C Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Minimum 23 5 23 4 23 6 23.8 23.5 24.0 23.6 23 6 23 8 23 4 23 7 23.8 Maximum 30.4 30 5 30.6 31.1 31.4 31.6 318 32.3 32.3 32.1 31.7 30.9 Extremes of temperature have also been recorded but provide little additional information. The annual minimum value of 21°C has occurred once in the record and the maximum of 34.4°C three times. The values reported by Blumenstock and Rex (1960) are less reliable because of shorter records. As mentioned above, it is possible that the maximum temperatures are overestimated. Similarly averaged relative humidity data shown in Fig. lb exhibit a bit more time dependence than the tempera- ture. There is a broad maximum in the early morning fol- lowed by decreasing values as the temp)erature rises in daylight. The afternoon minimum is lower but briefer in the dry season than it is in the wet season. These values are the best estimates available but may be underestimated during the day by several percent if the temperature is overestimated by 0.5°C. The dew point temperature — the temperature at which saturation will occur if the air is cooled — is less sensitive to such error and is a straightfor- ward indicator of the moisture content of the air. The dis- tribution shown in Fig. Ic seems quite different from, but is entirely consistent with, the data in Fig. la and b. The dew p)oint temp)eraturc is strongly dependent upon the time of year, with a broad minimum in the dry season and a broad maximum in the wet season. In both cases there is an increase during the daylight hours. This increase represents a slight increase in the water-vapor mixing ratio, which is consistent with increased evaporation during the day. It is this increase which prevents the relative humidity from decreasing more than it does in the afternoon. Precipitation Across most of the North Pacific, including the Marshall Islands area, the rainfall increases markedly from mid-latitudes to just north of the Equator. At Wake (19°N) the annual total is 940 mm, at Kwajalein (9°N) about 2400 mm, and at Jaluit (6°N) it exceeds 4000 mm. The highest values are in the equatorial trough near 3° to 6°N and the lowest in the subtropical high pressure area at about 25°N, well east of the dateline. Enewetak at 11°N lies near the northern edge of the zone of most rapid decrease of rainfall with latitude. The average rainfall of 1470 mm is not distributed uniformly through the year; about 85% comes during the wet season, which starts in April and ends in mid-November. The variability of the rainfall is remarkable, and this factor is a central theme in the discussion which follows. There is much contention in the meteorological litera- ture over the applicability of island-based rainfall data to open ocean conditions. This is not an issue here as we are concerned with the rainfall at the atoll, not in the environ- ment in its absence. We note briefly below the limited data available to discern gradients across the atoll. The rainfall distribution through the year, which is based on the archived data tabulated by Taylor (1973), is shown in Fig. 2a. The three measures shown for each month are the rainfall amount exceeded in 90, 50, and 10% of the years. Thus the amount expected (50% occurrence) in November, about 124 mm, is somewhat less than the 140 ram in the "rainiest December in 10 years." It can be seen at once that in certain months there is a very large range of variability. Although the average April has about 40 mm of rainfall, one year in ten may have less than 10 mm; another may have over 260 mm. The record represented here is 32 years long; there are missing data, but 26 to 29 years of monthly values are available in the various months. A longer series would probably not change the annual total much, and because it is inherent in tropical rainfall, the fluctuation evident in Fig. 2a would not be reduced by additional years of obser- vation. For this discussion, we let the dry season begin with December. That this is arbitrary can be seen by comparing the 10% value in Fig. 2a for December with the 50% value for November and the 90% value for November with the 10% value for December. When the wet season ends early, the November rainfall is less than 120 mm, whereas when it ends late, the December total exceeds 50 mm. The January, February, and March 50% values all lie between 20 and 35 mm, and the 90% values are uni- formly very small at <10 mm. Many of the dry season rain events are from small cumulus clouds; however, these affect the total amount less than the infrequent distur- bances. We follow the usual designation and let the wet season begin in April despite the small increase in the 50% value; this may be rationalized by the jump in the 10% figure — i.e., some Aprils are very wet. The increased rainfalls of May and June are followed by 50% values between 175 and 225 mm in July through September. The maximum is in October, which also has the highest average and the greatest 90% total. November is a transitional month. The number of days with measur- able (>0.25 mm) rainfall is greatest in August at 21 on the average; this figure varies between 10 and 21 for the wet season months, while it is 10 to 15 during the dry season. 74 MERRILL AND DUCE WET SEASON DRY SEASON (b) RAINFALL DEVIATION Fig. 2 Rainfall data for Enewetak. (a) Rainfall amounts for each month. The three measures correspond to the amount, in mm, exceeded in 90% (below shading). 50% (above) and 10% (top) of the months in the record, (b) Deviation frequen- cies for the wet and dry seasons. The curve shows the percent occurrence of deviations from the seasonal mean in units of a. the square root of the variance. The corresponding rain amounts, in mm are shown below the axis. Thus, as noted above, there are many small rain showers even in the dry season. The wet season amount includes such cumulus showers and a greater number of larger, sus- tained rainfalls during disturbances and during the infre- quent tropical storm or approach of the equatorial trough; these are much less common in the dry season. Another way to look at the variability of the precipita- tion is shown in Fig. 2b, where the frequency of occurrence of monthly amounts is given as a function of the deviation from the season average. Note first that in both regimes the most common value is significantly lower than the mean, about 0.5 a below. (Here a is the root mean square deviation of rainfall amounts for the season.) In the dry season, very dry months are common, but a few months with large amounts of rainfall do occur. For exam- ple, 2% of the years would be expected to have a "dry" month with 160 mm of rain, 2.5 a above the average. Note also that the overall occurrence of the large rainfall months is not dependent upon the season but that the amount of precipitation is. The infrequent 3 or 4 a cases for the wet season corresfsond to very substantial totals. These frequency distributions are typical of subtropical sites but are somewhat uncertain far from the mean value because of the limited length of record. Also, since these frequencies were averaged over the entire season, the month-to-month variation, which is large in the wet season, is lumped together with the interseasonal difference here. Insufficiency of data limits one's ability to document diurnal variation In precipitation amount. Nevertheless, Lavoie (1963), using primarily Enewetak data, presented convincing evidence of an early morning maximum in frequency of rain; the deviation is fjerhaps 15% at the peak. There is some evidence in the same data set for a broad and weak afternoon minimum in the rain occurrence. Lavoie considered several mechanisms in an attempt to rationalize these values and to explain some sig- nificant difficulties: the data have large scatter, and even the maximum does not appear at every station. It thus seems best to say that there is a tendency for a maximum in the rainfall occurrence in the early morning and a weak minimum in the afternoon. Even more limited are data giving the spatial distri- bution of rain about the atoll. Any variation is assumed to be primarily random because of the low relief of the atoll — i.e., the absence of orographic forcing. However, there could be sufficient disruption of the thermodynamic structure of the atmosphere by the presence of the lagoon to cause a discernable pattern. Data laboriously collected by Blumenstock and Rex (1960) for six special sites on islands around the atoll during 2-week periods, once in each season, have not to our knowledge been carefully analyzed in the literature. They reveal no systematic pat- tern of variation. The rainfall amounts at the various sta- tions are highly correlated only when the stations are close together, and there is always some difference among them. The record thus appears consistent with rain areas of vari- ous sizes unforced by the atoll itself. Nevertheless, this does not rule out some such forcing in other cir- cumstances. This record does not include any disturbed weather periods during which there could be a measurable difference of rainfall across the atoll. Cloud Cover and Solar Radiation Accurate estimates of cloud distribution and type are not easy to obtain, particularly at night and when low clouds obscure the sky. As discussed by Blumenstock and Rex (1960), there is likely a systematic bias — overestimation. Fortunately the overall cloud amount is least affected, and voluminous data exist for this quantity in the archive. Again, the average variation with time of day and time of the year is the main topic of discussion. METEOROLOGY AND ATMOSPHERIC CHEMISTRY 75 The fraction of the sky covered by clouds exceeded 75, 50, and 25% of the time as shown in Fig. 3, which is based on the tabulated three hourly frequencies of cloud cover in tenths. The cloudiness is more variable in the dry season than the wet; indeed, the daytime sky is covered Vio or more 75% of the time in July through October. More than 7,0 cover is common (^25%) in April through November, and this frequency does not depend u|x>n the time of day. The expected cloudiness (i.e., the 50% value) varies from about Yio, higher in daylight and lower at night, in the dry season to about '/lo, with somewhat less diurnal variation in the wet season. The sky is seldom clear, even in the dry season. Two-tenths of the sky is covered more than 75% of the time (although this may be biased at night). A different measure of cloud cover is obtained from satellite observations. Images obtained routinely from geo- stationary platforms show the aerial extent and temporal evolution of cloud systems. In addition, radiometric mea- surements of the cloud top temperature yield good esti- mates of the cloud top height. Accurate measurements below obscuring layers of high clouds are not yet obtain- able routinely. We are not aware of any published solar radiation data for the Marshall Islands area, but there are data for certain times in 1977 to 1979 (see "Sources of Additional Data"). It is obvious that the typical partial cloudiness and the high moisture content of the near-surface air significantly dimin- ish the incident sunlight. Working against this, however. Is the long (and unvarying) day. The interval between sunrise and sunset varies from 11 hours 29 minutes to 12 hours 46 minutes. The available data show these effects, with the average value exceeding 21 X 10^ j m~^ d~' in I z J*^' f'EB MAR APR MAY JUN 2 ° 10 10 10 10 10 •" '■ " I' M T I I, I 01 ■ ' 04 _l 07 5 '° O 13 —I 16 19 22 AUG SEP OCT NOV DEC Fig. 3 Cloud cover data for Enewetak. For each month the three measures are the fraction of the sky obscured, in tenths, at least 75% (left), 50%, and 25% (right) of the time for each three hour period of the day. many months; this value corresponds to 500 cal cm~^ d , a typical maximum total at mid-latitudes. Neverthe- less, during disturbances the flux can be reduced for periods of several days, and the value can drop below V^ of this figure for a day or two at a time. Surface Wind The surface wind data are shown in Fig. 4 as wind roses for each month (a) and for the entire year (b). As indicated in the key in Fig. 4b, in each rose the bar indi- cates the frequency of winds coming from that direction for each range of speed above calm. The numerical values beside each bar are the frequency, in percent, for wind from that direction and for that range of speed. The fre- quency of calms, to which no direction is assigned, is shown in the center of the circle. The frequency of occurrence of wind in each range of speed for all directions is shown, in percent, in the line below each rose. The wind JAN 42 37 1 1 Fig. 4 Surface wind data for Enewetak. Wind roses for each month (a) and for the year (b). Frequency of wind for each directon and range of speed is shown by the printed figure beside the bar. which shows the frequency of winds for that direction for all speeds above calm. Frequency of calms is shown in the center of the circle. Frequency of wind speeds for all directions shown below each rose. (Fig. 4a cont'd on next page) 76 MERRILL AND DUCE FEB APR I 1 I e I 42 ! 37 I 12 I X I I I X I .1 I 7 I 46 I 39 1 7 I X 1 MAR MAY I X I 1 I 8 I 46 1 36 I 10 I X I I | 1 | 2 I 1 2 I 49 | 3 1 | 6 1 X ; j Fig. 4a cont'd. (Fig. 4a cont'd on next page) METEOROLOGY AND ATMOSPHERIC CHEMISTRY 77 JUN AUG 1 1 4 ! 16 I 53 I 24 [ 3 ! X ' 4 I 15 I 31 I 40 7 1 X JUL SEP 2 I 9 I 28 I 48 i 11 1 1 I X : I | 6 | 18 | 34 | 33 | 7 | 1 | X | X | Fig. 4a cont'd. (Fig. 4a cont'd on next page) 78 MERRILL AND DUCE OCT NOV 5 I 17 I 31 I 35 I 7 I 1 I I 1 I 5 I 20 I 44 I 23 I 6 I X I X DEC X I 3 I 1 1 I 42 I 33 I 9 I 1 I X Fig. 4a cont'd. (Fig. 4 cont'd on next page) METEOROLOGY AND ATMOSPHERIC CHEMISTRY 79 ANNUAL 18 Lll. 24 speed increments are indicated in the figure and are the traditional Beaufort scale values. The data are from U. S. Air Force records, as are the temperature and humidity data above, and were collected at various intervals, hourly over most of the period, with the instrument mounted at a height of 40 feet above sea level. The data are from the years 1945 to 1969 and again correspond in number to 14.1 years of continuous hourly observation. Thus the representativeness of these figures is good and falls in the range that one would expect from a sample of 10 to 20 years of continuous measurement. We have smoothed the data to report them at eight compass points but were care- ful to use a weighted averaging that preserves the rapid falloff of wind occurrence away from the predominant east and northeast directions. This much-noted constancy of the wind is the first aspect of the rose data that we examine. During much of the year, the wind is from the northeast or east 95% or more of the time. From July through October, however, the peak broadens somewhat and moves a bit toward the south so that less than 20% of the winds are out of the northeast. The maximum frequency of winds from the less common directions (southeast around to north) is in August, September, and October, when disturbances are most common and when the equatorial trough is closest on 12 3 4 5 6 7 8 1-3 4-7 8-12 13-18 19-24 25-31 32-38 39-46 1-3 4-6 7-10 11-16 17-21 22-21 28-33 34-40 03-1.6 2-3 3.6-52 5.7-83 8^-11 11-14 14-17 18-21 FREQUENCY OF CALMS: PERCENTAGE IS SHOWN IN CENTER OF CIRCLE. DIRECTION FREQUENCY; BARS SHOW PERCENTAGE FROM EACH DIRECTION. EACH CIRCLE EQUALS 10%. 20% OF ALL WINDS FROM NE. SPEED FREQUENCY: FIGURES SHOW PERCENTAGE FROM EACH DIRECTION IN EACH SPEED RANGE. 6% OF WINDS WERE FROM BETWEEN 13 AND 18 mph. TABLE: FREQUENCY OF WIND FROM ALL DIRECTIONS FOR EACH SPEED RANGE. SPEEDS ARE THE BEAUFORT SCALE: MILES / HOUR NAUTICAL MILES / HOUR METERS / SEC Fig. 4b 80 MERRILL AND DUCE the average. The highest frequency of brisk winds is in the dry season, with over 45% of the hours having winds >8.5 ms~' (19 mi h~'). During the wet season the wind weakens substantially, particularly during August through October when >50% of the hours have wind <5.4 ms~' (12 mi h~^). Only during July through October are calms at all common, i.e., greater than 1% occurrence. The dry season months exhibit the greatest constancy of pattern: >50% from the east and >40% from the northeast, with >75% frequency of speeds between 5.8 ms"' and 10.7 ms"' (13 and 24 mi h^'). April does not differ much, except that the strength of the wind decreases slightly. In May and June the winds are strong out of the east, while in July through October the speed decreases and the direction varies more. In November the wind begins to shift back to the dry season pattern. As these are average winds, the pattern of variation with time is lost. There is a consistent shift in the wind associated with easterly waves, the most common distur- bance type in the wet season. The correlation of wind shifts with cloudiness and rainfall, obvious to anyone present during such events, is lost. The annual average wind rose shown in Fig. 4 is easily understood given the monthly distributions discussed previ- ously. Note that % of the time the wind is from 5.8 to 10.4 m/s (13 to 24 mi h '), and over 60% of the time the wind is from the east. Nevertheless, the annual aver- age shows at least 0.1% winds from every direction. Tropical Storms and Disturbances While tropical storms strike the Marshall Islands infre- quently, disturbances in the weather are a common and, on occasion, regular occurrence. Tropical storms of the greatest strength are called typhoons in the western Pacific, and they are, of course, extremely dangerous and destructive, particularly to exposed areas at low elevation such as Enewetak Atoll. Such storms grow from and are in fact the most fully developed form of tropical disturbance. We discuss the disturbances first because they are more numerous. Several types of tropical disturbances are recognized in the literature; nevertheless, it is often impossible to classify a given weather system as one of the several types, even given estimates of the thermal structure and the movement and growth of the system. We are concerned primarily with the surface manifestation, so we shall only summarize what is known about the most common disturbance types. During the wet season, particularly July through Sep- tember, westward propagating wave-like systems are com- mon in the tropics and have been observed and analyzed in the western Pacific and in the Caribbean and North Atlantic Ocean areas. In the western Pacific these easterly waves, on average, have a horizontal scale of 3500 to 4000 km and travel toward the west an average of 7° longitude per day (i.e., a mean velocity of 9 ms or 20 mi h^'); thus the disturbance affects a station for 4.5 to 5 days. During the passage of such a wave, there is a more or less systematic variation in the wind, cloud cover, and rainfall. The north-south component of the wind shifts, with maximum winds from the south of 1 to 2 ms"^ (2 to 5 mi h~') leading and maximum winds from the north following the center of the disturbance. The max- imum cloudiness and rainfall occur just after the pjassage of the center of the disturbance. There is a temperature fluc- tuation, but it is hardly discernable at the surface. These waves can be observed with satellite images and are now understood to be an inherent prop>erty of deep easterly flow. The structure and detailed dynamic characteristics of such waves in the Marshall Islands area were studied by Reed and Recker (1971) using radiosonde and satellite data. The waves are most common in the wet season because the upper level winds are most favorable for their growth then. About Vs of such waves increase in intensity sufficiently to become classified as depressions or storms, but this occurs most commonly well west of the Marshall Islands. Other types of disturbances are more uniformly dis- tributed through the year but are even less easily classi- fied. One type, the upp)er level cold-core low, is similar to the subtropical cyclone that is often observed in the Hawaiian area. In the Marshall Islands area, it may have no surface manifestation or may be accompanied by a weak but long-lived period of disturbed weather. In addi- tion, there arc squall lines and other short duration events which may produce strong winds and intense rainfall over limited areas as they pass. Although both the frequency and the destructive power of tropical storms are greater in the far western Pacific than in the Marshall Islands area, such storms can threaten any tropical location. A sense of the seasonal distribution and the range of impact possible can be obtained from Table 2, which summarizes the depressions and storms that affected Enewetak between 1959 and 1979. Of course, the highest overall probability of tropical storm for- mation in the area is during the wet season, particularly July to October. However, there have been strong storms well within the dry season (e.g., Alice in 1979). The high winds and waves that extend to the periphery of such storms can have devastating consequences. There is a sub- stantial body of literature on the effects of such storms on atolls, but the closest atoll so studied is Jaluit (Blumenstock, 1961). Specific data about individual storms are often sketchy, and prior to the operational use of satel- lite images, the tracking of past storms when far from land or shipping lanes may have been substantially in error. Nevertheless, there are useful data on several storms over the years, as indicated in Table 2. Winds Aloft The structure of the wind field above Enewetak Atoll is complex and variable. At time scales longer than 2 years, there are nearly periodic fluctuations at some levels, while at other levels there are short-period variations as impor- tant as those in mid-latitudes. In the following discussion. METEOROLOGY AND ATMOSPHERIC CHEMISTRY 81 TABLE 2 Tropical Storms and Disturbances Affecting Enewetak. 1959-1979 Name, Year dates (GMT) Remarl35 km) to at least 100 mbar (—16 km). This is now understood to be an interaction phenome- non illustrating the coupling between the troposphere and tropical stratosphere. Its discovery in the early 1960s illustrates how recently we have begun to learn about this area of the atmosphere. The "Krakatoa Easterlies," so named because they were first observed transporting debris from the spectacular 1883 volcanic eruption, are not as constant as had been thought. Sources of Additional Data Several of the important sources of additional meteoro- logical data for Enewetak have been referenced in the previous sections. Here we summarize briefly the availability of various types of data and indicate the best sources for discussions on sp>ecialized subjects. The archive of data used to prepare the figures in this chapter is the Revised Uniform Summan; of Surface Weather Observations for Eniwetok Marshall Islands. In addition to the wind, cloudiness, temperature, and humid- ity data presented here, it contains extensive information imfKDrtant primarily for aircraft operations, e.g., ceiling and visibility data. The document can be obtained for copying costs from the National Climatic Center, AsheviUe, North Carolina. A reliable and useful atlas of tropical wind and tem- perature charts is included in Newell et al. (1972) along with sophisticated discussions of the global tropical circula- tion in dynamical terms. The Pacific island rainfall data and analysis of Taylor (1973) are an excellent resource. There is a collection of marine meteorological observations (Sum- mary of Si/noptic Meteorological Observations, Volume 3, which includes "Area 10 — Eniwetok") available from the National Technical Information Service as AD-725 138, but the data are very sparse. Solar radiation data are available for certain periods beginning in May 1977 from the Department of Meteorol- 82 MERRILL AND DUCE ogy, University of Hawaii. They are tabulated as hourly totals of the shortwave radiant energy flux, in cal cm h~^ Because the data coverage is not continuous (no period longer than 8 months is available without extended interruption), it is not presented here. Many useful and interesting data were collected by Blumcnstock and Rex (1960) in addition to those discussed previously. ATMOSPHERIC CHEMISTRY OF ENEWETAK ATOLL Introduction During the period April to August 1979, an extensive program investigating the chemistry of atmospheric trace gases, particles, precipitation, and dry deposition was undertaken at Enewetak Atoll. The Sea/ Air Exchange Pro- gram, or SEAREX, was sponsored by the National Science Foundation and involved efforts by 11 institutions from the United States, France, and Great Britain. The impetus for this atmospheric chemistry study was the increasing interest in the possibility that significant quantities of both natural and anthropogenic substances may be transported to the ocean via the atmosphere in mid-ocean regions. An understanding of the importance of the atmosphere as a transport path is critical in determining the basic geochemi- cal cycles and budgets of a variety of naturally occurring substances and in predicting the near-global impact of anthropogenic material in open ocean regions. The objec- tives of the study were to investigate the concentrations and sources of selected inorganic and organic substances in the marine atmosphere at Enewetak, their flux into the ocean, and the mechanisms of their exchange with the ocean. Substances investigated included trace metals such as lead, cadmium, zinc, selenium, copf)er, iron, antimony, manganese, mercury, silver, aluminum, and the alkali and alkaline earth metals; soil dust; atmospheric sea salt; ^'"^b and its daughter ^'°Po; particulate organic carbon; and organic compounds such as PCBs, DDT, aliphatic hydro- carbons, phthalate plasticizers, fatty acids, fatty and poly- cyclic alcohols, and low molecular weight ketones and aldehydes. The atmospheric chemistry studies at Enewetak Atoll took place on Bokandretok Island, just north of Enewetak Island (Fig. 5). During late November and December 1978, an 18-meter-high walk-up sampling tower and three small buildings were constructed on Bokandretok. The sampling tower, located directly on the east coast of the island, was necessary to get above any local contamination from both man-made sources and natural sources such as erosion products and surf spray generated when waves strike the shoreline. Additional precautions were taken against local con- tamination. Sampling pumps were located on the ground and were connected to the collection systems on top of the tower by 20 meters of hose. The operation of the pumps N 1 ^ UJ t- I o n. CD o o lOU • • • •• Solt • . • • • 10 ■ ■ 10 ■ ■ ■ Dust ■ ■ 01 - 1 1 1 ■ ■ ■ I APRIL I MAY I JUNE I JULY I AUG I SEP I SAMPLE COLLECTION DATE, 1979 Fig. 7 Atmospheric concentrations of dust and sea salt at Enewetak l)€tween April and August, 1979. continental land mass, Asia, is about 5000 km to the northwest. The dramatic decrease in dust over the 5-month period was also unexpected, but both these obser- vations can be explained on the basis of the seasonal changes in the large-scale wind patterns over the North Pacific and the seasonal character of dust storm activity in the Takia Makan, Gobi, and Ordos Desert regions of China. Dust storm activity is apparently greatest in the spring in China due to the combined effects of low rainfall, the increased occurrence of high surface winds associated with strong cold fronts, and soil freshly plowed for plant- ing. The mean surface winds from March through May are strong easterlies over the western North Pacific between 30°N and the equator; north of 30°N, the surface winds are weak, with a tendency toward being westerly. How- ever, at 700 mbar (about 3000 m) there is very strong westerly flow north of about 20°N extending from well within Asia to the central North Pacific. Thus dust raised over China could easily be transported by the mean winds at this level to the region north of Enewetak. During June through August, however, conditions are not favorable for the transport of dust to the central North Pacific. Surface winds are easterly from Enewetak northward to about 40°N. At 700 mbar the northern boundary of the easter- lies is located at about 30°N. Persistent westerlies appear at 700 mbar only north of 40°N, and they are very weak. Thus we would generally expect much higher atmospheric soil dust concentrations and deposition rates to the ocean at Enewetak in the late winter and spring than the rest of the year. In corroboration of this, Ing (1972) documented an April 1969 dust storm over China, and satellite photos showed that dust cloud moving well out over the East China Sea. METEOROLOGY AND ATMOSPHERIC CHEMISTRY 85 The observed mass median radii for the Asian dust in the SEAREX study at Enewetak ranged from 0.7 to 1.0 /im, considerably smaller than the atmospheric salt particles. Eighty to 85% of the mass of the dust was present on particles with radii between 0.2 and 2 ^m. This is consistent with a very long atmospheric transport path. Removal of dust to the ocean by rain and dry deposi- tion was estimated at Enewetak through the analysis of rain samples and samples obtained by the exposure of flat plates on top of the tower. The total (wet and dry) deposi- tion of dust during May 1979 was estimated as about 4 ^g cm - Assuming this deposition was applicable for 3 to 5 months during the spring and early summer, with somewhat lower deposition the rest of the year, leads to an estimated annual atmospheric dust deposition to the ocean near Enewetak of 15 to 30 ^g cm~^ (Duce ct al., 1980). Settle and Patterson (1982) report dust in rain and dry deposition at Enewetak which converts to a yearly flux of about 13 and 1 /:ig cm~^ respectively, the latter being recycled in sea spray and not contributing to net input. These inputs can be compared with an estimate of the annual nonbiological marine sedimentation rate to the ocean floor in that region of about 50 ng cm^^ (M. Leinen, personal communication). Within the uncertain- ties in both estimates, it is clear that the atmosphere is a significant transport path for the nonbiological material found in marine sediments near Enewetak. It is also clear that the transport of Asian derived substances to the Enewetak region is seasonal. Lead-210 was also measured in the atmosphere at Enewetak. Lead-210 is a radioactive nuclide produced in the atmosphere by the decay of gaseous Rn, which in turn is derived from continental soils. Atmospheric ^'"Pb was found to decrease over the April to August 1979 period in a manner similar to the atmospheric Al concen- tration. Lead-210 in air ranged from about 4 dpm per 1000 m^ in April to 0.8 to 1.0 dpm per 1000 m^ in late July and August (Turekian and Cochran, 1981a, b). Using ^b as an indicator of Asian dust transport, Turekian and Cochran (1981a, b) calculated a dust deposition of about 10 /ig cm~ yr~ to the ocean at Enewetak. Trace Metals A number of trace metals were investigated on parti- cles in the atmosphere at Enewetak. Some of these trace metals, e.g., Na, Mg, K, and Ca, were clearly derived from the ocean as part of the atmospheric sea salt. Interelemen- tal concentration ratios among this group were the same as found in sea water. Another group of metals was clearly associated with the mineral aerosol or Asian dust. This was determined by using the Al content of the particles as a reference element for crustal weathering products and comparing the metal/Al ratio on the aerosols to the aver- age metal/Al ratio in the earth's crust. An enrichment fac- tor relative to the crust, EFj_i,,, can be defined as follows: where (X/AO^,, and (X/AO^^,,, refer to the mass ratio of metal X to aluminum in the Enewetak aerosols and the earth's crust, respectively. Values of EF^rust near 1 for any metal suggests that crustal weathering is likely its source in the particles (Duce et al., 1975; Rahn, 1976). EF<^, values for samples collected at Enewetak are given in Table 4. From this table it is clear that such elements as Al, Ta, Sc, Mn, Fe, Eu, Ni, Co, V, Hf, Cr, Th, Cu, and Rb are primarily found associated with mineral or soil aerosol particles at Enewetak. Metals with an EF,-^^ value higher than 4, e.g., Zn, Cs, Sb, Ag, Pb, Cd, and Se, apparently have some source other than continental weathering. TABLE 4 Geometric Mean EFct,,,, Values for Atmospheric Trace Metals at Enewetak* Metal EF Metal EF„ Ta 0.7 ± lit Cr 1.8 ± 1.2 Sc 0.8 ± 11 Th 2.0 ± 1.1 Mn 0.9 ± 11 Cu 2.3 ± 1.6 Al 1.0 Rb 3.0 ± 1.2 Fe 1.0 ± 1.1 Zn 4.6 ± 1.1 Eu 1.0 ± 11 Cs 4.8 ± 1.1 Ni 1.0 ± 1.1 Sb 27 ± 1.3 Co 10 ± 11 Ag 44 ± 1.8 V 1.6 ± 1.6 Pb 45 ± 2.9 Hf 1.6 ± 1.1 Cd Se 57 ± 4.8 3000 ± 1.9 E'cnist (X/Al), (X/A1)„ •From Duceet al., 1981. fGeometric standard deviation. On the basis of the measurements made at Enewetak in 1979, Table 5 presents the expected mean atmospheric concentrations for a number of trace metals during the March to June (high Asian dust) period and during the rest of the year (Duce et al., 1981). Concentration units are ng (10^^ g) and pg (10"'^ g) per cubic meter of air. Note that the concentration of all the metals is higher in the spring than the rest of the year, although the increase in concentration for many metals during the spring is not as great as for the metals clearly associated with the dust. For example, while the mean dust associated metals are —25 times higher in the spring, the difference for Pb is less than a factor of 2, Se is about 2, Cd is 5, etc. We assume the source of these "enriched" metals is also pri- marily continental regions. Metals associated primarily with the desert dust have considerably higher concentrations during the spring and early summer due to both stronger source functions (i.e., more frequent dust storms) and wind fields which are conducive to effective long-range transport to Enewetak during that period. However, the enriched elements may have continental sources which are not so seasonal in nature but which are 86 MERRILL AND DUCE TABLE 5 Mean Atmospheric Concentrations of Trace Metals at Enewetak* TABLE 6 Estimates of Annual Atmospheric Deposition of Trace Metals to the Ocean at Enewetak" March-June, Rest of year, ng m"' ng m"^ Al 75 3 Fe 50 2 Mn 1 0.04 Bat 1 0.02 pgm^ pgm"^ Sc 20 1 Cr 200 30 Co 25 1 Eu 2 0.1 Cs 15 0.5 Hf 5 0.2 Rb 200 (5500t) 40 (60t) Ta 15 0.1 Th 20 1 V 120 20 Zn 250 80 Cd 10 2 Cu 50 10 Pb 150(230t) 100(120t) Se 200 100 Sb 5 1 Ag 5 < 1 'From Duce et al., 1981, except as noted. tF. 'om Settle and Patterson. 1982. more uniformly distributed throughout the year. Thus their smaller change in concentration from spring to the rest of the year may largely reflect the changes in atmospheric cir- culation patterns for those time periods. From measurement of these trace metals in rain and dry deposition, estimates can be made of their atmospheric deposition to the ocean surface at Enewetak (Duce et al., 1981). The temfxsral variation in atmospheric concentra- tions shown in Table 5, the monthly rainfall amounts at Enewetak (Fig. 2a), and the measured concentrations of these metals in rain and dry deposition were taken into consideration when the total deposition rates given in Table 6 were calculated. Note that the data in Table 6 suggest that both wet and dry deposition arc important for all elements. There is evidence, however, that much of the measured dry deposition of at least some of these metals may be the result of metals being recycled from the sea surface on sea salt aerosols (Duce, 1982; Settle and Patterson, 1982). This would mean the dry deposition values do not represent a net input of these metals to the ocean. Thus the numbers presented in Table 6 probably represent an upper limit relative to net inputs from the atmosphere to the ocean. Atmospheric deposition Marine sedimen- tation Wet Dry Total Al, fig cm^^ 15 0.4 19 3.3 Fe, ng cm~^ 1.0 0.3 13 2.0 V, ng cm~^ 3 4 7 5t Sc, ng cm"^ 03 0.08 0.4 0.9t Cr. ng cm'^ 5 It 6 2t Eu, ng cm~^ 004 0.006t 0.05 0.05t Cs, ng cm~^ 0,3 0.08t 0.4 0.2t Th, ng cm~^ 0.7 0.2t 09 0.4t Ta, ng cm" 03 0.008 0.04 0.08t Hf, ng cm~^ 1 0.03 0.13 o.it Rb, ng cm'^ 6 1.6 8 4t Cu, ng cm~^ 2 6 8 7 Mn, ng cm~^ 13 5 18 250 Co, ng cm~^ 0.5 o.it 0.6 2 Pb, ng cm"^ 8(6:f) ~4(6t) ~12(12t) 0.61I§ Zn, ng cm"^ 15 -7 -22 7 Cd, ng cm~ 06 <0.8 -1 0.0211 Se. ng cm~^ 5 5" 10 0.00711 Sb, ng cm~^ 12 0.12" 0.24 0.0211 Ag, ng cm"^ 0.2 -0.5 -0.7 — 'From Duce et al , 1981. tEstimated from Al. tFrom Settle and Patterson, 1982 HEstimated from average marine clay composition. §Sum of 0.3 authigenic plus 0.3 silicate lattice. "Estimated from Pb. Marine sedimentation rates for these metals are presented in Table 6 and have been determined from the chemical analysis of surface sediments collected near 29°N 159°W. An estimate of the overall sedimentation rate near Enewetak was determined from mapping measured sedi- mentation rates over the entire North Pacific (M. Leinen, personal communication). Where the surface sediments were not analyzed for a psarticular metal, crustal ratios to Al were used for elements present in crustal abundance in the atmosphere, and average marine clay composition was assumed for the atmospherically enriched elements. It is apparent that the atmospheric deposition to the ocean and the marine deposition to the sediments are very close for Al, Fe, V, Sc, Cr, Eu, Cs, Th, Ta, Hf, Rb, and Cu, suggesting atmospheric transport is very important for marine sedimentation of these metals near Enewetak. Atmospheric input accounts for only a small part of the Mn and Co in the sediments. However, the atmospheric input of Pb, Zn, Cd, Se, and Sb to the ocean is apparently considerably greater than the deposition of these elements to the sediments. There are at least two pxjssible explana- tions for these latter results. First, this would be expected if the atmospheric concentrations and deposition rates of METEOROLOGY AND ATMOSPHERIC CHEMISTRY 87 these metals had resulted from pollution sources on the continents, since the marine sedimentation rates for these metals are mean rates applicable to approximately the past 15,000 years. Input of p)ollution-derived trace metals, which has developed significantly only in the past 50 years or so, would not be reflected in the measured marine sedi- mentation rates. Schaule and Patterson (1982) proposed that a shift may have occurred from principally fluvial inputs of lead to the oceans in earlier times to primarily atmospheric input in recent times. Second, these results would also be expected if a significant fraction of the atmospheric deposition of these trace metals came from their recycling from the ocean surface into the atmosphere and back to the ocean. Recent studies (Weisel, 1981) sug- gest that recycling of marine-derived metals probably does not account for more than a few percent of the mass of these metals in the atmosphere at Enewetak. However, since these ocean-derived metals would be found on the large sea salt particles, their dry dep>osition back to the ocean surface could be rather high (Duce, 1982). Thus, while it is believed that most of the mass of the enriched trace metals in the atmosphere at Enewetak is derived from the continents and very possibly from pollution sources, a significant fraction of the gross dry deposition of these metals into the ocean from the atmosphere may be due to recycled metals from the ocean surface, as mentioned above. Lead isotope ratios reported by Settle and Patterson (1982) confirm that, during the high dust period in April 1979, the pollution-derived Pb had an Asian origin (Tables 7a and b). However, as the Asian dust decreased, the TABLE 7a ^Pb/^'Pb Ratios in Filtered Air Sample at Enewetak' Collection date, 1979 *Pb/^Pb 4/22 to 5/09 5/09 to 5/15 7/12 to 8/10 1.170 1.196 1 205 ^°^Pb/^''^Pb ratio increased and became similar to that for pollution-derived Pb from North America. Thus some, if not most, of the small particle pollution-derived Pb found at Enewetak in the summer may have been transported from North America to Enewetak. On the basis of "Tb and stable lead measurements, Settle et al. (1982) and Set- tle and Patterson (1982) calculated a net atmospheric stable lead deposition rate of 4 to 10 ng cm~ yr at Enewetak. This agrees well with the value of 8 to 12 ng cm~^ yr~^ given in Table 6. The mass median radii (MMR) for the particles contain- ing the various trace metals are presented in Table 8. Note that the sea salt metals (Na, Mg, K, and Ca) have MMRs near 3.5 /zm while the crustally derived metals have MMRs of 0.75 to 1.0 nm. The enriched metals (Zn, Se, Sb, and Pb) have MMRs of <0.5 ^m, consistent with a possible pollution source for these elements. TABLE 8 Mean Particle Mass Median Radii (MMR) for Trace Metals at Enewetak* MMR. MMR. Element fim Element nm Na 3.4 Hf 0.80 Mg 3.5 Rb 1.1 K 3.4 Th 0.84 Ca 3.5 Ta 0.74 Al 0.80 Co 0.75 Fe 0.72 Eu 0.88 Mn 0.88 Ce 0.81 V 0.76 Pb 0.25 Cu 0.96 Se 0.53 So 0.70 Sb -0.35 Cs 0.79 Zn 0.04 'From Duce et al., 1981. Gaseous and particulate mercury were also investigated at Enewetak (Fitzgerald et al., 1981). The concentrations observed are given in Table 9. It is apparent that mercury exists almost entirely as a gas at Enewetak. The relatively small temporal variation in gaseous Hg concentration (and the fact that similar concentrations are found at other marine areas) suggests a relatively long atmospheric TABLE 7b ^Pb/^'Pb Expected from Major Continental Sources* Region *Pb/^Pb ratio Asia/ Japan 1.153 to 1.165 USA. 1.190 (1974) to 1.230 (1978) Mexico 1.187 'From Settle and Patterson, 1982. TABLE 9 Atmospheric Mercury at Enewetak' Collection period, 1979 Gaseous Hg, Particulate Hg, ng m ns m 4/27 to 5/21 6/28 to 8/6 1.6 ±0.6 1.7±0.5 0.0005 00012 •From Fitzgerald et al., 1981. 88 MERRILL AND DUCE residence time for the vapor phase. Studies of Hg specia- tion in the atmosphere at Enewctak indicate that the gas phase is principally inorganic mercury, of which elemental Hg is probably the major component. Mercury in rain at Enewetak was found to have a concentration of 2 ng 1 Apparently this concentration is derived primarily from the washout of the particulate Hg rather than the vapwr phase. are reported in Table 11. It is apparent that the vapor phase dominates the concentration of these compounds, at least from n-C2i to n-C^Q, and probably for the lower car- bon number alkanes as well. Duce and Gagosian (1982) used the concentration distribution in Table 11 to model the input of particulate nalkanes (n-Ci5 to n-C3o) and vapor phase nalkanes (n-Cjo to n-Cao) from the atmo- Organic Carbon The organic carbon concentration of atmospheric f)arti- cles at Enewetak was ~0.9 ^g m~^ (Chesselet et al., 1981) (Table 10). This is typical of marine regions, where the concentration generally ranges between 0.2 and 1.2 /xg m""^. Eighty to 85% of the mass of this organic carbon at Enewetak is found on particles with radii less than 1 nm. TABLE 10 Atmospheric Organic Carbon at Enewetak Concentration Mg m mg Reference Particulate 89 + 017 Rain Chesselet et al., 1981 0.64 ±0 48 Gagosian et al., 1981b Carbon isotope studies by Chesselet et al. (1981) have suggested strongly that the small particle (<1 ^m) organic carbon does not have a marine origin. By measuring both C and C, one can calculate 5 C as follows: 5'3c = \ ^/ *^/sample /13/-;12, C/ CJstandard -1 X 1000 5 C values calculated for the smallest particles (r < 1 /xm) are -26°/oo to -28°/oo. Chesselet et al. (1981) point out that this range is similar to b^^C values for continental vegetation, coal, and the products of jsetroleum combus- tion, — 26 ± 2°/oo, suggesting the small particle carbon is of continental origin. The d^'^C values for the larger parti- cles (r > l^im) are -187oo to -22° I ^. This is similar to the 6 C value for marine organic carbon, which is gen- erally -21 ± 2°/oo in low latitude regions (40°S to 50°N), suggesting the large particle carbon in the Enewetak marine atmosphere is of marine origin. The organic carbon content of rain at Enewetak aver- aged 1.2 mg l~^ during April and May and 0.3 mg 1 during July and August 1979 (Gagosian et al., 1981b). Organic Lipid Class Compounds Particulate and vapor phase heavy n-alkanes were measured independently in the atmosphere at Enewetak by two research groups in 1979. The observed concentrations TABLE 11 Atmospheric N-Alkancs at Enewetak Concentration Particulate.* ng m^ Gaseous N-alkane ng m~^" ng m *t nCi3 0.23 n-Ci4 0.19 n-Ci5 0.66 nCi6 0.13 n-Ci7 0.55 n-Ci8 0.07 n-C,9 0.07 n-C2o 007 n-Cji 0.0017 0.07 n-C22 0.0020 0,07 n-C23 00023 0.09 n-C24 00021 0.12 n-C25 0.0030 095 0.12 n-C26 0.0020 0.088 0.09 n-C27 00067 0055 007 n-C28 0.0037 0.024 n-C29 0.0170 0.019 n-Cjo 0.0033 0.013 'Gagosian et al , 1981b, 1982. fAtJas and Giam, personal communication, 1981. sphere to the ocean. Consideration was given to rain scavenging of both aerosol and vapor phase n-alkanes, dry depKDsition of aerosol n-alkanes, and direct gas exchange with the ocean of vap>or phase n-alkanes. Estimates of the atmospheric input of n-alkanes to the ocean at Enewetak are given in Table 12. Note that rain scavenging of n- alkanes on particles appears to be the primary method of n-alkane removal from the atmosphere. It can be seen from Table 11 that the odd carbon number n-alkanes on aerosols have higher concentrations than the adjacent even carbon number n-alkanes. This is observed for higher n-alkanes up to n-Cae as well (Gago- sian et al., 1980). The odd-to-even carbon preference index and the fact that the major alkanes are n-C27. n-C29, and n-Csi strongly suggest that the source of these heavier n-alkanes present on aerosols is vascular plant waxes, probably of Asian origin (Gagosian et al., 1981a). Concentrations of fatty alcohols, fatty acid esters, and fatty acid salts were also measured in the Enewetak atmo- sphere and are presented in Table 13 (Gagcsian et al.. METEOROLOGY AND ATMOSPHERIC CHEMISTRY 89 TABLE 12 Estimate of Annual Atmospheric Deposition of n-Cio to n-Cao Alkanes to the Ocean at Enewetak" Deposition mechanism Deposition rate, lO'^gcm^yr' Particulate: Wet Dry VafKjr Phase: Wet Dry Total 6.2 to 62 0.8 to 8 to 0.00001 Oto 1.4 7 to 71 'From Duce and Gagosian, 1982. TABLE 13 Concentration of Fatty Alcohols, Fatty Acid Esters, and Fatty Acid Salts on Atmospheric Particles at Enewetak" Organic substance Concentration range, pg m~^ Fatty alcohols C21"C32 Fatty acid esters C13-C20 C21~C32 Fatty acid salts C13-C20 C21~C32 2 to 85 58 to 210 34 to 290 6 to 91 87 to 4000 36 to 670 'From Gagosian et al., 1981b. 1981b, 1982). These authors suggest, on the basis of odd/even carbon concentration ratios and the concentra- tion distribution observed, that the fatty alcohols and the C21 to C32 fractions of the fatty acid esters and fatty acid salts have a natural terrestrial source, whereas the lighter Ci3 to C20 fatty acid esters and fatty acid salts in the Enewetak atmosphere likely have a marine source. Heavy Chlorinated Hydrocarbons and Other Synthetic Organics A number of synthetic organic substances were mea- sured in the Enewetak atmosphere, including PCBs and certain pesticides and plasticizers (Atlas and Giam, 1981). The concentrations observed in the vapKjr phase are presented in Table 14. Concentrations of these substances on aerosols were less than 10% of the vapor phase con- centrations. Altas and Giam (1981) point out that rain may not be the primary mechanism for removal of these vapor phase organic substances from the air. Direct vapor exchange with the ocean may be most important. By using a vapor deposition velocity of 8 m h ' for PCB 1242, the direct vapor exchange PCB flux to the ocean at Enewetak would be 35 X 10 '° g cm"^ yr"^ This is at least 50 times greater than the precipitation flux of <0.8 X 10"'° g cm^^ yr"' for Aroclor 1242, which can be calculated using the concentrations in Table 14 and the annual rainfall amount in Fig. 2a. The vapor phase flux would result in a PCB Aroclor 1242 atmospheric residence time of about 30 days, which would explain the relatively uniform concentration distribution of Aroclor 1242 over the Atlantic and Pacific Oceans and nonurban continental areas (Atlas and Giam, 1981). Phthalic acid esters are present in rather high concen- trations at Enewetak. These compounds are widely used as plasticizers. The concentrations are similar in air over the Atlantic Ocean, the Gulf of Mexico, and in a rural area in TABLE 14 Concentration of Synthetic Organic Compounds in the Air and in Rain at Enewetak' Air concentration. Rain concentration. ng m ngl ' Compound Mean Range Mean Range PCB, Aroclor 1242 0.54 0.35 to 1.02 <0.6 PCB, Aroclor 1254 0.06 Hexachlorobenzene 10 0.095 to 0.13 <0.03 a Hexachlorocyclohexane 0.25 0.075 to 0.57 3.1 1.3 to 6.8 7 Hexachlorocyclohexane 0.015 0.006 to 0.021 0.51 0.34 to 1.6 Chlordane (a and 7) 0.013 006 to 0.015 <0.02 Dieldrin 0.010 006 to 0.018 <0.02 p.p'-DDE 0.003 0.002 to 0.005 <0.02 Di-n-buthyl phthalate 0.87 0.40 to 1.8 31 2.6 to 73 Di-(2-ethylhexyl) phthalate 1.4 0.32 to 2.7 55 5.3 to 213 'From Atlas and Giam, 1981. 90 MERRILL AND DUCE Texas, suggesting relatively long atmospheric residence times. Urban area concentrations are about 100 times higher (Atlas and Giam, 1981). ACKNOWLEDGMENTS Wc thank the staff of the University of Hawaii's Mid- Pacific Research Laboratory, the Department of Energy, and Holmes and Narver, Inc. for field support in Enewetak. Paul Dellegatto assisted in the reduction of the wind rose and cloudiness data. Supported by NSF Grants OCE 77-13072, CX:E 77-13071, and OCE 81-11895. REFERENCES Atlas, E , and C S Giam, 1981, Global Transport of Organic Pollutants: Ambient Concentration in the Remote Marine Atmosphere, Science, 211: 163 165 Blanchard, D. C., 1963, Electrification of the Atmosphere by Par- ticles from Bubbles in the Sea, Progress in Oceanogr . 1: 71-202 Blumenstock, D 1, 1961, A Report on Typhoon Effects Upon Jaluit Atoll, Atoll Res. Bull., 75: 1105. , and D F. Rex, 1960, Microclimatic Observations at Eniwetol^, Atoll Res Bull , 71: 1158, Chesselet, R , M. Fontugne, P Buat-Menard, U. Ezat, and C. E. Lambert, 1981, The Origin of Particulate Organic Carbon in the Marine Atmosphere as Indicated by Its Stable Carbon Iso- topic Composition, Geophi^s Res Lett., 8: 345-348. Chapman, S., and R. S. Lindzen, 1970, Atmospheric Tides Ther- mal and Grauitational. Gordon and Breach Science Publishers, New York. Duce, R. A., 1982, Sea Salt and Trace Element Transport Across the Sea/Air Interface. Presented at the Joint Oceanographic Assembly, Halifax, Canada, Aug. 4 and R. B Gagosian, 1982, The Input of Atmospheric n-Cio to n-C3() Alkanes to the Ocean, J Geophys Res, 87: 71927200. , G. L Hoffman, and W. H Zoller, 1975, Atmospheric Trace Metals at Remote Northern and Southern Hemisphere Sites: Pollution or Natural?, Science. 187: 59-61 C. K. Unni, B. J. Ray, P J. Harder, A P Pszenny, and J. L. Fasching, 1981, The Atmospheric Concentration of Trace Elements and Their Deposition to the Ocean at Enewetak Atoll, Marshall Islands, in Svmposium on the Role of the Oceans in Atmospheric Chemistry, lAMAP Third Scien- tific Assembly, Hamburg, F.R.G. , C. K. Unni, B J Ray, J M. Prospero, and J. T Merrill, 1980, Long-Range Atmospheric Transport of Soil Dust from Asia to the Tropical North Pacific: Temporal Variability, Science. 209: 1522-1524 , and A. H. Woodcock, 1971, Difference in Chemical Compo- sition of Atmospheric Sea Salt Particles Produced in the Surf Zone and on the Open Sea in Hawaii, Tellus, 23: 427-435. Fitzgerald, W F., G. A. Gill, and A. D. Hewitt, 1981, Mercury, A Trace Atmosptieric Gas, in Symposium on the Role of the Oceans in Atmospheric Chemistr\^. lAMAP Third Scientific Assembly, Hamburg, F R.G. Gagosian, R B , E. T Peltzer, and O. C. Zafiriou, 1980, Und- Derived Organic Compounds in Enewetak Particulate Sam- ples, SEAREX Newsletter. 3: 10-15. , E T. Peltzer, and O. C. Zafiriou, 1981a, Atmospheric Transport of Continentally Derived Lipids to the Tropical North Pacific, Nature, 291: 312-314. E. T. Peltzer, and O C Zafiriou, 1981b, Organic Com- pounds in Vapor Phase and Rain Samples from the Enewetak Experiment, SEAREX Newsletter. 4: 31-35. O. C Zafiriou, E T. Peltzer, and J. B. Alford, 1982, Lipids in Aerosols from the Tropical North Pacific: Temporal Vari- ability, J Geophvs Res., 87: 11133-11144. Ing, G. K. T., 1972, A Duststorm over Central China, 1969, Weather. 27: 136-145. Lavoie, R L., 1963, Some Aspects of the Meteorology of the Tropical Pacific Viewed from an Atoll, Atoll Res Bull., 96: 1-80. McDonald, R. L., C K Unni, and R. A. Duce, 1982, Estimation of Atmospheric Sea Salt Dry Deposition: Wind Speed and Particle Size Dependence, J Geophys. Res . 87: 1246-1250. Newell, R. E., J. W. Kidson, D. W. Vincent, and G. J. Boer, 1972, The General Circulation of the Tropical Atmosphere and Interactions with Extratropical Latitudes, Vol. 1, MIT Press, Cambridge, Massachusetts. Rahn, K. A., 1976, The Chemical Composition of the Atmo- spheric Aerosol, Technical Report, Graduate School of Oceanography, University of Rhode Island, Kingston, Rhode Island Reed, R J , and E E Recker, 1971, Structure and Properties of Synoptic-Scale Wave Disturbances in the Equatorial Western Pacific, J Atmos Sc> . 28: 11171 133. Schaule, B K., and C. C. Patterson, 1981, Lead Concentrations in the Northeast Pacific: Evidence for Global Anthropogenic Perturbations, Earth Planet. Sci. Lett.. 54: 97-116. Settle, D. M., C. C. Patterson, K. K Turekian, and J. K. Cochran, 1982, Lead Precipitation Fluxes at Tropical Oceanic Sites Determined from ^"^b Measurements, J Geophys. Res . 87: 1239-1245 , and C C Patterson, 1982, Magnitudes and Sources of Pre- cipitation and Dry Deposition Fluxes of Industrial and Natural Leads to the North Pacific at Enewetak, J Geophys Res . 87: 8857-8869. Taylor, R. C, 1973, An Atlas of Pacific Islands Rainfall. Hawaii Institute of Geophysics, University of Hawaii, Honolulu, HIG- 73-9. Trenberth, K E., 1976, Spatial and Temporal Variations of the Southern Oscillation, Quart Jour. Roi;al Meteor. Soc , 102: 639-653. Turekian, K. K., and J. K. Cochran, 1981a, The Temporal Varia- tion of ^'"Pb Concentration in Surface Air and Precipitation at Enewetak and Its Use in the Prediction of the Asian Dust Flux to the Pacific, Nature. 292: 522-524 and J K Cochran, 1981b, ^"^b in Surface Air at Enewetak and the Asian Dust Flux to the Pacific: A Correction, Nature. 294: 670. Weisel, C. P., 1981, The Atmospheric Flux of Elements from the Ocean. Ph.D. Dissertation, University of Rhode Island, Kings- ton, Rhode Island. Woodcock, A. H., 1953, Salt Nuclei in Marine Air as a Function of Altitude and Wind Force, J. Met.. 10: 362-371. Chapter 7 Subtidal Environments and Ecologx; of Eneioetak Atoll PATRICK L. COLIN Motupore Island Research Station. Universittj of Papua New Guinea. Port Moresbi>. Papua New Guinea INTRODUCTION The area of subtidal marine environments of Enewetak far exceeds the intertidal and terrestrial habitats. Subtidal environments are the lagoon and outer reefs and the pas- sages between them which arc submerged at low tides. The lagoon and outer reefs are separated, except at passes, by the intertidal reef flat. Although closely con- nected to the subtidal habitat, the intertidal habitat of Enewetak is not discussed in this chapter except as it relates to circulation, energetics, and processes in the sub- tidal environment. With an area of about 930 km , the lagoon is more than 15 times the area of the intertidal reef flat and 140 times the area of the islands. The Enewetak Lagoon is relatively deep by atoll stan- dards (Wiens, 1962), with a mean depth of 48 m and a maximum depth of 64 m. Only about 10% of the lagoon is shallower than 18 m, and only 20% is less than 32 m (Emery et al., 1954). The open waters of the lagoon are a voluminous habitat of about 4.2 X 10^° m^ of water, most of which overlies the deep portions of the lagoon. In both area and volume, the lagoon is the largest subtidal unit at Enewetak. For this chapter, an arbitrary depth of 30 m has been selected to distinguish between the "deep" and "shallow" portions of the lagoon. The area of the reefs seaward of the reef flat and islands has never been accurately determined. Based on reef widths observed from the air compared to adjacent reef flat widths, the area of the seaward reefs is certainly less than that of the reef flat, perhaps by a factor of up to 3 or 4, but an accurate determination is not presently pos- sible. The outer slope of the atoll is quite different from the lagoon. The present discussion will include descriptive information on the outer slope to 300 to 400 m depth, but below those depths there is little detailed information concerning the biological communities or geological per- spectives. The subtidal environment of Enewetak contains a number of units, divisible on the basis of location, physical factors, substrate types, dominant organisms, and other factors. Biological communities can be similarly identified. The generally high diversity of marine organisms at Enewetak increases the complexity in describing individual communities and their relationship with the others. MARINE CONDITIONS Mean surface oceanic water temperatures at Enewetak range between 27° and 29°C (Atkinson et al., 1981; Coles et al., 1976), with March the coolest month and August the warmest. Temperature extremes during any one month vary about ± 1°C from the mean (Coles et al., 1976). Local conditions can alter these, with isolated tide pools in the midday sun reaching the low 40s C. Typical temperature-depth profiles for the seaward reefs of Enewetak during summer are shown in Fig. 1. From the surface to about 125 m in depth, the tempera- ture gradually decreases from 29° to about 25°C. A slight thermocline begins at about 125 to 150 m, and it changes from 25° to 20°C over a 30 to 40 m increase in depth. At 220 m it is about 13°C with the temperature gradually decreasing to about 9°C at 380 m. Atkinson et al. (1981) documented the isothermal and isohaline nature of the lagoon water column with no more than a 0.5°C variation in temperature and a 0.20 ppt range in salinity. Almost without exception the shallow waters of the open lagoon and ocean are ideal for eury- thermal tropical organisms. The salinity of the lagoon is essentially the same as that of the open ocean. Only in areas of restricted circulation and shallow depth are tem- peratures elevated significantly above genera! lagoon water. The atoll is located in the North Equatorial Current with a general westward movement of water past the atoll. Currents observed on the windward (east) ocean reefs agree with this generalized picture, even at depths of 100 to 200 m, with the ocean current seeming to split north and south near the easternmost extension of the atoll at Ananij. On the ocean side of Enewetak Island, the alongshore component of the current varies in speed but 91 92 COLIN TEMPERATURE C. 25° 30° 30° 30 30 Fig. 1 Tempcrature/dcpth profiles from four dives by the DSRV Makali'i at Enewetalt Atoll, summer 1981. Locations of profiles were as follows: a, Biken, west side of the atoll; b. east side of wide channel; c. Bokandretok, just north of Enewetak Island; and d. Runlt Island. Dual tracks represent an ascent and descent profile. has never been observed to move northward. It is only near passes into the lagoon that tidal currents potentially cause reversal of current direction on the ocean slope of the atoll. On the leeward (west) side of the atoll, currents are variable, and an eddy pattern often seems to exist. At the West Spit, the extreme northwest tip of the atoll, the currents from north and south seem to converge. Currents in the lagoon and passes are considerably dif- ferent and are covered in detail by Atkinson et al. (1981) and in Chapter 5 of this volume. In the lagoon the surface current is generally a wind-driven westerly flow with mid- depth return flow to the east. Most water enters the lagoon over the windward reefs and passes out the wide (south) channel. The deep channel has strong tidal flow but little or no net input into the lagoon. Water residence times have a mean of about 30 days but can vary between a few to more than 130 days. The residence time of water in the northern p)ortion of the lagoon is greater than the mean. The nutrient-poor oceanic water eastward of Enewetak is clear, with visibility normally exceeding 50 m. Such water visibility is typical of windward ocean reefs, but visi- bility changes when water enters the lagoon over the reef flat. Increased production and suspended particulates reduce the visibility of lagoon waters to 10 to 25 m and occasionally less. Aerial photographs of Enewetak have features on the ocean side reefs visible to depths of about 40 m, although this is limited to a maximum of about 15 to 18 m in lagoon waters. In the northern lagoon, particu- larly near the islands between Engebi and Bokoluo, two factors may contribute to visibilities at less than 10 m. First, lagoon water residence times in this area are near the maximum, hence high densities of phytoplankton and zooplankton can develop in this water. Second, the pres- ence of fine, easily suspended particulates produced as a result of nuclear tests and cratering in this area may greatly reduce water visibility. Phytoplankton blooms, man- ifested both as "brown water" and large, thick windrows of extremely dense phytoplankton, have been observed on several occasions in the northwestern lagoon. Visibility in such waters is consequently extremely limited. Trade wind conditions with steady 10 to 20 knot winds from the east to northeast prevail throughout most of the year at Enewetak. During the summer, trade winds are usually lighter than during the winter, and they can cease for periods of several days. The normal trade winds produce oceanic waves about 1 to 2 m high which hit the windward reef of the atoll. Within the lagoon, the margin on the windward side is relatively calm, being protected by islands and interisland reefs. At high tide, however, much more wave action comes across the reef flat from ocean to lagoon, making conditions choppier on the lagoon margin. The waves which do cross the reef are of small height and wave length, making the surface rough for small boats but having little effect below a few meters depth. Moving westward across the lagoon, a significant fetch is achieved quickly, and when trade winds prevail, the cen- tral and western areas of the lagoon are far from placid. Waves of 1.5 to 2 m with whitecaps can occur, and the lagoonward edge of the leeward reef can have significant surf breaking on it. Significant wave action can also reach the lagoon sides of the islands west of Engebi and the southwestern islands. The ocean side of the leeward reefs and islands is calm under trade wind conditions with the tide level affecting the wave action crossing the reef from lagoon to ocean. In the lee of islands it is extremely calm for an oceanic area. Waves in the Marshall Islands as a whole are from the east or northeast, a consequence of persistent trade winds. Waves exceeding 3.5 m high comprise fewer than 2% of waves in the Marshall Islands area (Japan Meteorological Agency [JMA], 1971 to 1978). Waves greater than 3.5 m can occur any time of the year and are generally associ- ated with (1) local storms or typhoons from the east through southwest and (2) more distant northern and southern hemisphere storms. The greatest wave amplitude observed (JMA report) was a long-period 6.5-m swell from the northeast. Local conditions can also greatly affect wave action. Where tidal currents lun against the trade winds, steep standing waves develop. The east channel at Enewetak can be treacherous under strong trade winds with the tide dropping sharply. A distinct central tongue of breaking waves extending out the channel to the ocean is visible from the air under these conditions. Similarly, the west SUBTIDAL ENVIRONMENTS AND ECOLOGY 93 side of the wide passage is an area of merging waves from the lagoon and refracted oceanic waves and swell with strong southerly currents producing standing waves and short, steep seas. During summer calms and at odd times during the rest of the year, the lagoon and surrounding ocean become smooth. At such times it is possible to swim off the reef flat to the windward ocean reefs because the surf is small and gentle Surface slicks are found in lagoon waters. The windrows of phytoplankton from blooms have been found during such calm periods. Swells from distant storms create a different situation in which shores exposed to the swell (which can be either windward or leeward) are heavily pummeled, whereas the waves produced by wind in the immediate area may be small. Most impressive are those rare times when large swells thunder against the reef while calm trade winds pro- duce an almost mirror like surface elsewhere. On shores normally lacking high surf, these waves can cause consider- able damage. Such swells can also enter the lagoon through the southern pass to break on the lagoon shore of Enewetak, Medren, and other islands. Several typhoons and near-typhoon strength storms have passed by or over Enewetak during the last decade. Although the atoll is generally not considered in the typhoon belt, these storms have had a significant and readily visible effect on Enewetak reefs. Often a storm whose center does not pass especially close to Enewetak can produce storm waves which severely damage reefs, although above-surface damage from winds may be very light. Deep Lagoon Biological Communities The lagoon bottom below 30 m depth consists largely of soft substratum with small to large reef structures (pin- nacle reefs) spread randomly throughout the area. Except for the pinnacle reefs, relatively little published information exists regarding in-situ observations of either the lagoon slope or deep lagoon benthic biological communities or geology. Nearly all published information is based on surface-lowered grab samplers or dredges. Emery et al. (1954) reported on results of samples taken by an "under- way" bottom sampler and about 50 photographs taken by a remote camera at unspecified locations in the deeper lagoon. More recent researchers (Nelson and Noshkin, 1973; Noshkin, 1980) have relied on grab samplers or short-core samplers to obtain bottom samples for analysis. The only published in-situ observations of the deep lagoon are those of Gilmartin (1960), made during deep scuba dives on a transect across the southern lagoon. Twelve of these stations were below 30 m in depth. All stations below 30 m had coral patches present within the range of visibility; these varied from only a few small heads in one instance to massive patch reefs rising 15 m or more above the surrounding bottom. Some stations had the substratum covered with "mounds of sand and cast- ings," but for most of the deeper stations, the presence or absence of mounds was not noted. This study confirmed that abundant algal communities exist in the deep areas of Enewetak Lagoon, many occurring at the deepest depths reached (62 m). Coral patches at these depths seemed more densely populated with algae than adjacent sand. Eight species of Halimeda were found primarily between 42 and 62 m, supporting previous reports that the genus was "most common and luxuriantly developed at the deeper levels." Gilmartin (1960) was the first person to realize the intense bioturbation of the soft substrate bot- toms of the lagoon, commenting that "the continual 'churn- ing' of the substratum by these benthic organisms has prevented algae, which might occur elsewhere on the same stations, from starting and growing to the fX)int where they would not be 'uprooted' or buried by the sand displace- ments." During 1980 and 1981, a distributional survey of deep lagoon benthic communities was conducted using a lowered camera system. During this "Enewetak Benthic Survey" over 2000 photographs covering 24 m each were taken at 190 stations throughout the deep lagoon (Colin, 1986). Additionally, in the summer of 1981 the submersible Makali'i was utilized for a series of dives in several areas of the deep lagoon to augment the photo- graphic survey. STATE OF KNOWLEDGE OF SUBTIDAL MARINE ENVIRONMENTS With certain exceptions, the subtidal marine environ- ments of Enewetak cannot be characterized as well known. Often our knowledge is based on studies in the southern lagoon close to the lee of the southern islands. The dis- tances from support bases, the generally rough conditions of the lagoon outside protected lee areas, and the rapidly increasing water depth have severely limited work in both the northern and central lagoon. Much of the work accom- plished in the southern lagoon is of equal applicability to the entire lagoon, but differences do exist between these areas. Even the southern areas of the lagoon below 20 to 30 m depth are poorly known. This is due to the limited access of scuba-diving scientists to the deep lagoon bot- tom, particularly at its most common depths of 40 to 60 m. Although a few hardy souls have ventured to dive in these areas, working time is limited, nearly precluding studies providing an understanding of overall conditions in the deep lagoon. Significant work on the deep lagoon floor requires either specialized instrumentation and recording methods or suitable vehicles for in-situ work. The present work is intended to provide descriptive information about subtidal habitats in the following order: (1) deep lagoon; (2) shallow lagoon; (3) lagoon-ocean passes; and (4) the seaward reefs, from the center of the lagoon outward. The information has been drawn from publications, annual reports, unpublished information in 94 COLIN MPRL files, and unpublished data from numerous scien- tists. No comprehensive descriptive account of the subtidal environments of Enewetak has been attempted. Only a few specialists have endeavored to discuss atollwide distribu- tion and contributions of their restricted groups of organ- isms. Cuffey (1973 to 1978) examined the role of bryozoa at Enewetak with comparison to other reef areas. He listed three major marine benthic "macrohabitats" at Enewetak: the coral-dominated, the bedrock-dominated, and the sediment-dominated. He distinguished between "larger patch reefs (from 25 to more than 150 ft height)" and "smaller coral knolls (from 1 to 25 ft high)" in examining bryozoan distribution at Enewetak. He also distinguished between biohcrms ("coral-dominated macrohabitats with significant height") and biostromes ("coral-dominated macrohabitats lacking significant height"), such as his "coral pavement." Allen (1972), in his work on anemonefishes, provides brief descriptions of some Enewetak habitats. The major physiographic features of environments from outer reef slope, reef flat, shallow lagoon margin, and deep lagoon arc mentioned. He thought that the deep lagoon floor "appears to be of a fairly uniform nature" and had "large stretches of sand with orcasional small patch reefs." Deep lagoon pinnacles were described as "an oasis, rising from the barren lagoon floor" and harboring "an extraordinary wealth of marine organisms." The deep lagoon can be characterized as dominated by sediment substrates but with reefs of widely varying size and vertical relief, distributed fairly evenly throughout the lagoon. The soft substrate supports several different biolog- ical communities, often occurring within short distances of other soft substrata and reef substrata. Extensive distur- bance of the sediments is evident in many of the benthic photographs and from submersible dives. Based on point counts of the benthic survey photo- graphs, approximately 859b of the deep lagoon has soft substrata, with the remaining 15% hard substratum. Nearly half of the stations had 100% coverage of soft substrata, more than 60% were 90% or more soft substrata, and more than 75% were 75% or more soft substrata cover- age. If the individual photographs are considered, rather than entire stations, slightly higher percentages of 100% and 90% soft substrate coverage are found. The soft substrate biological communities comprise four identifiable types. These include (1) open sand sub- strate without a visible algal mat, (2) sand substrate with visible algal mat on its surface ("algal film"), (3) sand sub- strate with macioalgae, particularly species of Halimeda, on its surface ("algal flat"), and (4) sand with large popula- tions of an unattached Fungiid coral, Cxjclosersis and Diaseris spp. ("button corals"). Typical views of these com- munities from benthic survey photographs are shown in Figs. 2 and 3. Interpretation of the benthic survey photo- graphs has been facilitated by observations and photo- graphs from the Enewetak submersible project and scuba dives in shallower areas where similar communities occur. The soft substrate communities often intergrade, for example, the macroalgae of the "algal flat" community decreasing in density until only open sand remains. Arbi- trary points at which one community "becomes" another have been used in interpreting the photographs, but abso- lute distinctions among types of soft substrate communities are often impossible. The distributions of community types in the deep lagoon as based on benthic survey photographs are shown in Figs. 4 and 5. Deep lagoon sediment substrata with no visible algal cover are qualitatively similar to areas of the lagoon mar- gin as shallow as 15 m. They are usually heavily bioturbated, dominated by the conical mounds produced by callianassid shrimp. The occurrence of open sand sul>- strates, based on benthic survey photographs, is shown in Fig. 4. Although this covers only a limited number of sta- tions, it does indicate "barren" soft substratum can occur throughout the lagoon. Open sand substratum, however, can change within a few meters horizontally to soft sub- stratum covered with macroalgae. Such short-scale horizon- tal changes among soft-substratum communities and hard substrata are common throughout the lagoon. It is possible that the rapid sediment turnover in open sand areas is responsible for the lack of dark algal films of macroalgae. However, algal mats over 1 m in diameter do occur in heavily bioturbated areas but are capable of form- ing in only a few days time. Biological sediment overturn is concentrated at specific points in the short term (days) pro- ducing "splotching" of algal mats when viewed from above. Other factors affecting density of algal films (nutrients in water or sediment, water clarity, standing crop already present, etc.), may be critical in controlling the presence of dense algal mats. The presence of algal films, particularly diatoms and blue-green algae, on sand bottom without visible algal mat is well known (J. T. Harrison, personal communication). The population level at which an algal mat becomes visible in a photograph or to a human observer is dependent on the standing crop per unit area and the plants involved. Water visibility over open sand bottoms is often limited to only 5 to 10 m, even in the deep lagoon. Considerable amounts of suspended particulates were observed from the submersible Makali'i, using its lights, in the deep lagoon; but relative densities varied considerably from day to day at one location. It was noted, however, that suspended particulates were often elevated closer to the bottom than near the lagoon surface. Similar observations commonly have been made by scuba divers over open sand substrates at depths around 20 m. Sediment will often have a clearly visible thin layer of microalgae on its surface. Algal films are seen in the lagoon margin areas as shallow as 15 m. At depths of 15 to 30 m, small dense algal mats (only a few centimeters in diameter) are often seen on otherwise clear bottoms. Where a film of algae exists, any recent disturbance of the sediment is clearly indicated by lack of, or disturbance of, the algal mat. This relationship has been closely examined at diving depths from 15 to 30 m and has been verified in SUBTIDAL ENVIRONMENTS AND ECOLOGY 95 f.- Fig. 2 Enewetak deep lasoon soft bottom communities. Bar equals approximately 1 m. a, Algal film community photographed by vertical benthic camera. Considerable bioturbation (pale splotches) is visible in this photograph (56 m depth); b, Algal flat com- munity, vertical l)€nthic camera, dominated by species of Halimeda (51 m depth); c. Algal flat community dominated by mixed macroalgae, vertical benthic camera (55 m depth); d. Algal flat community at 57 m depth, central lagoon, dominated by mixed macroalgae, with small Halimeda sp. thalli. Some mounds produced by bioturbation are visible (diver photograph); e, Same gen- eral area as above with sponges (dark objects near center), Halimeda sand and abundant macroalgae visible, 57 m depth (diver photograph): f. Near vertical view of area of Fig. 2d and 2e, 57 m depth. Significant bioturbation is evident in this diver photo- graph. [Figures from Colin, 1986.] 96 COLIN Fig. 3 Enewetak deep lagoon communities. Bar equals approximately 1 m length, a. Algal flat community dominated by Caulerpa, vertical benthic camera photograph (51 m depth): b, Oblique view of sediment bottom with Caulerpa sp. and small patch reefs in left background. 44 m depth. Note the mounds produced by callianassids on sediment. Diver photograph; c, Benthic community with "button corals" and Caulerpa algae, vertical benthic camera (55 m depth); d, Dense "button coral" community (Cycloscris and/or Diaseris) at 56 m depth, vertical benthic camera; e. Hard substratum community with relatively barren rock surfaces, vertical benthic camera (44 m depth); f, Hard substratum community with patch reefs with stony corals and gorgonians, vertical benthic camera (47 m depth). [Figures from Colin, 1986.] SUBTIDAL ENVIRONMENTS AND ECOtOGY 97 Soft Substrate Coverage O >50% O >90% Algal Flat Algal Film C >2S% 3 >25% Algal Flat with Algal Film • >25% each Button Corals O>50-100 m-2 to o 9 "^ ^N^W^TA^ ATOLC r\j| iM IM (M Ml .359 256 -ll- JO' .254 .252 ix> o Sl,250 |J4S .24t, Fig. 4 Distribution of soft substratum communities at benthic photographic station in the deep lagoon (<30 m depth), Enewetak Atoll (Colin, 1986). 98 COLIN Hard Substrate Coverage *>20% >50% V lO , _^ ; . I \fN^WETAf: _ ^ATQlI^ , _ . , i ' i .1 C ^ ecies occurring at Enewetak are documented, other biological information is usually not known. In one of the few instances where more than the base essentials are known, several species of irregular sea urchins occur buried in, or on, sediments. The density of given sp>ecies In apparently similar areas of the lagoon margin has been documented to vary by well over an order of magnitude (V. S. Frey, unpublished data). Similar population varia- tion has also been observed at a single station over severed months. Although these variations have been documented, the many factors determining papulation structure of infau- nal organisms are poorly understood. The smaller organisms dwelling in sediment bottoms are more jxxjrly known. For example, using a technique where an area of bottom is covered by a plastic sheet and rotenone, or another toxicant, introduced beneath the sheet for a time, lancets (Branchiostomldae) have been col- lected recently at a density of approximately 100 individu- als m~^ on sediment bottoms below 15 m at Enewetak (Suchanek and Colin, 1986). Schultz et al. (1952), in spite of their collecting efforts in the Marshedl Islands, took only a single specimen of lancet at Bikini Atoll. Approximately 50 small unidentified ghost shrimps were collected per square meter using this technique, a density far greater than imagined. The only visible evidence for the presence of these small calllanassids is small-scale conical mounds present in combination with larger mounds produced by larger species. Also collected were stomatopods, sipuncu- lids, molluscs, and echinoids (Suchanek and Colin, 1986). Interestingly, in sediment-leveling experiments the number of small-scale mounds (less than 5 cm diameter) was an order of magnitude or more greater than large-scale mounds, supp>orting evidence of the high populations of small callianassids (Suchanek et al., 1986). Pinnacle Reefs of the Deep Lagoon It is impossible to draw an absolute line where the f)atch reefs on the margin of the lagoon and pinnacle reefs begin. A working distinction can be made between "patch reefs" which rise from a surrounding sediment or rock bot- tom which is visible from the surface under normal condi- tions and "pinnacle reefs" rising from depths where the 100 COLIN surrounding bottom is not visible. Emery et al. (1954) used the term "coral knoll" for such structures, but this author thinks it is not truly a descriptive term in this case. The pinnacle reefs of Enewetak Lagoon cover only a small percent of the bottom area but are areas of great biological diversity and interest. Their presence in the deep lagoon, appearing as light areas among the dark waters, parallels on a smaller scale the presence of atolls in the deep ocean. Pinnacle reefs vary greatly in size, from a few tens of meters to over 1 km in diameter at their base. Emery et al. (1954) pointed out that among the 20 largest pinnacle reefs, they are quite evenly spaced throughout the lagoon. For several reasons the distribution of smaller pinnacles, though, is not as well known. The tops of most are not visible from the surface and because of their small size, they are easily missed by echo sounding surveying. Emery et al. (1954) estimated there were about 3000 coral pinnacles in the Enewetak Lagoon but ignored any which did not have a relief of more than 4 m. There are about 150 to 180 pinnacles which should be visible from the surface (<18 m depth), rising from depths of about 35 m or more. The surface-visible pinnacles are the best known because they can be located relatively easily for diving and are shallow enough for prolonged scuba diving. They have been used as sites for a variety of studies, but their origins and underlying structure are not well known. The slope of the sides of pinnacle reefs can vary greatly. In general, the smaller a pinnacle reef diameter, the steeper its slope. On small pinnacles much of the sloF>e is nearly vertical. The largest pinnacles are somewhat flat on top for much of their diameter but still slope to the lagoon floor at an angle of at least 10 to 20°. Those pinnacles closest to Enewetak Island are best known because of their closeness to MPRL. Figure 6 indi- cates the location of many of these and the names applied to them. There is, however, considerable variation in the biological communities between pinnacles, even among those of similar size and shape. A few pinnacle reefs are described subsequently in greater detail. An example of a well-developed small, but not typical, pinnacle reef is "Pole Pinnacle," so named because of a toppled marker pole and anchor block on its upper sur- face. It is located 1.6 km from Jedrol Island (Fig. 6). Pole Pinnacle actually rests on the edge of the deep channel on an extension of the wedge of shallow reef produced by the split of the deep channel west of Jedrol. The entire upper surface of the pinnacle is dominated by the coral Pontes rus, the P. iuxjyamaensis of Wells (1954) (Veron and Pichon, 1982). On the upper surface at 3 to 5 m depth, the columnar form of P. rus occurs, but on the sides of the pinnacle where P. rus also dominates, the plate- columnar form occurs. The vertical distribution of P. rus varies on different sides of the pinnacle. On the northern face, little occurs below 8 m, whereas on the south side a solid cover is found above 12 m. The eastern face has its first colonies of P. rus at about 26 m, with large patches starting at 18 m. The western face has some large clumps as deep as 15 m. Below the steeply sloping upper portion, the bottom becomes less steep, having an angle of about 45° to depths of 30 m. The bottom around the base of the pinnacle becomes relatively flat with coarse Ha/imeda-dominated sediment and occasional small reefs. On the eastern side, which abuts the side of the deep channel, the bottom slopes away farther to about 40 m. Below the depth of P. rus dominance, the coral cover is low. The bottom is largely rocky substrate with shelves on which considerable quantities of sediment are retained. Hillis-Colinvaux (1980) reported that Pole Pinnacle "pos- sessed the same high Halimeda species richness" encoun- tered in some shallow water interisland channels. She felt the Halimeda species populations of the sides of all pinna- cles "may well be principal suppliers of carbonate to the reef floor." In light of recent information on "Halimeda meadows" and the occurrence of Halimeda in the deep lagoon, pinnacle reefs may be less important as carbonate producers than previously suspected, but they are still sig- nificant. Both small and large pinnacles are definitely Halimeda spp. sediment source points; their sloping sides and shallow depths producing a potentially radial dispersal of Halimeda plates from shallower depths to the deep lagoon. A well-known example of a "larger" pinnacle reef is South Medren Pinnacle, located 1.7 km west of the south end of Medren (Fig. 4). It is about 100 m in diameter, roughly circular, and slopes off at about a 30° angle to the lagoon floor at 35 to 40 m. Its upper surface is rugged, with coral ridges and heads interspersed with deeper rub- ble areas. Coral coverage is not as high as Pole Pinnacle but seems average (10 to 30%) for most pinnacle reefs. Coral distribution on the tops and flanks of pinnacles, particularly larger ones, seems somewhat patchy (Fig. 7). Definite sediment downfall areas exist on large pinnacles which restrict corals. Medren Pinnacle has several on its southern face, and near the base of the pinnacle at 35 to 40 m only isolated areas of reef exist. These small patch reefs are generally of low relief, somewhat rounded with abundant macroalgae populations. Here the large blue tubular to vasiform sponge, Cribochalina olemda, is often common. Gilmartin (1966) found the green alga, Tydemania expeditionis, along with species of Caulerpa, Halimeda, and Dictiiota to form the bulk of algal biomass on the deep lagoon coral patches at depths greater than 40 m. Previ- ous dredging work on 7. expeditionis had indicated it to be uncommon, but Gilmartin (1966) found it to be first or second in abundance among algae on deep lagoon coral patches, equal to or exceeded only by Halimeda at 51 to 62 m depth. Ship Channel #1 Pinnacle (not shown in Fig. 6), located some 6 km west of Ananij Island, is unusual. It is a fairly small pinnacle, about 100 m in diameter, rising within about 3 m of the surface with the lagoon about 40 m deep around it. The eastern end of its top is dom- inated by Porites rus, similar to that found at Pole and Tunnel Pinnacles, while its western end has almost exclusively table Acropora corals and appears to have been devastated by a storm several years ago. SUBTIDAL ENVIRONMENTS AND ECOLOGY 101 N *^ * 2. '■ iKjj4 * ^' Medren Enewetak Fig. 6 Locations of lagoon pinnacle reefs in the southeastern portion of Enewetak Atoll. Many of these pinnacles have been important collection localities and are not named on any other published charts. Names used are as follows: 1. Tunnel Pinnacle. 2. unnamed. 3. unnamed. 4. Cucumber Patch. 5. Dead Pinnacle, 6. Pole Pinnacle, 7. unnamed, 8. unnamed. 9. unnamed, 10. unnamed. 11. unnamed. 12. Medren pseudopinnacle. 13. south Medren Pinnacle. 14. Reefer 8, 15. Sand Island *1, 16. Sand Island *2, 17. north Enewetak Pinnacle, 18. Marine Pier Pinnacle, 19. unnamed, 20. unnamed, 21. Harry's Patch, 22. Gemini Pinnacle, 23. Power Plant Pinnacle. 24. Friendly Fish (Bubblebut), 25. Mini Power Plant Pinnacle, 26. unnamed, 27. Garbage Pier Pinnacle. Asparagopsis taxiformis is perhaps, after Halimeda spp., the most common algae on pinnacle reefs {Fig. 7). Its upright thalli protrude from most rocky areas, often in dense stands. Schleck (MPRL, 1978) found that A. taxi- formis grew in a band about 1 m wide and several hun- dred meters in length along leeward island lagoon shores. In deeper water, Schleck reported it formed an abundant. but scattered, community with a vertical distribution to at least 20 to 30 m. The Lagoon Margin The lagoon between 30 m depth and islands or the reef flat, the "lagoon margin," the shore of is an area of 102 COLIN Views of Lagoon pinnacle reefs. Upper left and lower left: Coral development (Pontes rus) on the western side of "Tunnel Pinna- cle" (Fig. 6) with extensive development of the plate-like growth form of this cora! from about 5 to 18 m depth. Upper right: Typical view of lagoon pinnacle (Tunnel Pinnacle) at about 25 m depth with the coral Pauona cactus and the sponge Cribochalina olemda visible. Much of the substratum is devoid of cora) and has an algal community growing on the rock surfaces. Lower right: The algae Asparagopsis taxiformis which is abundant on most lagoon pinnacle and margin patch reefs. great transition. The width of the lagoon nriargin varies considerably from only a few hundred meters at the south end of Enewetak Island to over 1 km from Lojwa north to Engebi and Boken. The deeper portions are similar to the deep lagoon, and because of their accessibility to scuba divers, are an excellent area for studies relevant to the deep lagoon. From about 6 m to 15 to 20 m depth, the bottom has areas of relatively steep sediment slopes, often at the angle of repose; abundant patch reefs, often with relatively high vertical relief, high coral diversity, and abun- dant fish populations. The windward lagoon margin is strongly influenced by the reef flat. Areas of high water transport across the reef flat ("rips"), found at the ends of islands and also along interisland reef flats, affect the distribution of sediments and patch reefs on the lagoon margin. In the lee of the large islands (Enewetak, Medren, Runit) patch reefs are somewhat "dead," with relatively low coverage of corals. SUBTIDAL ENVIRONMENTS AND ECOLOGY 103 A different situation exists on the lagoon margin on the southwestern, western, and northwestern sides (leeward). Because of exposure to prevailing winds across the fetch of the lagoon, these areas often possess an almost barrier- reef type structure with small patch reefs inside it. The sediment bottom often slopes upward steeply near this structure. This is discussed subsequently. Hiatt and Strasburg (1960), in their classic study of reef fish feeding ecology, presented a brief summary of Enewetak reefs. They reported that in the lagoon "in pro- tected areas there is a discontinuous series of irregular patch reefs which extend from nearshore to the outer reef slope leading to the deeper parts of the lagoon." On the western side of the lagoon, "the lagoon reefs are better developed and frequently are continuous, because they receive fairly strong waves engendered by the prevailing winds" across the lagoon. In some respects, they come to resemble reefs of the windward shore. Hiatt and Strasburg (1960) provide drawings of typical reef environments (tidal pools, seaward reef flat, spur and groove surf zone, patch reefs and coral heads, mid-water) with the characteristic fishes found there. The patch reefs of the windward lagoon margin have particularly well-developed coral communities where the water flow across the reef is unimpeded by islands. The vertical relief of the reef generally increases with size, but in many cases small reefs have a relief about one-half their diameter, up to a maximum of about 6 m relief. Table Acropora sp. corals are abundant on these patch reefs, whereas other corals grow well on the sides of the patch reefs and even under overhangs because of the reflection of light from the white bottom. Relatively few soft corals occur in such areas. Sand areas in between the lagoon rim patch reefs are areas of high grazing pressure by surgeonfishes and parrot fishes. Burrowing activity in the sediments is also high, mainly through the activities of a variety of fishes. An important factor determining the distribution of windward lagoon margin patch reefs is the effect of lagoon- ward sediment and rubble movement from the reef flat. Between Enewetak and Medren such patches are abun- dant, but they are best developed in areas protected from sediment "overwash." Leeward of Bokandretok is an area of numerous patch reefs, whereas north and south of this the island rips have covered the area with sediment where the reefs occur. Farther north along the reef, areas of sedi- ment overwash have at best reduced numbers of patch reefs. In areas protected by structures diverting the cross- reef flow of sediment, patch reefs are better developed, coming close in behind the reef flat. Nolan (1975) used a large series of patch reefs in the lee of "Isaac's Island," a small rock and sand spit, for his fish community studies. Nolan (1975) described some patch reefs between Medren and Enewetak Islands where he analyzed and manipulated reef fish populations on these and artificial reefs. He felt coral development was particularly luxurious on the patch reefs on the lagoon side of Isaac's Island. Nolan (1975) pointed out that many of the patch reefs to leeward of Enewetak and Medren Island were predom- inantly dead coral. He provided a detailed map locating his study reefs and chose reefs of about 3 X 3 X 3 m in size, which were abundant, in depths of 5 to 7 m. He noted that the reefs in the lee of Isaac's Island were pro- tected from the full brunt of the cross-reef currents but that an eddy pattern existed on the leeward side of this small outcropping which provided abundant water circula- tion. Nolan's (1975) study reefs were predominated by mas- sive "table" Acropora cythera, but during his study in 1972, heavy surge from the leeward side of the atoll dislodged many of these corals on his study reefs. Sand in this area was also removed and deposited in shallow water creating a 3 m high sand bar continuous from Medren to Enewetak. This sand ridge was destroyed and moved into the lagoon with the resumption of normal trade wind weather and sea swell. Similar destruction of A c\^thera on patch reefs was observed during southwesterly to westerly storms in March 1981 and July 1982. The tables of A. cythera were bro- ken loose at their bases and moved. Many specimens ended up on island beaches with the corallum nearly intact, testament to the strength of this form. North from Japtan to Ananij, no significant lagoon mar- gin patch reefs exist between islands. The bottom slopes relatively steeply into the lagoon, and the reef from ocean to lagoon is narrow. The zonation across the reef is dis- tinct (Fig. 10) and is described subsequently. Chinimi, the only island interrupting this 4 km stretch of open wind- ward reef, has the lagoon margin protected from reef flat "outwash," and patch reefs are well developed in the lee of the island. The change in zonation of the lagoon margin is really visible north from Japtan. Island rips occur north and south of Chinimi and lagoonward depth contours veer close to Chinimi 's shore in its lee. This cusping of the atoll rim behind islands is seen in other areas of the windward side. The area on the northern lagoon margin of Chinimi has one of the best developed reefs along the shore of any windward island, with lovely microatolls, although less than 100 m north the reef seems limited by the island rip and sediment outwash. Ananij similarly has a large number of lagoon margin patch reefs in its lee and has the most developed island rip system of any island at Enewetak. Between it and Runit, 8 km farther north, cross reef zonation is similar to that south of Ananij, but more islands are found on the reef. The island cusping effect, however, is evident with many patch reefs in their lee. A good example of a well-developed lagoon margin patch reef is "Choptop Reef," located just north of "Isaac's Island" between Enewetak and Medren (Fig. 8). It is large for a lagoon margin patch reef, but smaller reefs adjacent to it are similar and provide easy comparison. Choptop has high coral cover and diversity and high fish popula- tions (Fig. 8). It is located on the margin of a reef flat rip, and although not in the strongest portion of the current coming off the reef flat, it is in a well-flushed area. An 104 COLIN Fig. 8 A lagoon margin patch reef. "Choptop Reef," from the air. The main reef (center of photograph) is surrounded by smzdier "Satellite Reefs," some of which are Pontes cylindrica colonies probably broken from the main reef by storm waves. The reef flat is seen in the upper left with a sediment/rubble bar, produced by a cross reef "rip" seen in the upper center. Water depth around Choptop Reef is about 6 m. Upper right: Typical view of coral development on a lagoon margin patch reef (Choptop ReeO with the sediment floor surrounding the reef visible in the background. Lower left: View of upper surface of a lagoon margin patch reef (Choptop Reef) with abundant coral and fishes visible. Depth on the top of the reef is approximately 2 m. Lower right: "Satellite Reer' located about 15 m away from the main portion of Choptop Reef. This reef is simply a smaller version of Choptop with a vertical relief of about 4 m. aerial photograph of the reef is shown in Fig. 8, with the rubble bar and outwash area of the reef flat rip clearly visi- ble. There are several smaller "satellite" reefs close to Choptop which may have resulted from storm fragmenta- tion of the larger reef (Fig. 9). The sediment around lagoon margin patch reefs, like Choptop, is coarse. Cal- careous macroalgae, such as Halimeda spp., occur sporadically on the lagoon margin (Fig. 9), not in large beds as is found in the deeper lagoon. Coral heads on the upper surface of lagoon margin patch reefs often rise to near the surface, but at Enewetak, patch reefs are not planar at about mean to low water levels. At Ujilang Atoll, 200 km southwest, lagoon patch reefs were planar on top, reaching low water level, because of growth of coralline algae. Enewetak patch reefs lack abundant coralline algae on the upper surfaces which may account for these differences. Encrusting corallines are abundant within interstices of Enewetak patch reefs, but the difference, compared to Ujilang, in the amount of exposed corallines is striking. Where the internal structure of patch reefs is exposed, such as in caves or recent fractures, it appears to be com- posed of accumulations of coral skeletons that are poorly cemented internally. Dead branches of coral plates have interstices where small sclerosponges are common. Smith (MPRL, 1972) reported that an explosive blast on a lagoon pinnacle west of Jedrol "exposed unconsolidated to poorly consolidated coral material more or less in growth posi- tion." Sclerosponges, one of the prominent inhabitants of the unlighted holes in the reef, were abundant. The sediment in these lagoon rim areas is not neces- sarily stable. At some coral patches, the sediment is scoured away at the base of the patch reef. Likewise, in SUBTIDAL ENVIRONMENTS AND ECOLOGY 105 Fig. 9 Environments of lagoon margin patch reefs. Upper left: Coarse carbonate sand bottom with Hallmeda spp. and other macroalgae. depth 6 m, near Choptop Reef. Upper right: Sediment/rubble bar near Choptop Reef produced by cross reef "rip." Lower left: "Satellite Reef near Choptop Reef, comprised of a single colony of Porites cylindrica, probably torn from the main reef by storms. Several other satellite reefs are visible in the background. Depth on the bottom is 6 m. Lower right: Small patch reefs on the lagoon margin. A large cable from the atomic testing period is draped over a small patch reef (indicating an age of at least 20 to 30 years) with a colony of Porites edouxi;i which has grown on the cable, depth 5 m. some areas sand can be piled against the reef-killing corals or other sessile invertebrates. Coral colonies with half their surface buried and dead and the upper half healthy can be found at the point of reef-sediment contact on some patch reefs. Alteration of normal tradewind sea conditions can radically alter shallow water sediment distributions. Beaches grow or recede, islands change, and shallow sand bars on the lagoon margin appear or vanish with changes produced by passage of cyclonic storms (Nolan, 1975). It is not necessary for storms to pass close to Enewetak because the swell produced by a distant storm can accom- plish the listed changes without high winds. Lagoon Margin Zonation The area immediately lagoonward of the reef flat is quite variable and of considerable biological interest. Vari- ous authors have described this zone, usually in combina- tion with a description of a cross-reef flat transect. Odum and Odum (1955) described the zonation of the interisland reef about 400 m north of Japtan Island. In many respects this is typical of windward interisland reefs of the central and southeastern portions of Enewetak. They described six zones from ocean reef to lagoon (a dis- tance of about 450 m). These were (1) a buttress zone, (2) the algal-ridge, (3) an encrusting zone, (4) a zone of small coral heads, (5) a zone of small patch reefs "larger heads," and (6) a sand and shingle zone. Typical views of the bottom on the Odum and Odum (1955) transect are shown in Figs. 10 and 11. They make the point that the interisland reefs generally had more "vigorous" communi- ties as opposed to reefs seaward of islands ("island reefs") where living corals were limited to the outmost portions of the reef. They believed this was due largely to different 106 COLIN Fig. 10 Upper: Aerial view of the windward reef in the area of the Odum and Odum (1955) transect north of Japtan Island. The ocean is to the right and the lagoon to the left. The reef flat and associated reefs lie in the middle of the photograph. The photo- graph was taken while flying over the island of Japtan: Chinimi Island is visible with Ananij Island behind It. A normid surf is breaking on the windward reef with the lagoon margin very calm. Lower left: View of the zone of small coral heads on the Odum and Odum transect, depth approximately 1 m. Lower right: Junction of large coral head zone with the sand shingle zone of the Odum and Odum transect. water circulation patterns. Johannes and Gerber (1974) illustrated a simplified cross section of reef near the tran- sect of Odum and Odum (1955). In the Odum and Odum (1955) study area, the bottom slopes gradually lagoonward from the encrusting zone. Indi- vidual coral colonies grow upward to a level limited by low water. In some corals the central portion of the colonies are dead with the sides continuing to thrive, producing structures known as "microatolls" (Fig. 12). These have been examined further on Enewetak reefs by Highsmith (1979) and will be commented on later. Often a distinct lagoonward edge to the reef flat pavement exists, and in many places, water flowing across the reef flat has eroded away and undercut the sediment beneath this lagoonward edge (Fig. 12). This has caused the reef flat pavement to collapse or buckle in places. This is most evident in areas where reef flat rips pass the edge of the pavement. The swift currents combined with the effects of dropping off the pavement have scoured deep potholes (as deep as 4 to 5 m) down into the sediments. The pavement is usually undercut on these edges. The shallow reefs of the northern lagoon are p>oorly known. From Engebi west to Bokoluo, the reef Is broad, as much as 1 to 1.5 km across, unlike southern reefs. Its zonatlon can be seen In aerial photographs but has not been investigated In detail. There Is a reef flat about 100 m wide, then a broad (to 1 km) shallow area with coral heads. This coral head area on the west side of Engebi was examined. There were large microatolls of Pontes lutea and acroporld corals on a level sandy bottom. To the west of Bokoluo lies the open expanse of the northwest reef tract. It runs fairly straight to the northwest corner of the atoll at the West Spit. The gentle arc of the northwest reef is about 1.5 to 1.7 km across from the ocean to the deepening lagoon. From aerial photographs there appear to be four major zones: (1) a reef flat, (2) a coral head zone, (3) a clustered coral head zone, and (4) a patch reef zone. The reef flat Is estimated to be about 150 m across, merging with a deeper coral head zone toward the lagoon. The coral head zone appears about 800 m across and Is complex, with three visible com- ponents to It. The middle one-third of the coral head zone appears deepest, whereas the lagoonward one-third appears shallow. The density of coral heads in this area Is high. Density data from photographs Indicate there are at least 500,000 coral heads in this "coral head zone" between Bokoluo and the West Spit. There is scarcely any open sand of more than a few tens of meters between any SUBTIDAL ENVIRONMENTS AND ECOLOGY 107 Fig. 11 Views of tiic sand-shingle zone of tfie Odum and Odum (1955) transect at Enewetak Atoll. Upper left: Coarse carbonate rubble and sand immediately behind the large coral head zone (depth 1.5 m). Upper right: General view of rubble area behind the large coral zone. Lower left: Carbonate sand farther lagoonward from the large coral head zone, depth 3 m. Lower right: Break in slope of sand-shingle zone where the slope increases considerably (to the right) toward the deep lagoon. of them. From aerial photographs it appears many of the coral heads are arranged in a serial fashion across the reef with large numbers of them resembling striations across the bottom. The clustered coral head zone is about 600 m across and has a lower density of coral heads than the previous zone. Those present are grouped together somewhat. Finally there is a zone of large patch reefs about 400 m wide. These patch reefs appear comparable in size to the larger patch reefs of the windward lagoon margin. Channels Between Northern Islands The channels between the closely spaced northern islands are of special beauty and biological interest. They are not true passes from ocean to lagoon because they draw their flow from the shallow reef flats to seaward but channelize the flow of water off the reef flat between islands. Viewed from the air, their bottom features show strong orientation to the current which funnels between the islands from ocean to lagoon, with reefs often elongated with the current and sediment washed out between patch reefs. These "passes" have a reef flat on their seaward end, but the cross-reef flat flow from an area of reef front several times broader than the channel is funneled into each one. The channels are often deep, but where current flow slows on their lagoonward end, they usually have a shallow, delta-like bottom. A good example of a northern island channel is that between Lojwa and Aomen. At very low tides water flow across the reef flat is completely eliminated, with no current in the channel. At high tides with strong waves pumping, the current is swift, sufficient to deeply churn sediment from around patch reefs in the channel. The gaps between reefs have the sediment scoured away, appearing darker blue when viewed from above, whereas areas on the sheltered, downcurrent side of the patch reef have 108 COLIN Fig. 12 Upper left: MIcroatolls (Porites lobata) at the north end of Chinimi Island, Enewetak. At low tide the water Is essentially at the upper level of the microatolls. Secondary growth is also occurring In the central area of the top of the mlcroatoUs. Upper right: Aerial view of lagoonward edge of the reef flat showing erosion at the end of the reef pavement caused by water flowing across the reef flat. Lower left: Typical views of patch reefs in the Lojwa-Aomon interisland channel. Extensive sculpturing of the sediment bottom is caused by currents which course through this channel at high tide. Lower right: Area of the Lojwa-Aomon channel with sand built up behind (down current side of) a large patch reef. white sand built up. The width of these "tails" of sediment decreases downcurrent of the reef. The upcurrent sides of the patch reefs have the sand washed away to depths equal to those on the sides of the reef. Corals and benthic invertebrates are usually well developed on the upcurrent end and sides of reefs. The deepest portions of the channel are 6 to 7 m, and some patch reefs are emergent at low tide (Fig. 12). The reefs in this channel have changed little in the last 32 years based on aerial photographs taken in 1949 and 1981. The major patch reefs are identifiable, but some of the lagoonward patch reefs seem to have been somewhat buried by sediment. Other interisland channels are similar. Rock surfaces are heavily grazed by herbivorous fishes. Small caves and overhangs off the patch reefs are lined with encrusting coralline algae. These patch reefs are one of the few places within the lagoon where branching coralline algae are found. Sediments are coarse, with predominantly large foram tests, coral, and Halimeda bits. The reefs of the channel between Lojwa and Alembel seem to have been devastated by a storm during the last decade. Very little live coral and few benthic invertebrates arc on them. Allen (1972) used this channel as a primary study site for his anemonefish work. One patch reef in the channel had more than 75 clusters of 10 to 30 individuals of Ph^isobrachia douglasi, with larger numbers of Amphiprion melanopus. in an area of only 700 m^. In the summer of 1981, this area was re-examined for anemones and Amphiprion, no anemones or anemonefishes of any type were found. In channels farther north, corals and other inver- tebrates seem healthy. Some of the channels were noted for their abundance of large tridacnid clams, but many of these clams have been eliminated since the repatriation of the Enewetak people. SUBTIDAL ENVIRONMENTS AND ECOLOGY 109 Passes There are three passes from ocean to lagoon with suffi- cient water depth for boats to regularly traverse them. They are the "deep passage" (east) between Medren and Japtan, the "wide passage" (south) between Enewetak and Igurin, and the "southwest passage" between Kidrenen (south) and Biken. Various details of these passes have been discussed in Chapter 3, this volume. The biological communities of the deep channel and its margins have not been well described. Hobson and Chess (1978) discussed the patch reefs and plankton communities in the area between Japtan and Jedrol Islands which are affected by currents coursing through the deep channel, but their study site was not in the deep channel proper. The northern side of the deep channel slop>es steeply from depths of only a few meters. To the east of Jedrol there is actually a "barrier" reef awash at low water which is constantly exposed to oceanic swells entering the lagoon through the deep channel. The northern slope of the deep channel to depths of 30 to 40 m is a nearly 45° angle rocky slope with abundant corals and reef-associated inver- tebrates. At depths of 25 to 40 m, the bottom levels and the central portions of the channel are probably relatively flat. There is a downslope sediment transport along this face, and below 30 m where the bottom begins to level, sediments also begin to dominate the bottom compared to exposed rock outcrops. The easternmost extension of the shallow wedge where the channel splits is distinct, the "cutting edge" being only a few meters wide and descending at about a 45^ angle from 6 m to depths below 30 m. The coral communities of the shallow reef and slope are rich. The fish communities of the north side of the channel are diverse and abundant with zooplanktivores more dominant than in other areas. The south side of the deep channel is different from the north, with the bottom sloping gradually as a sediment slope with little or no exposed rocky substratum. A shelf between 30 and 36 m in depth extends a kilometer or more northwest from Medren into the lagoon. Little is known about the area of the wide channel. Aerial photographs show large patch reefs on a sandy bottom scattered across the entire 9.3 km width. The crest and outer slope of the sill was examined about 1.6 km west of Enewetak and had large, rocky patch reefs, not unlike large lagoon margin patch reefs at 18 to 20 m depth (Fig. 13). The patch reefs had relatively little live coral but had abundant Halimeda spp. and Asparagopsis taxiformis. The most common corals were Pocilhpora spp. The sediment was coarse, dominated by Halimeda. with small ripples at 22 m depth. There were small rocks between the much larger reefs but little grew on them. To seaward, the sediment bottom sloped perceptively. At 30 m, it was nearly all sediments with only a few rock patches and sloped at an angle of about 15° (Fig. 13). Below that depth, the slope increased to about 20° at 40 m and more with increasing depth. The southeast passage consists of sandy channels between elevated fingers of reef for 6.5 km southeast of Biken. Atkinson et al. (1981) estimated the cross-sectional area of the southwest passage as only 26% of the deep passage and 6% of the wide passage with no net inflow or outflow. The reef fingers have well-developed coral com- munities which do not differ greatly from the interisland Fig. 13 Views of the bottom, wiae ^soutn; passage, Enewetak Atoll. Upper: Rubble substratum at about 20 m depth looking downslope. Middle: Juncture of rubble and sand substratum at 30 m depth, looking downslope. Lower Sand slope substratum with isolated coral l>oulders at 40 m depth. There is considerable evidence of downslope transport of sedi- ment in this view. 110 COLIN patch reefs fin the windward side. The sand channels shoal gradually from the lagoon to their shallowest fxsint, then again gradually deepen to seaward. Near the precipitous reef edge to seaward, the channels quickly steepen, then plunge down the near vertical slope. Sediment is trans- ported over the drop-off here with heavy scouring of the reef face below the sand chutes. Algal Ridge Before considering the true seaward reefs, it is worthwhile to mention the zone marginal to the reef flat. This is the "algal ridge" which is truly intertidal but has extreme relevance to subtidal seaward areas. The seaward reef on the windward side of Enewetak is mostly devoid of live coralline algal ridges. Live algal ridge (often termed "Lithothamnion ridge" by earlier authors) occurs only along one section of windward reef about 200 m in length at Ananij Island. This section is readily distinguished by its pink coloration, produced by the abun- dance of Porolithon species, as compared to the dull sur- face of the algal ridge dominated by macroalgae. Three species of Porolithon, as identified by Lee (1967), have been found on the Ananij algal ridge. Large portions of the surfaces of the spurs are covered with crus- tose corallines, probably Porolithon onkodes. Distinct colo- nies of Porolithon craspedium, often with blunt fingers forming a lobate mass, occur scattered on the upper sur- face of the spur. Porolithon gardineri seems the least com- mon species, although its colonies are often irregular masses 20 cm or more across. It appears limited to the sides of the spurs, not being found on the upper surface among P craspedium^ Within the sponge-like structure of the spurs at Ananij, virtually all visible internal surfaces are covered by coralline algae, but the species involved are not known. Inshore from the live algal ridge at Ananij is a slight depression of the reef flat where colonies of Acropora sp. flourish. Small patch reefs occur on the hard pavement here which has water on it even at low tides. The Acropora sp. colonies are emergent at low tides. The small coralline algae Neogoniolithon rutescens is occasionally found among these patch reefs but not on the more exposed spur and groove areas. Seaward Reefs Smith and Harrison (1977) described the windward reef slope off Chinimi Island, and since their study other areas have been examined. The spurs are relatively flat on top and occasionally have undercut, overhanging edges (Fig. 14). Algae and invertebrates are abundant on the sides of these spurs. Sea urchins have eroded elongate grooves in the rocks on the sides of the spurs which afford protection from wave action and predatory fishes (Fig. 15). The bases of the grooves are floored with boulders and cobbles, precluding any significant benthic invertebrate populations (Fig. 14). The walls of the grooves, however. have on them small corals and invertebrates adapted to withstand the wave surge. On the upper surfaces of the spurs, small corals grow with an abundant film of algae on rock surfaces (Fig. 14). Smith (MPRL, 1972) dissected a spur and groove sys- tem north of Japtan using explosives. "The spur proved to be dense, well-cemented coral rubbles covered by a veneer of live encrusting coralline algae." He felt that, except for relatively minor growth by the coralline algae, the spur and groove systems are erosional features. On windward reefs the spur and groove zone and the area immediately seaward of it are areas of very high fish abundance (Fig. 15). Herbivorous parrot fishes and sur- geonfishes feed in this productive area and range on to the algal ridge and reef flat from there. At low tide these shal- lower areas are dry, requiring their exploiting fish popula- tions to move elsewhere. The spur and groove zone seaward of the areas of Porolithon algal ridge at Ananij is different from other areas examined where the ridge is "dead." The cover of benthic invertebrates appears higher there. This is the only area on the windward shore where the club-spined urchin Heterocentrodus trigonarius is known to be abundant, both in holes on the sides of the spurs and on the algal ridge. A form of branched Acropora sp. coral with other corals and Halimeda sp. algae with distinct laminations occurs there. This form of Acropora has not been seen elsewhere (Fig. 15). Off the north end of Enewetak Island, the sides of the i spurs are lined with grazed macroalgae and occasional patches of coralline algae. The rock-boring urchin, Echinometra methaei. is abundant in grooves in the sides of the spurs. In small caves and on overhangs a wide variety of benthic invertebrates occurs. On the sides of the spurs' upper surface are small head corals and soft corals. There is less coral on the tops of the spurs, and the area is more dominated by macroalgae. | At the seaward end of the spur, colonies of stony corals, Heliopora caerulea, and soft corals are common. These are larger than those of the top or sides of the spur. Some sizeable encrusting sponges may also occur in this area. Several herbivorous fishes are characteristic of this j spur and groove zone. The surgeonfishes, Acanthurus achilles. A. guttatus, A. thostegus, and especially A. lin- eatus are generally found in any abundance only in this area on the windward shore. One small damselfish, Plectrogliiphidodon phoenixensis, is common on the wind- ward shore and occurs only in the spur and groove area. Seaward of the spur and groove, the rocky bottom lev- els somewhat with only a slight seaward slope (Fig. 14). The bottom often has minor undulations of its surface, occasionally with small shallow grooves oriented perpendic- ular to the reef front, but generally it has few distinguish- ing features. The irrejular pits and grooves of rock-boring sea urchins, Echinometra mathaei. and lesser numbers of some diademnid urchins (Fig. 16) are often abundant. A few small- to medium-sized corals occasionally occur on i this "barren" zone (Fig. 15). Viewed from the air, this zone - SUBTIDAL ENVIRONMENTS AND ECOLOGY 111 Fig. 14 Typical views of the spur and groove zone off Enewetal( Island. Upper left: Shallow groove with large coral boulders In its center which are set in motion during periods of high waves. These effectively keep the grooves free of sessile benthic macroor- ganisms. Upper right: A variety of herbivorous fishes at seaward end of a spur. Lower left: Spur seaward of the Enewetak Island reef flat with breakers rolling over it. Lower right: Seaward end of a spur with a breaker forming at its prow. The shallower water depth on the spurs causes waves to break there sooner than over the grooves. appears to be a uniform light color. The rock shelf width varies around the atoll. Off Enewetak Island it is relatively wide, about 200 to 300 m, but farther north it becomes narrower, probably less than 100 m wide. The rock surface of the shelf often has evidence of extensive boring by clionid sponges. Large areas of substrate may have the tiny, dark oscula visible, but these are not apparent on superficial examination (Fig. 16). Smith and Harrison (1977) have described a windward reef slope from off Jinimi Island in connection with estimates of calcium carbonate production there. The reef crest had essentially no corals. Moving seaward from the reef crest and spur and groove zone, the bottom slopes gradually from 4 to 5 m depth to about 8 m and is essen- tially a rocky pavement with minor surface undulations. Smith and Harrison (1977) estimated only 10% coral cov- erage in their study area at 7 m depth. Seaward, the amount of coral cover increased with depth, although the slope may increase only slightly with 15, 20, and 25% at 11, 15, and 21 m depth, respectively. At 50 m, coral cov- erage was virtually zero. Smith and Harrison (1977) found that the vasiform Acropora cvthera was the most conspicu- ous coral in their study area, with its nearly flat upper sur- face well adapted for capturing sunlight. They performed coral and coralline algae incubations using clear acrylic domes, where possible, at depths to 21 m. Steadily decreasing rates of calcification with increasing depth were found. Overall they believed the seaward slope of wind- ward reefs at Enewetak (the "mare incognition" of Ladd, 1961) has only a small role in the CaCOa mass balance of the atoll. Large numbers of vasiform Acropora cilthera colonies, up to 2 m in diameter, were found by Smith and Harrison (1977) at 15 to 25 m at their study area (Fig. lA of that paper). Colonies had a maximum of 13 growth bands (annual), and they considered that the major typhoon in late 1962 (their observations were in late 1976) may have devastated Acropora corals in that area. Smith and 112 COLIN Fig. 15 Upper leh: Larger shoal of Acanthurus triostegus in the spur and groove zone, Ananij Island, depth 4 m. Upper right Unusual growth form of Acropora sp. found seaward of the area of live edgal ridge, Ananij Island, Enewetak Atoll. Lower left: Grooves eroded in the side of spurs by sea urchin Echinometra mathaei, windward reefs, Enewetak Atoll. Lower right: Isolated coral head located to seaward of the spur and groove zone, windward shore of EnewetiU( Island, depth 7 m. Harrison's (1977) study area was disrupted by a severe typhoon in January 1979 (Alice) in which all the large Acropora colonics at 15 to 25 m were reduced to rubble (Fig. 16), confirming their suspicion that typhoon-strength storms are capable of such disruption to depths near 20 to 25 m. The outer slope or "drof>-off" begins at depths of 18 to 23 m as a distinct change, from a gentle slope of a few degrees to an angle of approximately 30° to 45°. This slope rapidly increases with depth (Fig. 17). The deep reefs of the windward side have been severely damaged by storms so that there is relatively little live coral and tremendous amounts of rubble at 15 to 30 m depth (Fig. 16). Along Enewetak Island to Medren, there is gen- erally a sandy zone at 30 to 40 m which appears as an irregular light band from the air. Below this depth sand channels alternating with reef can be seen on the outer slope when viewed from the air; this sandy zone is not apparent from the air on reefs of the islands farther north. Vosburgh (1977) experimentally determined that waves of near 5 m height did not produce sufficient water motion at depths of 9 to 21 m to cause breakage of the skeleton of large, healthy Acropora Ci/thera. He reported that although this species is found at less than 2 m depth in sheltered areas of the lagoon, it occurs commonly on the windward reefs only at depths below 8 to 10 m. Sheltered lagoon colonies were generally larger than those on the windward reef, and depth distribution and colony size are related to wave exposure. Although his estimates of near 5 m waves are based on the highest 1% of waves observed during the windiest portion of the year, he points out that typhoon waves, not considered in his study, "might cause catastrophic breakage over the entire species range on the (windward) terrace." The steady seaward slope of the windward reef gen- erally prevents accumulation of large amounts of sedimen- tary material. At the slope break at about 18 to 20 m depth, some sediment-bottomed channels occur which can SUBTIDAL ENVIRONMENTS AND ECOLOGY 113 '-^u ? \ ■W > •,. v "•c ^. ' . ■+'■ ; ^ ' ■ J^ft" *-■-'. */' %• ", '!_' ^ ■ V , V r 1 :, i f /,--.^ SsAi.;*. , >i' ' ;^' m^y^ut- W. Fig. 16 Upper: Carbonate rock substratum heavily bored by the sponge Cliona sp. on the windward reef, Enewet€U< Island, depth 8 m. The dark oscula of the sponges are visible over much of the substratum, although the tissue of the sponge is located internally beneath the surface of the rock. Middle: The area of the shelf edge break (20 m depth) off Enewetak Island. There is very little live coral in this area with only a single table Acropora visible among large amounts of coral rubble. Lower: Outer slope at 25 m depth. Enewetak Island, coral rubble going down the slope into deeper water. serve to transport sediment into deep)er water. Below the shelf break, larger amounts of sediment are visible on relatively horizontal areas, but the slope limits the amount of build up. The reefs of the leeward side have extremely steep slopes. The distance between the reef crest and the steep slope into deep water changes with location. Along the southwestern islands (Ikurin through Kidrenen) there is a narrow shelf sloping gently from about 3 to 15 to 18 m. This shelf is generally about 100 to 150 m wide and has a well-developed coral community on the rocky shelf. Most of the corals are small, less than 10 years old, implying recent devastation, probably by storm waves. Sand chan- nels occur perpendicular to the reef front which is at the head of reentrants on the reef face. The change to a steep slope occurs at about 15 m where it becomes a 45° to 60° slope to the limit of scuba diving. A typical profile of a southwest island reef is shown in Fig. 17. To the west of Kidrenen, the reef remains unbroken until the southwest passage. The bottom slopes gently, then progressively becomes steeper with virtually no shelf to a near-vertical face at about 10 m depth. The horizontal distance from water a few meters deep to the vertical face is less than 50 m. This extremely steep profile is even more pronounced on the reef north of Biken to the West Spit. Reentrants penetrate the reef face with Halimeda dominated sediments on shelves on a steep slope into the deep water (Fig. 17). The leeward reef crest near the island of Ikuren has a healthy cover of coralline algae on its upper surface, even though on the leeward side of the atoll, small to moderate surf usually occurs, which is produced by the long, low swell from the west. Large numbers of herbivorous fishes occur here, essentially the same species as are found on the windward spur and groove areas. The two areas are similar; but near the southwest islands the grooves, strength of surf, and various invertebrates are lesser developed. Seaward of the reef flat are often small high relief rocky structures with flattened tops and abundant coral (Fig. 18). Species of Acropora, Pocillopora, and Heliopora axe common on the edges of the coralline flat. The cidaroid sea urchin, Heterocentrotus trigor)arius, is found deep in small caves and crevices of the outlying rock structures among coralline-covered fossil coral branches. Around and to seaward of these structures is often a bot- tom at 5 m depth composed of large coral boulders and shingle. Much of the hard substrate in this area not covered by hard or soft corals has coralline algae growing on it. These algae often have large numbers of grazing marks almost cerlainly from parrot fish (Fig. 18). The alga, Asparagopsts taxiformis, is extremely abundant; its upright tufts in evidence on nearly all rocky surfaces (Figs. 19 and 20). A rock substrate begins within 20 to 40 m of the reef flat with occasional large vertical knobs of rock covered with hard and soft corals. Urchin grooves are evident in the rock, but diademnid urchins were seen much more often in them than Echinometra mathaei. 114 COLIN 30 60 120 DISTANCE, m 180 240 300 360 420 5X Vertical Exaggeration Fig. 17 Typical slope profiles of Enewetak Atoll seaward reefs. The profiles, which are vertically exaggerated, are from a wind- ward reef off Enewetak Island (upper), a leeward reef off Ikuren (southwest islands) (middle), and a leeward reef north of Biken (lower). Dotted lines represent the bottom in areas of sand channel reentrants of the reef face. Waves shown on the surface reflect the normal wave conditions on these different areas. The rocky shelf slopes gradually seaward, and at about 8 to 10 m depth sand channels begin to appear on its sur- face. There is considerable relief between the reef fingers at about 9 to 10 m and the channels at 12 to 14 m (Figs. 19 and 20). The sides of many of the fingers are nearly vertical and often undercut. These overhanging walls have dense coverage of coralline algae and abundant Haliweda. The sediment in the channels is coarse, derived largely from Halimeda flakes and often has wave ripples on its sur- face from the long period swells. The upper portions of the reef fingers have dense coral on their tops and sides. Coral coverage at 12 m depth on the top of the fingers at the shelf break is 80 to 90% in some areas. A few large head corals occur but most are small to medium acro- porids. They are at most 25 to 40 cm across and probably reflect recruits after storm destruction of most of the previ- ous acroporids (Fig. 19). The bottom slopes away at the shelf break (12 to 15 m) at a 45 to 60° angle. Most of the sand channels continue down the slope as sediment chutes into deep water. These chutes are cut back into the reef face and have sediment down them to the limit of visibility (Fig. 20). Adjacent rock surfaces have abundant corals, the same types of species that occur in shallower water. Live Halimeda is abundant all down the slope to over 60 m. Hillis-Colinvaux (1980) found four species of Halimeda on the seaward reef off Mut at 10 to 15 m depth. She estimated cover of Halimeda on this bottom as about 15% and commented that Halimeda was much more conspicu- ous on the spur reef structure than she would have expected on a reef buttress in Jamaica. Halimeda flakes dominate the sediments of all leeward side oceanic reefs. Below about 20 m depth, sediment builds up on any nearly horizontal surface, particularly near the reentrants which are the primary "down chutes" for sediment. There are many overhangs and small caves formed by coral plates on the leeward reefs. Incredibly delicate large colonies of stylasterine corals grow in their dim recesses. Three species of sclerosponges — Astrosclera willei/ana, Acanthochaetetes welisi, and one unidentified species (Basile et al., 1984) — are found in caves along the reef SUBTIDAL ENVIRONMENTS AND ECOLOGY 115 Fig. 18 Typical views of seaward reefs off the southwestern islands, leeward side of Enewetak. Upper left: Area of heavy graz- ing, probably by parrot fishes, on rock substrata. Upper right: Seaward end of the reef flat with short "grooves" going Into the reef flat. Lower left: Heavily greized substrata, with parrot fish tooth marks, depth 2 m. Lower right: Seaward end of the reef flat with some large coral colonies. face, but they are small and do not produce significant amounts of calcium carbonate. There are many large fan-like gorgonians along the vertical face, in addition to widely scattered colonies of antipatharians (black coral). Large black coral "trees" are rare in these (and all other) areas. On the leeward side of Enewetak Atoll there is an algal ridge-type structure which is not well known. Marsh (1970) reported one area at Igurin to have "a relatively good growth" of coralline algae. The leeward ridge is in many places slightly submerged at low tides, but never as emer- gent as the windward reef flat. The outer slope of Enewetak below scuba diving depths was examined to a depth of 365 m with the research submersible Makali'i during the summer of 1981. Twenty-two dives were made on the seaward face from Biken around the southern end of the atoll to south of Runit (Colin et al., 1986). The seaward reefs of the north- ern half of the atoll were not examined. The depth profiles of five areas on the seaward margin are shown in Fig. 21. The profile of the outer slojie of Enewetak is very steep, an angle of about 60° between 90 and 360 m with the leeward slope being slightly steeper. Emery et al. (1954) and subsequent writers have com- mented on the steep slopes of atolls in the northern Marshalls. Their opinions were based on echo soundings and were confirmed by observations from the submersible Makali'i. To depths of about 300 m the slojje is generally rock with small accumulations of sediment. Every near- horizontal surface has a dusting of sediment, and small ledges have accumulations varying with the area where sediment can rest. There is little significant accumulation of talus to depths of 200 to 300 m from upper areas as the slope remains steep enough to prevent talus accumula- tion. Incised, highly polished vertical grooves occur in the rock face serving for downslope transport of sediment. At depths between 200 and 300 m, large talus begins to occur in the form of broken colonies of coral and reef plate brought down the slope. In some locations a Halimeda sediment-dominated slope began at about 270 to 300 m depth with a slight decrease in slope. Along with this were 116 COLIN Fig. 19 Typical views of seaward reefs off the southwestern islands off Enewetak. Upper left: Rubble area to seaward of the reef flat, depth 5 m. Upper right: Area of isolated coral heads seaward of rubble area. Lower left: Small acroporid corals on the outer reef face, southwestern islands, depth 20 m. These corals are probably less than 10 years old and may represent recruits after major storm damage to the community. Lower right: Outer slope along the southwest islands, depth 20 m. The bottom slopes away at about a 45° angle to great depths. often mounds or ridges of talus and carbonate blocks more than 1 m across. There was some relief on the rock face at 100 m to about 180 m, often with the surface pitted with shallow depressions less than 50 cm across. There were occasional small caves, seldom penetrating more than 1 m into the reef face. Stony corals were observed to grow relatively deep. Below about 60 m only flattened forms were found. Sparse coral communities occurred to at least 90 m depth, with individual colonies occurring to slightly more than 100 m depth. Similarly, attached and living Halimeda colonies were found at more than 120 m (HillisColinvaux, 1986). Green algae were found to almost 150 m and coralline algae to nearly 200 m. Some differences in biological zonation were noted between the windward and leeward slopes. The windward areas have more coral at 60 to 90 m depths, larger popu- lations and diversity of small reef fishes from 60 to 200 m, and generally more benthic invertebrates. In the wide channel area, there seemed to be much down slope transport of sediment, although again the steep slope at 100 to 200 m trapped relatively little sediment on the face. Below about 200 m, huge slopes of Halimeda with seapens growing on them were found (Colin et al., 1986). At the eastern edge of the wide channel this uncon- solidated slope was alternating elevated areas of talus and the sand between "pure" sand slopes. Lagoon Water Column The waters of the lagoon have not received adequate attention. Recent work has examined the circulation of the lagoon (Atkinson et al., 1981), the relationship between reef-produced organic material and lagoon plankton (reviewed in Gerber and Marshall, 1982) and plankton SUBTIDAL ENVIRONMENTS AND ECOLOGY 117 Fig. 20 Reentrants of the reef face, southwest islands, Enewetak Atoll. Upper left: Sand channel in between elevated reef fingers, depth on sand approximately 12 m. Upper right: upper portion of a reentrant on the outer slope at 30 m depth. Lower left and right: Sediment transport via reentrant down the outer reef face, southwestern islands, depth 40 m. composition (Gerber, 1981), but these are only a bare beginning. The open lagoon is generally rough and less than ideal for working in a small boat. Navigation and posi- tioning are difficult because the islands of the atoll rim are so low that in the center of the lagoon little land is visible. Water column productivity within the lagoon has not been well documented. The author has seen, on several occasions, large blooms of phytoplankton in the northern and western lagoon. These were sharply differentiated areas of "brown water" many kilometers in length and, on two occasions, as surface slicks many centimeters thick. The surface slicks occurred under extremely calm condi- tions and were nearly linear masses of tan phytoplankton, tens of meters broad and over 1 km in length. Thickness was not determined but was believed to be at least 30 cm. The blooms may be associated with water of lengthy residence time in the lagoon since they have been observed only from areas where this is typically the case. Dense swarms of zooplankton were often observed in the lagoon by scuba divers, often within a discrete portion of the water column. During the summer, particularly huge numbers of salps and ctenophores were observed many times on the reef. Gerber and Marshall (1982) documented a "bloom" of pteropods and a subsequent population decrease in the central lagoon during a 4-week period. The unidirectional flow of water from windward reefs across the reef flat to the lagoon is significant not only in the physical flushing of the lagoon but as a mechanism by which increased nitrogen, produced by nitrogen fixation on the shallow reef flat (Webb et al., 1975), reaches the lagoon. Webb et al. (1975) felt there were three important routes by which Calothrix Crustacea fixed N2 enters the remaining reef ecosystem. First, fish grazing and the low assimilation efficiency (Chartoch, 1972) of herbivorous fishes makes the fixed nitrogen available. Second, fragmen- tation of Calothrix and lagoon transport makes it available to herbivores and detritivores in the lagoon. Third, 40 to 60% of the nitrogen fixed is released in solution and is available for other organisms. Gerber and Marshall (1974) have shown that detritus flowing off the shallow reefs forms a major component of 118 COLIN Fig. 21 Views of lagoon reef coral colonies. Upper left: Turbinaria sp. Upper right: Large colony of Pontes rus with areas removed, possibly by coral damaging fishes. Lower left: Pontes nts, large colony showing a shift from columnar to plate-like growth with depth and exposure to less light. Lower right: Fan-like growth form of the hydrozoan Millepora sp. on a lagoon margin. ingested material in two abundant lagoon zooplankters. Furthermore, lagoon copepods have also been known to ingest and assimilate such particulate matter (Gerber and Gerber, 1979). Gerber and Marshall (1982) suggest that the occurrence of planktonic organisms in the central lagoon results mostly from production and consumption in the water-column community. They indicate that the reef com- munities are the sources for a large percentage of the car- bon and nitrogen fixed and present in lagoon waters. The phytopiankton community of the lagoon is also important as a food chain base, but the relative importance of each is not well understood. Gerber (1981) documented the diversity and abun- dance of zooplankton at two stations in the lagoon. Ninety six species of copepods and species of chaetognaths, larva- SUBTIDAL ENVIRONMENTS AND ECOLOGY 119 ceans, mysids, euphausiids, amphipods, siphonophores, pteropods, dinoflagellates, medusae, other planktonic Crus- tacea and larval forms were found. One station, near the Enewetak-Medren reef flat had lower abundance, fewer species of typically planktonic organisms, and more mero- planktonic and benthic forms than the mid-lagoon station. There was considerable variation in densities of zoo- plankters among samples taken during the same study periods of a few to several weeks at the mid-lagoon sta- tion. Gerber and Marshall (1982) reported that lagoon con- centrations of copepods, pteropods, and larvaceans were higher during their summer sampling period. Phytoplank- ton biomass in mid-lagoon in summer was also about twice that of the winter. Individual components of the zooplank- ton changed their densities considerably during periods of several weeks during the summer. Copepods and larva- ceans increased 1.5 to 3 times. Pteropods increased 20 times in 4 weeks, then declined rapidly. Coles and Strathmann (1973) collected mucus floes from the water column at Enewetak and other areas and found them to represent substantial quantities of organic matter when compared to particulate organic material in the water. They noted that under calm conditions few mucus floes were seen in the water at Enewetak, but after a storm abundant large floes were seen passing into the lagoon. ASPECTS OF MARINE COMMUNITIES AT ENEWETAK Reef Growth and Destruction Coral reef growth is a balance of factors: the accretion of calcium carbonate by stony corals and other calcifying organisms in addition to the consolidation of these materi- als into a cohesive structure versus the erosive effects of grazing pressure, physical weakening, and destruction of the reef structure. Much work at Enewetak has focused on questions related to the growth and maintenance of reefs. Not all is summarized here but some environmental factors concerning reef growth are. Calcification of corals and other organisms can be affected by environmental conditions, such as light, tem- perature, and water movement over the range of condi- tions under which the organism ca.i survive. Stony corals are also known to "compete," albeit in a relatively slow manner. Methods include overgrowth, reducing the light necessary for calcification and growth of competitors, and by "extracoelentric digestion" in which mesentarial fila- ments are extended to "attack" and kill tissue of other species growing close by. The range of conditions inhabi*3d by a single species or genera of corals is often broad. The genus Pocillopora is illustrative. Stimson (1978) reports that the Pocillopora species at Enewetak occur over a broad range of depths but are most abundant on reef flats and in water <5 m deep with currents. He reported P. verrucosa to reach 15 m depth on pinnacles and windward and leeward reef slopes and to occur in the "small head zone" (Odum and Odum, 1955) north of Japtan. Pocillopora uerrucosa is also common in spur and groove areas of the windward reefs. In eastern Australia the species is found in areas of regular water movement and good illumination, and its growth variations are less diverse than those of P. damicornis (Veron and Pichon, 1976). Pocillopora damicornis occupies potentially a greater range of habitats than any other coral at Enewetak. Veron and Pichon (1976) have figured the wide variation in corallum morphology and documented the broad range of conditions this species inhabits in eastern Australia. Pocillopora eudoxt^i occurs deeper than any other branch- ing coral at Enewetak, to approximately 60 m on the sea- ward slope. Members of Acropora are similar. Some are limited to very shallow water. Stimson (1978) found A. aspera and A. humilis only in water less than 2 m deep. Acropora digi- tifera and A. aspera are sometimes exposed and killed by extreme low tides. Others, such as A. s^ringoides, are re- stricted to water deeper than 5 m. Acropora s\;ringoides is abundant on the flanks of patch reefs and pinnacles near Enewetak and Medren. Other species have broad depth distributions. Acropora hyacinthus and A. nasuta occur from 1 to 20 m depth. Coral growth rates have been examined for a number of species of stony corals at Enewetak. The technique of x radiography of slabbed coral specimens was first applied to Enewetak coral specimens and used to verify the annual nature of the density banding observed (Knutson et al., 1972). Autoradiographic exposures of coral slabs show dis- tinct bands of activity from atomic test series and, there- fore, serve as bench marks in coral growth chronology. Knutson et al. (1972) also presented evidence that the high density bands seen were formed during the rainy season at Enewetak. Buddemeier et al. (1974) examined skeletal growth rates of 15 species of corals, including the same species from various locations at Enewetak. They reported growth rates of generally 4 to 12 mm per year with some exceptions above and below these figures. Not all coral species examined showed variation in growth rates with depth. Porites lutea did show a negative correla- tion between growth rate and increasing depth, with about one-half the rate at 25 to 30 m as was at 4 to 10 m depth. However, colonies of Goniastrea sp. collected in deep water grew as fast as those from shallow water. Although Buddemeier et al. (1974) focused attention on obtaining large symmetric head corals, no specimens examined indicated ages before 1952 and 1953. Whether the nuclear tests of 1952, particularly the "Mike" test, had any effect on this is uncertain. Stimson and Polacheck (MPRL, 1977) reported that growth rates of Acropora and Pocillopora at four different depths from 1 to 15 m on lagoon pinnacles and patch reefs were statistically indistinguishable. Three species of common shallow water Acropora had annual increments in 120 COLIN diameter of the colony of about 5 to 6 cm, whereas Pocillopora in shallow water had an annual growth in diam- eter of about 4 cm. Smith and Harrison (1977) reported table Acropora colonies to increase their diameter 15 cm or more per year once they had reached the stage where they transform from a vasiform to tabulate corallum. Haggerty (1980) found that with increasing water depth both Fauia pallida and F. stelligera had more widely spaced corallites, a slower linear skeletal growth rate and a decrease in the annual skeletal growth rate per square cen- timeter. Fauia pallida had a hemispherical colony form in all environments at 3 to 41 m depth, but deep water populations possessed more septa per corallite than shal- low water. Fauia stelligera changed its colony morphology with depth, from "lobate or hummocky" in shallow water to "columnar with a slight basal skirt" (Haggerty, 1980) in deeper water. Stimson (MPRL, 1973) looked at the interactions via "extracoelentric digestion" between closely adjacent corals of various species at Enewetak. In the hierarchy of Enewetak corals, based on the species which could be successfully "attacked," Astreopora mi/riopthalrna ranked the highest, with Acropora acuminata second, and Pontes lutea third. Pocillopora spp. were lowest, being killed on contact with other species. Stimson (1978) studied the timing of planulation by species of Pocillopora and Acropora at Enewetak. He found Enewetak colonies to produce planulae primarily during the new and first quarters of the moon. He also suggested that planulation by Acropora may be more sea- sonal than Pocillopora because about twice as many colonies planulated during the summer than in the winter. Among pocilloporids, colonies 6 to 8 cm in diameter (15 to 30 cm in volume) were the smallest observed to planu- late and estimated to be 1 to 2 years old. Acropora colonies as small as 50 cm'' planulated, but most were greater than 1000 cm^ in volume. Pocilloporids generally produced more planulae than acroporids at Enewetak. There can also be geographic variation in lunar timing of planulation. The lunar periodicity of planulation in P. dam- icornis is the same in Palau as Enewetak but is reversed from Hawaii (Stimson, 1978). Stimson (1978) felt that shallow-water corals at Enewetak were in a more "disturbed" environment than in deeper water and that species found predominantly there would have high reproductive rates. He has measured annual mortality rates as high as 20% for shallow-water corals. Most of these species produce planulae rather than smaller eggs and may do so to facilitate rapid settlement in the current-swept reef flat areas. The large table Acropora (A. hiiacinthus?) produce shaded area beneath them. Stimson and Polacheck (MPRL, 1977, 1979) found the shaded area to be less than 1 m^ per colony at 30 to 80 cm from the substrate. The density and number of other coral species beneath table Acropora. both dead and alive, were less than in controlled unshaded areas. The genera of corals occurring in the shaded areas v^ere Stiilocoer)ieUa, Montipora, Seriatopora. and various massive species Species of Acropora and Pocillopora piedominated the adjacent unshaded areas. Kastendick (MPRL, 1975, 1976) examined the habitat differences among eight species of fungiid corals which grow unattached on lagoon coral pinnacles and patch reefs. The young of two species were attached (Fur\gia fungites and Halomitra pileus) and found almost exclusively at the upper limit of adult distribution. It is likely that as they age, fungiids move passively down the slope. Kastendick observed invasions of colonies onto the foot area of several pinnacles after removal of these corals the previous year. Fur^gia spp. were found exclusively on coral rubble, whereas H pileus was most abundant on sandy substrate. Translocation of individuals up and down the pinnacle slope indicates that F. fungites has the most restricted habitat requirements, with H. pileus less so. Storms during the summer of 1972 (Nolan, 1975; Stimson, MPRL, 1974, 1976) destroyed large areas of coral growth on reefs with a southern exposure, even within the lagoon. Only massive species of Porites survived in any quantity on damaged reefs. First recolonizers were Acropora striata and A. s^ringoides. Stimson (MPRL, 1975, 1976) also noted that Sarcoph^ton sp., a soft coral, was an important colonist and component of the benthic fauna on storm-damaged reefs. As the hard coral commu- nity recovered, he believed that Sarcoph\^ton sp. would become progressively rarer. It was observed shading many corals, including P, damicornis and Seriatopora hystrix. Highsmith (1981a) suggested that corals with high skeletal density are less able to recolonize dead areas on their skeletons by tissue growth than less dense species. For example, he reports Porites lutea. with a relatively low density (1.4 to 1.5 g cm~^), is able to rapidly grow over dead skeletal regions, whereas Goniastrea retiformis (1.6 to 2.0 g cm~^) requires considerable skeletal deposi- tion and polyp growth reorientation to overgrow dead areas. Calcareous material produced by organisms other than stony corals is important in both the reef framework and sedimentary material. Animals, other then Scleractinia, which might make a significant contribution are the Fora- minifera, Mollusca, Bryozoa, Sclerosponges, and other Cin- daria. The occurrence of foraminifera tests in sedimentary material in the lagoon and beach sands at Enewetak is well documented (Emery et al., 1954; Odum and Odum, 1955; Deutsch and Lipps, 1976). Forams may consititute a sig- nificant percentage of lagoon sediment grains, but they are believed insignificant in reef growrth. Mollusc shells similarly constitute a minor component of lagoon sediments but do not contribute to reef growth. Cuffey (1973) found no bryozoans on the coralline algal ridge of Enewetak and very few in the area (which he terms the "back-ridge trough") immediately shoreward of it. The reef flat, similarly, has almost no bryozoa occurring on it. Areas between islands with abundant coral in shallow water also had relatively few bryozoa. Howpver, in the lagoon margin area, where larger patch reefs begin to SUBTIDAL ENVIRONMENTS AND ECOLOGY 121 occur, bryozoans increase in abundance, particularly in the patch reefs at depths below 9 m. Cuffey (1973) believed the floor of the deep lagoon, accessible to him only by dredging, lacked any diverse bryozoan fauna and only a few "small detrital fragments" of bryozoa were taken. The pinnacle reefs of the deep lagoon, however, contained an abundant and diverse complement of bryozoans. The steep leeward slope of the atoll apparently had the most diverse community of bryozoa, particularly below 9 m. Cuffey (1973) found bryozoans more abundant in Ber- muda than Enewetak, where they were infrequent in depths less than 9 m. He suggested that the considerably higher diversity of Enewetak corals might adversely influ- ence the relative success of bryozoa when compared to Bermuda. He makes the interesting observation that "the leeward (southwestern) side of Eniwetok Atoll harbors noticeably more bryozoans (both taxa and individuals) than does its windward (northeastern) side. Bryozoan distribu- tion on Eniwetok thus parallels sponge distribution within Pacific atolls, as described by De Laubenfels (1954)." In addition to not being principal frame builders on Enewetak reefs, bryozoans do not contribute any significant amounts of classic detritus to the sediments of the reef (Cuffey, 1973). Cuffey (1973), in considering the bryozoa of Enewetak, found that most species inhabited the undersides of corals and rocks on reefs. The most abundant bryozoans at Enewetak were encrusting cheilostomes which grow as thin, sheet-like crusts on the undersurfaces of corals or rocks. Most bryozoa inhabited these sheltered microhabi- tats and "function primarily as 'hidden encrusters,' adding small quantities of calcareous skeletal material to the reef framework." Cavity-filling tendency by bryozoa was noted in Bermuda but not at Enewetak. Hydrozoans of the genus Millepora are extremely important calcifying and reef-building organisms at Enewetak. In many areas, such as the large coral head zone of Odum and Odum (1955), Millepora spp. can form heads several meters across which grow to the low tide level where they form flat-topped structures (Fig. 12). In deeper water — including the ocean slopes of all sides, lagoon margin patch reefs, and lagoon pinnacles — Millepora spp. form large delicately branched, often fan- like, colonies (Fig 21). The stylastcrine hydrozoans, unlike Millepora, are not important carbonate producers at Enewetak. The delicate fan-like species of Sfy/aster are found beneath overhangs and within caves of patch reefs, pinnacles, and on outside reefs. Similar, but more robust, are two species of Distichopora which occur in similar areas but are often more exposed. These are extremely common on the leeward reef slope but do not produce reef framework. Tubipora musica is not common at Enewetak, being found only occasionally on reef fronts or on lagoon pinna- cles, and therefore does not produce significant reef struc- ture. Calcareous green algae, particularly members of Halimeda, are extremely important in sedimentation and reef building. The distribution of Halimeda in most subtidal environments at Enewetak is well documented (Hillis- Colinvaux, 1977, 1980, 1986; Emery et al., 1954; Colin, 1986). Borings at various atolls (Funafuti, Enewetak, Bikini, reviewed by HillisColinvaux, 1980) have shown Halimeda segments to be the major identifiable component of unconsolidated lagoon deposits. Milliman (1974) indi- cates that among sand-sized components of lagoon sedi- ments in Pacific and Atlantic atolls, Halimeda segments are generally the first or second most common material. Hillis- Colinvaux (1980) cites evidence in Couch et al. (1975) that Halimeda segments make a significant contribution not only to unconsolidated lagoon sediments but also to material underlying the reef rim. The fate of Halimeda plates in sediments varies. Some are shed intact, but a few species (H. macroph^sa and H, favulosa at Enewetak) have delicate segments that are easily broken (HillisColinvaux, 1980). Carbonate production rates by Halimeda at Enewetak are not well known, depending on plant density, generation time, and shedding rates. HillisColinvaux (1980) reports that population densities in Halimeda can vary by two or- ders of magnitude with concurrent effects on carbonate production. Turnover rates are perhaps lower than some published data (HillisColinvaux, 1980) since Halimeda is predominantly a long-lived alga. One experiment at Enewetak indicated that 70% of the original thalli were still present after 4 months (HillisColinvaux, 1980). Dense populations of Halimeda at Enewetak and else- where have about 100 plants m^^ of the H. incra^^sata- c\;lindrica type thallus. The rock-growing H. opuntia type can have higher densities of plant material, although abso- lute numbers of plants may be less. HillisColinvaux (1980) estimates that the H. incrassatact^Hndrica types would pro- duce only about 10% of the total carbonate accumulation in the lagoon (Smith and Kinsey, 1976) if they covered the major portion of the lagoon bottom. She was not aware at that time of the presence of the "Halimeda meadows" and the estimated percent coverage of the deep lagoon bottom predominantly by Halimeda. The contribution of Halimeda segments from lagoon pinnacles may be smaller than HillisColinvaux (1980) calculated when a comparison was made to Halimeda from flat lagoon bottoms. Bioerosion of Reefs The agents of bioerosion at Enewetak act in a variety of ways. Some, such as the boring sponges of the genus Cliona, excavate chambers on the carbonate skeletons of living corals and virtually any other carbonate substrate. The shells of molluscs, coral rubble, and other small car- bonate fragments can be attacked. Other organisms, in the course of feeding activities, rasp away the surface layers of carbonate while grazing the thin film of algae which covers such surfaces. The parrot fishes, surgeonfishcs, various echinoderms, and other such herbivores generally pass the carbonate material through their gut, subjecting it not only to mechanical effects but also to chemical effects. Other 122 COLIN organisms may prey directly on calcifying organisms and in the proces% often damage the skeleton. A few species of fishes vigorously attack coral skele- tons, biting off and ingesting the tips of branched species. Randall (1974) observed the pufferfish, Arothoron nigro- punctatus, feeding heavily (85 to 100% of gut contents) on corals, particularly Acropora and Montipora. Hiatt and Strasburg (1960) found corals in the guts of nine plectog- nath fishes (two triggerfishes, three filefishes, three puffers, and one sharpnose puffer). Most, but not all individuals of any species, had ingested branched coral tips in various amounts. Although none of these fishes are obligate coral predators, many contain coral tips in such quantity that these must constitute a regular part of their diet at Enewetak (Randall, 1974). Large portions of coral skeleton will, on occasion, have the ends of the branches removed, often with piles of coral fragments left in the depression. This is seen in Pontes rus at Enewetak (Fig. 21) and in other species. It is assumed this phenomenon results from the activities of fishes which feed on coral branches, but the feeding by some of these fishes is seldom observed. Other coral-feeding fishes tend to eat only the polyps, leaving the skeleton essentially intact. In such cases, the polyp normally regenerates. A number of butterflyfishes (Chaetodontidae) and damselfishes of the genus Plectrogli^phidodon feed on corals in this matter (Motta, 1980; Randall, 1974; Reese, 1973, 1975, 1977) and are discussed elsewhere. Randall (1974) notes also that the blenny Ecsenius bicolor at Enewetak has been observed feeding on Acropora. Some herbivorous fishes occasionally scrape at the sur- face of living corals doing more damage than the chaeto- dontids. Scarids produced a characteristic scrape mark on corals with an elongate furrow, often with a slight ridge along its midline where the two sides of the beak fuse. Hiatt and Strasburg (1960) found some species of scarids at Enewetak had fed on corals. Randall (1974) has reviewed the question of parrot fish grazing on live corals and discusses an apparent disparity between published data on coral feeding by scarids at Heron Island, Great Barrier Reef, and Hiatt and Strasburg's (1960) information. He found no obvious reason for the differences observed but suggested that local coral cover may influence how much coral is ingested by parrot fishes. Although some scarids do graze live corals, the impact of this behavior is probably minor compared to the effect on sediment pro- duction and deposition. Randall (1974), Ogden (1977), and others have docu- mented the role of scarids in sediment production. The rasping of rock or coral for its algal film is the first step. This material is then ground to a fine consistency by the pharyngeal mill of the parrot fish, passed through the gut, and eventually expelled. The rain of sedimentary material shed when parrot fishes defecate is impressive, and the amount of sediment produced from hard substrates by this mechanism is enormous. Also important in sediment production are fishes which reduce the hard parts of invertebrates (mollusc shells, echi- noid tests and spines, crustaceans, etc.) to bits. Randall (1974) reports that plectognaths with their fused or but- tressed teeth, lethrinids with molariform teeth, labrids with pharyngeal teeth, and dasyatid and myliobatid rays with plate-like jaws are well adapted for this purpose. Massive corals at Enewetak are attacked by a number of biological agents. Although seldom visible, these agents weaken the skeleton to the point that physical factors can break the colony loose or cause it to crumble. Highsmith (1981a) reports that clinoid sponges accounted for 70 to 80% of skeletal damage in various massive corals from Enewetak. They did not burrow deeply into the skeleton, only a few millimeters, but extended interconnected chambers laterally beneath dead surfaces of the coral colony. Highsmith (1981a) reported that 65 to 95% of the boring was within the "dead area" of skeletons. In a mas- sive coral this "dead area" includes the area around the basal attachment and dead spots on the colony surface. Similarly, these dead areas are heavily eroded by grazing organisms. When exposed to light or scraped (as when overlying skeletal material is removed), clinoid sponges engage in rapid burrowing (Ruetzler, 1975) Heavy grazing pressure, combined with this response, may produce rapid erosion rates at basal attachments. Highsmith (1981a) points out that skeletal weakening at the base, combined with storm-induced water motion, may not be sufficient to dislodge most massive colonies. However, coral rubble on the bottom can be put into motion by storm waves and, to a point, may be the most significant force in breaking heads loose. Eventually though, "as massive corals grow, they become more sus- ceptible to breakage by storm currents and less susceptible to breakage by suspended rubble or to biocrosion detach- ment." The alpheid shrimps occurring in deep grooves on Goniastrea retiformis apparently form the grooves, not by boring or erosion, but by preventing growth of coral in that area while the remainder of the colony continues to increase in size (Fig. 22). These grooves, though, provide dead areas which penetrate deeply into the G. retiforrr}is head and are penetrated by boring organisms (Highsmith, 1981a). Highsmith (1981b) suggested that bioerosional damage to corals is positively correlated with increasing skeletal density. Five species of Enewetak corals {Ouloph\^lha crispa. Fauia pallida, Goniastrea retiformis. Pavona clauus, and Pontes lutea) had a positive correlation between bioerosion and density. This correlation did not correspond to differences in growth rates. The slowest growing species, F. pallida, was the least bored. Among molluscs, the boring bivalve Lithophaga curta preferentially colonized the coral Montipora berrvi (Highsmith, 1980). Boring bivalves in general have thin, weak shells and, if exposed, are easily eaten by fish preda- tors. Highsmith (1981a), for example, reported that the SUBTIDAL ENVIRONMENTS AND ECOLOGY 123 Fig. 22 Heads of Goniastrea retiformis with deep grooves formed through the activities of alpheid shrimps. wrasse Thalassoma lutescens readily took bivalves exposed during collecting. Bivalves were not common borers of Enewetak massive corals. Highsmith (1981a) found only four bivalves in more than 100 coral heads and contrasts with other areas where they produce significant boring. Polychaetcs are also significant borers of corals (Highsmith, 1981a) but are often believed to occupy empty sponge chambers. He found 280 polychaetes in a single Pontes lutea head; the diversity of polychaetes exceeded any other infaunal organisms. Although they arc common, they are probably not as important borers as are sponges. Sipunculans were imp>ortant borers of coral rub- ble, rather than live coral (Highsmith, 1981a). Highsmith (1981b) discussed the role of endolithic algae, Ostreobium spp., in several species of Enewetak corals. They occur as one or more dark green bands in the upper few centimeters of the coral skeleton. He found Ostreobium in every coral sampled from the surface to 30 m depth. No significant effect by the filamentous algae on the integrity of coral skeletons was detected. In some species of corals there was an inverse correlation, with considerable variation, between water depth of a coral and the depth of its outermost algal band. Algal bands are believed to occur where and when conditions are suitable for vigorous growth. DiSalvo (1969) isolated bacteria from within the skele- ton of the coral Pontes lobata. Bacteria were cultured from light brown discolored regions revealed when the corals were split of>en. Attempts to culture bacteria from adja- cent, nondiscolored skeleton were not successful. Some of 124 COLIN the isolated bacteria were able to digest chitin in vitro, and DiSalvo suggested these might weaken the skeleton by breaking down the organic matrix. DiSalvo (1969) also found that sediments in proximity to the bases of corals had 10^ to 10* bacteria g dry wt^' of which 10 to 20% were chitin-digesting varieties. Thus there is a ready source of suitable bacteria close to the coral's skeleton. Invertebrate Coral Predators Some invertebrates are also coral predators at Encwetak. The crownofthorns starfish, Acanthaster plana, is found in many areas. Most lagoon pinnacles have one or more A planci, and evidence of their feeding activities on corals is apparent. The status of A. planci populations at Enewetak, population changes, and impact on the reefs in recent times arc not well known. Population levels are cer- tainly below "plague" levels. Allen (1972b) noted that "during the summer of 1970 the author observed an increasing number of Acanthaster at Eniwetok. Prior to this date relatively few were observed." Starck (MPRL, 1971, 1972) found "no unusual populations of Acanthaster." About a Ihour (one dive) search during the day would result in one to two individuals, whereas a nighttime search of about only one-quarter the area revealed 10 to 12 specimens. Starck found a wrasse, Cheilinus undulatus, of about 50 kg weight (of four examined) with a large, nearly intact A. planci in its gut contents. Starck, in Ander- son (1979), reported a sizeable population, perhaps as many as 50 to 100 A planci on Pole Pinnacle, but stated that there was no extensive damage to the coral there. The juveniles of A planci apparently occur beneath rubble on reefs. Lisa Boucher and Scott Johnson (personal communication) report finding numerous examples from a few centimeters to less than 1 cm disk diameter on pinna- cles near Enewetak Island. Although they never found these A. planci to be very common, a distinct increase in the numbers of juveniles encountered was noted in April 1982. Storm Destruction of Reefs The effect of subtyphoon storms (tropical storms, tropi- cal depressions) on subtidal environments can be devastat- ing. Many such storms occur compared to full strength typhoons and are often not noted in historical records. Damage from wind and rain to terrestrial areas may be minor, but the swell produced by such storms can wreak havoc in shallow-water communities. The production of boulder ramparts by storms is well known, and such struc- tures occur on the southwest islands of Enewetak. Since the ocean shores of these islands are normally in the lee and the reef slope is steep and close to shore, the infre- quent reversal of wind and waves can cause catastrophic destruction of corals in shallow water, moving vast quanti- ties of material onto the shore or into deeper water. The movement and effect of ocean swells in the lagoon are important. The wide pass at Enewetak is sufficiently deep to allow ocean swell from the southeast to southwest to enter the lagoon. Ocean swell is also refracted at the pass so that wave trains moving from the west and southwest can come through the pass and proceed north to northeast to reach lagoon shores. These long period swells have no direct effect on the deep lagoon communi- ties. However, when they reach the lagoon shore of wind- ward islands or shallow pinnacle or patch reefs, they can turn these shallow-water communities into churning mael- stroms of breaking waves. One such period of swells from the southwest to west for 3 days in July 1982 turned the lagoon shore of Enewetak Island and other southern islands into a mass of dark brown water (with essentially zero visibility) above 6 m depth with breakers to 2 m high where depths were less than about 4 m. Significant swells and breakers persisted for nearly 1 week. Many fishes, molluscs, and other invertebrates were killed and cast up on the beaches. In places sediment and rubble were eroded away as much as 1 m or more. Many of the delicate corals on shallow reefs {Pocillopora edouiixi. Millepora spp.) were broken to stubs. Carbonate material from shallow water, particularly large pieces such as coral boulders, can be deposited on the islands by storm waves, transported into the lagoon or transported downslope on the seaward reefs. Various islands of Enewetak are densely covered with recently deposited coral boulders, and boulder ramparts are evident on the seaward beaches of some islands. Less visible, but perhaps more significant, is downslope transport of rubble on seaward reefs. Talus was evident at many locations examined by the submersible Makali'i. and shallow-water coral rubble was extremely evident in the material photo- graphed. At 300 to 360 m, the maximum depths visited, the slope of the bottom was generally too steep for extremely large talus accumulations. Larger accumulations of talus should lie below those depths where the bottom slope is less steep. Storm swell within the lagoon may be a major factor controlling the morphology of lagoon margin patch reefs and shallow pinnacle reefs. Choptop Reef had moderate damage from swells entering the lagoon in July 1982. Some large coral heads, their bases weakened, were tum- bled over. Pieces of Porites c^ilindrica colonies as much as a meter across were torn loose from larger colonies and rolled a few meters over the bottom. Although individual branches were often broken, such pieces formed satellite patches of P. cfjiindrica which survived and grew. In another instance, a tunnel torn through a huge mass of P. ci;lindhca at Choptop Reef during a tropical storm in March 1981 was collapsed by the July 1982 storm. In both instances the total structure was fractured. Swell within Enewetak Lagoon seems capable of breaking apart patch reef features which reach too near the surface. Where the internal structure of lagoon margin patch reefs is visible, they seem little more than accumulations of poorly cemented coral rubble. One well-known lagoon pin- nacle, "Tunnel Pinnacle" (Fig. 6), has had the "tunnel" col- lapsed, almost certainly by storm swell, during the past few years. Reese (1981) provides a description of the SUBTIDAL ENVIRONMENTS AND ECOLOGY 125 effects of storms on the corals and butterflyfishes (chaeto- lontidae) of "Tunnel Pinnacle" and other pinnacles at Enewetak. The lack of significant cementation on upper surfaces of patch reefs and pinnacles by coralline algae may reduce greatly the amount of wave energy required to crumble the structure. Protection of the lagoon from storm swells by a complete or nearly complete atoll rim with no deep passages may also be important. At Ujilang Atoll, which has no major passes allowing ocean swell to enter the southern lagoon, lagoon margin patch reefs examined were near planar on top at about the level of spring low tides with near vertical edges dropping to a few meters in depth. The patch reefs were well cemented by coralline algae on their upper surfaces. Ujilang is exposed to essen- tially the same oceanic conditions (wind, waves, and storms) as Enewetak, yet patch reefs of such morphology are not found at Enewetak. Herbivory in Subtidal Communities Much work on herbivory and its impact on the ecosys- tems at Enewetak has been undertaken on intertidal areas because of the large, accessible area of such environments at the atoll and the abundance of herbivores there. Her- bivory is a major factor influencing subtidal communities in the lagoon and on the seaward margin. Evidence of intense grazing pressure can be found in many subtidal areas, both on hard and soft substrata. Unlike the intertidal areas, subtidal areas are accessible to herbivores at all times. On exposed rock substrates, both seaward and lagoonward of the reef flat at Enewetak, tooth marks from the action of grazing fishes arc nearly ubiquitous in areas to at least 15 to 20 m depth. Many show considerable erosion from grazing (Fig. 23) with angular facets on the rock, and deep tooth marks, particu- larly from large parrot fishes, are often densely grouped. The principal grazers of hard substrates at Enewetak are fishes, particularly parrot fishes (Scaridae) and sur- geonfishes (Acanthuridae). In general, algae on subtidal rock surfaces are close-cropped except in the case where the species may be heavily calcified {Halimeda spp.) or potentially toxic or distasteful {Li;ngbia sp). In this respect subtidal rock surfaces are not qualitatively different from the intertidal reef flat. Macroalgae and algal films are found at all depths in the lagoon. Surgeonfishes show a well-defined zonation among this largely herbivorous family. Acanthurus triostequs. A. achil- les. A guttatus. and A. lineatus are principally found on windward and leeward reef flats, back reef areas, and the spur and groove zone — all shallow-water environments. In somewhat deeper water on the seaward reefs, lagoon patch reefs, and pinnacles are found species of Ctenochaetus, Zebrasoma, Acanthurus nigrofuscus. and A. olivaceus. Two species of Acanthurus at Enewetak, A. thompsoni and A. bleekeri, feed on zooplankton as do most species of Naso. The species of surgeonfishes with a well-developed gizzard-like stomach commonly feed on sediment bottoms and ingest, along with the algae, consid- erable sand (Randall, 1956; Hiatt and Strasburg, 1960) (Fig. 23). Herbivory occurs on sediment bottoms where macroal gae and microalgac occur. Macroalgae can occur as dense stands, as exemplified by the species of Halimeda and Caulerpa, and many are probably unpalatable to her- bivores. Microalgae can occur as nearly invisible films on sediment grains on the surface of the sediment bottom, but easily apparent films (algal mats) covering many square meters are often found from a few meters depth to the deepest portion of the lagoon. Epiphytic algae also grow upon larger algae and are often more desirable to her- bivores than the plants on which they grow. Invertebrate grazers of rock surfaces are not as impor- tant at Enewetak as in the western Atlantic. In many Caribbean locations sea urchins, particularly Diadema antil- larum, are extremely abundant and as herbivores have an impact equaling, or exceeding, that of fishes (Ogden, 1976; Ogden and Lobcl, 1978). Sea urchins, particularly diademnids, are not nearly as abundant, except in localized areas, on reefs at Enewetak. Fishes are important herbivores of sediment bottoms at Enewetak. The principal herbivore families of hard substrata, parrot fishes and surgeonfishes, range onto sedi- ment bottoms also (Fig. 23). Different species from those that remain on the reef are often involved. As distance increases from the shelter of the reef, the grazing pressure of reef-based herbivores decreases. They are exposed to increasing risk of predation with increasing distance from shelter. Therefore, soft bottoms near reef shelter are more heavily grazed, often to the extent that no visible bcnthic plant growth except the less desirable species mentioned previously occur near the reef. This results in a phenomenon most easily visible from the air in which reefs arc surrounded by a light colored band, compared to sedi- ment substrata farther away, representing the denuded substatum close to the reef. This area has been termed a "halo" (Randall, 1965) or the "Randall zone" (Ogden, 1976) and is a feature found near both Indo-West Pacific and Atlantic reefs. Other herbivores, particularly invertebrates, exist far from the reef, either remaining on or above the sediment surface or burying and burrowing in the bottom. Dense stands of macroalgae provide excellent shelter for small herbivores, both fishes and invertebrates which can hide among the thalli. Although these macroalgae are not pri- mary foods for these herbivores, the environment created provides abundant spaces for epiphytic algae (and cpizoic organisms also) w)iich are suitable for the small herbivores. In areas without dense algal cover, burrowing organ- isms can function as herbivores without the need of shelter. Irregular sea urchins (Spatangoidea) occur abun- dantly in Enewetak Lagoon sediment bottoms and apparently process sediment grains for the algal matter on their surfaces. These and other "sediment processors" must pass relatively large amounts of sediment to obtain sufficient organic matter. 126 COLIN % 1>- Fig. 23 Upper left: Heavily grazed rock sub- strata, seaward reefs. Right: Heavily grazed dead coral skeleton on seaward reefs. Lower left: Sur- geonfishes, probably Acanthurus mata, feeding on algal films on sediment substrata near lagoon margin patch reefs. Other herbivores are found on the sediment surface. The gastropod, Strombus luhuanus, can occasionally occur in localized high densities over open sediment in water 2 to 10 m deep. Densities more than 10 individuals per m^ with distinct (advancing?) edges to the population were often seen with adjacent areas lacking S. luhuanus. High densities of S. luhuanus have been found at stations where only a few weeks previously the species was absent. Numerous species of sea cucumbers (Holothuroidea) are found on sediment — sometimes near reefs, but not always. They process sediment through their gut and are relatively immune to predation, probably because of their toxin (holothurin). Lamberson (1978) found the relatively large species (up to % m in length) Thelenota anax in relative abundance at Encwetak at 5 to 30 m depth. This species was found on lagoon pinnacles and patch reefs, on sandy bottoms near reefs, and on the vertical slope off the leeward side near Biken. Holothurians are important sand processors of reef areas. Bakus (1963) reported that Holothuria dificilus in- gested sediment particles up to 2 mm in size, but about 80% were less than 250 microns in diameter. Holothuria atra fed on even larger rubble, up to 20 mm in size. Bakus (1973) indicates that beyond selection of suitable size, there is little specificity among tropical holothurians for sediments ingested. Hammond (1981) found that among West Indian holothurians and echinoids (irregular) that sig- nificant carbonate dissolution and sediment grain-size modification did not occur during passage of sediment through the guts of five species of tropical deposit-feeding cchinoderms. A sirr' r situation probably exists for Enewetak species. Irregular urchins are important herbivores of open sandy areas. Extremely high densities (more than 50 m^^) of moderately large (more than 30 to 35 mm test length) species have been observed over large areas. This implies SUBTIDAL ENVIRONMENTS AND ECOLOGY 127 that a significant amount of algal production must be avail- able for them to survive for any period. Population size of irregular urchins seems influenced more by recruitment success than by food availability (V. S. Frey, unpublished data). There are many other organisms living in the sediment which ultimately make their living from the algal produc- tion occurring on sediment surfaces or passing over the sediment. The callianassid crustaceans, mentioned previ- ously for their bioturbational activities, almost certainly process prodigeous quantities of sediment to winnow the organics present on the surface of the sediment grains. They may additionally exploit the algal fragments which enter their burrow systems. On hard substratum, some herbivores at Enewetak live within a limited area which they maintain as a territory. Some damselfishes, particularly S(egas(es nigricans, estab- lish and maintain an "algal lawn" of filamentous algae. The algal lawn is often found on basal dead parts of the fine branches of Acropora spp. corals and is strongly defended against intruding herbivores. This action by S. nigricans is identical to the western Atlantic Stegasfes planifrons, the first species for which algal lawn maintenance was described (Kaufmann, 1977). It is likely that S. nigricans can kill coral polyps in expanding its algal plot, and large numbers can significantly damage Acropora spp. corals. The darkened areas of damselfish algal plots are common features of Acropora spp. thickets at Enewetak (Fig. 24). The long-range effect of these areas of dead coral has not been examined, though areas a few meters square of dead Acropora are often found in the midst of dense thickets (Fig. 24). These may potentially represent old algal plots eroded away by other herbivores. The general lack of herbivores as significant as fishes at Enewetak presents an interesting contrast to reefs in some other areas of the world. In the tropical western Atlantic, sea urchins, particularly Diadema antillarum, play a role as herbivores equal to or superior to that of fishes Fig. 24 Algal "lawns" on Acropora sp. corals produced by the herbivorous damselfish Stegastes nigricans in the Enewetak Island quarry. Upper left: Large dead area in the Acropora sp. coral possibly produced by the presence of an algal lawn. Upper right: Aerial view of Enewetak quarry Acropora sp. with many algal lawns (dark spots) established on the coral. Lower left: Acropora sp. coral in the quarry with algal lawns. Lower right: Stegastes nigricans with its algal lawn (dark area to right of fish). 128 COLIN (Ogden, 1976). Sea urchins are abundant and conspicuous elements of the reef fauna. At Enewetak and much of the Indo-West Pacific, sea urchins are considerably less abun- dant. Diademnids are particularly less conspicuous, being small and deeply hidden in the reef. One possible explana- tion for this difference is a higher population of fishes which prey on sea urchins in the Indo-West Pacific (Fricke, 1971). in general, western Pacific fish faunas are consider- ably more diverse and "highly evolved" than the Atlantic, and more species may be adapted for exploiting sea urchins (among other things) as food. Gilmartin (1960) indicated that herbivores have a much smaller influence on benthic algal communities at 19 to 63 m at Enewetak than they do on shallower communities. Bakus (1967) felt that grazers influence the benthic biota most in water <10 m deep. With heavy grazing pressure from herbivores, the pres- ence of dense abundant algae implies some reason for its avoidance by herbivores. For example, the filamentous strands of the blue-green algae, L[^ngbia sp., occur extremely abundantly on projections, particularly corals; on many reefs at depths below 6 to 9 m on windward lagoon reefs; and as shallow as 1 m in protected areas. The alga covers large areas of the substratum, streaming from corals and resembling long reddish hair. Often it virtually covers the coral with a tangle of filaments that is extremely diffi- cult to remove. Lf^ngbia sp. often seems to have detrimen- tal effects on the live coral with the coral tissue beneath the algae appearing unhealthy. Often coral areas beneath the alga are dead, but whether the alga is the causative agent or simply grows on available substratum is not known. At some lagoon pinnacle reefs, such as Medren Pinnacle, Liingbia sp. appears to have a significant impact on the total reef and may be significant in coral mortality there. Li/ngbia sp. is also abundant on some lagoon margin patch reefs below 6 m depth but is absent on the shal- lower portions of the same reefs. The small sea hare, Sfylocheilus longicauda, occurs abundantly on the Li^ngbia sp. Sarver (MPRL, 1976) reported it feeds almost exclusively on Liingbia sp. and spends its life, exclusive of larval stages, on the alga. The sea hare accumulates an antitumor agent, debromoaplysia- toxin, from L^ingbia sp. and Oscillatoria sp. at Enewetak. This poisonous lipid was first isolated from the digestive tract of S. longicauda but has its origin from the blue-green algae (Moore, MPRL, 1976). Sarver found that adult S. longicauda (about 3 to 7 g) reproduce rapidly, at about 30 days age, and consume about 75 to 100 g of the alga during their lifetime. Bioturbation in the Deep Lagoon A high level of bioturbation in sediment bottoms throughout the deep lagoon has been verified by recent work (Suchanek and Colin, 1986; Suchanek et al., 1986). Gilmartin (1960) first noted, based on in-situ observations, significant bioturbation of deep lagoon bottoms, but several other authors have commented on it Emery et al. (1954) noted, in shallow lagoon photographs, disturbance of the bottom and burrowing. Hillis Colinvaux (1980) noted a "relative prominence of animal mounds and castings on the lagoon floor near the base of pinnacles in 40 m." Bioturbation of the deep lagoon is evidenced by the ubiquitous presence of "lebensspuren," a term designating any sedimentary structure produced by a living organism (Hantzschcl, 1962). A wide variety of lebensspuren occurs on sediment bottoms at Enewetak, but the conical mounds of ghost shrimps (Callianassids) that are as much as 1 m in diameter and 30 to 40 cm high are the most apparent type. The conical mounds represent the excurrent open- ings of complex burrow systems which penetrate deep into lagoon sediments and underlie nearly all the sediment bottoms. Photographs from the Enewetak Lagoon benthic sur- vey, observations from the submersible Makali'i, and scuba diving on the lagoon margin have confirmed the near pan- lagoon (below 10 m depth) distribution of callianassid mounds. The basic morphology of the burrow system, pumping rates, and sediment processing have been exam- ined and will be discussed subsequently. Since the lagoon sediments are the major repository of remaining radionu- clides at Enewetak, an understanding of the mixing and resuspension abilities of lebensspuren-producing organisms, particularly callianassids, is of basic relevance in any con- sideration of the future fate of long-lived radionuclides in the marine environment. Callianassid mounds are often referred to as "vol- canoes" because of their conical shape with steeply sloping sides, a small apical depression (crater), and the resulting eruption when water and sediment are pumped out of the apical depression at irregular intervals. These volcanoes can be so dense that essentially no level substratum can be found in a large area, the bottom being comprised solely of volcanoes and incurrent depressions of the callianassid bur- row systems. It is estimated, based on the photographic survey, submersible work, and diving observations that about one volcano per square meter occurs overall in the lagoon below 10 m depth. Since approximately 85% of the lagoon bottom is soft substratum and covers about 8 X 10 m , on the order of 10 callianassid volcanoes occur in the Enewetak Lagoon. Densities may vary from place to place by a factor of 10, and several species of callianassids are certainly involved. A typical callianassid burrow system at Enewetak con- sists of three major elements: (1) conical depressions on the surface where sediment enters the system, (2) a com- plex network of horizontal and vertical burrows, and (3) conical mounds (volcanoes) where sediment and water are discharged. The incurrent openings to the system (Suchanek and Colin, 1986), in which sediment and water are drawn into the system, appear as a conical depression many centimeters in diameter. The excurrent side of the system is represented by the volcanoes, each of which is fed by a vertical tube at its center through which sand and water are pumped by the action of ghost shrimp in the SUBTIDAL ENVIRONMENTS AND ECOLOGY 129 tunnels below. Burrow systems linking the down holes with the volcanoes are complex, often consisting of a series of interconnected horizontal tunnels (as much as 4 to 5 cm in diameter), and sloping to vertical tunnels connecting differ- ent levels. As much as 1300 g of sediments were ejected per day from each volcano. Callianassids alter grain size distribution of processed sediment to produce a very con- sistent sediment size fraction which is depleted (compared to some other Enewetak sediments) in both coarse (>2 mm) and fine (<90 microns) sediments (Suchanek and Colin, 1985). As much as 3 liters of water evolves daily from volcanoes during sediment-pumping activities (Colin et al., 1986). Volcano water contained suspended particu- lates >0.45 microns in diameter at levels at least five times that of water immediately overlying the sediment, which itself has elevated particulates compared to "aver- age" lagoon water (Colin et al., 1986). "Tagged" (painted with fluorescent paint) sediment experiments have demonstrated that most large particles, more than 1.5 to 2 mm in diameter, entering the ghost shrimp burrow system are not returned to the surface (Suchanek et al., 1986). Probably such particles are too large to be temporarily suspended by pumping, and it is believed that callianassids "store" large particulates in unused portions of the burrow system. A constant "disturbance" effect occurs in bioturbation at Enewetak. The constant grazing by herbivores, the digging into the sediment by carnivores looking for prey, and the ingesting and then expelling of sediment by the surface-dwelling species cause the upper few centimeters of sediment to be constantly disturbed (Suchanek and Colin, 1986). The sediment surface, through this action, is a continuous mosaic of small pits and mounds, disturbed places and tracks, all from this surface reworking. In the lagoon at depths shallower than about 5 to 6 m, this evi- dence is quickly obliterated by wave ripples, but below that depth the disturbances remain apparent for some time. FISH COMMUNITIES The fish fauna of Enewetak is quite diverse, numbering more than 800 species. There are, certainly, a number of species yet to be recorded from the atoll. Fish species are not evenly distributed around the atoll; many occur com- monly in only one type of habitat. These preferences result in general fish communities which are identifiable assem- blages of species. Hiatt and Strasburg (1960) made the first (and still best) attempt to characterize fish communi- ties of various areas of Enewetak and to examine their trophic relationships. Subsequent researchers have exam- ined feeding by various portions of the Enewetak fish fauna (Hobson and Chess, 1978; Randall, 1980; Bakus, 1967; Gerber and Marshall, 1974; Smith and Paulson, 1974; Reese, 1975, 1977; and others), but a definitive study of the overall trophic dynamics of fishes has never been undertaken. Randall has reviewed records of fishes since Schultz and collaborators (1953 to 1966), and a checklist is included in Chapter 27 of Volume 11, this publication. Although this may be approaching a definitive list of fishes for the Marshall Islands for shallow-water species, it was apparent from the observations and photographs from the submersible Makali'i that a significant number remain to be recorded (many undoubtedly undescribed) from depths greater than those usually penetrated by scuba divers. The species of fishes inhabiting a given location at Enewetak are strongly influenced by environmental factors. Primary among these are substratum types (hard or soft, variations of these), depth, current, wave action, and oth- ers. The food of Enewetak fishes is based on two different sources: primary production from atoll bottoms and waters and oceanic zooplankton and phytoplankton. The relative importance of these two pathways has never been rigorously compared, but the high productivity of reef flat and lagoon versus the low density of phytoplankton and zooplankton in oceanic water upcurrent of Enewetak imply that the former is of considerable significance. Seeing the immense numbers of large, herbivorous fishes on spur and groove, reef flat, and shallow patch reefs impresses one with the amount of fish life supported by algae growing on the substratum. Predators of mid- and upper-water lagoon areas arc varied. Hiatt and Strasburg (1960) differentiate mid- and surface-water communities but p>oint out that some large carnivores enter both areas. They felt that surface water communities contained various sizes of zooplankton, small plankton-feeding fishes (round herring and silversides), larger macroplankton-feeding fishes (such as halfbeaks), and piscivores (needlefish, tunas, barracuda, jacks). Randall (1980) and Hiatt and Strasburg (1960) have discussed the food habits and general habits of many of these. The Carangidae are important predators at Enewetak. The members of the genus Caranx are largely fish eaters, occasionally taking cephalopods or crustaceans. Caranx ignobi/is, the largest species of the genus, reaches 80 kg, and as Randall (1980) notes "may be encountered any- where in the atoll environment including water surprisingly shallow for such a large fish." Caranx melampi^gus is very common on Enewetak reefs and feeds largely on reef fishes, including some such as Caracanthus sp., that live deep within the branches of living corals. The rainbow runner, Elagatts bipinnulatus, is a mid-water feeding caran- gid which occurs in schools above reefs. Among scombrid fishes, the dogtooth tuna, G\;mnosarcla unicolor. is the only tuna commonly seen around lagoon pinnacles. It also occurs on outer reefs and is a predator on free-swimming fishes, including Naso spp., Caesio, Ptercxaesio, and Decapterus (Randall, 1980). Hiatt and Strasburg (1960) reported it and the common lit- tle tunny, Euthynnus affinis, as slashing through the dense schools of round herring. Some species of moderately large fishes are detri- tivores. Mullets are common on shallow reefs, both lagoon- ward and seaward of the reef flat. Crenimugil crenilabis has 130 COLIN been seen to expel sand through the gills after feeding, a process in eommon with many smaller fishes, and appears to feed on fine algae detritus (Randall, 1980). The largest planktivorous bony fish at Enewetak is probably the milkfish, Chanos chanos, which occurs occa- sionally on outer reefs and in the lagoon. The largest planktivore, at least within the lagoon, is the manta ray, Manta alfredi. They frequent lagoon margin areas, often in water 3 to 6 m deep, and the wide channel area. On one occasion, a group of more than 100 M. alfredi with a 2 m or more span were seen from the air in the deep ocean just beyond the wide channel. Randall (1980) has summarized the food habits of larger groupers (Serranidae), snappers (Lutjanidae), and emperors (Lcthrinidae). All are benthic predators, although some groupers and snappers will rise to a lure in mid- waters. Large oceanic predatory fishes occur commonly around Enewetak. Tunas, wahoo (Acanthoc\,;bium solanderi), dolphin (Cor^/phaena hippurus), and billfishes are known to frequent waters within a few kilometers of shore (Schultz et al., 1952). Hiatt and Strasburg (1960) note that the presence of larval fishes and crustaceans produced by reef- and shore-dwelling adults, "supplementing the usual high seas forage species, probably is significant in attracting tunas (and other large pelagic fishes) to mid-ocean islands." There are numerous fishes that are highly specialized in their food habits. For example, in the Chaetodontidae, Reese (1975, 1977, 1981) found that at Enewetak, 10 of 17 species are coral-feeders, whereas two are planktivores and five are "omnivores " Among coral feeders, four were believed to be obligate coral predators, with fine comb-like teeth for biting off coral polyps. One species, Chaetodon ornatissimus, appears to eat coral mucus with its fleshy lips rather than biting off the polyps like other species. Other coral-feeding species at Enewetak ingested other ani- mal matter as food. The other extreme is C unimacuhtus which even ingests fragments of septa as it feeds on polyps. Other coral-polyp feeders include Oxiimonocanthus lon- girostris, Labnchth\js unilineata, and Labropsis spp. Some damselfishes, such as Plectrogi;lphidodon johnstonianus and P dickii, have been observed to feed on coral polyps (Ran- dall, 1974). Few Enewetak fishes feed on sfjonges. Hiatt and Stras- burg (1960) recorded only Arothron mappa, a puffer, as having eaten sponges. They examined, however, only one species of Pomacanthidae, a group shown to contain sponge-feeding species in the West Indies (Randall and Hartmann, 1967). Among fishes there are several "cleaners" at Enewetak, those species which eat ectoparasites and con- sume mucus from the bodies of other, usually larger, fishes. Most important are members of the wrasse genus Labroides, particularly L. dimidiatus. There are also various invertebrate cleaners, usually shrimps, on Enewetak reefs. Some fishes associate with invertebrates that are avoided by predators as one method of gaining protection. Anemonefishes associate with sea anemones; in spite of this, they are occasionally eaten by other fishes. Hiatt and Strasburg (1960) found a juvenile Amphiprion melanopus in Apogon noLiem/asciafus. Allen (1972a) reported that disoriented Amphiprion (due to "fin-clipping" manipula- tions) were sometimes eaten by groupers, particularly An[^perodon leucogrammicus. Allen (1972b) described a cardinalfish, Siphamia fuscolineata, sheltering among the venomous spines of the crown-of-thorns starfish, Acanthaster planci. Between eight and 31 fish were found with each of four A. planci; however, only a small percent- age of starfish had the apogonid associated with it. Species of Siphamia are more often found associated with diadema- tid sea urchins. A small group of fishes shelter among branched corals, some never leaving the coral. Hiatt and Strasburg (1960) illustrate some which include the gobies of Gobiodon and Paragobiodon , plus the scorpaenoid genus Caracanthus. There are similar invertebrate associates, particularly crabs of the genus Trapezia and some alpheid shrimps. A much greater number of fishes temporarily shelter in branched corals when danger threatens. The hundreds of Chromis caerulea. C atripectoralis. Dasc^illus reticulatus, and D aruanus stationed above small heads of Pocillopora corals which can take nearly instant shelter on that head (Hobson and Chess, 1978) are astounding. Hobson and Chess (1978) have examined the feeding relationship between zooplankton and planktivorous fishes of the lagoon margin. At their two study areas, one northeast of Jedrol within the strong influence of currents in the deep channel and the second in the lee of Bokandre- tok where currents are weak, they found that current pat- terns sharply affected trophic relationships. The plankters ingested by diurnal and nocturnal planktivores were quite different. There was an abundance of suitable zooplankton in strong current areas, whereas areas of weak currents were poor in zooplankton. These poor areas in the lee of reefs and islands were, however, rich in debris from the reefs and, among diurnal planktivores, many fishes here were adapted to feeding on algal fragments. Some species, common in both strong and weak current areas, showed a shift in diet between areas reflecting the type of food items in the water column. Nocturnally, fish planktivores were more abundant in weak current areas feeding on larger zooplankton which is absent from the water column during the day. Much of this zooplankton shelters on or in the substrate during the day, rising into the water column at night. During the day many nocturnal planktivores shelter on reefs. Horch (1973) found both M\jripristis violaceus and M. pralinius common in shallow water during the day, coexisting in coral caves of patch reefs and reefs fringing some islands. At night they left their shelters and often fed in mid-water within a meter of the water's surface. Hobson and Chess (1978) found a clear-cut differentia- tion in the distance that various planktivorous reef fishes move away from reef shelter to feed in the water column. On windward lagoon margin patch reefs, they found that SUBTIDAL ENVIRONMENTS AND ECOLOGY 131 species stationed farther from the reef had more cylindrical bodies with deeply incised caudal fins than species remain- ing relatively close. The fish communities on lagoon margin patch reefs were examined in detail by Nolan (1975) for reefs between Enewetak and Medren. These reefs are typical of those found throughout the lagoon margin on the windward side. He divided the fish community into four assemblages: 1 . The patch reef assemblage (about 25 species) 2. The roving fish assemblage (about 25 species) 3. The sand assemblage 4. The rubble assemblage (3 and 4 together about 50 species) The most numerous fishes living on the patch reefs were cardinalfishes (Apogonidae) and damselfishes (Pomacentri- dae) A large percentage of these individuals are mid-water plankton feeders, relying on items brought by the steady cross reef flat currents from ocean to lagoon. Nolan (1975) found that fish species composition of lagoon margin patch reefs on the windward side visually censused at about 100-day intervals fluctuated considerably over 2'/2 years of observations. Among five "control reefs," each had 20 to 24 species at the end (mean 21). Individual reefs varied by as many as 10 species during the study. The numbers of individuals, however, varied by as much as a factor of 10 during the study. One reef went from about 100 individu- als to 970 because of juvenile recruitment of two species of apogonids and pomacentrids. If increases related to juvenile recruitment were not considered (or in the case of those reefs where massive juvenile recruitment did not occur), numbers of individuals were much more consistent, varying by less than a factor of 2. Nolan (1975) found considerable movement among reef fishes between lagoon margin patch reefs on the windward side of Enewetak. He reported various surgeonfishes, wrasses, and parrot fishes as ranging freely between reefs. Although identifiable assemblages oi' fishes occur in a particular environment, there is small-scale variation in spe- cies composition. Nolan (1975) constructed artificial reefs, made of cement pipe "modules," on the lagoon margin between Enewetak and Medren to provide identical shelter to reef fishes which inhabited those reefs. Artificial reefs reached species equilibrium in 100 to 200 days, a figure equivalent to defaunated natural reefs, but the colo- nization pattern differed from natural reefs. About 10 spe- cies occurred on the artificial reefs (versus about 20 for small natural reefs), which had limited habitat diversity, and variation over time was much higher for artificial reefs than natural reefs. Gladfelter et al. (1980) examined the fish communities of lagoon margin patch reefs between Enewetak and Medren and near the deep channel but utilized reefs over an order of magnitude larger than those studied by Nolan (1975), averaging 150 to 200 m^ in area. Compared to western Atlantic patch reefs of similar size, Enewetak reefs are steef)€r sided with greater vertical relief and more com- plex surface topography. The number of species (visually censused) on the Enewetak patch reef varied between 76 and 109, with about 500 to 900 individuals per reef. Con- sidering trophic categories, diurnal planktivores were more abundant on Enewetak than Virgin Island patch reefs, probably because of the consistent ocean to lagoon cross- reef currents. Herbivores were more diverse among Enewetak reefs with fewer individuals. The Virgin Islands reefs were surrounded by sea grass beds, a habitat lacking at Enewetak, and had more nocturnally foraging invertebrate-feeding fishes. The interrelationships between reef fishes on Enewetak patch reefs are complex. Competition for space and food can be intense between species and among conspecif- ics. Factors controlling initial recruitment of juveniles and their eventual growth to adults are additional controllers of ultimate community composition. Nolan (1975) describes numerous instances of unique interactions among fish species inhabiting small patch reefs on the lagoon margin. Many of these interactions were the result of experimental manipulation, but others were simply the result of long- term careful observation of the environment. Nolan's (1975) record is, perhaps, the best such record of relation- ships and occurrences among a diverse group of fishes on small reefs over time. Allen (1972a, b) described instances where removal of adult anemonefish from their host anemones was followed within a few weeks or months by recruitment of large numbers of juvenile Amphiphon. Anemones unoccupied by Amphiprion were not encountered by Allen (1972a, b), and he felt that anemone availability was one of the major factors limiting anemonefish populations at Enewetak. The situation has not changed since Allen's work; anemones remain relatively uncommon and Amphiphon populations are limited compared to other tropical Pacific areas. The fish communities of the outer reefs, deep lagoon, and open ocean around Enewetak are not as well docu- mented. Even the nearshore spur and groove is poorly known because of its normally hazardous surf conditions. The movement of herbivores and predators onto the reef flat with rising tides is well documented (Hiatt and Stras- burg, 1960; and others). Population sizes, movement along the reef face, and foraging dynamics are not well known. FISH REPRODUCTION AND RECRUITMENT Most reef fishes reproduce by either laying demersal eggs on the substrate or releasing planktonic eggs in the water column. Both hatch as planktonic larvae. Larval life ranges from a minimum of 2 to 3 weeks to as much as 2 to 3 months. Lack of proper substrate for metamorphosis may greatly extend this time. Some information exists con- cerning the spawning habits of demersal- and planktonic- egged species at Enewetak. Major families producing demersal eggs include p>oma- centrids, gobiids, and blennies. Swerdloff (1970) and Allen (1972a, 1975) have described various aspects of pomacen- trid spawning at Enewetak. 132 COLIN Most of the larger fishes at Enewetak produce plank- tonic eggs. Relatively little has been published about the spawning of larger fishes at Enewetak. What has been writ- ten is limited to the papers by Helfrich and Allen (1975), Thresher (1982), and Bell and Colin (1986). There are considerable unpublished data of Colin and Bell. Spawning habits of about 60 species are known and, although gen- eral patterns are known for these, there are exceptions to every generalization. Many planktonic-egged species can spawn at any time during the day in certain locations when tidal conditions are correct. This is generally true for the labrids and parrot fishes, but other families, such as the Pomacanthidae, are believed to spawn only near sunset (Thresher, 1982; Bell and Colin, 1986). In spite of the abundance of large pis- civores, predation on spawning fishes appears to be rare. Predation on eggs immediately after release by particulate plankton-feeding fishes is also uncommon, occurring in only a few percent of spawning releases. Planktonic eggs and larvae from both demersal and planktonic eggs are carried by currents during their development. Larvae produced on the windward side of the lagoon, particularly the northern part, would have an excellent chance of undergoing their entire development within the lagoon, since water residence times in that area are above the mean of about 30 days, reaching as much as 4 months. The mid-depth water return mechanisms of the lagoon would ensure return of larvae to the windward side in spite of the westward surface drift. There is no distinct seasonality known in spawning of fishes at Enewetak, but relatively small differences cannot be ruled out. Gerber (1981) found approximately a two- fold increase in the mean number of fish eggs in mid- lagoon plankton tows during summer as opposed to winter periods, but considerable variability in individual collections indicated different means were not significant. Given the transitory nature of fish spawning, the observed patchiness of eggs is not surprising. A similar situation existed for fish larvae (Gerber, 1981). Higher concentrations of fish eggs and larvae at significant levels were found at Gerber's (1981) "behind reef" station than in the mid-lagoon during winter and may be the result of distance from sources (reefs and their immediate vicinity) of eggs and larvae. Other larvae are undoubtedly carried out to sea, but their potential fate is not well known. The presence of down- current eddies (in this case to the west) behind islands (and atolls) is well documented and may serve to return larvae to the vicinity of Enewetak after a p)eric)d of days or weeks. More work is needed on this phenomenon. Many larvae are certainly lost into the general westward drift of the North Equatorial Current, but sufficient numbers of lar- vae develop within the lagoon or are returned by eddies to maintain fish populations at the atoll. A limited number of recruits must originate east of Enewetak, from Bikini and other atolls of the northern Marshalls, but in terms of numbers are probably overwhelmed by locally produced offspring. Nearly all Enewetak fishes recruit as free-swimming lar- v.Tse. Exceptions would include elasmobranchs (sharks, rays) bearing live young and a limited number of reef fishes which have live young (Brotulidae, Ophidiidae) or a greatly modified larval life (Syngnathidae). Most of the reef fishes have planktonic larvae which must make a transition when becoming juveniles, often moving into a reef environment crowded with others of their species and other species. There was no significant evidence for seasonality of reef fish recruitment to artificial reefs in Nolan's (1975) study. Some species, however, did not recruit at all sea- sons. Possibly, this was because of the relative scarcity of those species, but one common species Apogon 'nouae- guinae" (the species identified as nouaeguinea by Lachner, Schultz, and collaborators, 1953, appears to be A. c^iano- soma though seemingly subspccifically different) did not recruit during the summer. Since it did appear in small numbers on natural reefs, Nolan (1975) attributed this, potentially, to reduced recruitment during the summer. Year-round spawning activity and reproductive colora- tion were observed in some apogonids and pomacentrids. Female chaetodontids with ripe ovaries were noted at all seasons by E. S. Reese (personal communication). The role of predators in limiting the numbers of some small reef fishes on patch reefs has been amply demon- strated by Nolan (1975). He found that when additioijal pomacentrids {Chromis and Dascy//us) were added to reefs already at saturation levels with conspecifics, the new arrivals were readily eaten by cruising piscivores. One artificial reef already at equilibrium had additional damsel- fishes added. Within a day or two almost all additions had perished. Shelter is a factor which limits absolute numbers of such reef fishes; the excess individuals which cannot find a refuge are easily taken by the abundant predators of Enewetak reefs. CIGUATERA Ciguatera is the most common tropical fish jjoisoning known in the Marshall Islands, including Enewetak. Randall (1980) has reviewed the historical reports of ciguatera in the Marshall Islands. Of relevance was information pro- vided by Iroij Johannes Peter that before 1946, some reef fishes from certain areas of Enewetak were pxDisonous to eat. Randall (1980) described instances of ciguatera poison- ing at Enewetak. The internal organs (which are consider- ably more toxic than the flesh) of 47 species of large reef fishes were tested using a mongoose bioassay for toxicity. At least one individual of five species produced the strong- est reaction (death within 48 hours), whereas 31 species produced at least some response by mongooses to inges- tion. Even in the species producing the most frequent reac- tion, the percentage of individuals producing a response is relatively small. Ciguatoxic fishes at Enewetak were found to fit the recognized pattern of being generally large indi- viduals, mostly roving predators, and largely piscivorous (Randall, 1980). No evidence exists that the occurrence of SUBTIDAL ENVIRONMENTS AND ECOLOGY 133 ciguatera at Enewetak is related to radiation in the environment. Disturbance of the marine environment (dredging, construction, wrecks, etc.) has been strongly implicated in producing ciguatera (Randall, 1980). The probable causative organism of ciguatera, a dinoflagellate Gambierdiscus tox'cus, has been identified and the toxin collected and purified. Sharks Several species of sharks are common at Enewetak. They range from nearly harmless to extremely dangerous. Some are found in only one environment, whereas others are nearly ubiquitous The blacktip reef shark, Carcharhinus melanopterus, is abundant on the reef flats all around Enewetak. Hiatt and Strasburg (1960) reported C. melanopterus was the most common shark on windward and leeward reefs. Hobson (1963) reported blacktip sharks were most often observed on sand and coral rubble flats in shallow water. It often penetrates into water so shallow the dorsal fin and back are well exposed. Small C melanopterus individuals are most common on the reef flat. Larger individuals cruise the spur and groove zone offshore and are often seen around lagoon margin pinnacles. The whitetip reef shark, Triaenodon obesus, is f)€rhaps the next most commonly seen species. It is most abundant in the lagoon along the marginal sandy areas and reefs but is also found on seaward reefs. Hobson (1963) found T. obesus most often on patch reefs and coral ledges around the margin of the lagoon. Randall (1977) reported that T. obesus feeds largely on reef fishes, especially scarids and acanthurids, plus octopuses. Also found in the lagoon is the lemon shark, Negaprion brevirosths, which although large, penetrates into relatively shallow water. The author once nearly stepped down onto the back of a 1.5 m lemon shark while wading ashore on Ikuren in knee-deep water. The most studied and the most dangerous shark at Enewetak is the gray reef shark, Carcharhinus ambli;rh\;n- chos. It is found throughout the lagoon and on the sea- ward reefs. Hobson (1963) felt it was most abundant in the deeper waters of the lagoon and passes. Attacks on humans are discussed subsequently. Johnson and Nelson (1973) described in detail the threat display of the gray reef shark, which often precedes an attack. Sharks placed in a situation of a diver potentially restricting its escape produced the most intense displays: an exaggerated, often rolling, swimming motion with back arched, pectoral fins dropped, and snout lifted. Starck (MPRL, 1971 to 1972) elicited attack responses on a small wet submersible by pursuing C ambl\;rh\^nchos. A more detailed account is presented in Anderson (1980). Subsequently, this attack and its preceding threat display have been investigated by Nelson (MPRL, 1978, 1979). He found that the shark usu- ally attacks after displaying if the object or person contin- ues to approach. The attacks are sudden, high-speed strikes, often with the mouth open. He believed that "oriented pursuit" by the small submarine was of primary importance in releasing an attack. A straight-line pass near the shark never released an attack, although it did produce the threat display. For more information on gray reef shark behavior, see Nelson et al. (1986). Randall (1980) reported that C. ambliirht;nchos from Enewetak and other localities fed mostly on reef fishes and, to a lesser extent, on cephalofxxls. It is the most common shark seen on seaward reefs. Off the southwest islands and on the leeward reef face, it usually app>ears before the silvertip shark, C. albimarginatus, and out- numbers the latter shark two or three to one. In the lagoon it is common around mid-lagoon pinnacles where it seems particularly aggressive. Often when a boat stops in mid-lagoon on a calm day, one or more C. ambl^rhi^nchos will rise to investigate the boat from water 50 to 60 m deep. The movements of C ambl\^rhi>nchos tagged with ultra- sonic transmitters have been investigated by Nelson (MPRL, 1978, 1979). He has determined that gray reef sharks often move surprisingly long distances around Enewetak. Deep-water gray reef sharks tagged on or near the drop-off of the seaward reefs ranged as much as 16 km along the reef in one night. They were not as predictably home ranging as lagoon gray reef sharks, and Nelson (MPRL, 1979) suggested they might represent a more nomadic segment of the population. Lagoon grays were tracked for as long as 21 days, and although some stayed in one area, others moved considerably. One individual tagged at the mid-lagoon "dome" pinnacle sfsent the daylight hours near that pinnacle but ranged widely at night. Its home range was estimated at about 53 km . McKibben and Nelson (1986) discussed movements of tagged gray reef sharks at Enewetak. Other seaward reef sharks are the silvertip shark, Carcharhinus albimarginatus, and Galapagos shark, C galapagensis. The silvertip shark is found normally on seaward slopes below 20 to 30 m, although Randall (1980) observed one individual in the lagoon in water 2 m deep. There are reports of C. albimarginatus as deep as 400 m (Randall, 1980). Silvertip sharks feed almost exclusively on fishes, both reef and open water. Randall (1980) also found at Enewetak a gray reef shark over 60 cm in total length in the stomach of a C. albimarginatus that was 1.6 m in total length. The Galapagos shark is a large, dangerous species, but fortunately it is uncommon at Enewetak. Randall (1980) collected only a single specimen, but little is known of its habits beyond feeding on fishes (including sharks) and cephalopods. Probably the largest dangerous shark in Enewetak waters is the tiger shark, Galecerdo cuvier. Randall (1980) examined two specimens from Enewetak, 1.7 and 2.4 m precaudal length (length minus the caudal fin), of 72 and 174 kg, respectively. McNair (1975), an accurate and experienced shark observer, while diving on the leeward seaward reef, observed a huge tiger shark pass above him 134 COLIN which he estimated was longer than the 21 -ft boat it passed by. Tiger sharks are seldom seen by divers and, therefore, are not as much of a hazard as some smaller, dangerous species. Randall (1980) found the scutes of a green turtle shell, shark vertebrae, bird feathers, digested shark fins, and pieces of a porpoise in the stomachs of Enewetak G. cuvier. There have been several instances of shark attacks at Enewetak. Most have involved the gray reef shark and, in some, injury occurred to the human involved. Hobson et al. (1961) documented two incidents with gray reef sharks in which spearfishing probably stimulated aggressive behavior. Fortunately neither instance resulted in injuries. Not so lucky was another individual whose head was slashed by the upper jaw of C. ambl\;rh[jnchos after the powerhead he was using to try to kill this shark failed to detonate on impact (Randall, 1980). In April 1978, another attack by C. ambli/rhi;nchos occurred in which a 1.5-m (5-ft) long individual severely mauled the right arm of a diver and attacked his diving partner (M. V. deGruy and P. Light, unpublished report). In this case deGruy approached the shark, which was ex- hibiting the threat posture, in an attempt to photograph it. When deGruy triggered the electronic strobe of his cam- era, the shark turned, rushed toward deGruy, and seized his arm. Seconds later the shark tore a chunk from one diving fin. As the diver's companion, Light, swam to his assistance, the shark bolted toward Light and badly slashed his hand. The shark disappeared. Both divers sub- sequently recovered from their wounds. The most recent attack by C. ambli^rhimchos on humans occurred in January 1982 when one of the repa- triated Marshallesc. while spearfishing, had his left arm mauled by what was probably a gray reef shark. Several sharks were around this fisherman and his two compan- ions, who were carrying a considerable quantity of dead fishes. Randall and Helfman (1973) reported two instances of C. melanopterus menacing humans at Enewetak. It is interesting to note that despite repeated success in producing threat displays and attacks by C. ambli;rhi;nchos by pursuit with small wet submersibles, similar attempts have failed to produce the threat or attack by C albimar- ginatus, C. melanopterus, and T. obesus (Nelson, MPRL, 1979; Starck, MPRL, 1971 to 1972). Crater Life Nolan et al. (1975) described bottom substrata and fishes inhabiting the two small craters (Cactus and Lacrosse) at the north end of Runit Island. Hard substrata were restricted to the upper 4 m and the sides of both craters sloped quickly to a sediment plain at about 12 m deep. The bottoms of the craters were "heavily excavated by several species of gobies and burrowing shrimps." Other bioturbating organisms were also present. Colonies of Hahmeda and Derbesia minima were abundant on the sediment bottom. Hillis-Colinvaux (1980) found no Halimedae in Lacrosse crater but found a pure, dense strand of Halimeda incrassata in the murky center of Cactus crater at 11 m depth. This merged peripherally with Caulerpa ad serrulata. but no loose plants were seen on the sides of the crater. Halimeda incrassata was rare at Enewetak (Hillis-Colinvaux, 1980), and the Cactus crater population was the only dense (about 200 or more thalli m~^) strand found at Enewetak. Hillis-Colinvaux (1980) suggested that the extremely soft and fine sediment of Cactus crater might have promoted the growth of this dense strand, possibly from a limited number of vegetative propagules. Nolan et al. (1975) found little living coral in the Runit craters but reported that "molluscs, crustaceans, poly- chaetes, zooplankton, algae, and phytoplankton found in the craters seemed typical of the fauna and flora occurring in the adjacent lagoon or upon the reef flat." Eighty-four species of fish were observed or collected in the craters; the number is incomplete for various cryptic families. Cen- suses at near high and low tides indicate fewer individuals of species at low tides than at high tides. The third small atomic test crater, Seminole, on Bokcn (North) Island has not been examined biologically. Adjacent sand flats seem to be an area with high numbers of small blacktip sharks. Circulation into Seminole crater is much more restricted than circulation into either Runit atomic crater. The three large thermonuclear craters in the north lagoon have not previously been described biologically. During the Enewetak Submersible Project, several dives were made in Oak crater using the submarine, and addi- tional scuba dives were made on the crater slope. The level bottom of Oak crater (Ristvet et al., 1978; Chapter 4, this volume) was heavily bioturbated, almost certainly by callianassids, with a mound density equaling that observed anywhere else at Enewetak. In addition, the irregular urchin Maretia planulata occurred in high densities of 10 m on the surface of sediment at the bottom of the crater. Similar bioturbation was evident in Koa crater, although at a lesser depth. Nelson and Noshkin (1973) did not consider that biologically mediated presentation of radionuclides from within the sediment column to the water column was occurring in Koa (and Mike) craters but that the "principal loss of activity from the deposits may only be from the slow release to the overlying waters." Smith and Brock (MPRL, 1976) found that the Mike- Koa and Oak craters have large amounts of rubble in the vicinity of the craters which provide an unfavorable sub- stratum for coral growth. The locations of the craters are described in Chapter 3 of this volume. Why Are There No Sea Grasses at Enewetak? Tsuda et al. (1977) have summarized the known distri- bution of sea grasses in Micronesia. Only one species, Thalassia hemprichii, is known from the Marshall Islands. Records exist for it from Ailinglapalap, Jaluit, and Ujilang SUBTIDAL ENVIRONMENTS AND ECOLOGY 135 Atolls. At the last atoll, Fosburg (1955) reported "a rather extensive strip" of T hemprichii along the lagoon shore of Ujilang Island. Thalassia hemprichii was found by the author only along the lagoon shore of Ujilang Island in July 1982. None was seen on several lagoon shores on the windward side visited or on two coral pinnacles with extensive sandy areas above 15 to 20 m depth. The strip along Ujilang Island occurred only at depths <1 m. Reliable information exists that 7 hemprichii (or any other sea grass) does not occur at Bikini, Rongelap, or Rongerik (Emery ct al., 1954). Fosburg (1955) reported on visits to many north Marshall atolls (Kwajalein, Lae, Ujae, Wotho, Likiep, Aihik, Bikar, Pokak. and Ujilang) with Ujilang the only of these where sea grasses were noted. At Kwajalein, signifi- cant diving and collecting activities by knowledgeable marine biologists over large portions of the atoll have failed to reveal 7. hemprichii. Tsuda (Chapter 1, Volume II) documents that no record of sea grasses from Enewetak exists. Because of the large amount of marine work carried out at this atoll, it is reasonable to say they do not occur here. The presence of T. hemprichii at Ujilang, only 200 km away, is intriguing. The decline, however, in the numbers of sea grass species eastward through Micronesia (Tsuda et al., 1977) and the probable absence of T. hemprichii at most — if not all — other Marshall atolls, indicate that per- haps there has been no opportunity for transport of T. hemprichii to Enewetak. The areas upcurrent of the atoll are similar atolls without sea grasses. A similar condi- tion has been noted for Sargassum (Tsuda, 1976) with no records from any northern Marshall atoll, including Enewetak. The means of dispersal of 7. hemprichii are lim- ited. Potentially it could be transported as drift material torn loose during storms or as drifting seeds or seed cap- sules. Both potential mechanisms are current dependent, which would work against transport to Enewetak. Atolls farther south, in the influence of the Equatorial Counter- current, may have received their populations via this cur- rent. Zoogeographic Considerations One interesting phenomenon is that many marine animals that arc relatively scarce at Enewetak are much more common elsewhere. This seems true even within the Marshall Islands where disparity exists between Enewetak and the more southerly Marshall Island atolls— such as Kwajalein, Majuro, and Arno — which have had a signifi- cant collecting effort. This scarcity is true among fishes. For example, Allen (1972a) commented that most anemone species were rela- tively scarce at Enewetak as compared to his personal observations in Tahiti and literature from the Nicobor Islands. "Several hours of intensive searching may at best result in finding four or five widely scattered sp>ecimens (of anemones) of the variety which harbor Amphiprion." As has been pointed out elsewhere, even when present, the occurrence of anemones may be transitory. Hiatt and Strasburg (1960) indicate that a number of fish species are uncommon or rare at Enewetak (and often Bikini) compared to Arno Atoll. They felt that Arno was a more productive atoll than either Enewetak or Bikini because it is located in an area of upwelling where the North Equatorial Current and Equatorial Countercurrent meet and because it has a higher rainfall than the other two atolls. Whether this has an effect on the abundance of reef fishes or whether the differences observed are pro- duced by sources of larval recruits, etc., is not known. Species of Plesiops and Pseudogramma are among fishes that are less common at Enewetak. Randall (1986) lists a number of species from Kwajalein Atoll which, in spite of comparable collecting effort, are not known from Enewetak. Within the overall picture of Indo-West Pacific shore fish distribution, the Enewetak fauna is less diverse than the "core" areas of the Indo-Malayan Archipelago. This is well known for individual families (Allen, 1975), but the Enewetak fish fauna is at a level of diversity "expected" when compared to adjacent areas. The differences exist with respect to abundance of quite a number of species. Why Are There No Mangroves at Enewetak? ACKNOWLEDGMENTS Wiens (1962) has summarized much of the information on mangroves on atolls. He cites records of "mangroves" on several southern Marshall Island atolls. Hatheway (1953) described stands of Sonrieratia caseolahs and Bruguiera conjugata on Arno Atoll. Wiens (1956) observed a tidal inlet on Ailinglapalap Atoll with three species of mangroves. Fosburg (1955) reports B. conjugata on Utirik, Ailuk, and Lae to be rare and limited to "wet depressions." Otherwise, he does not record any "mangrove" species, particularly those of the Rhizophori- dae from the northern Marshall Islands. Again the situation is similar to sea grasses. Mangroves can certainly survive at atolls like Enewetak, but it is likely their transport mechanisms have never allowed their intro- duction. The staff of the Mid-Pacific Research Laboratory (MPRL) and its predecessor institutions made possible the vast majority of the marine research undertaken at Enewetak Atoll since 1954. Many of the people involved in this work have been cited in the preface to this volume. I would like to thank in particular the following MPRL staff members for their help in my own fieldwork and that of others: L. J. Bell, L. M. Boucher, V. S. Frey, S. Johnson, J. T. Harrison, III, and R. M. Richmond. The operation of MPRL would not have been possible without their dedica- tion and perseverance under extremely difficult cir- cumstances. I would like to thank the following for their comments on the manuscript: L. J. Bell, J. T. Harrison, III, A. Kohn, and J. E. Randall. 136 COLIN REFERENCES Allen, G. R., 1972a, The Anemonefishes: Their Classification and Biology, TFH Publications, Neptune City, New Jersey. , 1972b, Observations on a Commensal Relationship Between Siphamia fuscolineata (Apogonidae) and the Crown-of-Thorns Starfish, Acanthaster planci. Copeia. 1972: 595-597. , 1975, Damselfishes of the South Seas. TFH Publications, Neptune City, New Jersey. Atkinson, M., S V. Smith, and E. D. Stroup, 1981, Circulation in Enewetak Atoll Lagoon, Limnol. Oceanogr, 26: 10741083. Bakus, G. J., 1963, The Effects of Fish-Grazing on Invertebrate Evolution in Shallow Tropical Waters, Allan Hancock Founda tion Occas. Pap . 27: 1-29. — , 1967, The Feeding Habits of Fishes and Primary Production at Eniwetok. Marshall Islands, Micronesica. 3: 135-149 — , 1973, The Biology and Ecology of Tropical Holothurians, Vol. 2, Biol., Bio/ogy and Geolog\^ of Coral Reefs. O. A. Jones and R. Endean (Eds.), Academic Press, New York, 1: 325367. Basile, L. L., R. J. Cuffey, and D F Kosich, 1984, Sclerosponges, Pharetronids, and Sphinctozoans (Relict Cryp- tic Hard-Bodied Porifera) in the Modern Reefs of Enewetak Atoll, J Paleontoi. 58: 636-650 Bell, L. J., and P. L. Colin, 1986, Mass Spawning of Caesio teres (Pisces: Caesionidae) at Enewetak Atoll, Marshall Islands, Environ. Biol Fishes. 15: 69-74. Buddemeier, R. W., J. E. Maragos, and D W. Knutson, 1974, Radiographic Studies of Reef Coral Exoskeletons: Rates and Patterns of Coral Growth, J Exp Mar. Biol. Ecol . 14: 179-200. Chartoch, M., 1972, The Role of Detritus in a Tropical Marine Ecosystem: Niche Separation in Congeneric Ophiuroids, Food Partitioning in Cr\;ptic Invertebrates, and Herbivore Detritus Production at Eniwetok. Marshall Islands. Ph D. dissertation, University of Southern California. Coles, S. L., and R. Strathmann, 1973, Observations on Coral Mucus "Floes" and Their Potential Trophic Significance, Limnol Oceanogr . 18: 673-678. — , P. L. Jokiel, and C. R. Lewis, 1976, Thermal Tolerance in Tropical Versus Subtropical Pacific Reef Corals, Pac. Sci., 30: 159-166. Colin, P. L., 1986, Benthic Community Distribution in the Enewetak Atoll Lagoon, Marshall Islands, Bull Mar. Sci.. 38: 129143. , T. H. Suchanek, and G. McMurtry, 1986, Water Pumping and Particulated Resuspension by Callianassids (Crustacea: Thalasslnidea) at Enewetak and Bikini Atolls, Marshall Islands, Bull. Mar. Sci.. 38: 19-24. Cuffey, R. J., 1973, The Roles of Bryozoans in Modem Coral Reefs, Geol. Rundschau. 61: 542 550. — , 1978, A Note on Bryozoans in the Modern Reefs of Eniwetok Atoll and the Australian Great Barrier Reef, Pac Geol., 13: 65-76. Deutsch, S., and J. H. Lipps, 1976, Test Structure of the Fora- minifer Carterina. J. Paleontoi. 50: 312-317 DiSalvo, L. H., 1969, Isolation of Bacteria from the Corallum of Porites lobata (Vaughan) and Its Possible Significance. Am. Zooi, 9: 735-740. Emery, K. O., J. I. Tracey, Jr., and H. S. Ladd, 1954, Geology of Bikini and Nearby Atolls, U. S. Geol Surv Prof Pap 260-A: 1-265. Fosburg, F R., 1955, Northern Marshall Islands Expedition, 1951-1952, Land Biota: Vascular Plants, Atoll Res. Bull.. 39: 1-22. Fricke, H. W., 1971, Fische als Feinde Tropischer Seeigel, Mar. Biol. 9: 328-338 Gerber, R P., 1981, Species Composition and Abundance of Lagoon Zooplankton at Enewetak Atoll, Marshall Islands, Atoll Res Bull.. 247: 1-22 , and M B. Gerber, 1979, Ingestion of Particulate Organic Material and Subsequent Assimilation, Respiration, and Growth by Tropical Lagoon Zooplankton, Mar. Biol.. 52: 33-43. , and N. Marshall, 1974, Reef Pseudoplankton in Lagoon Trophic Systems, in Proceedings of the Second International Coral Reef Symposium. Great Barrier Reef, Australia, 1: 105-107. , 1982, Characterization of the Suspended Particulate Organic Matter and Feeding by the Lagoon Zooplankton at Enewetak Atoll, Bull. Mar. Sci . 32: 290-300. Gilmartin, M., 1960, The Ecological Distribution of Deep Water Algae of Eniwetok Atoll, Ecology. 41: 210-221. , 1966, Ecology and Morphology of Tydemania expeditionis. a Tropical Deep-Water Siphonous Green Alga, J PhycoL. 2: 100-105. Gladfelter, W. B., J. C. Ogden, and E. H. Gladfelter, 1980, Simi- larity and Diversity Among Coral Reef Fish Communities: A Comparison Between Tropical Western Atlantic (Virgin Islands) and Tropical Central Pacific (Marshall Islands) Patch Reefs, Ecology. 61: 1156-1168. Haggerty, J. A.. J. N. Weber, R. J. Cuffey, and P. Deines, 1980, Environment-Related Morphologic and Geochemical Variability in the Modern Reef Corals Favia pallida and Favia stelligera on Enewetak Atoll, Pacific Geology. 14: 95-113 Hammond, L. S , 1981, An Analysis of Grain-Size Modification in Biogenic Carbonate Sediments by Deposit-Feeding Holothuri- ans and Echinoids, Limnol. Oceanogr . 26: 898-906. Hantzschel, W., 1962, Trace Fossils and Problematica. R. C. Moore (Ed), Geol. Soc. Amer. and Univ. Kansas Press, Vol. W, pp. 177-245. Harrigan, J. F., 1972, The Planula Larva of PcKillopora dam- icomis. Lunar Periodicity of Swarming and Substratum Selec- tion Behavior. Ph.D. dissertation. University of Hawaii. Hatheway, W. H., 1953, The Land Vegetation of Arno Atoll, Marshall Islands, Atoll. Res Bull.. 16: 1-68. Helfrich, P., and P. M. Allen, 1975, Observations on the Spawn- ing of Mullet, Crenimugil crenilabus (Forskal), at Enewetak, Marshall Islands, Micronesica. 11: 219-225. Hiatt, R. W., and D. W Strasburg, 1960, Ecological Relation- ships of the Fish Fauna on Coral Reefs of the Marshall Islands, £co/ Monogr.. 30: 65-127. Highsmith, R. C, 1979, Coral Growth Rates and Environmental Control of Density Banding, J Exp. Mar. Biol. Ecol. 37; 105-125. , 1980, Burrowing by the Bivalve Mollusc Lithophaga curta in the Living Coral Montipora beryi and a Hypothesis of Recipro- cal Larval Recruitment, Mar. Biol. 56; 155-162. , 1981a, Agents and Dynamics of Coral Bioerosion at Enewetak, Int. Rev. Cesamten Hydrobiol. 66: 335-375. , 1981b, Lime-Boring Algae in Hermatypic Coral Skeletons, J. Exp Mar. Biol Ecol, 55: 267-281. Hillis-Colinvaux, L., 1977, Halimeda and Tydemania: Distribution, Diversity, and Productivity at Enewetak, in Proceedings of the Third International Coral Reef Symposium, Miami, Florida, I: 365-370. SUBTIDAL ENVIRONMENTS AND ECOLOGY 137 , 1980, Ecology and Taxonomy of Halimeda: Primary Pro ducer of Coral Reefs, Adv Mar. Bioi. 17; 1-327 1986, Deep Water Populations of Halimeda in tfie Economy of an Atoll, Bull Mar Sci . 38: 155-169 Hobson, E. S., 1963, Feeding Behavior in Three Species of Sharks, Pac Sci , 17: 171- 194 , and J R. Chess, 1978, Trophic Relationships Among Fishes and Plankton in the Lagoon at Enewetak Atoll, Marshall Islands, Fish Bull, 76: 133-153 , F. Mautin, and E S. Reese, 1961, Two Shark Incidents at Eniwetok Atoll, Marshall Islands, Pac. Sci , 15: 605-609. Horch, K, W,, 1973, Adaptations to the Acoustic Environment by the Squirrelfishes Myripristis violaceus and M pralinus. Mar Behav. Physio/,, 2: 121-139. Johannes, R E., and R Gerber, 1974, Import and Expxsrt of Net Plankton by an Eniwetok Coral Reef Community, in Proceedings of the Second International Coral Reef Siimpo sium. Great Barrier Reef, Australia, 1: 97-104 Johnson, D W., and D R Nelson, 1973, Agonistic Display in the Gray Reef Shark, Carcharhinus menisorrah. and Its Rela- tionship to Attacks on Man, Copeia, 1973: 76-84. Kaufman, L. S., 1977, The Threespot Damselfish: Effects on Benthic Biota of Caribbean Coral Reefs, in Proceedings of the Second International Coral Reef Symposium, Great Barrier Reef, Australia, 1: 559-564 Knutson, D. W., R. W Buddemeier, and S. V. Smith, 1972, Coral Chronometers: Seasonal Growth Bands in Reef Corals, Science. 177: 270-272. Ladd, H S., 1961, Reef Building, Science, 134: 703-715 Lamberson, J. O., 1978, Notes on the Morphology, Ecology, and Distribution of Thelenota anax H. L. Clark (Holothuroidea: Stichopodidae), Micronesica, 14: 115122 Lee, R. K. S., 1967, Taxonomy and Distribution of the Melobesioid Algae on Rongalap Atoll, Marshall Islands, Can J Bot, 45: 985-1001. Marsh, J. A., Jr., 1970, Primary Productivity of Reef-Building Calcareous Red Algae, Ecology, 51: 255-263. McKibben, J N., and D R Nelson, 1986, Patterns of Movement and Grouping of Gray Reef Sharks, Carcharhinus amblyrhyn chos, at Enewetak, Marshall Islands, Bull Mar. Sci , 38: 89110. McMurtry, G. M , R. C. Schneider, P L. Colin, R. W. Bud- demeier, and T. H. Suchanek, 1985, Redistribution of Fallout Radionuclides in Enewetak Atoll Lagoon Sediments by Calli- anassid Bioturbation, Nature, 313: 67 '1-677. Milliman, J. D , 1974, Marine Carbonates. Recent Sedimentary Carbonates. Part 1. Springer-Verlag, Berlin, Heidelberg, and New York Motta, P. J., 1980, Functional Anatomy of the Jaw Apparatus and the Related Feeding Behavior of the Butterflyfishes (Chae todontidae) Including a Review of Jaw Protrusion in Fishes, Ph.D. dissertation. University of Hawaii. , 1985, Functional Morphology of the Head of Hawaiian and Mid-Pacific Butterflyfishes (Perciformes: Chaetodontidae), Environ. Biol. Fishes. 13(4): 253-276. Nelson, D. R., R. R. Johnson, J N. McKibben, and G. G. Pit- tenger, 1986, Agonistic Attacks on Divers and Submersibles by Gray Reef Sharks, Carcharhinus amblyrhpnchos : Antipreda- tory or Competitive?, Bull. Mar. Sci.. 38: 68-88. Nelson, V., and V. Noshkin, 1973, Marine Programs, Enewetak Radiological Survey. U. S. Atomic Energy Commission, Nevada Operations Office, Las Vegas, Nevada, NVO140, 131: 1 224. Nolan, R. S., 1975, The Ecology of Patch Reef Fishes. Ph.D. dis- sertation. University of California, San Diego. , R. R. McConnaughy, and C. R. Steams, 1975, Fishes Inhab- iting Two Small Nuclear Test Craters at Enewetak Atoll, Marshall Islands, Micronesica, 11: 205-217. Noshkin, V. E., 1980, Transuranium Radionuclides in Com- fjonents of the Benthic Environment of Enewetak Atoll, Transuranium Elements in the Environment. W. C. Hanson (Ed.), Technical Information Center, U. S. Dept. Energy. Odum, H. T , and E P. Odum, 1955, Trophic Structure and Pro- ductivity of a Windward Coral Reef Community on Eniwetok, Ecol. Monogr.. 25: 291-320. Ogden, J. C, 1976, Some Aspects of Herbivore-Plant Relation- ships on Caribbean Reefs and Seagrass Beds, Aquatic Botany, 2: 103-116 1977, Carbonate-Sediment Production by Parrot Fish and Sea Urchins on Caribbean Reefs, Stud. Geol. 4: 281-288 , R. Brown, and N Salesky, 1973, Grazing by the Echinoid Diadema antillarum Philippi: Formation of Halos Around West Indian Patch Reefs, Science, 182: 715 717. and P. S Lobel, 1978, The Role of Herbivorous Fishes and Urchins in Coral Reef Communities, Environ. Biol. Fishes. 3: 49-63 Randall, J. E., 1956, A Revision of the Surgeonfish Genus Acanthurus. Pac Sci . 10: 159-235. , 1965, Grazing Effect on Seagrass by Herbivorous Reef Fishes in the West Indies, Ecology. 46: 255-260. , 1974, The Effect of Fishes on Coral Reefs, in Proceedings of the Second International Coral Reef Symposium. Great Barrier Reef, Australia. , 1977, Contribution to the Biology of the Whitetip Reef Shark (Triaenodon obesus), Pac. Sci., 31: 143-164. , 1980, A Survey of Ciguatera at Enewetak and Bikini, Marshall Islands, with Notes on the Systematics and Food Habits of Ciquatoxic Fish, Fish Bull., 78(2): 201-249. , 1986, 108 New Fish Records from the Marshall Islands, Bull Mar. Sci,, 38: 170-252. , and W. Hartman, 1968, Sponge-Feeding Fishes of the West Indies, Mar Biol. 1(3): 216-225. , and G. S Helfman, 1973, Attacks on Humans by the Black- tip Reef Shark (Carcharhinus melanoptsrus). Pac. Sci.. 27: 226-238. Reese, E. S., 1973, Duration of Residence of Coral Reef Fishes on "Home" Reefs, Copeia. 1973.1: 145-149. 1975, A Comparative Field Study of the Social Behavior and F?elated Ecology of Reef Fishes of the Family Chaetodon- tidae, Z Tierpsychol . . 37: 37-61. 1977, Coevolution of Corals and Coral Feeding Fishes of the Family Chaetodontidae, in Proceedings of the Third Interna- tional Coral Reef Symposium. Miami, Florida, 1: 267-274. , 1981, Predation on Corals by Fishes of the Family Chaeto- dontidae: Implications for Conservation and Management of Coral Reef Ecosystems, Bull Mar Sci , 31: 594-604 Ristvet, B. L., 1978, Geologic and Geophysical Investigations of the Enewetak Nuclear Craters, Air Force Weapons Lab, Air Force Systems Command, Kirtland Air Force Base, New Mex- ico, AFWLTR 77 242. Rutzler, K., 1975, The Role of Burrowing Sponges in Bioeroslon, Oecologia, 19: 203-216. Schultz, L. P., et al.. 1953-1956, Fishes of the Marshall and Marianas Islands, Bull U. S Natl. Mus. Smith, S v., and J. T. Harrison, III, 1977, Calcium Carbonate 138 COLIN Production pf the Mare incognituw, the Upper Windward Reef Slope, at Enewetak Atoll, Science, 197: 556-559. , and D W. Kinsey, 1976, Calcium Carbonate Production, Coral Reef Growth, and Sea Level Change, Science. 194: 937-939. , and A. C. Paulson, 1975, Carbonic Anhydrase in Some Coral Reef Fishes: Adaptation to Carbonate Ingestion?, Comp. Bicxhem Physio/ , 50A: 131 134 Stimson, J. S., 1978, Mode and Timing of Reproduction of Some Common Hermatypic Corals of Hawaii and Enewetak, Mar- Biol 48: 173-184. Suchanek, T, H,, and P. L. Colin, 1986, Rates and Effects